CN119840468A

CN119840468A - Charging pile control method based on multi-mode sensing and edge calculation

Info

Publication number: CN119840468A
Application number: CN202510113562.5A
Authority: CN
Inventors: 周明星
Original assignee: Shenzhen Haide New Energy Technology Co ltd
Current assignee: Shenzhen Haide New Energy Technology Co ltd
Priority date: 2025-01-24
Filing date: 2025-01-24
Publication date: 2025-04-18

Abstract

The present application relates to the technical field of battery charging of electric vehicles, and provides a charging pile control method based on multimodal perception and edge computing, the method comprising: obtaining multimodal perception information corresponding to a target charging pile, the multimodal perception information including environmental information, charging pile working information, and information of a vehicle to be charged; performing edge computing on the multimodal perception information through an edge computing unit pre-configured around the target charging pile to obtain an edge computing result; inputting the edge computing result to a cloud server, so that the cloud service unit adjusts the power allocation and charging priority of the target charging pile according to the edge computing result, obtains a target charging plan, and feeds back the target charging plan to the edge computing unit to control the target charging pile to charge the vehicle to be charged. The scheme in the present application improves the charging efficiency of the charging pile and reduces the fluctuation of the power load.

Description

Charging pile control method based on multi-mode sensing and edge calculation

Technical Field

The application relates to the technical field of battery charging of electric vehicles, in particular to a charging pile control method based on multi-mode sensing and edge calculation.

Background

With the increase of the number of electric vehicles, the demand for charging piles is increasing.

Existing charge stake management systems typically rely on a single data source for charge scheduling, such as vehicle battery status or user subscription information. However, the system lacks comprehensive consideration of environmental factors, cannot effectively cope with emergency situations such as fluctuation of power supply in the area where the charging pile is located, traffic condition change near the charging pile and the like, and moreover, the centralized data processing mode causes higher network delay and data processing cost. The existing charging pile management mode can lead to low charging efficiency of the charging pile and large fluctuation of electricity load.

Therefore, how to improve the charging efficiency of the charging pile and reduce the fluctuation of the power load is a problem that needs to be solved at present.

Disclosure of Invention

In view of the above problems in the prior art, an object of the present application is to provide a charging pile control method based on multi-mode sensing and edge calculation, so as to at least solve the technical problems of how to improve the charging efficiency of the charging pile and reduce the fluctuation of the power load.

To achieve the above and other related objects, the present application provides a charging pile control method based on multi-modal sensing and edge calculation, the method comprising:

The method comprises the steps of obtaining multi-mode sensing information corresponding to a target charging pile, wherein the multi-mode sensing information comprises environment information, charging pile working information and vehicle information to be charged;

performing edge calculation on the multi-mode sensing information through an edge calculation unit which is arranged around the target charging pile in advance to obtain an edge calculation result;

And outputting the edge calculation result to a cloud server, so that the cloud service unit adjusts the power distribution and the charging priority of the target charging pile according to the edge calculation result to obtain a target charging plan, and feeds back the target charging plan to the edge calculation unit to control the target charging pile to charge the vehicle to be charged.

In some embodiments, the environmental information includes people traffic density, traffic flow, weather conditions;

the charging pile working information comprises a current load, a working mode and a fault condition;

The vehicle information to be charged comprises a vehicle battery state, a vehicle type to be charged and vehicle user reservation information to be charged.

In some embodiments, the obtaining the multi-mode sensing information corresponding to the target charging pile includes:

acquiring a first video stream through cameras arranged around the target charging pile, and detecting and tracking pedestrians in the first video stream to obtain the traffic density;

acquiring a second video stream through cameras arranged on a road around the target charging pile, and detecting and tracking vehicles in the second video stream to obtain the traffic flow;

The weather conditions are obtained by weather sensors arranged around the target charging pile and calling a third party weather interface;

The method comprises the steps that real-time voltage and real-time current are obtained through a voltage sensor and a current sensor which are arranged on a power supply line of a target charging pile, and the current load is obtained according to the real-time voltage and the real-time current;

The working mode is obtained through a state detection sensor arranged in the target charging pile;

And acquiring working states of the power supply unit, the communication unit and the charging interface through fault detection sensors arranged on the power supply unit, the communication unit and the charging interface of the target charging pile, and determining the fault condition according to the working states of the power supply unit, the communication unit and the charging interface.

In some embodiments, the performing edge calculation on the multi-mode sensing information to obtain an edge calculation result includes:

Performing standardized processing on the environment information, performing independent heat coding on the charging pile working information, and performing standardized processing on the vehicle information to be charged to obtain the preprocessed multi-mode sensing information;

extracting the characteristics corresponding to each mode in the preprocessed multi-mode sensing information through a pre-constructed neural network, and fusing the characteristics corresponding to each mode to obtain fused characteristics;

and analyzing the fusion characteristics to obtain an edge calculation result.

In some embodiments, the extracting the features corresponding to each mode in the preprocessed multi-mode sensing information, and fusing the features corresponding to each mode to obtain fused features includes:

extracting environmental characteristics corresponding to the environmental information in the preprocessed multi-mode sensing information through a convolutional neural network;

extracting working characteristics corresponding to the charging pile working information in the preprocessed multi-mode sensing information through a fully connected neural network;

Extracting vehicle characteristics to be charged corresponding to vehicle information to be charged in the preprocessed multi-mode sensing information through an embedding layer;

Respectively carrying out linear change on the environmental characteristics, the working characteristics and the characteristics of the vehicle to be charged to obtain inquiry and keys;

According to the query and the key, calculating an attention score, and normalizing the attention score to obtain attention weights corresponding to the environmental characteristics, the working characteristics and the vehicle characteristics to be charged;

And weighting the environmental characteristics, the working characteristics and the vehicle characteristics to be charged according to the attention weights corresponding to the environmental characteristics, the working characteristics and the vehicle characteristics to be charged to obtain the fusion characteristics.

In some embodiments, the analyzing the fusion feature to obtain an edge calculation result includes:

taking an object to be analyzed in the fusion characteristic as a node, wherein the object to be analyzed comprises the target charging pile and the vehicle to be charged, and the characteristic vector of the node is the characteristic vector in the fusion characteristic;

Determining edges among a plurality of nodes according to the relation structure of the object to be analyzed;

Initializing feature vectors of all the nodes through the fusion features to obtain initialized feature vectors of all the nodes;

For each node, acquiring information from adjacent nodes, and updating the initialization feature vector according to the information to obtain a target feature vector of each node;

and transmitting the target feature vector of each node to a pre-trained regression model to obtain the edge calculation result.

In some embodiments, the adjusting the power distribution and the charging priority of the target charging pile according to the edge calculation result, to obtain a target charging plan includes:

Determining the state of an environment and executable actions according to the edge calculation result, wherein the state of the environment comprises the power state of the target charging pile and the state of the vehicle to be charged, and the executable actions comprise adjusting the power distribution of the charging pile and setting the charging priority;

Evaluating the effect of the executable action according to a preset rewarding function, wherein the preset rewarding function comprises a charging efficiency rewarding function, a user satisfaction rewarding function and a cost benefit rewarding function;

Determining an expected return corresponding to the executable action taken in the state of the environment, and determining a target action according to the effect of the executable action and the expected return;

and obtaining the target charging plan based on the target action.

In an embodiment of the present application, there is also provided a charging pile control device based on multi-modal sensing and edge calculation, the device including:

the multi-mode sensing information acquisition module is used for acquiring multi-mode sensing information corresponding to the target charging pile, wherein the multi-mode sensing information comprises environment information, charging pile working information and vehicle information to be charged;

the edge calculation module is used for carrying out edge calculation on the multi-mode sensing information through an edge calculation unit which is arranged around the target charging pile in advance to obtain an edge calculation result;

and the charging plan module is used for inputting the edge calculation result to a cloud server so that the cloud service unit can adjust the power distribution and the charging priority of the target charging pile according to the edge calculation result to obtain a target charging plan, and feeding the target charging plan back to the edge calculation unit so as to control the target charging pile to charge the vehicle to be charged.

In an embodiment of the present application, there is further provided a computer readable storage medium, where the computer readable storage medium includes a stored computer program, and the computer program executes the above-mentioned charging pile control method based on multi-modal sensing and edge calculation.

In an embodiment of the present application, there is also provided an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the above-mentioned charging pile control method based on multi-modal sensing and edge calculation through the computer program.

The invention has the beneficial effects that:

The method comprises the steps of obtaining multi-mode sensing information corresponding to a target charging pile, wherein the multi-mode sensing information comprises environment information, charging pile working information and vehicle information to be charged, carrying out edge calculation on the multi-mode sensing information through an edge calculation unit which is arranged around the target charging pile in advance to obtain an edge calculation result, and finally, transmitting the edge calculation result to a cloud server to enable the cloud service unit to adjust power distribution and charging priority of the target charging pile according to the edge calculation result to obtain a target charging plan, and feeding the target charging plan back to the edge calculation unit to control the target charging pile to charge the vehicle to be charged. According to the method, the edge computing unit can process multi-mode sensing information in real time at a place close to a data source, delay of data transmission is reduced, the charging piles can be rapidly adjusted according to the latest environment and working state information, response speed and decision accuracy are improved, the charging piles can be dynamically adjusted through real-time data processing, unnecessary waiting time and resource waste are avoided, accordingly charging efficiency is improved, future electricity consumption requirements and environment changes can be predicted through analysis of environment information, working modes of the charging piles are adjusted in advance, potential problems can be timely found and solved through monitoring of the working information of the charging piles, the charging piles are guaranteed to be always in an optimal working state, personalized charging services can be provided for different types of vehicles through analysis of the vehicle information to be charged, a charging plan is optimized, a cloud server performs global optimization according to edge computing results, a target charging plan is generated, the cloud server can balance loads of the charging piles through a global view angle, local overload or idle conditions are avoided, overall charging efficiency is improved, and the cloud server can dynamically adjust power distribution and charging priority of the charging piles according to real-time data to ensure efficient utilization of resources. In addition, the cloud server can perform load balancing according to global data, the condition that some charging piles are overloaded and other charging piles are idle is avoided, the power consumption load can be smoothed through reasonably distributing the charging tasks, fluctuation is reduced, sudden power consumption demands can be flexibly dealt with through dynamically adjusting the power distribution and the charging priority of the charging piles, and the load surge in a short time is avoided.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:

Fig. 1 is a schematic view showing an application environment of a charging pile control method based on multi-modal sensing and edge calculation according to an exemplary embodiment of the present application;

FIG. 2 is a flow chart of a method for controlling a charging pile based on multi-modal sensing and edge calculation according to an exemplary embodiment of the present application;

FIG. 3 is a schematic diagram of a charging pile control device based on multi-modal sensing and edge calculation according to an exemplary embodiment of the present application;

Fig. 4 is a schematic structural view of an electronic device according to an exemplary embodiment of the present application.

Detailed Description

In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

In an embodiment of the present application, a method for controlling a charging pile based on multi-modal sensing and edge calculation is provided, optionally, as an optional implementation manner, the method for controlling a charging pile based on multi-modal sensing and edge calculation may be applied, but not limited to, the environment shown in fig. 1. Fig. 1 is a schematic diagram of an application environment of a charging pile control method based on multi-modal sensing and edge computing according to an exemplary embodiment of the present application, and referring to fig. 1, the implementation environment includes a charging pile 101, a charging pile control terminal 102, and a cloud server 103, where the charging pile 101, the charging pile control terminal 102, and the cloud server 103 may communicate, but are not limited to, through a network, and the cloud server 103 may perform operations, such as a data writing operation or a data reading operation, on a database. The charging pile control terminal 102 may include, but is not limited to, a man-machine interaction screen, a processor and a memory. The human-machine interaction screen described above may be used, but is not limited to, for displaying a target charging plan. The processor may be, but is not limited to, configured to perform a corresponding operation in response to the man-machine interaction operation, or generate a corresponding instruction, and send the generated instruction to the cloud server 103. The memory is used for storing relevant stored data such as environment information, charging pile working information and vehicle information to be charged. The charging pile control terminal 102 may also include, but is not limited to, an edge calculation unit and a data processing terminal.

Alternatively, the data may be collected by the charging stake control terminal 102, for example, to collect and pre-process multi-modal awareness information corresponding to the target charging stake.

As an alternative, the following steps in the charging pile control method based on multi-modal awareness and edge calculation may be performed on the charging pile control terminal 102:

As an alternative manner, the cloud server 103 may adjust the power allocation and the charging priority of the target charging pile according to the edge calculation result, obtain the target charging plan, and feed back the target charging plan to the edge calculation unit, so as to control the target charging pile to charge the vehicle to be charged.

According to the method, the edge computing unit can process multi-mode sensing information in real time at a place close to a data source, delay of data transmission is reduced, the charging piles can be rapidly adjusted according to the latest environment and working state information, response speed and decision accuracy are improved, the charging piles can be dynamically adjusted through real-time data processing, unnecessary waiting time and resource waste are avoided, accordingly charging efficiency is improved, future electricity consumption requirements and environment changes can be predicted through analysis of environment information, working modes of the charging piles are adjusted in advance, potential problems can be timely found and solved through monitoring of the working information of the charging piles, the charging piles are guaranteed to be always in an optimal working state, personalized charging services can be provided for different types of vehicles through analysis of the vehicle information to be charged, a charging plan is optimized, a cloud server performs global optimization according to edge computing results, a target charging plan is generated, the cloud server can balance loads of the charging piles through a global view angle, local overload or idle conditions are avoided, overall charging efficiency is improved, and the cloud server can dynamically adjust power distribution and charging priority of the charging piles according to real-time data to ensure efficient utilization of resources. In addition, the cloud server can perform load balancing according to global data, the condition that some charging piles are overloaded and other charging piles are idle is avoided, the power consumption load can be smoothed through reasonably distributing the charging tasks, fluctuation is reduced, sudden power consumption demands can be flexibly dealt with through dynamically adjusting the power distribution and the charging priority of the charging piles, and the load surge in a short time is avoided.

Optionally, in this embodiment, the network may include, but is not limited to, a wireless network, where the wireless network includes bluetooth, WIFI, and other networks that implement wireless communications. The cloud server may be a single server or a server cluster composed of a plurality of servers. The above is merely an example, and is not limited in any way in the present embodiment.

As an optional example, the present embodiment does not limit the execution subject of the above-described charging pile control method based on multi-modal sensing and edge calculation, and part or all of the steps of the above-described charging pile control method based on multi-modal sensing and edge calculation may be executed on the charging pile control terminal 102.

In an embodiment of the application, a charging pile control method based on multi-modal sensing and edge calculation is provided. Fig. 2 is a flowchart illustrating a method for controlling a charging pile based on multi-modal sensing and edge calculation according to an exemplary embodiment of the present application, referring to fig. 2, the method for controlling a charging pile based on multi-modal sensing and edge calculation includes the following steps S210 to S230:

In step S210, multi-mode sensing information corresponding to the target charging pile is obtained, where the multi-mode sensing information includes environment information, charging pile working information, and vehicle information to be charged.

The multi-modal awareness information refers to information obtained from a plurality of different types of data sources, which together describe various environments and states around the target charging pile. The goal of the multimodal awareness information is to provide comprehensive, real-time data support for more accurate analysis and decision making.

The environmental information is data describing the surrounding environment conditions of the charging pile, and comprises, but is not limited to, people flow density, people flow near the charging pile, traffic flow, vehicle traffic conditions, weather conditions and weather parameters such as temperature, humidity, wind speed and the like near the charging pile, wherein the people flow is monitored in real time through a camera or other sensors, the people movement conditions are known, the traffic flow is monitored in real time through the camera or other sensors, and the weather parameters such as the temperature, the humidity, the wind speed and the like near the charging pile are monitored in real time through the weather sensors.

The charging pile working information refers to data describing the working state of the charging pile, and comprises, but is not limited to, current loads, working modes, fault conditions, historical data, historical working data, and fault records, wherein the current loads are used for monitoring the current load condition of the charging pile in real time through current and voltage sensors to know the power output of the charging pile, the working modes are used for monitoring the working modes (such as standby, charging and fault) of the charging pile in real time through state monitoring sensors, the key component states of the charging pile are monitored in real time through fault detection sensors to know whether faults exist or not, and the historical working data of the charging pile are recorded, wherein the historical working data comprise past loads, working modes and fault records.

The vehicle information to be charged refers to related data describing the vehicle to be charged and a user thereof, including but not limited to vehicle battery State, SOC (State of charge), SOH (State of health) and temperature of a battery monitored in real time through a Battery Management System (BMS) of the vehicle, vehicle type, type of the vehicle to be charged identified through a camera or other sensors, user reservation information, reservation information of the user acquired through a user application program, including reservation time, vehicle type, charging requirement, and the like.

In step S220, the multi-mode sensing information is subjected to edge calculation by an edge calculation unit pre-configured around the target charging pile, so as to obtain an edge calculation result.

The edge computing unit is a computing device disposed around a proximity data source (such as a charging pile) and is used for processing and analyzing data collected from a sensor and other data sources in real time, and the main characteristics and functions of the edge computing unit include close-range disposition, the edge computing unit is generally disposed around a data generation position, such as the charging pile to reduce delay of data transmission, high-performance computing capability, the edge computing unit is provided with enough computing resources to perform complex real-time data processing and analysis, low-power design, the edge computing unit is generally provided with low-power design to prolong service life in order to adapt to outdoor or embedded environments, network connection is provided with network connection function, and processing results can be transmitted to a cloud server or other central nodes, and security is provided with security mechanisms, such as encryption transmission, access control and the like, so as to protect security and privacy of the data.

The edge calculation refers to data processing and analysis on edge equipment for data generation, and all data are not transmitted to a central server or a cloud server for processing.

In step S230, the edge calculation result is input to a cloud server, so that the cloud service unit adjusts the power distribution and the charging priority of the target charging pile according to the edge calculation result, and a target charging plan is obtained, and the target charging plan is fed back to the edge calculation unit, so as to control the target charging pile to charge the vehicle to be charged.

The power distribution is to reasonably distribute the output power of the charging pile according to the working state of the current charging pile, the environmental condition and the requirement of the vehicle to be charged, and the reasonable power distribution can improve the charging efficiency, optimize the resource utilization and reduce the energy waste. The power distribution method comprises the steps of dynamically adjusting output power of each charging pile according to real-time data to ensure efficient utilization of resources, balancing loads of each charging pile through a global optimization algorithm to avoid local overload or idle conditions, reducing energy consumption as much as possible on the premise of ensuring charging requirements, improving energy efficiency, and automatically adjusting power of other charging piles when faults of the charging piles are detected to ensure continuity of charging tasks.

The charging priority refers to determining which vehicles should be charged with priority according to factors such as charging requirements of the vehicles, reservation information of users, environmental conditions and the like. The reasonable charging priority can improve user satisfaction, optimize charging experience, determine priority according to reservation information and charging requirements of users, for example, emergency charging requests can be processed preferentially, determine priority according to battery states and charging requirements of vehicles, for example, vehicles with low battery capacity can be charged preferentially, adjust charging priority according to environment information (such as people flow density, traffic flow and weather conditions), for example, more charging requests can be processed preferentially in peak hours, predict future charging requirements according to charging history data of users, and adjust charging priority in advance.

According to the embodiment of the application, the edge computing unit can process multi-mode sensing information in real time at a place close to a data source, delay of data transmission is reduced, the charging pile can be rapidly adjusted according to the latest environment and working state information, response speed and decision accuracy are improved, the charging pile can dynamically adjust a charging strategy through real-time data processing, unnecessary waiting time and resource waste are avoided, accordingly charging efficiency is improved, future electricity consumption requirements and environment changes can be predicted through analysis of environment information, working modes of the charging pile are adjusted in advance, potential problems can be timely found and solved through monitoring of the working information of the charging pile, the charging pile is always in an optimal working state, personalized charging service can be provided for different types of vehicles through analysis of the vehicle to be charged, a charging plan is optimized, a cloud server is globally optimized according to an edge computing result, a target charging plan is generated, loads of each charging pile can be balanced through a global view angle, local overload or idle condition is avoided, overall charging efficiency is improved, and the cloud server can dynamically adjust the charging pile and power distribution and efficient utilization of the power priority of the charging pile according to real-time data. In addition, the cloud server can perform load balancing according to global data, the condition that some charging piles are overloaded and other charging piles are idle is avoided, the power consumption load can be smoothed through reasonably distributing the charging tasks, fluctuation is reduced, sudden power consumption demands can be flexibly dealt with through dynamically adjusting the power distribution and the charging priority of the charging piles, and the load surge in a short time is avoided.

In an embodiment of the present application, the environmental information includes people flow density, traffic flow, weather conditions;

In an embodiment of the present application, the obtaining multi-mode sensing information corresponding to the target charging pile includes:

The method comprises the steps of acquiring a first video stream of an area around a charging pile through a camera, and identifying and tracking pedestrians in the video by using a pedestrian detection and tracking algorithm. And acquiring a second video stream through cameras arranged on the road around the charging pile, and identifying and tracking vehicles in the video by using a vehicle detection and tracking algorithm. And (3) acquiring weather conditions, including temperature, humidity, wind speed and the like, near the charging pile by calling a third party weather API (interface). The method helps to predict the use requirement of the charging pile, optimize the charging plan and ensure that the equipment works in a proper environment.

The voltage and current of the charging pile are monitored in real time through a voltage sensor and a current sensor which are arranged on a power supply line of the charging pile, and the current load is calculated. The working mode (such as standby, charging, fault and the like) of the charging pile is monitored in real time through a state detection sensor arranged in the charging pile, the state detection sensor (such as a switch sensor and a state register) is used for collecting data, and the working mode is determined through logic judgment. The working states of the components are monitored in real time through fault detection sensors arranged on the charging pile power supply unit, the communication unit and the charging interface, and whether faults exist or not is determined. Data is collected by using fault detection sensors (such as temperature sensors, current sensors and state registers), and fault conditions are determined through logic judgment and threshold comparison.

In this embodiment, through real-time data processing and analysis, the charging pile can quickly make adjustments according to the latest environment and working state information, so as to avoid unnecessary waiting time and resource waste, thereby improving charging efficiency. The load of each charging pile can be balanced by dynamically adjusting the power distribution and the charging priority of the charging pile, so that the local overload or idle condition is avoided, and the resource utilization is optimized. Through global optimization and dynamic adjustment, the power utilization load can be smoothed, the load surge in a short time is reduced, the power utilization load fluctuation is reduced, and the stability of the power grid is improved. Through personalized charging service and optimized charging plan, the charging requirement of the user can be met, and the user satisfaction degree and charging experience are improved. Through the fault condition of real-time supervision fills electric pile and in time handles, can ensure to fill electric pile's normal operating, reinforcing reliability. Through the edge calculation and the safe communication protocol, the safety and privacy of the data can be protected, and the risk of data leakage is reduced.

In an embodiment of the present application, the performing edge calculation on the multi-mode sensing information to obtain an edge calculation result includes:

and analyzing the fusion characteristics to obtain an edge calculation result.

Numerical characteristics (such as people flow density, traffic flow, temperature, humidity, wind speed and the like) in the environment information are converted into standard normal distribution, so that the average value is 0, and the standard deviation is 1. The standardization processing can enable the features with different scales to have comparability, and the effects of subsequent feature extraction and model training are improved.

Wherein, category characteristics (such as working mode, fault code, etc.) in the charging pile working information are converted into One-Hot Encoding (One-Hot Encoding). The one-time thermal encoding may convert class features into numerical features, making it suitable for machine learning models.

Wherein the numerical feature in the vehicle information to be charged is converted into a range of 0 to 1 using a min-max scaler. The normalization processing can enable the features with different scales to have comparability, and the effects of subsequent feature extraction and model training are improved.

The method comprises the steps of constructing a characteristic extraction network of multi-mode sensing information, extracting characteristics of each mode sensing information by using a full connection layer or an attention mechanism, and fusing the characteristics to generate comprehensive characteristic representation. The complex relation between the multi-mode information can be captured by the feature extraction and fusion, the comprehensive feature representation is generated, and the prediction and decision making capability of the model is improved.

The fusion characteristics are analyzed by using a regression model or a classification model, and edge calculation results, such as a power distribution scheme of the charging pile, a charging priority list and the like, are generated. The edge calculation result can be directly used for controlling the behavior of the charging pile, and the instantaneity and the response speed are improved.

In the embodiment, through standardization, single thermal coding and normalization processing, dimensional differences among different features are eliminated, and the efficiency and accuracy of data processing are improved. The deep learning model is used for extracting the characteristics of the multi-mode perception information, and characteristic fusion is carried out, so that complex relations between data can be captured, comprehensive characteristic representation is generated, and the prediction and decision making capability of the model is improved. The edge computing unit processes and analyzes the real-time data in a place close to the data source, so that the delay of data transmission is reduced, and the real-time performance and response speed are improved.

In an embodiment of the present application, the extracting the features corresponding to each mode in the preprocessed multi-mode sensing information, and fusing the features corresponding to each mode to obtain the fused features includes:

Wherein, using multi-layer convolution layer and pooling layer, gradually extracting local and global features in the environment information. And capturing complex modes in the environment information, such as the spatial distribution of the traffic density, the change trend of the traffic flow and the like.

Wherein advanced features in the charge pile operation information are extracted using a fully connected neural network (FCN). And capturing complex relations in the working information of the charging pile, such as the change trend of the current load, the conversion rule of the working mode and the like.

Wherein discrete class features (e.g., vehicle type, user subscription information) are mapped to a low-dimensional space using an embedding layer (Embedding Layer) to generate continuous feature vectors. The discrete category features are converted into numerical features, so that the numerical features are suitable for subsequent feature extraction and fusion.

Wherein feature vectors are converted into queries (Query) and keys (Key) by linear transformations (e.g., full-join layer) to provide input for subsequent attention mechanisms. A Query (Query) and a Key (Key) are two important parts in the attention mechanism for calculating an attention score, wherein a Query vector and a Key vector respectively represent different aspects of a feature vector, and the attention score is calculated through the inner product of the Query and the Key to reflect the similarity between features.

When the attention score is calculated through the inner product of the query vector and the key vector, the dot product of the query vector and the key vector is calculated, and an attention score matrix is generated to reflect the similarity and importance among different features.

The attention score is converted into attention weight through normalization processing, and the attention weight represents the importance degree of different characteristics and is used for subsequent weighting processing.

Wherein the attention weights are multiplied by the corresponding feature vectors and then summed to generate a fused feature, generating a comprehensive feature representation for subsequent analysis and decision.

In this embodiment, the environmental features, the working features and the features of the vehicle to be charged are extracted through the convolutional neural network, the fully connected neural network and the embedded layer respectively, so that complex modes and relations in the multi-mode sensing information can be captured, and advanced feature representation is generated. The attention mechanism is used for weighting the features of different modes, so that important features can be highlighted, unimportant features can be restrained, and more comprehensive and effective fusion features can be generated. The fusion characteristics can better reflect the overall characteristics of the multi-mode perception information, and the accuracy and the robustness of subsequent analysis and decision making are improved. The edge computing unit processes and analyzes the real-time data in a place close to the data source, so that the delay of data transmission is reduced, and the real-time performance and response speed are improved.

In an embodiment of the present application, the analyzing the fusion feature to obtain an edge calculation result includes:

In which a relationship structure between the objects to be analyzed is defined, i.e. which objects there are associations or interactions between them. The Graph structure (Graph) is used for representing the relation between objects to be analyzed, the nodes represent the objects to be analyzed (such as target charging piles and vehicles to be charged), and the edges represent the relation between the nodes (such as connection between the charging piles and the vehicles), so that the model is helped to understand the interaction between different objects, and the analysis accuracy is improved.

Wherein, when the characteristic information is acquired from the neighboring node directly connected with the current node, the characteristic information is acquired from the neighboring node using a message passing mechanism in a Graph Neural Network (GNN). And capturing the dependency relationship among the nodes, and enriching the characteristic representation of the current node.

The aggregation and updating mechanism in the Graph Neural Network (GNN) is used for updating the initialization feature vector of the current node according to the information acquired from the adjacent nodes to generate the target feature vector, so that richer and more accurate node feature representation can be generated, and the prediction capability of the model is improved.

The regression model is a machine learning model for predicting continuous values, such as linear regression, decision tree regression, neural network regression, and the like. And constructing and training a regression model by using a deep learning framework, and predicting the power distribution scheme, charging priority and other edge calculation results of the charging pile according to the target feature vector of the node.

And predicting the node characteristic vector by using a regression model to generate results such as a power distribution scheme, charging priority and the like of the charging pile. And generating a specific edge calculation result for controlling the behavior of the charging pile, and improving the instantaneity and the response speed.

In this embodiment, by using a message passing mechanism of the graph neural network, information is obtained from neighboring nodes and node feature vectors are updated, so that richer and more accurate feature representations can be generated, and the prediction capability of the model is improved. The graph structure is used for representing the relation among the objects to be analyzed, so that the interaction among different objects can be better captured, comprehensive characteristic representation is generated, and the robustness of the model is improved. The updated node feature vector is predicted through the regression model, a specific edge calculation result is generated, and the prediction precision and generalization capability of the model can be improved.

In an embodiment of the present application, the adjusting the power allocation and the charging priority of the target charging pile according to the edge calculation result to obtain a target charging plan includes:

and obtaining the target charging plan based on the target action.

The state of the environment describes various parameters of the current system running condition, including the power state of the target charging pile and the state of the vehicle to be charged. The executable action is an action that can be taken to adjust the operational state of the charging stake.

Wherein the reward function, the function used to evaluate the effect of the executable action, guides the decision by quantifying the performance of the different aspects. And (3) evaluating the charge quantity of the charging pile in unit time, wherein the higher the charge efficiency reward function is, the better the charge quantity is. The user satisfaction rewarding function evaluates the user's charging experience, such as waiting time, charging speed, etc., the higher the better. Cost benefit rewarding function, namely, evaluating the cost benefit of the charging process, such as electricity charge, equipment maintenance charge and the like, the lower the better. When evaluating the effect of the executable action according to the preset reward function, the value of each reward function can be calculated according to the current environment state and the action taken to help to select the optimal action so as to achieve the optimal overall effect.

Where the expected return refers to the total rewards expected to be available after taking some action in the current environmental state. By means of dynamic programming or reinforcement learning methods, long-term benefits after taking a certain action are estimated, long-term effects of different actions can be estimated, and optimal actions can be selected. And comprehensively considering the effect and the expected return of the current action, and selecting the optimal action. The action with the highest expected return is selected by using an optimization algorithm, so that the action can obtain the best overall effect in the current environment. And generating a specific charging plan according to the selected target action, converting the target action into a specific power distribution scheme and a charging priority list, guiding the actual operation of the charging pile, and realizing optimized charging management.

For example, it is assumed that in one intelligent charging station, the edge calculation unit generates the target charging plan by adjusting the power distribution and the charging priority of the charging stake according to the edge calculation result by:

a. Determining the state of the environment and the actions that can be performed:

the current power state of the charging pile A is that the output power is 50kW, the working mode is charging, and the load is 80%;

The State of the vehicle V1 is that the battery SOC (State of Charge) is 30%, the SOH (State of health) is 90%, and the reservation time is 10:00;

The state of the vehicle V2 is that the battery SOC is 40%, the SOH is 85%, and the reservation time is 10:30;

the state of the vehicle V3 is that the battery SOC is 50%, the SOH is 80%, and the reservation time is 11:00;

adjusting the power distribution of the charging pile A, namely distributing power to V1, V2 and V3;

setting charging priority, namely determining the charging sequence of V1, V2 and V3;

b. According to the reward function, evaluating the effect of the executable action:

user satisfaction rewarding function:

Action 1, user satisfaction is 75%;

Action 2, user satisfaction is 85%;

cost benefit rewarding function:

Action 1, cost effectiveness is 70%;

action 2, the cost benefit is 80%;

the effect of action 1 is that power is distributed to V1, V2 and V3, and the charging efficiency is 85%;

The effect of action 2 is that the charging V1 is prioritized, and the charging efficiency is 90%;

C. determining an expected return:

The expected return for action 1, r1=0.85×0.75×0.70;

The expected return for action 2, r2=0.90×0.85×0.80;

d. determining a target action:

select target action-select action 2 with highest expected return.

E. Generating a target charging plan:

the power distribution of the charging pile A is V150kW, V220kW and V310kW.

Charging priority V1> V2> V3.

In this embodiment, the charging efficiency of the charging pile is improved by evaluating the charging efficiency reward function and selecting an optimal power distribution scheme. And the charging priority is optimized by evaluating the user satisfaction rewarding function, so that the waiting time of the user is reduced, and the user satisfaction is improved. By evaluating the cost-benefit rewarding function, the most cost-effective charging scheme is selected, reducing the cost of the charging process. And comprehensively considering the charging efficiency, the user satisfaction and the cost effectiveness, generating an optimal charging plan, optimizing the resource utilization and avoiding the local overload or idle condition. The edge computing unit processes and analyzes the real-time data in a place close to the data source, so that the delay of data transmission is reduced, and the real-time performance and response speed are improved.

According to the embodiment, the edge computing unit can process multi-mode sensing information in real time at a place close to a data source, delay of data transmission is reduced, the charging piles can be rapidly adjusted according to the latest environment and working state information, response speed and decision accuracy are improved, the charging piles can dynamically adjust a charging strategy through real-time data processing, unnecessary waiting time and resource waste are avoided, accordingly charging efficiency is improved, future electricity consumption requirements and environment changes can be predicted through analysis of the environment information, working modes of the charging piles are adjusted in advance, potential problems can be timely found and solved through monitoring of the working information of the charging piles, the charging piles are always in an optimal working state, personalized charging services can be provided for different types of vehicles through analysis of the vehicle to be charged, a charging plan is optimized, a cloud server performs global optimization according to an edge computing result, a target charging plan is generated, the cloud server can balance loads of the charging piles through a global view angle, local overload or idle conditions are avoided, overall charging efficiency is improved, the cloud server can dynamically adjust the power distribution of the charging piles and the power distribution of the charging piles according to real-time data, and high efficiency of the charging piles is guaranteed. In addition, the cloud server can perform load balancing according to global data, the condition that some charging piles are overloaded and other charging piles are idle is avoided, the power consumption load can be smoothed through reasonably distributing the charging tasks, fluctuation is reduced, sudden power consumption demands can be flexibly dealt with through dynamically adjusting the power distribution and the charging priority of the charging piles, and the load surge in a short time is avoided.

In an embodiment of the present application, a charging pile control device based on multi-modal sensing and edge calculation is also provided. Fig. 3 is a schematic view of a charging pile control device based on multi-modal sensing and edge calculation according to an exemplary embodiment of the present application, and referring to fig. 3, the device includes:

The multi-mode sensing information acquisition module 301 is configured to acquire multi-mode sensing information corresponding to a target charging pile, where the multi-mode sensing information includes environment information, charging pile working information, and vehicle information to be charged;

the edge calculation module 302 is configured to perform edge calculation on the multi-mode sensing information through an edge calculation unit that is pre-configured around the target charging pile, so as to obtain an edge calculation result;

And the charging plan module 303 is configured to output the edge calculation result to a cloud server, so that the cloud service unit adjusts the power distribution and the charging priority of the target charging pile according to the edge calculation result to obtain a target charging plan, and feeds back the target charging plan to the edge calculation unit to control the target charging pile to charge the vehicle to be charged.

According to the charging pile control device based on multi-mode sensing and edge computing, the edge computing unit can process multi-mode sensing information in real time at a place close to a data source, delay of data transmission is reduced, the charging pile can be quickly adjusted according to the latest environment and working state information, response speed and decision accuracy are improved, the charging pile can be dynamically adjusted through real-time data processing, unnecessary waiting time and resource waste are avoided, accordingly charging efficiency is improved, future electricity consumption requirements and environment changes can be predicted through analysis of environment information, working modes of the charging pile can be adjusted in advance, potential problems can be timely found and solved through monitoring of the working information of the charging pile, the charging pile is always in an optimal working state, personalized charging service can be provided for different types of vehicles through analysis of vehicle information to be charged, the charging plan is optimized, a cloud server is globally optimized according to an edge computing result, loads of the charging pile can be balanced through a global view angle, local overload or idle condition is avoided, overall charging efficiency is improved, and the cloud server can be enabled to adjust the dynamic power distribution of the charging pile according to the global view angle. In addition, the cloud server can perform load balancing according to global data, the condition that some charging piles are overloaded and other charging piles are idle is avoided, the power consumption load can be smoothed through reasonably distributing the charging tasks, fluctuation is reduced, sudden power consumption demands can be flexibly dealt with through dynamically adjusting the power distribution and the charging priority of the charging piles, and the load surge in a short time is avoided.

The specific embodiments of the charging pile control device based on multi-modal sensing and edge calculation in the present application may refer to the examples shown in the charging pile control method based on multi-modal sensing and edge calculation, and the examples are not described herein.

In an embodiment of the present application, an electronic device for implementing the above-mentioned charging pile control method based on multi-modal sensing and edge calculation is also provided. The electronic device comprises a memory in which a computer program is stored, and a processor arranged to execute the above-described charging pile control method based on multi-modal awareness and edge calculation by means of the computer program.

Referring to fig. 4, fig. 4 is a schematic structural view of an electronic device according to an exemplary embodiment of the present application. The computer system 400 includes a central processing unit (Central Processing Unit, CPU) 401 that can perform various appropriate actions and processes, such as performing the methods described in the above embodiments, according to a program stored in a Read-Only Memory (ROM) 402 or a program uploaded from a storage portion 408 into a random access Memory (RandomAccess Memory, RAM) 403. In the RAM 403, various programs and data required for the system operation are also stored. The CPU 401, ROM 402, and RAM 403 are connected to each other by a bus 404. An Input/Output (I/O) interface 405 is also connected to bus 404.

Connected to the I/O interface 405 are an input section 406 including a keyboard, a mouse, and the like, an output section 407 including a display such as a Cathode Ray Tube (CRT), a Liquid crystal display (Liquid CRYSTAL DISPLAY, LCD), and a speaker, a storage section 408 including a hard disk, and the like, and a communication section 409 including a network interface card such as a LAN (LocalAreaNetwork) card, a modem, and the like. The communication section 409 performs communication processing via a network such as the internet. The drive 410 is also connected to the I/O interface 405 as needed. A removable medium 411 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is installed on the drive 410 as needed, so that a computer program read therefrom is installed into the storage section 408 as needed.

In particular, according to embodiments of the present application, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising a computer program for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion 409 and/or installed from the removable medium 411. When executed by a Central Processing Unit (CPU) 401, performs the various functions defined in the system of the present application.

It should be noted that, the computer readable medium shown in the embodiments of the present application may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium may be, for example, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-Only Memory (ROM), an erasable programmable read-Only Memory (Erasable Programmable Read Only Memory, EPROM), a flash Memory, an optical fiber, a portable compact disc read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer-readable signal medium may comprise a data signal propagated in baseband or as part of a carrier wave, with a computer-readable computer program embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, etc., or any suitable combination of the foregoing.

The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Where each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units involved in the embodiments of the present application may be implemented by software, or may be implemented by hardware, and the described units may also be provided in a processor. Wherein the names of the units do not constitute a limitation of the units themselves in some cases.

In another aspect, the present application further provides a computer readable storage medium, where the computer readable storage medium includes a stored computer program, where the computer program executes the above-mentioned charging pile control method based on multi-modal sensing and edge calculation. The computer-readable storage medium may be included in the electronic device described in the above embodiment or may exist alone without being incorporated in the electronic device.

Another aspect of the application also provides a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the charging pile control method based on the multi-modal sensing and the edge calculation provided in the above-described respective embodiments.

The above embodiments are merely illustrative of the principles of the present invention and its effectiveness, and are not intended to limit the invention. Modifications and variations may be made to the above-described embodiments by those skilled in the art without departing from the spirit and scope of the invention. It is therefore intended that all equivalent modifications and changes made by those skilled in the art without departing from the spirit and technical spirit of the present invention shall be covered by the appended claims.

Claims

1. A charging pile control method based on multi-modal sensing and edge calculation, the method comprising:

2. The charging pile control method based on multi-modal sensing and edge calculation as set forth in claim 1, wherein:

the environment information comprises people flow density, traffic flow and weather conditions;

3. The method for controlling a charging pile based on multi-modal sensing and edge calculation as set forth in claim 2, wherein the obtaining multi-modal sensing information corresponding to the target charging pile includes:

4. The method for controlling a charging pile based on multi-modal sensing and edge calculation as set forth in claim 1, wherein the performing edge calculation on the multi-modal sensing information to obtain an edge calculation result includes:

and analyzing the fusion characteristics to obtain an edge calculation result.

5. The method for controlling a charging pile based on multi-modal sensing and edge calculation as set forth in claim 4, wherein the extracting the characteristics corresponding to each modality in the preprocessed multi-modal sensing information and fusing the characteristics corresponding to each modality to obtain the fused characteristics includes:

6. The method for controlling a charging pile based on multi-modal sensing and edge calculation as set forth in claim 4, wherein the analyzing the fusion features to obtain an edge calculation result includes:

7. The method for controlling a charging pile based on multi-modal sensing and edge calculation as set forth in claim 1, wherein said adjusting the power allocation and the charging priority of the target charging pile according to the result of the edge calculation to obtain a target charging plan includes:

and obtaining the target charging plan based on the target action.

8. A charging pile control device based on multi-modal sensing and edge calculation, the device comprising:

9. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored computer program, wherein the computer program when run performs the multi-modal awareness and edge calculation based charging pile control method of any one of claims 1 to 7.

10. An electronic device comprising a memory and a processor, characterized in that the memory has stored therein a computer program, the processor being arranged to execute the multi-modal awareness and edge calculation based charging pile control method according to any one of claims 1 to 7 by means of the computer program.