WO2019085567A1 - 机器人的行走预测及控制方法 - Google Patents

机器人的行走预测及控制方法 Download PDF

Info

Publication number
WO2019085567A1
WO2019085567A1 PCT/CN2018/098912 CN2018098912W WO2019085567A1 WO 2019085567 A1 WO2019085567 A1 WO 2019085567A1 CN 2018098912 W CN2018098912 W CN 2018098912W WO 2019085567 A1 WO2019085567 A1 WO 2019085567A1
Authority
WO
WIPO (PCT)
Prior art keywords
robot
point
line
predicted point
reference direction
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2018/098912
Other languages
English (en)
French (fr)
Inventor
李永勇
肖刚军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhuhai Amicro Semiconductor Co Ltd
Original Assignee
Zhuhai Amicro Semiconductor Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhuhai Amicro Semiconductor Co Ltd filed Critical Zhuhai Amicro Semiconductor Co Ltd
Priority to JP2020523718A priority Critical patent/JP7075994B2/ja
Priority to KR1020207012135A priority patent/KR102445731B1/ko
Priority to EP18871985.0A priority patent/EP3705968B1/en
Priority to US16/649,145 priority patent/US11526170B2/en
Publication of WO2019085567A1 publication Critical patent/WO2019085567A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0268Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means
    • G05D1/0274Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means using mapping information stored in a memory device
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/20Control system inputs
    • G05D1/24Arrangements for determining position or orientation
    • G05D1/246Arrangements for determining position or orientation using environment maps, e.g. simultaneous localisation and mapping [SLAM]
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/005Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 with correlation of navigation data from several sources, e.g. map or contour matching
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/20Instruments for performing navigational calculations
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3804Creation or updating of map data
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/60Intended control result
    • G05D1/617Safety or protection, e.g. defining protection zones around obstacles or avoiding hazards
    • G05D1/622Obstacle avoidance
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/60Intended control result
    • G05D1/644Optimisation of travel parameters, e.g. of energy consumption, journey time or distance
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D2101/00Details of software or hardware architectures used for the control of position
    • G05D2101/10Details of software or hardware architectures used for the control of position using artificial intelligence [AI] techniques

Definitions

  • the invention relates to the field of robots, and in particular to a method for predicting and controlling walking of a robot.
  • the present invention provides a walking prediction and control method for a robot, which can quickly and accurately predict a path situation in front of the robot, and can control the robot to adopt different walking modes for different road conditions to improve the walking efficiency of the robot.
  • the specific technical solutions of the present invention are as follows:
  • a walking prediction and control method for a robot includes the steps of: constructing a grid map based on a grid unit marked with a state; and establishing a dynamic detection model based on the current position of the robot as a reference point based on the grid map; The dynamic detection model is described to predict the path condition in front of the robot; and based on the prediction result, the walking mode of the robot is controlled.
  • the invention has the beneficial effects that: based on the grid map formed by the grid cells marked with various states, the robot can quickly and accurately predict the path ahead by the dynamic detection model during the walking process, thereby according to different path conditions. Reasonably control the walking mode of the robot, avoid the problem that the robot rushes out of the cliff or repeatedly enters the dangerous area, and improves the walking efficiency of the robot.
  • FIG. 1 is a flow chart of a walking prediction and control method for a robot according to the present invention.
  • FIG. 2 is a schematic diagram of a dynamic detection model according to the present invention.
  • FIG. 3 is a schematic diagram of converting local coordinates into global coordinates according to the present invention.
  • the robot of the invention is a kind of intelligent household appliances, and can automatically walk in some occasions automatically with certain artificial intelligence.
  • the mobile robot of the present invention comprises the following structure: a robotic body capable of autonomous walking with a driving wheel, a human-computer interaction interface is arranged on the body, and an obstacle detecting unit is arranged on the body.
  • An inertial sensor is disposed inside the body, and the inertial sensor includes an accelerometer and a gyroscope.
  • the driving wheel is provided with an odometer (generally a code wheel) for detecting the walking distance of the driving wheel, and is also provided with a parameter capable of processing the relevant sensor. And can output a control signal to the control module of the execution component.
  • the method for predicting and controlling a walking of a robot includes the steps of: constructing a grid map based on a grid unit marked with a state; and based on the grid map, based on a current position of the robot
  • the point establishes a dynamic detection model; based on the dynamic detection model, predicts the path condition in front of the robot; and based on the prediction result, controls the walking mode of the robot.
  • the grid unit with the marked state refers to the grid unit that the robot has normally traveled as the unit that has been traveled, and the grid unit that detects the obstacle is marked as an obstacle unit by the robot, and the robot detects that the robot is stuck or slipped.
  • the grid unit is marked as a dangerous unit
  • the grid unit in which the robot detects the cliff is marked as a cliff unit
  • the grid unit that the robot has not traveled is marked as an unknown unit.
  • the grid unit in which the robot normally walks refers to an abnormal situation in which the robot walks through the grid unit without slipping, jamming, hitting an obstacle or rushing out of a cliff, and these abnormal conditions may cause the robot to Walking causes errors and even damages the robot.
  • the robot will walk smoothly and orderly at a predetermined speed and direction.
  • the method of the present invention is based on a grid map formed by grid cells marked with various states (the raster map is a map constructed and saved after the robot has previously traversed all regions), and the robot can be in the walking process.
  • the raster map is a map constructed and saved after the robot has previously traversed all regions
  • the establishing a dynamic detection model based on the current position of the robot as a reference point comprises the steps of: constructing a first circular arc with a radius of the first length and a radius of the second length with the reference point as a center Constructing a second circular arc line, wherein the first length is smaller than the second length; determining an area between the first circular arc line and the second circular arc line as a first prediction area; determining the The area other than the second arc line is the second prediction area; the first prediction point, the second prediction point, the third prediction point, and the first prediction area are determined by using the current direction of the robot walking as a reference direction.
  • the first predicted point is located in the reference direction, and a line connecting the first predicted point and the center of the circle and the reference direction constitute a first angle;
  • the second predicted point And the fourth predicted point is located at one side of the first predicted point, and the line connecting the second predicted point and the center of the circle and the line connecting the first predicted point and the center of the circle constitute a second angle, the first a line connecting the predicted point to the center of the circle and the first prediction a line connecting the point and the center of the circle constitutes a fourth angle;
  • the third predicted point and the fifth predicted point are located on the other side of the first predicted point, and the line connecting the third predicted point and the center of the circle
  • the line connecting the first predicted point and the center of the circle constitutes a third angle, the line connecting the fifth predicted point and the center of the circle and the line connecting the first predicted point and the center of the circle constitute a fifth angle;
  • the current direction in which the robot walks is Determining, in the reference direction, a sixth prediction point, a seventh
  • the first length and the second length may be correspondingly set according to actual conditions.
  • the first length may be set to 1.1 times to 1.5 times the radius of the robot body; the second length may be set. It is 1.6 times to 2.1 times the radius of the robot body.
  • the first predicted point is M point
  • the second predicted point is L point
  • the third predicted point is N point
  • the fourth predicted point is K point
  • the fifth predicted point is point O
  • the sixth predicted point is
  • the seventh predicted point is the G point
  • the eighth predicted point is the I point.
  • Point P is the reference point (ie, the center of the circle), and a circle outside the point P represents the body of the robot, and a semi-circular arc having a smaller inner diameter outside the circle is the first circular arc, and a larger inner half outside the circle
  • the arc is the second arc line. Since the PM direction is the current direction of the robot (ie, the reference direction) and the M point is in the reference direction, the ⁇ MPM is the first angle and the angle value is 0°. Since the H point is also in the reference direction, the angle value of the sixth angle ⁇ HPM is also 0°.
  • ⁇ LPM is the second angle
  • ⁇ NPM is the third angle
  • ⁇ KPM is the fourth angle
  • ⁇ OPM is the fifth angle
  • ⁇ GPM is the seventh angle
  • ⁇ IPM is the eighth angle
  • the angle values of these angles can be Make the appropriate settings according to the actual situation.
  • the angle values of the second angle and the third angle range between 20° and 40°
  • the angle values of the fourth angle and the fifth angle range between 45° and 65°
  • the angle values of the seventh angle and the eighth angle range between 15° and 40°.
  • the first length is 1.2 times the radius of the robot body, and the second length is twice the radius of the robot body.
  • the values of the first length and the second length are suitable. If the setting is too short, when the abnormal situation such as a dangerous area or an obstacle is predicted, the robot can't avoid or adjust the walking state, and the prediction is not carried. The beneficial effect is that if it is set too long, it will consume relatively more computing resources, which leads to the inefficiency of prediction.
  • the first predicted point is located at an intersection of the first circular arc line and the reference direction (ie, M point in the figure); and the second predicted point is located at the first vertical The intersection of the line and the first parallel line; the third predicted point is located at an intersection of the first vertical line and the second parallel line; the fourth predicted point is located in the first line segment of the third parallel line; The fifth predicted point is located in the second line segment of the fourth parallel line.
  • the first vertical line is a vertical line that passes through the first prediction point and is perpendicular to the reference direction (ie, line ac in the figure);
  • the first parallel line is located on one side of the reference direction
  • the vertical distance from the reference direction is a parallel line of the robot body radius (ie, line gk in the figure);
  • the second parallel line is located on the other side of the reference direction,
  • the vertical distance from the reference direction is a parallel line of the radius of the robot body (ie, line hm in the figure);
  • the third parallel line is on one side of the reference direction, and parallel
  • a vertical distance from the reference direction is a parallel line of a first length (ie, a line ab in the figure);
  • the fourth parallel line is located on the other side of the reference direction, and parallel to the a reference direction, a vertical distance from the reference direction is a parallel line of a first length (ie, a line cd in the figure);
  • the first line segment is an intersection
  • the distance between point e and point P is half the radius of the robot body.
  • the K point can be set to any point in the ab line segment according to the actual situation; the O point can be set to any point in the cd line segment according to the actual situation.
  • the sixth predicted point is located at an intersection of the second circular arc line and the reference direction (ie, H point in the figure); the seventh predicted point is located at a third line segment of the third vertical line The eighth predicted point is located in the fourth line segment of the third vertical line.
  • the third vertical line is a vertical line that passes through the sixth prediction point and is perpendicular to the reference direction (ie, line fi in the figure);
  • the third line segment is the first parallel line and the a line segment between the intersection of the third vertical line and the intersection of the fifth parallel line and the third vertical line (ie, the fg line segment in the figure);
  • the fourth line segment is the second parallel line and the first a line segment between the intersection of the three vertical lines to the intersection of the sixth parallel line and the third vertical line (ie, the hi line segment in the figure);
  • the fifth parallel line is located on one side of the reference direction and parallel to The reference direction,
  • the vertical distance from the reference direction is a parallel line of the sum of the second length minus the difference of the first length and the first length (ie, the line fj in the figure);
  • the sixth parallel The line is located on the other side of the reference direction and parallel to the reference direction, and the vertical distance from the reference direction is a parallel line of the second length minus the difference between the
  • the lines described in the above embodiments are all virtual lines, which are cited for convenience in explaining the architecture of the dynamic detection model. In the actual operation of the robot, the above-mentioned lines are not present, but the robot is in the model architecture. Forecasts made within the scope of the.
  • the predicting a path condition in front of the robot based on the dynamic detection model includes the following steps: taking a current position of the robot as a local coordinate origin, and a current direction as a local Y-axis direction, and establishing an XY-axis local coordinate system;
  • the local coordinates in the XY-axis local coordinate system in which the first predicted point to the eighth predicted point are located are converted into global coordinates corresponding to the XY-axis global coordinate system;
  • the global coordinates are converted into raster coordinates;
  • the grid coordinates and the grid map are determined, and the state of the grid cells corresponding to the first predicted point to the eighth predicted point is determined. Only when the local coordinates of the predicted points of the dynamic detection model are converted into global coordinates can the grid coordinates in the grid map be matched, so that the robot can accurately predict the grid cells in front of the walking direction during the actual walking process. status.
  • the XQY coordinate system is an XY-axis global coordinate system
  • the X'PY' coordinate system is an XY-axis local coordinate system.
  • the circle represents the robot
  • the P point is the current position of the robot, and also serves as the origin of the local coordinate system.
  • the coordinate position of the P point in the global coordinate system is known, assuming (x, y).
  • the PM direction is the current direction of the robot, and can also be obtained from the detection data of the gyroscope, assuming that the angle of the current direction is ⁇ (ie, the angle between the PM direction and the Y-axis direction).
  • the N point is the third prediction point of the dynamic detection model, and the point is taken as an example to describe a method of converting local coordinates into corresponding global coordinates.
  • the local coordinate of the N point is (x3, y3)
  • the angle of the PM as the angle between the current direction and the Y axis is ⁇
  • the shortest distance from the intersection point of the X-axis direction is the xr3.
  • xr3 (x3*cos ⁇ -y3*sin ⁇ )
  • the path in front of the grid cannot be determined by the global coordinates
  • the path condition in the grid map is marked by the state of the grid unit. Therefore, the global coordinates need to be converted into the corresponding grid unit, and then the corresponding grid unit is determined. The state of the road to finalize the situation in front of the path to achieve the predicted effect.
  • the controlling the walking manner of the robot based on the prediction result includes the following steps: determining whether the grid unit corresponding to the first predicted point, the second predicted point, or the third predicted point is a dangerous unit If yes, controlling the robot to walk in the first walking manner; if not, determining whether the grid unit corresponding to the first predicted point, the second predicted point, or the third predicted point is a cliff unit or a barrier unit; if yes, controlling the robot to walk in a second walking manner; if not, determining whether the grid unit corresponding to the fourth predicted point or the fifth predicted point is a cliff unit or an obstacle unit; And controlling the robot to walk in a third walking manner; if not, determining whether the sixth predicted point, the seventh predicted point, or the eighth predicted point is a cliff unit or an obstacle unit; if yes, controlling the robot Walk in the fourth walking mode; if not, keep the robot's current walking mode.
  • the danger unit is a grid unit that detects that a robot is stuck or slipped
  • the obstacle unit is a grid unit in which the robot detects an obstacle
  • the cliff unit is a grid unit in which the robot detects the cliff.
  • the controlling robot walks in the first walking manner, comprising the steps of: controlling the robot to stop walking; determining whether the robot is in the bow type walking phase; if yes, controlling the robot to turn around; if not, determining that the robot is walking along the edge Stage, and control the robot to bypass the dangerous unit and continue along the edge.
  • the controlling robot walks in the second walking manner, and includes the following steps: if it is determined that the grid unit is an obstacle unit, the controlling robot reduces the walking speed by a first ratio; if the grid unit is determined to be a cliff The unit controls the robot to reduce the walking speed by a second ratio.
  • the controlling robot walks in the third walking manner, and includes the following steps: controlling the robot to reduce the walking speed by a third ratio.
  • the controlling robot walks in the fourth walking manner, and includes the following steps: controlling the robot to reduce the walking speed by a fourth ratio.
  • the ratio may be set according to actual conditions.
  • the first ratio is 0.5; the second ratio is 0.3; the third ratio is 0.7; and the fourth ratio is 0.8.
  • the robot When it is determined that there are dangerous units among the three predicted points closest to the robot in the first prediction area, the robot needs to avoid the corresponding dangerous area. To avoid the dangerous area, it is necessary to adopt different avoidance methods according to the walking stage of the robot.
  • the walking mode of the sweeping robot it is mainly divided into a bow type walking stage and an edge side walking stage.
  • the robot When the robot is in the bow-shaped walking stage, you can directly turn around when you encounter the dangerous area, and continue to sweep the other paths while the bow-shaped walking; when the robot is in the walking phase, you cannot turn around because the U-turning path will be repeated after the U-turn If the dangerous area is detected again and the U-turn is turned again, the repeated edges between the two dangerous areas will be formed.
  • the grid unit corresponding to the three predicted points closest to the robot is a walking unit that has been normally traveled, determining whether the grid unit corresponding to the fourth predicted point or the fifth predicted point is a cliff unit or The obstacle unit, because the two predicted points are on the outside of the robot, even if it is a dangerous unit, it can pass through the middle, so it is not necessary to judge whether the two predicted points are dangerous units.
  • the speed of the robot needs to be reduced to 0.7 times, because the garbage of the obstacle or the edge of the cliff is generally more, so it is required to pass slowly to improve the cleaning effect.
  • the purpose of the prediction is to control the speed of the robot in advance. When detecting one of them is a cliff unit or an obstacle unit, the speed of the robot is reduced to 0.8 times. If not, keep the robot's current walking mode and continue cleaning.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Manipulator (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

本发明涉及的机器人的行走预测和控制方法,包括如下步骤:基于标示过状态的栅格单元构建栅格地图;基于所述栅格地图,以机器人的当前位置为基准点建立一个动态检测模型;基于所述动态检测模型,预测机器人前方的路径情况;基于预测结果,控制机器人的行走方式。该方法基于标示有各种状态的栅格单元所构成的栅格地图,可以让机器人在行走过程中通过动态检测模型快速准确地预测前方的路径情况,从而根据不同的路径情况合理地控制机器人的行走方式,避免机器人冲出悬崖或者重复进入危险区域等问题,提高了机器人的行走效率。

Description

机器人的行走预测及控制方法 技术领域
本发明涉及机器人领域,具体涉及一种机器人的行走预测和控制方法。
背景技术
目前的机器人预测前方是否有障碍物,一般需要用红外检测,而危险区或者悬崖只能用视觉检测,这些方式都需要增加硬件成本。但是,如果不对前方的情况作出预测,有很多情况可能导致比较糟糕的结果,比如危险区,上次已经进入过了,这次因为没有预测而再次进入,会显得非常不智能。还有悬崖,如果不进行预测,机器人的速度比较快时,就有可能会冲过去,导致机器人跌落。
发明内容
为解决上述问题,本发明提供了一种机器人的行走预测和控制方法,可以快速准确地预测机器人前方的路径情况,并能控制机器人针对不同的路况采取不同的行走方式,以提高机器人的行走效率。本发明的具体技术方案如下:
一种机器人的行走预测和控制方法,包括如下步骤:基于标示过状态的栅格单元构建栅格地图;基于所述栅格地图,以机器人的当前位置为基准点建立一个动态检测模型;基于所述动态检测模型,预测机器人前方的路径情况;基于预测结果,控制机器人的行走方式。
本发明的有益效果在于:基于标示有各种状态的栅格单元所构成的栅格地图,可以让机器人在行走过程中通过动态检测模型快速准确地预测前方的路径情况,从而根据不同的路径情况合理地控制机器人的行走方式,避免机器人冲出悬崖或者重复进入危险区域等问题,提高了机器人的行走效率。
附图说明
图1为本发明所述机器人的行走预测和控制方法的流程图。
图2为本发明所述的动态检测模型的示意图。
图3为本发明所述局部坐标转换成全局坐标的示意图。
具体实施方式
下面结合附图对本发明的具体实施方式作进一步说明:
本发明所述的机器人是智能家用电器的一种,能凭借一定的人工智能,自动在某些场合自动进行行走。机器人的机体上设有各种传感器,可检测行走距离、行走角度、机身状态和障碍物等,如碰到墙壁或其他障碍物,会自行转弯,并依不同的设定,而走不同的路线,有规划地行走。本发明所述的移动机器人包括如下结构:带有驱动轮的能够自主行走的机器人机体,机体上设有人机交互界面,机体上设有障碍检测单元。机体内部设置有惯性传感器,所述惯性传感器包括加速度计和陀螺仪等,驱动轮上设有用于检测驱动轮的行走距离的里程计(一般是码盘),还设有能够处理相关传感器的参数,并能够输出控制信号到执行部件的控制模块。
如图1所示,本发明所述的机器人的行走预测和控制方法,包括如下步骤:基于标示过状态的栅格单元构建栅格地图;基于所述栅格地图,以机器人的当前位置为基准点建立一个动态检测模型;基于所述动态检测模型,预测机器人前方的路径情况;基于预测结果,控制机器人的行走方式。所述标示过状态的栅格单元是指将机器人正常行走过的栅格单元标示为已行走单元,将机器人检测到障碍物的栅格单元标示为障碍单元,将机器人检测到卡住或者打滑的栅格单元标示为危险单元,将机器人检测到悬崖的栅格单元标示为悬崖单元,将机器人未行走过的栅格单元标示为未知单元。其中,所述机器人正常行走过的栅格单元指的是机器人走过该栅格单元不会出现打滑、卡住、碰到障碍物或者冲出悬崖等异常情况,而这些异常情况会使机器人的行走产生误差,甚至导致机器人受损。机器人正常行走时会按预定的速度和方向平稳有序地行走。本发明所述的方法,基于标示有各种状态的栅格单元所构成的栅格地图(该栅格地图是机器人先前已经遍历所有区域后所构建并保存的地图),可以让机器人在行走过程中通过动态检测模型快速准确地预测前方的路径情况,从而根据不同的路径情况合理地控制机器人的行走方式,避免机器人冲出悬崖或者重复进入危险区域等问题,提高了机器人的行走效率。
优选的,所述以机器人的当前位置为基准点建立一个动态检测模型,包括如下步骤:以所述基准点为圆心,以第一长度为半径构建第一圆弧线,以第二长度为半径构建第二圆弧线,其中,所述第一长度小于所述第二长度;确定所述第一圆弧线和所述第二圆弧线之间的区域为第一预测区;确定所述第二圆弧线之 外的区域为第二预测区;以机器人行走的当前方向为参考方向,确定所述第一预测区中的第一预测点、第二预测点、第三预测点、第四预测点和第五预测点;所述第一预测点位于所述参考方向上,且所述第一预测点与圆心的连线和所述参考方向构成第一角度;所述第二预测点和所述第四预测点位于所述第一预测点的一侧,且所述第二预测点与圆心的连线和所述第一预测点与圆心的连线构成第二角度,所述第四预测点与圆心的连线和所述第一预测点与圆心的连线构成第四角度;所述第三预测点和所述第五预测点位于所述第一预测点的另一侧,且所述第三预测点与圆心的连线和所述第一预测点与圆心的连线构成第三角度,所述第五预测点与圆心的连线和所述第一预测点与圆心的连线构成第五角度;以机器人行走的当前方向为参考方向,确定所述第二预测区中的第六预测点、第七预测点和第八预测点;所述第六预测点位于所述参考方向上,且所述第六预测点与圆心的连线和所述参考方向构成第六角度;所述第七预测点位于所述第六预测点的一侧,且所述第七预测点与圆心的连线和所述第六预测点与圆心的连线构成第七角度;所述第八预测点位于所述第六预测点的另一侧,且所述第八预测点与圆心的连线和所述第六预测点与圆心的连线构成第八角度。其中,所述第一长度和所述第二长度可以根据实际情况进行相应设置,优选的,所述第一长度可以设置为机器人机身半径的1.1倍至1.5倍;所述第二长度可以设置为机器人机身半径的1.6倍至2.1倍。如图2所示,第一预测点为M点,第二预测点为L点,第三预测点为N点,第四预测点为K点,第五预测点为O点,第六预测点为H点,第七预测点为G点,第八预测点为I点。P点为所述基准点(即圆心),P点外的圆圈表示机器人的机身,圆圈外的内径较小的半圆弧为所述第一圆弧线,圆圈外的内径较大的半圆弧为所述第二圆弧线。由于PM方向为机器人的当前方向(即参考方向),而M点又在参考方向上,所以,∠MPM为第一角度,且角度值为0°。由于H点也在参考方向上,所以,第六角度∠HPM的角度值也为0°。此外,∠LPM为第二角度,∠NPM为第三角度,∠KPM为第四角度,∠OPM为第五角度,∠GPM为第七角度,∠IPM为第八角度,这些角的角度值可以根据实际情况进行相应设置。优选的,所述第二角度和所述第三角度的角度值范围在20°至40°之间,所述第四角度和所述第五角度的角度值范围在45°至65°之间,所述第七角度和所述第八角度的角度值范围在15°至40°之间。
优选的,所述第一长度为机器人机身半径的1.2倍,所述第二长度为机器人机身半径的2倍。所述第一长度和所述第二长度的值要适合,如果设置得太短,则预测到危险区域或者障碍物等异常情况时,机器人来不及躲避或者调整行走状态,就起不到预测所带来的有利效果;如果设置得太长,又要耗费相对较多的运算资源,导致预测所带来的效率不够高。
优选的,如图2所示,所述第一预测点位于所述第一圆弧线与所述参考方向的交点(即图中的M点)上;所述第二预测点位于第一垂直线和第一平行线的交点上;所述第三预测点位于第一垂直线和第二平行线的交点上;所述第四预测点位于第三平行线中的第一线段内;所述第五预测点位于第四平行线中的第二线段内。
其中,所述第一垂直线为穿过所述第一预测点且垂直于所述参考方向的垂直线(即图中的线ac);所述第一平行线为位于所述参考方向一侧,且平行于所述参考方向,与所述参考方向的垂直距离为机器人机身半径的平行线(即图中的线gk);所述第二平行线为位于所述参考方向另一侧,且平行于所述参考方向,与所述参考方向的垂直距离为机器人机身半径的平行线(即图中的线hm);所述第三平行线为位于所述参考方向一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第一长度的平行线(即图中的线ab);所述第四平行线为位于所述参考方向另一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第一长度的平行线(即图中的线cd);所述第一线段为所述第一垂直线与所述第三平行线的交点到第二垂直线与所述第三平行线的交点之间的线段(即图中的ab线段);所述第二线段为所述第一垂直线与所述第三平行线的交点到第二垂直线与所述第三平行线的交点之间的线段(即图中的cd线段);所述第二垂直线为位于所述第一预测点与圆心之间,并垂直于所述参考方向,且距圆心的最短距离为机器人机身半径的一半的垂直线(即图中的线bd)。图中,e点与P点的距离为机器人机身半径的一半。K点可以根据实际情况设置为ab线段中的任意一点;O点可以根据实际情况设置为cd线段中的任意一点。
优选的,所述第六预测点位于所述第二圆弧线与所述参考方向的交点(即图中的H点)上;所述第七预测点位于第三垂直线中的第三线段内;所述第八预测点位于第三垂直线中的第四线段内。
其中,所述第三垂直线为穿过所述第六预测点且垂直于所述参考方向的垂直线(即图中的线fi);所述第三线段为所述第一平行线与所述第三垂直线的交点到第五平行线与所述第三垂直线的交点之间的线 段(即图中的fg线段);所述第四线段为所述第二平行线与所述第三垂直线的交点到第六平行线与所述第三垂直线的交点之间的线段(即图中的hi线段);所述第五平行线为位于所述参考方向一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第二长度减去第一长度的差值的一半与第一长度之和的平行线(即图中的线fj);所述第六平行线为位于所述参考方向另一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第二长度减去第一长度的差值的一半与第一长度之和的平行线(即图中的线in)。
以上实施例所述的线路,均是虚拟的线路,是为了便于说明清楚动态检测模型的架构而引用的,在机器人的实际运行中,是不存在上述线路的,但机器人却是在该模型架构的范围内进行的预测。
优选的,所述基于所述动态检测模型,预测机器人前方的路径情况,包括如下步骤:以机器人的当前位置为局部坐标原点,当前方向为局部Y轴方向,建立XY轴局部坐标系;将所述第一预测点至所述第八预测点所在的XY轴局部坐标系中的局部坐标转换成对应于XY轴全局坐标系中的全局坐标;将所述全局坐标转换成栅格坐标;基于所述栅格坐标和栅格地图,确定所述第一预测点至所述第八预测点所对应的栅格单元的状态。只有将动态检测模型的预测点的局部坐标转换成全局坐标,才能对应到栅格地图中的栅格坐标,如此,机器人才可以在实际行走过程中,准确地预测行走方向前方的栅格单元的状态。
优选的,所述将所述第一预测点至所述第八预测点所在的XY轴局部坐标系中的局部坐标转换成对应于XY轴全局坐标系中的全局坐标,包括如下步骤:确定机器人的当前位置在所述XY轴全局坐标系中的全局坐标为(x,y);确定机器人的当前方向与所述XY轴全局坐标系中的Y轴的夹角为θ;确定所述第一预测点在所述XY轴局部坐标系中的局部坐标为(x1,y1);确定所述第一预测点投影至XY轴全局坐标系中的X轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的X轴的投影点之间的距离为xr1=(x1*cosθ-y1*sinθ),且确定所述第一预测点投影至XY轴全局坐标系中的Y轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的Y轴的投影点之间的距离为yr1=(x1*sinθ+y1*cosθ);确定所述第一预测点的全局坐标为(xw1=x+xr1,yw1=y+yr1);确定所述第二预测点在所述XY轴局部坐标系中的局部坐标为(x2,y2);确定所述第二预测点投影至XY轴全局坐标系中的X轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的X轴的投影点之间的距离为xr2=(x2*cosθ-y2*sinθ),且确定所述第二预测点投影至XY轴全局坐标系中的Y轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的Y轴的投影点之间的距离为yr2=(x2*sinθ+y2*cosθ);确定所述第二预测点的全局坐标为(xw2=x+xr2,yw2=y+yr2);以此类推,至完成所述第八预测点的全局坐标的确定。
如图3所示,XQY坐标系为XY轴全局坐标系,X'PY'坐标系为XY轴局部坐标系。圆圈表示机器人,P点为机器人当前所在的位置,也作为局部坐标系的原点,通过陀螺仪和里程计等传感器的检测数据,P点在全局坐标系中的坐标位置是已知的,假设为(x,y)。PM方向为机器人的当前方向,也可以通过陀螺仪的检测数据得出,假设当前方向的角度为θ(即PM方向与Y轴方向的夹角)。N点为动态检测模型的第三预测点,以该点为例进行说明局部坐标转换成对应的全局坐标的方法。N点的局部坐标为(x3,y3),PM作为当前方向与Y轴的夹角的角度为θ,N点投影至X轴方向时,与X轴方向的交点距离P点的最短距离为xr3,且xr3=(x3*cosθ-y3*sinθ),与Y轴方向的交点距离P点的最短距离为yr3,且yr3=(x3*sinθ+y3*cosθ)。则,从图中可以看出,N点的全局坐标为(xw3,yw3),xw3=x+xr3=x+x3*cosθ-y3*sinθ,yw3=y+yr3=y+x3*sinθ+y3*cosθ。由于动态检测模型中,各预测点之间的相对位置是固定不变的,所以,随着机器人当前方向的变化,各预测点的转动角度是相同的,所以,各预测点转换成全局坐标的方式是一样的,其它预测点与上述N点的转换方式相同,在此不再赘述。
优选的,所述将所述全局坐标转换成栅格坐标,包括如下步骤:确定所述栅格单元是边长为L的正方形;确定所述第一预测点的栅格坐标为(xg1=xw1/L,yg1=yw1/L),且xw1/L和yw1/L的结果取整数;确定所述第二预测点的栅格坐标为(xg2=xw2/L,yg2=yw2/L),且xw2/L和yw2/L的结果取整数;以此类推,至确定完所述第八预测点的栅格坐标。由于通过全局坐标并不能确定前方的路径情况,栅格地图中的路径情况都是栅格单元的状态进行标示,所以,需要把全局坐标转换成对应的栅格单元,再通过判断相应栅格单元的状态来最终确定前方路径情况,以此达到预测效果。通过上述转换方式,可以准确的确定动态检测模型中各预测点对应哪个栅格单元,运算简单,准确性高。
优选的,所述基于预测结果,控制机器人的行走方式,包括如下步骤:判断所述第一预测点、所述第二预测点或者所述第三预测点所对应的栅格单元是否为危险单元;如果是,则控制机器人按第一行走方式 行走;如果否,则判断所述第一预测点、所述第二预测点或者所述第三预测点所对应的栅格单元是否为悬崖单元或者障碍单元;如果是,则控制机器人按第二行走方式行走;如果否,则判断所述第四预测点或者所述第五预测点所对应的栅格单元是否为悬崖单元或者障碍单元;如果是,则控制机器人按第三行走方式行走;如果否,则判断所述第六预测点、所述第七预测点或者所述第八预测点是否为悬崖单元或者障碍单元;如果是,则控制机器人按第四行走方式行走;如果否,则保持机器人当前的行走方式。其中,所述危险单元为机器人检测到卡住或者打滑的栅格单元,所述障碍单元为机器人检测到障碍物的栅格单元,所述悬崖单元为机器人检测到悬崖的栅格单元。通过由近及远的方式,逐步判断各预测点对应的栅格单元的状态,来确定前方路径情况,从而控制机器人按不同的行走方式行走,以此提高机器人的行走效率,避免进入危险区所导致的卡住或者打滑,强力撞击障碍物或者冲出悬崖等问题。
优选的,所述控制机器人按第一行走方式行走,包括如下步骤:控制机器人停止行走;判断机器人是否处于弓字型行走阶段;如果是,则控制机器人掉头;如果否,则确定机器人处于沿边行走阶段,并控制机器人绕过所述危险单元后继续沿边。和/或,所述控制机器人按第二行走方式行走,包括如下步骤:如果判断所述栅格单元为障碍单元,则控制机器人按第一比例降低行走速度;如果判断所述栅格单元为悬崖单元,则控制机器人按第二比例降低行走速度。和/或,所述控制机器人按第三行走方式行走,包括如下步骤:控制机器人按第三比例降低行走速度。和/或,所述控制机器人按第四行走方式行走,包括如下步骤:控制机器人按第四比例降低行走速度。其中,所述比例可以根据实际情况进行设置,优选的,所述第一比例为0.5;所述第二比例为0.3;所述第三比例为0.7;第四比例为0.8。
当确定第一预测区域内离机器人最近的三个预测点中,有危险单元,则机器人需要避开对应的危险区域。而避开危险区域又需要根据机器人的行走阶段采取不同的避让方式。按扫地机器人的行走方式,主要分为弓字型行走阶段和沿边行走阶段。当机器人处于弓字型行走阶段时,碰到危险区域可以直接掉头,继续弓字型行走清扫其它路径;当机器人处于沿边行走阶段时,不能掉头,因为掉头后会重复清扫已沿过边的路径,且如果再次检测到危险区域,又再次掉头,就会形成两个危险区之间反复沿边的情况,所以,在沿边过程中碰到危险区域,需要采用绕过危险区域的方式,在绕过危险区域后继续沿边,如此才能最终完成全部沿边阶段的清扫。当确定第一预测区域内离机器人最近的三个预测点中,没有危险单元,如果有障碍单元,则将机器人的行走速度降低至0.5倍,避免机器人速度过快而强力撞击障碍物。如果有悬崖单元,则将机器人的速度降低至0.3倍,使机器人的速度更快降低下来,避免冲出悬崖的风险。
如果离机器人最近的三个预测点所对应的栅格单元为正常行走过的已行走单元,则判断所述第四预测点或者所述第五预测点所对应的栅格单元是否为悬崖单元或者障碍单元,因为这两个预测点在机器人两外侧,即使是危险单元,也可以从中间通过,所以,不需要判断这两个预测点是否是危险单元。当确定其中一个点是悬崖单元或者障碍单元时,则需要把机器人的速度降低至0.7倍,因为障碍物或者悬崖边缘的垃圾一般比较多,所以需要慢速通过,以提高清扫效果。
如果第四预测点或者所述第五预测点所对应的栅格单元都为正常行走过的已行走单元,则判断所述第六预测点、所述第七预测点或者所述第八预测点是否为悬崖单元或者障碍单元。因为这三个点离机器人的距离相对较远,所以预测的目的是为了提前控制机器人的速度。当检测其中之一是悬崖单元或者障碍单元时,则将机器人的速度降低至0.8倍。如果都不是,则保持机器人当前的行走方式,继续进行清扫。
以上实施例仅为充分公开而非限制本发明,凡基于本发明的创作主旨、未经创造性劳动的等效技术特征的替换,应当视为本申请揭露的范围。

Claims (10)

  1. 一种机器人的行走预测和控制方法,其特征在于,包括如下步骤:
    基于标示过状态的栅格单元构建栅格地图;
    基于所述栅格地图,以机器人的当前位置为基准点建立一个动态检测模型;基于所述动态检测模型,预测机器人前方的路径情况;
    基于预测结果,控制机器人的行走方式。
  2. 根据权利要求1所述的方法,其特征在于:所述以机器人的当前位置为基准点建立一个动态检测模型,包括如下步骤:
    以所述基准点为圆心,以第一长度为半径构建第一圆弧线,以第二长度为半径构建第二圆弧线,其中,所述第一长度小于所述第二长度;
    确定所述第一圆弧线和所述第二圆弧线之间的区域为第一预测区;
    确定所述第二圆弧线之外的区域为第二预测区;
    以机器人行走的当前方向为参考方向,确定所述第一预测区中的第一预测点、第二预测点、第三预测点、第四预测点和第五预测点;所述第一预测点位于所述参考方向上,且所述第一预测点与圆心的连线和所述参考方向构成第一角度;所述第二预测点和所述第四预测点位于所述第一预测点的一侧,且所述第二预测点与圆心的连线和所述第一预测点与圆心的连线构成第二角度,所述第四预测点与圆心的连线和所述第一预测点与圆心的连线构成第四角度;所述第三预测点和所述第五预测点位于所述第一预测点的另一侧,且所述第三预测点与圆心的连线和所述第一预测点与圆心的连线构成第三角度,所述第五预测点与圆心的连线和所述第一预测点与圆心的连线构成第五角度;
    以机器人行走的当前方向为参考方向,确定所述第二预测区中的第六预测点、第七预测点和第八预测点;所述第六预测点位于所述参考方向上,且所述第六预测点与圆心的连线和所述参考方向构成第六角度;所述第七预测点位于所述第六预测点的一侧,且所述第七预测点与圆心的连线和所述第六预测点与圆心的连线构成第七角度;所述第八预测点位于所述第六预测点的另一侧,且所述第八预测点与圆心的连线和所述第六预测点与圆心的连线构成第八角度。
  3. 根据权利要求2所述的方法,其特征在于:所述第一长度为机器人机身半径的1.2倍,所述第二长度为机器人机身半径的2倍。
  4. 根据权利要求2所述的方法,其特征在于:
    所述第一预测点位于所述第一圆弧线与所述参考方向的交点上;所述第二预测点位于第一垂直线和第一平行线的交点上;所述第三预测点位于第一垂直线和第二平行线的交点上;所述第四预测点位于第三平行线中的第一线段内;所述第五预测点位于第四平行线中的第二线段内;
    其中,所述第一垂直线为穿过所述第一预测点且垂直于所述参考方向的垂直线;所述第一平行线为位于所述参考方向一侧,且平行于所述参考方向,与所述参考方向的垂直距离为机器人机身半径的平行线;所述第二平行线为位于所述参考方向另一侧,且平行于所述参考方向,与所述参考方向的垂直距离为机器人机身半径的平行线;所述第三平行线为位于所述参考方向一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第一长度的平行线;所述第四平行线为位于所述参考方向另一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第一长度的平行线;所述第一线段为所述第一垂直线与所述第三平行线的交点到第二垂直线与所述第三平行线的交点之间的线段;所述第二线段为所述第一垂直线与所述第三平行线的交点到第二垂直线与所述第三平行线的交点之间的线段;所述第二垂直线为位于所述第一预测点与圆心之间,并垂直于所述参考方向,且距圆心的最短距离为机器人机身半径的一半的垂直线。
  5. 根据权利要求4所述的方法,其特征在于:
    所述第六预测点位于所述第二圆弧线与所述参考方向的交点上;所述第七预测点位于第三垂直线中的第三线段内;所述第八预测点位于第三垂直线中的第四线段内;
    其中,所述第三垂直线为穿过所述第六预测点且垂直于所述参考方向的垂直线;所述第三线段为所述第一平行线与所述第三垂直线的交点到第五平行线与所述第三垂直线的交点之间的线段;所述第四线段为所述第二平行线与所述第三垂直线的交点到第六平行线与所述第三垂直线的交点之间的线段;所述第五平行线为位于所述参考方向一侧,且平行于所述参考方向,与所述参考方向的垂直距离为第二长度减去第一长度的差值的一半与第一长度之和的平行线;所述第六平行线为位于所述参考方向另一侧,且平行于所述 参考方向,与所述参考方向的垂直距离为第二长度减去第一长度的差值的一半与第一长度之和的平行线。
  6. 根据权利要求2所述的方法,其特征在于:所述基于所述动态检测模型,预测机器人前方的路径情况,包括如下步骤:
    以机器人的当前位置为局部坐标原点,当前方向为局部Y轴方向,建立XY轴局部坐标系;
    将所述第一预测点至所述第八预测点所在的XY轴局部坐标系中的局部坐标转换成对应于XY轴全局坐标系中的全局坐标;
    将所述全局坐标转换成栅格坐标;
    基于所述栅格坐标和栅格地图,确定所述第一预测点至所述第八预测点所对应的栅格单元的状态。
  7. 根据权利要求6所述的方法,其特征在于:所述将所述第一预测点至所述第八预测点所在的XY轴局部坐标系中的局部坐标转换成对应于XY轴全局坐标系中的全局坐标,包括如下步骤:
    确定机器人的当前位置在所述XY轴全局坐标系中的全局坐标为(x,y);
    确定机器人的当前方向与所述XY轴全局坐标系中的Y轴的夹角为θ;
    确定所述第一预测点在所述XY轴局部坐标系中的局部坐标为(x1,y1);
    确定所述第一预测点投影至XY轴全局坐标系中的X轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的X轴的投影点之间的距离为xr1=(x1*cosθ-y1*sinθ),且确定所述第一预测点投影至XY轴全局坐标系中的Y轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的Y轴的投影点之间的距离为yr1=(x1*sinθ+y1*cosθ);
    确定所述第一预测点的全局坐标为(xw1=x+xr1,yw1=y+yr1);
    确定所述第二预测点在所述XY轴局部坐标系中的局部坐标为(x2,y2);
    确定所述第二预测点投影至XY轴全局坐标系中的X轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的X轴的投影点之间的距离为xr2=(x2*cosθ-y2*sinθ),且确定所述第二预测点投影至XY轴全局坐标系中的Y轴的投影点,相对于机器人的当前位置投影至XY轴全局坐标系中的Y轴的投影点之间的距离为yr2=(x2*sinθ+y2*cosθ);
    确定所述第二预测点的全局坐标为(xw2=x+xr2,yw2=y+yr2);
    以此类推,至完成所述第八预测点的全局坐标的确定。
  8. 根据权利要求7所述的方法,其特征在于:所述将所述全局坐标转换成栅格坐标,包括如下步骤:
    确定所述栅格单元是边长为L的正方形;
    确定所述第一预测点的栅格坐标为(xg1=xw1/L,yg1=yw1/L),且xw1/L和yw1/L的结果取整数;
    确定所述第二预测点的栅格坐标为(xg2=xw2/L,yg2=yw2/L),且xw2/L和yw2/L的结果取整数;
    以此类推,至确定完所述第八预测点的栅格坐标。
  9. 根据权利要求6所述的方法,其特征在于:所述基于预测结果,控制机器人的行走方式,包括如下步骤:
    判断所述第一预测点、所述第二预测点或者所述第三预测点所对应的栅格单元是否为危险单元;
    如果是,则控制机器人按第一行走方式行走;
    如果否,则判断所述第一预测点、所述第二预测点或者所述第三预测点所对应的栅格单元是否为悬崖单元或者障碍单元;
    如果是,则控制机器人按第二行走方式行走;
    如果否,则判断所述第四预测点或者所述第五预测点所对应的栅格单元是否为悬崖单元或者障碍单元;
    如果是,则控制机器人按第三行走方式行走;
    如果否,则判断所述第六预测点、所述第七预测点或者所述第八预测点是否为悬崖单元或者障碍单元;
    如果是,则控制机器人按第四行走方式行走;
    如果否,则保持机器人当前的行走方式;
    其中,所述危险单元为机器人检测到卡住或者打滑的栅格单元,所述障碍单元为机器人检测到障碍物的栅格单元,所述悬崖单元为机器人检测到悬崖的栅格单元。
  10. 根据权利要求9所述的方法,其特征在于:
    所述控制机器人按第一行走方式行走,包括如下步骤:控制机器人停止行走;判断机器人是否处于弓字型行走阶段;如果是,则控制机器人掉头;如果否,则确定机器人处于沿边行走阶段,并控制机器人绕过所述危险单元后继续沿边;
    和/或,
    所述控制机器人按第二行走方式行走,包括如下步骤:如果判断所述栅格单元为障碍单元,则控制机器人按第一比例降低行走速度;如果判断所述栅格单元为悬崖单元,则控制机器人按第二比例降低行走速度;
    和/或,
    所述控制机器人按第三行走方式行走,包括如下步骤:控制机器人按第三比例降低行走速度;
    和/或,
    所述控制机器人按第四行走方式行走,包括如下步骤:控制机器人按第四比例降低行走速度。
PCT/CN2018/098912 2017-10-30 2018-08-06 机器人的行走预测及控制方法 Ceased WO2019085567A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
JP2020523718A JP7075994B2 (ja) 2017-10-30 2018-08-06 ロボットの走行予測及び制御方法
KR1020207012135A KR102445731B1 (ko) 2017-10-30 2018-08-06 로봇의 주행 예측 및 제어 방법
EP18871985.0A EP3705968B1 (en) 2017-10-30 2018-08-06 Robot path prediction and control method
US16/649,145 US11526170B2 (en) 2017-10-30 2018-08-06 Method for detecting skidding of robot, mapping method and chip

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711036010.0 2017-10-30
CN201711036010.0A CN107807643B (zh) 2017-10-30 2017-10-30 机器人的行走预测和控制方法

Publications (1)

Publication Number Publication Date
WO2019085567A1 true WO2019085567A1 (zh) 2019-05-09

Family

ID=61591636

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/098912 Ceased WO2019085567A1 (zh) 2017-10-30 2018-08-06 机器人的行走预测及控制方法

Country Status (6)

Country Link
US (1) US11526170B2 (zh)
EP (1) EP3705968B1 (zh)
JP (1) JP7075994B2 (zh)
KR (1) KR102445731B1 (zh)
CN (1) CN107807643B (zh)
WO (1) WO2019085567A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111107505A (zh) * 2019-12-10 2020-05-05 北京云迹科技有限公司 位置预估方法、空间变换判断方法、装置、设备及介质

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107807643B (zh) * 2017-10-30 2019-09-03 珠海市一微半导体有限公司 机器人的行走预测和控制方法
CN108628312B (zh) * 2018-05-14 2021-11-19 珠海一微半导体股份有限公司 机器人被卡的检测方法和脱卡的控制方法及芯片
CN109514581B (zh) * 2018-12-20 2021-03-23 珠海市一微半导体有限公司 一种基于智能移动机器人的安全提醒方法
CN109597385B (zh) * 2018-12-26 2021-08-20 芜湖哈特机器人产业技术研究院有限公司 一种栅格地图及基于栅格地图的多agv动态路径规划方法
CN111368760B (zh) * 2020-03-09 2023-09-01 阿波罗智能技术(北京)有限公司 一种障碍物检测方法、装置、电子设备及存储介质
CN111329399B (zh) * 2020-04-09 2021-09-10 湖南格兰博智能科技有限责任公司 一种基于有限状态机的扫地机目标点导航方法
CN112859862B (zh) * 2021-01-15 2024-09-24 珠海一微半导体股份有限公司 一种利用充电桩进行地图修正的方法及系统
CN114952853B (zh) * 2022-06-15 2025-03-11 仁洁智能科技有限公司 机器人定位方法、系统、光伏电站及终端设备
CN115890676B (zh) * 2022-11-28 2024-08-13 深圳优地科技有限公司 机器人控制方法、机器人及存储介质
JPWO2024195180A1 (zh) * 2023-03-23 2024-09-26

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103679301A (zh) * 2013-12-31 2014-03-26 西安理工大学 基于复杂地形的最优路径寻找方法
CN105182979A (zh) * 2015-09-23 2015-12-23 上海物景智能科技有限公司 一种移动机器人障碍物检测及避让方法和系统
CN105509729A (zh) * 2015-11-16 2016-04-20 中国航天时代电子公司 一种基于仿生触角的机器人自主导航方法
WO2016067640A1 (ja) * 2014-10-28 2016-05-06 シャープ株式会社 自律移動装置
CN107807643A (zh) * 2017-10-30 2018-03-16 珠海市微半导体有限公司 机器人的行走预测和控制方法

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5400244A (en) * 1991-06-25 1995-03-21 Kabushiki Kaisha Toshiba Running control system for mobile robot provided with multiple sensor information integration system
US7663333B2 (en) * 2001-06-12 2010-02-16 Irobot Corporation Method and system for multi-mode coverage for an autonomous robot
JP4256812B2 (ja) 2004-04-26 2009-04-22 三菱重工業株式会社 移動体の障害物回避方法及び該移動体
JP2007148591A (ja) 2005-11-24 2007-06-14 Funai Electric Co Ltd 自走式掃除機
JP4745149B2 (ja) 2006-06-30 2011-08-10 セコム株式会社 移動ロボット
US7211980B1 (en) * 2006-07-05 2007-05-01 Battelle Energy Alliance, Llc Robotic follow system and method
KR100901013B1 (ko) * 2007-04-17 2009-06-04 한국전자통신연구원 경로 탐색 시스템 및 그 방법
JP2011128899A (ja) 2009-12-17 2011-06-30 Murata Machinery Ltd 自律移動装置
US8930019B2 (en) 2010-12-30 2015-01-06 Irobot Corporation Mobile human interface robot
US8798840B2 (en) * 2011-09-30 2014-08-05 Irobot Corporation Adaptive mapping with spatial summaries of sensor data
CN103914068A (zh) * 2013-01-07 2014-07-09 中国人民解放军第二炮兵工程大学 一种基于栅格地图的服务机器人自主导航方法
CN105955262A (zh) * 2016-05-09 2016-09-21 哈尔滨理工大学 一种基于栅格地图的移动机器人实时分层路径规划方法
EP3611589B1 (en) * 2017-04-11 2021-11-17 Amicro Semiconductor Co., Ltd. Method for controlling motion of robot based on map prediction

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103679301A (zh) * 2013-12-31 2014-03-26 西安理工大学 基于复杂地形的最优路径寻找方法
WO2016067640A1 (ja) * 2014-10-28 2016-05-06 シャープ株式会社 自律移動装置
CN105182979A (zh) * 2015-09-23 2015-12-23 上海物景智能科技有限公司 一种移动机器人障碍物检测及避让方法和系统
CN105509729A (zh) * 2015-11-16 2016-04-20 中国航天时代电子公司 一种基于仿生触角的机器人自主导航方法
CN107807643A (zh) * 2017-10-30 2018-03-16 珠海市微半导体有限公司 机器人的行走预测和控制方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3705968A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111107505A (zh) * 2019-12-10 2020-05-05 北京云迹科技有限公司 位置预估方法、空间变换判断方法、装置、设备及介质
CN111107505B (zh) * 2019-12-10 2021-09-24 北京云迹科技有限公司 位置预估方法、空间变换判断方法、装置、设备及介质

Also Published As

Publication number Publication date
KR102445731B1 (ko) 2022-09-21
KR20200053612A (ko) 2020-05-18
EP3705968B1 (en) 2023-06-07
EP3705968C0 (en) 2023-06-07
EP3705968A4 (en) 2020-12-09
US20200293048A1 (en) 2020-09-17
JP2021500688A (ja) 2021-01-07
JP7075994B2 (ja) 2022-05-26
US11526170B2 (en) 2022-12-13
CN107807643B (zh) 2019-09-03
CN107807643A (zh) 2018-03-16
EP3705968A1 (en) 2020-09-09

Similar Documents

Publication Publication Date Title
WO2019085567A1 (zh) 机器人的行走预测及控制方法
US11832774B2 (en) Method for detecting skidding of robot, mapping method and chip
CN103335658B (zh) 一种基于弧线路径生成的自主车辆避障方法
CN106598054B (zh) 机器人路径调整方法及装置
CN105043376B (zh) 一种适用于非全向移动车辆的智能导航方法及系统
JP2022532920A (ja) レーダーデータからのヨーレート
WO2021008611A1 (zh) 机器人被困检测及脱困方法
CN108274467B (zh) 机器人走直线的控制方法和芯片及机器人
CN102323819A (zh) 一种基于协调控制的智能轮椅室外导航方法
CN114932562B (zh) 一种基于激光雷达的地下电缆隧道巡检机器人及实现方法
CN108544490B (zh) 无人驾驶智能机器人道路避障方法
US20170168492A1 (en) Autonomous travel vehicle and reproduction travel method
Kato et al. Autonomous robot navigation system without grid maps based on double deep Q-network and RTK-GNSS localization in outdoor environments
Csaba et al. Mobil robot navigation using 2D LIDAR
CN118435141A (zh) 机器人及其避障方法,机器人系统及存储介质
CN114003036A (zh) 机器人避障控制方法、装置、设备和介质
Takahashi et al. Development of prediction based emergency obstacle avoidance module by using LIDAR for mobile robot
CN119744375B (zh) 机器人及机器人行进方法、机器人系统及计算机存储介质
CN112882479A (zh) 一种基于内外角累积状态的路径规划方法
CN113448325A (zh) 一种扫地机器人控制方法及扫地机器人
CN121028842A (zh) 路径规划方法、系统及清洁机器人
EP4524678A1 (en) Robot local planner selection
CN118533165A (zh) 一种基于狭窄空间感知的移动巡检机器人自主导航方法
CN107498559B (zh) 基于视觉的机器人转向的检测方法和芯片
Ariffin et al. Real-time obstacle avoidance for humanoid-controlled mobile platform navigation

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18871985

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 20207012135

Country of ref document: KR

Kind code of ref document: A

Ref document number: 2020523718

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2018871985

Country of ref document: EP

Effective date: 20200602