JPH07200030A - Robot motion planning device, controller, and optimal trajectory generation method - Google Patents

Abstract

(57) [Summary] [Purpose] In a robot motion planning device, obtain the optimum trajectory of a robot arm at high speed. [Structure] Input / output unit 1, robot command generation unit 2 for generating a robot command according to an input from the input / output unit 1, and robot command interpretation for generating a start point, an end point and a movable range of a trajectory of a robot arm from the robot command. Means 3, physical parameter storage means 9 for storing physical parameters of the robot arm, offline table data generation means 7 for offline calculation of data calculated by the trajectory generation means 4, and table for storing the table data. The data storage means 8 and the trajectory generation means 6 for obtaining the optimum trajectory by referring to the table data in the table data storage means 8. [Effect] In the robot motion planning device, the optimum trajectory of the robot arm can be obtained at high speed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a robot motion planning device for planning the motion of a robot arm, a robot controller connected to the robot motion planning device for controlling a robot, and a method for generating an optimum trajectory of the robot arm. .

[0002]

2. Description of the Related Art As a method for improving the efficiency of robot work, there is optimization of the trajectory of a robot arm. As a first conventional example, a method of planning the shortest time trajectory of a robot arm in consideration of the dynamics of the robot arm is described in 1987, Institute of Instrument and Control Vol. 23, No. 11 "Planning of manipulator approximation shortest time trajectory using B-spline" (by Hiroaki Ozaki et al.). This is a method for efficiently searching for an optimal trajectory by the local controllability of the B-spline.

As a second conventional example, a general method for solving the optimization problem of the trajectory of the robot arm is "Adapti".
ve Control Processes: a guided Tour "(1961 Pri
Richard Bellman, published by nc−eton University Press)
Explained. This is known as dynamic programming.

As a third conventional example, a method for obtaining the trajectory of a robot arm mounted on an artificial satellite is described in 1992, Vol. 28, No. 3 "About the motion trajectory of the hand of the space robot and the posture change of the main body"
(Katsuhiko Yamada et al.). This uses a model that ignores the mass of the arm and simply considers only the mass of the satellite body and the payload, and minimizes the attitude variation of the satellite body and the weighted linear sum of the robot arm tip travel distance. This is a method of obtaining a trajectory.

As a fourth conventional example, there is an example in which the trajectory of a robot arm mounted on an artificial satellite is obtained.
No. 158540. This is because the locus is limited to a linear locus in Cartesian coordinates or joint coordinates, and the velocity pattern is limited so that each acceleration during acceleration and deceleration becomes a quadratic curve. And is a method of obtaining the acceleration time and the deceleration time such that the influence on the satellite body is minimized.

[0006]

If the above-mentioned first or second conventional example is simply adapted to a robot arm, thousands of arithmetic operations are required at the time of trajectory planning,
Since the evaluation function is repeatedly calculated using the equation of motion, even if a computer is used, it takes a huge amount of calculation time to obtain the optimum trajectory.

In the third conventional example, the locus of the robot arm can be obtained at a high speed because it is analytically obtained, but the locus without a payload is obtained because the model is simplified. However, there is a drawback that only a locus that minimizes the evaluation function can be obtained, and it is difficult to adapt to a general evaluation function.
Further, in the fourth conventional example, the locus and the speed pattern are limited, which is not general.

An object of the present invention is to obtain an optimum trajectory of a robot arm at high speed using offline calculation data.

[0009]

The above objects of the present invention are as follows.
As a first means, an input / output means, a robot command generation means connected to the input / output means for generating a work command to the robot, and a robot arm trajectory connected to the robot command generation means to interpret the work command. Robot command interpreting means for generating the start and end points and the movable area, trajectory generating means for generating the trajectory of the robot arm connected to the robot command interpreting means, and physics for storing physical parameters of each link of the robot arm. In a robot motion planning device including parameter storage means, calculation data of an operation cost of the robot arm or an intermediate calculation value of the calculation data which is connected to the physical parameter storage means and is required by the trajectory generation means is offline. Offline table data generation with table data calculated in advance by And a table data storage means connected to the offline table data generation means for storing the table data and delivering the table data to the trajectory generation means. The trajectory generation means uses the table data to determine an optimum trajectory. It is achieved by generating.

Further, as a second means, a communication means for receiving a work command, a robot command interpreting means for interpreting the work command for the robot and generating a start point and an end point of the trajectory of the robot arm, and a movable area, In the robot controller, the trajectory generation means is connected to robot command generation means, and generates trajectory of the robot arm, and physical parameter storage means that stores physical parameters of each link of the robot arm. And the table data storing means for storing the table data in which the calculation data of the operation cost of the robot arm required in advance is stored offline and delivering it to the trajectory generating means. It is achieved by using to generate an optimal trajectory.

Further, as a third means, in the optimum trajectory generating method for generating the trajectory of the robot arm, the optimum trajectory generating method divides the operation range of the robot arm with a mesh table, and the robot arm separates the mesh table from the robot table. Of the radial unit trajectories set for the respective lattice points and the table data necessary for calculating the operation cost of the robot arm tabulated by the lattice points are calculated off-line in advance and stored in the table data storage means. The method and the method of calculating the optimum trajectory of the robot arm by the trajectory generation means using the table data from the start point and the end point of the trajectory of the robot arm and the movable range.

[0012]

According to the first and third means, the off-line table data generating means refers to the physical parameters of each link of the robot arm stored in the physical parameter storage means, and intermediates the data calculated by the trajectory generating means. After the calculated values are calculated off-line in advance, they are tabulated and stored in the table data storage means. The robot command generation means generates a work command according to the input from the input / output means. The robot command interpreting means generates a start point and an end point of the trajectory of the robot arm and an operable area from the work instruction. The trajectory generation means refers to the table data of the table data storage means and the physical parameters of each link of the robot arm stored in the physical parameter storage means,
The optimum trajectory can be obtained at high speed.

Further, according to the second means, since the robot motion planning device and the robot controller are provided with the communication means, only the work command from the robot motion planning device is received, and the robot controller is operated by the robot motion planning device. Since it holds the same data as the table data owned by, the robot controller can obtain the optimum trajectory at high speed. Moreover, once the table data is transmitted to the robot controller, after that, only the work command need be transmitted, so that the communication amount can be reduced.

[0014]

FIG. 1 shows a first embodiment of the present invention. The robot motion planning apparatus 14 of the present embodiment is a robot command which is a work command which describes input / output means 1 provided for the robot to send and receive data to and from a person and a work of the robot connected to the input / output means 1. 16, and a robot command interpreting means 3 which is connected to the robot command generating means 2 and interprets the robot command 16 to generate a start point and an end point of a trajectory of a robot arm (not shown) and an operable area 17. And physical parameter storage means 9 for storing physical parameters such as the size, mass, moment of inertia, and position of the center of gravity of each link of the robot arm, and a trajectory of the robot arm connected to the physical parameter storage means 9 is generated. Offline table data generation that allows you to calculate the table data 20 that is sometimes needed in advance offline Means 7, a table data storage means 8 connected to the offline table data generation means 7 for storing the table data 20, a table data storage means 8, the robot command interpretation means 3, and a physical parameter storage means 9. A trajectory generating means 4 which is connected to the robot arm and calculates an optimum trajectory 18 of the robot arm by a dynamic programming method by referring to the table data 20 and the physical parameters according to the start and end points of the trajectory of the robot arm and the movable area 17; RW command value generation for outputting a RW command value 19 to a reaction wheel (hereinafter referred to as RW) installed in an artificial satellite (not shown) connected to the orbit generation means 4 and the physical parameter storage means 9 by the physical parameters and the optimum orbit 18. Means 5, the RW command value generating means 5, the trajectory generating means 4, and the object A simulator 6 for simulation calculation for confirming the operation prior to being connected to the parameter storing unit 9 actually operating the robot arm, is configured to include a.

The off-line table data generating means 7 pre-calculates the off-line table data 20 necessary for generating the trajectory of the robot arm. The table data 20 is stored in the table data storage means 8, and the table data 20 is delivered to the trajectory generation means 4 at the time of trajectory generation. The table data 20 stored in the table data storage means 8 enables high-speed trajectory generation. When the optimum trajectory 18 of the robot arm is generated, the angular momentum of the robot arm or the integral of the angular momentum along the trajectory is required. Therefore, the offline table data generating means 7
Refers to physical parameters such as the size, mass, and moment of inertia of each link of the robot arm stored in the physical parameter storage unit 9, and after integrating the intermediate calculated value of the angular momentum along the trajectory, the table data It is delivered to the table data storage means 8 as 20.

Here, the reason why the intermediate calculation value is used is that the robot arm has a payload, and when the artificial satellite body is docked with another artificial satellite or the like, the physical parameters change and there is a portion that cannot be calculated. Is. Details of the calculation method will be described later.

The input / output means 1 is a part for exchanging information with a person, and is a device such as a display, a keyboard, a mouse and the like. The robot command generating means 2 generates a robot command 16 which describes the work of the robot. In the robot command interpreting means 3, the robot command 16
To generate the start and end points of the trajectory of the robot arm and the movable area 17.

The trajectory generating means 4 refers to the table data 20 of the intermediate calculation values and the physical parameters of the artificial satellite body and the robot arm according to the start and end points of the trajectory of the robot arm and the movable area 17, and refers to the robot. The optimal trajectory of the robot arm is calculated by the dynamic programming method by calculating the operation cost, which is the optimal path among the paths for moving the arm.
Plan. The operation cost will be described later.

As described above, in the off-line table data generation means 7, the part that can be calculated is calculated in advance and stored as the table data 20 in the table data storage means 8.
The optimum trajectory 18 can be calculated at high speed by referring to. Details of the calculation method will be described later.

In the RW command value generating means 5, RW1 when the robot arm is operated along the optimum trajectory 18
The angular velocity command value 19 of 2 is generated. Here, the angular velocity command value 19 of the RW 12 is generated so as to compensate the angular momentum of the robot arm and keep the attitude of the artificial satellite body constant.

The simulator 6 carries out a simulation calculation for confirming the operation before actually operating the robot arm. The simulator 6 has a dynamic model of an artificial satellite equipped with a robot arm, and calculates the operation of the artificial satellite equipped with the robot arm from the robot arm optimum trajectory 18 and the RW command value 19.

A second embodiment of the present invention is shown in FIG. In this embodiment, a communication means 13 is added to the robot operation planning apparatus 14 of the first embodiment to add a robot operation planning apparatus 2 for artificial satellites.
It is applied to 1.

The artificial-satellite robot operation planning apparatus 21 of the present embodiment comprises a robot operation of the first embodiment, which is connected to the robot command generating means 2 and includes a communication means 13 for transmitting a robot command 16 to the artificial satellite. A planning device 14, an artificial satellite body 10 in which the RW 12 is installed, a robot arm 11 mounted on the artificial satellite body 10,
It is configured to include the robot arm 11 and a robot controller 15 connected to the artificial satellite body 10.

The robot controller 15 has a communication means 13A for communicating with the communication means 13 arranged in the robot motion planning apparatus 14 of the first embodiment, and a robot command connected to the communication means 13A for interpreting the robot command 16A. Interpretation means 3A and the robot command interpretation means 3
The optimum trajectory 1 for the robot arm 11 is connected to A, the table data storage means 8A and the physical parameter storage means 9A.
Trajectory generating means 4A for generating 8A, and the trajectory generating means 4
RW12 connected to A and the physical parameter storage means 9A
And RW command value generating means 5A for outputting the RW command value 19A.

When the robot arm 11 is operated, the artificial satellite body 10 having the robot arm 11 causes a posture change due to the reaction of the robot arm 11. Therefore, the attitude control unit controls the attitude so that the attitude of the artificial satellite body 10 is kept constant. The artificial satellite body 10 of the present embodiment is provided with the RW 12 as attitude control means. The robot operation planning apparatus 14 is additionally provided with a communication unit 13 that communicates with the robot controller 15. The robot controller 15 is mounted on the artificial satellite, and according to the robot command 16A sent from the robot operation planning device 14, the robot arm 11 and
Control the reaction wheel 12.

The robot controller 15 stores the same data as that of the robot motion planning apparatus 14, and is provided with a robot command interpreting means 3A, a trajectory generating means 4A, a physical parameter storing means 9A, and a table data storing means 8A which perform the same operation. Therefore, the robot motion planning device 1
By sending the robot command 16 from 4, the robot arm 11 and the RW 12 can be operated.

A concrete method of generating an optimum trajectory according to the third embodiment of the present invention will be described. As described above, in the case of the artificial satellite body 10 having the robot arm 11, the attitude of the artificial satellite body 10 changes due to the operation of the robot arm 11. The attitude variation of the artificial satellite body 10 can be calculated as in Equation 1 using the law of conservation of angular momentum.

[0028]

[Equation 1]

When the robot arm 11 is moved while keeping the attitude of the artificial satellite body 10 constant, the RW 12 needs to be controlled so as to output an angular momentum having the same sign as that of the robot arm 11 but the same sign. However, RW
Since the angular velocity that can be output by 12 is limited, the robot arm 11 must be operated with an angular momentum within the range that can be output by the RW 12. Since the angular momentum of the robot arm 11 changes depending on the orbit to be operated, when the robot arm 11 is operated while controlling the attitude of the artificial satellite body 10, a linear orbit that connects the start point and the end point of the orbit with a straight line is the operating time. It will not be the shortest orbit.

Here, the attitude control is possible, and
As an example, a method for obtaining the optimum trajectory having the shortest operation time will be described.

[0031]

[Equation 2]

The evaluation function J is such that when the angular momentum of the robot arm 11 can be compensated by the RW 12, the robot arm 11 is operated at the maximum speed, and when the robot arm 11 is operated at the maximum speed, the angular momentum is calculated by the RW 12. When the compensation is not possible, the operation time of the robot arm 11 when the robot arm 11 is operated at a speed that can be compensated by the RW 12 is shown. Further, the following may be performed by multiplying by a coefficient of 1 or less.

[0033]

[Equation 3]

Here, the maximum acceleration is not taken into consideration. The optimal trajectory that minimizes the operation time cannot be found unless the spatial path and the velocity pattern are generated at the same time considering the maximum acceleration, but the calculation amount increases if the velocity pattern is also generated at the same time considering the maximum acceleration. . Here, first, a spatial path that minimizes Equation 3 is generated, and then an approximate optimal trajectory generation method that generates an optimal velocity pattern when operating along the spatial path is adopted. If the trajectory is long enough to reach the maximum speed of the robot arm, the operation time may be taken into consideration. Also, a method for simultaneously generating a spatial path and a velocity pattern will be described later.

Since the calculation formula of the angular momentum of the robot arm is non-linear, it is difficult to analytically obtain the spatial path that minimizes the evaluation function described above. Therefore, FIGS.
As shown in the table, the motion range of the robot arm
Unit trajectories 402 are set radially at each grid point on the mesh 401 table, and a spatial route is numerically searched using dynamic programming.

A procedure for obtaining a trajectory will be described with reference to FIGS. 6 and 7. First, the operation of the offline table data generating means 7 shown in FIG. 6 will be described. In the step number 600 of the flowchart of FIG. 6 (hereinafter, only the number is displayed at the end of the step such as (600))}, the minimum coordinates of each mesh are put in variables x, θ, and z, and the unit trajectory number Put 1 in n.

Next, the matrix C of the unit orbit whose coordinates of the starting point is [x, θ, z] and the unit orbit number is n is calculated by the equation 4, and the table data storage means 8 is used as table data on the unit orbit. It is stored in (601).

[0038]

[Equation 4]

Here, the matrix C is obtained by integrating the intermediate calculation value matrix D of the angular momentum along the unit trajectory.

Next, the unit orbit number n is compared with n1 representing the number of unit orbits (603). If the unit orbit number n is not equal to n1, 1 is added to n (602), and the operation (60) is performed.
Return to 1), and if equal, proceed to the next operation (604).

Next, the joint angles of the arm corresponding to the coordinates [x, θ, z] of the arm tip are stored in the table data storage means 8 as table data relating to lattice points (6).
04). Here, the joint angle of the arm corresponding to the coordinates [x, θ, z] of the arm tip is generated during the operation (601).

Next, the unit orbit number n is returned to 1 (60
5). Next, the variable x is compared with the maximum x coordinate value of the mesh (607), and if the variable x is not equal to the maximum x coordinate value of the mesh, the step size in the x direction is added to the variable x (606),
Returning to the operation (601), if equal, the variable x is returned to the minimum x coordinate value of the mesh (608), and the next operation (61).
Move to 0). Next, the variable θ is compared with the maximum θ coordinate value of the mesh (610), and if the variable θ is not equal to the maximum θ coordinate value of the mesh, the step size in the θ direction is added to the variable θ (60
9), return to the operation (601), and if equal, the variable θ
Is returned to the minimum θ coordinate value of the mesh (611), and the next operation (610) is performed. Next, the variable z and the maximum z coordinate value of the mesh are compared (613). If the variable z is not equal to the maximum z coordinate value of the mesh, the step size in the z direction is added to the variable z (612), and the operation ( Return to 601), and if equal, end.

That is, for all the unit trajectories 402 of all the lattice points of the mesh 401, the matrix C, which is the integral of the intermediate calculation value matrix D of the angular momentum along the unit trajectory, is calculated by the equation 4, and the unit is calculated. The table data of the matrix C regarding the trajectory and the table data of the arm joint angles regarding the lattice points are generated. The calculation result is stored in the table data storage means 8 and referred to as table data by the trajectory generation means 4. By generating the table data for the grid points of the mesh and the unit trajectories set at the respective grid points of the mesh, the table data can be constructed accurately and efficiently.

The trajectory generating means 4 shown in FIG. 7 obtains the optimum trajectory by the following procedure. The vector s depending on the physical parameters of the artificial satellite body 10 and the link is calculated according to Step 1 Expression 5. (614)

[0045]

[Equation 5]

Step 2 In Step 2, the evaluation function table 500 shown in FIG. 5 is created. In the evaluation function table 500, all grid points [x, θ,
z], the end point of the trajectory [x (T), θ (T), z
(T)] to operation point f of the optimum trajectory from the grid point
(X) and the optimum unit orbit selected at the lattice point are entered. Fill in all columns in this evaluation function table.
In the procedure, first, the values of the operation costs f (x) in all the columns of the evaluation function table are filled with infinite values and initialized (61).
5). Next, the operation cost f (x) in the column of the end point position of the trajectory
The value of is filled with 0 (616). Next, the variable x is the end point x
The value obtained by subtracting the step size in the x direction from the coordinate is put into the variables θ and z as the minimum coordinate value of the mesh (617).

Next, using the matrix C stored in the table data storage means 8 and the vector s calculated in Step 1, the integral of the angular momentum along the unit trajectory is calculated by the formula 6. Using the integral of this angular momentum along the unit trajectory and the joint angle stored in the table data storage means 8, the movement of the optimum trajectory from the end point of the trajectory to the lattice point [x, θ, z] is calculated by Equation 7. The cost value is obtained, and the operation cost value and the optimum unit trajectory selected at that time are filled in the evaluation function table shown in FIG. 5 (61).
8).

[0048]

[Equation 6]

[0049]

[Equation 7]

Next, the variable z is compared with the maximum z coordinate value of the mesh (620). If the variable z is not equal to the maximum z coordinate value of the mesh, the step size in the z direction is added to the variable z (62
1), return to the operation (618), and if equal, the variable z
Is returned to the minimum z-coordinate value of the mesh (619) and the next operation (623) is performed. Next, the variable θ and the maximum θ coordinate value of the mesh are compared (623). If the variable θ is not equal to the maximum θ coordinate value of the mesh, the step size in the θ direction is added to the variable θ (624), and the operation ( 618) and if they are equal,
The variable θ is returned to the minimum θ coordinate value of the mesh (622), and the process proceeds to the next operation (625). Next, the variable x is compared with the x coordinate value of the starting point of the trajectory (625), and the variable x is the x of the starting point of the trajectory.
If it is not equal to the coordinate value, the step size in the x direction is reduced from the variable x (626), and the process returns to the operation (618). If they are equal, the next operation (627) is performed.

That is, the matrix C stored in the table data storage means 8 and the vector s calculated in Step 1
Equation 7 and Equation 8 are repeatedly calculated using, and the evaluation function table 500 shown in FIG. 5 is completely filled. The order of the directions 501 for filling the evaluation function table 500 is the arrow direction, that is, x = x
Fill from (T) toward x = x (0).

Step 3 In Step 3, the evaluation function table 500 is referred to, and the optimum spatial path of the robot arm is generated. First, 0 is set to the coordinate value N of the starting point of the trajectory as the initial value of the point sequence of the spatial path (627). Next, in the evaluation function table, the optimum unit trajectory in the column corresponding to the coordinate value of the N-th element of the point sequence of the trajectory is referred to, and the N + 1-th element of the point sequence of the spatial path is obtained by Expression 8 (628 ).

[0053]

[Equation 8]

Next, 1 is added to N (629). The coordinate value of the N-th element of the point sequence of the trajectory and the coordinate value of the end point of the trajectory are compared. If the coordinate value of the N-th element of the trajectory point sequence is not equal to the coordinate value of the end point of the trajectory, the operation of (628) above. Return to and end if equal.

That is, the spatial path is obtained by referring to the evaluation function table and repeatedly calculating the equation 8 from the starting point of the trajectory. Here, in some cases, the values between the grid points of the mesh are required in Step 2 and Step 3, but the values between the grid points of the mesh are generated by linear complementation.

The optimum spatial path has been obtained by the above operation. Next, a method of generating a velocity pattern will be described. First, from the starting point, along the optimum spatial path obtained by the method of FIG. 7, acceleration is performed at maximum acceleration, and when the maximum velocity is reached, a trajectory that operates at maximum velocity is generated. However, it is necessary to consider the maximum angular velocity and the maximum angular acceleration of the RW here. The maximum velocity and the maximum acceleration described here are the maximum angular velocity and the maximum angular acceleration of all the joints of the robot arm and all the axes of the RW. Considering these, the maximum velocity and the maximum acceleration of the robot arm tip that do not exceed them. Similarly, from the end point,
By accelerating with the maximum acceleration along the optimal spatial path obtained by the method of (1), when the maximum velocity is reached, a trajectory that operates at the maximum velocity is generated. Next, connect the two orbits at the points where the velocities are equal. From the above, the optimum trajectory can be obtained.

A concrete method of generating an optimum trajectory according to the fourth embodiment of the present invention will be described with reference to FIGS. 8 and 9. Figure 6
Similarly to the method shown in FIG. 7, the operation range is divided by the mesh 401 as shown in FIGS. 3 and 4, the space is discretized, and the radial unit trajectory is set at each lattice point of the mesh. Further, in order to consider the acceleration, an axis representing the magnitude V of the operation speed of the arm is provided and discretized. The orbit search is this x-θ-
Perform in zV space.

Step 1A Operation of the offline table data generating means 7, and
The operation of Step 1A of the trajectory generating means is the same as that of the third embodiment which is the method shown in FIG. 6 as shown in FIG.

The operations of Steps 2A and 3A corresponding to the operations of Steps 2 and 3 of the trajectory generating means 4 of the third embodiment described above will be described. Step2A In Step2A, the evaluation function table 500 is created. In the evaluation function table 500, the end points of the trajectory [x (T), θ (T), z (T), for all the lattice points [x, θ, z, V].
0] to the lattice point, the operating cost f of the optimum trajectory
(X), the optimum unit trajectory selected at the lattice point, and the optimum speed change are entered. This evaluation function table 50
The procedure for filling in 0 is as follows.

First, the operation cost values in all columns of the evaluation function table 500 are filled with infinity values and initialized (70).
0). Next, the end point of the trajectory [x (T), θ (T), z
The value of the operation cost in the column of (T), 0] is filled with 0 (70
1). Next, the value obtained by subtracting the step size in the x direction from the x coordinate of the end point is put into the variable x, the minimum coordinate value of the mesh is put into the variables θ and z, and 0 is put into the variable V (702).

Next, using the equation 9, the operating cost value of the optimum trajectory from the end point of the trajectory to the lattice point [x, θ, z, V], the optimum unit trajectory, and the optimum speed change are obtained and the evaluation function is obtained. Fill the table. However, the joints of the robot arm, and
If the angular velocity or the angular acceleration of one of the RW axes exceeds the maximum value, the value of the operation cost is set to infinity. Here, the joints of the robot arm and the angular velocities and angular accelerations of the RW axes are obtained by equations 10 and 11. In the middle of this calculation, the matrix D stored in the table data storage means 8 and the joint angle data of the robot arm are used. (703)

[0062]

[Equation 9]

[0063]

[Equation 10]

[0064]

[Equation 11]

Next, the variable V is compared with the maximum velocity of the arm tip (705). If the variable V is not equal to the maximum velocity of the arm tip, the step size of the velocity is added to the variable V (706),
Return to the operation (703). If they are equal, the variable V is returned to 0 (704) and the next operation (708) is performed. Next, the variable z and the maximum z coordinate value of the mesh are compared (708). If the variable z is not equal to the maximum z coordinate value of the mesh, the step size in the z direction is added to the variable z (709), and the operation ( 70
Returning to 3), if they are equal, the variable z is returned to the minimum z coordinate value of the mesh (707), and the next operation (711) is performed. Next, the variable θ and the maximum θ coordinate value of the mesh are compared (71
1) If the variable θ is not equal to the maximum θ coordinate value of the mesh, the step size in the θ direction is added to the variable θ (712) and the process returns to the operation (703). If they are equal, the variable θ is set to the minimum θ coordinate of the mesh. The value is returned (710) and the next operation (713) is performed. Next, the variable x is compared with the x coordinate value of the starting point of the trajectory (713), and if the variable x is not equal to the x coordinate value of the starting point of the trajectory, the step size in the x direction is reduced from the variable x (71
4) Return to the operation (703), and if they are equal, move to the next operation (715).

Step 3A Here, with reference to the evaluation function table 500, the trajectory is obtained from the starting point of the trajectory by using Expression 12. First, the coordinate value of the starting point of the trajectory is entered as the initial value of the point sequence of the trajectory, 0 is entered in the magnitude of the velocity at that time, and 0 is entered in N (715). Next, referring to the optimal unit trajectory and the optimal velocity change in the column corresponding to the N-th element of the point sequence of the trajectory in the evaluation function table, Equation 12
The N + 1th element of the point sequence of the orbit is calculated by (71
6). Next, 1 is added to N (717). N of orbital point sequence
If the coordinate value of the N-th element of the sequence of points of the trajectory is not equal to the coordinate value of the endpoint of the trajectory, the procedure returns to the operation of (716) above, and if it is equal, ends. To do.

[0067]

[Equation 12]

Here, in Step 2A and Step 3A, there may be a case where the mesh grid points are required, but the values between the mesh grid points are generated by linear interpolation.

An operation method for setting the optimum trajectory which is the fifth embodiment of the present invention will be described. FIG. 10 is an example of a graphics screen for optimizing the trajectory. Reference numeral 800 is a robot operation display window for displaying the operation of the robot.
A robot command display window 810 displays the robot command taught by the operator. A command window 820 displays commands that can be instructed by the operator. An operation instruction window 830 displays an explanation of a teaching method for the operator. A cursor 840 is moved by a mouse.

Reference numeral 801 denotes an object A, 80 operated by the robot.
Reference numerals 2 and 803 denote a port B and a port C to which the object A can be connected.

Reference numerals 811 and 812 are robot commands indicating the work taught by the operator. grasp (A) is a robot command indicating the work of grasping the object A, and attach (B) is a robot command indicating the work of connecting the object to the port B. The line connecting the robot commands shows the state when the work is completed. 81
3 is the initial state before starting work, 814 is grasp (A)
815 is atta when the work is finished
It shows the state when the work of ch (B) is completed.
Reference numerals 816 and 817 are icons for scrolling the screen.

Reference numeral 821 is a command designated when performing optimization. Reference numeral 822 is a command designated when teaching the passing point of the trajectory. Reference numeral 823 is a command specified when performing a simulation. A command 824 is designated when returning to the command of the previous stage.

The operation for optimizing the trajectory will be described. 1. The cursor is moved to the optimization command 821 and the mouse button is clicked.

2. A message "Which work arm trajectory do you want to optimize?" Is displayed in the operation command 830, and the robot command display window 82 is displayed.
0 indicates which robot command will optimize the trajectory of the work. For example, when optimizing the arm trajectory of the grasp (A), move the cursor to 813 and click.

3. A message "Please instruct the movable range of the arm." Is displayed in the operation command 830, so move the robot arm 11 with the mouse and specify a rectangular parallelepiped with all sides parallel to the coordinate axis several times. Specify the operable range by increasing or decreasing the range. Since a rectangular parallelepiped whose all sides are parallel to the coordinate axis can be specified by designating two diagonal points, as shown in the figure, the rectangular parallelepiped is designated by designating these two points. For example, the rectangular parallelepiped 804 is designated by designating points 805 and 806.

At this time, the robot arm 1 is parallel to the coordinate axes.
If 1 is moved, it is easy to understand the range of interference and it is convenient. Such an operation is necessary because the calculation time becomes longer if the interference check between the objects is performed when the optimum trajectory of the robot arm 11 is obtained. Further, in the designated work, when the robot arm 11 is grasping an object, the movable range is designated while grasping the object.

Further, in the case of a simulation for confirming the operation, the simulation command 823 is selected, and which work to which work is to be simulated is designated in the robot command display window. At that time, it is also possible to open and display a window that displays a graph such as the angular momentum of the arm. Also, 814 and 815
If you specify a line that connects robot commands such as, the status at that time can be displayed.

When the passing point instruction command 822 is selected, the passing point can be taught while moving the robot arm 11 in the robot operation display window 800. Also, the trajectory may be corrected based on the optimum trajectory.

[0079]

According to the present invention, in the robot motion planning apparatus and method, the optimum trajectory of the robot arm can be obtained at high speed. Further, the amount of communication between the robot motion planning device and the robot controller can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram of a robot motion planning apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram of a robot operation planning device and a controller according to a second embodiment of the present invention.

FIG. 3 is a plan view showing a method of meshing an operation range of a robot arm of a robot operation planning device according to a third exemplary embodiment of the present invention.

FIG. 4 is a front view showing a method of meshing an operation range of a robot arm of a robot operation planning device according to a third embodiment of the present invention.

FIG. 5 is a diagram showing an evaluation function table of the robot motion planning apparatus according to the third embodiment of the present invention.

FIG. 6 is a flowchart showing an optimal trajectory generation method by an offline dynamic programming method of a robot motion planning apparatus according to a third exemplary embodiment of the present invention.

FIG. 7 is a flowchart continued from FIG. 6 showing an optimal trajectory generation method by the off-line dynamic programming of the robot motion planning apparatus according to the third embodiment of the present invention.

FIG. 8 is a flowchart showing an optimal trajectory generation method by another off-line type dynamic programming method of the robot motion planning apparatus of the fourth exemplary embodiment of the present invention.

FIG. 9 is a flowchart following FIG. 8 showing an optimal trajectory generation method by another off-line dynamic programming method of the robot motion planning apparatus according to the fourth exemplary embodiment of the present invention.

FIG. 10 is a block diagram showing an optimization designation screen displayed on the input / output device of the robot motion planning apparatus according to the fifth embodiment of the present invention.

[Explanation of symbols]

 1 Input / Output Means 2 Robot Command Generation Means 3, 3A Robot Command Interpretation Means 4, 4A Trajectory Generation Means 5, 5A RW Command Value Generation Means 6 Simulator 7 Offline Table Data Generation Means 8, 8A Table Data Storage Means 9, 9A Physical Parameters Storage means 10 Artificial satellite body 11 Robot arm 12 Reaction wheel 13,13A Communication means 14,14A Robot operation planning device 15 Robot controller. 16,16A Robot command 17,17A Trajectory start and end points and movable area 18,18A Optimum trajectory 19,19A RW command value 20,20A Table data 21 Artificial satellite robot motion planner 401 Mesh 402 Unit trajectory 500 Evaluation function table 501 Direction for filling evaluation function table 800 Robot operation display window 801 Object A 802 Port B 803 Port C 804 Rectangular solid 805 points 806 points 810 Robot command display window 811 Robot command 812 Robot command 813 Initial state 814 Work end state 815 Work end state 816 Icon 817 Icon 820 Command display window 821 Optimization command 822 Pass point command 823 Simulation command 824 Previous command 830 Operation Instruction window 840 Cursor

Claims (14)

[Claims] 1. An input / output unit, a robot command generation unit connected to the input / output unit for generating a work command to a robot, and connected to the robot command generation unit to interpret the work command to interpret a trajectory of a robot arm. Robot command interpreting means for generating a start point and an end point and a movable area, trajectory generating means for generating a trajectory of a robot arm connected to the robot command interpreting means, and a physical parameter for storing physical parameters of each link of the robot arm. In a robot operation planning device including a storage means, the calculation data of the operation cost of the robot arm or an intermediate calculation value of the calculation data required by the trajectory generation means connected to the physical parameter storage means is offline. An offline table data generating means for calculating the table data in advance and And a table data storage unit connected to the offline table data generation unit for storing the table data and delivering the table data to the trajectory generation unit, and the trajectory generation unit generates an optimum trajectory using the table data. A robot motion planning device characterized by being a thing. 2. The table data divides an operating range of the robot arm using a mesh table, and a table is formed by radial unit trajectories and grid points set from the robot arm to each grid point of the mesh table. The robot motion planning apparatus according to claim 1, wherein the robot motion planning apparatus is the converted data. 3. The table data is data obtained by integrating the angular momentum of the robot arm along a unit trajectory, or data obtained by integrating an intermediate calculated value of the angular momentum of the robot arm along the unit trajectory. The robot motion planning apparatus according to claim 2, wherein: 4. The robot motion planning apparatus according to claim 2, wherein the table data is data including a joint angle of the robot arm. 5. The trajectory generation means is a means for controlling a reaction wheel that is installed in the artificial satellite body and controls the attitude of the artificial satellite body, and generates a trajectory with the shortest operation time of the robot arm. The robot motion planning apparatus according to any one of claims 1 to 4, which is a means. 6. A simulator that is connected to the physical parameter storage means, the trajectory generation means, and the reaction wheel command value generation means, and has a simulator capable of confirming the operation of the robot in advance based on the data from the respective means. The robot motion planning apparatus according to claim 1, wherein: 7. The robot motion planning apparatus according to claim 1, wherein the input / output unit has a robot command display unit for displaying the robot arm and a surrounding environment of the robot arm as a graphic. 8. A communication means for receiving a work command, a robot command interpreting means for interpreting the work command to the robot and generating a start point and an end point of the trajectory of the robot arm, and an operable area, and connected to the robot command generation means. A robot controller comprising: trajectory generating means for generating a trajectory of the robot arm; and physical parameter storage means for storing a physical parameter of each link of the robot arm, wherein the robot required by the trajectory generating means. Table data storage means for storing calculation data of arm operation cost or intermediate calculation value of the calculation data off-line in advance and delivering it to the trajectory generation means, wherein the trajectory generation means is Characterized by generating an optimal trajectory using table data Robot controller. 9. A satellite comprising a robot having a robot arm, a reaction wheel for attitude control, and a robot motion planning device used for motion planning of the robot arm, wherein the robot motion planning device comprises:
An artificial satellite comprising the robot operation planning apparatus according to any one of claims 1 to 5. 10. An optimal trajectory generation method for generating a trajectory of a robot arm, wherein the optimal trajectory generation method divides an operation range of the robot arm with a mesh table and directs the robot arm to each grid point of the mesh table. The data necessary for calculating the operation cost of the robot arm tabulated by the radial unit trajectories and grid points set by the above or an intermediate calculation value of the calculation data is calculated off-line in advance and stored in the table data storage means. And a method of calculating the optimum trajectory of the robot arm by the trajectory generating means using the table data from the start point and the end point of the trajectory of the robot arm and the movable area. . 11. The optimal trajectory generation method is used to generate an optimal trajectory that can control the attitude of the artificial satellite body and has the shortest operation time of a robot arm mounted on the artificial satellite body. The optimal trajectory generation method according to claim 10. 12. The table data is obtained by a method of integrating an angular momentum of the robot arm or an intermediate calculated value of the angular momentum along a unit trajectory. , The optimal trajectory generation method according to any one of items. 13. The table data is obtained by a method of tabulating the joint angles of the robot arm by using grid points of a mesh table.
The optimal trajectory generation method according to any one of 0 and 11. 14. A robot arm and a robot command display means for graphically displaying the surrounding environment of the robot arm are used to move the movable area of the robot arm on the screen of the robot command display means. The optimal trajectory generation method according to claim 10, wherein the trajectory generated by the trajectory generation means can be corrected.