WO2024154250A1

WO2024154250A1 - Trajectory generation device and trajectory generation method

Info

Publication number: WO2024154250A1
Application number: PCT/JP2023/001320
Authority: WO
Inventors: 将吾東
Original assignee: 株式会社Fuji
Priority date: 2023-01-18
Filing date: 2023-01-18
Publication date: 2024-07-25

Abstract

Provided is a trajectory generation device that generates a trajectory of an operation of an articulated robot arm and comprises: a generation unit which searches for via points which can achieve an operation from a start point to a target point and generates a plurality of trajectory candidates each connecting these points with each other and having set movement parameters each between these points; an optimization unit which optimizes, for each of the plurality of trajectory candidates, the positions of the via points and the movement parameters such that a total time of a movement time of a hand tip of the robot arm to move from the start point to the target point and a convergence time for hand tip vibration to converge at the target point is shorter; and an evaluation/selection unit which evaluates the plurality of optimized trajectory candidates in consideration of the total time and selects the trajectory of the operation on the basis of the evaluation.

Description

Trajectory generation device and trajectory generation method

This specification discloses a trajectory generation device and a trajectory generation method.

　Conventionally, there have been proposals for generating trajectories for the movements of robot arms. For example, Patent Document 1 describes a method for generating a trajectory for moving the tip of a robot arm from a starting point to a target point in such a way that the movement time is minimized while satisfying constraints on the speed and acceleration of the tip.

JP 2015-58497 A

As mentioned above, when generating a trajectory for the movement of a robot arm, it is required to make the movement time as short as possible. However, shortening the movement time can easily increase vibration in the hand, which can have effects such as lengthening the convergence time from when the hand reaches the target point until the hand vibration converges. In such cases, it ultimately takes longer to complete the movement of the robot arm. Also, it is difficult for an operator to create a trajectory that reduces the effects of hand vibration while shortening the movement time, and it takes a lot of time to create it.

The primary objective of this disclosure is to quickly generate an appropriate trajectory that reduces the movement time of the robot arm's hand while minimizing the effects of hand vibration.

This disclosure takes the following steps to achieve the above-mentioned primary objective:

The trajectory generating device of the present disclosure is
A trajectory generating device for generating a trajectory of a motion of a multi-joint robot arm,
a generation unit that searches for possible waypoints from a start point of the motion to a target point, and generates a plurality of trajectory candidates that connect the points and have movement parameters between the points set;
an optimization unit that optimizes the positions of the via points and the movement parameters for each of the plurality of trajectory candidates so as to shorten a total time of a movement time for the hand of the robot arm to move from the start point to the target point and a convergence time for the hand vibration to converge at the target point;
an evaluation and selection unit that evaluates the plurality of trajectory candidates that have been optimized, including the total time, and selects the trajectory of the movement based on the evaluation;
The gist of the invention is to provide the following:

In the trajectory generation device disclosed herein, for each of a number of trajectory candidates, the positions of the via points and the movement parameters are optimized so as to shorten the total time of the movement time for the robot arm's hand to move from the start point to the target point and the convergence time for the hand vibration to converge at the target point. In addition, the multiple optimized trajectory candidates are evaluated, including the total time, and the movement trajectory is selected based on the evaluation. This makes it possible to quickly generate an appropriate trajectory that reduces the effects of hand vibration while shortening the movement time of the robot arm's hand.

FIG. 1 is a diagram showing an outline of the configuration of a robot system 10. 13 is a flowchart showing an example of a trajectory generation process. FIG. 13 is a schematic diagram showing an example of a manner in which a trajectory is generated. FIG. 13 is a schematic diagram showing an example of a manner in which a trajectory is generated. FIG. 13 is a schematic diagram showing an example of a manner in which a trajectory is generated. FIG. 13 is a schematic diagram showing an example of a manner in which a trajectory is generated. FIG. 13 is a schematic diagram showing an example of a manner in which a trajectory is generated. 11 is a flowchart showing an example of an optimization process. FIG. 11 is an explanatory diagram showing an example of a moving time and a convergence time.

Embodiments of the present disclosure will be described with reference to the drawings. FIG. 1 is a schematic diagram showing the configuration of a robot system 10. The robot system 10 includes a robot 20 and a trajectory generating device 40. The robot 20 is configured as, for example, a six-axis vertical articulated robot, and performs a predetermined task, such as picking up a workpiece (not shown) and placing it in a predetermined position.

The robot 20 comprises a base 21 fixed to the workbench 11, a robot arm 22 including a plurality of links connected in series via six joint axes (first to sixth joint axes J1 to J6), and a control device 30 that controls the operation of the robot 20. The robot arm 22 is provided with servo motors 23a to 23f that rotate the first to sixth joint axes J1 to J6, respectively, and encoders (rotary encoders) 24a to 24f that detect the rotation angles of the servo motors 23a to 23f, respectively.

The end effector 25 is detachably attached to the tip link of the robot arm 22 as a work tool, and a camera (not shown) can also be attached. The end effector 25 is composed of a suction nozzle that attracts the workpiece by negative pressure, a mechanical chuck that grips the workpiece with a pair of claws, an electromagnetic chuck that attracts the workpiece with an electromagnet, etc., and is appropriately selected according to the shape and material of the workpiece to be worked on. In FIG. 1, a mechanical chuck is shown as an example of the end effector 25. The end effector 25 is provided with an actuator 26 that drives the pair of claws 25a to open and close, and an encoder (linear encoder) 27 that detects the open/close position of the actuator 26.

The robot 20 also includes a power supply circuit 32 that converts AC power from a commercial power source (not shown) into DC power and supplies it to each part of the robot 20, and amplifiers 34a to 34f that drive the servo motors 23a to 23f with power from the power supply circuit 32. Each amplifier 34a to 34f drives each of the servo motors 23a to 23f by driving a switching element (not shown).

The control device 30 is configured as a microprocessor centered around a CPU (not shown), and in addition to the CPU, it is equipped with ROM, RAM, an input/output interface, etc. Detection signals from the encoders 24a to 24f of the servo motors 23a to 23f, detection signals from the encoder 27 of the actuator 26, etc. are input to the control device 30. The control device 30 also outputs control signals to the amplifiers 34a to 34f, control signals to the actuator 26, etc. The control device 30 is also configured to be able to communicate with the trajectory generating device 40, and controls the drive of the robot arm 22 based on the trajectory generated by the trajectory generating device 40.

The trajectory generating device 40 is a computer that includes a control device 41, a storage device 46, an input device 47, and a display device 48, and generates a trajectory for the robot arm 22 of the robot 20. The control device 41 is configured as a microprocessor centered around a CPU (not shown), and in addition to the CPU, includes ROM, RAM, an input/output interface, etc. The storage device 46 is configured, for example, by a HDD, and stores various data and programs required to generate the trajectory of the robot arm 22. The input device 47 is, for example, a keyboard, a mouse, etc., through which the operator performs input operations. The display device 48 is, for example, an LCD display, etc., that displays various information.

The control device 41 includes a condition setting unit 42, a generation unit 43, an optimization unit 44, and an evaluation and selection unit 45 as functional blocks for generating a trajectory of the movement of the robot arm 22. The condition setting unit 42 sets various conditions such as constraint conditions when generating a trajectory. The generation unit 43 searches for waypoints from the start point S of the movement to the target point G by changing search parameters under the constraint conditions set by the condition setting unit 42, and generates a base trajectory that serves as the basis for evaluation and selection by the evaluation and selection unit 45, and trajectory candidates based on the base trajectory. The generation unit 43 uses, for example, an RRT-Connect-based search method that is an extension of RRT as a search method based on RRT (Rapidly exploring Random Tree). The optimization unit 44 performs necessary trajectory adjustments and optimization on the base trajectory and trajectory candidates generated by the generation unit 43. The evaluation and selection unit 45 evaluates the trajectory that has been subjected to necessary trajectory adjustments and optimization by the optimization unit 44, and selects a trajectory based on the evaluation. The trajectory selected by the evaluation and selection unit 45 is transmitted to the robot 20, and the drive control of the robot arm 22 is performed based on the trajectory. The optimization unit 44 may also optimize the trajectory selected by the evaluation and selection unit 45. The trajectory generated in this manner is stored in the storage device 46.

The following is an explanation of the trajectory generation process of the trajectory generation device 40. FIG. 2 is a flowchart showing an example of the trajectory generation process. This trajectory generation process is executed by the control device 41 of the trajectory generation device 40 using the functions of each of the functional blocks described above.

In the trajectory generation process, the control device 41 first acquires various information related to the trajectory to be generated, such as the start point S and target point G of the movement in the joint angle space, information indicating the area of the obstacle B, and constraint conditions (S100). In this embodiment, the start point S, target point G, waypoints between them, and information indicating the area of the obstacle B are expressed by the joint angles (movable angles) of each joint in the joint angle space, rather than by position coordinates in the XYZ space. Since the robot arm 22 has six joints, the first to sixth joint axes J1 to J6, the position information of each point, such as the start point S, target point G, and waypoints in the joint angle space, is expressed by the parameters of six joint angles. For this reason, even if it is possible to set a waypoint by searching for the position of the hand of the robot arm 22 in the XYZ space, it is possible to prevent problems such as going outside the movable angle range of the joints of the robot arm 22 while moving to the waypoint. In addition, various constraint conditions are defined based on the specifications of the robot 20 and the environment around the robot 20, such as the movable angle range of each joint, torque constraints, speed constraints, and acceleration constraints.

Next, the control device 41 searches for waypoints using an RRT-Connect-based search method to generate a base trajectory (S110). Figures 3 to 7 are schematic diagrams showing an example of how a trajectory is generated. As described above, each point is represented in joint angle space, but for convenience of illustration, the schematic diagrams are shown in a two-dimensional manner. As shown in the figure, for example, from a starting point S on the left to a target point G on the right, waypoints (white circles) are searched for so as not to interfere with three obstacles B, and a base trajectory is generated that connects each point. Note that the position of the waypoint indicates, for example, the position of the hand of the robot arm 22. Also, the space between each point connected by a straight line is called a section.

In S110, the control device 41 starts searching for waypoints from both the start point S side and the target point G side, and ends the search when one waypoint searched for on the start point S side can be connected with one waypoint searched for on the target point G side by a straight line (dotted lines in Figures 3 and 4). This is because the operation of the robot arm 22 at each point and between them is guaranteed by representing each point in joint angle space, so if a waypoint on the start point S side and a waypoint on the target point G side are connected by a straight line, the robot arm 22 can operate between them. For this reason, in the example of Figure 4, it is possible to generate a base trajectory that omits the search for waypoints along the way compared to the example of Figure 3. This allows the control device 41 to shorten the time required to search for waypoints and quickly generate a base trajectory.

Furthermore, in S110, the control device 41 can change the search distance. Specifically, the control device 41 can change the search distance so that the longer the distance from the search source point to the obstacle B, the longer the search distance, and the shorter the distance from the search source point to the obstacle B, the shorter the search distance. The control device 41 first calculates the distance from the search source point to each obstacle B, and derives the shortest distance among them. For example, when the search source point is the start point S, the control device 41 calculates the distance from the start point S to each obstacle B1, B2, and B3, and derives the distance to obstacle B1 as the shortest distance among them, and searches for waypoints with a search distance D1 according to that distance. Note that in the search direction from the start point S toward the obstacle B1, waypoints are searched for so as not to interfere with the obstacle B1, and as a result, the search distance D1' is shorter than the search distance D1.

When searching for the next via point from via point P in FIG. 5, the control device 41 calculates the distance from via point P to each of the obstacles B1, B2, and B3, derives the distance to obstacle B2 as the shortest distance, and searches for the via point at a search distance D2 corresponding to that distance. Similarly, when searching for a via point from target point G, the control device 41 searches for the via point at a search distance D3. Note that in the search direction from target point G to obstacle B2, the via point is searched for so as not to interfere with obstacle B2, and as a result, the search distance D3' is shorter than the search distance D3. In FIG. 5, search distances D1, D2, and D3 are illustrated, but the search distance is set steplessly according to the distance from the search source to obstacle B. Alternatively, the search distance may be set stepwise according to the distance from the search source to obstacle B. In this way, by making the search distance variable, the control device 41 can efficiently search for via points and quickly generate a base trajectory.

The control device 41 searches for waypoints in this way to generate one base trajectory, for example the base trajectory R0 in Figure 6. Next, the control device 41 generates multiple trajectory candidates based on the generated base trajectory (S120). In this embodiment, multiple trajectory candidates are generated by randomly changing the positions of the waypoints of the base trajectory and movement parameters such as speed, acceleration/deceleration, and synthesis rate. As shown in Figure 7, each waypoint of the base trajectory R0 (dotted line) is changed and the movement parameters are changed to generate, for example, three trajectory candidates R1 to R3. Note that the control device 41 only needs to generate multiple trajectory candidates, and the trajectory candidates may include the base trajectory R0.

Here, the values of the movement parameters such as the speed, acceleration/deceleration, and synthesis rate are determined for each waypoint of the base trajectory, and these values are used to perform trajectory adjustment including the moving average process and synthesis process described below. The moving average process is a process of deriving a command value by taking the moving average of the original speed for each calculation period within a section, for example, in order to make the speed change in each section gentle. Note that the original speed is determined to be a predetermined speed, such as the maximum speed of the robot arm 22 under constraint conditions. The synthesis process is performed after the moving average process, and is a process of multiplying the command value of the previous section and the command value of the following section by a ratio (synthesis rate) respectively set so that the speed does not decrease in consecutive sections (connections between adjacent sections). These processes are well known, so detailed explanations are omitted, but the trajectory after the synthesis process may have some parts that are inwardly oriented compared to the original trajectory that was generated initially, i.e., a trajectory that connects the waypoints with a straight line.

Then, the control device 41 executes an optimization process to optimize each of the generated trajectory candidates (S130). FIG. 8 is a flowchart showing an example of the optimization process. In the optimization process, the control device 41 selects one trajectory candidate to be processed (S200), and searches for the positions of each waypoint and each movement parameter by, for example, a gradient method so as to shorten the total time of movement time and convergence time on the trajectory candidate to be processed (S210). Note that the control device 41 is not limited to the gradient method, and may search for the positions of each waypoint and each movement parameter by using other methods such as particle swarm optimization or a genetic algorithm.

Here, FIG. 9 is an explanatory diagram showing an example of the movement time and convergence time, with the horizontal axis showing time and the vertical axis showing the position of the hand of the robot arm 22. As shown in the figure, the movement time Ta is the time from when the hand of the robot arm 22 starts moving from the starting point S to when it reaches the target point G via each waypoint at time t0. The convergence time Tb is the time from time t0 to when the vibration of the hand of the robot arm 22 at the target point G converges. The issuance of the operation command value for the robot arm 22 is completed at time t0. The convergence time Tb is the time from when the hand of the robot arm 22 reaches the target point G to when the vibration of the hand converges within the allowable range A, and is also called the stabilization time. The time that combines the movement time Ta and the convergence time Tb is called the total time. The movement of the robot arm 22 along the trajectory is completed in the time that combines the movement time Ta and the convergence time Tb, so the total time is also called the operation time.

In S210, the control device 41 calculates the total time by calculating the movement time Ta and the convergence time Tb when the positions of the via points and the movement parameters are changed, and optimizes the positions of the via points and the movement parameters so as to shorten the total time. Based on the changed via points and movement parameters, the control device 41 calculates the position from the start point S to the target point G for each predetermined time and the hand vibration at that position, and calculates the movement time Ta until the target point G is reached. The control device 41 also stores in the storage device 46 a well-known vibration model that models the size and weight of each element of the robot arm 22 and the size and weight of the workpiece to be grasped by the end effector 25, and when the positions and movement parameters for each predetermined time are given, the control device 41 uses the vibration model to calculate the hand vibration for each predetermined time. The control device 41 then calculates the time required for the hand vibration generated at the target point G in the vibration model to converge within the above-mentioned allowable range A as the convergence time Tb.

Then, the control device 41 executes the above-mentioned moving average processing and synthesis processing (S220). As described above, by executing these processing, the trajectory may turn inward and interfere with the obstacle B. Therefore, the control device 41 checks interference with the obstacle B in the trajectory candidate on which the moving average processing and synthesis processing have been executed (S230), and judges whether the total time is shorter without interference (S240). If the control device 41 judges in S240 that there is interference, or that the total time is not short even without interference, it returns to S210. When the control device 41 returns to S210, it executes the processing of S210 to S240, so that the processing of searching for the position and movement parameters of each way point so that the trajectory candidate to be processed does not interfere with the obstacle B and the total time is shorter is repeated. On the other hand, if the control device 41 judges in S240 that the total time is shorter without interference, it retains the way point and movement parameters searched for in S210 (S250). The total time for the retained waypoints and movement parameters is the shortest total time for that trajectory candidate at that point in time, and is therefore compared with the total time when the determination in S240 is made thereafter. Note that the process of optimizing the trajectory by searching for waypoints and movement parameters so as not to interfere with obstacle B and to shorten the total time is also called trajectory adjustment.

Next, the control device 41 judges whether the search for the trajectory candidates (each waypoint and movement parameter) to be processed has been completed based on whether a predetermined termination condition has been met (S260). The predetermined termination condition may be, for example, a condition that the number of executions (repetitions) of S210 to S250 has reached a predetermined number, or a condition that the gradient has converged to a predetermined range when the gradient method is used in S210. If the control device 41 judges in S260 that the search has not been completed, it returns to S210 and executes the processing from S210 onwards again. As described above, in S240, it is judged whether the total time is shorter than the total time based on the waypoints and movement parameters held in the previous S250. On the other hand, when the control device 41 judges in S260 that the search for the trajectory candidates to be processed has been completed, it judges whether the search for all the trajectory candidates has been completed (S270). If the search for all the trajectory candidates has not been completed, the control device 41 returns to S200, selects a trajectory candidate and performs processing. On the other hand, when the control device 41 determines in S270 that the search for all trajectory candidates has been completed, it ends the optimization process.

In the trajectory generation process of FIG. 2, when the optimization process of S130 is performed, the control device 41 performs a trajectory selection process (S140). In S140, the control device 41 evaluates the total time of multiple optimized trajectory candidates and selects one trajectory candidate with the shortest total time as the trajectory of the robot arm 22's operation (S140). Note that if the hand is moved quickly to shorten the movement time Ta, the hand vibration increases when the target point G is reached, and the convergence time Tb increases, which may result in a longer total time. Conversely, if the hand is moved slowly to reduce the hand vibration when the target point G is reached in order to shorten the convergence time Tb, the movement time Ta may increase, which may result in a longer total time. Also, depending on whether the robot arm 22 is in an extended state, i.e., depending on the position of the via point to be passed, the hand vibration may change and the convergence time Tb may vary greatly. Therefore, the control device 41 of this embodiment optimizes multiple trajectory candidates so that the total time of the travel time Ta and the convergence time Tb is shortened, and further selects the trajectory candidate with the shortest total time from the multiple trajectory candidates. This allows the control device 41 to select a trajectory that reliably shortens the total time.

Then, the control device 41 stores the trajectory selected in S140 in the storage device 46 (S150), and ends the trajectory generation process. The control device 41 also outputs the trajectory generated (selected) in the trajectory generation process to the control device 30 of the robot 20. The robot 20 stores the trajectory in the storage unit of the control device 30, and performs drive control of the robot arm 22 based on the generated trajectory or trajectories to perform the above-mentioned predetermined task. The generated trajectory may be used for simulating the operation of the robot arm 22. In other words, the generated trajectory is not limited to the actual operation of the robot 20 (robot arm 22), and may be used for evaluating or verifying the operation. The trajectory generation device 40 may also have a function of executing such a simulation.

Here, the correspondence between the components of this embodiment and the components of this disclosure will be clarified. The control device 41 (generation unit 43) of the trajectory generation device 40 that executes S110 and S120 of the trajectory generation process of this embodiment corresponds to the generation unit of this disclosure, the control device 41 (optimization unit 44) that executes S130 of the trajectory generation process corresponds to the optimization unit, and the control device 41 (evaluation selection unit 45) that executes S140 of the trajectory generation process corresponds to the evaluation selection unit. Note that this embodiment also clarifies an example of the trajectory generation method of this disclosure by explaining the processing of the trajectory generation device 40.

The trajectory generation device 40 (control device 41) of this embodiment described above optimizes the positions of the waypoints and the movement parameters for multiple trajectory candidates so that the total time of the movement time Ta of the hand of the robot arm 22 and the convergence time Tb of the hand vibration is shortened. Then, a movement trajectory is selected based on the result of evaluating the total time of the multiple trajectory candidates. This makes it possible to quickly generate an appropriate trajectory that reduces the effects of hand vibration while shortening the movement time of the hand of the robot arm.

In addition, the trajectory generation device 40 selects the trajectory candidate with the shortest total time of the movement time Ta and the convergence time Tb of the hand vibration, so that it can more reliably generate a trajectory with the shortest total time.

The trajectory generating device 40 also calculates the hand vibration and movement time Ta for each predetermined time period until the hand moves from the starting point S to the target point G, and calculates the time from when the hand reaches the target point G until the hand vibration converges to the allowable range A as the convergence time Tb. This makes it possible to prevent the convergence time Tb from becoming unnecessarily long and appropriately optimize the trajectory candidates.

The trajectory generating device 40 generates a base trajectory by searching for the positions of via points in the joint angle space, connecting each point and setting movement parameters, and generates multiple trajectory candidates by changing the positions of the via points in the base trajectory and the movement parameters, so that trajectory candidates can be generated quickly. Furthermore, the via points searched for in the joint angle space are in positions that the robot arm 22 can reach, so there is no need for an operator to adjust the positions of the via points.

It goes without saying that this disclosure is in no way limited to the above-described embodiments, and can be implemented in various forms as long as they fall within the technical scope of this disclosure.

For example, in the embodiment, the hand vibration is calculated for each predetermined time period until the hand moves from the starting point S to the target point G, but this is not limited to this, and in order to calculate the convergence time Tb, it is sufficient to calculate at least the hand vibration when the hand reaches the target point G. In other words, the hand vibration at the target point G may be calculated without calculating the hand vibration during movement.

In the embodiment, the trajectory candidate with the shortest total time is selected as the movement trajectory, but this is not limited to the above. For example, the movement trajectory may be selected based on the results of evaluation including the total time, such as selecting a trajectory with the shortest total time from among trajectory candidates without interference. Also, depending on the object, smaller vibrations may be preferable even if the total time is slightly longer. In such a case, if the difference in total time is within a predetermined time, the trajectory candidate with the shorter convergence time Tb may be selected. In other words, the time including the total time is evaluated, and one of the trajectory candidates may be selected as the movement trajectory based on the evaluation value.

In the embodiment, the search for the waypoint is started from both the start point S and the target point G, but this is not limited, and the search for the waypoint may be started from either the start point S or the target point G. In the embodiment, the search distance is changed according to the distance to the obstacle B, but the search distance may be constant without being changed. In addition, the search method is not limited to the RRT-Connect based search method, and other search methods such as the RRT method and the potential method may be used. In the embodiment, the waypoint is searched in the joint angle space, but this is not limited, and the waypoint may be searched in, for example, the XYZ space. In the embodiment, the position of the waypoint in the base trajectory and the movement parameters are randomly changed to generate multiple trajectory candidates, but multiple trajectory candidates may be generated by searching for the waypoints and the movement parameters respectively.

In the embodiment, a six-axis vertical articulated robot is exemplified, but this is not limited to this and other vertical articulated robots such as a five-axis robot may be used, or a horizontal articulated robot may be used. In the embodiment, the trajectory generation device 40 generates the trajectory, but the control device 30 of the robot 20 may generate the trajectory. In addition, the trajectory generation device 40 and the control device 30 may work together to generate the trajectory.

This specification also discloses the technical idea of changing "the trajectory generating device according to claim 1 or 2" in claim 4, as originally filed, to "the trajectory generating device according to any one of claims 1 to 3."

This disclosure can be used in the technical field of generating trajectories for multi-joint robot arms.

10 robot system, 11 workbench, 20 robot, 21 base, 22 robot arm, 23a to 23f servo motors, 24a to 24f encoders, 25 end effector, 25a claw, 26 actuator, 27 encoder, 30 control device, 32 power supply circuit, 34a to 34f amplifier, 40 trajectory generator, 41 control device, 42 condition setting unit, 43 generation unit, 44 optimization unit, 45 evaluation and selection unit, 46 storage device, 47 input device, 48 display device, J1 first joint axis, J2 second joint axis, J3 third joint axis, J4 fourth joint axis, J5 fifth joint axis, J6 sixth joint axis, B obstacle, G target point, S start point.

Claims

A trajectory generating device for generating a trajectory of a motion of a multi-joint robot arm,
a generation unit that searches for possible waypoints from a start point of the motion to a target point, and generates a plurality of trajectory candidates that connect the points and have movement parameters between the points set;
an optimization unit that optimizes the positions of the via points and the movement parameters for each of the plurality of trajectory candidates so as to shorten a total time of a movement time for the hand of the robot arm to move from the start point to the target point and a convergence time for the hand vibration to converge at the target point;
an evaluation and selection unit that evaluates the plurality of trajectory candidates that have been optimized, including the total time, and selects the trajectory of the movement based on the evaluation;
A trajectory generating device comprising:
the evaluation and selection unit selects, as the trajectory of the movement, the trajectory candidate having the shortest total time from among the plurality of trajectory candidates that have been optimized.
The trajectory generating device according to claim 1 .
The optimization unit calculates the hand vibration and the movement time for each predetermined time period until the hand moves from the start point to the target point based on the position of the via point and the movement parameters, and calculates the time from when the hand reaches the target point until the hand vibration converges to an allowable range as the convergence time.
The trajectory generating device according to claim 1 or 2.
the generation unit generates a base trajectory by searching for positions of the via points in a joint angle space, connecting the points and setting the movement parameters, and generates a plurality of the trajectory candidates by changing the positions of the via points and the movement parameters in the base trajectory.
The trajectory generating device according to claim 1 or 2.
A trajectory generation method for generating a trajectory of a motion of a multi-joint robot arm, comprising the steps of:
(a) searching for possible waypoints from a start point of the motion to a target point, and generating a plurality of trajectory candidates in which each of the waypoints is connected and movement parameters between each of the points are set;
(b) optimizing the position of the via point and the movement parameters for each of the plurality of trajectory candidates so as to shorten the total time of a movement time for the hand of the robot arm to move from the start point to the target point and a convergence time for the hand vibration to converge at the target point;
(c) evaluating a plurality of the optimized trajectory candidates, including the total time, and selecting a trajectory for the movement based on the evaluation;
A trajectory generation method comprising: