JPH02277101A - Track generating method for manipulator - Google Patents

Abstract

PURPOSE:To move an object to its target position with a practical calculated variable by setting a safe area that does not change regardless of the positions set before and after the movement of the object carried out by a manipulator. CONSTITUTION:A manipulator 1 consists roughly of an arm part 2 and a hand part 3 and contains six joints (freedom degrees) which are defined as the 1st - 3rd axes (for part 2) and the 4th - 6th axes (for part 3) respectively. In addition, a three-dimensional space is secured as a safe area 6 where no interference is produced between the 4th - 6th axes and obstacles (base 4, structure 5) despite the attitudes of these three axes when the 4th axis is included in the area 6 and no change is produced for the area 6 regardless of the positions set before and after the movement of an object. When the track of the manipulator 1 is generated, a track covering the initial position of the manipulator 1 through the area 6 is obtained and then a track covering a target position through the area 6 is obtained. Thus a track connecting each point in the area 6 is obtained. As a result, the object can be moved to its target position with a small calculated variable.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a manipulator trajectory generation method for driving the manipulator in space without interfering with obstacles.

(Prior Art) Robots and manipulators, which only have arms, can perform a variety of complex tasks, unlike conventional machines that only perform monotonous tasks. However, most of the tasks that are taught to robots still use the teaching playback method. In the teaching playback method, a human actually moves the robot, then memorizes and reproduces the movement as sensor information. stop,
Humans have to teach it.

This was fine in the past when the same work was repeated over and over again in a factory, but from now on robots will be required to produce a greater variety of products in small quantities, and will also be required to go outside the factory to perform non-routine work. , it becomes difficult and even impossible for humans to actually teach the movements one by one.

Therefore, we create a model of the environment around the robot inside a computer, and based on that model, the computer automatically creates a trajectory for the manipulator that can transport the cargo to the target position without colliding with obstacles. Research is currently underway to generate this. For example, Proceedings of the Society of Instrument and Control Engineers VoN, 22. In No. 8, "Planning the collision avoidance motion of a manipulator using a free space classification representation method", by considering the manipulator separately into the arm part and the hand part, it is possible to calculate the posture of the hand in the free space of the arm. Classify into completely free space that does not collide with objects, and other partial free spaces,
In the partial free space, move while maintaining the hand posture,
By changing the posture in completely free space, it is possible to explore the obstacle avoidance problem of a six-degree-of-freedom articulated manipulator in three-dimensional space. Therefore, although the determined trajectory may be a circuitous one, a considerable reduction in the amount of calculation can be achieved. However, in this method, the free space must be determined in advance for all of the spaces to be searched (configuration space), so when an obstacle moves, it is necessary to recalculate all the free spaces. Therefore, when moving an object using a manipulator, the free space must be recalculated each time, resulting in problems such as an increase in the amount of calculation.

(Problems to be Solved by the Invention) As described above, conventionally, in order to reliably generate a trajectory in which a manipulator moves to a target position while avoiding obstacles,
Generally, a large amount of calculation was required. In view of the above-mentioned points, the present invention aims to provide a method for generating a trajectory of a manipulator that allows the manipulator to move to a target position with a practical amount of calculation.

[Structure of the Invention] (Means for Solving the Problem) In order to achieve the above object, the present invention uses a manipulator that has at least six joint axes and whose posture is determined by three joint axes on the distal end side. In a manipulator trajectory generation method, the manipulator is driven so as to seek a space in which the manipulator does not interfere with obstacles no matter what the manipulator is in, and to change the manipulator's attitude within this space, wherein the space is created by the manipulator. The setting is made so that it does not change depending on the position of the object being moved before the movement starts and after the movement ends.

(Function) If the manipulator trajectory generation method is as described above,
No matter what attitude the manipulator is in, the size of the space where it does not interfere with obstacles is smaller than before, but even when the manipulator moves an object, this object does not become an obstacle. Therefore, there is no need to recalculate the space each time an obstacle moves, so the amount of calculation is significantly reduced, and the trajectory of the manipulator to the target position can be reliably generated.

(Example) Hereinafter, the present invention will be explained using the drawings. FIG. 1 shows an example (simulation) in which the manipulator trajectory generation method of the present invention is applied to a large space manipulator. Here, the manipulator 1 has six degrees of freedom and is roughly divided into an arm section 2 and a hand section 3. The arm portion 2 has 30 degrees of freedom and is characterized by a wide range of motion due to the long distance between each joint. Hand part 3
It also has three degrees of freedom, but the distance between each joint is smaller than that of the arm section 2, and its operating range is narrow. Further, the hand section 3 can hold a payload, and various tasks in all directions can be performed by changing the orientation of the joints of the arm section 2, that is, by changing the posture of the hand section 1.

The manipulator 1 having such a configuration is fixed to the base body 4, and the manipulator 1 is automatically driven by control inside the base body 4. The six joints of the manipulator 1 are arranged in order from the connection part with the base body 4. These will be referred to as the 6th axis.

Next, we will discuss setting the safety area. The safe area can be set from an environment model created by inputting the positions of obstacles and the like in advance. The safe area here means that when the 4th axis is within the area, no matter what posture the 4th to 6th axes (that is, the hand part 3) take, there will be no obstacles (here, the base 4 or the structure). 5) is an area in three-dimensional real space that is guaranteed not to interfere with any obstacle.For example, the area 6 is a space at a certain distance from obstacles, such as the shaded area 6 in Figure 1. shall be called.

Such a safety area 6 can be set extremely easily from the environment model. In addition, in FIG. 1, the coordinate system in the figure is set as Z>3500 fins and X>3500 mm. As shown in the figure, the safe area 6 may be set in a part of the area that can be easily specified.

Further, when setting the safety area 6, it is necessary to determine the positions of the payload 7, which is moved by the hand unit 3, before the movement starts and after the movement ends.

This can be determined from the environment models before the start of movement of the payload 7 and after the end of movement of the payload 7, and may be set with enough margin to prevent interference with the payload 7. In addition, in consideration of the case where payload 7 is once moved to another location and then moved again, payload 7
It is desirable to create an environmental model for all situations in which movement occurs.

Further, the safety area 6 may be directly set by the operator via a monitor or the like, as long as the work environment is simple enough to ensure that the manipulator 1 does not interfere with obstacles by visual inspection.

After setting the safety area 6 in this way, a trajectory is determined so that the manipulator 1 is driven according to the following procedure.

I: Find the trajectory from the initial position a of the manipulator 1 to the safety area 6.

■: Obtain the trajectory from the target position of the manipulator 1 to the safety area 6.

Note that l may be a trajectory in which the manipulator 1 moves from its initial position a to any point in the safety area 6.

Similarly, in case (2), the trajectory may be such that the manipulator 1 moves from any point in the safety area 6 to the target position. Also,
In I and n, the trajectory from the first axis to the third axis is determined under the condition that the posture of the hand portion 3 of the manipulator 1 is kept constant. Trajectories are searched in configuration space.

nl: Create a trajectory connecting each point within the safety area 6 used in I and II.

In case (2), since the trajectory is within the safe area 6, the posture of the hand section 3 can be changed in any way, but if the trajectory is a straight line, each axis from the 4th to the 6th axis related to the posture of the hand section 3 is changed. The amount of change in joint angle was set as a trajectory that linearly captures the amount of change.

Incidentally, the procedure for generating the above trajectory is shown in FIG.

FIGS. 2 and 3 show the trajectory of moving the manipulator 1 from the initial position to the target position in the model shown in FIG. 1 using the method of the present invention (FIG. 2) and the conventional method (FIG. 3). This figure shows the simulation results obtained by Here, the conventional method is a method of searching in a configuration space in which the six joint axes of the manipulator 1 are used as coordinate axes. Table 1 below shows the amount of change in joint angle.

Table 1 Table 2 also shows the total number of spatial cells in which the interference between the manipulator 1 and an obstacle was checked in each case of the present invention and the conventional example. Here, each spatial cell was divided into a lattice shape using 5'' as a quantization unit.

Table 2 As described above, although the trajectory of the manipulator 1 is somewhat detoured, the amount of calculation is significantly reduced, and as a result, the working time of the manipulator is significantly reduced.

Also, 1. When searching as in (2) above, instead of searching the entire configuration space, a heuristic (heuristic) function is used to first check whether the cell is free space or not, starting with the cell with the smallest value. By doing this, the search range is limited. Here, the heuristic function h(c) is expressed as follows.

f (e) = g (c) + h (c) However, g: Distance traveled in the configuration space from the starting point to the current posture (θ ~ θ3) [degl h: (Current 4th axis Distance from the position to the safe area!y
)x3 [m] A decrease in the value of the evaluation function f (c) means that the distance to the target position decreases, and the trajectory search by the manipulator proceeds in the direction that approaches the target position as much as possible. This was done as it progressed. "3" above
is multiplied by weighting to prioritize h over g, for example, X increases by 3 and y decreases by 2 (
-2 increase), it becomes f--3, but X decreases by 2 (
-2 increase) If y3 increases, it becomes f-7, and the value of the evaluation function becomes sufficiently smaller as it approaches the target position.

Although g and h have different units, the multiplier "3" can be used here as an appropriate value that can adjust these units. Of course, other multipliers may also be used.

[Effects of the Invention] As described above, according to the present invention, the amount of calculation is significantly reduced;
Furthermore, a manipulator trajectory generation method that can reliably generate a manipulator trajectory to a target position is realized.

[Brief explanation of drawings]

FIG. 1 is a diagram showing an example of the safety area used in the present invention, FIGS. 2 and 3 are diagrams showing an example of the manipulator trajectory according to the method of the present invention and the conventional method, and FIG. 3 is a flowchart illustrating the method. 1... Manipulator, 4... Base (obstacle), 5
...Structure (obstacle), 6...Safety area, 7...
Payload (obstacle)

Claims (1)

[Scope of Claims] A manipulator having at least six joint axes and whose posture is determined by three joint axes on the distal end side is used, and a space is created in which the manipulator does not interfere with obstacles no matter what posture the manipulator is in. In the manipulator trajectory generation method, the manipulator is driven to change the attitude of the manipulator in this space, wherein the space is defined by the position of an object moved by the manipulator before the movement starts and after the movement ends. A manipulator trajectory generation method, characterized in that the manipulator trajectory is set so that it does not change.