JPH03139704A - Method and device for obstacle evasion control over robot - Google Patents

Abstract

PURPOSE:To enable the robot to easily perform operation for obstacle evasion by adjusting the quantity of displacement for attitude matching and the quantity of displacement for obstacle evasion while judging the current attitude of the robot and whether or not the distance to the obstacle is long or short by a fuzzy rule. CONSTITUTION:The articulated robot 1 is equipped with a 1st arm 3 which is connected to a 1st joint 2 and a 2nd joint 5 which connects the 1st arm 3 and a 2nd arm 4 and the 1st joint 2 and 2nd joint 5 are both rotatable in a plane 6, so the articulated robot 1 has two degrees of freedom. The robot finds the distance to the obstacle by itself and employes the fuzzy rule based upon this distance as a condition, so the robot has a proper displacement quantity autonomously by combining attitude matching operation to a target attitude with obstacle evading operation and takes an attitude matching the target attitude eventually. Consequently, the robot can evade obstacles impeding its movement in order when put in the attitude matching operation.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) The present invention eliminates the need to set a movement route in advance even when there are obstacles in the robot's working environment, and allows the robot to easily The present invention relates to an obstacle avoidance control method for a robot and an obstacle avoidance control device for a robot that can perform obstacle avoidance operations.

(Prior art) When disassembling, inspecting, and assembling various equipment using a work robot, there may be obstacles in the work environment such as equipment or equipment other than the work target, such as an articulated robot. If so, the robot itself may collide with these obstacles.

Conventionally, the robot operator has either set a travel route that takes obstacles into account and inputs it into the robot, or inputs a detailed map of the working environment into the robot.
A method was used in which the robot divided this into several areas and automatically set the movement route (path planning).

(Problems to be Solved by the Invention) However, it requires a great deal of effort for the operator to set a travel route while taking into account the positions and sizes of obstacles.

Furthermore, even when providing information about the work environment to a robot, it is similarly difficult to create a map that shows the location and size of obstacles in detail, and the placement of obstacles is complicated for the robot as well. When this happens, there are problems such as the time required for path planning.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a robot obstacle avoidance control method that allows the robot to easily perform obstacle avoidance operations in a work environment, and a device used in this method. With the goal.

[Structure of the invention]

(Means for Solving the Problems) In order to solve the above problems, the present invention provides (1) a step of inputting a target posture of the robot, and (2) a step of measuring the distance between the current posture of the robot and an obstacle. and (3) a displacement amount for posture adjustment to match the robot's current posture to the target posture, and a displacement amount for avoiding obstacles when moving the robot from the current posture to match the target posture. and (4) adjusting the displacement amount for posture adjustment and the displacement amount for obstacle avoidance while determining the length of the distance between the current posture of the robot and the obstacle using fuzzy rules. The present invention provides a method for controlling obstacle avoidance of a robot, which comprises a step of repeating the steps (2) to 4) until the robot's posture matches the target posture.

The present invention also provides an input device for inputting a target posture of the robot, a distance measuring device for measuring the distance between the current posture of the robot and an obstacle, and a posture adjustment device for adjusting the posture of the robot to match the current posture to the target posture. The amount of displacement and the amount of displacement for avoiding obstacles when moving the robot from the current posture to match the target posture are calculated, and the distance between the current posture of the robot and the obstacle is calculated using fuzzy technology. The present invention also provides an obstacle avoidance control device for a robot, which includes an arithmetic processing unit that adjusts the displacement amount for posture adjustment and the displacement amount for obstacle avoidance while making a judgment based on rules.

(Operation) In the robot obstacle avoidance control method according to the present invention, for example, a target posture of the robot is inputted into the input device of the above-mentioned robot obstacle avoidance control device, and then a distance measuring device is used to compare the current posture of the robot and the obstacle. Measure the distance to.

Then, the arithmetic processing unit calculates the amount of displacement for posture adjustment to match the robot's current posture to the target posture, and the displacement amount for avoiding obstacles when moving the robot from the current posture to match the target posture. The amount of displacement is calculated. The arithmetic processing unit further adjusts the displacement amount for posture adjustment and the displacement amount for obstacle avoidance while determining the length of the distance between the current posture of the robot and the obstacle using a fuzzy rule. After this, operations (2) to 4) are repeated until the robot's posture matches the target posture.

In this way, in the present invention, the robot itself calculates the distance to the obstacle and incorporates a fuzzy rule that uses the distance between parentheses as a condition, so that the robot can autonomously adjust its posture to the target posture and By taking an appropriate amount of displacement in combination with avoidance movements, it is possible to finally take a posture that matches the target posture. Furthermore, the present invention eliminates the need for path planning in advance because the system sequentially avoids obstacles that impede movement when performing posture adjustment operations.

(Example) The present invention will be described in detail below with reference to the accompanying drawings.

Figure 1 shows a two-arm, two-arm system to which the method and device of the present invention are applied.
FIG. 3 is a diagram showing changes in posture of a jointed articulated robot until it avoids an obstacle and unites with a target posture.

The articulated robot 1 has a first joint connected to a first joint 2.
It has an arm (length (1) 3) and a second joint 5 that connects the first arm 3 and the second arm (length 12) 4.The first joint 2 and the second joint 5 both have a flat surface 6. Since it can rotate within the
It has several degrees of freedom.

This articulated robot 1 rotates around the first arm 3 from an initial posture in which the first arm 3 and the second arm 4 extend in a straight line within a plane 6 represented by two axes, X and Y. Similarly, the first arm 3 and the second arm 4 extend in a straight line, and the tip coordinates of the second arm 4 (-1, 19, 0, 60
), but if the first arm 2 and the second arm 4 were to rotate counterclockwise around the first joint while remaining in a straight line, the obstacle 7 collide with

Therefore, in this embodiment, the articulated robot 1
This obstacle 7 according to the algorithm shown in the flowchart of fig.
Rotate counterclockwise to reach the target position while avoiding.

In this algorithm, first, the target posture, in this case the coordinates of the tip of the second arm 4 (-1, 19, 0°60), is input into the input device of the robot's obstacle avoidance control device.

Then, the robot's obstacle avoidance control bag and the distance measuring device located on the first arm 2. The distance d between the second arm 4 and the obstacle 7 is measured.

Next, in the arithmetic processing unit of the robot's obstacle avoidance control device, the displacement amount (rotation angle) Δθ of the first joint 2 and the second joint 5 for obstacle avoidance, (j is the j-th joint av, J) is calculated.

In this embodiment, first, the displacement amount Δ of the second joint 5 is
θ8,2 is determined, and then the displacement amount Δθ8,1 of the first joint 2 is determined by the relational expression cosΔθ − (ji
+Acosθay 2)/avl 1
2'! 1 +j! 2 +2j! 1j! 2 cosθa
y, 2. ]/2 2 Therefore, the displacement amount Δθ39,2 of the second joint 5 is determined in the opposite direction. Therefore, if the length of the first joint 2 (and the length 12 of the second joint 5 are equal), Δθav, 1 = 1 / 2Δθ8
v, becomes 2.

On the other hand, in this embodiment, in the arithmetic processing unit of the obstacle avoidance control device of the robot, the amount of displacement (rotation angle) Δθ (j is the j-th ad, indicating that it is the i-th joint) is calculated.

Setting of this attitude alignment control can be configured by proportional integral (PI) control.

That is, first, for each joint, the deviation from the target attitude Ej (Ej - θ□G, j - θj; θ□G, i is the target angle of the j-th joint, and θj is the current angle of the j-th joint. ), and after setting the displacement ΔE of this deviation E, each joy J
The displacement amount Δθ for each point is expressed as Δθad, i”Ki・
ad, j E, +K −ΔE, determined according to (K, and KJl)
J
11) are the integral gain and proportional gain, respectively)
.

The posture adjustment control settings are based on proportional integral derivative (PID)
It can also be configured with control.

After setting the obstacle avoidance control and attitude adjustment control in this way, the next step is to determine how to perform the obstacle avoidance control and attitude adjustment control according to the following fuzzy rules (Rule 1) in the arithmetic processing unit of the robot's obstacle avoidance control device. In other words, it is determined how to adjust Δθ and Δθad,i as the displacement amount Δθ of the j-th joint from the current position.

Rule 1: If d is SMALL.

THEN Δθ, = Δθ.

J av, I ELSE Δθ・=Δθad, j (If d is SMALL, 80, = to 0,; otherwise Δθ, −ΔθIld, i) J
aV,I
JHere, the satisfaction degree DOFs of SMALL (fuzzy label) in the conditional clause (clause 11) (0≦D.

Fs≦1), membership function tt Small (d
) is calculated by (DOFs=μSmall (d)
), and the degree of satisfaction DOFe (0≦DOFe≦1) of ELSE (not SMALL) is determined by the expression DOF by the membership function p NOT Small (d).
e = μNOT Small (d) =1. -D
Obtained by OFs.

FIG. 3 shows the membership function t for the size of the distance d.
t Small (d) and p NOT Small (
The shape of each of d) is shown.

After determining DOFs and DOFe in this way, use them to calculate Δθ, Δθ86. j is weighted according to the following formula (I)vg, and Δθ31. j and Δθad, j
Find the amount of displacement Δθ, which is reasonably adjusted.

Δθ, = DOFS・80, +DOFe・ΔθJ
aV, 18.1j...
(1) Figures 4 (A) and (B) show the articulated robot 1 of this embodiment.
Figures 5 (A) and 5 (B) show the changes in posture when simulating obstacle 7 avoidance for FIG. 7 is a diagram illustrating a change in posture when the alignment control and obstacle avoidance operation are switched and unified with the target posture.

In Figure 4 (A) and Figure 5 (A), the distance on the vertical axis is 4
．． It is normalized based on 00. Further, the time on the horizontal axis is a unit of steps, which are equal divisions of a certain period of time. Furthermore, Figure 4 (
In A) and Fig. 5(A), curves 11, 12, and 13 represent the angle change of the first arm, the angle change of the second arm (+ is shown in reverse), and the obstacle, respectively. In Figure 4 (B) and Figure 5 (B), curve 2
1.22 and 23.24 are the target posture of the first arm, the angle of the first arm, the target posture of the second arm (indicating the deviation from the first arm), and the angle of the second arm (indicating the deviation from the first arm), respectively. ).

As can be seen from these figures, when fuzzy rules are not used, posture adjustment control and obstacle avoidance operation are alternatively selected based on a certain threshold value based on the distance to the obstacle. Therefore, at first, a posture adjustment action is taken in the step where the threshold is maintained, but when the distance to the obstacle decreases beyond this threshold, an obstacle avoidance action is performed to maintain the threshold. Take. However, in the next step, since the threshold value is maintained, posture adjustment control is performed instead. As a result, the distance to the obstacle is significantly reduced beyond the threshold value.

In the following, two types of control are repeated alternately in the same way, and finally the obstacle is avoided and the target attitude is achieved. During this time, the first arm and the second arm repeat expansion and contraction in response to the switch between the obstacle avoidance operation and the attitude adjustment operation.

On the other hand, in this embodiment, based on fuzzy rules,
Because the arm moves by weighting the displacement for posture adjustment and the displacement for obstacle avoidance, the arm bends to the maximum at the point of closest contact with the obstacle, and after avoiding the obstacle, it moves toward the target. The arm gradually extends towards the posture. Therefore, for one obstacle, the first arm and the second arm only need to be expanded and contracted once, allowing smooth movement as a whole.

〔Effect of the invention〕

As explained above, the robot obstacle avoidance control method according to the present invention includes (1) a step of inputting a target posture of the robot, and (2) a step of measuring the distance between the current posture of the robot and the obstacle. , (3) a displacement amount for posture adjustment to match the robot's current posture to the target posture, and a displacement amount for avoiding obstacles when moving the robot from the current posture to match the target posture. (4)
) a step of adjusting the displacement amount for posture adjustment and the displacement amount for obstacle avoidance while determining the length of the distance between the robot's current posture and the obstacle using fuzzy rules; The above (2) is repeated until the posture matches the target posture.
) or 4) are repeated.

Therefore, according to the present invention, the robot can capture an obstacle between the target posture and the obstacle without the operator setting a movement route in advance or preparing a detailed working environment map. In order to achieve an appropriate amount of displacement by combining posture adjustment to the target posture and obstacle avoidance while the robot judges the distance from The difficult work of preparation becomes unnecessary. Furthermore, in the method of the present invention, when performing posture adjustment operations, obstacles that impede movement are sequentially avoided, so there is no need for path planning in advance.

The obstacle avoidance control device for a robot according to the present invention includes an input device for inputting a target posture of the robot, a distance measuring device for measuring the distance between the current posture of the robot and the obstacle, and a distance measuring device for measuring the distance between the robot's current posture and the obstacle. Calculate the amount of displacement for posture adjustment to match the posture and the amount of displacement for avoiding obstacles when moving the robot from the current posture toward matching the target posture, and The above-mentioned method is realized by comprising an arithmetic processing unit that adjusts the displacement amount for posture adjustment and the displacement amount for obstacle avoidance while determining the length of the distance to the obstacle using fuzzy rules. be able to.

Fig. 1 is a schematic diagram showing the posture change of an articulated robot according to an embodiment of the method of the present invention, Fig. 2 is a flowchart of the algorithm in the embodiment of Fig. 3, and Fig. 3 is a diagram showing the distance to an obstacle. Figure 4 (A) and (
B) and FIGS. 5(A) and 5(B) are graphs showing posture changes when an obstacle is avoided using the fuzzy rules of the present invention and when an obstacle is avoided without using the fuzzy rules, respectively. be.

1... Multi-jointed robot, 2... First joint, 3
...First arm, 4...Second arm, 5...Second arm
Joint, 7...obstacle.

Claims (1)

[Claims] 1. (1) inputting a target posture of the robot;
) Measuring the distance between the robot's current posture and the obstacle; (3) the amount of displacement for posture adjustment to match the robot's current posture to the target posture; and (3) the process of adjusting the displacement amount to match the robot's current posture to the target posture. (4) calculating the amount of displacement for avoiding obstacles when moving toward the robot; and (4) determining the length of the distance between the robot's current posture and the obstacle using fuzzy rules while adjusting the posture. A step of adjusting the amount of displacement and the amount of displacement for obstacle avoidance, and performing the steps (2) or (2) above until the robot's posture matches the target posture.
A robot obstacle avoidance control method that repeats the process of 4). 2. An input device for inputting a target posture of the robot, a distance measuring device for measuring the distance between the current posture of the robot and an obstacle, and an amount of displacement for posture alignment to match the robot's current posture to the target posture. , calculate the displacement amount for avoiding obstacles when moving the robot from the current posture to match the target posture, and determine the length of the distance between the robot's current posture and the obstacle using fuzzy rules. while doing,
An obstacle avoidance control device for a robot, comprising: an arithmetic processing unit that adjusts the displacement amount for posture adjustment and the displacement amount for obstacle avoidance.