JPH11254380A - Collision detecting method for industrial robot - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform quick and highly reliable collision detection by operating an industrial robot in a non-collision state and automatically setting a prescribed value by multiplying the maximum value of the collision component of external disturbance torque by a specified margin torque. SOLUTION: An industrial robot is actually operated according to an operation program, and during this period, the maximum value of the collision component of external disturbance torque is first initialized in a robot controller. Then, the collision component of the external disturbance torque is calculated based on a minimum dimensional observer and, if the collision component of the external disturbance torque is larger than the maximum value at this time, the collision component of the external disturbance torque is replaced by a new maximum value. Then, when a maximum peak value is obtained as a maximum value, a value obtained by multiplying the maximum value by a specified margin value is stored as a prescribed value for a collision detector 7. Thus, a proper prescribed value is uniquely selected for detecting the occurrence of collision different for each driving shaft, and a method for performing quick and highly reliable collision detection for the industrial robot is provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a collision of an industrial robot which can minimize the load when the arm or end effector constituting the industrial robot collides with an obstacle or the like.

[0002]

2. Description of the Related Art When an arm constituting an industrial robot or an end effector gripped by the arm collides with an obstacle, a drive shaft motor for driving each arm continues to rotate in accordance with a preset movement command. Trying to
As a result, the drive shaft motor is in a constrained state, and continues to generate a large torque. If this condition continues for a long time, the mechanism of the arm including the drive shaft motor and the reduction gear may be damaged.Therefore, the occurrence of a collision is detected by some method, and the movement command of the drive shaft motor is immediately interrupted. And other measures were taken.

For example, in the method disclosed in Japanese Patent Application Laid-Open No. 6-131050, a disturbance estimation observer estimates a disturbance torque in consideration of a friction torque, and a collision component of the disturbance torque calculated based on the estimated disturbance torque is calculated. When the value exceeds the specified value, it is determined that a collision or the like has occurred as a load abnormality. This method detects the occurrence of a collision by processing on software without using a special detector for collision detection, such as a force sensor, and shuts off the power supply to the drive shaft motor, thereby turning off the arm. An emergency stop can be made immediately.

[0004]

However, as for the above-mentioned specified value, the smaller the value is, the quicker the occurrence of a collision is detected. However, if the value is extremely small, the collision will occur even if no collision actually occurs. A situation arises in which it is erroneously determined that an error has occurred. In detail, FIG. 3 is a graph showing the time change of the collision component of the disturbance torque when no collision occurs. Even when no collision occurs, the robot model used by the disturbance estimation observer and the actual machine Since there are mechanical errors of the robot, errors due to the environment such as temperature changes, and errors in setting the hand and workpiece set by the operator, accurate disturbance torque cannot always be estimated. However, as shown in FIG. 3, the collision component of the disturbance torque does not become completely zero as shown in FIG.

[0005] For this reason, the time change of the collision component of the disturbance torque before and after the occurrence of the collision is shown in a graph shown in FIG. 4. ), A situation may occur in which the collision component of the disturbance torque exceeds a specified value. on the other hand,
If the specified value is too large, the collision component of the disturbance torque will not exceed the specified value even at the time of a collision, and in this case, occurrence of a collision will not be detected. Therefore, in order to perform quick and highly reliable collision detection, a method is required that can uniquely select an appropriate prescribed value.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and a rapid and highly reliable industrial robot can be obtained by uniquely selecting an appropriate prescribed value for detecting the occurrence of a collision. It is an object of the present invention to provide a collision detection method.

[0007]

In order to achieve the above object, the present invention has a structure in which a drive shaft motor for driving a joint is connected to an arm via a speed reducer, and an observer is used. By calculating the estimated disturbance torque received by the drive shaft motor, a known disturbance torque is subtracted from the estimated disturbance torque to calculate a collision component of the disturbance torque, and the collision component of the disturbance torque exceeds a predetermined value. In the collision detection method of an industrial robot in which the occurrence of a collision is detected when the collision occurs, the industrial robot is operated in a state in which no collision occurs, and the maximum value of the collision component of the disturbance torque at this time. And calculating the specified value automatically by multiplying the maximum value by a predetermined margin value. And it provided (claim 1).

In order not to detect the occurrence of a collision when no collision actually occurs, the specified value must always be larger than the collision component of the disturbance torque when no collision occurs. Therefore, the industrial robot is operated in a state where no collision occurs, and the maximum value is extracted from the collision component of the disturbance torque calculated based on the disturbance torque at this time. A value calculated by multiplying the maximum value of the collision component of the disturbance torque in a state where no collision occurs by a predetermined margin value is selected as a specified value.
This makes it possible to automatically set different specified values for each drive axis. The margin value should be considered so as not to detect the occurrence of a collision when no collision actually occurs, and when the collision component of the disturbance torque at the time of the collision is small, such as when colliding with an obstacle at a low speed or when the When an object is an elastic body, it is necessary to select an object in consideration of the fact that the occurrence of a collision can be reliably detected.

The applicant of the present invention has conducted experiments on the operation of an industrial robot under various conditions, and found that if the margin value is set within a range of 1.5 or more and 2.5 or less, a collision actually occurs. It has been found that the occurrence of a collision is not detected in a state where no collision occurs, and that the occurrence of a collision can be reliably detected even when the collision component of the disturbance torque at the time of the collision is small.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a servo system constituting a drive shaft control device according to an embodiment of the present invention. In the figure, 1 is a servomotor as a drive shaft motor for driving the drive shaft of each arm of the industrial robot, 2 is an encoder as a position detector for detecting the position of the drive shaft, and 3 is an amplifier as an amplifier. It is a power amplifier. The servo loop is composed of a triple loop of a current control device 4, a speed control device 5, and a position control device 6, from the inside. Reference numeral 10 denotes a command position generator that outputs a command position of each drive shaft, which is obtained every moment, for the industrial robot to operate properly based on the target position, starting position, required speed, acceleration limit value, and the like. . Reference numeral 7 denotes a collision detection device to which the collision detection method of the present invention is applied and which detects when the arm 23 or an end effector (not shown) gripped by the arm 23 collides with an obstacle (not shown). 8
Is a position loop gain changing device that changes the position loop gain when a collision is detected. Reference numeral 9 denotes a command position changing device that changes the command position when a collision is detected. On the other hand, FIG. 2 is a conceptual diagram of a spring-mass system showing a relationship between the servomotor 1, the rotational speed reducer 22, and the arm 23.

The function of this servo system will be described.
The position control device 6 includes a momentary command position of the servo motor 1 according to an operation program or the like generated by the command position generation device 10 and a position feedback as a current position read from the encoder 2 attached to the servo motor 1. The position command 14 is obtained by calculating the difference from the position command 11, and outputs a speed command 15 obtained by multiplying the position difference 14 by a position loop gain. The speed control device 5
The difference between the speed command 15 output from the position control device 6 and the speed feedback 12 obtained by differentiating the current position read from the encoder 2 with the differentiator S is taken as a speed deviation. The current command 16 is output based on the current command. Current control device 4
Calculates the difference between the current command 16 output from the speed control device 5 and the current feedback 13 as an actual current flowing to the servomotor 1 detected by the current detector 17, and based on this, the motor is fed to the servomotor 1. It is configured to output a current.

The collision detection device 7 is a device to which the collision detection method of the present invention is applied, estimates a disturbance torque by a disturbance estimation observer, and calculates a collision component of the disturbance torque calculated based on the estimated disturbance torque. Is greater than a specified value, it is determined that a collision has occurred. At the time of collision, the servo loop tries to output a larger torque (current command 16) to the servomotor 1 than usual, but since the position of the servomotor 1 hardly changes due to the collision, the value of the speed feedback 12 is It is almost zero. Therefore, the speed control device 5
Command 16 and speed feedback 12 output from
Is monitored, the amount of torsion is calculated based on these values, and the amount of torsion is converted into a disturbance torque applied to the arm 23.
From this converted value, by subtracting a known disturbance torque such as a so-called shaft interference torque generated by unbalance torque generated by gravity or centrifugal force, Coriolis force, or inertia force generated by the movement of another drive shaft, The increase due to the collision of the disturbance torque, that is, the collision component of the disturbance torque can be calculated.

FIG. 4 is a graph showing the timing of collision detection in the present embodiment, and shows the time change of the collision component of the disturbance torque. The horizontal axis is time, and the vertical axis is the collision component of the disturbance torque. As described above, the collision component of the disturbance torque is calculated by using the observer to calculate the estimated disturbance torque received by the drive shaft motor. From the estimated disturbance torque, the unbalance torque generated by gravity and the motion of other drive shafts are calculated. By subtracting a known disturbance torque such as a so-called shaft interference torque originating from a centrifugal force, a Coriolis force, an inertial force, or the like generated as a result, an increase in the disturbance torque due to collision can be obtained. T1 is the collision occurrence time, and T2
Is the time when the collision component of the disturbance torque as the collision detection time exceeds a specified value.

Here, an embodiment of a method for automatically setting the specified value will be described with reference to a flowchart shown in FIG. The industrial robot is operated in accordance with the actual operation program, and during that time, the processing specified by the flowchart shown in FIG. 5 is performed in the robot controller. First, the maximum value _Tmax of the collision component of the disturbance torque is initialized (step 31). Next, the collision component T of the disturbance torque is calculated based on the minimum dimension observer (step 32). If the collision component T of the disturbance torque is larger than the maximum value _Tmax at this time (step 33Y), the collision component T of this disturbance torque is calculated. Collision component T
_Is replaced with a new maximum value _Tmax (step 34).
When the collision component T of the disturbance torque is smaller than the maximum value _Tmax at this time (step 33N), the process directly proceeds to step 35. Then, in step 35, if the operation program has not been completed, the processing after step 32 is executed again. On the other hand, if the operation program has been completed, the process proceeds to step 36.

Finally, in step 36, it is determined that the maximum value _Tmax at this time is the maximum value in this operation program. That is, FIG. 6 is a graph showing the relationship between the time change of the collision component of the disturbance torque and the maximum value _Tmax when no collision occurs, but by performing the processing shown in the flowchart of FIG. Is obtained as the maximum value _Tmax . Then, a value obtained by multiplying the maximum value _Tmax by a predetermined margin value is stored as a specified value.

Here, the margin value is a numerical value of 1 or more, and is set in consideration of a certain allowable rate. That is, the margin value is considered so as not to detect the occurrence of a collision when no collision actually occurs, and
When the collision component of the disturbance torque at the time of the collision is small, for example, when the vehicle collides with an obstacle at a low speed, or when the obstacle is an elastic body, the selection is made in consideration of reliably detecting the occurrence of the collision. When the applicant of the present application performed an operation experiment of an industrial robot under various conditions, if the margin value was set within a range of 1.5 or more and 2.5 or less, a collision actually occurred. It has been found that the occurrence of a collision is not detected in a state where no collision occurs, and that the occurrence of a collision can be reliably detected even when the collision component of the disturbance torque at the time of the collision is small.

FIG. 7 is a graph showing the relationship between the time change of the collision component of the disturbance torque and the specified value obtained by multiplying the maximum value _Tmax by a predetermined margin value when a collision occurs. By setting an appropriate margin value, the collision occurrence time T1 can be ensured while ensuring the reliability of collision detection.
And the time lag between the collision detection time T2, that is, the time lag, can be reduced.

When a collision is detected by the collision detection device 7, the command position change device 9 inputs the current position of the servo motor 1 from the encoder 2 and uses this current position as a command position at the time of collision. Output to 10 As described above, the command position generating device 10 normally generates the command position of the servo motor 1 every moment according to the operation program or the like, but the current position of the servo motor 1 input from the command position changing device 9 is detected when a collision is detected. The position is output to the position control device 6 as a command position. As a result, at the time of a collision, the position deviation 14, which is the difference between the command position and the current position, becomes zero, and the speed command 15 obtained by multiplying the position deviation by the position loop gain also becomes zero. Is generated, the operation of the arm 23 is immediately stopped. Therefore, the penetration of the arm 23 or the end effector (not shown) into an obstacle is minimized.

When the collision detecting device 7 detects a collision, the position loop gain changing device 8 outputs a position loop gain setting value to the position control device 6 at the time of collision.
The position control device 6 receives the collision set value and changes the initial position loop gain value stored in the position control device 6 to the collision set value. The position loop gain is a proportionality constant used when calculating the speed command 15 from the position deviation 14. The larger this value is, the higher the stiffness of the drive shaft is, and the better the followability of the drive shaft to the command position is. Conversely, the load on the servomotor 1 and the reduction gear 22 increases. While the arm 23 is in operation, it is desirable to increase the position loop gain in order to improve the followability of the drive shaft. , The life of the reduction gear 22 is shortened, or in the worst case, the drive system including the reduction gear 22 is damaged.

Therefore, in the event of a collision, the speed command is reduced by changing the position loop gain from the initial set value during arm operation to a smaller set value during the collision, thereby reducing the rigidity of the drive shaft. As a result, the arm 23 that has collided with the obstacle is naturally pulled back to the collision position by the restoring force, so that the biting state is eliminated and the load on the reduction gear 22 is reduced.

The set value of the position loop gain at the time of collision needs to be small enough that the arm 23 that has collided with the obstacle is naturally pulled back to the collision position by the restoring force.
If it is made extremely small, the weight of the arm itself will not be able to counter gravity and, in the worst case, the arm 23 will fall. For this reason, it is necessary to compensate at least the influence of gravity on the set value of the position loop gain at the time of collision. With this in mind,
Although the predetermined value at the time of collision of the position loop gain may be defined in advance, a predetermined ratio is defined in advance,
At the time of a collision, the position loop gain may be changed by multiplying the initial value of the position loop gain by this predetermined ratio.

Specifically, by giving a predetermined load to the end effector and actually generating a collision state,
The collision set value is determined experimentally so that the arm 23 is naturally pulled back to the collision position by the restoring force, and the arm 23 is not dropped by the influence of gravity. Alternatively, if a plurality of data are obtained by repeating the same experiment, it is possible to obtain the ratio of the set value at the time of collision to the initial set value. Therefore, at the time of a collision, multiply the initial value of the position loop gain by this ratio. Thus, the position loop gain is changed.

The embodiment of the present invention has been described above. Although the above embodiment describes the case where the present invention is applied to the drive shaft of an industrial robot, the drive shaft motor that drives the joint operates the arm or a member corresponding thereto via a speed reducer. As long as it is a form, the technology of the present invention can be easily applied to a machine other than the industrial robot, and can be applied to, for example, a machine tool using a servomotor and a speed reducer for a drive shaft.

[0024]

According to the present invention, a drive shaft motor for driving a joint has a structure connected to an arm via a speed reducer, and an estimated disturbance torque received by the drive shaft motor is calculated by using an observer. Then, a collision component of the disturbance torque is calculated by subtracting the known disturbance torque from the estimated disturbance torque, and the occurrence of a collision is detected when the collision component of the disturbance torque exceeds a predetermined value. In the collision detection method for an industrial robot, the industrial robot is operated in a state where no collision occurs, a maximum value of a collision component of a disturbance torque at this time is calculated, and a predetermined margin is set to the maximum value. Since the specified value is automatically set by multiplying the value, an appropriate specified value for detecting occurrence of a different collision is uniquely selected for each drive shaft. Theft can now. Therefore, it has become possible to provide a method for quickly and reliably detecting a collision of an industrial robot.

In particular, if the margin value is set within the range of 1.5 or more and 2.5 or less as in the invention according to the second aspect, it is possible to prevent the collision in a state where no collision actually occurs. The occurrence is not detected, and the occurrence of the collision can be reliably detected even when the collision component of the disturbance torque at the time of the collision is small.

[Brief description of the drawings]

FIG. 1 is a block diagram of a servo system of an industrial robot to which an industrial robot collision detection method according to the present invention is applied.

FIG. 2 is a conceptual diagram of a spring-mass system, showing a relationship among a servomotor 1, a rotation speed reducer 22, and an arm 23.

FIG. 3 is a graph showing a time change of a collision component of a disturbance torque when no collision occurs.

FIG. 4 is a graph showing a time change of a collision component of a disturbance torque before and after a collision occurs.

FIG. 5 is a flowchart showing one embodiment of a method for automatically setting a specified value in the present invention.

FIG. 6 is a graph showing a relationship between a time change of a collision component of a disturbance torque and a maximum value T _max when a collision does not occur in the present invention.

FIG. 7 is a graph showing a relationship between a temporal change of a collision component of a disturbance torque and a specified value obtained by multiplying a maximum value T _max by a predetermined margin value when a collision occurs in the present invention. .

[Explanation of symbols]

 Reference Signs List 1 drive shaft motor (servo motor) 2 position detector (encoder) 3 power amplifier 4 current control device 5 speed control device 6 position control device 7 collision detection device 8 position loop gain change device 9 command position change device 10 command position generation device DESCRIPTION OF SYMBOLS 11 Position feedback 12 Speed feedback 13 Current feedback 14 Position deviation 15 Speed command 16 Current command 17 Current detector 22 Reduction gear 23 Arm

Claims (2)

[Claims] A drive shaft motor for driving a joint has a structure connected to an arm via a speed reducer, and an estimated disturbance torque received by the drive shaft motor is calculated by using an observer. The collision component of the disturbance torque is calculated by subtracting the known disturbance torque from the torque, and when the collision component of the disturbance torque exceeds a predetermined value, it is determined that a collision has been detected. In the collision detection method for an industrial robot, the industrial robot is operated in a state where no collision occurs, and a maximum value of a collision component of the disturbance torque at this time is calculated,
A method for detecting a collision of an industrial robot, wherein the specified value is automatically set by multiplying the maximum value by a predetermined margin value. 2. The collision detection method for an industrial robot according to claim 1, wherein said margin value is set within a range of 1.5 or more and 2.5 or less.