Online setting method for industrial robot controller parameters

Abstract

The invention discloses an online setting method for industrial robot controller parameters, which comprises the following steps: building a rigid body dynamic model of a multi-joint series industrial robot on the premise of not considering the uncertainty item of the robot and the interference of friction torque; substituting the actual rotation angle of the driving joint into a rigid body dynamic model of the robot to obtain an expected input torque of the driving joint, taking an absolute value of the expected input torque as input, and obtaining a parameter initial value of a feedback controller through a one-dimensional fuzzy algorithm; the rotation angle error of the driving joint and the error change rate serve as input, and the parameter increment of a feedback controller is obtained through a two-dimensional fuzzy algorithm module; a final feedback controller parameter is obtained through calculation of the parameter initial value and the parameter increment, and online setting of the feedback controller parameter is achieved; the operation stability of the multi-joint series industrial robot in the whole joint space is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot automatic control, in particular to an online tuning method for industrial robot controller parameters. Background Art

[0002] Multi-joint serial industrial robots typically consist of a series of joint links connected in series from the base to the end effector, with each joint driven by a servo motor. The feedback controller for the drive joints of multi-joint serial industrial robots generally uses a PID controller, which features a simple structure, high reliability, and strong anti-interference capabilities. Multi-joint serial industrial robots are nonlinear, multi-input, multi-output, time-varying systems, and the joint torques vary significantly with joint configuration. Using a PID controller with fixed parameters makes it difficult to ensure the stability of the drive joints throughout the joint space and the consistency of control quality. Fuzzy PID incorporates the technician's experience into the fuzzy rule table, which can support real-time operation during robot operation. Therefore, there is an urgent need for an online tuning method for the controller parameters of multi-joint serial industrial robots. This method combines fuzzy PID with the robot's rigid body dynamics, allowing the multi-joint serial industrial robot to adapt to changes in joint torque with joint configuration, thereby improving the operational stability of the multi-joint serial industrial robot throughout the joint space. Summary of the Invention

[0003] In order to solve the problems existing in the prior art, the purpose of the present invention is to provide an online tuning method for controller parameters of a multi-joint serial industrial robot that combines fuzzy PID with robot rigid body dynamics to adapt to changes in joint torque with joint configuration.

[0004] To achieve the above object, the technical solution adopted by the present invention is: an online tuning method for industrial robot controller parameters, comprising the following steps:

[0005] Step 1: Build a rigid body dynamics model of a multi-joint serial industrial robot without considering the robot uncertainty term and the interference of friction torque;

[0006] Step 2: Substitute the actual rotation angle of the driving joint into the rigid body dynamics model of the robot to obtain the expected input torque of the driving joint. The absolute value of the expected input torque is used as input to obtain the initial value of the feedback controller parameters through a one-dimensional fuzzy algorithm;

[0007] Step 3: Using the rotation angle error and error change rate of the driving joint as input, the parameter increment of the feedback controller is obtained through the two-dimensional fuzzy algorithm module;

[0008] Step 4: The final feedback controller parameters are obtained by calculating the initial parameter values and the parameter increments, thereby realizing online tuning of the feedback controller parameters.

[0009] As a further improvement of the present invention, in step 1, the rigid body dynamics model is specifically as follows:

[0010]

[0011] Where, θ=(θ1 θ2 … θ n )∈ n , are the angle, angular velocity and angular acceleration vector of each driving joint, respectively, M(θ)∈ n×n is the symmetric inertia matrix, is the Coriolis vector, G(θ) is the gravitational moment vector, τ=(τ1 τ2 …τ n )∈ n is the input torque vector of each driven joint.

[0012] As a further improvement of the present invention, the step 2 is specifically as follows:

[0013] During the real-time operation of the robot, the actual angle of the motor fed back by the encoder is converted into the actual rotation angle θ of the driving joint. a,i The actual rotation angle of the driving joint is substituted into the dynamic model of the robot to obtain the expected input torque of the driving joint. The absolute value of the expected input torque is used as input to obtain the initial value of the feedback controller parameter ρ through the one-dimensional fuzzy algorithm. Oi =(K PO,i K IO,i K DO,i ) T (i=1,2,…,n), where i is the i-th driven joint, K PO,i , K IO,i and K DO,i are the initial proportional coefficient, initial integral coefficient and initial differential coefficient of the feedback controller of the i-th driven joint respectively.

[0014] As a further improvement of the present invention, the one-dimensional fuzzy algorithm implementation process is specifically as follows:

[0015] The absolute value of the expected input torque of the driving joint is used as input, and the initial value of the feedback controller parameter is used as output. The input and output values are divided into multiple regions. The absolute value of the expected input torque is substituted into the triangular membership function to obtain the membership of each region. The output is obtained by searching the one-dimensional fuzzy rule table based on the membership of each region.

[0016] As a further improvement of the present invention, the step 3 is specifically as follows:

[0017] The actual angle of the motor fed back by the encoder is converted into the actual rotation angle θ of the driven joint a,i , minus the desired rotation angle θ at this timed,i Get the rotation angle error e of each driving joint i , for e i Perform differential operation to obtain the error change rate ec i , to drive the joint’s rotation error e i and error change rate ec i As input, the parameter increment Δρ of the feedback controller is obtained through the two-dimensional fuzzy algorithm module i =(ΔK P,i ΔK I,i ΔK D,i ) T (i=1,2,…,n), where i is the i-th driven joint, ΔK P,i , ΔK I,i and ΔK D,i are the proportional coefficient increment, integral coefficient increment and differential coefficient increment of the feedback controller of the i-th driven joint respectively.

[0018] As a further improvement of the present invention, the two-dimensional fuzzy algorithm is implemented as follows:

[0019] The angular error e of the driven joint i and error change rate ec i As input, the parameter increment of the feedback controller is output, the input and output values are divided into multiple regions, and the angular error e of the driven joint is i and error change rate ec i Substitute the triangle membership function to obtain the membership of each area, and search the two-dimensional fuzzy rule table according to the membership of each area to obtain the output.

[0020] As a further improvement of the present invention, in step 4, by formula ρ i =ρ Oi +Δρ i The final feedback controller parameters are calculated.

[0021] This invention employs a "1+2" dimensional fuzzy control strategy. Using the absolute value of the desired torque calculated from the robot's rigid-body dynamics model as input, a one-dimensional fuzzy algorithm is used to obtain the initial values of the feedback controller's parameters. Furthermore, using the errors and their rates of change of each driven joint as input, a two-dimensional fuzzy algorithm is used to obtain the incremental values of the feedback controller's parameters. Finally, the two are summed to obtain the final feedback controller parameters, enabling online tuning of the feedback controller parameters. Compared to existing methods, this method uses real-time fuzzy control to enable industrial robots to adapt to changes in joint torque as joint configuration changes, improving the robot's operational stability throughout the entire joint space.

[0022] The beneficial effects of the present invention are:

[0023] The present invention does not require real-time identification of the rigid body dynamics model parameters of the driving joints. Instead, it utilizes an ideal rigid body dynamics model combined with a traditional two-dimensional fuzzy PID controller to enable the multi-joint serial industrial robot to adapt to changes in joint torque with joint configuration, thereby improving the operational stability of the multi-joint serial industrial robot in the entire joint space. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A strategy diagram of an embodiment of the present invention;

[0025] Figure 2 Graph showing a triangle membership function of a one-dimensional fuzzy algorithm module according to an embodiment of the present invention;

[0026] Figure 3 This is a triangle membership function curve diagram of the two-dimensional fuzzy algorithm module in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0028] Example

[0029] like Figure 1 As shown, an online tuning method for controller parameters of a multi-joint serial industrial robot includes the following steps:

[0030] (1) Build a rigid body dynamics model of a multi-joint serial industrial robot, ignoring the robot uncertainty term and the interference of friction torque;

[0031]

[0032] Where, θ=(θ1 θ2 … θ n )∈ n , are the angle, angular velocity and angular acceleration vector of each driving joint, respectively, M(θ)∈ n×n is the symmetric inertia matrix, is the Coriolis vector, G(θ) is the gravitational moment vector, τ=(τ1 τ2 …τ n )∈ n is the input torque vector of each driven joint.

[0033] (2) During the real-time operation of the robot, the actual angle of the motor fed back by the encoder is converted into the actual rotation angle θ of the driving joint. a,i , and substitute the actual rotation angle of the driving joint into the robot dynamics model of step 1 to obtain the expected input torque of the driving joint. The absolute value of the expected input torque is used as input to obtain the initial value of the feedback controller parameter ρ through the one-dimensional fuzzy algorithm Oi =(KPO,i K IO,i K DO,i ) T (i=1,2,…,n), i is the i-th driving joint, K PO,i , K IO,i and K DO,i are the initial proportional coefficient, initial integral coefficient and initial differential coefficient of the feedback controller of the i-th driven joint respectively;

[0034] (3) The actual angle of the motor fed back by the encoder is converted into the actual rotation angle θ of the driving joint a,i , minus the desired rotation angle θ at this time d,i Get the rotation angle error e of each driving joint i , for e i Perform differential operation to obtain the error change rate ec i , to drive the joint’s rotation error e i and error change rate ec i As input, the parameter increment Δρ of the feedback controller is obtained through the two-dimensional fuzzy algorithm module i =(ΔK P,i ΔK I,i ΔK D,i ) T (i=1,2,…,n), where i is the i-th driven joint, ΔK P,i , ΔK I,i and ΔK D,i are the proportional coefficient increment, integral coefficient increment and differential coefficient increment of the feedback controller of the i-th driven joint respectively;

[0035] (4) Through the formula ρ i =ρ Oi +Δρ i The final feedback controller parameters are calculated and the online tuning of the feedback controller parameters is realized.

[0036] The one-dimensional fuzzy algorithm implementation process in step (2) is as follows:

[0037] The absolute value of the desired input torque of the driving joint is used as input, and the initial value of the feedback controller parameter is used as output. The input and output values are divided into five regions, namely "extremely low (EL)", "low (LO)", "moderate (ME)", "high (HI)", and "extremely high (EH)". The corresponding fuzzy subset domain is {1, 2, 3, 4, 5}. Substitute the absolute value of the desired input torque into Figure 2 The triangular membership function shown in FIG. 1 is used to obtain the membership of each region, and the one-dimensional fuzzy rule table shown in Table 1 is searched according to the membership of each region to obtain the output.

[0038] In addition, the two-dimensional fuzzy algorithm implementation process in step (3) is as follows:

[0039] The angular error e of the driven joint i and error change rate ec i As input, the parameter increment of the feedback controller is used as output. The input and output values are divided into 7 regions, namely "Negative Large (NB)", "Negative Medium (NM)", "Negative Small (NS)", "Zero (ZO)", "Positive Small (PS)", "Positive Medium (PM)", and "Positive Large (PB)". The corresponding fuzzy subset domain is {-3, -2, -1, 0, 1, 2, 3}. The angular error e of the driven joint is i and error change rate ec i Substitute into Figure 3 The triangular membership function shown in FIG. 1 is used to obtain the membership of each region, and the two-dimensional fuzzy rule table shown in Table 2 is searched according to the membership of each region to obtain the output.

[0040] Table 1

[0041]

[0042] Table 2

[0043]

[0044] The above-described embodiments merely represent specific implementations of the present invention. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, and all such variations and improvements fall within the scope of protection of the present invention.

Claims

1. An online tuning method for industrial robot controller parameters, characterized in that: The following steps are involved: Step 1: Build a rigid body dynamics model of a multi-joint serial industrial robot without considering the robot uncertainty term and the interference of friction torque; Step 2: Substitute the actual rotation angle of the driving joint into the rigid body dynamics model of the robot to obtain the expected input torque of the driving joint. The absolute value of the expected input torque is used as input to obtain the initial value of the feedback controller parameters through a one-dimensional fuzzy algorithm; Step 3: Using the rotation angle error and error change rate of the driving joint as input, the parameter increment of the feedback controller is obtained through the two-dimensional fuzzy algorithm module; Step 4: The final feedback controller parameters are obtained by calculating the initial parameter values and the parameter increments, thereby realizing online tuning of the feedback controller parameters.

2. The online tuning method for industrial robot controller parameters according to claim 1, characterized in that: In step 1, the rigid body dynamics model is as follows: Where, θ=(θ1 θ2 … θ n )∈ n , are the angle, angular velocity and angular acceleration vector of each driving joint, respectively, M(θ)∈ n×n is the symmetric inertia matrix, is the Coriolis vector, G(θ) is the gravitational moment vector, τ=(τ1τ2…τ n )∈ n is the input torque vector of each driven joint.

3. The online tuning method for industrial robot controller parameters according to claim 2, characterized in that: The step 2 is specifically as follows: During the real-time operation of the robot, the actual angle of the motor fed back by the encoder is converted into the actual rotation angle θ of the driving joint. a,i The actual rotation angle of the driving joint is substituted into the dynamic model of the robot to obtain the expected input torque of the driving joint. The absolute value of the expected input torque is used as input to obtain the initial value of the feedback controller parameter ρ through the one-dimensional fuzzy algorithm. Oi =(K PO,i K IO,i K DO,i ) T (i=1,2,…,n), where i is the i-th driven joint, K PO,i , K IO,i and K DO,i are the initial proportional coefficient, initial integral coefficient and initial differential coefficient of the feedback controller of the i-th driven joint respectively.

4. The online tuning method for industrial robot controller parameters according to claim 3 is characterized in that: The one-dimensional fuzzy algorithm implementation process is as follows: The absolute value of the expected input torque of the driving joint is used as input, and the initial value of the feedback controller parameter is used as output. The input and output values are divided into multiple regions. The absolute value of the expected input torque is substituted into the triangular membership function to obtain the membership of each region. The output is obtained by searching the one-dimensional fuzzy rule table based on the membership of each region.

5. The online tuning method for industrial robot controller parameters according to claim 3, characterized in that: The step 3 is specifically as follows: The actual angle of the motor fed back by the encoder is converted into the actual rotation angle θ of the driven joint a,i , minus the desired rotation angle θ at this time d,i Get the rotation angle error e of each driving joint i , for e i Perform differential operation to obtain the error change rate ec i , to drive the joint’s rotation error e i and error change rate ec i As input, the parameter increment Δρ of the feedback controller is obtained through the two-dimensional fuzzy algorithm module i =(ΔK P,i ΔK I,i ΔK D,i ) T (i=1,2,…,n), where i is the i-th driven joint, ΔK P,i , ΔK I,i and ΔK D,i are the proportional coefficient increment, integral coefficient increment and differential coefficient increment of the feedback controller of the i-th driven joint respectively.

6. The online tuning method for industrial robot controller parameters according to claim 5, characterized in that: The two-dimensional fuzzy algorithm implementation process is as follows: The angular error e of the driven joint i and error change rate ec i As input, the parameter increment of the feedback controller is output, the input and output values are divided into multiple regions, and the angular error e of the driven joint is i and error change rate ec i Substitute the triangle membership function to obtain the membership of each area, and search the two-dimensional fuzzy rule table according to the membership of each area to obtain the output.

7. The online tuning method for industrial robot controller parameters according to claim 5, characterized in that: In step 4, by formula ρ i =ρ Oi +Δρ i The final feedback controller parameters are calculated.

Images