Electromechanical system kinetic parameter identification method and system based on inverse model

Abstract

The invention belongs to the related technical field of electromechanical control, and discloses an electromechanical system kinetic parameter identification method and system based on an inverse model. The method comprises the following steps: constructing a driving torque inverse model of a mechanical arm to be processed; actual motion parameters of the to-be-processed mechanical arm moving under the expected motion parameters are collected, and expected driving torque and actual driving torque are calculated through the inverse model; and an error function is constructed by utilizing the expected driving torque and the actual driving torque, an optimization model is constructed by taking the actual rotational inertia of the to-be-processed mechanical arm as a variable and taking the minimum error function as a target, and the optimization model is solved to obtain the corresponding actual rotational inertia when the error function is minimum, so that the identification of the kinetic parameters of the to-be-processed mechanical arm is realized. According to the invention, the problems of low calculation efficiency, poor robustness and incapability of adapting to a variable working condition scene when an electromechanical system with a strong nonlinear characteristic is processed by the existing parameter identification technology are solved and overcome.

Description

Technical Field

[0001] The present invention belongs to the technical field related to electromechanical control, and more specifically, relates to a method and system for identifying dynamic parameters of an electromechanical system based on an inverse model. Background Art

[0002] Mechatronic systems refer to systems composed of mechanical devices and electrical equipment, which are widely used in automation, robotics, aerospace, intelligent manufacturing and other fields. The performance of mechatronic systems is closely related to their dynamic parameters (such as inertia, friction, stiffness, damping, etc.), which directly affect the system's motion response, stability, control accuracy and anti-interference ability. The purpose of dynamic parameter identification of mechatronic systems is to improve the performance and reliability of the system by observing the input and output data of the system and identifying the unknown dynamic parameters in the model. By accurately identifying the dynamic parameters of the mechatronic system, the control accuracy, response speed and adaptability of the system to different working conditions can be effectively improved.

[0003] Traditional parameter identification methods estimate the dynamic parameter values ​​by constructing mathematical models and combining them with observed data, mainly including the least squares method and the maximum likelihood estimation method. However, these methods are usually used when the system model is assumed to be linear or approximately linear, and only consider error minimization, ignoring the complex nonlinear characteristics of the electromechanical system, showing shortcomings such as strong dependence on initial values, sensitivity to external disturbances of the system, and high computational complexity. In order to overcome the shortcomings of traditional parameter identification methods, parameter identification methods based on optimization algorithms use optimization algorithms such as genetic algorithms and particle swarm algorithms to find the optimal dynamic parameter values ​​that can minimize the objective function. Although they can effectively deal with nonlinear problems, they often have problems such as slow convergence and easy to fall into local optimality in practical applications; parameter identification methods based on data-driven fit the dynamic parameter values ​​through machine learning algorithms, which can conveniently deal with nonlinear problems, but the identification accuracy depends on the scale and quality of the observed data, and the generalization is poor, and the identification accuracy will drop significantly when the working conditions change. Summary of the invention

[0004] In view of the above defects or improvement needs of the prior art, the present invention provides a method and system for identifying the dynamic parameters of an electromechanical system based on an inverse model, which overcomes the problems of low computational efficiency, poor robustness and inability to adapt to variable working conditions of the existing parameter identification technology when processing electromechanical systems with strong nonlinear characteristics.

[0005] To achieve the above object, according to one aspect of the present invention, a method for identifying dynamic parameters of an electromechanical system based on an inverse model is provided, the method comprising the following steps: Simplify the robot arm to be processed into a single-axis robot arm, so as to construct an inverse model of the driving torque of the robot arm to be processed with respect to the moment of inertia; Setting expected motion parameters of the robot arm to be processed, collecting actual motion parameters of the robot arm to be processed when it moves under the expected motion parameters, and calculating the expected driving torque and the actual driving torque using the inverse model; The error function is constructed using the expected driving torque and the actual driving torque. The actual moment of inertia of the robot to be processed is taken as the variable and the optimization model is constructed with the minimum error function as the goal. The optimization model is solved to obtain the actual moment of inertia corresponding to the minimum error function, so as to realize the identification of the dynamic parameters of the robot to be processed.

[0006] Further preferably, the driving torque inverse model is as follows: ; in, and are the joint drive torque and load torque of the robot arm, J is the rated joint moment of inertia of the robot arm, b is the joint friction coefficient of the robot arm, m and l is the mass and length of the connecting rod of the robot arm, g is the acceleration due to gravity, , and are the desired joint rotation angle, desired joint angular velocity and desired joint angular acceleration of the robot arm, Set by humans, and Obtained by difference method.

[0007] Further preferably, the method of simplifying the robot arm to be processed into a single-axis robot arm is: simplifying the structure of the robot arm to be processed into a single-axis robot arm including a connecting rod and a rotating joint, and the multi-axis is simplified to the load torque at the end of the connecting rod.

[0008] Further preferably, the expected motion parameters are an expected rotation angle, an expected angular velocity and an expected angular acceleration of the connecting rod, and the actual motion parameters are an actual rotation angle, an actual angular velocity and an actual angular acceleration of the connecting rod.

[0009] Further preferably, the optimization model is as follows: ; in, and is the actual joint moment of inertia of the robot and its estimated value, N is the number of sampling points, subscript i Indicates i sampling points, , and It is the actual joint rotation angle, actual joint angular velocity and actual joint angular acceleration of the robot arm.

[0010] Further preferably, the optimization model is solved by using a gradient descent iterative method.

[0011] Further preferably, the gradient descent iteration method is performed according to the following steps: (1) Set the initial value of the actual moment of inertia and calculate the error function and gradient; (2) Update the estimated moment of inertia and learning rate; (3) Repeat step (2) until the preset number of iterations is reached.

[0012] Further preferably, the updating of the moment of inertia is performed according to the following formula: ;

[0013] in, α is the learning rate, which controls the step size of each iteration of the gradient descent method, k Indicates k Iterations, and Respectively represent k +1 and k The actual joint inertia of the robot arm at the iteration The estimated value of yes The simplified form of k The error function at the iteration for Find the gradient of the partial derivative.

[0014] According to another aspect of the present invention, a system for identifying dynamic parameters of an electromechanical system based on an inverse model is provided. The system includes an actuator for executing the above-mentioned method for identifying dynamic parameters of an electromechanical system based on an inverse model.

[0015] According to another aspect of the present invention, there is provided a computer-readable storage medium, which includes a computer program, and when the computer program is executed, the method for identifying dynamic parameters of an electromechanical system based on an inverse model is implemented.

[0016] In general, the above technical solutions conceived by the present invention have the following beneficial effects compared with the prior art: 1. The present invention applies the inverse model technology to model parameter identification, uses the inverse model to establish an error minimization optimization model, and reversely models the nonlinear relationship between system output and input by combining the inverse model, thereby effectively eliminating the influence of nonlinear error propagation in traditional model parameter identification methods. It is particularly suitable for electromechanical systems with strong nonlinear characteristics, explicitly reflects the core physical laws of the system, helps to reduce the interference of redundant parameters and external disturbances, has strong robustness and generalization, and is particularly advantageous when dealing with variable operating conditions.

[0017] 2. This method can effectively improve the identification accuracy and significantly reduce the calculation time. Traditional electromechanical system parameter identification methods usually rely on the quantity and quality of observation data, the setting of iterative initial values, and are easily affected by nonlinear error accumulation and external disturbances when dealing with systems with strong nonlinear characteristics. By combining the inverse model to reversely model the nonlinear relationship between the system output and input, a more intuitive and interpretable model framework is provided, thereby reducing the impact of nonlinear error propagation in traditional methods and improving the accuracy of parameter identification. In addition, by combining numerical differentiation methods and inverse models, the calculation process is optimized, the calculation efficiency is significantly improved, and unnecessary redundant calculations are avoided, thereby shortening the time of the entire parameter identification process.

[0018] 3. This method has strong robustness and adaptability, and is particularly suitable for variable operating conditions. Traditional parameter identification methods often require retraining the model or adjusting parameters when faced with changes in the environment or operating conditions, resulting in reduced recognition accuracy or reduced computational efficiency. The inverse model only models the key parameters related to the system state feedback. The identification method based on the inverse model can eliminate the interference caused by operating condition changes and external disturbances by constructing an inverse model. With this method, good recognition performance can be maintained even under complex and dynamically changing operating conditions, ensuring that the dynamic parameters of the electromechanical system can be efficiently and accurately identified under different working environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the main flow of the inverse model-based electromechanical system dynamic parameter identification method constructed according to the preferred embodiment of the present invention.

[0020] Figure 2 It is a schematic diagram of the structure of a single-axis robotic arm system according to an inverse model-based electromechanical system dynamics parameter identification method constructed in accordance with a preferred embodiment of the present invention.

[0021] Figure 3 It is a schematic diagram of the main process of iteratively optimizing and solving the kinetic parameters of an exemplary embodiment using a gradient descent algorithm constructed according to a preferred embodiment of the present invention.

[0022] Figure 4 It is a diagram of an experimental scene of a six-axis robotic arm system according to an inverse model-based electromechanical system dynamics parameter identification method constructed in accordance with a preferred embodiment of the present invention.

[0023] Figure 5 : It is a parameter identification result diagram of the moment of inertia of each joint of the six-axis robot under different load conditions constructed according to the preferred embodiment of the present invention, wherein (a) is the parameter identification result diagram of the moment of inertia of the 6th joint when the angles of the other joints of the six-axis robot are 0°; (b) is the parameter identification result diagram of the moment of inertia of the 5th joint when the angles of the other joints of the six-axis robot are 0°; (c) is the parameter identification result diagram of the moment of inertia of the 5th joint when the 6th joint of the six-axis robot is 5° and the angles of the other joints are 0°; (d) is the parameter identification result diagram of the moment of inertia of the 5th joint when the 6th joint of the six-axis robot is 45° and the angles of the other joints are 0°.

[0024] Figure 6 It is a parameter identification result diagram of the moment of inertia of the 4th joint of the six-axis robot under different load conditions constructed according to the preferred embodiment of the present invention, wherein (a) is the parameter identification result diagram of the moment of inertia of the 4th joint of the six-axis robot when the angles of the other joints are 0°; (b) is the parameter identification result diagram of the moment of inertia of the 4th joint of the six-axis robot when the 5th joint is 5° and the angles of the other joints are 0°; (c) is the parameter identification result diagram of the moment of inertia of the 4th joint of the 6-DOF robot when the 5th joint is 20° and the angles of the other joints are 0°; (d) is the parameter identification result diagram of the moment of inertia of the 4th joint of the six-axis robot when the 5th joint is 45° and the angles of the other joints are 0°. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0026] like Figure 1 As shown, a method for identifying dynamic parameters of an electromechanical system based on an inverse model specifically includes the following steps: S1. Analyze the physical structure of the electromechanical system, derive the forward dynamics model of the electromechanical system based on the Lagrangian dynamics equation, and establish the inverse dynamics model of the electromechanical system on this basis.

[0027] The electromechanical system to be studied is simplified to a single-axis robotic arm. Figure 1 As shown in the figure, the single-axis robot arm consists of a rigid link and a revolute joint. The joint position of the robot arm, that is, the joint rotation angle, is expressed by i The joint driving torque of the robot arm is t , the load torque at the end of the robot arm is t L The single-axis robot arm consists of the following basic elements: Base: The bottom fixed part of the robot arm, usually connected to a supporting platform or the ground.

[0028] Joint: The axis of rotation of the robot arm, which is its only degree of freedom. The joint is usually driven by a motor or servo, allowing the robot arm to rotate around the joint axis.

[0029] Link: The connecting part from the base to the end effector, usually a rigid rod that can bear torque and force.

[0030] End-effector: The component at the end of the robot arm, used to perform various tasks. In a single-axis robot arm, the end-effector is usually fixed to the end of the connecting rod and can be designed into different shapes according to the task requirements.

[0031] Load: Usually the object mounted on the end effector, or the mass of the end effector itself.

[0032] like Figure 2 As shown in the figure, the kinetic energy of a single-axis robot arm mainly comes from the rotation of the robot arm. T It can be expressed as: .

[0033] in, J is the moment of inertia of the robot arm, is the joint angular velocity of the robot arm.

[0034] The potential energy of a single-axis robot arm mainly comes from the gravity of the robot arm. Assume that the gravity acts on the center of mass of the robot arm, which is located at the midpoint of the connecting rod and the distance from the rotation axis is l / 2. Potential energy of the robotic arm V It can be expressed as: .

[0035] in, m is the mass of the connecting rod of the robot arm, g is the acceleration due to gravity, l is the connecting rod length of the robot arm, i is the joint rotation angle of the robot arm.

[0036] Lagrangian function L Kinetic Energy T With potential energy V Difference: .

[0037] The Lagrange equations are of the form: .

[0038] in, F It is a general force. and The calculation of is as follows: ; .

[0039] Then, the time derivative part of the Lagrange equation is calculated according to equation (5): .

[0040] Substituting equations (6) and (7) into equation (4), the Lagrangian equation of the single-axis manipulator is obtained as follows: .

[0041] for Figure 2 The single-axis robot arm shown, the generalized force F Including joint driving torque, friction force and load torque at the end of the robot. Ignoring the external disturbance of the system, the dynamic equation of the robot can be deduced from equation (8): .

[0042] in, is the joint angular acceleration of the robot arm, b is the joint friction coefficient of the robot arm, t L It is the load torque at the end of the robot arm, which comes from the robot arm's operating task or external load.

[0043] Through the Lagrangian method, the mathematical model of the single-axis robot arm is obtained. The model includes the influence of joint inertia, friction, gravity and external load torque. This equation can be used to analyze and control the movement of the robot arm, especially in the subsequent inverse dynamics solution and parameter identification.

[0044] For a single-axis robot, the direct dynamics problem is: the joint drive torque applied by the actuator to the robot at a certain moment is known. t , and the instantaneous joint angle i and joint angular velocity According to formula (9), the joint angular acceleration of the robot arm at this time is calculated ; Inverse dynamics problem refers to: the joint rotation angle of the robot arm at a certain moment is known i , joint angular velocity and joint angle acceleration According to formula (9), the driving torque applied by the actuator on the robot arm at this time is t .

[0045] Put the unknown quantity to be solved on the left side of the equation and the known quantity on the right side of the equation. The forward model and inverse model of the single-axis robot arm can be described as follows: .

[0046] S2. Set different control signals for the electromechanical system, and perform experimental data collection and calculation based on sensors and numerical differentiation methods.

[0047] S2.1 Set different control signals for the electromechanical system (i.e., the expected rotation angle of the system), calculate the expected angular velocity and expected angular acceleration of the system based on the numerical differentiation method, and substitute them into the inverse dynamics model established in step S1 to calculate the expected input of the system (i.e., the expected driving torque of the system). Input different control signals to the single-axis manipulator, calculate the required experimental data through the numerical differentiation method, and collect the actual output observation value of the system through the sensor.

[0048] In this embodiment, it is the expected rotation angle of the joint of the single-axis manipulator. Substitute the set control signal and the rated structural parameters of the single-axis manipulator into the inverse model of the single-axis manipulator system established in step S1 (10) to calculate the expected input of the system, that is, calculate the driving torque that the actuator needs to apply to the joint of the manipulator in order to drive the single-axis manipulator to reach the expected rotation angle of the joint. The specific implementation steps are as follows: Set the control signal to the desired rotation angle of the robot's joints i d , and They are the expected angular acceleration of the joint and the expected acceleration of the joint angle. In the motion analysis of the robot arm, the numerical differentiation method can be used to approximate the angular velocity and angular acceleration at each moment based on the rotation angle. Numerical differentiation methods include forward difference method, backward difference method and central difference method, among which the central difference method can achieve more accurate calculation. Use the central difference method to calculate The formula is as follows: .

[0049] Assume that the total number of time series is N time steps, where i ∈[0, N -1] is the index of the time series, i d,i is the expected rotation angle of the joint at the corresponding moment, Δ t = t i+1 - ti is the time interval between each moment.

[0050] After calculating the expected angular velocity of the joint, the central difference method can be used to further calculate the expected angular acceleration of the joint: .

[0051] Likewise, is the expected angular velocity of the joint at the corresponding moment.

[0052] For the first moment ( i =0) and the last moment ( i = N -1), since there is no data at the previous and next moments, the forward difference method or the backward difference method can be used to calculate the angular velocity and angular acceleration.

[0053] For the first time step, the angular acceleration is calculated based on the current and next values ​​of the angular velocity using the forward difference method: .

[0054] For the last time step, the angular acceleration is calculated based on the current and previous values ​​of the angular velocity using the backward difference method: .

[0055] Set the rated joint moment of inertia of the robot arm to J ,Will i d And calculated and Substituting into the inverse model of equation (10), the system input is calculated, that is, the driving torque applied by the actuator to the joint of the manipulator t for: .

[0056] S2.2 The desired driving torque of the system calculated in step S2.1 is applied to the electromechanical system through the actuator, and the actual rotation angle of the robot arm of the system at each time step is collected and recorded through the sensor.

[0057] Step 2: According to the system expected input calculated in step S2.1, the joint driving torque is applied to the real single-axis robot system through the actuator. t , observe the response of the real robotic arm equipment and record the relevant real observation values. The specific implementation steps are as follows: The joint driving torque calculated in step S2.1 is tThe actuator is applied to the rotary joint of the real single-axis robot arm. The real robot arm responds to the input joint torque and rotates around the joint axis. The angle sensor collects the actual rotation angle of the joint of the single-axis robot arm at each moment. i a,i , and refer to equations (12) to (14) to obtain the actual angular velocity of the joint of the single-axis robot at each moment through numerical differentiation method and the actual angular acceleration of the joint .

[0058] Through sensor acquisition and numerical differential calculation, the expected rotation angle of the joint of the single-axis robot arm at each moment is recorded. i d,i 、Joint expected angular velocity and the desired angular acceleration of the joint , Actual rotation angle of joint i a,i , Actual angular velocity of joint , Actual angular acceleration of joint , used for subsequent parameter identification operations.

[0059] S3, selecting the dynamic parameters of the electromechanical system to be identified, defining the error function of the electromechanical system according to the parameters to be identified, eliminating the working condition terms based on the inverse model, and defining the optimization objective function on this basis; S3.1. Based on the system inverse model constructed in step S1, respectively calculate the inverse model solution results of the system expected rotation angle and the system actual rotation angle, analyze the reason for the deviation between the two results, and thus select the dynamic parameters that need to be identified.

[0060] In the inverse model parameter identification, the actual rotation angle of the joint i a Substituting into the inverse model calculation formula (10), we get: .

[0061] Theoretically, the left sides of equations (15) and (16) are t , so the right sides of the two equations should be equal, that is .

[0062] Solved i d =θ a , but this result does not conform to the actual situation: the actual rotation angle of the single-axis robot arm joint in actual operation i a Always the same angle as the joint's expected rotation i d There is a certain deviation between them, so formula (16) does not hold. .

[0063] It can be inferred that this is the actual device's moment of inertia J a With rated value J There is a bias (i.e. J a ≠ J ). Therefore, the joint moment of inertia of the single-axis robot arm is selected as the dynamic parameter to be identified.

[0064] Actual equipment robot arm joint moment of inertia J a is the unknown parameter to be identified, and equation (16) is changed to ;

[0065] S3.2 Define an error function based on the parameters to be identified selected in step S3.1, eliminate the operating condition terms based on the inverse model, and determine the optimization target and define the optimization target function based on the error function.

[0066] Kinetic parameters to be identified selected according to step S3.1 J a , defined with J a The error function of the related single-axis robot is obtained by eliminating the load term t L To eliminate the impact of changes in load conditions, determine the optimization target based on the error function, and define the optimization objective function, the following are the specific implementation steps: Combining equation (15) and equation (18), we define J a The error function E( J a ) to measure The parameter estimation error is: ;

[0067] Among them, E( J a ) represents the total error over all data points, N is the number of sampling points. In this error function, the moment i The expected rotation angle of the joint i d,i Given by humans, the time i The actual rotation angle of the joint i a,i The actual single-axis robot system i d,i Respond to get.

[0068] In addition, it can be seen from formula (19) that the error function E( Ja ) finally eliminates the load term t L , that is, the error function is not affected by changes in working conditions. The identification technology based on the inverse model can effectively realize the identification of the dynamic parameters of the single-axis robot under variable load conditions.

[0069] In order to find the closest actual moment of inertia J a Estimated moment of inertia of the single-axis manipulator joint , we need to minimize the error function E( J a ), so the optimization objective function is defined as: ;

[0070] Usually, the error function E( J a )right J a The actual moment of inertia is obtained by taking the derivative of J a The best estimate of .

[0071] S4, based on the gradient descent method, perform iterative optimization of the parameters to be identified to obtain the best estimated values ​​of the parameters; S4.1 Based on the error function constructed in step S3.2, the gradient of the error function is calculated by taking the derivative of the parameter to be identified, and the minimization condition is set to construct the minimization condition equation.

[0072] The optimization objective function constructed in step S3.2 is iteratively solved using the gradient descent method to obtain the best estimate of the kinetic parameter to be identified selected in step S3.1. S4 includes two sub-steps S4.1 and S4.2, wherein step S4.1 calculates the gradient of the error function defined in step S3 by taking the first-order derivative of the kinetic parameter to be identified, and solves the minimization boundary condition by setting the gradient to zero. The specific implementation steps are as follows: Calculate the error function E( J a )about J a Gradient , can be calculated by E( J a )about J a The first-order derivative of is: ; make ; Formula (21) can be simplified as: ; According to the chain rule, equation (23) can be expanded as: ; in, , Substituting formula (22) into formula (24), we can get: ;

[0073] Set the gradient to zero to solve the minimization condition, that is, let ∇ Ja E( J a ) = ∂E( J a ) / ∂ J a = 0, substituting into formula (25) we can get: ;

[0074] By solving the minimization condition equation (26), we find the error function E( J a )The smallest optimal parameter estimate .

[0075] Specifically, the minimization condition equation can usually be solved by linearization method. If there are multiple unknown quantities (such as b or l / 2 ·mg ), which can be treated as constants to solve for multiple parameters simultaneously.

[0076] However, in electromechanical system problems with strong nonlinear characteristics, the minimization condition equation contains nonlinear terms, and analytical solutions are not easy to achieve. Therefore, numerical optimization methods, such as gradient descent or Newton's method, are usually used to iteratively solve the best parameter estimates of this equation. Among them, the gradient descent method is very suitable for dealing with such nonlinear optimization problems. The main advantages include: the gradient descent method can handle complex dynamic equations, especially parameter identification problems; the gradient descent method only requires the gradient information of the objective function, which is easy to calculate and implement; the gradient descent method can be easily extended to problems with multiple parameters (multi-dimensional optimization).

[0077] S4.2 The minimization condition equation constructed in step S4.1 is solved iteratively using the gradient descent method, and finally the best estimate of the parameter to be identified is obtained when the error function converges or the maximum number of iterations is reached.

[0078] The actual moment of inertia parameters are estimated using the gradient descent method based on the minimization condition equation (26) obtained in step S4.1. Iterative optimization solution. Figure 3 As shown, the following are the specific implementation steps: Initialization parameters: Start with an initial value (usually random or based on empirical estimates), i.e. ; Calculate the gradient: Calculate the error function according to formula (25) right Gradient , and in simplified form Make a substitution; Update parameter estimates: Update parameter estimates based on gradient and learning rate according to formula (27): ;

[0079] in, α is the learning rate, which controls the step size of each iteration, k Indicates k Iterations; Update the learning rate: Substitute the current parameter estimate and update it according to the error function Update the learning rate according to the value size; Repeat the iteration until the error function Convergence or reaching the set maximum number of iterations M ; The beneficial effects of the present invention are verified below in conjunction with specific embodiments: A six-axis robotic arm system is selected to verify the inverse model-based electromechanical system dynamic parameter identification method. The experimental scene is set up as follows: Figure 4 The object to be parameterized is a serial robot with six degrees of freedom. An inverse model of the robot is built based on the Lagrangian dynamics equation, and a method for estimating the actual moment of inertia of the robot joint is built. J a The actuator is used to convert the joint driving torque obtained by solving the inverse model into t Applied to the robot arm, an angle sensor is used to collect the actual rotation angle of the robot arm's joints at each moment i a The controller sets the control signal, i.e. the desired rotation angle of the robot joint. i d For parabola y ( t )=1000 ·t 2 Perform simulation.

[0080] The actual moment of inertia of one joint of the six-axis robot arm is identified, and the structures connected to this joint in series can be regarded as the load of this joint. At this time, the system can be simplified as follows Figure 2 The one shown has a load moment t L The 1st to 6th joints of the robot can be set to have different angles, that is, the robot joints have different load conditions.

[0081] like Figure 5 and 6 As shown, the electromechanical system dynamics parameter identification method based on the inverse model provided in the embodiment of the present application is iteratively optimized and can eventually gradually converge to a stable estimation value of the moment of inertia parameter of each joint when the six-axis robot arm is under different load conditions. .

[0082] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying dynamic parameters of an electromechanical system based on an inverse model, characterized in that: The method comprises the following steps: Simplify the robot arm to be processed into a single-axis robot arm, so as to construct an inverse model of the driving torque of the robot arm to be processed with respect to the moment of inertia; Setting expected motion parameters of the robot arm to be processed, collecting actual motion parameters of the robot arm to be processed when it moves under the expected motion parameters, and calculating the expected driving torque and the actual driving torque using the inverse model; The error function is constructed using the expected driving torque and the actual driving torque. The actual moment of inertia of the robot to be processed is taken as the variable and the optimization model is constructed with the minimum error function as the goal. The optimization model is solved to obtain the actual moment of inertia corresponding to the minimum error function, so as to realize the identification of the dynamic parameters of the robot to be processed.

2. The method for identifying the dynamic parameters of an electromechanical system based on an inverse model according to claim 1, characterized in that: The driving torque inverse model is as follows: ； in, and are the joint drive torque and load torque of the robot arm, J is the rated joint moment of inertia of the robot arm, b is the joint friction coefficient of the robot arm, m and l is the mass and length of the connecting rod of the robot arm, g is the acceleration due to gravity, , and are the desired joint rotation angle, desired joint angular velocity, and desired joint angular acceleration of the robot arm, Set by humans, and Obtained by difference method.

3. A method for identifying dynamic parameters of an electromechanical system based on an inverse model as claimed in claim 2, characterized in that: The method for simplifying the robot arm to be processed into a single-axis robot arm is: simplifying the structure of the robot arm to be processed into a single-axis robot arm including a connecting rod and a rotating joint, and the multi-axis is simplified to the load torque at the end of the connecting rod.

4. A method for identifying dynamic parameters of an electromechanical system based on an inverse model as claimed in claim 1 or 2, characterized in that: The expected motion parameters are the expected rotation angle, expected angular velocity and expected angular acceleration of the connecting rod, and the actual motion parameters are the actual rotation angle, actual angular velocity and actual angular acceleration of the connecting rod.

5. A method for identifying the dynamic parameters of an electromechanical system based on an inverse model as claimed in claim 1 or 2, characterized in that: The optimization model is as follows: ； in, J a and is the actual joint moment of inertia of the robot and its estimated value, N is the number of sampling points, subscript i Indicates i sampling points, , and It is the actual joint rotation angle, actual joint angular velocity and actual joint angular acceleration of the robot arm.

6. A method for identifying dynamic parameters of an electromechanical system based on an inverse model as claimed in claim 1 or 2, characterized in that: The optimization model is solved by using a gradient descent iterative method.

7. The method for identifying the dynamic parameters of an electromechanical system based on an inverse model according to claim 6, characterized in that: The gradient descent iteration method is performed according to the following steps: (1) Set the initial value of the actual moment of inertia and calculate the error function and gradient; (2) Update the estimated moment of inertia and learning rate; (3) Repeat step (2) until the preset number of iterations is reached.

8. The method for identifying the dynamic parameters of an electromechanical system based on an inverse model according to claim 7, characterized in that: The updating of the moment of inertia is performed according to the following formula: ； in, α is the learning rate, which controls the step size of each iteration of the gradient descent method, k Indicates k Iterations, and Respectively represent k +1 and k The actual joint inertia of the robot arm at the iteration The estimated value of yes The simplified form of k The error function at the iteration for Find the gradient of the partial derivative.

9. A system for identifying dynamic parameters of an electromechanical system based on an inverse model, characterized in that: The system comprises an actuator, which is used to execute the inverse model-based electromechanical system dynamic parameter identification method described in any one of claims 1 to 8.

10. A computer-readable storage medium comprising a computer program, characterized in that: When the computer program is executed, the method for identifying dynamic parameters of an electromechanical system based on an inverse model as described in any one of claims 1 to 8 is implemented.

Images