Spherical electric joint module model prediction control system and method based on generalized proportional integral observer
By using a generalized proportional integral observer and model prediction control system in the spherical electric joint module movement device, the external interference is estimated and compensated in real time, and the existing control algorithm cannot effectively deal with the influence of complex nonlinear factors and multivariable interactions is achieved, and a high-precision, stable and anti-interference trajectory tracking effect is achieved.
Patent Information
- Application Number
- CN202510217266.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-26
- Publication Date
- 2025-05-30
AI Technical Summary
Existing multi-degree-of-freedom motion device control algorithms, such as PID control, adaptive control and sliding mode control, cannot effectively deal with the influence of complex nonlinear factors and multivariable interactions, resulting in low control accuracy, slow response speed and weak anti-interference ability.
A spherical electric joint module model prediction control system based on a generalized proportional integral observer is adopted to estimate external interference online in real time and compensate it for feedforward to the model prediction controller to optimize the control torque.
It significantly improves the system's trajectory tracking accuracy and stability, enhances anti-interference ability and robustness, reduces control torque, and improves the overall performance of the system.
Smart Images
Figure CN120065693A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of robot control, and particularly to a model predictive control system and method for a spherical electric joint module based on a generalized proportional integral observer. Background Art
[0002] With the continuous development of modern industrial technology, the demand for multi-degree-of-freedom motion devices is increasing. Traditional multi-degree-of-freedom devices are limited by their structure, usually being large in volume, occupying a large amount of space, and having limited integration. Although innovative multi-degree-of-freedom spherical motors solve the volume problem, their output torque is still insufficient to meet the requirements of large-scale industrial applications. Conventional three-degree-of-freedom joints composed of a series-parallel combination of multiple motors and linkages have a relatively complex system, and inevitable situations such as decreased coordination between components, delayed response, and reduced accuracy will occur due to their complicated structure.
[0003] In this context, the three-degree-of-freedom spherical electric joint module provides a new perspective and solution for solving the above problems with its unique design concept and excellent three-dimensional motion ability. The spherical electric joint module mainly consists of a bridge-shaped bracket, a hemispherical rotating bracket, a hemispherical stationary bracket, an output shaft, and three high-torque-density rotating joints in orthogonal directions. By integrating three mutually perpendicular rotating joints into a compact spherical housing and through precise control strategies, the rotation angles in these three degrees of freedom can be adjusted independently and synchronously, thereby precisely controlling the spatial position and attitude of the end effector. This multi-degree-of-freedom motion device can achieve multi-degree-of-freedom motion in a narrow space, and has advantages such as high integration, small volume, simple structure, and high precision. It is crucial for achieving high-precision positioning, complex trajectory tracking, and flexible operation in a dynamic environment, providing strong technical support for the progress of multiple fields such as industrial automation, intelligent manufacturing, medical technology, and space exploration, and having important research significance and engineering application prospects in the fields of robotics, medicine, and aerospace.
[0004] The spherical electric joint module has significant application value and can be used in multiple fields in the future. However, due to its fine spatial structure, it is difficult to establish a model, adding many challenges to motion control. As a non-linear multivariable system, the spherical electric joint module not only has modeling errors but is also affected by factors such as friction and external disturbances. The core of the motion control of this multi-degree-of-freedom motion device focuses on several key points. First, it is necessary to accurately construct its kinematic and dynamic models; second, it is necessary to integrally design a motor drive system with a position detection function; finally, it is necessary to construct a complete trajectory tracking control strategy. At present, thanks to the unremitting efforts of researchers, the motion control system has successfully achieved a major transformation from open-loop to closed-loop and made remarkable progress. In the field of controlling multi-degree-of-freedom motion devices, common algorithms include PID control, adaptive control, robust control, and sliding mode variable structure control, etc. These control methods have their own unique advantages and corresponding limitations, and need to be selected and applied according to specific control requirements.
[0005] Multi-degree-of-freedom motion devices often involve multiple variables and complex interactions. In this case, the PID control algorithm cannot effectively handle the interactions and constraint conditions between each degree of freedom, resulting in poor control effects.
[0006] The adaptive control algorithm mainly adjusts based on the current state and lacks the ability to predict the future state of the system. In multi-degree-of-freedom devices, due to the high complexity and non-linear characteristics of the system, when the system state changes, the adaptive control needs a long time to re-adjust the control parameters, and this lag will affect the stability and control accuracy of the system.
[0007] The core strategy of the sliding mode control algorithm is to make the system state slide on the sliding mode surface through the switching control term. However, in the application of multi-degree-of-freedom motion devices, the sign function in the switching control term will cause the control signal to switch rapidly, resulting in high-frequency oscillations. This chattering will not only cause high-frequency noise in the system, affecting the control effect, but may also damage the hardware of the system.
[0008] In summary, the PID control algorithm cannot provide sufficient control accuracy and robustness to meet the requirements; the adaptive control algorithm has the disadvantages of response lag, large steady-state error, and low control accuracy; the chattering phenomenon of sliding mode control will cause the system output to fluctuate near the equilibrium point, affecting the control accuracy, aggravating the impact and wear of mechanical components, and shortening the service life of the system. Summary of the Invention
[0009] To solve the problems existing in the prior art and improve the control accuracy and response speed of the spherical electric joint module, the present invention proposes a model predictive control system and method for the spherical electric joint module based on a generalized proportional-integral observer. This method can handle problems with complex non-linear factors and multi-variable interaction effects, and has strong adaptability and flexibility. Moreover, by using the generalized proportional-integral observer, the system uncertainty caused by external disturbances is reduced, the steady-state error of system trajectory tracking is decreased, and the anti-interference ability and robustness of the system are further enhanced.
[0010] The technical solutions adopted by the present invention include:
[0011] A model predictive control system for a spherical electric joint module based on a generalized proportional-integral observer, comprising: a generalized proportional-integral observer, a model predictive controller, and a prototype of the spherical electric joint module. Among them, the generalized proportional-integral observer is used to perform real-time online estimation of the state variables and external disturbances of the prototype of the spherical electric joint module, so as to obtain an estimated value of the external disturbance, and feed forward and compensate the estimated value of the external disturbance estimated by the generalized proportional-integral observer into the model predictive controller. The model predictive controller is used to obtain the control torque of the prototype of the spherical electric joint module based on the optimization result obtained by the model predictive control algorithm and in combination with the feed forward compensation term of the external disturbance from the generalized proportional-integral observer.
[0012] A model predictive control method for a spherical electric joint module based on a generalized proportional-integral observer, which is executed by the model predictive control system for the spherical electric joint module based on the generalized proportional-integral observer, and comprises the following steps:
[0013] Step 1: Use the generalized proportional-integral observer to perform real-time online estimation of the state variables and external disturbances of the spherical electric joint module system, so as to obtain an estimated value of the external disturbance;
[0014] Step 2: Feed forward and compensate the estimated value of the external disturbance estimated by the generalized proportional-integral observer into the model predictive controller. The model predictive controller obtains the control torque of the prototype of the spherical electric joint module based on the optimization result obtained by the model predictive control algorithm and in combination with the feed forward compensation term of the external disturbance from the generalized proportional-integral observer.
[0015] The beneficial effects brought by the technical solutions of the present invention:
[0016] The present invention provides a model predictive control system and method for a spherical electric joint module based on a generalized proportional-integral observer, aiming to optimize the trajectory tracking performance of the spherical electric joint module. The present invention adopts a generalized proportional-integral observer, which estimates external disturbances in real time and online, and feeds the obtained disturbance estimation value forward to compensate the model predictive controller. This design effectively weakens the uncertain factors in the control input, significantly enhances the system's ability to cope with various disturbances, thereby reducing the control torque required to maintain trajectory tracking, and greatly improving the accuracy of trajectory tracking and the stability of the overall system. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 Shows a block diagram of the model predictive control system for the spherical electric joint module based on the generalized proportional-integral observer of the present invention.
[0018] Figure 2 Is the control torque diagram obtained by using a linear sliding mode controller (LSMC).
[0019] Figure 3 Is the control torque diagram obtained by using a model predictive controller (MPC).
[0020] Figure 4 Is the control torque diagram obtained by using the model predictive controller (MPC-GPIO) based on the generalized proportional-integral observer in the present invention.
[0021] Figure 5 Is the tracking error generated when using a linear sliding mode controller for trajectory tracking.
[0022] Figure 6 Is the tracking error generated when using a model predictive controller for trajectory tracking.
[0023] Figure 7 Is the tracking error generated when using the model predictive controller based on the generalized proportional-integral observer in the present invention for trajectory tracking.
[0024] Figure 8 Shows the mean square error (MSE) obtained by using three control algorithms to perform trajectory tracking on three joint angles respectively. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0025] Hereinafter, the exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms, and the present disclosure should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully communicated to those skilled in the art.
[0026] The implementation details of the present invention will be elaborated in detail below in conjunction with the accompanying drawings.
[0027] The present invention provides a model predictive control system for a spherical electric joint module based on a generalized proportional-integral observer. Figure 1 The block diagram of the system is shown. The system mainly consists of three parts: a generalized proportional-integral observer, a model predictive controller, and a prototype of the spherical electric joint module. Among them, the generalized proportional-integral observer is used to estimate the state variables and external disturbances of the spherical electric joint module prototype in real time and online, so as to obtain an estimated value of the external disturbance, and feed forward and compensate the estimated value of the external disturbance estimated by the generalized proportional-integral observer into the model predictive controller. The model predictive controller is used to obtain the composite control law of the spherical electric joint module prototype according to the optimal result obtained by the model predictive control algorithm and in combination with the feed forward compensation term of the external disturbance from the generalized proportional-integral observer.
[0028] The working process of the model predictive controller is described below. The model predictive controller is constructed through the state space equation of the spherical electric joint module prototype. At each sampling moment, the model predictive controller predicts the state of the spherical electric joint module prototype in a future period according to the current position and velocity information of the spherical electric joint module prototype, designs a cost function according to the control objective of the model predictive controller, and minimizes the value of the cost function by solving the optimization problem, so as to obtain the optimal control sequence at the current moment, and applies the first control quantity in the obtained optimal control sequence to the spherical electric joint module prototype. This process is repeated and optimized in a rolling manner to make it move along the desired joint angle position.
[0029] The working process of the generalized proportional-integral observer is described below. The generalized proportional-integral observer is used to estimate the state variables and external disturbances of the spherical electric joint module prototype in real time and online. The generalized proportional-integral observer estimates by expanding the external disturbance into the state of the spherical electric joint module prototype and adjusts the corresponding observer order according to the order of the disturbance model, and can better estimate the time-varying type of disturbance, feed forward and compensate the time-varying type of disturbance into the model predictive controller, provide accurate feed forward information for subsequent control, and thus further offset the influence of external disturbances on the system and improve the stability and accuracy of the system.
[0030] Figure 1 In , represent the desired joint angle position vector and velocity vector respectively; , represent the actual joint angle position vector and velocity vector respectively; is the joint control torque, is the unknown external disturbance; is the estimated value of the external disturbance obtained by the generalized proportional-integral observer.
[0031] As can be seen from Figure 1 the model predictive controller is based on the desired joint angle position vector and velocity vector and the actual joint angle position vector and velocity vector feedback from the actual operation of the spherical electric joint module prototype , by adjusting internal parameters such as the prediction horizon length, weight coefficient, and sampling period, solving the optimal control sequence within a prediction horizon through an optimization algorithm, and selecting the first control element to act on the control process, that is, the control torque applied to the spherical electric joint module .
[0032] The generalized proportional-integral observer is based on the actual operation trajectory and control torque of the spherical electric joint module. By high-order processing of the system state and using the information of each order derivative of the disturbance, it estimates the external disturbance acting on the spherical electric joint module prototype, and feeds back the estimated value of this external disturbance as a feedforward compensation to the model predictive controller, so as to make the actual motion trajectory of the spherical electric joint module prototype as close as possible to the desired trajectory
[0033] The dynamic model of the spherical electric joint module prototype is:
[0034] (1)
[0035] In the formula, , , are the angular position, angular velocity, and angular acceleration vectors of the spherical electric joint module respectively; is the inertia force matrix, is the centrifugal force and Coriolis force matrix, is the gravity vector, is the torque vector, is the unknown external disturbance.
[0036] (1) For the spherical electric joint module prototype without considering external disturbances, the design method of the model predictive controller is as follows:
[0037] Define the state variables of the spherical electric joint module prototype, and the system can obtain the following state equation:
[0038] (2)
[0039] According to the principle of feedback linearization, let , then formula (2) can be equivalent to the following linear system:
[0040] (3)
[0041] Step a: Build a prediction model for the spherical electric joint module prototype:
[0042] Write the prediction model in the following state-space form:
[0043] (4)
[0044] Among them, represents a 3×3 zero matrix, represents a 3×3 identity matrix. Denote the sampling period as T, and discretize formula (4) using the Euler method at time k:
[0045] (5)
[0046] In the formula, , , then the discrete state-space model of the spherical electric joint module prototype can be sorted out as:
[0047] (6)
[0048] In the formula, , ,
[0049] Among them, is the identity matrix.
[0050] Step b: Calculate the predicted output for the spherical electric joint module
[0051] Backward prediction of the future state of the spherical electric joint module prototype for N steps at time k can be obtained as:
[0052] (7)
[0053] Write the predicted output in matrix form:
[0054] (8)
[0055] In the formula, is the predicted output sequence at time k, is the predicted control input sequence, , .
[0056] Step c: Set the reference trajectory .
[0057] Step d: Design the cost function as follows:
[0058] (9)
[0059] In the formula, , Q is the weight of the predicted tracking error, and R is the weight of the control input.
[0060] Step e: Obtain the optimal sequence that satisfies . According to the design principle of model predictive control, take the first element in and apply it to the control process of the predictive control method of the present invention. At each sampling moment, the optimization problem needs to be updated and solved as described above until the entire control process ends.
[0061] (2) The design of the generalized proportional-integral observer for the spherical electric joint module of the present invention is as follows:
[0062] (10)
[0063] The system can be transformed into the following continuous state-space representation form:
[0064] (11)
[0065] In the formula, is the estimated value of the system state , is the adjustment parameter of the observer, , , is the inertia force matrix.
[0066] Let the observation error , then:
[0067] (12)
[0068] After arranging it into matrix form, we can get:
[0069] (13)
[0070] In the formula, , .
[0071] The characteristic polynomial of matrix Y is:
[0072] (14)
[0073] Among them, is the eigenvalue of matrix Y, is the identity matrix.
[0074] The roots of the characteristic polynomial determine the stability of the system. By choosing appropriate gain coefficients such that all the observer poles are located in the left half-plane, the generalized proportional-integral observer can be made input-output bounded stable. At this time, it is assumed that the external disturbance in the spherical electric joint module prototype is twice differentiable, and the second derivative According to it can be obtained that when t approaches infinity, Z = 0, then .
[0075] According to the Hurwitz criterion, a Hurwitz matrix based on the coefficients of the characteristic polynomial is constructed so that the real parts of all eigenvalues are negative. Then, as time goes by, the error decays with time, and the estimated value of the system state will converge to the true value within a finite time, that is: .
[0076] The model predictive controller combines the optimal result obtained from the model predictive control algorithm with the feedforward compensation term of the disturbance to obtain the control torque of the spherical electric joint module prototype as follows:
[0077] (15)
[0078] wherein, is the estimated value of the external disturbance received by the spherical electric joint module prototype by the generalized proportional-integral observer.
[0079] According to the embodiments of the present invention, a model predictive control method for a spherical electric joint module based on a generalized proportional-integral observer is provided. This method is executed by the above-mentioned model predictive control system for a spherical electric joint module based on a generalized proportional-integral observer, and includes the following steps:
[0080] Step 1, using the generalized proportional-integral observer to perform real-time online estimation on the state variables and external disturbances of the spherical electric joint module system, so as to obtain the estimated value of the external disturbance.
[0081] Step 2, feedforward compensate the estimated value of the external disturbance estimated by the generalized proportional-integral observer into the model predictive controller, so as to further reduce the chattering and magnitude of the control input and improve the stability and tracking accuracy of the system. The model predictive controller combines the optimal result obtained from the model predictive control algorithm with the feedforward compensation term of the external disturbance from the generalized proportional-integral observer to obtain the control torque of the spherical electric joint module prototype.
[0082] To verify the effectiveness of the method proposed in the present invention, the method is now applied to a kinetic model for numerical simulation. External disturbances are added during the simulation for testing and compared with the linear sliding mode control algorithm. Among them, the parameters of the model predictive controller are set as follows: prediction time domain , sampling period , weight coefficient matrix of the prediction tracking error , weight coefficient matrix of the control input , the given external disturbance is , and the desired trajectory is set as follows:
[0083] (16)
[0084] Figure 2 , Figure 3 and Figure 4 are the control torque diagrams obtained by using the linear sliding mode controller (LSMC), the model predictive controller (MPC), and the model predictive controller based on the generalized proportional integral observer (MPC-GPIO) in the present invention, respectively. As can be seen from Figure 2 , the control torque generated by the linear sliding mode control algorithm shows obvious high-frequency jitter phenomenon, and the torque value switches rapidly between positive and negative. Figure 3 , Figure 4 After using the model predictive controller, the degree of chattering of the control torque is significantly reduced, and Figure 4 compared with the other two control algorithms, the model predictive control algorithm with the generalized proportional integral observer added has a further reduced fluctuation range of the control torque, a smoother output, and stronger robustness.
[0085] Figure 5 , Figure 6 and Figure 7 are the tracking errors generated when the three joint angles use the linear sliding mode controller, the model predictive controller, and the model predictive controller based on the generalized proportional integral observer in the present invention for trajectory tracking, respectively. All three control algorithms show different degrees of error fluctuations in the trajectory tracking task. The waveform of the linear sliding mode control algorithm shows a relatively large fluctuation range, indicating that there is a large deviation between the actual trajectory and the desired trajectory during the trajectory tracking process. In contrast, the waveforms of the model predictive control algorithm and the model predictive control algorithm with the generalized proportional integral observer added are smoother, with a smaller fluctuation range, showing higher tracking accuracy and stability.
[0086] To more clearly show the advantages and disadvantages of the three controllers in terms of control performance, Figure 8 The mean square error (MSE) obtained by using three control algorithms to perform trajectory tracking on three joint angles respectively is shown. From the variation of the MSE data of the three joint angles, it can be seen that the performance differences of the algorithms shown by different joint angles may vary, but the model predictive control algorithm with a generalized proportional-integral observer always shows the lowest MSE value. This indicates that the control algorithm proposed by the present invention has significant advantages in reducing trajectory tracking errors, can more flexibly handle various trajectory tracking tasks, has significant anti-interference ability, and its high precision, stability and adaptability make it an ideal choice for designing a high-precision trajectory tracking control system.
[0087] In summary, the control algorithm proposed by the present invention can achieve precise tracking of the motion trajectory of the spherical electric joint module. When complex situations such as external disturbances occur, it can still maintain a stable trajectory tracking effect, make the motion trajectory smoother, reduce jitter and mutation phenomena, and greatly improve the reliability and stability of the system.
[0088] The present invention proposes a model predictive control strategy based on a generalized proportional-integral observer for a spherical electric joint module. By using a generalized proportional-integral observer, the disturbance is extended as a state of the system for estimation. In cooperation with the model predictive control strategy, a prediction model is constructed to predict the motion state of the spherical electric joint module within a future period of time, and an optimal control instruction sequence is calculated according to the prediction results. The disturbance estimation value obtained by the generalized proportional-integral observer is fed forward and compensated to the model predictive controller. By utilizing the accurate estimation ability of the observer for the system state and external disturbances, the steady-state error of the system and the fluctuation range of the control torque are reduced, and the accuracy and robustness of the control algorithm are improved.
[0089] In the specification provided herein, a large number of specific details are described. However, it can be understood that the embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and technologies are not shown in detail so as not to obscure the understanding of this specification.
[0090] Although the present invention is described in terms of a limited number of embodiments, those skilled in the art in this technical field will understand, based on the above description, that other embodiments can be envisioned within the scope of the present invention thus described. In addition, it should be noted that the language used in this specification is mainly selected for the purpose of readability and teaching, and is not selected for the purpose of interpreting or limiting the subject matter of the present invention.
Claims
1. A spherical electric joint module model predictive control system based on a generalized proportional integral observer, characterized in that: include: Generalized proportional-integral observer, model predictive controller, spherical electric joint module prototype, wherein the generalized proportional-integral observer is used to perform real-time online estimation of the state variables and external disturbances of the spherical electric joint module prototype, thereby obtaining an estimated value of the external disturbance, and the estimated value of the external disturbance obtained by the generalized proportional-integral observer is feedforward compensated to the model predictive controller. The model predictive controller is used to obtain the control torque of the spherical electric joint module prototype based on the optimization result obtained by the model predictive control algorithm and the feedforward compensation term of the external disturbance from the generalized proportional-integral observer.
2. The spherical electric joint module model predictive control system based on generalized proportional integral observer according to claim 1, characterized in that: The model predictive controller is constructed through the state space equation of the spherical electric joint module prototype, wherein the model predictive controller is used to predict the state of the spherical electric joint module prototype in the future based on the current position and speed information of the spherical electric joint module prototype, and design a cost function according to the control objective of the model predictive controller. The value of the cost function is minimized by solving the optimization problem, thereby obtaining the optimal control sequence at the current moment, and the first control quantity in the solved optimal control sequence is applied to the spherical electric joint module prototype. The above steps are repeated to make the spherical electric joint module prototype move along the desired joint angle position.
3. The spherical electric joint module model predictive control system based on generalized proportional integral observer according to claim 2, characterized in that: The generalized proportional-integral observer estimates the external disturbance by expanding it into the state of the spherical electric joint module prototype, and adjusts the corresponding observer order according to the order of the disturbance model to estimate the time-varying disturbance, and feed-forward compensates the time-varying disturbance into the model predictive controller.
4. The spherical electric joint module model predictive control system based on generalized proportional integral observer according to claim 2, characterized in that: The process of model predictive control algorithm includes: Step a: Build a prediction model: Define the state variables of the spherical electric joint module prototype , according to the feedback linearization principle, let , the prediction model is written in the following state space form: ， In the formula, , , They are the angular position, angular velocity and angular acceleration vector of the spherical electric joint module; is the inertia force matrix, is the centrifugal force and Coriolis force matrix, is the gravity vector, is the torque vector, represents a zero matrix of 3 rows and 3 columns, Represents the identity matrix with 3 rows and 3 columns; The sampling period is T, and the Euler method is used for discretization at time k. The discrete state space model of the spherical electric joint module prototype is organized as follows: ， In the formula, , ; in, It is T times the identity matrix with 3 rows and 3 columns; Step b: Calculate the predicted output: At time k, the future state of the spherical electric joint module prototype is predicted N steps backward to obtain: ， Write the prediction output in matrix form: ， In the formula, is the predicted output sequence at time k, For predictive control input sequence, , ; Step c: Set the reference trajectory ; Step d: Design the cost function as follows: ， In the formula, , Q is the weight of the predicted tracking error, and R is the weight of the control input; Step e: Find the satisfaction The optimal sequence ,Pick The first element in Acting on the spherical electric joint module prototype; At each sampling moment, execute the above steps be until the entire control process is completed.
5. The spherical electric joint module model predictive control system based on generalized proportional integral observer according to claim 4, characterized in that: The design form of the generalized proportional integral observer is as follows: ， In the formula, This is the state of the spherical electric joint module prototype The estimated value of is the adjustment parameter of the generalized proportional-integral observer, , , is the inertia force matrix.
6. The spherical electric joint module model predictive control system based on generalized proportional integral observer according to claim 5, characterized in that: By choosing a suitable gain factor , so that the observer poles are all located in the left half plane.
7. The spherical electric joint module model predictive control system based on generalized proportional integral observer according to claim 5, characterized in that: The model predictive controller obtains the optimal result obtained by the model predictive control algorithm and combines the feedforward compensation term of the disturbance from the generalized proportional integral observer to obtain the control torque of the spherical electric joint module prototype, as shown below: ， In the formula, It is the estimated value of the external disturbance on the spherical electric joint module prototype by the generalized proportional integral observer.
8. A spherical electric joint module model predictive control method based on a generalized proportional integral observer, characterized in that: The method is performed by a spherical electric joint module model predictive control system based on a generalized proportional integral observer according to any one of claims 1 to 7, and comprises the following steps: Step 1, using a generalized proportional integral observer, the state variables and external disturbances of the spherical electric joint module system are estimated online in real time, thereby obtaining an estimated value of the external disturbance; In step 2, the estimated value of the external disturbance obtained by the generalized proportional-integral observer is feed-forwarded into the model predictive controller. The model predictive controller obtains the optimization result of the model predictive control algorithm and combines the feed-forward compensation term of the external disturbance from the generalized proportional-integral observer to obtain the control torque of the spherical electric joint module prototype.
9. The spherical electric joint module model predictive control method based on generalized proportional integral observer according to claim 8, characterized in that: The model predictive controller is constructed through the state space equation of the spherical electric joint module prototype, wherein the model predictive controller predicts the state of the spherical electric joint module prototype in the future based on the current position and speed information of the spherical electric joint module prototype, and designs a cost function based on the control objective of the model predictive controller. The value of the cost function is minimized by solving the optimization problem, thereby obtaining the optimal control sequence at the current moment, and the first control quantity in the solved optimal control sequence is applied to the spherical electric joint module prototype. The above steps are repeated to make the spherical electric joint module prototype move along the desired joint angle position.
10. The spherical electric joint module model predictive control method based on generalized proportional integral observer according to claim 8, characterized in that: The generalized proportional-integral observer estimates the external disturbance by expanding it into the state of the spherical electric joint module prototype, and adjusts the corresponding observer order according to the order of the disturbance model to estimate the time-varying disturbance, and feed-forward compensates the time-varying disturbance into the model predictive controller.