Reconfigurable robot control method and system for zero-sum game optimization event triggering mechanism

Abstract

The invention discloses a reconfigurable robot control method and system for a zero-sum game optimization event triggering mechanism. The method comprises the steps that a reconfigurable robot dynamic model based on joint torque feedback is established, an error dynamic state space is established, and a preset performance function is set to restrain the transient performance of the error dynamic state space; the error dynamics state space constraining the transient performance is mapped to the unconstrained error dynamics augmented system state space, and an event triggering mechanism is introduced, so that the unconstrained error dynamics augmented system state space meets the zero-sum game of maximizing the event triggering interval and minimizing the control input; setting a performance index function based on a zero-sum game, and obtaining an optimal control law under an event triggering mechanism through the performance index function; according to the invention, the precision and stability of robot system control are improved through the predetermined performance function; the zero-sum game of an event triggering mechanism is introduced to obtain an optimal control law under a Nash equilibrium condition, so that communication resources and calculation burdens are effectively reduced; and the control torque of each subsystem is smoother under the action of the optimal control law, so that the energy consumption is reduced and the service life of the motor is prolonged.

Description

Technical Field

[0001] This invention relates to the field of robot control, and in particular to a reconfigurable robot control method and system based on a zero-sum game-optimized event triggering mechanism. Background Technology

[0002] With the rapid development of artificial intelligence, computer technology, and control theory, robot systems have been widely applied in industrial production, daily life services, disaster relief, and maritime and aerial exploration. In particular, the emergence of AI-based strategy generation methods has enabled robots to exhibit greater autonomy and intelligence. However, the control precision of existing intelligent robots, both domestically and internationally, is generally lower than that of robots using traditional control schemes. Since control methods are the supporting technologies for robots to execute strategies, achieving high intelligence in robots requires significantly improving the accuracy and stability of robot controllers. Simultaneously, for reconfigurable robots typically used in complex environments, communication resources and computational burden must also be considered.

[0003] Optimal control, based on the Pontryagin maximum principle, is a modern control theory that seeks to optimize one or more indices during a dynamically evolving control process. For complex nonlinear systems such as reconfigurable robots, the "curse of dimensionality" problem exists in solving the Hamiltonian-Jacobi-Essex partial differential equations to obtain the optimal control law, making analytical solutions difficult to obtain. Adaptive dynamic programming algorithms can approximate the cost function using structures such as neural networks to solve the aforementioned partial differential equations, thereby obtaining an optimal control law that simultaneously optimizes error and energy consumption. Summary of the Invention

[0004] The purpose of this invention is to provide a reconfigurable robot control method and system based on a zero-sum game-based optimized event-triggered mechanism. The event-triggered mechanism is a control strategy that triggers controller actions based on specific events by monitoring the system state and set thresholds, saving communication resources and reducing computational burden. Traditional event-triggered mechanisms involve setting a trigger threshold and constantly monitoring trigger conditions; when the trigger condition reaches the threshold, the control law is calculated and the control input is updated. This invention treats the trigger threshold and control strategy as participants in a two-player zero-sum differential game, seeking stable optimal control under the conditions of maximum trigger threshold and minimum control law. Its advantage lies in that this variable threshold optimization method can significantly reduce the number of controller updates under conditions of system stability and energy saving. Pre-defined performance control is a control technique that allows for the pre-setting of transient performance. By designing a pre-defined performance function for constraint representation and exploring the mapping relationship of the state space, the constrained state space is mapped to an unconstrained augmented system state space. Optimal control based on the event-triggered mechanism can then be performed in the mapped state space. The control action satisfies the pre-defined constraints through the mapping relationship, thus solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a reconfigurable robot control method based on a zero-sum game-optimized event triggering mechanism, comprising:

[0006] A reconfigurable robot dynamics model based on joint torque feedback is established, an error dynamics state space is created, and a predetermined performance function is set to constrain the transient performance of the error dynamics state space.

[0007] The error dynamics state space constrained by transient performance is mapped to the state space of the unconstrained error dynamics augmented system. An event triggering mechanism is introduced to make the state space of the unconstrained error dynamics augmented system satisfy a zero-sum game that maximizes the event triggering interval and minimizes the control input.

[0008] Based on a zero-sum game theory approach, a performance index function is set up. Through this performance index function, the optimal control law under the event-triggered mechanism is obtained.

[0009] Furthermore, based on the reconfigurable robot dynamics model using joint torque feedback, an error dynamics state space is established, and a predetermined performance function is set to constrain the transient performance of the error dynamics state space, including:

[0010] Generate the state space of a reconfigurable robot system based on a dynamic model;

[0011] When the tracking error approaches zero, an error dynamics state space is established based on the state space of the reconfigurable robot system;

[0012] A predetermined performance function is constructed by limiting the steady-state error, maximum overshoot, and convergence rate, and the upper and lower limits of the predetermined performance function are defined to constrain the range of transient performance.

[0013] Furthermore, the error dynamics state space constrained by transient performance is mapped to the state space of the unconstrained error dynamics augmented system, including:

[0014] A nonlinear error space transformation function is constructed based on transient performance and the upper and lower bounds of a predetermined performance function.

[0015] Based on the transformation function between the error dynamics state space and the nonlinear error space, the state space of the unconstrained error dynamics augmented system is constructed.

[0016] Furthermore, an event-triggered mechanism is introduced to make the state space of the unconstrained error dynamics augmented system satisfy a zero-sum game that maximizes the event trigger interval and minimizes the control input. This also includes:

[0017] The difference between the second control law and the first control law is set as the triggering error. The control law that is time-triggered in the state space of the unconstrained error dynamic augmented system before the introduction of the time triggering mechanism is defined as the first control law, and the control law that is time-triggered in the state space of the unconstrained error dynamic augmented system after the introduction of the time triggering mechanism is defined as the second control law.

[0018] The triggering error is introduced into the unconstrained error dynamics based on the time triggering mechanism to augment the system state space, and the triggering error is maximized while the second control law is minimized, so as to achieve a zero-sum game.

[0019] Furthermore, a performance index function is set based on a zero-sum game. Through this function, the optimal control law and the optimal event triggering threshold law under the event triggering mechanism are obtained. The event triggering conditions are then obtained based on the optimal event triggering threshold law, including:

[0020] The Hamilton-Jacobi-Essex equations are obtained based on the performance index function and the error dynamics state space.

[0021] Based on the maximum principle and the Hamilton-Jacobi-Essex equation, the optimal control law and the optimal event triggering threshold law under the event triggering mechanism in Nash equilibrium are obtained.

[0022] Furthermore, an evaluation network is established through adaptive dynamic programming;

[0023] Based on the performance index function and the evaluation network, an approximate optimal performance index function is obtained.

[0024] Based on the approximation optimal performance index function and the optimal control law, solve for the approximation optimal control law.

[0025] Furthermore, the optimal event triggering threshold law is obtained through the performance index function, and the event triggering conditions are obtained based on the optimal event triggering threshold law;

[0026] Define the error used for neural network updates, and obtain the weight update law for the optimal control law;

[0027] Based on event triggering conditions, a weight update law is applied in the state space of the unconstrained error dynamic augmented system, so that the optimal control law is updated in real time.

[0028] Furthermore, the event triggering conditions include:

[0029] If the event triggering error is less than the optimal event triggering threshold law, then the optimal control law from the previous moment will be used.

[0030] If the event triggering error is greater than the optimal event triggering threshold law, then the optimal control law is updated based on the weight update law.

[0031] Furthermore, based on the reconfigurable robot dynamics model using joint torque feedback, an error dynamics state space is established, including:

[0032] A reconfigurable robot dynamic model of the subsystem is constructed based on the reconfigurable robot's moment of inertia, acceleration ratio, friction, and the cross-coupling terms between the reconfigurable robot subsystems.

[0033] Based on the dynamics model of the reconfigurable robot, establish the actual state space of the reconfigurable robot system;

[0034] Establish the reference state space of the reconfigurable robot system based on the reconfigurable robot dynamics model and reference signals;

[0035] When the tracking error approaches zero, an error dynamic state space is constructed based on the actual state space and the reference state space.

[0036] On the other hand, a reconfigurable robot control system based on a zero-sum game optimization event triggering mechanism is provided. The reconfigurable robot control method applying the zero-sum game optimization event triggering mechanism described above is characterized by comprising:

[0037] The model building module is used to establish the error dynamics state space and set a predetermined performance function to constrain the transient performance of the error dynamics state space for the reconfigurable robot dynamics model based on joint torque feedback.

[0038] The zero-sum game module maps the error dynamics state space constrained by transient performance to the state space of the unconstrained error dynamics augmented system, and introduces an event triggering mechanism to make the state space of the unconstrained error dynamics augmented system satisfy a zero-sum game that maximizes the event triggering interval and minimizes the control input.

[0039] The optimal control law generation module sets a performance index function based on zero-sum game theory, and obtains the optimal control law under the event-triggered mechanism through the performance index function.

[0040] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention provides a technical path for optimal trajectory tracking of reconfigurable robots, enabling optimized control performance; from the perspective of tracking performance, this invention can pre-set steady-state error, convergence rate, and maximum overshoot through predetermined performance functions, thereby improving the accuracy and stability of robot system control; from the perspective of triggering frequency, this invention introduces an event triggering mechanism and treats the event triggering threshold and control law as participants in a zero-sum game, thus obtaining the optimal control law and optimal triggering threshold law under Nash equilibrium conditions, effectively reducing communication resources and computational burden; from the perspective of control torque, the control torque of each subsystem under the action of this invention is smoother, which can reduce energy consumption and extend motor life; therefore, this invention improves the control performance of reconfigurable robots compared to existing technologies, achieving higher accuracy. Attached Figure Description

[0041] Figure 1 This is a flowchart of a reconfigurable robot control method based on a zero-sum game optimization event triggering mechanism in an embodiment of the present invention.

[0042] Figure 2 This is a connection block diagram of a reconfigurable robot control system with a zero-sum game optimization event triggering mechanism in an embodiment of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] In the description of this invention, unless otherwise stated, "a plurality of" means two or more; the terms "upper," "lower," "left," "right," "inner," "outer," "front end," "rear end," "head," "tail," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0045] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0046] Please refer to the accompanying drawings in the specification. This invention provides a technical solution: such as... Figure 1 As shown, a reconfigurable robot control method based on a zero-sum game optimization event triggering mechanism includes the following steps:

[0047] S102. Based on joint torque feedback, a reconfigurable robot dynamics model is established, an error dynamics state space is created, and a predetermined performance function is set to constrain the transient performance of the error dynamics state space.

[0048] Optionally, a reconfigurable robot dynamics model of the subsystem is constructed based on the reconfigurable robot's moment of inertia, acceleration ratio, friction, and the cross-linking coupling terms between the reconfigurable robot subsystems; the actual state space of the reconfigurable robot system is established based on the reconfigurable robot dynamics model; the reference state space of the reconfigurable robot system is established based on the reconfigurable robot dynamics model and the reference signal; and when the tracking error approaches zero, an error dynamics state space is constructed based on the actual state space and the reference state space.

[0049] In the above embodiments, the reconfigurable robot dynamics model based on joint torque feedback is as follows:

[0050] ——Formula (1);

[0051] in, Let be the angle, angular velocity, and angular acceleration of the i-th joint subsystem, respectively. These are the moment of inertia and deceleration ratio of the i-th joint subsystem, respectively; , These are the joint feedback torque and control torque of the i-th joint subsystem, respectively; Let be the position, velocity, and acceleration vectors of the overall robot system, respectively; the frictional torque of the i-th joint subsystem is described as a nonlinear function of the joint position and velocity.

[0052] ——Formula (2);

[0053] in, , , Let represent the viscous friction, Coulomb friction, static friction, and Stribect friction of the i-th joint subsystem, respectively. This represents position-dependent friction of the i-th joint subsystem;

[0054] ——Formula (3);

[0055] Linearizing the above friction yields:

[0056] ——Formula (4);

[0057] In the formula, ——Formula (4-1);

[0058] in, , , and Let represent the estimated values ​​of viscous friction, Coulomb friction, static friction, and Stribect friction of the i-th joint subsystem, respectively. For an uncertain coefficient of friction, This is the velocity error matrix.

[0059] The cross-linking coupling term of the i-th joint subsystem in Equation (1) can be expressed as:

[0060] ——Formula (5);

[0061] in, , and These are the unit vectors of the i-th joint and the m-th and k-th joints adjacent to it around the rotation axis, respectively.

[0062] Optionally, the state space of the reconfigurable robot system is generated based on the dynamic model; when the tracking error approaches zero, an error dynamic state space is established based on the state space of the reconfigurable robot system; a predetermined performance function is constructed by limiting the steady-state error, maximum overshoot and convergence rate, and the upper and lower limits of the predetermined performance function are defined to constrain the range of transient performance.

[0063] In the above embodiments, equation (1) can be rearranged to obtain:

[0064] ——Formula (6);

[0065] In the formula, ;

[0066] Define the state-space description of a reconfigurable robot system as follows:

[0067] ——Formula (7);

[0068] In the formula, select It is a state variable.

[0069] ——Formula (7.1);

[0070] ——Formula (7.1.1);

[0071] ——Formula (7.1.2);

[0072] in, For accurate measurement, the system model is the system drift dynamics model. This represents the uncertainty of the global model.

[0073] Define reference signal Its state-space description in terms of dynamics:

[0074] ——Formula (8);

[0075] in, ——Formula (8.1);

[0076] In the formula, For reference control, when the tracking error approaches 0, Approaching 0.

[0077] make Combining (7) and (8), the state-space description of the error dynamics model is established:

[0078] ——Formula (9);

[0079] Design predetermined performance functions to constrain transient performance (mainly including steady-state error, maximum overshoot, and convergence rate) by designing upper and lower bound functions to constrain the error state.

[0080] Design the following predetermined performance function:

[0081] ——Formula (10);

[0082] To achieve predetermined performance control, upper and lower bound error constraints are defined, including the predetermined performance function:

[0083] ——Formula (11);

[0084] ——Formula (11.1);

[0085] ——Formula (11.2);

[0086] in, ——Formula (12);

[0087] In the formula, The upper limit of the predetermined performance function, This is the lower bound of the predetermined performance function; It is a user-defined constant. ;

[0088] Among the above constraints, and as well as Together, they are used to limit overshoot and steady-state error; Used to limit the convergence rate; used to constrain the upper and lower limits of the predetermined performance function of the error system via parameters. adjust.

[0089] S104. Map the state space of the error dynamics constrained by transient performance to the state space of the unconstrained error dynamics augmented system, and introduce an event triggering mechanism to make the state space of the unconstrained error dynamics augmented system satisfy a zero-sum game that maximizes the event triggering interval and minimizes the control input.

[0090] Optionally, a nonlinear error space transformation function is constructed based on transient performance and the upper and lower bounds of a predetermined performance function; and the state space of an unconstrained error dynamic augmented system is constructed based on the error dynamic state space and the nonlinear error space transformation function.

[0091] In the above embodiments, the following nonlinear error space transformation function is introduced:

[0092] ——Formula (13);

[0093] In the formula: ——Formula (13.1);

[0094] ——Formula (13.2);

[0095] To obtain the augmented system state space, we differentiate equation (13):

[0096] ——Formula (14);

[0097] In the formula: ——Formula (14.1);

[0098] ——Formula (14.2);

[0099] The augmented system state space is designed as follows:

[0100] ——Formula (15);

[0101] In the formula: ——Formula (15.1);

[0102] ——Formula (15.2);

[0103] in, It is the augmented system state vector. It is an augmented system state-space control strategy;

[0104] ——Formula (15.3);

[0105] in, It is an adjustable parameter in the predetermined performance function.

[0106] Optionally, the difference between the second control law and the first control law is set as the trigger error. The control law that is time-triggered in the state space of the unconstrained error dynamic augmented system before the introduction of the time trigger mechanism is defined as the first control law, and the control law in the state space of the unconstrained error dynamic augmented system after the introduction of the time trigger mechanism is defined as the second control law. The trigger error is introduced into the state space of the unconstrained error dynamic augmented system based on the time trigger mechanism, and the trigger error is maximized while the second control law is minimized to achieve a zero-sum game.

[0107] In the above embodiments, the performance index function of the continuous augmentation system of the i-th sub-joint is typically expressed as follows:

[0108] ——Formula (16);

[0109] In the formula: ——Formula (16.1);

[0110] in, Indicates the initial state. It is a normal number.

[0111] Defined as a monotonically increasing sequence at the trigger time ,in, The state at the sampling time updated by the controller is defined as , Then, the system (15) under the event-triggered mechanism can be written as:

[0112] ——Formula (17);

[0113] The difference between the control law at the triggering moment in the event-triggered mechanism and the control law under time-triggered mechanism is defined as the event-triggered error:

[0114] ——Formula (18);

[0115] Therefore, by introducing the event triggering error into the model, and further refining equation (17) according to equation (18), we can obtain:

[0116] ——Formula (19);

[0117] Equation (19) serves as a mapping system for the actual robot system in an unconstrained augmented state space. There exists a zero-sum game relationship in it, which means that under the premise of satisfying performance requirements and control stability, it is desirable to maximize the event trigger interval and minimize the control input.

[0118] S106. Based on zero-sum game theory, set a performance index function, and obtain the optimal control law under the event-triggered mechanism through the performance index function.

[0119] In order to represent this zero-sum game relationship in the performance index function, the performance index function of system (19) is defined as:

[0120] ——Formula (20);

[0121] in, It is a normal number.

[0122] Optionally, the Hamilton-Jacobi-Essex equations are obtained based on the performance index function and the error dynamics state space; the optimal control law and the optimal event triggering threshold law under the event triggering mechanism in Nash equilibrium are obtained based on the maximum principle and the Hamilton-Jacobi-Essex equations.

[0123] In the above embodiments, the Hamilton-Jacobi-Essex equations are written, and the optimal control law and optimal event triggering threshold law under the event triggering mechanism are obtained, ultimately determining the event triggering conditions. According to equations (19) and (20), the Hamiltonian function is defined as:

[0124] ——Formula (21);

[0125] Based on the maximum principle, the combined Hamilton-Jacobi-Essex equations are expressed as follows:

[0126] ——Formula (22);

[0127] in, Then, according to the equilibrium condition The optimal control law and the optimal event triggering threshold law under Nash equilibrium can be derived:

[0128] ——Formula (23);

[0129] ——Formula (24);

[0130] Substituting equations (23) and (24) into (22) yields the optimal Hamilton-Jacobi-Essex equation under event triggering:

[0131] ——Formula (25);

[0132] Optionally, the method further includes step S108: establishing an evaluation network through adaptive dynamic programming; obtaining an approximate optimal performance index function based on the performance index function and the evaluation network; and solving the approximate optimal control law based on the approximate optimal performance index function and the optimal control law.

[0133] In the above embodiments, an evaluation network based on an RBF network is defined:

[0134] ——Formula (26);

[0135] in, Indicates the weight. Represents radial basis functions. The derivative of the approximation error is defined as:

[0136] ——Formula (27);

[0137] Therefore, the approximate optimal control strategy can be obtained:

[0138] ——Formula (28);

[0139] The estimated Hamiltonian equation can be written as:

[0140] ——Formula (29);

[0141] Optionally, the method further includes step S110: obtaining the optimal event triggering threshold law through the performance index function, and obtaining the event triggering condition based on the optimal event triggering threshold law; defining the error for neural network update, and obtaining the weight update law of the optimal control law; and applying the weight update law in the state space of the unconstrained error dynamic augmented system based on the event triggering condition, so that the optimal control law is updated in real time.

[0142] In the above embodiments, the error used for neural network updates is defined. The weight estimation error can be expressed as And thus Minimize the objective function based on gradient descent Therefore, the weights are based on... Update. The weight update law can be further written as:

[0143] ——Formula (30);

[0144] Finally, based on the weight update law, in the unconstrained state... The optimal control law applied in the state space is updated in real time, and constrained control is applied to the actual physical system according to the mapping relationship.

[0145] Furthermore, the event triggering conditions include: if the event triggering error is less than the optimal event triggering threshold law, then the optimal control law of the previous moment is used; if the event triggering error is greater than the optimal event triggering threshold law, then the optimal control law is updated based on the weight update law.

[0146] In the above embodiments, the event triggering condition is expressed as follows:

[0147] ——Formula (31);

[0148] If the trigger condition is less than the trigger threshold, the control law from the previous moment is used; otherwise, the control law is updated, thereby reducing the error and bringing the trigger condition down below the threshold.

[0149] On the other hand, such as Figure 2 As shown, a reconfigurable robot control system based on a zero-sum game optimization event triggering mechanism is provided. The reconfigurable robot control method applying the zero-sum game optimization event triggering mechanism described above includes:

[0150] The model building module is used to establish the error dynamics state space and set a predetermined performance function to constrain the transient performance of the error dynamics state space for the reconfigurable robot dynamics model based on joint torque feedback.

[0151] The zero-sum game module maps the error dynamics state space constrained by transient performance to the state space of the unconstrained error dynamics augmented system, and introduces an event triggering mechanism to make the state space of the unconstrained error dynamics augmented system satisfy a zero-sum game that maximizes the event triggering interval and minimizes the control input.

[0152] The optimal control law generation module sets a performance index function based on zero-sum game theory, and obtains the optimal control law under the event-triggered mechanism through the performance index function.

[0153] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A reconfigurable robot control method of a zero-sum game optimization event triggering mechanism, characterized in that, The method comprises the following steps: a reconfigurable robot dynamics model based on joint torque feedback is established, an error dynamics state space is established, and a predetermined performance function is set to constrain the transient performance of the error dynamics state space; the error dynamics state space with the constrained transient performance is mapped to an unconstrained error dynamics augmented system state space, and an event-triggered mechanism is introduced to make the unconstrained error dynamics augmented system state space satisfy a zero-sum game of maximizing an event-triggered interval and minimizing a control input; a performance index function is set based on the zero-sum game, and an optimal control law under the event-triggered mechanism is obtained through the performance index function.

2. The reconfigurable robot control method of claim 1, wherein, The method comprises the following steps: a reconfigurable robot dynamics model based on joint torque feedback is established, an error dynamics state space is established, and a predetermined performance function is set to constrain the transient performance of the error dynamics state space, including: a state space of a reconfigurable robot system is generated based on the dynamics model; when a tracking error approaches zero, an error dynamics state space is established based on the state space of the reconfigurable robot system; 3. The reconfigurable robot control method of claim 2, wherein, a predetermined performance function is constructed by limiting a steady-state error, a maximum overshoot and a convergence rate, and an upper limit and a lower limit of the predetermined performance function are defined to constrain the range of transient performance. the error dynamics state space with the constrained transient performance is mapped to an unconstrained error dynamics augmented system state space, including: a nonlinear error space conversion function is constructed based on the transient performance and the upper limit and the lower limit of the predetermined performance function; 4. The reconfigurable robot control method of claim 3, wherein, an unconstrained error dynamics augmented system state space is constructed according to the error dynamics state space and the nonlinear error space conversion function. An event-triggered mechanism is introduced to make the unconstrained error dynamics augmented system state space satisfy a zero-sum game of maximizing an event-triggered interval and minimizing a control input, and the method further comprises the following steps: a trigger error is set as a difference between a second control law and a first control law, wherein the control law of the unconstrained error dynamics augmented system state space before the event-triggered mechanism is introduced is defined as the first control law, and the control law of the unconstrained error dynamics augmented system state space after the event-triggered mechanism is introduced is defined as the second control law; 5. The reconfigurable robot control method of a zero-sum game optimization event triggering mechanism according to claim 4, wherein, the trigger error is introduced into the unconstrained error dynamics augmented system state space based on the time-triggered mechanism, and the trigger error is maximized and the second control law is minimized to realize the zero-sum game. A performance index function is set based on the zero-sum game, and an optimal control law under the event-triggered mechanism and an optimal event-triggered threshold law are obtained through the performance index function, and an event-triggering condition is obtained according to the optimal event-triggered threshold law, including: a Hamilton-Jacobi-Isaacs equation is obtained according to the performance index function and the error dynamics state space; 6. The reconfigurable robot control method of zero-sum game optimization event triggering mechanism according to claim 1, characterized in that, the optimal control law under the event-triggered mechanism and the optimal event-triggered threshold law under the Nash equilibrium are obtained according to the maximum principle and the Hamilton-Jacobi-Isaacs equation. The method comprises the following steps: a critic network is established through adaptive dynamic programming; an approximate optimal performance index function is obtained based on the performance index function and the critic network; 7. The reconfigurable robot control method of claim 4, wherein, an approximate optimal control law is solved according to the approximate optimal performance index function and the optimal control law. The method comprises the following steps: An optimal event-triggered threshold law is obtained by a performance index function, and an event-triggered condition is obtained according to the optimal event-triggered threshold law; An error for neural network updating is defined, and a weight updating law of the optimal control law is obtained; Based on the event-triggered condition, the weight updating law is applied in the unconstrained error dynamics augmented system state space, so that the optimal control law is updated in real time.

8. The reconfigurable robot control method of a zero-sum game optimization event triggering mechanism according to claim 7, characterized in that, The event-triggered condition includes: If the event-triggered error is less than the optimal event-triggered threshold law, the optimal control law of the last moment is used; If the event-triggered error is greater than the optimal event-triggered threshold law, the optimal control law is updated based on the weight updating law.

9. The reconfigurable robot control method of claim 1, wherein, Based on the joint torque feedback of the reconfigurable robot dynamics model, an error dynamics state space is established, including: According to the moment of inertia, acceleration ratio, friction and cross-coupling terms between the subsystems of the reconfigurable robot, a reconfigurable robot dynamics model of the subsystem is constructed; According to the reconfigurable robot dynamics model, an actual state space of the reconfigurable robot system is established; According to the reconfigurable robot dynamics model and the reference signal, a reference state space of the reconfigurable robot system is established; When the tracking error tends to zero, the error dynamics state space is constructed according to the actual state space and the reference state space.

10. A reconfigurable robot control system with a zero-sum game optimization event-triggered mechanism, applying the reconfigurable robot control method with a zero-sum game optimization event-triggered mechanism according to any one of claims 1 to 9, characterized in that, It includes: A model construction module is configured to establish an error dynamics state space based on a joint torque feedback of a reconfigurable robot dynamics model, and set a predetermined performance function to constrain transient performance of the error dynamics state space; A zero-sum game module is configured to map the error dynamics state space with constrained transient performance to an unconstrained error dynamics augmented system state space, introduce an event-triggered mechanism, and make the unconstrained error dynamics augmented system state space satisfy a zero-sum game of maximizing event-triggered intervals and minimizing control inputs; An optimal control law generation module is configured to set a performance index function based on the zero-sum game, and obtain an optimal control law under the event-triggered mechanism through the performance index function.

Images