Multi-agent consensus control method and control system
By utilizing the sliding sector characteristics and a fixed-time control protocol with nonlinear compensation terms, combined with a disturbance observer, high-precision synchronous control of a multi-agent system is achieved. This solves the problems of slow response speed and difficulty in verifying stability in traditional methods, meeting the rapid response requirements of industry.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGSU STARING TECHNOLOGY CO LTD
- Filing Date
- 2026-03-06
- Publication Date
- 2026-06-05
AI Technical Summary
Traditional multi-agent consistency control is difficult to achieve rapid response and intuitive stability verification. The stability analysis of existing fixed-time consistency control for motors is complex and cannot meet the needs of industrial scenarios.
A fixed-time control protocol is designed using sliding sector characteristics. Combined with nonlinear compensation terms, the input control signal and the disturbance observer state vector are adjusted by the error value to achieve closed-loop feedback control. A third-order fixed-time observer is used to estimate the load disturbance.
It achieves high-precision synchronous control of multiple agents, quickly tracks the leader's expected parameters, has strong anti-interference ability, and the convergence time can be predetermined, meeting the time-scheduling requirements of industrial scenarios.
Smart Images
Figure CN122151518A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent agent control technology, and in particular to a method and system for synchronous control of multiple intelligent agents. Background Technology
[0002] Multi-agent system consensus control aims to converge the states of all agents to the same value or track the leader state through local information interaction. It has been widely used in ground vehicle formation, missile cooperative guidance, and multi-motor synchronization. Traditional consensus control often achieves asymptotic convergence, and the convergence speed depends on the initial conditions, making it difficult to meet the rapid response requirements in industrial scenarios.
[0003] Moreover, existing stability analyses of motor fixed-time consistency control mostly rely on complex phase diagrams, making it difficult to intuitively verify system stability.
[0004] There is an urgent need for a multi-agent consensus control method and control system that can solve the above problems. Summary of the Invention
[0005] The purpose of this invention is to provide a multi-agent consensus control method and control system that can achieve high-precision synchronous control of multiple agents.
[0006] To achieve the above objectives, this invention provides a multi-agent consensus control method, comprising: having n agents, where n is greater than or equal to 2; initializing the agent parameters of the agents; and setting the expected parameters of the leader. Collect the current action parameters of the intelligent agent. Based on the action parameters of the i-th agent Action parameters of the j-th agent Calculate the error value of the i-th agent using the expected parameter x0. Let i = 1, 2, ..., n, j = 1, 2, ..., n and Error value ; where, it is determined whether a communication link exists between the i-th agent and the j-th agent; if so... =1, if not =0; Determine whether the i-th agent has obtained the expected parameter x0. If yes... =1, if not =0; based on the error value of the i-th agent. Adjust the input control signal of the i-th agent , Let i be the action parameters of the i-th agent. Let be the action parameters of the j-th agent.
[0007] Preferably, the desired parameter Including expected mechanical value and mechanical value Expected speed Action parameters of the i-th agent Including the mechanical value of the i-th agent and mechanical speed Action parameters of the j-th agent Including the mechanical value of the j-th agent and mechanical speed The error value of the i-th agent Including the mechanical error value of the i-th agent and mechanical speed error value ; ; .
[0008] More preferably, based on the error value of the i-th agent Adjust the input control signal of the i-th agent Including: based on preset control parameters and error values Obtain the corresponding error control signal Based on the error control signal Adjust the input control signal of the i-th agent Error control signal , , , , These are preset control parameters. The following conditions must be met: ,in This scheme uses the boundary of the sliding sector to define the control parameters, ensuring that the system trajectory has no singular points within the sector, that is, the system operates stably when error control is performed.
[0009] Specifically, control parameters , The following conditions must be met: , , , The parameters of the first layer are defined by the stability formula, and the parameters of the second layer are set by the parameters of the first layer, so that the entire error control signal... When stability is achieved, it is used to adjust the input control signal, which also makes the entire adjustment process stable and enables the agent to quickly track the leader's desired angle, allowing the agent to quickly reach the target's desired parameters.
[0010] Preferably, the desired parameters Also includes expected acceleration Based on the error value of the i-th agent Adjust the input control signal of the i-th agent At the same time, it is also based on the expected acceleration. Compensation Expected Acceleration Term The input control signal of the i-th agent is adjusted based on the compensated acceleration term. This approach allows the agent's action parameters to quickly reach the desired values.
[0011] Preferably, each agent is equipped with an observer, which is used to observe the perturbation factors of the i-th agent and estimate the load perturbation of the i-th agent. Based on the error value The load disturbance is estimated by the trend of change. equivalent acceleration disturbance value Based on the equivalent acceleration disturbance value Adjust the input control signal This causes the input control signal to be subjected to an equivalent acceleration disturbance value. The adjustment introduces a compensation term for disturbance factors when adjusting the input control signal, which directly offsets the influence of load disturbances at the acceleration level, thereby achieving precise feedforward compensation for disturbances.
[0012] Specifically, the multi-agent consensus control method further includes the step of: based on the error value of the i-th agent... Adjust the disturbance observer state vector of the observer corresponding to the i-th agent. Adjust the disturbance factor of the observer based on the error value to achieve closed-loop feedback control.
[0013] More preferably, the perturbation observer state vector of the i-th agent includes the load perturbation. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value The disturbance observer state vector is used to estimate load disturbances. This scheme enables the observer of the present invention to estimate load disturbances using a third-order fixed-time observer, which can effectively achieve anti-disturbance control and achieve consensus control of multiple agents within a fixed time.
[0014] Preferably, the intelligent agent is a motor, and the input control signal q-axis current The intelligent agent parameters include the motor torque constant and the motor moment of inertia, and the desired mechanical value. From the perspective of expectation Expected speed Desired angular velocity Mechanical value For mechanical angle Mechanical value speed Mechanical angular velocity Mechanical value For mechanical angle Mechanical value speed Mechanical angular velocity Mechanical error value Angular error value Mechanical speed error value Mechanical angular velocity error value ; ; .
[0015] Specifically, the desired parameter x0 also includes the desired acceleration. q-axis current in, Let be the motor torque constant of the i-th motor. Let be the moment of inertia of the i-th motor; compensate for nonlinear terms. The friction term of the i-th motor , It is the viscous friction coefficient of the i-th motor, and the friction term of the j-th motor. , It is the viscous friction coefficient of the j-th motor; the compensation desired acceleration term , Let be the motor torque constant of the j-th motor. Let be the moment of inertia of the j-th motor. The q-axis current of the j-th motor; error control signal. , , , , These are preset control parameters. This scheme, combining the characteristics of the sliding sector (stable control operation characteristics), designs a fixed-time control protocol with nonlinear compensation terms, using the compensation nonlinear terms. To compensate for the leader acceleration term, not only can the motor quickly achieve the desired parameters, but the entire control process is also accurate and stable.
[0016] Specifically, the desired parameter x0 also includes the desired acceleration. q-axis current ,in, Let be the motor torque constant of the i-th motor. Let be the moment of inertia of the i-th motor. The observer corresponding to the i-th motor is based on the angular velocity error. The load disturbance of the i motors is obtained by observing the changing trend. The equivalent acceleration disturbance value; compensation nonlinear term The friction term of the i-th motor , It is the viscous friction coefficient of the i-th motor; the compensation desired acceleration term. The desired parameter x0 also includes the desired acceleration. , The q-axis current of the j-th motor; error control signal. , , , , These are the preset control parameters.
[0017] Specifically, the observer state vector is adjusted according to the following equation: [Equation omitted for brevity]
[0018] Among them, system aggregation items for: ; , , These are the preset observer parameters; , , Error value Including speed error value The disturbance observer state vector includes the equivalent speed disturbance value of the i-th motor. The equivalent acceleration disturbance value of the i-th motor and the equivalent jerk disturbance value of the i-th motor. In this scheme, the disturbance observer state vector of the observer is adjusted by the error value to achieve closed-loop feedback control. Furthermore, the observer uses a third-order fixed-time observer to estimate the load disturbance, which can effectively achieve anti-interference control and realize consistent control of multiple motors within a fixed time.
[0019] The present invention also provides a multi-agent consensus control system, including multiple agent modules and a host computer connected to at least one of the agent modules, wherein the agent modules are connected together in pairs for communication; the host computer includes a processor, a memory, and one or more operating programs, wherein the one or more operating programs are stored in the memory and executed by the processor to implement the multi-agent consensus control method as described above.
[0020] Specifically, each of the intelligent agent modules includes an intelligent agent driver and an intelligent agent, and the intelligent agent drivers in the intelligent agent modules are connected together in pairs for communication.
[0021] Compared with existing technologies, this invention defines a tracking error variable that integrates communication topology information, transforms the consistency problem into an error system stabilization problem, and introduces a variable related to the agent's communication state. and As a coefficient for error calculation, it simplifies the complex problem of stability analysis and enables high-precision synchronous control of multiple agents. Attached Figure Description
[0022] Figure 1 This is a flowchart of the multi-agent consistency control method of the present invention.
[0023] Figure 2 This is a structural diagram of the multi-agent consensus control system of the present invention.
[0024] Figure 3 This is a flowchart of the multi-motor consistency control method of the present invention.
[0025] Figure 4 This is a structural diagram of the multi-motor consistency control system of the present invention.
[0026] Figure 5 This is a state diagram of the desired angle and mechanical angle during the control process of the multi-motor consistency control method of the present invention.
[0027] Figure 6 This is a state diagram of the error value during the control process of the multi-motor consistency control method of the present invention.
[0028] Figure 7 This is a state diagram showing the change of the q-axis current of the control input over time during the control process of the multi-motor consistency control method of the present invention.
[0029] Figure 8 This is a state diagram of the disturbance observer state vector in the control process of the multi-motor consistency control method of the present invention. Detailed Implementation
[0030] To illustrate the technical content, structural features, objectives, and effects of the present invention in detail, the following description is provided in conjunction with the embodiments and accompanying drawings.
[0031] Example 1: refer to Figure 1 This invention discloses a multi-agent consensus control method, including steps S1 to S5. There are n agents, where n is greater than or equal to 2. In this embodiment, there are 3 agents (i.e., n=3), and they communicate with each other in pairs.
[0032] S1, Initialize the agent's agent parameters. These agent parameters include those related to the agent's action parameters.
[0033] S2, Set the leader's expected parameter x0. The expected parameter x0 includes the expected mechanical value. and mechanical value Expected speed .
[0034] S3, Collect the agent's current action parameters x m The action parameters of the i-th agent. Including the mechanical value of the i-th agent and mechanical speed , i=1,2...n, where n is the number of agents.
[0035] S4, based on the action parameters of the i-th agent. Action parameters of the j-th agent Calculate the error value of the i-th agent using the expected parameter x0. Let i = 1, 2, ..., n, j = 1, 2, ..., n and ; .
[0036] Specifically, it determines whether a communication link exists between the i-th agent and the j-th agent; if so... =1, if not =0; Determine whether the i-th agent has obtained the expected parameter x0. If yes... =1, if not =0.
[0037] In this embodiment, agents communicate with each other to form a communication topology (the leader and n agents form a directed spanning tree) to ensure that the leader's information (expected parameter x0) can be transmitted to all agents.
[0038] Specifically, the expected parameters Including expected mechanical value and mechanical value Expected speed Action parameters of the i-th agent Including the mechanical value of the i-th agent and mechanical speed The error value of the i-th agent Including the mechanical error value of the i-th agent and speed error value Action parameters of the j-th agent Including the mechanical value of the j-th agent and mechanical speed The error value of the i-th agent Including the mechanical error value of the i-th agent and mechanical speed error value ; ; .
[0039] Here, mechanical value refers to the numerical value of a parameter that can be collected or calculated and corresponds to the agent's motion parameters, representing the agent's current motion amplitude or motion state. For example, the current position or angle of movement. Desired speed refers to the expected speed of the mechanical value.
[0040] S5, based on the agent parameters and error value of the i-th agent. Adjust the input control signal of the i-th agent . Let i be the action parameters of the i-th agent. Let be the action parameters of the j-th agent.
[0041] Step S5 specifically includes, based on preset control parameters and error values... Obtain the corresponding error control signal Based on the error control signal Adjust the input control signal of the i-th agent .
[0042] Error control signal , , , , These are the preset control parameters.
[0043] Control parameters The following conditions must be met: ,in This scheme uses the boundary of the sliding sector to define the control parameters, ensuring that the system trajectory has no singular points within the sector, that is, the system is running stably while error control is being performed.
[0044] Specifically, control parameters , The following conditions must be met: , , , In this embodiment, . , , , .
[0045] In the above embodiment, a novel sliding sector region is proposed. Avoid singularities in traditional terminal sliding mode.
[0046] Ideally, the desired parameter x0 also includes the desired acceleration. In step S5, the desired acceleration is also determined. Compensation Expected Acceleration Term Based on the input control signal of the i-th agent after compensation of the acceleration term .
[0047] In the initial stage prior to step S1, a multi-agent system and leader dynamics model are also constructed, in which the first... The dynamic model of an agent is as follows: ,in For the nonlinear function of the system, The input control signal is used to control the input.
[0048] The leader dynamics model is as follows: .
[0049] In this embodiment, the input control signal ,in, These are preset values corresponding to the parameters of the i-th agent. These are preset values corresponding to the parameters of the j-th agent. Compensation for nonlinear terms. , Compensation for the leader's acceleration term, For virtual control signals, where and , , , .
[0050] Each agent is equipped with an observer, which is used to observe the perturbation factors of the i-th agent and estimate the load perturbation of the i-th agent. Based on the error value The load disturbance is estimated by the trend of change. equivalent acceleration disturbance value .
[0051] Preferably, in step S5, the equivalent acceleration disturbance value is also considered. Adjust the input control signal This makes the input control signal... Subject to equivalent acceleration disturbance value Adjustment, in the input control signal The adjustment incorporates a compensation term for disturbance factors, which directly counteracts the impact of load disturbances at the acceleration level, achieving precise feedforward compensation for disturbances.
[0052] The perturbation observer state vector of the i-th agent includes the load perturbation. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value This scheme enables the observer of the present invention to estimate load disturbances using a third-order fixed-time observer, which can effectively achieve anti-interference control and realize multi-agent consensus control within a fixed time.
[0053] in, .
[0054] More preferably, the multi-agent consensus control method further includes step S6: based on the error value of the i-th agent... Adjust the disturbance observer state vector of the observer corresponding to the i-th agent. Adjust the disturbance observer state vector of the observer based on the error value to achieve closed-loop feedback control.
[0055] Specifically, the observer state vector is adjusted according to the following equation: [Equation omitted for brevity]
[0056] Among them, system aggregation items Based on the communication topology and known dynamic model of a multi-agent system (e.g., a multi-motor system), this method provides the observer with prior information about the system, enabling the observer to focus on estimating unknown external disturbances. This is done based on the desired parameters, agent parameters, and input control signals. , and Computing system aggregate items , , , These are the preset observer parameters; , , The disturbance observer state vector includes load disturbances. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value .
[0057] The upper bound of the convergence time of this invention is independent of the initial conditions and can be predetermined, thus meeting the time-scheduling requirements of industrial scenarios.
[0058] Among them, by constructing Lyapunov functions Prove that the system state converges to the leader state within a fixed time, with the upper bound of the convergence time being: .
[0059] Fixed-time observers can quickly estimate and accurately compensate for external disturbances, suppressing the impact of disturbances on synchronization performance.
[0060] refer to Figure 2 The present invention also discloses a multi-agent consensus control system, including multiple agent modules 10 and a host computer 20 connected to at least one of the agent modules. The agent modules 10 are connected together in pairs. The host computer 20 includes a processor 21, a memory 22, and one or more operation programs 23. The one or more operation programs 23 are stored in the memory 22 and executed by the processor 21 to realize the multi-agent consensus control method as described above.
[0061] Each of the aforementioned intelligent agent modules 10 includes an intelligent agent driver 11 and an intelligent agent 12, wherein the intelligent agent drivers 11 in the intelligent agent module 10 are connected together in pairs.
[0062] In this embodiment, there are three intelligent agent modules 10. Of course, there can also be two, four, five, or other intelligent agent modules 10.
[0063] Example 2: Using electric motors as intelligent agents, this invention also discloses a multi-motor consistency control method and control system.
[0064] refer to Figure 3 The multi-motor consistency control method includes steps S1a to S5a. The number of motors is n, where n is greater than or equal to 2. In this embodiment, n=3.
[0065] S1a initializes the motor parameters. The motor parameters include the motor torque constant and the motor moment of inertia.
[0066] S2a, Set the leader's expected parameters The desired parameter x0 includes the desired angle. and expected angular velocity .
[0067] S3a, Collect the current operating parameters of the motor. The operating parameters of the i-th motor Including the mechanical angle of the i-th motor and mechanical angular velocity .
[0068] In this embodiment, the mechanical value specifically refers to the mechanical angle, and the mechanical velocity refers to the mechanical angular velocity. The desired mechanical value is the desired angle, and the desired velocity is the desired angular velocity.
[0069] S4a, based on the operating parameters of the i-th motor The operating parameters of the j-th motor Calculate the error value of the i-th motor using the expected parameter x0. Let i = 1, 2, ..., n, j = 1, 2, ..., n and ; .
[0070] Specifically, it determines whether a communication link exists between the i-th motor and the j-th motor; if so... =1, if not =0; Determine if the i-th motor has obtained the desired parameter x0. If yes... =1, if not =0.
[0071] Wherein, the error value of the i-th motor Including the angle error value of the i-th motor and angular velocity error value .
[0072] ; .
[0073] Mechanical error value Angular error value Mechanical speed error value Angular velocity error value .
[0074] S5a, based on the motor parameters and error value of the i-th motor. Adjust the q-axis current of the i-th motor . These are the operating parameters of the i-th motor. These are the operating parameters of the j-th motor.
[0075] Step S5a includes: based on preset control parameters and error values. Obtain the corresponding error control signal Based on the error control signal Adjust the q-axis current of the i-th motor .
[0076] Error control signal , , , , These are the preset control parameters.
[0077] Control parameters The following conditions must be met: ,in This scheme uses the boundary of the sliding sector to define the control parameters, ensuring that the system trajectory has no singular points within the sector, that is, the system operates stably when error control is performed.
[0078] Specifically, control parameters , The following conditions must be met: , , , In this embodiment, . , , , .
[0079] Among them, the q-axis current in, Let be the motor torque constant of the i-th motor. Let be the moment of inertia of the i-th motor; compensate for nonlinear terms. The friction term of the i-th motor , It is the viscous friction coefficient of the i-th motor, and the friction term of the j-th motor. , It is the viscous friction coefficient of the j-th motor; the compensation desired acceleration term The desired parameter x0 also includes the desired acceleration. , Let be the motor torque constant of the j-th motor. Let be the moment of inertia of the j-th motor. The q-axis current of the j-th motor; error control signal. , , , , These are the preset control parameters.
[0080] Ideally, the desired parameter x0 also includes the desired acceleration. In step S5a, the desired acceleration is also considered. Compensation Expected Acceleration Term Based on the compensated acceleration term, the q-axis current of the i-th motor .
[0081] Each motor is equipped with an observer, which is used to observe the disturbance factors of the i-th motor and estimate the load disturbance of the i-th motor. Based on the angular velocity error value The load disturbance is estimated by the trend of change. equivalent acceleration disturbance value .
[0082] Preferably, in step S5a, the equivalent acceleration disturbance value is also considered. Adjust the q-axis current This causes the q-axis current to experience an equivalent acceleration disturbance. The adjustment introduces a compensation term for disturbance factors during q-axis current adjustment, which directly counteracts the influence of load disturbances at the acceleration level, achieving precise feedforward compensation for disturbances.
[0083] Among them, the q-axis current ,in, Let be the motor torque constant of the i-th motor. Let be the moment of inertia of the i-th motor. The observer corresponding to the i-th motor is based on the angular velocity error. The changing trend of load disturbance The equivalent acceleration disturbance value; compensation nonlinear term The friction term of the i-th motor , It is the viscous friction coefficient of the i-th motor; the compensation desired acceleration term. The desired parameter x0 also includes the desired acceleration. , The q-axis current of the j-th motor; error control signal. , , , , These are the preset control parameters.
[0084] The disturbance observer state vector of the i-th motor includes the load disturbance. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value This scheme enables the observer of the present invention to estimate load disturbances using a third-order fixed-time observer, which can effectively achieve anti-interference control and achieve consistent control of multiple motors within a fixed time.
[0085] More preferably, the multi-motor consistency control method further includes step S6a: based on the angular velocity error value of the i-th motor Adjust the disturbance observer state vector of the observer corresponding to the i-th motor. Adjust the disturbance observer state vector of the observer based on the error value to achieve closed-loop feedback control.
[0086] Specifically, the observer state vector is adjusted according to the following equation: [Equation omitted for brevity]
[0087] Among them, system aggregation items for: ; , , These are the preset observer parameters; , , The disturbance observer state vector includes load disturbances. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value .
[0088] Among them, system aggregation items Based on the communication topology of multi-agent systems (such as multi-motor systems) and the construction of known dynamic models, it is used to provide the observer with known prior information about the system, so that the observer can focus on estimating unknown external disturbance factors.
[0089] In this embodiment, a Lyapunov function is constructed. Prove that the system state converges to the leader state within a fixed time, with the upper bound of the convergence time being: .
[0090] refer to Figure 4 The present invention also discloses a multi-motor consistency control system, including multiple motor modules 10a and a host computer 20 connected to at least one of the motor modules. The motor modules 10a are connected together in pairs for communication. The host computer 20 includes a processor 21, a memory 22, and one or more operation programs 23. The one or more operation programs 23 are stored in the memory 22 and executed by the processor 21 to realize the multi-motor consistency control method as described above.
[0091] Each motor module 10a includes a motor driver 11a and a motor 12a, and the motor drivers 11a in each motor module 10a are connected together in pairs for communication. In this embodiment, the motor modules 10a communicate with each other to form a communication topology (the host computer and n motor modules 10a form a directed spanning tree) to ensure that the desired parameter x0 in the host computer can be transmitted to all motor modules 10a.
[0092] In this embodiment, there are three motor modules 10a. Of course, there can also be two, four, five, or other motor modules 10a.
[0093] In this embodiment, each motor is considered as an intelligent agent, and the virtual motor is the leader.
[0094] Leader dynamics model: ;in, From the perspective of leader expectations, For the desired angular velocity, The desired acceleration is set. In this embodiment, it is set... and This refers to the controller's runtime.
[0095] Follower (following motor) dynamics model:
[0096] in, For the first The mechanical angle of TE Connectivity Let be the mechanical angular velocity of the i-th motor, and be the real-time sensor signal. Let be the moment of inertia of the i-th motor. Let be the motor torque constant of the i-th motor. For the i-th motor Shaft current (control input). Let be the friction term for the i-th motor. Let be the load disturbance of the i-th motor.
[0097] In this embodiment, three motors are used as an example. System sampling frequency .
[0098] In step S2, in this embodiment, the following is set and This refers to the controller's runtime.
[0099] Assume that the three motors communicate with each other and each obtains the desired parameter x0. , .
[0100] In this embodiment, the sliding sector parameters are set as follows: .
[0101] In this embodiment, the control protocol parameters are set as follows: , , , .
[0102] Observer parameter settings: , , , , , .
[0103] In this embodiment, a permanent magnet synchronous motor (PMSM) is selected.
[0104] After setting the above parameters, perform multi-motor consistency control as described above. Figure 5 It can converge the motor's mechanical angle to the desired angle in a very short time (e.g., within 0.15 seconds), achieving rapid tracking. (Reference) Figure 6 In this embodiment, the synchronization error can be controlled to be very small, for example ( This meets industrial precision requirements. (Reference) Figure 7 The entire control input q-axis current is continuous and free of jitter, avoiding motor torque fluctuations. (Reference) Figure 8 In this embodiment, the observer can quickly estimate the load disturbance, and the impact of the disturbance on the synchronization performance after compensation is negligible.
[0105] Of course, the intelligent agent in this invention is not limited to motors, but can also be robots, vehicles, missile launch systems, etc.
[0106] The above-disclosed embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, any equivalent variations made in accordance with the scope of the present invention are still within the scope of the present invention.
Claims
1. A multi-agent consensus control method, characterized in that: include: There are n agents, where n is greater than or equal to 2. Initialize the agent parameters of the agents. Set the leader's expected parameters ; Collect the current action parameters of the intelligent agent ; Based on the action parameters of the i-th agent Action parameters of the j-th agent and expected parameters Calculate the error value of the i-th agent. Let i = 1, 2, ..., n, j = 1, 2, ..., n and , Error value ; Specifically, it determines whether a communication link exists between the i-th agent and the j-th agent; if so... =1, if not =0; Determine whether the i-th agent has obtained the expected parameter x0. If yes... =1, if not =0; Based on the error value of the i-th agent Adjust the input control signal of the i-th agent , Let i be the action parameters of the i-th agent. Let be the action parameters of the j-th agent.
2. The multi-agent consensus control method as described in claim 1, characterized in that: The expected parameters Including expected mechanical value and expected speed Action parameters of the i-th agent Including mechanical values and mechanical speed Action parameters of the j-th agent Including mechanical values and mechanical speed The error value of the i-th agent Including the mechanical error value of the i-th agent and mechanical speed error value ; ; 。 3. The multi-agent consensus control method as described in claim 2, characterized in that: Based on the error value of the i-th agent Adjust the input control signal of the i-th agent Including: based on preset control parameters and error values Obtain error control signal Based on the error control signal Adjust the input control signal of the i-th agent ; Error control signal , , , , These are preset control parameters; Control parameters Eligible conditions: ,in .
4. The multi-agent consensus control method as described in claim 3, characterized in that: Control parameters , The following conditions must be met: , , , .
5. The multi-agent consensus control method as described in claim 1, characterized in that: Expected parameters Also includes expected acceleration Based on the error value of the i-th agent Adjust the input control signal of the i-th agent At the same time, it is also based on the expected acceleration. Compensation Expected Acceleration Term The input control signal of the i-th agent is adjusted based on the compensated acceleration term. .
6. The multi-agent consensus control method as described in claim 1, characterized in that: Each agent is equipped with an observer, which is used to observe the perturbation factors of the i-th agent and estimate the load perturbation of the i-th agent. Based on the error value The load disturbance is estimated by the trend of change. equivalent acceleration disturbance value Based on the error value of the i-th agent Adjust the input control signal of the i-th agent At the same time, it is also based on the equivalent acceleration disturbance value. Adjust the input control signal .
7. The multi-agent consensus control method as described in claim 6, characterized in that: It also includes the step of: based on the error value of the i-th agent. Adjust the perturbation observer state vector of the observer corresponding to the i-th agent, wherein the perturbation observer state vector includes the load perturbation. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value .
8. The multi-agent consensus control method as described in claim 2, characterized in that: The intelligent agent is a motor, and the input control signal q-axis current The intelligent agent parameters include the motor torque constant and the motor moment of inertia, and the desired mechanical value. From the perspective of expectation Expected speed Desired angular velocity Mechanical value For mechanical angle Mechanical value speed Mechanical angular velocity Mechanical value For mechanical angle Mechanical value speed Mechanical angular velocity Mechanical error value Angular error value Mechanical speed error value Mechanical angular velocity error value ; ； 。 9. The multi-agent consensus control method as described in claim 8, characterized in that: The desired parameter x0 also includes the desired acceleration. ; q-axis current in, Let be the motor torque constant of the i-th motor. Let be the moment of inertia of the i-th motor; Compensation for nonlinear terms The friction term of the i-th motor , It is the viscous friction coefficient of the i-th motor, and the friction term of the j-th motor. , It is the viscous friction coefficient of the j-th motor; Compensation Expected Acceleration Term The desired parameter x0 also includes the desired acceleration. , Let be the motor torque constant of the j-th motor. Let be the moment of inertia of the j-th motor. Let be the q-axis current of the j-th motor; Error control signal , , , , These are the preset control parameters.
10. The multi-agent consensus control method as described in claim 8, characterized in that: The desired parameter x0 also includes the desired acceleration. ; q-axis current ,in, Let be the motor torque constant of the i-th motor. Let be the moment of inertia of the i-th motor. The observer corresponding to the i-th motor is based on the angular velocity error. The load disturbance of the i motors is obtained by observing the changing trend. The equivalent acceleration disturbance value; Compensation for nonlinear terms The friction term of the i-th motor , It is the viscous friction coefficient of the i-th motor; Compensation Expected Acceleration Term , Let be the q-axis current of the j-th motor; Error control signal , , , , These are the preset control parameters.
11. The multi-agent consensus control method as described in claim 10, characterized in that: Adjust the observer state vector for external disturbances according to the following equation: Among them, system aggregation items for: ； , , These are the preset observer parameters; , , The disturbance observer state vector includes load disturbance. equivalent velocity disturbance value Load disturbance equivalent acceleration disturbance value and load disturbance equivalent jerk disturbance value .
12. A multi-agent consensus control system, characterized in that: It includes multiple intelligent agent modules and a host computer connected to at least one of the intelligent agent modules, with the multiple intelligent agent modules communicating with each other in pairs; The host computer includes a processor, a memory, and one or more operating programs. The one or more operating programs are stored in the memory and executed by the processor to implement the multi-agent consensus control method as described in any one of claims 1-11.