A method for optimizing human-machine collaborative dual control parameters of a cluster system based on fuzzy decision
Patent Information
- Application Number
- CN202611179623.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-05
- Publication Date
- 2026-10-27
AI Technical Summary
[0003]现有的人机协同控制方法多针对单机系统设计，尚未有效推广至集群系统
[0079]现有基于最优隶属度函数的人机协同控制方法仅优化单一控制增益，无法兼顾系统标称跟踪性能与鲁棒抑制性能的协同最优。本发明将最优隶属度函数的设计从单参数情形拓展至双参数情形，同时优化与人类输入相关的两个控制增益——其中主导系统对标称系统的约束跟踪性能，
主导系统对于不确定性的抑制能力，使两个控制通道各自的隶属度函数在同一个情境变量输入下均达到最优；相比单参数优化，双参数优化为操作员提供了独立调节跟踪精度与鲁棒强度的决策自由度，能够适应更复杂的任务需求。此外，本发明所得最优隶属度函数具有闭式解析表达式，由边界条件唯一确定，其解析形式透明度高，可直接用于理论分析与性能证明。鲁棒控制律保证了系统在不确定性下的一致有界性与一致最终有界性，使系统兼具人类决策的灵活性与可证明的鲁棒最优性。
Smart Images
Figure CN122984758A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a human-machine collaborative control method for a cluster system, specifically to a dual control parameter optimization method for human-machine collaborative cluster systems based on fuzzy decision-making, belonging to the field of intelligent control and decision-making for cluster systems. Background Technology
[0002] Cluster systems, through collaborative operations among multiple agents, can compensate for the limitations of individual capabilities, accomplishing complex tasks such as search and rescue, containment, and area monitoring that are difficult for a single agent to handle. However, in actual deployment, cluster systems face highly uncertain environments and complex, ever-changing task requirements, making it difficult to cope with all scenarios using only pre-set automatic control algorithms. In these situations, the real-time perception and decision-making capabilities of human operators become particularly important—operators can make flexible judgments based on the on-site situation, compensating for the shortcomings of pure algorithmic logic. Therefore, integrating human decision-making into the cluster control framework has significant practical implications.
[0003] Existing human-machine collaborative control methods are mostly designed for single-machine systems and have not been effectively extended to cluster systems. Furthermore, the integration of human decision-making relies heavily on fuzzy rule bases built from expert experience, and typically only uses a single control gain to adjust system performance, which lacks flexibility and makes it difficult to achieve coordinated optimization among multiple performance indicators.
[0004] In view of this, it is urgent to extend the human-machine collaborative control method to the cluster system and break through the limitation of the existing single-parameter structure that can only be adjusted in one way. By adjusting the weights of nominal tracking and robust suppression separately through dual-parameter optimization, the tracking accuracy and robustness of the cluster system can be optimized in a coordinated manner, in order to solve at least some of the above problems. Summary of the Invention
[0005] The purpose of this invention is to provide a dual-control parameter optimization method for human-machine collaboration in cluster systems based on fuzzy decision-making. This method introduces contextual variables to represent the human operator's perception and judgment of complex environments, and models human decision-making behavior as two independent fuzzy membership functions. and They are respectively mapped to two-parameter robust control laws. The two control gains and Define the cost functional that characterizes the overall sensitivity of system performance to changes in context. The solution of the two-parameter optimal membership function is transformed into a functional minimization problem. The Euler-Lagrange equations in the variational method are used to solve this functional extremum problem, yielding a closed-form analytical expression for the two-parameter optimal membership function, whose uniqueness is proven by boundary conditions. This two-parameter optimal membership function is then substituted into a robust control law to achieve globally optimal control in a human-machine collaborative manner under two parameters. This invention achieves decoupled adjustment of system tracking accuracy and robustness through a two-parameter adjustment mechanism, effectively improving the system's adaptability to complex operating conditions and overall control quality while ensuring global consistency and boundedness.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is: a method for optimizing dual control parameters of a cluster system based on fuzzy decision-making, which includes the following steps:
[0007] S1. Adjustable parameters related to human input in the controller of the cluster system. Define a functional that comprehensively reflects the size of the system's uniformly bounded region and its sensitivity to changes in context variables, representing the overall performance cost. This transforms the solution to the optimal human decision into a functional optimization problem;
[0008] S2. Solve the functional optimization problem based on the Euler-Lagrange equations in the variational method, and seek a set of optimal fuzzy membership functions that minimize the functional of the overall system performance cost. ;
[0009] S3. Combining the existence condition of functional extrema and boundary constraints, prove the uniqueness of the optimal fuzzy membership function and determine its analytical expression symmetric about the center of the universe of discourse;
[0010] S4. Substitute the optimal fuzzy membership function into the human control term of the robust control law. This leads to the globally optimal human-machine collaborative control design.
[0011] As a further preferred embodiment of this scheme, in step S1, the controller of the cluster system is:
[0012]
[0013] in, and For adjustable controllable gain related to human input, , The control law for the nominal system in the equality constraint control term is expressed as follows:
[0014]
[0015] This is the control law used to handle uncertainties in the cluster system within the equality constraint control terms. Its expression is:
[0016]
[0017] in, For generalized position vectors, These are generalized velocity and acceleration, respectively. This is the nominal part of the system's inertia matrix. For the constraint matrix, , To constrain the following error, Let be an upper bound function representing the uncertainty of the system.
[0018] As a further preferred embodiment of this scheme, the design cost process for functional optimization in step S1 is as follows:
[0019] Choose the Lyapunov function:
[0020]
[0021] in, ;
[0022] According to the Lyapunov stability proof:
[0023]
[0024] in, This is a lower bound for the system's inertial uncertainty. The upper bound vector of the magnitude of the system uncertainty is its relation to... Together they describe the uncertainty of the system;
[0025] Furthermore, we deduce that the system is uniformly bounded:
[0026]
[0027] in, and Represent matrices respectively The minimum and maximum eigenvalues;
[0028] Further, we deduce that the system is uniformly eventually bounded: for ,in ,satisfy:
[0029]
[0030] Construct the following optimal fuzzy membership function: Let and The upper and lower bounds are respectively, satisfying Consider continuous membership functions ,in ;
[0031] propose The selection method is as follows:
[0032]
[0033] in Representing biological signals or human perception, representing the context of decision-making, and mapping values. This corresponds to the outcome of human decision-making; this mapping reflects human perception of the situation. The way they react, and thus determine The value of ;
[0034] therefore, Represented as:
[0035]
[0036] in, ;
[0037] make:
[0038]
[0039] get:
[0040]
[0041] Consider the following cost functional:
[0042]
[0043] The cost functional The size of the region representing consistent boundedness and consistent ultimate boundedness varies with the context variable. The overall sensitivity to variation across the entire domain. The smaller the value, the less sensitive the system performance is to situational fluctuations, and the stronger the system robustness; accordingly, the solution for optimal human decision-making is transformed into minimizing the defined cost functional.
[0044] As a further preferred embodiment of this scheme, in step S2, the optimal design formulation for human decision-making is based on solving the extremum problem of the functional using the variational method:
[0045] Consider the following Lagrange function:
[0046]
[0047] To derive the multidimensional Euler-Lagrange equations, we first calculate their approximation with respect to... Partial derivatives:
[0048]
[0049] Next, calculate its relation to Partial derivatives:
[0050]
[0051] Regarding the above two equations Taking the total differential and simplifying, we obtain the following Euler-Lagrange equations:
[0052]
[0053] in,
[0054]
[0055] The above system of equations can be expressed in matrix form:
[0056]
[0057] Calculate the determinant of the coefficient matrix:
[0058]
[0059] because Therefore, the determinant is not zero, that is:
[0060]
[0061] Therefore, the homogeneous linear system of equations has only the zero solution:
[0062]
[0063] This proves and All are about A linear function, that is:
[0064]
[0065] in, These are undetermined constants;
[0066] Using the boundary conditions, the solution is obtained as follows:
[0067]
[0068] The optimal membership function Input cost functional The optimal cost functional is obtained as follows: .
[0069] As a further preferred embodiment of this scheme, the uniqueness and analytical expression of the optimal fuzzy membership function in step S3 are proven as follows:
[0070] Combined with boundary conditions and The slope and intercept of a linear function are uniquely determined by the boundary conditions; therefore, the optimal membership function... It exists uniquely under the stated boundary conditions and smoothness conditions;
[0071] because Satisfying the centrality of discourse Symmetry, that is ,in Then the symmetric optimal membership function In the complete domain The closed analytic expression for is:
[0072]
[0073] As a further preferred embodiment of this scheme, in step S4, the globally optimal human-machine collaborative control design is as follows:
[0074]
[0075] To quantitatively evaluate system performance, the following comprehensive performance index is defined:
[0076]
[0077] in, This represents the total simulation time. To constrain the following error in the horizontal formation of the cluster system; and These represent the cumulative error and peak error of the cluster system during the simulation period, respectively. They characterize the overall tracking accuracy of the system throughout the entire task cycle and the maximum instantaneous deviation that the system may exhibit during dynamic response, in order to evaluate the safety and robustness of the system under sudden disturbances or adverse initial conditions.
[0078] Compared with the prior art, the present invention has the following technical effects:
[0079] Existing human-machine cooperative control methods based on optimal membership functions only optimize a single control gain, failing to achieve a synergistic optimization of nominal tracking performance and robust suppression performance. This invention extends the design of the optimal membership function from the single-parameter case to the two-parameter case, simultaneously optimizing two control gains related to human input—wherein… The constraint tracking performance of the dominant system over the nominal system The dominant system's ability to suppress uncertainty ensures that the membership functions of both control channels reach their optimal values under the same context variable input. Compared to single-parameter optimization, two-parameter optimization provides operators with the decision-making freedom to independently adjust tracking accuracy and robustness, adapting to more complex task requirements. Furthermore, the optimal membership function obtained in this invention has a closed-form analytical expression, uniquely determined by boundary conditions. Its analytical form is highly transparent and can be directly used for theoretical analysis and performance verification. The robust control law guarantees the system's uniform boundedness and uniform eventual boundedness under uncertainty, giving the system both the flexibility of human decision-making and provable robust optimality. Attached Figure Description
[0080] Figure 1 This is a flowchart of the human-machine collaborative control method for the cluster system of the present invention.
[0081] Figure 2 This is a schematic diagram of the optimal membership function.
[0082] Figure 3 It is a graph showing the control gain as a function of situational variables.
[0083] Figure 4 This is a comparison chart of the cluster system trajectory under different scenario gains.
[0084] Figure 5 This is a bar chart comparing the cumulative constraint following error of the cluster system under different scenario gains.
[0085] Figure 6 This is a bar chart comparing the maximum constraint following error of the cluster system under different scenario gains. Detailed Implementation
[0086] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.
[0087] like Figure 1 The present invention provides a method for optimizing dual control parameters of a cluster system based on fuzzy decision-making in a human-machine collaborative manner, comprising the following steps:
[0088] S1. Adjustable parameters related to human input in the controller of the cluster system. Define a functional that comprehensively reflects the size of the system's uniformly bounded region and its sensitivity to changes in context variables, representing the overall performance cost. This transforms the solution to the optimal human decision into a functional optimization problem.
[0089] S2. Solve the functional optimization problem based on the Euler-Lagrange equations in the variational method, and seek a set of optimal fuzzy membership functions that minimize the functional of the overall system performance cost. .
[0090] S3. Combining the existence condition of functional extrema and boundary constraints, prove the uniqueness of the optimal fuzzy membership function and determine its analytical expression symmetric about the center of the universe of discourse.
[0091] S4. Substitute the optimal fuzzy membership function into the human control term of the robust control law. This leads to the globally optimal human-machine collaborative control design.
[0092] In step S1, the controller of the proposed cluster system is:
[0093]
[0094] in, and An adjustable controllable gain related to human input. . The control law for the nominal system in the equality constraint control term is expressed as follows:
[0095]
[0096] This is the control law used to handle uncertainties in the cluster system within the equality constraint control terms. Its expression is:
[0097]
[0098] in, For generalized position vectors, These are generalized velocity and acceleration, respectively. This is the nominal part of the system's inertia matrix. For the constraint matrix, , To constrain the following error, Let be an upper bound function representing the uncertainty of the system.
[0099] Choose the Lyapunov function:
[0100]
[0101] in, .
[0102] According to the Lyapunov stability proof, we can conclude that:
[0103]
[0104] in, This is a lower bound for the system's inertial uncertainty. The upper bound vector of the magnitude of the system uncertainty is its relation to... Together they describe the uncertainty of the system.
[0105] It can be deduced that the system is uniformly bounded.
[0106]
[0107] in, and Represent matrices respectively The minimum and maximum eigenvalues.
[0108] Furthermore, it can be deduced that the system is uniformly eventually bounded: for ,in ,satisfy
[0109]
[0110] The following optimal fuzzy membership function is constructed: Let and The upper and lower bounds are respectively, satisfying Consider continuous membership functions. ,in .
[0111] propose The selection method is as follows:
[0112]
[0113] in It can represent biological signals or human perception, represent the context in which a decision is made, and the mapping value. This corresponds to the outcome of human decision-making. This mapping reflects human perception of the situation. The way it reacts, and thus determines The value of .
[0114] therefore, It can be represented as
[0115]
[0116] in, .
[0117] make
[0118]
[0119] It can be obtained
[0120]
[0121] Consider the following cost functional:
[0122]
[0123] The cost functional The size of the region characterizing uniform boundedness and uniform eventual boundedness varies with the context variable. Total sensitivity to changes across the entire domain. The smaller the value, the less sensitive the system performance is to situational fluctuations, and the stronger the system robustness.
[0124] In step S2, consider the following Lagrange function.
[0125]
[0126] To derive the multidimensional Euler-Lagrange equations, we first calculate their approximation on... partial derivatives
[0127]
[0128] Next, calculate its relation to partial derivatives
[0129]
[0130] Regarding the above two equations Taking the total differential and simplifying, we obtain the following Euler-Lagrange equations:
[0131]
[0132] in,
[0133]
[0134] The above system of equations can be expressed in matrix form:
[0135]
[0136] Calculate the determinant of the coefficient matrix:
[0137]
[0138] because Therefore, the determinant is not zero, that is
[0139]
[0140] Therefore, the homogeneous linear equation system has only the zero solution:
[0141]
[0142] This proves and All are about A linear function, that is:
[0143]
[0144] in, is an undetermined constant.
[0145] Using the boundary conditions, we obtain the following solution:
[0146]
[0147] The optimal membership function In the cost functional The optimal cost functional is obtained as follows: .
[0148] In step S3, the boundary conditions are considered. and The slope and intercept of the linear function are uniquely determined by the boundary conditions; therefore, the optimal membership function... It exists only under the stated boundary conditions and smoothness conditions.
[0149] because Satisfying the centrality of discourse Symmetry, that is ,in Then the symmetric optimal membership function In the complete domain The closed analytic expression for is:
[0150]
[0151] In step S4, the optimal human-machine collaborative control design is as follows:
[0152]
[0153] To quantitatively evaluate system performance, the following comprehensive performance index is defined:
[0154]
[0155] in, This represents the total simulation time. To constrain the following error in the horizontal formation of the cluster system. and These represent the cumulative error and peak error of the cluster system during the simulation period, respectively. They characterize the overall tracking accuracy of the system throughout the entire task cycle and the maximum instantaneous deviation that may occur during the dynamic response process, in order to evaluate the system's safety and robustness under sudden disturbances or adverse initial conditions.
[0156] The above methods will be further applied in an example:
[0157] This example uses a swarm system of five unmanned surface vehicles (USVs) as the application object, including one leader and four followers, to perform formation tasks in a two-dimensional water surface environment. In the simulation, the nominal inertial mass matrix of the USV swarm system is taken as... The positive definite matrix of the Lyapunov function is taken The upper and lower bounds of the control gain are respectively set as The domain of situational variables is taken The feasibility of the proposed control method was verified through MATLAB simulation.
[0158] The overall process of this method is as follows: Figure 1 As shown, the specific implementation steps are as follows:
[0159] Step 1: Apply the above human-machine collaborative control method to the USV cluster system, and define the total system cost functional as:
[0160]
[0161] Minimize the functional to obtain the optimal membership function, and select several representative sets of human decision terms under different situational variables under the above simulation model parameters: when the situational variables When, corresponding ; When, corresponding ; When, corresponding .
[0162] Simulation verification Figures 2 to 6 The simulation results of the human-machine collaborative control method proposed in this invention are presented. Figure 2 The optimal membership function is shown. A curve on the complete universe of discourse; the function lies at the center of the universe of discourse. The maximum value is 1 at the boundary and the minimum value is 1 at the edge. The overall structure exhibits a symmetrical linear piecewise form, verifying the correctness and uniqueness of the analytical expression for the optimal membership function. Figure 3 Demonstrates control gain and Context variables The curve changes, and the gain varies with The linear increase verifies the mapping relationship. The effectiveness. Figure 4 The study demonstrates a comparison of the motion trajectories of the cluster system under different scenario gains. Figure 5 and Figure 6Bar charts comparing the cumulative constraint following error and the peak value of the constraint following error of the cluster system under different scenario gains are presented. The results show that there are significant differences in system performance corresponding to different gains—operators can obtain the corresponding control gain through the designed optimal membership function mapping by inputting appropriate scenario variables based on actual task requirements and real-time environmental perception, thereby adapting the system performance to the requirements of different scenarios.
[0163] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the scope of protection of the present invention in any way, and all technical solutions obtained by equivalent substitution or other means fall within the scope of protection of the present invention. Parts not covered in this invention are the same as or can be implemented using existing technology.
Claims
1. A method for optimizing dual control parameters of a cluster system based on fuzzy decision-making in a human-machine collaborative manner, characterized in that, Includes the following steps: S1. Adjustable parameters related to human input in the controller of the cluster system. Define a functional that comprehensively reflects the size of the system's uniformly bounded region and its sensitivity to changes in context variables, representing the overall performance cost. This transforms the solution to the optimal human decision into a functional optimization problem; S2. Solve the functional optimization problem based on the Euler-Lagrange equations in the variational method, and seek a set of optimal fuzzy membership functions that minimize the functional of the overall system performance cost. ; S3. Combining the existence condition of functional extrema and boundary constraints, prove the uniqueness of the optimal fuzzy membership function and determine its analytical expression symmetric about the center of the universe of discourse; S4. Substitute the optimal fuzzy membership function into the human control term of the robust control law. This leads to the globally optimal human-machine collaborative control design.
2. The method for optimizing dual control parameters of a cluster system based on fuzzy decision-making for human-machine collaboration, as described in claim 1, is characterized in that... In step S1, the controller of the cluster system is: in, and An adjustable controllable gain related to human input. , The control law for the nominal system in the equality constraint control term is expressed as follows: This is the control law used to handle uncertainties in the cluster system within the equality constraint control terms. Its expression is: in, For generalized position vectors, These are generalized velocity and acceleration, respectively. This is the nominal part of the system's inertia matrix. For the constraint matrix, , To constrain the following error, Let be an upper bound function representing the uncertainty of the system.
3. The method for optimizing dual control parameters of a cluster system based on fuzzy decision-making, as described in claim 1 or 2, is characterized in that... In step S1, the design cost process of functional optimization is as follows: Choose the Lyapunov function: in, ; According to the Lyapunov stability proof: in, This is a lower bound for the system's inertial uncertainty. The upper bound vector of the magnitude of the system uncertainty is its relation to... Together they describe the uncertainty of the system; Furthermore, we deduce that the system is uniformly bounded: in, and Represent matrices respectively The minimum and maximum eigenvalues; Further, we deduce that the system is uniformly eventually bounded: for ,in ,satisfy: Construct the following optimal fuzzy membership function: Let and The upper and lower bounds are respectively, satisfying Consider continuous membership functions ,in ; propose The selection method is as follows: in Representing biological signals or human perception, representing the context of decision-making, and mapping values. This corresponds to the outcome of human decision-making; this mapping reflects human perception of the situation. The way they react, and thus determine The value of ; therefore, Represented as: in, ; make: get: Consider the following cost functional: The cost functional The size of the region representing consistent boundedness and consistent ultimate boundedness varies with the context variable. The overall sensitivity to variation across the entire domain. The smaller the value, the less sensitive the system performance is to situational fluctuations, and the stronger the system robustness; accordingly, the solution for optimal human decision-making is transformed into minimizing the defined cost functional.
4. The method for optimizing dual control parameters of a cluster system based on fuzzy decision-making according to claim 3, characterized in that, In step S2, the optimal design formulation for human decision-making is based on solving the extremum problem of the functional using the variational method: Consider the following Lagrange function: To derive the multidimensional Euler-Lagrange equations, we first calculate their approximation with respect to... Partial derivatives: Next, calculate its relation to Partial derivatives: Regarding the above two equations Taking the total differential and simplifying, we obtain the following Euler-Lagrange equations: in, The above system of equations can be expressed in matrix form: Calculate the determinant of the coefficient matrix: because Therefore, the determinant is not zero, that is: Therefore, the homogeneous linear system of equations has only the zero solution: This proves and All are about A linear function, that is: in, These are undetermined constants; Using the boundary conditions, the solution is obtained as follows: The optimal membership function Input cost functional The optimal cost functional is obtained as follows: .
5. The method for optimizing dual control parameters of a cluster system based on fuzzy decision-making according to claim 4, characterized in that, In step S3, the uniqueness and analytical expression of the optimal fuzzy membership function are proven as follows: Combined with boundary conditions and The slope and intercept of a linear function are uniquely determined by the boundary conditions; therefore, the optimal membership function... It exists uniquely under the stated boundary conditions and smoothness conditions; because Satisfying the centrality of discourse Symmetry, that is ,in Then the symmetric optimal membership function In the complete domain The closed analytic expression for is:
6. The method for optimizing dual control parameters of a cluster system based on fuzzy decision-making for human-machine collaboration, as described in claim 1, is characterized in that... In step S4, the globally optimal human-machine collaborative control design is as follows: To quantitatively evaluate system performance, the following comprehensive performance index is defined: in, This represents the total simulation time. To constrain the following error in the horizontal formation of the cluster system; and These represent the cumulative error and peak error of the cluster system during the simulation period, respectively. They characterize the overall tracking accuracy of the system throughout the entire task cycle and the maximum instantaneous deviation that the system may exhibit during dynamic response, in order to evaluate the safety and robustness of the system under sudden disturbances or adverse initial conditions.