A heavy-load robot human-machine collaborative control method based on double-layer game

Abstract

The present application belongs to the technical field of heavy-load manipulator human-machine collaborative control system, and in particular to a heavy-load manipulator human-machine collaborative control method based on double-layer game, comprising the following steps: first, the operator and the automatic control system respectively send control signals according to the current environment; then, the upper-layer game permission allocation system acquires optimal control permission according to the control signals of the two parties of collaborative control and the current environment state, and sends the optimal control permission to the lower-layer game human-machine interaction model; then, the human-machine interaction model optimizes the control signals of the operator and the automatic control system in real time by solving Nash equilibrium; finally, the optimal control permission is used to fuse the optimal control signals, and the fused control signals are output to the heavy-load manipulator, so as to realize the heavy-load manipulator human-machine collaborative control based on double-layer game. The present application can effectively reduce the influence of human-machine conflict on the performance of heavy-load manipulator human-machine collaborative control, and maximally improve the operation efficiency and safety of the heavy-load manipulator.

Description

Technical Field

[0001] This invention relates to the field of heavy-duty robotic arm human-machine collaborative control system technology, specifically a heavy-duty robotic arm human-machine collaborative control method based on two-level game theory. Background Technology

[0002] Heavy-duty robotic arms, due to their superior efficiency in high-load tasks, have been widely used in unstructured environments such as mining, rescue operations, and construction. However, compared to conventional robotic arms in industrial settings, heavy-duty robotic arms have larger moments of inertia, more degrees of freedom, and are subject to frequent heavy load impacts. This not only increases the workload of operators but also greatly increases the risk of operational errors, posing significant challenges to the operational efficiency and safety of heavy-duty robotic arms. To address this, researchers have attempted to apply automation technology to heavy-duty robotic arms to reduce operator workload and improve operational efficiency and safety. However, in the face of complex and ever-changing unstructured environments, the dynamic adaptability and rapid decision-making capabilities of existing heavy-duty robotic arm automatic control systems are insufficient, making it difficult to effectively cope with highly dynamic environmental changes.

[0003] Human-machine collaborative control technology integrates the operator's adaptability to complex environments with the high-precision control capabilities of automatic control systems. By leveraging the complementary strengths of both, it is expected to significantly improve the operational efficiency and safety of heavy-duty robotic arms. Existing research on heavy-duty robotic arms primarily focuses on achieving human-machine collaborative control by considering environmental risks and operator states, but rarely addresses the conflict between the operator and the automatic control system. In unstructured work environments, operators typically face high workloads, are prone to distraction and fatigue, leading to discrepancies between their understanding of the task environment and the automatic control system, thus triggering human-machine conflict. This not only affects the human-machine collaborative control performance of heavy-duty robotic arms but may even pose potential safety hazards. Therefore, there is an urgent need for an effective human-machine collaborative control method for heavy-duty robotic arms that resolves human-machine conflict, fully leveraging the advantages of both to improve the operational efficiency and safety of heavy-duty robotic arms.

[0004] Game theory has significant advantages in dealing with problems involving multiple participants and their interactions or conflicts of interest. In human-machine conflict scenarios, the interaction between the operator and the automatic control system exhibits characteristics consistent with differential games. Therefore, game theory is widely used in modeling interactive behaviors under conflict conditions in multi-agent systems, providing a theoretical foundation for designing safer and more efficient human-machine collaborative control systems. Currently, game-based human-machine collaborative control methods mainly construct the human-machine interaction relationship between the operator and the automatic control system through game models, optimizing the control signals of both systems to reduce human-machine conflict. However, these studies typically employ fixed allocation strategies for human-machine collaborative control permissions, which cannot adaptively adjust in complex and variable unstructured environments, thus failing to maximize the performance of heavy-duty robotic arm human-machine collaborative control. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a heavy-duty robotic arm human-machine collaborative control method based on a two-layer game theory approach. It comprehensively considers human-machine conflict, the degree of participation of the operator and the automatic control system, and collision avoidance risks. An upper-layer game theory model for heavy-duty robotic arm human-machine collaboration is constructed to adaptively acquire human-machine collaborative control permissions. A lower-layer human-machine interaction model based on game theory is constructed. By understanding the interaction behavior between the operator and the automatic control system, it provides assistance more suited to the operator's characteristics, thereby effectively resolving human-machine conflict issues while achieving collaborative control. Through upper-layer and lower-layer game theory, the impact of human-machine conflict on the performance of collaborative control is minimized, achieving the optimal effect of heavy-duty robotic arm human-machine collaborative control. This maximizes the efficiency and safety of the heavy-duty robotic arm while reducing the operator's workload, solving the problems of low efficiency and safety hazards caused by human-machine conflict in heavy-duty robotic arm human-machine collaborative control systems.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention specifically adopts the following technical solution:

[0009] A human-machine collaborative control method for a heavy-duty robotic arm based on two-level game theory includes the following steps:

[0010] S1, the operator and the automatic control system send control signals to the heavy-duty robotic arm according to the current environment. , ;

[0011] S2, the upper-level game-theoretic permission allocation system constructs utility functions for the operator and the automatic control system based on the control signals from the operator and the automatic control system, as well as the current environmental state. , The optimal control authority for the current operator and the automatic control system is determined using game theory. , ), and send it to the lower-level game-playing human-computer interaction model;

[0012] S3, the lower-level game-theoretic human-computer interaction model combines the current control authority of the operator and the automatic control system, and constructs a cost function that conforms to the control characteristics of the operator and the automatic control system based on the distributed model predictive control algorithm. It also constructs the human-computer interaction behavior between the operator and the automatic control system based on Nash game theory, and solves for the Nash optimal control signal of the operator and the automatic control system through Nash equilibrium. , ;

[0013] S4. The system integrates the human-machine collaborative control authority of the heavy-duty robotic arm with the Nash optimal control signals of the operator and the automatic control system, and outputs the integrated control signals to the heavy-duty robotic arm to achieve human-machine collaborative control of the heavy-duty robotic arm based on two-layer game theory.

[0014] Furthermore, by combining the control signals from the operator and the automatic control system with the current environmental state, we construct sub-utilities related to collision avoidance risk, human-machine conflict, and participation regarding control authority. The function;

[0015] Before constructing the utility function, we first construct the discrete-space state equations for the heavy-duty robotic arm's human-machine collaborative control, as shown below:

[0016] (1)

[0017] In the formula: Let be the system state vector. The system output vector includes the position information of the robotic arm's end effector. The system state matrix, Input matrix to the system, For the system output matrix, The control input of an automatic control system For the operator's control input.

[0018] The collision avoidance risk sub-utility function is shown below:

[0019] Keeping the current control signals and control permissions unchanged, iteratively (1) construct the future first... The position status of the end effector of the heavy-duty robotic arm in the step, and its control permissions. Functional relationship ;

[0020] (2);

[0021] Constructing the collision avoidance risk sub-utility function:

[0022] (3)

[0023] In the formula: , These are the collision avoidance risk sub-utility functions for the operator and the automatic control system, respectively. This is the location information of the obstacle.

[0024] The human-machine conflict sub-utility function is shown below:

[0025] Considering the deviation between the operator's and the automatic control system's control signals, a corresponding human-machine conflict sub-utility function is constructed as follows:

[0026] (4)

[0027] In the formula: , These represent the human-machine conflict sub-utility functions of the operator and the automatic control system, respectively.

[0028] The participation sub-utility function is shown below:

[0029] To optimize the control process and improve safety, it is desirable for both the operator and the automatic control system to participate in collaborative control, without granting either party complete control authority. Therefore, the participation sub-utility functions of the operator and the automatic control system are constructed as follows:

[0030] (5)

[0031] In the formula: , These are the participation sub-utility functions for the operator and the automatic control system, respectively. For coefficients;

[0032] Furthermore, by merging the aforementioned sub-utility functions, a utility function for the operator and the automatic control system is constructed:

[0033] (6)

[0034] In the formula: , These are the utility functions for the operator and the automatic control system, respectively. These are the normalization coefficients;

[0035] Construct the Nash product according to equation (6) And solve for the optimal control authority for human-machine collaboration in heavy-duty robotic arms. :

[0036] (7).

[0037] Furthermore, in the process of human-machine collaborative control, both the operator and the automatic control system will do their best to make the heavy-duty robotic arm track their respective target trajectories, and will consider the intervention behavior of the other party in collaborative control before executing control actions.

[0038] The future can be obtained through iterative formula (1). The predicted output expression for the step:

[0039] (8)

[0040] in, Depend on The values ​​at each prediction step are composed of... , , , and These are vectors and matrices obtained through iteration;

[0041] With the objective of minimizing the deviation between the control signal and the actual position from the target trajectory, cost functions for both the operator and the automatic control system can be constructed:

[0042] (9)

[0043] In the formula: and These are the penalty coefficient matrices for the target trajectory deviation and the control quantity, respectively. and These represent the cost functions of the automatic control system and the operator, respectively.

[0044] In non-cooperative games, there exists a set of control combinations such that unilaterally changing the control inputs does not allow any participant to gain more benefits. Similarly, in this study, the operator and the automatic control system also have a Nash equilibrium solution. By using the distributed model predictive control method, the final Nash equilibrium strategy can be solved.

[0045] (10)

[0046] in, This represents the Nash equilibrium strategy of the human-machine collaborative control system for heavy-duty robotic arms;

[0047] By integrating the optimal control permissions allocated at the current moment. and the Nash equilibrium solution for both human and machine Obtain the optimal collaborative control signal for the heavy-duty robotic arm. To achieve human-machine collaborative control of heavy-duty robotic arms based on two-level game theory;

[0048] (11)

[0049] in, This is the optimal collaborative control signal for the heavy-duty robotic arm at the current moment.

[0050] (III) Beneficial Effects

[0051] Compared with existing technologies, this invention provides a human-machine collaborative control method for heavy-duty robotic arms based on two-layer game theory, which has the following beneficial effects:

[0052] This invention comprehensively considers human-machine conflict, participation, and collision risk between the operator and the automatic control system. It obtains the optimal control authority through upper-level game theory and optimizes the control signals of the operator and the automatic control system based on the lower-level Nash game human-machine interaction model. It integrates the optimal control signals of the lower-level operator and the automatic control system with the human-machine collaborative control authority, thereby realizing human-machine collaborative control of heavy-duty robotic arms based on two-level game theory. This not only effectively solves the human-machine conflict problem in the human-machine collaborative control of heavy-duty robotic arms, but also improves the operating efficiency and safety of heavy-duty robotic arms. Attached Figure Description

[0053] Figure 1 This is a research roadmap for the human-machine collaborative control method for heavy-duty robotic arms based on two-layer game theory, as described in this invention.

[0054] Figure 2 This invention presents a research roadmap for upper-level human-machine collaborative control authority allocation based on game theory.

[0055] Figure 3 This is a schematic diagram of the human-machine interaction model of the lower-level heavy-duty robotic arm based on game theory in this invention. Detailed Implementation

[0056] 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.

[0057] Example

[0058] like Figure 1 , Figure 2 and Figure 3 As shown in the figure, an embodiment of the present invention proposes a human-machine collaborative control method for a heavy-duty robotic arm based on two-layer game theory, comprising the following steps:

[0059] S1, the operator and the automatic control system send control signals to the heavy-duty robotic arm according to the current environment. , ;

[0060] S2, the upper-level game-theoretic permission allocation system constructs utility functions for the operator and the automatic control system based on the control signals from the operator and the automatic control system, as well as the current environmental state. , The optimal control authority for the current operator and the automatic control system is determined using game theory. , ), and send it to the lower-level game-playing human-computer interaction model;

[0061] S3, the lower-level game-theoretic human-computer interaction model combines the current control authority of the operator and the automatic control system, and constructs a cost function that conforms to the control characteristics of the operator and the automatic control system based on the distributed model predictive control algorithm. It also constructs the human-computer interaction behavior between the operator and the automatic control system based on Nash game theory, and solves for the Nash optimal control signal of the operator and the automatic control system through Nash equilibrium. , ;

[0062] S4. The system integrates the human-machine collaborative control authority of the heavy-duty robotic arm with the Nash optimal control signals of the operator and the automatic control system, and outputs the integrated control signals to the heavy-duty robotic arm to achieve human-machine collaborative control of the heavy-duty robotic arm based on two-layer game theory.

[0063] First, by combining the control signals from the operator and the automatic control system with the current environmental state, we construct sub-utilities related to collision avoidance risk, human-machine conflict, and participation regarding control authority. The function;

[0064] Before constructing the utility function, we first construct the discrete-space state equations for the heavy-duty robotic arm's human-machine collaborative control, as shown below:

[0065] (1)

[0066] In the formula: Let be the system state vector. The system output vector includes the position information of the robotic arm's end effector. The system state matrix, Input matrix to the system, For the system output matrix, The control input of an automatic control system For the operator's control input.

[0067] The collision avoidance risk sub-utility function, human-machine conflict sub-utility function, and participation sub-utility function are shown below:

[0068] (1) Collision avoidance risk sub-utility function:

[0069] Keeping the current control signals and control permissions unchanged, iteratively (1) construct the future first... The position status of the end effector of the heavy-duty robotic arm in the step, and its control permissions. Functional relationship ;

[0070] (2);

[0071] Constructing the collision avoidance risk sub-utility function:

[0072] (3)

[0073] In the formula: , These are the collision avoidance risk sub-utility functions for the operator and the automatic control system, respectively. This is the location information of the obstacle.

[0074] (2) Human-machine conflict sub-utility function:

[0075] Considering the deviation between the operator's and the automatic control system's control signals, a corresponding human-machine conflict sub-utility function is constructed as follows:

[0076] (4)

[0077] In the formula: , These represent the human-machine conflict sub-utility functions of the operator and the automatic control system, respectively.

[0078] (3) Participation degree sub-utility function:

[0079] To optimize the control process and improve safety, it is desirable for both the operator and the automatic control system to participate in collaborative control, without granting either party complete control authority. Therefore, the participation sub-utility functions of the operator and the automatic control system are constructed as follows:

[0080] (5)

[0081] In the formula: , These are the participation sub-utility functions for the operator and the automatic control system, respectively. For coefficients;

[0082] By combining the aforementioned sub-utility functions, a utility function for the operator and the automatic control system is constructed:

[0083] (6)

[0084] In the formula: , These are the utility functions for the operator and the automatic control system, respectively. These are the normalization coefficients;

[0085] Construct the Nash product according to equation (6) And solve for the optimal control authority for human-machine collaboration in heavy-duty robotic arms. :

[0086] (7).

[0087] During human-machine collaborative control, both the operator and the automatic control system will do their best to make the heavy-duty robotic arm track their respective target trajectories, and will consider the intervention behavior of the other party in collaborative control before executing control actions.

[0088] The future can be obtained through iterative formula (1). The predicted output expression for the step:

[0089] (8)

[0090] in, Depend on The values ​​at each prediction step are composed of... , , , and These are vectors and matrices obtained through iteration;

[0091] With the objective of minimizing the deviation between the control signal and the actual position from the target trajectory, cost functions for both the operator and the automatic control system can be constructed:

[0092] (9)

[0093] In the formula: and These are the penalty coefficient matrices for the target trajectory deviation and the control quantity, respectively. and These represent the cost functions of the automatic control system and the operator, respectively.

[0094] In non-cooperative games, there exists a set of control combinations such that unilaterally changing the control inputs does not allow any participant to gain more benefits. Similarly, in this study, the operator and the automatic control system also have a Nash equilibrium solution. By using the distributed model predictive control method, the final Nash equilibrium strategy can be solved.

[0095] (10)

[0096] in, This represents the Nash equilibrium strategy of the human-machine collaborative control system for heavy-duty robotic arms;

[0097] By integrating the optimal control permissions allocated at the current moment. and the Nash equilibrium solution for both human and machine Obtain the optimal collaborative control signal for the heavy-duty robotic arm. To achieve human-machine collaborative control of heavy-duty robotic arms based on two-level game theory;

[0098] (11)

[0099] in, This is the optimal collaborative control signal for the heavy-duty robotic arm at the current moment.

[0100] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1.A method for human-robot collaborative control of heavy-duty manipulator based on double-layer game, characterized in that, Includes the following steps: S1, the operator and the automatic control system respectively send control signals to the heavy-duty robot arm according to the current environment , ; S2, the upper-level game-theoretic permission allocation system constructs utility functions for the operator and the automatic control system based on the control signals from the operator and the automatic control system, as well as the current environmental state. , The optimal control authority for the current operator and the automatic control system is determined using game theory. , ), and send it to the lower-level game-playing human-computer interaction model; S3, the lower-level game-theoretic human-computer interaction model combines the current control authority of the operator and the automatic control system, and constructs a cost function that conforms to the control characteristics of the operator and the automatic control system based on the distributed model predictive control algorithm. It also constructs the human-computer interaction behavior between the operator and the automatic control system based on Nash game theory, and solves for the Nash optimal control signal of the operator and the automatic control system through Nash equilibrium. , ; S4. Integrate the human-machine collaborative control authority of the heavy-duty robotic arm with the Nash optimal control signals of the operator and the automatic control system, and output the integrated control signals to the heavy-duty robotic arm to realize human-machine collaborative control of the heavy-duty robotic arm based on two-layer game theory. By combining the control signals from the operator and the automatic control system with the current environmental state, we can construct sub-utilities related to collision avoidance risk, human-machine conflict, and participation regarding control authority. The function; Before constructing the utility function, we first construct the discrete-space state equations for the heavy-duty robotic arm's human-machine collaborative control, as shown below: (1)； In the formula: Let be the system state vector. The system output vector includes the position information of the robotic arm's end effector. The system state matrix, Input matrix to the system, For the system output matrix, The control input of an automatic control system For the operator's control input; The collision avoidance risk sub-utility function is as follows: Keeping the current control signals and control permissions unchanged, iteratively (1) construct the future first... The position status of the end effector of the heavy-duty robotic arm in the step, and its control permissions. Functional relationship ; (2)； Constructing a collision avoidance risk sub-utility function: (3)； In the formula: , These are the collision avoidance risk sub-utility functions for the operator and the automatic control system, respectively. This refers to the location information of the obstacle. The human-machine conflict sub-utility function is shown below: Considering the deviation between the operator's and the automatic control system's control signals, a corresponding human-machine conflict sub-utility function is constructed as follows: (4)； In the formula: , These represent the human-machine conflict sub-utility functions of the operator and the automatic control system, respectively. The participation sub-utility function is shown below: To optimize the control process and improve safety, it is desirable for both the operator and the automatic control system to participate in collaborative control, without granting either party complete control authority. Therefore, the participation sub-utility functions of the operator and the automatic control system are constructed as follows: (5)； In the formula: , These are the participation sub-utility functions for the operator and the automatic control system, respectively. For coefficients; By combining the aforementioned sub-utility functions, a utility function for the operator and the automatic control system is constructed: (6)； In the formula: , These are the utility functions for the operator and the automatic control system, respectively. These are the normalization coefficients; Construct the Nash product according to equation (6) And solve for the optimal control authority for human-machine collaboration in heavy-duty robotic arms. : (7)。 2. The method for human-machine collaborative control of a heavy-duty robotic arm based on two-layer game theory according to claim 1, characterized in that: In the process of human-machine collaborative control, both the operator and the automatic control system will make their best efforts to make the heavy-duty robotic arm track their respective target trajectories, and will consider the intervention behavior of the other party in collaborative control before executing control actions. The future can be obtained through iterative formula (1). The predicted output expression for the step: (8)； in, Depend on The values ​​at each prediction step are composed of... , , , and These are vectors and matrices obtained through iteration; With the objective of minimizing the deviation between the control signal and the actual position from the target trajectory, cost functions for both the operator and the automatic control system can be constructed: (9)； In the formula: and These are the penalty coefficient matrices for the target trajectory deviation and the control quantity, respectively. and These represent the cost functions of the automatic control system and the operator, respectively. In non-cooperative games, there exists a set of control combinations such that unilaterally changing the control inputs does not allow any participant to gain more benefits. Similarly, in this study, the operator and the automatic control system also have a Nash equilibrium solution. By using the distributed model predictive control method, the final Nash equilibrium strategy can be solved. (10)； in, This represents the Nash equilibrium strategy of the human-machine collaborative control system for heavy-duty robotic arms; By integrating the optimal control permissions allocated at the current moment. and the Nash equilibrium solution for both human and machine Obtain the optimal collaborative control signal for the heavy-duty robotic arm. To achieve human-machine collaborative control of heavy-duty robotic arms based on two-level game theory; (11)； in, This is the optimal collaborative control signal for the heavy-duty robotic arm at the current moment.

Images