Multi-person cooperation teleoperation method based on self-adaptive authority and robot

Abstract

The invention discloses a multi-person cooperation teleoperation method based on self-adaptive permission and a robot. The method comprises the steps that according to a first motion state corresponding to a local end actuator and a second motion state corresponding to a remote end robot, an interval type-2 polynomial fuzzy model is constructed based on a multi-person cooperation kinetic equation; based on the evaluator neural network and the interval type-2 polynomial fuzzy model, dynamically allocating execution instruction weights to the external input of the local end executor; performing deviation state conversion processing on the first motion state and the second motion state to obtain a converted first motion state and a converted second motion state; according to the transformed first motion state and the transformed second motion state, a controller is constructed through an interval type-2 fuzzy logic system and a multi-person cooperation kinetic equation; and performing reverse adjustment on the controller parameters based on the feedback state of the controller to generate a stable controller. Therefore, a solution with robustness and safety can be provided for multi-person cooperative teleoperation.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of robot control, and particularly relates to a multi-person collaborative teleoperation method based on adaptive authority and a robot. BACKGROUND

[0002] In recent years, the stability of traditional teleoperation relying on single remote operation has led to the wide application of adaptive control and passive methods, however, these methods rely on the accuracy of a single operator, increasing the cognitive load and fatigue. Multi-lateral teleoperation systems reduce the cognitive and physical workload of single operation, thereby reducing human operation errors caused by subjective factors. Their control framework can be classified according to the communication signals used. One method involves each human operator communicating with a remote robot through different signal streams, exploring a dual-local-single-remote teleoperation structure to achieve motion and force control in the subsystem. The hybrid position and force module in dual-hand teleoperation may encounter challenges related to multi-sampling rates. The second type of method mainly involves human operators sharing interaction information within a remote environment, combining local commands with weights, and normalization techniques are most commonly used to assign constant weights to local instructions through an authority factor in a dual-user haptic system. However, the constant authority factor used above is sensitive to large weight error commands, which can disrupt the coordination of multiple operators.

[0003] Irregular environmental geometry often results in asymmetric Cartesian position constraints. Traditional barrier Lyapunov functions avoid these state constraints by directly incorporating pre-set state constraints into stability analysis; for example: logarithmic and tangent barrier-Lyapunov functions ensure that the system output is always within the constraint range, although these methods are effective for symmetric constraints, they are tailored for single / one-sided teleoperation systems and cannot be directly used to handle multi-person collaborative teleoperation with asymmetric constraints. Therefore, it is still challenging to address asymmetric constraints in multi-person collaborative teleoperation.

[0004] The prior art discloses a multi-person collaborative mobile robot parallel multi-instruction remote control method and system. The method includes acquiring external environment information of the mobile robot and extracting a target point position therefrom; according to the target point position, N subjects collaboratively plan a local path of the mobile robot, and assign a current control instruction to each subject; in response to the result of the currently assigned control instruction, each subject completes a coding task of the corresponding control instruction according to the corresponding paradigm; real-time acquisition of electroencephalogram signals of each subject, parallel decoding of subject intentions according to the acquired electroencephalogram signals, and conversion of the subject intentions into a plurality of continuous control instructions to realize multi-instruction remote control of the mobile robot.

[0005] The disadvantages of the prior art are mainly the limitations of the multi-person collaborative interaction method: 1) irregular environments are prone to asymmetric Cartesian position constraints, and traditional barrier Lyapunov functions (for example: logarithmic, tangent, exponential) are only applicable to single-person single-side teleoperation systems and cannot be directly adapted to multi-side collaborative operation systems; 2) traditional normalization techniques assign constant authority factors, which are sensitive to large-weight error instructions and disrupt multi-operator collaboration. SUMMARY

[0006] The application provides a multi-person collaborative teleoperation method and robot based on adaptive authority.

[0007] According to a first aspect of an embodiment of the application, a multi-person collaborative teleoperation method based on adaptive authority is provided, which is applied to any local end effector in the multi-person collaborative teleoperation: the method comprises: constructing an interval type-2 polynomial fuzzy model based on multi-person collaborative dynamics equations according to a first motion state corresponding to the local end effector and a second motion state corresponding to a remote end robot; dynamically assigning execution instruction weights to local end effector external inputs based on an evaluator neural network and the interval type-2 polynomial fuzzy model; performing bias state transformation processing on the first motion state and the second motion state respectively to obtain transformed first motion state and transformed second motion state; constructing a controller through the interval type-2 fuzzy logic system and the multi-person collaborative dynamics equations according to the transformed first motion state and the transformed second motion state; and performing reverse adjustment on the controller parameters based on the feedback state of the controller to generate a stable controller.

[0008] Optionally, according to the first motion state corresponding to the local end executor and the second motion state corresponding to the remote end robot, an interval type-2 polynomial fuzzy model is constructed based on multi-person cooperation dynamics equation; the first motion state corresponding to the local end executor is taken as a first premise variable corresponding to a first polynomial fuzzy set in a rule base of an interval type-2 fuzzy logic system, and upper and lower membership degrees of a rule trigger strength are determined based on the first premise variable; the second motion state corresponding to the remote end robot is taken as a second premise variable corresponding to a second polynomial fuzzy set in the rule base of the interval type-2 fuzzy logic system, and a rule trigger strength corresponding to the second premise variable is determined based on the second premise variable; an environmental force of the remote end robot corresponding to the second premise variable and a human-machine interaction force of the local end executor corresponding to the first premise variable are obtained from the rule base of the interval type-2 fuzzy logic system; based on the human-machine interaction force corresponding to the first premise variable and the upper and lower membership degrees, an interval type-2 polynomial fuzzy model for determining an output of the local end executor is constructed according to the multi-person cooperation dynamics equation; based on the environmental force corresponding to the second premise variable and the upper and lower membership degrees, an interval type-2 polynomial fuzzy model for determining an output of the remote end robot is constructed according to the multi-person cooperation dynamics equation.

[0009] Optionally, the dynamic distribution of execution instruction weights of external inputs of the local end executor based on the evaluator neural network and the interval type-2 polynomial fuzzy model comprises: determining a local position error based on the local end executor position and the remote end robot position output by the interval type-2 polynomial fuzzy model, and taking the local position error as a position synchronization index of the local end executor; determining a local interaction force error based on the human-machine interaction force of the local end executor and the environmental force felt by the remote end robot output by the interval type-2 polynomial fuzzy model, and taking the local interaction force error as a force tracking composite index of the local end executor; constructing an evaluator neural network corresponding to the local end executor based on the position synchronization index and the force tracking composite index of the local end executor; determining a periodic performance index corresponding to the local end executor based on the evaluator neural network; and determining the instruction weights corresponding to the external inputs of the local end executor based on the periodic performance index corresponding to each of the local end executors in the local end; wherein the external inputs at least include human-machine interaction forces.

[0010] Optionally, the deviation state transformation processing is performed on the first motion state and the second motion state respectively to obtain a transformed first motion state and a transformed second motion state; the processing comprises: performing symmetric constraint transformation processing on the motion variables in the first motion state to generate a transformed first motion state adapted to EBLF processing; and performing symmetric constraint transformation processing on the motion variables in the second motion state to generate a transformed second motion state adapted to EBLF processing.

[0011] Optionally, the controller is constructed based on the transformed first motion state and the transformed second motion state through the interval type-2 fuzzy logic system and the multi-player collaborative dynamics equation; the transformed first motion state is taken as a third premise variable corresponding to a polynomial fuzzy set in a rule base of the interval type-2 fuzzy logic system; the transformed second motion state is taken as a fourth premise variable corresponding to the polynomial fuzzy set in the rule base of the interval type-2 fuzzy logic system; a local end executor corresponding interval type-2 fuzzy feedback controller is obtained through fuzzy reasoning and trigger strength calculation based on the third premise variable and the multi-player collaborative dynamics equation; a remote end robot corresponding interval type-2 fuzzy feedback controller is obtained through fuzzy reasoning and trigger strength calculation based on the fourth premise variable and the multi-player collaborative dynamics equation.

[0012] Optionally, the method further comprises: constructing a local end executor corresponding feedforward controller in a steady state stage and a transient state stage respectively through parameter estimation compensation based on the local end executor corresponding interval type-2 fuzzy feedback controller and the multi-player collaborative dynamics equation; constructing a local end executor corresponding feedforward controller in a steady state stage and a transient state stage respectively through parameter estimation compensation based on the remote end robot corresponding interval type-2 fuzzy feedback controller and the multi-player collaborative dynamics equation.

[0013] Optionally, the controller parameter is adjusted reversely based on the feedback state of the controller to generate a stable controller; the adjusting comprises: updating a time delay derivative upper bound estimation value and a dynamics parameter estimation value of the multi-player collaborative dynamics equation respectively based on the feedback state of the controller; constructing a time delay compensation term of the controller based on the updated time delay derivative upper bound estimation value; constructing a parameter compensation term of the controller based on the updated dynamics parameter; updating the controller parameter based on the time delay compensation term and the parameter compensation term to generate a stable controller.

[0014] Optionally, the controller parameter is updated based on the time delay compensation term and the parameter compensation term to generate a stable controller; the updating comprises: updating the controller parameter based on the time delay compensation term and the parameter compensation term to generate an updated controller; performing stability verification on the updated controller to generate a stable controller.

[0015] According to a second aspect of the embodiments of the present application, there is further provided a multi-player collaborative teleoperation device based on adaptive permissions, which is applied to any local end executor in the multi-player collaborative teleoperation, and comprises: a model construction module, configured to construct an interval two-type polynomial fuzzy model based on a multi-player collaborative dynamics equation according to a first motion state corresponding to the local end executor and a second motion state corresponding to a remote end robot; an instruction weight distribution module, configured to dynamically distribute an execution instruction weight of a local end executor external input based on an evaluator neural network and the interval two-type polynomial fuzzy model; a state transformation processing module, configured to perform bias state transformation processing on the first motion state and the second motion state respectively to obtain a transformed first motion state and a transformed second motion state; a controller construction module, configured to construct a controller by an interval two-type fuzzy logic system and a multi-player collaborative dynamics equation according to the transformed first motion state and the transformed second motion state; and a generation module, configured to perform reverse adjustment on controller parameters based on a feedback state of the controller to generate a stable controller.

[0016] According to a third aspect of the embodiments of the present application, there is further provided a computer readable medium having a computer program stored thereon, the program being executed by a processor to implement the method according to the first aspect.

[0017] According to a fourth aspect of the embodiments of the present application, there is further provided a robot, which performs the multi-player collaborative teleoperation method according to the first aspect.

[0018] The embodiments of the present application provide a multi-player collaborative teleoperation method and device based on adaptive permissions and a computer readable medium. The method comprises the following steps. Firstly, an interval two-type polynomial fuzzy model is constructed based on a multi-player collaborative dynamics equation according to a first motion state corresponding to a local end executor and a second motion state corresponding to a remote end robot. Secondly, an execution instruction weight of a local end executor external input is dynamically distributed based on an evaluator neural network and the interval two-type polynomial fuzzy model. Thirdly, bias state transformation processing is performed on the first motion state and the second motion state respectively to obtain a transformed first motion state and a transformed second motion state. Finally, a controller is constructed by an interval two-type fuzzy logic system and a multi-player collaborative dynamics equation according to the transformed first motion state and the transformed second motion state. Finally, the controller parameters are adjusted in reverse based on the feedback state of the controller to generate a stable controller. This embodiment effectively captures uncertainties in human-computer interaction and robot-environment interaction through an interval type-2 polynomial fuzzy model. By using the allocation mechanism of the evaluator neural network, the position synchronization index and force tracking composite index determined by the interval type-2 polynomial fuzzy model are used to adjust the command weights corresponding to the local actuators. This allows for real-time adjustment of the command weights of each operator, avoiding the problem of erroneous commands dominating control under fixed command weights and reducing human error. Asymmetric constraints are transformed into symmetric constraints through deviation state transformation processing, and a stable controller is constructed by combining the interval type-2 fuzzy logic system with inverse adjustment. This specifically addresses the core pain points of traditional teleoperation systems in multi-person collaborative scenarios, such as asymmetric constraints and time delay uncertainties. This embodiment provides a robust and safe solution for multi-person collaborative teleoperation. Attached Figure Description

[0019] The following sections will describe some specific embodiments of the invention in detail by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 This is a flowchart illustrating a multi-user collaborative teleoperation method based on adaptive permissions, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of a multi-user collaborative teleoperation device based on adaptive permissions, provided in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0021] like Figure 1 The diagram shown is a flowchart illustrating a multi-user collaborative teleoperation method based on adaptive permissions, according to an embodiment of the present invention.

[0022] A method for collaborative teleoperation based on adaptive permissions, the method being applied to any local executor in the collaborative teleoperation: the method includes at least the following steps: S101, constructing an interval type-2 polynomial fuzzy model based on a multi-person collaborative dynamics equation according to a first motion state corresponding to the local end effector and a second motion state corresponding to the remote end robot; S102, dynamically distributing an execution instruction weight to the local end effector based on an evaluator neural network and the interval type-2 polynomial fuzzy model; S103, performing bias state transformation processing on the first motion state and the second motion state respectively to obtain a transformed first motion state and a transformed second motion state; S104, constructing a controller through an interval type-2 fuzzy logic system and the multi-person collaborative dynamics equation according to the transformed first motion state and the transformed second motion state; S105, performing reverse adjustment on controller parameters based on a feedback state of the controller to generate a stable controller.

[0023] It should be noted that the multiple local ends in the present application include two scenarios: (1) one operator simultaneously operates multiple local end effectors through natural body control; and (2) multiple operators respectively operate their own single local end effectors.

[0024] In S101, an interval type-2 polynomial fuzzy logic system is introduced to construct an interval type-2 polynomial fuzzy model based on a multi-person collaborative dynamics equation according to a first motion state corresponding to the local end effector and a second motion state corresponding to the remote end robot.

[0025] For example, the first motion state corresponding to the local end effector is taken as a first premise variable corresponding to a first polynomial fuzzy set in a rule base of the interval type-2 fuzzy logic system, and the upper and lower membership degrees of the corresponding rule trigger strength are determined based on the first premise variable; the second motion state corresponding to the remote end robot is taken as a second premise variable corresponding to a second polynomial fuzzy set in the rule base of the interval type-2 fuzzy logic system, and the corresponding rule trigger strength is determined based on the second premise variable; the environmental force of the remote end robot corresponding to the second premise variable and the human-machine interaction force of the local end effector corresponding to the first premise variable are obtained from the rule base of the interval type-2 fuzzy logic system; based on the human-machine interaction force corresponding to the first premise variable and the upper and lower membership degrees, an interval type-2 polynomial fuzzy model for determining the output of the local end effector is constructed according to the multi-person collaborative dynamics equation; based on the environmental force corresponding to the second premise variable and the upper and lower membership degrees, an interval type-2 polynomial fuzzy model for determining the output of the remote end robot is constructed according to the multi-person collaborative dynamics equation.

[0026] For example, the multi-person collaborative dynamics equation can be expressed in the task space as formula (1) as follows:

[0027]

[0028] Equation (1); wherein, , is the number of local actuators. , , are the position, velocity and acceleration of the local actuators, respectively. is the inertia matrix of the local actuators, is the centripetal force and Coriolis torque of the local actuators, is the gravitational torque of the local actuators. and are the feedback control input and human-robot interaction force of the local actuators, respectively, and are the feedforward control input of the local interface to be designed and the feedforward control input of the remote robot, respectively, and are the feedback control torque and environmental force of the remote robot, respectively, is the inertia matrix of the remote robot, is the centripetal force and Coriolis torque of the remote robot, is the gravitational torque of the remote robot; , , are the position, velocity and acceleration of the remote robot, respectively. The interval type-2 fuzzy logic system rule base includes p rules; here, only the rules in the p rules are exemplarily listed.

[0029] Rule m: if belongs to , …, and is then ; Rule i: if belongs to , …, and belongs to then ; wherein, is the first motion state corresponding to the local actuators, is the second motion state corresponding to the remote robot; is the first polynomial fuzzy set corresponding to the first premise variable in the rule m, is the number of elements in the first polynomial fuzzy set, which is the element function of the first premise variable; is the second polynomial fuzzy set corresponding to the second premise variable corresponding second polynomial fuzzy set, is the number of elements in the second polynomial fuzzy set, which is the element function of the second antecedent variable, and respectively represent the maximum eigenvalue in the equivalent inertia matrix, damping matrix and stiffness matrix of the local end effector, is the bounded exogenous force related to the local end effector position. and respectively represent the maximum eigenvalue in the equivalent inertia matrix, damping matrix and stiffness matrix of the remote end robot,

[0030] is the external force related to the environment position of the remote end robot. the upper membership degree and the lower membership degree of the first antecedent variable are determined based on the corresponding element function of the first antecedent variable; the upper membership degree and the lower membership degree of the second antecedent variable are determined based on the corresponding element function of the second antecedent variable. The triggering strength of the rule m and the rule i is the following interval set:

[0031] wherein, and respectively represent the lower membership function and the upper membership function, satisfying . Similarly, the relevant variable with subscript r is the corresponding variable of the remote end robot, which is not repeated here.

[0032] The expression of the interval bivariate polynomial fuzzy model of the local end effector output can be various, for example: expressed by , or . The expression of the interval bivariate polynomial fuzzy model of the remote end robot output can be various, for example: expressed by , or .

[0033] wherein, and satisfy , satisfy . .

[0034] ​​​​The embodiment can simultaneously model the intra-individual and inter-individual uncertainties through a polynomial fuzzy set. In the interactive process of multi-person collaborative teleoperation, an interval biquadratic polynomial fuzzy logic system is introduced to model the multi-person collaborative teleoperation. The model can capture the uncertainty of the nonlinear and time-varying parameters in the force sensation and the environmental force. Compared with the T-S fuzzy system which needs a large number of rules to approximate the nonlinearity, the interval biquadratic polynomial fuzzy model can realize stronger nonlinearity representation with fewer rules, which can effectively reduce the computational complexity when describing the environmental interaction of the remote robot, thereby improving the real-time control response speed of the remote robot.

[0035] In S102, the evaluator neural network is a practical method of reinforcement learning, which takes the state S as input and outputs the expected value. The main function of the evaluator neural network is to evaluate the value of the current state, thereby providing guidance for the optimization of the strategy. In reinforcement learning, the actions of the robot and the environmental feedback have randomness. If the actual reward value of each time is directly used to update the strategy, it will cause large uncertainty due to random fluctuations, thereby leading to low learning efficiency of the robot. The evaluator neural network does not look at the single random reward, but estimates the expected reward value to provide a more stable direction for strategy optimization. For example, "the expected value of this state is high, so adjust the action in this direction more", which avoids being biased by single random results, makes the learning process of the robot smoother and more focused on effective strategies, thereby improving the overall efficiency. That is, the evaluator neural network is a practical tool in reinforcement learning to "find the optimization direction" for the strategy, which reduces the interference caused by randomness by estimating the state value, and makes the learning more efficient.

[0036] For example, according to the local end effector position output by the interval biquadratic polynomial fuzzy model and the remote end robot position, the local position error is determined, and the local position error is taken as the position synchronization index of the local end effector; based on the local end effector interaction force output by the interval biquadratic polynomial fuzzy model and the environmental force felt by the remote end robot, the local interaction force error is determined, and the local interaction force error is taken as the force tracking composite index of the local end effector; based on the position synchronization index and the force tracking composite index of the local end effector, the evaluator neural network corresponding to the local end effector is constructed; based on the evaluator neural network, the periodic performance index corresponding to the local end effector is determined; based on the periodic performance index corresponding to each local end effector in the local end, the instruction weight corresponding to the external input in the local end effector is determined; wherein the external input at least includes human-robot interaction force.

[0037] For example, the expression of the evaluator neural network is as follows: Formula (4); wherein, is the approximation error, is the ideal weight matrix with estimation, Select as the Gaussian activation function, , is the number of neurons; local position error , local interaction force error .

[0038] The periodic performance index corresponding to the jth local end effector is Let , which represents the sum of the performance indexes of all local end effectors except the jth local end effector. By normalizing the relevant value function, the instruction weight corresponding to the jth local end effector can be obtained, as shown in the following formula.

[0039] Formula (5).

[0040] The embodiment combines the position synchronization index and the force tracking composite index to dynamically balance the instruction weight distribution of multiple operators, adaptively strengthens the influence of high-quality operations, weakens the negative impact of incorrect instructions, and avoids the sensitivity of constant weight to incorrect instructions. Moreover, the embodiment is based on the periodic task performance of the operator, reduces the interference of instantaneous errors through periodic evaluation, and improves the collaborative stability.

[0041] In S103, the motion variables in the first motion state are subjected to symmetric constraint transformation processing to generate a transformed first motion state suitable for EBLF processing; and the motion variables in the second motion state are subjected to symmetric constraint transformation processing respectively to generate a transformed second motion state suitable for EBLF processing.

[0042] For example: by introducing auxiliary variables, the asymmetric constraint is converted into a symmetric constraint, which is suitable for EBLF processing and simplifies the control design. The symmetric constraint transformation processing relationship is shown in the following formula.

[0043] Formula (6); , , and are bias constraint functions with continuous differentiability and monotonic convergence. Using the block matrix technique, the following formula can be obtained from the above formula and , , and are n-dimensional unit matrices and 0 matrices, respectively. is the Moore-Penrose inverse matrix of matrix p. Similarly, the following formulas can be obtained and .

[0044] In S104, the construction method of the controller is not limited here. In an embodiment of the present application, only a feedback controller is constructed. In another embodiment of the present application, a feedback controller and a feedforward controller are constructed simultaneously.

[0045] Exemplarily, the controller is constructed by the interval type-2 fuzzy logic system and the multi-player cooperative dynamics equation according to the transformed first motion state and the transformed second motion state; including: taking the transformed first motion state as a third premise variable corresponding to a polynomial fuzzy set in a rule base of the interval type-2 fuzzy logic system; and taking the transformed second motion state as a fourth premise variable corresponding to a polynomial fuzzy set in the rule base of the interval type-2 fuzzy logic system; obtaining an interval type-2 fuzzy feedback controller corresponding to the local end effector based on the third premise variable and the multi-player cooperative dynamics equation through fuzzy reasoning and trigger strength calculation; and obtaining an interval type-2 fuzzy feedback controller corresponding to the remote end robot based on the fourth premise variable and the multi-player cooperative dynamics equation through fuzzy reasoning and trigger strength calculation.

[0046] For example, the design of the interval type-2 fuzzy feedback controller is as follows.

[0047] Consistent with the interval type-2 polynomial fuzzy model modeling rule, an interval type-2 fuzzy feedback controller containing p rules is designed to generate a control quantity adapting to multiple constraints and uncertainties, and to fuse fuzzy logic and polynomial model to process nonlinearity and uncertainty. There are various representation methods for the interval type-2 fuzzy feedback controller corresponding to the local end effector and the interval type-2 fuzzy feedback controller corresponding to the remote end robot. Hereinafter, the interval type-2 fuzzy feedback controller corresponding to the local end effector is represented by The interval type-2 fuzzy feedback controller corresponding to the local end effector is represented by The interval type-2 fuzzy feedback controller corresponding to the remote end robot is represented by the following formula.

[0048] Formula (7); Wherein, is a premise variable fusing the states of the local end effector and the remote end robot, is a polynomial stiffness matrix, is a known regression matrix, is a parameter estimation value, is a time delay compensation term containing a sign function for anti-interference, is a time delay derivative compensation term, is a controller parameter, guarantees the effectiveness of the deviation state transformation, compensates for the influence of centrifugal force and Coriolis force, ensures that the EBLF energy attenuation improves stability.

[0049] Exemplarily, according to the interval type-2 fuzzy feedback controller corresponding to the local end executor and the multi-person cooperation dynamics equation, a corresponding feedforward controller of the local end executor in the steady state stage and the transient state stage is respectively constructed in a manner of parameter estimation compensation; according to the interval type-2 fuzzy feedback controller corresponding to the remote end robot and the multi-person cooperation dynamics equation, a corresponding feedforward controller of the local end executor in the steady state stage and the transient state stage is respectively constructed in a manner of parameter estimation compensation.

[0050] For example: feedforward controller design: the feedforward controller is divided into a feedforward controller corresponding to the local end executor and a feedforward controller corresponding to the remote end robot , both of which adopt a “fuzzy rule weighting + parameter estimation compensation” structure, and the feedforward control is used to offset the known disturbances such as gravity and external force, and is mainly divided into two stages of transient state and remote stage.

[0051] In the local end executor, the feedforward control in the transient state stage mainly offsets the known exogenous force, such as gravity and human-computer interaction force, and the feedforward control in the steady state stage mainly matches the remote stiffness term to enhance the force feedback effect. In the remote end robot, the feedforward control in the transient state stage mainly offsets the gravity and environmental force, and the feedforward control in the steady state stage only retains the gravity compensation. The calculation formulas of the two stages are as follows: ; ; formula (8) wherein, represents a set of positive integers of [1, p], , , can be directly obtained from the inertia matrix, wherein, . is a user-defined constraint function with n-order continuous derivative. is the adjustment time extracted from the constraint function and . The fuzzy rule triggering strength is consistent with the rule triggering strength of the interval type-2 fuzzy feedback controller, and is represented as follows;

[0052] formula (9) wherein, and are embedded membership functions, satisfying and . is a time function, satisfying and .

[0053] In S105, based on the adaptive update rate, the controller parameters are adjusted inversely according to the feedback state of the controller, and a stable controller is generated.

[0054] For example, based on the feedback state of the controller, the time delay derivative upper bound estimate and the dynamic parameter estimate of the multi-person cooperation dynamics equation are updated in real time, respectively; based on the updated time delay derivative upper bound estimate, a time delay compensation term of the controller is constructed; based on the updated dynamic parameters, a parameter compensation term of the controller is constructed; based on the time delay compensation term and the parameter compensation term, the controller parameters are updated, and a stable controller is generated.

[0055] Specifically, first, the time delay derivative upper bound estimate and the dynamic parameter estimate are performed based on the adaptive law design, and the parameters of the controller are dynamically optimized; second, the constructed global function is used to verify that the robot state is bounded, the constraint is reasonable and the error converges, and the final controller parameters and adaptive law parameters are output, so as to generate a stable controller.

[0056] The present application explicitly extends the exponential barrier Lyapunov function to realize the multi-constraint control algorithm. In addition, the bias constraint function And can effectively handle the asymmetric constraint problem without changing the control framework.

[0057] For example: (1) adaptive law parameter update The core of the adaptive update rate design is "dynamic adjustment based on state feedback". Two types of adaptive laws are used to handle time delay uncertainty and dynamic parameter uncertainty, respectively. The update rate parameters (such as learning rate, decay term) are optimized through stability derivation to ensure that the system stability is not destroyed while tracking uncertainty; its execution is deeply coordinated with the controller and constraint processing throughout the whole process, which is the key support for the system to realize high-precision position synchronization and force tracking. The estimation object of the adaptive law is the time delay derivative upper bound estimate and the dynamic parameter estimate. According to the auxiliary variable which reflects the constraint satisfaction degree and the error size, the estimate value is dynamically adjusted to ensure the real-time and accuracy of the feedforward compensation and feedback control. The unknown parameters of the feedforward controller are updated in real time through the adaptive law, forming a "compensation-estimation-optimization" closed loop. The specific formula is as follows: Equation (10); The adaptive law adjusts according to the auxiliary variable which reflects the constraint satisfaction degree and the tracking error, and the feedforward controller compensates the uncertainty of the updated parameters, and the two cooperate to reduce the system error.

[0058] Learning rate: ; Adapt the output characteristics of regression matrix Y to balance convergence speed and robustness. Decay term coefficient ; Suppress parameter estimation oscillation and ensure boundedness of estimation value. Regression matrix : composed of known system states (e.g. error e, auxiliary variable v, bias constraint function ), establish the association between parameter estimation and system dynamics. Another estimation object of adaptive update rate, the upper bound of time delay derivative update rate is expressed as equation (11):

[0059] Equation (11); Learning rate : adjust the update sensitivity to control the speed of estimation value tracking unknown parameters. Decay term coefficient : introduce exponential decay to prevent unbounded growth of estimation value and ensure stability. Input terms: reflect the system state error and constraint satisfaction degree to ensure that the update is associated with the actual working condition of the system. By calculating the rate of change of the estimation value and , and updating the current estimation value, the estimation value is fed back to the controller: the updated is used to construct the time delay compensation term, is used for parameter compensation in feedforward compensation and feedback control, directly optimizing the control amount of force.

[0060] (2) Constraint adaptation and stability check The key to constraint adaptation is "bias transformation + EBLF guarantee", which converts asymmetric constraints into symmetric constraints that can be processed by the controller; the core of stability check is "fuzzy Lyapunov-Krasovskii functional construction and derivative analysis", which verifies the boundedness of system state, constraint satisfaction and performance convergence. Both of them constitute the "constraint-stability" double guarantee of the system, which is the core support for the reliable operation of the multi-local-single-remote teleoperation system under time delay and uncertainty conditions. Constraint processing provides "operable boundaries" for the controller: through bias state transformation, asymmetric constraints are converted into symmetric constraints, so that the controller does not need to directly process complex asymmetric boundaries, and only needs to ensure that the auxiliary variable does not exceed the boundary.

[0061] The controller guarantees constraint satisfaction through "active suppression + compensation": the constraint guarantee term in the feedback controller directly suppresses the auxiliary variable from exceeding the boundary, and the feedforward controller compensates for uncertainties to reduce error accumulation and avoid triggering constraint violation due to excessive error. The feedforward controller needs to cooperate with the bias state transformation + EBLF to avoid the control amount exceeding the actuator range or violating the position constraint.

[0062] Stability closed loop verification: The feedforward controller is incorporated into the fuzzy Lyapunov-Krasovskii functional V=V1 (multi-constrained EBLF) +V2 (parameter estimation error term) +V3 stability auxiliary term, through derivation , it is verified that the feedforward compensation does not destroy the system stability, and can promote the force tracking error , ensuring the dual goals of "constraint satisfaction" and "system stability".

[0063] Most teleoperation strategies focus on the time-delay stability of single local-single remote teleoperation systems, but lack mechanisms to correct operation errors caused by time delay. In addition, state constraints further increase the complexity of control design. To solve these problems, the embodiment proposes an effective solution for a multi-local-single remote teleoperation system with constraint requirements. First, the human-machine interaction process and the robot environment interaction process are modeled as an interval type-2 polynomial fuzzy model, wherein the nonlinearity and parameter uncertainty are effectively captured by the interval type-2 polynomial fuzzy model. A distribution mechanism based on the evaluator neural network uses periodic position synchronization indicators and force tracking composite indicators to adjust the operator's instruction weight. The adaptive interval type-2 polynomial fuzzy model control framework converts asymmetric constraints into easy-to-handle symmetric constraints through a novel bias state transformation technique, while strengthening the transient / steady-state performance and ensuring the satisfaction of asymmetric constraints. This paper strictly proves that the position synchronization and force tracking performance can be guaranteed by introducing a fuzzy Lyapunov-Krasovskii functional.

[0064] The present application proposes a multi-operator collaboration method for a multi-edge collaborative teleoperation system in the field of robot control, which realizes adaptive collaborative control framework. Through interval type-2 polynomial fuzzy model system modeling, evaluator neural network permission setting, exponential barrier Lyapunov constraint processing, the system state is guaranteed to be bounded, asymmetric constraints are satisfied, and force tracking performance is guaranteed.

[0065] As Figure 2 shown, it is a structural schematic diagram of a multi-person collaborative teleoperation device based on adaptive permissions provided by an embodiment of the present application.

[0066] A multi-person collaborative teleoperation device based on adaptive authority, the device is applied to any local end executor in the multi-person collaborative teleoperation: the device 200 comprises: a model construction module 201, configured to construct an interval bivariate polynomial fuzzy model based on a multi-person collaborative dynamics equation according to a first motion state corresponding to the local end executor and a second motion state corresponding to a remote end robot; an instruction weight distribution module 202, configured to dynamically distribute an execution instruction weight of a local end executor external input based on an evaluator neural network and the interval bivariate polynomial fuzzy model; a state transformation processing module 203, configured to perform bias state transformation processing on the first motion state and the second motion state respectively to obtain a transformed first motion state and a transformed second motion state; a controller construction module 204, configured to construct a controller by an interval bivariate fuzzy logic system and a multi-person collaborative dynamics equation according to the transformed first motion state and the transformed second motion state; and a generation module 205, configured to perform reverse adjustment on controller parameters based on a feedback state of the controller to generate a stable controller.

[0067] In a preferred embodiment of the present embodiment, the model construction module comprises: a first determination unit configured to take the first motion state corresponding to the local end executor as a first premise variable corresponding to a first polynomial fuzzy set in a rule base of an interval bivariate fuzzy logic system, and determine upper and lower membership degrees of a rule trigger intensity based on the first premise variable; a second determination unit configured to take the second motion state corresponding to the remote end robot as a second premise variable corresponding to a second polynomial fuzzy set in the rule base of the interval bivariate fuzzy logic system, and determine a rule trigger intensity corresponding to the second premise variable; an acquisition unit configured to acquire an environmental force of the remote end robot corresponding to the second premise variable and a human-machine interaction force of the local end executor corresponding to the first premise variable from the rule base of the interval bivariate fuzzy logic system; a first construction unit configured to construct an interval bivariate polynomial fuzzy model for determining an output of the local end executor based on the human-machine interaction force corresponding to the first premise variable and the upper and lower membership degrees according to a multi-person collaborative dynamics equation; and a second construction unit configured to construct an interval bivariate polynomial fuzzy model for determining an output of the remote end robot based on the environmental force corresponding to the second premise variable and the upper and lower membership degrees according to the multi-person collaborative dynamics equation.

[0068] In the preferred implementation of the embodiment, the instruction weight distribution module comprises: a first determination unit configured to determine a local position error based on the local end effector position output by the interval bivariate polynomial fuzzy model and the remote end robot position, and take the local position error as a position synchronization indicator of the local end effector; a second determination unit configured to determine a local interaction force error based on the local end effector robot interaction force output by the interval bivariate polynomial fuzzy model and the environment force sensed by the remote end robot, and take the local interaction force error as a force tracking composite indicator of the local end effector; a construction unit configured to construct an evaluator neural network corresponding to the local end effector based on the local position error of the local end effector and the local interaction force error of the local end effector; a third determination unit configured to determine a periodic performance indicator corresponding to the local end effector based on the evaluator neural network; and a fourth determination unit configured to determine an instruction weight corresponding to an external input of the local end effector based on the periodic performance indicator corresponding to each of the local end effectors in the local end.

[0069] In the preferred implementation of the embodiment, the state transformation processing module comprises: a first transformation processing unit configured to perform symmetric constraint transformation processing on the motion variables in the first motion state to generate a transformed first motion state adapted to EBLF processing; and a second transformation processing unit configured to perform symmetric constraint transformation processing on the motion variables in the second motion state respectively to generate a transformed second motion state adapted to EBLF processing.

[0070] In the preferred implementation of the embodiment, the controller construction module comprises: a determination unit configured to take the transformed first motion state as a third premise variable corresponding to a polynomial fuzzy set in a rule base of an interval bivariate fuzzy logic system, and take the transformed second motion state as a fourth premise variable corresponding to a polynomial fuzzy set in a rule base of an interval bivariate fuzzy logic system; a first calculation unit configured to obtain an interval bivariate fuzzy feedback controller corresponding to the local end effector by fuzzy reasoning and trigger strength calculation based on the third premise variable and the multi-person collaborative dynamics equation; and a second calculation unit configured to obtain an interval bivariate fuzzy feedback controller corresponding to the remote end robot by fuzzy reasoning and trigger strength calculation based on the fourth premise variable and the multi-person collaborative dynamics equation.

[0071] In the preferred implementation of the embodiment, the controller constructing module further comprises: a first constructing module for constructing, in a manner of parameter estimation compensation, a corresponding feedforward controller of the local end executor in a steady state stage and a transient state stage respectively according to the corresponding interval two-type fuzzy feedback controller of the local end executor and the multi-player cooperation dynamics equation; and a second constructing module for constructing, in a manner of parameter estimation compensation, a corresponding feedforward controller of the local end executor in a steady state stage and a transient state stage respectively according to the corresponding interval two-type fuzzy feedback controller of the remote end robot and the multi-player cooperation dynamics equation.

[0072] In the preferred implementation of the embodiment, the generating module comprises: a constructing unit for updating the upper limit estimation value of the time delay derivative and the dynamic parameter estimation value of the multi-player cooperation dynamics equation respectively in real time based on the feedback state of the controller; constructing a time delay compensation term of the controller based on the updated upper limit estimation value of the time delay derivative; and constructing a parameter compensation term of the controller based on the updated dynamic parameter; and a generating unit for updating the controller parameter based on the time delay compensation term and the parameter compensation term, and generating a stable controller.

[0073] In the preferred implementation of the embodiment, the generating unit comprises: an updating subunit for updating the controller parameter based on the time delay compensation term and the parameter compensation term, and generating an updated controller; and a correction subunit for performing stability verification on the updated controller, and generating a stable controller.

[0074] The multi-player cooperation teleoperation device based on adaptive permissions can execute the multi-player cooperation teleoperation method based on adaptive permissions provided by an embodiment of the application, has the corresponding function modules and beneficial effects of executing the multi-player cooperation teleoperation method based on adaptive permissions. Technical details not described in detail in the embodiment can be referred to the multi-player cooperation teleoperation method based on adaptive permissions provided by an embodiment of the application.

[0075] The application further provides an electronic device comprising: a processor; a memory for storing executable instructions of the processor; and the processor is configured to read the executable instructions from the memory and execute the instructions to implement the multi-player cooperation teleoperation method based on adaptive permissions.

[0076] In addition to the above-mentioned method and device, the embodiments of the application can also be a computer program product comprising computer program instructions, which, when executed by a processor, cause the processor to perform the steps of the methods according to various embodiments of the application described in the above “Exemplary Method” section of the specification.

[0077] The computer program product can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computing device, partly on the user's computing device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device or entirely on the remote computing device or server. The embodiments of the present application are not limited by the

[0078] Furthermore, embodiments of the present application can also be a computer readable storage medium, having stored thereon computer program instructions which, when run by a processor, cause the processor to perform steps of the method according to embodiments of the present application as described in the above "Exemplary Methods" section of the specification.

[0079] The computer readable storage medium can be any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0080] The above describes basic principles of the present application in conjunction with specific embodiments, but it should be noted that the advantages, benefits, effects and the like mentioned in the present application are only examples and are not limiting, and these advantages, benefits, effects and the like cannot be considered as necessary for each embodiment of the present application. In addition, the above specific details are only for the purpose of example and understanding, and the above details do not limit the present application to the specific details described above.

[0081] The block diagrams of the devices, apparatuses, equipment, systems referred to in this application are only illustrative examples and are not intended to require or imply that the connection, arrangement, configuration must be as shown in the block diagrams. As those skilled in the art will recognize, the devices, apparatuses, equipment, systems can be connected, arranged, configured in any manner. Words such as "include", "comprise", "have", etc. are open-ended words that are intended to mean "including but not limited to", and are to be taken in a non-exclusive or a non-limiting sense. The words "or" and "and" as used herein are intended to mean "and / or" unless explicitly indicated to the contrary. The word "such as" as used herein is intended to mean "such as but not limited to", and is to be taken in a non-exclusive or a non-limiting sense.

[0082] It is also to be noted that in the devices, apparatuses and methods of the present application, the various components or steps can be split into and / or combined from other components or steps. These splits and / or combinations are to be considered as equivalents of the present application.

[0083] The above description of the disclosed aspects is given for illustrative purposes and is not intended to limit the application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0084] The above description has been given for illustrative and descriptive purposes. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, changes, additions and sub-combinations.

[0085] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Also, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine the features of different embodiments or examples described in the specification and the features of different embodiments or examples, without contradiction, if necessary.

[0086] In addition, the terms "first", "second", etc. are used only for the purpose of description, and should not be understood as indicating or implying relative importance or implying the number of the technical features indicated. Therefore, the features defined as "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0087] The above description is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A multi-user collaborative teleoperation method based on adaptive permissions, characterized in that, The method is applied to any local actuator in the multi-user collaborative teleoperation: the method includes: Based on the first motion state corresponding to the local actuator and the second motion state corresponding to the remote robot, an interval type II polynomial fuzzy model is constructed based on the multi-person collaborative dynamic equation. The execution instruction weights are dynamically assigned to the external input of the local executor based on the evaluator neural network and the interval type II polynomial fuzzy model. The first motion state and the second motion state are respectively subjected to deviation state transformation processing to obtain the transformed first motion state and the transformed second motion state; Based on the transformed first motion state and the transformed second motion state, a controller is constructed using an interval type II fuzzy logic system and multi-person collaborative dynamic equations; The controller parameters are adjusted in reverse based on the feedback state of the controller to generate a stable controller.

2. The method according to claim 1, characterized in that, Based on the first motion state corresponding to the local actuator and the second motion state corresponding to the remote robot, an interval type II polynomial fuzzy model is constructed based on the multi-person collaborative dynamic equation. The first motion state corresponding to the local executor is used as the first premise variable corresponding to the first polynomial fuzzy set in the rule base of the interval type II fuzzy logic system, and the upper and lower membership degrees of the corresponding rule triggering intensity are determined based on the first premise variable. The second motion state corresponding to the remote robot is used as the second premise variable corresponding to the second polynomial fuzzy set in the rule base of the interval type II fuzzy logic system, and the corresponding rule triggering strength is determined based on the second premise variable. Obtain the environmental forces of the remote robot corresponding to the second premise variable and the human-machine interaction forces of the local actuator corresponding to the first premise variable from the rule base of the interval type II fuzzy logic system; Based on the human-computer interaction force and the upper and lower membership degrees corresponding to the first premise variable, an interval type II polynomial fuzzy model is constructed according to the multi-person collaborative dynamic equation to determine the output of the local actuator. Based on the environmental forces and upper and lower membership degrees corresponding to the second premise variable, an interval type II polynomial fuzzy model is constructed according to the multi-person collaborative dynamic equation to determine the output of the remote robot.

3. The method according to claim 1, characterized in that, The dynamic allocation of execution instruction weights to the external input of the local executor based on the evaluator neural network and the interval type-II polynomial fuzzy model includes: Based on the position of the local actuator and the position of the remote robot output by the interval type II polynomial fuzzy model, the local position error is determined and used as the position synchronization index of the local actuator; based on the human-machine interaction force of the local actuator output by the interval type II polynomial fuzzy model and the environmental force sensed by the remote robot, the local interaction force error is determined and used as the force tracking composite index of the local actuator. Based on the position synchronization index and force tracking composite index of the local actuator, an evaluator neural network corresponding to the local actuator is constructed. Based on the evaluator neural network, the periodic performance index corresponding to the local actuator is determined; Based on the periodic performance index corresponding to each local executor in the local terminal, the instruction weight corresponding to the external input in the local executor is determined; wherein, the external input includes at least human-computer interaction force.

4. The method according to claim 1, characterized in that, The step of performing deviation state transformation processing on the first motion state and the second motion state respectively to obtain the transformed first motion state and the transformed second motion state includes: The motion variables in the first motion state are subjected to symmetric constraint transformation to generate a transformed first motion state adapted to EBLF processing. The motion variables in the second motion state are subjected to symmetric constraint transformation to generate a transformed second motion state adapted to EBLF processing.

5. The method according to claim 1, characterized in that, The step of constructing a controller based on the transformed first motion state and the transformed second motion state, using the interval type-II fuzzy logic system and multi-person collaborative dynamic equations, includes: The transformed first motion state is used as the third premise variable corresponding to the polynomial fuzzy set in the rule base of the interval type II fuzzy logic system; and the transformed second motion state is used as the fourth premise variable corresponding to the polynomial fuzzy set in the rule base of the interval type II fuzzy logic system. Based on the third premise variable and the multi-person collaborative dynamic equation, the interval type II fuzzy feedback controller corresponding to the local actuator is obtained through fuzzy inference and trigger intensity calculation. Based on the fourth premise variable and the multi-person collaborative dynamic equation, the interval type II fuzzy feedback controller corresponding to the remote robot is obtained through fuzzy inference and trigger intensity calculation.

6. The method according to claim 5, characterized in that, Also includes: By using parameter estimation compensation, based on the interval type II fuzzy feedback controller and the multi-person collaborative dynamic equation corresponding to the local actuator, feedforward controllers are constructed for the steady-state and transient phases of the local actuator, respectively. By using parameter estimation and compensation, feedforward controllers are constructed for the local actuator to handle the steady-state and transient phases, based on the interval type II fuzzy feedback controller corresponding to the remote robot and the multi-person collaborative dynamic equations.

7. The method according to claim 1, characterized in that, The step of adjusting the controller parameters in reverse based on the feedback state of the controller to generate a controller includes: Based on the feedback state of the controller, the upper bound estimate of the time delay derivative and the dynamic parameter estimate of the multi-person collaborative dynamic equation are updated in real time; based on the updated upper bound estimate of the time delay derivative, the time delay compensation term of the controller is constructed; based on the updated dynamic parameters, the parameter compensation term of the controller is constructed. Based on the time delay compensation term and parameter compensation term, the controller parameters are updated to generate a stable controller.

8. The method according to claim 7, characterized in that, The step of updating the controller parameters based on the time delay compensation term and parameter compensation term to generate a stable controller includes: Based on the delay compensation term and parameter compensation term, the controller parameters are updated to generate the updated controller; The updated controller is then subjected to stability verification to generate a stable controller.

9. A multi-user collaborative teleoperation device based on adaptive permissions, characterized in that, The device is applied to any local actuator in the multi-person collaborative teleoperation; the device includes: The model building module is used to construct an interval type II polynomial fuzzy model based on the multi-person collaborative dynamic equation, according to the first motion state corresponding to the local actuator and the second motion state corresponding to the remote robot. The instruction weight allocation module is used to dynamically allocate execution instruction weights to the external input of the local executor based on the evaluator neural network and the interval type II polynomial fuzzy model. The state transformation processing module is used to perform deviation state transformation processing on the first motion state and the second motion state respectively to obtain the transformed first motion state and the transformed second motion state. The controller construction module is used to construct a controller based on the transformed first motion state and the transformed second motion state through an interval type II fuzzy logic system and multi-person collaborative dynamic equations. The generation module is used to adjust the controller parameters in reverse based on the feedback state of the controller to generate a stable controller.

10. A computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method as claimed in any one of claims 1-8.

11. A robot, characterized in that, The robot performs the multi-person collaborative teleoperation method as described in any one of claims 1-8.

Images