Human-robot collaborative spatial tethered robot deployment method and system based on parameter optimization

Abstract

The application is a kind of parameter optimization based human-robot collaborative space tethered robot deployment method and system, belongs to the field of human-robot collaborative control based space tethered robot deployment;The method steps are: the related state initialization of physical human-computer interaction system and space tethered robot system;Carrying out physical human-computer interaction, obtaining the control intention of operator, so that the whole space tethered robot deployment process reaches the expected dynamic performance and steady-state performance;Based on the rolling horizon optimization method, the feasible control parameters are optimized to ensure the ultimate consistent asymptotic stability of the closed-loop system;According to the optimized control parameters, the deployment task of space tethered robot is executed;If the deployment task is not completed, return to repeat until the deployment of space tethered robot is completed.The application fully gives play to the cognition of operator to the scene and subjective initiative, makes the deployment process of space tethered robot according to the expected way of people, improves the reliability and safety of space task.

Description

Technical Field

[0001] The present invention belongs to the field of space tethered robot deployment based on human-machine collaborative control, and specifically relates to a human-machine collaborative space tethered robot deployment method and system based on parameter optimization. Background Art

[0002] Deployment of a tethered space robot involves towing a scientific payload from a satellite to its desired on-orbit location using a tethered space robot. This is a crucial step in enabling tethered space robot missions. However, deploying a tethered space robot presents a complex control problem, primarily due to the inherent underactuated dynamics of the system, making it difficult to find a suitable diffeomorphism to derive a cascade paradigm. Furthermore, the system's sole control input is unidirectional tension, which complicates the deployment process considerably.

[0003] At present, most control methods focus on the feasibility of the control effect when designing the control parameters of the system, and cannot take into account the control performance during the deployment process. In addition, once the control parameters of the system are fixed, the control performance of the entire system will also be fixed. These fixed parameter strategies are difficult to cope with the complex requirements of space missions and greatly affect the development of space tethered robots. If human cognition and evaluation of the scene are introduced into the decision-making of space tethered robot deployment and the system parameter strategy is dynamically adjusted, the reliability and safety of space missions can be greatly improved. Therefore, the present invention proposes a human-machine collaborative space tethered robot deployment method based on parameter optimization. Summary of the Invention

[0004] Technical issues to be solved:

[0005] In order to avoid the shortcomings of the existing technology, the present invention provides a human-machine collaborative space tethered robot deployment method and system based on parameter optimization, which introduces the physical human-machine interaction control method into the space tethered robot deployment process, gives full play to the operator's cognition of the scene and subjective initiative, and makes the space tethered robot deployment process proceed in the way people expect, thereby improving the reliability and safety of space missions.

[0006] The technical solution of the present invention is: a method for deploying a human-machine collaborative space tethered robot based on parameter optimization, the specific steps of which are as follows:

[0007] Initialize the relevant states of the physical human-machine interaction system and the space tethered robot system, and determine the feasible control parameter set;

[0008] Performing physical human-machine interaction to obtain the operator's control intention so that the entire deployment process of the space tethered robot achieves desired dynamic performance and steady-state performance, wherein control parameters during the deployment process are within the range of the control parameter set;

[0009] Based on the rolling horizon optimization method, feasible control parameters are optimized to ensure the final consistent asymptotic stability of the closed-loop system;

[0010] Execute the deployment mission of the space tethered robot according to the optimized control parameters;

[0011] If the deployment task is not completed, the process returns to repeat the steps of physical human-machine interaction, optimizing feasible control parameters, and executing the deployment task of the space tethered robot until the deployment of the space tethered robot is completed.

[0012] A further technical solution of the present invention is as follows: in the physical human-machine interaction system, a fast logarithmic sliding surface is initialized to ensure the stability of the human-machine interaction system at the initial moment, and the neural network parameters used to estimate the interaction force between the operator and the hand controller are initialized; in the spatial tethered robot system, a state synchronization scale between the fully driven state and the underdriven state is initialized, and a feasible control parameter set is calculated according to the rolling horizon optimization method. and To ensure the stability of the closed-loop system.

[0013] A further technical solution of the present invention is that the operator's control intention is acquired through a hand controller, and during the human-computer interaction process, the hand controller's dynamic model is expressed in Cartesian space using the Lagrangian method as follows:

[0014]

[0015] in,

[0016] is the acceleration of the hand controller end,

[0017] is the speed of the hand controller end, is the position of the end of the hand control; is the inertia matrix of the hand controller end,

[0018] is the centripetal force / Coriolis force matrix at the end of the hand controller, The gravity vector at the end of the hand controller; represents the control law generated by the controller to complete the trajectory tracking task; Represents the interaction force behavior between the human and the end of the hand controller.

[0019] A further technical solution of the present invention is that the method for achieving the desired dynamic performance and steady-state performance of the entire space tethered robot deployment process is: the estimated force in the interaction process is obtained by the interaction force estimation method based on a neural network. Reference trajectory obtained by impedance control method The motion trajectory of the hand controller end indirectly carries the operator's control intention, and the desired dynamic performance and steady-state performance during deployment are obtained by controlling the end trajectory;

[0020] Control parameter c3 during deployment and motion trajectory of the hand controller end The relationship between them is expressed as:

[0021]

[0022] in A feasible control parameter set The lower boundary of A feasible control parameter set The upper boundary of ; S(x) = κ||x|| is the scaling function with parameter κ>0.

[0023] A further technical solution of the present invention is: the specific process of optimizing feasible control parameters is as follows:

[0024] The control parameters The optimization criteria of c3 are transformed into solving [t0,t f ] time period, where t0 represents the initial time, t f Represents the end time; select T s As the time interval, the whole optimization process is decomposed into N=(t f -t0) / T s optimization subproblems, where N represents [t0,t f ]The number of time intervals in the time period, that is, the number of optimization subproblems; in the time period [t i ,t i+1 ], i=0,1,...,N-1, control parameters The optimization problem of and c3 is expressed as:

[0025]

[0026] The optimization problem needs to satisfy the following constraints:

[0027]

[0028]

[0029]

[0030] u(t)∈[-T max ,-T min ], (11)

[0031]

[0032] in, is the state vector, is a known nonlinear equation vector, is a positive definite diagonal matrix;

[0033] Formula (7) is the loss function of the optimization problem. Its optimization goal is to make the state vector υ(t) as close to 0 as possible while satisfying the constraints, and the loss function (8) It can be calculated by the integral of constraint (8); constraint (9) is the control input of the system, which is the state vector υ(t) and the parameter Equation; Formulas (10), (11) and (12) are the constraints of the sliding surface arrival stage, the actual control input constraints and the system asymptotic stability constraints respectively;

[0034] During the whole parameter optimization process, each time the rolling horizon optimization method is used, the parameters c3 and near-optimal value.

[0035] A further technical solution of the present invention is: if the optimized state synchronization scale parameter c3 obtained by the human-computer interaction part cannot ensure the stability of the closed-loop system by solving the optimization problem, then the state synchronization scale parameter c3 that can ensure the stable operation of the system at the previous moment is selected. The parameter c3 is derived from the feasible control parameter set, thereby carrying out the deployment process of the space tethered robot.

[0036] A further technical solution of the present invention is: to ensure the eventual consistent asymptotic stability of the entire system, the following conditions must always be met:

[0037]

[0038] Where ξ is the expression of the deployment degree of the space tethered robot.

[0039] A further technical solution of the present invention is that the process of executing the deployment task of the space tethered robot is as follows:

[0040] Establish a dynamic model of a space tethered robot;

[0041] Determining the final deployment goal as deploying the space tethered robot to a desired length toward the center of the Earth;

[0042] According to the state synchronization scale between the fully actuated state and the under-actuated state, a state synchronization term is constructed and a sliding surface is designed;

[0043] Designing a control law based on the sliding surface and the dynamic model;

[0044] The control rate is then introduced into the dynamic model of the system to obtain the under-driven part and the fully driven part in the dynamics.

[0045] A further technical solution of the present invention is: the under-driven part is Where θ is the opening and closing angle of the tethered robot, The full drive part is Where u is the force between the ropes,

[0046] A human-machine collaborative space tethered robot deployment system based on parameter optimization, comprising a human-machine interaction system and a space tethered robot system;

[0047] In the human-machine interaction system, the operator's control intention is obtained through the hand controller, and the entire spatial tethered robot deployment process is made to achieve the desired dynamic performance and steady-state performance through the control system;

[0048] In the space tethered robot system, an optimizer is used to optimize the control parameters determined by the human-machine interaction system, and then a control law is obtained through a sliding mode controller design. The control law and the dynamic model are combined through a control system to obtain the under-actuated part and the fully-actuated part in the dynamics, and the deployment of the space tethered robot is executed.

[0049] Beneficial effects

[0050] The beneficial effects of the present invention are as follows: the present invention is aimed at the deployment scenario of a space tethered robot based on human-machine collaborative control, and a sliding surface using a mixed state is designed according to the characteristics of the dynamic model of the space tethered robot to achieve synchronization between the under-driven state and the fully-driven state. In order to give full play to the operator's cognition of the scene and subjective initiative during the deployment of the space tethered robot, the present invention introduces a physical human-machine interactive control method into the process, and uses a neural network-based interactive force estimation method to obtain the operator's control intention, so that the deployment process of the space tethered robot proceeds in the way people expect. In addition, the present invention optimizes the control parameters based on the rolling time domain optimization method and generates a feasibility parameter adjustment strategy. The feasibility parameter adjustment strategy ensures that the space tethered robot system can always maintain asymptotic stability when the system control parameters change. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1This is a system block diagram of a method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to the present invention;

[0052] Figure 2 The impact of different c3 on the deployment process;

[0053] Figure 3 It is the state trajectory of the system during the deployment process under physical human-computer interaction;

[0054] Figure 4 It is the state trajectory of the end effector of the hand controller under physical human-computer interaction. DETAILED DESCRIPTION

[0055] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.

[0056] Based on the difficulties in deploying space tethered robots in the prior art, the present invention proposes a method and system for deploying human-machine collaborative space tethered robots based on parameter optimization. This method introduces a physical human-machine interaction system into the space tethered robot system, and uses a neural network-based interaction force estimation method to obtain the operator's control intention, so that the deployment process of the space tethered robot proceeds in the way people expect. In addition, the present invention optimizes the control parameters based on the rolling time domain optimization method and generates a feasibility parameter adjustment strategy. This feasibility parameter adjustment strategy ensures that the space tethered robot system can always maintain asymptotic stability when the system control parameters change. The specific technical solution of this method is as follows:

[0057] The dynamic model of the space tethered robot can be constructed as follows based on the energy of the system using the Lagrangian method:

[0058]

[0059]

[0060] Where u is the force between the tethers, θ is the opening and closing angle of the tethered robot, and ξ is the expression for the deployment degree of the space tethered robot. The relationship between this and the actual deployment length can be expressed as ξ = l / L-1, where L is the desired total deployment length. When ξ = -1, it means that the tethers are not deployed; when ξ = 0, it means that the tethers are deployed. The ultimate deployment goal of this system is to deploy the space tethered robot to the desired length toward the center of the Earth, which can be expressed as follows:

[0061]

[0062] The dynamic model of the tethered space robot demonstrates that, compared to the state deployment degree ξ, the tethered robot's opening and closing angles represent an underactuated system state, and its changes cannot be directly controlled by system inputs. This invention addresses this system characteristic by linking the fully actuated and underactuated states through a scale parameter, constructing a state synchronization term. A linear sliding surface is designed based on this state synchronization term, ensuring the asymptotic stability of the entire system.

[0063] The sliding surface designed according to the state synchronization term can be expressed as:

[0064]

[0065] Where c1 and c2 are sliding surface parameters, satisfying c1, c2 > 0; c3 is the state synchronization scale, the size of which affects the dynamic and steady-state performance of the system. To ensure the asymptotic stability of the tethered space robot system, the following control law can be designed based on the designed sliding surface and dynamic model:

[0066]

[0067] where u eq is the equivalent control force, u sw The switching term is designed to ensure the gradual stability of the system. The selection of will be discussed later; f1 and f2 can be expressed as:

[0068]

[0069]

[0070] Bringing the above control rate into the system's dynamic model, we can get is the underactuated part of the dynamics, This is the full drive part. To ensure the eventual consistent asymptotic stability of the entire system, the following conditions must always be met:

[0071]

[0072] It can be found that the choice of the state synchronization scale parameter c3 directly affects the final consistent asymptotic stability of the system, and the selection of this parameter is usually made by the operator based on his own experience or task requirements, which has strong uncertainty. And when the state synchronization scale changes, the stability of the system will change. It is necessary to adjust the control parameters The overall optimization of c3 is performed to ensure the stability of the entire system during the deployment of the space tethered robot.

[0073] The present invention aims at the above problems The parameter optimization criteria of c3 are transformed into solving [t0,t f ] time period, where t0 represents the initial time, t f Represents the end time, select T s As the time interval and using the rolling horizon optimization method, the entire optimization process is decomposed into N = (t f -t0) / T s optimization subproblems, where N represents [t0,t f ]The number of time intervals in the time period, that is, the number of optimization sub-problems, is the number of parameters and c3 to perform feasibility optimization and give a more reasonable parameter criterion to ensure that the stability condition (6) of the closed-loop system is established. i ,t i+1 ](i=0,1,...,N-1), parameters The optimization problem of and c3 can be expressed as follows:

[0074]

[0075] The optimization problem needs to satisfy the following constraints:

[0076]

[0077]

[0078]

[0079] u(t)∈[-T max ,-T min ], (11)

[0080]

[0081] in is the state vector, is a known nonlinear equation vector, is a positive definite diagonal matrix. Formula (7) is the loss function of the optimization problem, and its optimization goal is to make the state vector υ(t) as close to 0 as possible while satisfying the constraints, and the loss function It can be calculated by integrating the constraint condition (8). The constraint condition (9) is the control input of the system, which is the state vector υ(t) and the parameter Equation; Formulas (10), (11) and (12) are the constraints of the sliding surface arrival stage, the actual control input constraints and the system asymptotic stability constraints. In the whole parameter optimization process, each time the rolling horizon optimization method is used, the parameters c3 and near-optimal value.

[0082] In order to give full play to the human's understanding of the scene and subjective initiative during the deployment of the space tethered robot, the operator can change the degree of synchronization between the full-drive and under-drive states according to his or her own expectations, thereby obtaining the desired dynamic and steady-state performance. This invention introduces physical human-machine interaction into the deployment process of the space tethered robot, uses the PhantomOmni haptic hand controller to collect human-machine interaction information, and the control parameter c3 during the deployment process is related to the motion trajectory of the hand controller end. The relationship between can be expressed as:

[0083]

[0084] in There is a lower bound for a feasible solution to the above optimization problem. There is an upper bound on the feasible solution of the above optimization problem, and S(x)=κ||x|| is a scaling function with κ>0 as a parameter.

[0085] The dynamic model of the 3-DOF hand controller during human-computer interaction can be organized into the following form in Cartesian space using the Lagrangian method:

[0086]

[0087] in, is the acceleration of the hand controller end, is the speed of the hand controller end, is the position of the end of the hand control; is the inertia matrix of the hand controller end, is the centripetal force / Coriolis force matrix at the end of the hand controller, The gravity vector at the end of the hand controller; represents the control law generated by the controller to complete the trajectory tracking task; Represents the interaction force behavior between the human and the end of the hand controller.

[0088] In order to ensure that the hand controller can accurately and smoothly track the desired trajectory during human-computer interaction, this paper designs a fast logarithmic sliding mode control method. The designed sliding mode surface can be expressed as:

[0089]

[0090] Where e = xx r represents the trajectory tracking error of the hand controller during human-computer interaction. Parameters p and q are positive odd numbers and satisfy 1<p / q<2; β=diag(β1,β2,β3) and satisfy β i >0i,=1,, γ=diag(γ1,γ2,γ3) and satisfy γi >0,i=1,2,3, η=diag(eta1,eta2,eta3) and satisfy eta i >0,i=1,2,3; represents a vector of all ones; sgn(*), |*|, ln(*), and diag(*) represent the sign function, absolute value, logarithmic function, and diagonal matrix of the vector *, respectively.

[0091] To ensure the stability of the system during human-computer interaction, the present invention designs the following control law based on the above-mentioned fast logarithmic sliding surface:

[0092]

[0093] The above control rate includes the control input of the hand controller end effector and the estimated force of the interaction behavior, where For the control input at the end of the hand controller, is a diagonal matrix satisfying ρ i >0, i=1,2,3. Since the end effector of the hand controller does not have a corresponding force sensor, the actual operating force f cannot be obtained. h The present invention uses a neural network-based method to obtain the estimated force during the interaction process. f h and The mathematical expressions of the two are as follows:

[0094]

[0095] in represents the weight matrix of the neural network, is the basis function of the neural network, d is the number of hidden layers, is the input basis vector of the neural network, which specifically refers to the trajectory tracking error and velocity tracking error of the end effector of the hand controller in this invention. ∈ is a bounded approximation error satisfying ||∈||<∈ h ,∈ h is a positive number. W * The estimated value is used to generate the estimated value of the operating force behavior during the human-computer interaction process. In order to more accurately estimate the operating force during the human-computer interaction process and ensure the stability of the interactive system, the present invention designs the weight update rate of the neural network based on the control law (16)

[0096]

[0097] in and is a diagonal matrix of all positive numbers.

[0098] Operation force during physical human-computer interaction Expected trajectory and reference trajectory The relationship between the three can be expressed in Cartesian space using the admittance or impedance model:

[0099]

[0100] in represents the artificially set interaction inertia matrix, represents the artificially set interaction damping matrix, Represents the artificially set elastic matrix. When no interaction force is applied to the end of the hand controller, the hand controller follows the pre-set desired trajectory. In the present invention, the desired trajectory is a fixed position; during the human-computer interaction process, the hand controller tracks the reference trajectory under the influence of the interaction force.

[0101] The technical solution is further analyzed and explained below with reference to the accompanying drawings:

[0102] Figure 2 The effects of varying the state synchronization scale parameter designed in this invention on the system's dynamic and steady-state performance are demonstrated. It can be seen that under different state synchronization scale parameters, both the fully actuated and underactuated states approach zero, achieving the deployment goal of the tethered space robot. Furthermore, the underactuated state θ / c3 also approaches the fully actuated state ξ, indicating that the state synchronization scale parameter achieves synchronous control of the fully actuated and underactuated states. Furthermore, as the parameter increases, the system's dynamic performance improves, resulting in a faster response speed at the initial deployment moment. However, this is accompanied by a larger overshoot, which increases the system's adjustment time.

[0103] Figure 3 and Figure 4 The relevant state trajectories of the spatial tethered robot system and the hand controller end effector in the physical human-machine interaction mode are shown. Figure 3 This indicates that the state synchronization scale parameters generated under physical human-machine interaction do not necessarily guarantee the stability of the system. In this case, the spatial tethered robot needs to be deployed according to pre-optimized parameters, so that the operator can obtain the desired dynamic and steady-state performance by interacting with the hand controller end within an appropriate range. Figure 4 This shows that the fast logarithmic sliding mode control method based on neural network interaction force estimation designed in this paper can always ensure the stability of the interactive system during physical human-computer interaction and accurately convey the operator's control intention. That is, the actual trajectory x of the hand controller can accurately follow the reference trajectory x generated by the estimated force. r .

[0104] Example:

[0105] The implementation process of a method for deploying a human-machine collaborative space tethered robot based on parameter optimization in this embodiment is as follows:

[0106] Step 1: Initialize the relevant states of the physical human-machine interaction system and the space tethered robot system;

[0107] In a physical human-machine interaction system, a fast logarithmic sliding surface is initialized to ensure the stability of the human-machine interaction system at the initial moment. The neural network parameters used to estimate the interaction force between the operator and the hand controller are initialized. In a space tethered robot system, the state synchronization scale between the fully actuated state and the underactuated state is initialized, and a feasible control parameter set is calculated using the rolling horizon optimization method. and To ensure the stability of the closed-loop system.

[0108] Step 2: Conduct physical human-machine interaction to obtain the operator's control intention, so that the entire space tethered robot deployment process achieves the desired dynamic performance and steady-state performance;

[0109] The present invention uses the Phantom Omni Haptic hand controller to collect the operator's control intention during human-computer interaction. The 3-DOF three-degree-of-freedom hand controller dynamic model in the human-computer interaction process can be organized into the following form in Cartesian space using the Lagrangian method:

[0110]

[0111] in, is the acceleration of the hand controller end, is the speed of the hand controller end, is the position of the end of the hand control; is the inertia matrix of the hand controller end, is the centripetal force / Coriolis force matrix at the end of the hand controller, The gravity vector at the end of the hand controller; represents the control law generated by the controller to complete the trajectory tracking task; Represents the interaction force behavior between the human and the end of the hand controller.

[0112] To ensure that the hand controller can accurately and smoothly track the desired trajectory during human-computer interaction, this embodiment designs a fast logarithmic sliding mode control method. The designed sliding mode surface can be expressed as:

[0113]

[0114] Where e = xx rrepresents the trajectory tracking error of the hand controller during human-computer interaction. Parameters p and q are positive odd numbers and satisfy 1<p / q<2; β=diag(β1,β2,β3) and satisfy β i >0i,=1,, γ=diag(γ1,γ2,γ3) and satisfy γ i >0i,=1,, η=diag(η1, η2, η3) and satisfy η i >0, i=1,2,3, the subscript i in the β,γ,η parameters represents the three dimensions in Cartesian space; represents a vector of all ones; sgn(*), |*|, ln(*), and diag(*) represent the sign function, absolute value, logarithmic function, and diagonal matrix of the vector *, respectively.

[0115] To ensure the stability of the system during human-computer interaction, this embodiment designs the following control law based on the above-mentioned fast logarithmic sliding surface:

[0116]

[0117] The above control rate includes the control input of the hand controller end effector and the estimated force of the interaction behavior, where For the control input at the end of the hand controller, is a diagonal matrix satisfying ρ i >0, i=1,2,3. Since the end effector of the hand controller does not have a corresponding force sensor, the actual operating force f cannot be obtained. h This embodiment uses a neural network-based method to obtain the estimated force during the interaction process. and use the estimated force in actual control As the operating force f h Execute. h and The mathematical expressions of the two are as follows:

[0118]

[0119] in represents the weight matrix of the neural network, is the basis function of the neural network, d is the number of hidden layers, is the input basis vector of the neural network, which specifically refers to the trajectory tracking error and velocity tracking error of the end effector of the hand controller in this embodiment. ∈ is a bounded approximation error satisfying ||∈||<∈ h ,∈ h is a positive number. The estimated value of is used to generate the estimated value of the operating force behavior during the human-computer interaction process. In order to more accurately estimate the operating force during the human-computer interaction process and ensure the stability of the interactive system, this embodiment designs the weight update rate of the neural network based on the control law (16)

[0120]

[0121] in and is a diagonal matrix of all positive numbers. Integrating can get the corresponding neural network weights

[0122] Estimating forces during physical human-computer interaction Expected trajectory and reference trajectory The relationship between the three can be expressed in Cartesian space using the admittance or impedance model:

[0123]

[0124] in represents the artificially set interaction inertia matrix, represents the artificially set interaction damping matrix, Represents the artificially set elastic matrix. When no interaction force is applied to the end of the hand controller, the hand controller follows the pre-set desired trajectory. In this embodiment, the desired trajectory is a fixed position; during the human-computer interaction process, the hand controller tracks the reference trajectory under the influence of the interaction force.

[0125] The estimated force obtained by the above-mentioned neural network-based interaction force estimation method And the reference trajectory obtained by the impedance control method The motion trajectory of the hand controller end indirectly carries the operator's control intention, so the desired dynamic and steady-state performance during deployment can be obtained by controlling the end trajectory. The relationship between can be expressed as:

[0126]

[0127] in There is a lower bound for a feasible solution to the above optimization problem. There is an upper bound on the feasible solution for the above optimization problem. S(x)=κ||x|| is a scaling function with parameter κ>0.

[0128] Step 3: Based on the receding horizon optimization method, optimize the feasible control parameters to ensure the final consistent asymptotic stability of the closed-loop system;

[0129] According to the above physical human-computer interaction part, the desired state synchronization scale parameter c3 can be obtained, but this parameter cannot guarantee the stability of the system, so the parameter needs to be adjusted. and c3 are optimized. The parameter optimization criteria of c3 are transformed into solving [t0,t f ] time period, where t0 represents the initial time, t f Represents the end time, select T s As the time interval and using the rolling horizon optimization method, the entire optimization process is decomposed into N = (t f -t0) / T s There are optimization sub-problems, for the parameters Perform feasibility optimization and give more reasonable parameter criteria to ensure the stability condition (6) of the closed-loop system. i ,t i+1 ](i=0,1,...,N-1), parameters The optimization problem of and c3 can be expressed as follows:

[0130]

[0131] The optimization problem needs to satisfy the following constraints:

[0132]

[0133]

[0134]

[0135] u(t)∈[-T max ,-T min ], (31)

[0136]

[0137] in is the state vector, is a known nonlinear equation vector, is a positive definite diagonal matrix. Formula (28) is the loss function of the optimization problem, and its optimization goal is to make the state vector υ(t) as close to 0 as possible while satisfying the constraints. Constraint (29) is the control input of the system, u(t) is the state vector υ(t) and the parameter Equation; Formulas (30), (31) and (32) are the constraints of the sliding surface arrival stage, the actual control input constraints and the system asymptotic stability constraints respectively. In the whole parameter optimization process, each time the rolling horizon optimization method is used, the parameters c3 and If the optimized state synchronization scale parameter c3 obtained by the human-computer interaction part cannot ensure the stability of the closed-loop system by solving the optimization problem, the state synchronization scale parameter c3 that can ensure the stable operation of the system at the previous moment is selected. This parameter c3 comes from the parameter set obtained by the rolling horizon optimization method in step 1, thereby carrying out the deployment process of the space tethered robot.

[0138] Step 4: Execute the deployment task of the space tethered robot according to the optimized control parameters;

[0139] The dynamic model of the space tethered robot can be constructed into the following dimensionless form based on the energy of the system using the Lagrangian method:

[0140]

[0141]

[0142] Where u is the force between the tethers, θ is the opening and closing angle of the tethered robot, and ξ is the expression for the deployment degree of the space tethered robot. The relationship between this and the actual deployment length can be expressed as ξ = l / L-1, where L is the desired total deployment length. When ξ = -1, it means that the tethers are not deployed; when ξ = 0, it means that the tethers are deployed. The ultimate deployment goal of this system is to deploy the space tethered robot to the desired length toward the center of the Earth, which can be expressed as follows:

[0143]

[0144] The dynamic model of the tethered space robot demonstrates that, compared to the state deployment degree ξ, the tethered robot's opening and closing angles represent an underactuated system state, and its changes cannot be directly controlled by system inputs. This invention addresses this system characteristic by linking the fully actuated and underactuated states through a scale parameter, constructing a state synchronization term. A linear sliding surface is designed based on this state synchronization term, ensuring the asymptotic stability of the entire system.

[0145] The sliding surface designed according to the state synchronization term can be expressed as:

[0146]

[0147] Where c1 and c2 are sliding surface parameters, satisfying c1, c2 > 0; c3 is the state synchronization scale, the size of which affects the dynamic and steady-state performance of the system. This parameter is obtained through the physical human-computer interaction mentioned above. To ensure the asymptotic stability of the spatial tethered robot system, the following control law can be designed based on the designed sliding surface and dynamic model:

[0148]

[0149] where u eq is the pricing control power, u sw The switching term is designed to ensure the gradual stability of the system. According to the above rolling horizon optimization method, f1 and f2 can be expressed as:

[0150]

[0151]

[0152] Substituting the above control rate into the system's dynamic equation, we can obtain is the underactuated part of the dynamics, For the full drive part.

[0153] Step 5: If the deployment task is not completed, return to step 2 until the deployment of the space tethered robot is completed.

[0154] This embodiment provides a human-machine collaborative space tethered robot deployment system based on parameter optimization, including a human-machine interaction system and a space tethered robot system;

[0155] In the human-machine interaction system, the operator's control intention is obtained through the hand controller, and the entire spatial tethered robot deployment process is made to achieve the desired dynamic performance and steady-state performance through the control system;

[0156] In the space tethered robot system, an optimizer is used to optimize the control parameters determined by the human-machine interaction system, and then a control law is obtained through a sliding mode controller design. The control law and the dynamic model are combined through a control system to obtain the under-actuated part and the fully-actuated part in the dynamics, and the deployment of the space tethered robot is executed.

[0157] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and purpose of the present invention.

Claims

1. A method for deploying a human-machine collaborative space tethered robot based on parameter optimization, characterized in that The specific steps are as follows: Initialize the relevant states of the physical human-machine interaction system and the spatial tethered robot system, and determine a feasible control parameter set; in the physical human-machine interaction system, initialize the fast logarithmic sliding surface to ensure the stability of the human-machine interaction system at the initial moment, and initialize the neural network parameters used to estimate the interaction force between the operator and the hand controller; in the spatial tethered robot system, initialize the state synchronization scale between the fully driven state and the underdriven state, and calculate a feasible control parameter set based on the rolling horizon optimization method and To ensure the stability of the closed-loop system; Performing physical human-machine interaction to obtain the operator's control intention so that the entire deployment process of the space tethered robot achieves desired dynamic performance and steady-state performance, wherein control parameters during the deployment process are within the range of the control parameter set; Based on the rolling horizon optimization method, feasible control parameters are optimized to ensure the final consistent asymptotic stability of the closed-loop system; Execute the deployment mission of the space tethered robot according to the optimized control parameters; If the deployment task is not completed, the process returns to repeat the steps of physical human-machine interaction, optimizing feasible control parameters, and executing the deployment task of the space tethered robot until the deployment of the space tethered robot is completed.

2. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 1, characterized in that: The operator's control intention is acquired through the hand controller. During the human-machine interaction process, the hand controller dynamic model is expressed in Cartesian space using the Lagrangian method as follows: in, is the acceleration of the hand controller end, is the speed of the hand controller end, is the position of the end of the hand control; is the inertia matrix of the hand controller end, is the centripetal force / Coriolis force matrix at the end of the hand controller, The gravity vector at the end of the hand controller; represents the control law generated by the controller to complete the trajectory tracking task; Represents the interaction force behavior between the human and the end of the hand controller.

3. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 2, characterized in that: The method for achieving the desired dynamic performance and steady-state performance of the entire space tethered robot deployment process is: the estimated force in the interaction process is obtained by the interaction force estimation method based on the neural network Reference trajectory obtained by impedance control method The motion trajectory of the hand controller end indirectly carries the operator's control intention, and the desired dynamic performance and steady-state performance during deployment are obtained by controlling the end trajectory; Control parameter c3 during deployment and motion trajectory of the hand controller end The relationship between them is expressed as: in A feasible control parameter set The lower boundary of A feasible control parameter set The upper boundary of ; S(x) = κ||x|| is the scaling function with parameter κ>

0.

4. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 3, characterized in that: The specific process of optimizing feasible control parameters is as follows: The control parameters The optimization criteria of c3 are transformed into solving [t0,t f ] time period, where t0 represents the initial time, t f Represents the end time; select T s As the time interval, the whole optimization process is decomposed into N=(t f -t0) / T s optimization subproblems, where N represents [t0,t f ]The number of time intervals in the time period, that is, the number of optimization subproblems; in the time period [t i ,t i+1 ], i=0,1,...,N-1, control parameters The optimization problem of and c3 is expressed as: The optimization problem needs to satisfy the following constraints: u(t)∈[-T max ,-T min ],(11) in, is the state vector, is a known nonlinear equation vector, is a positive definite diagonal matrix; u eq is pricing control power, u sw For the designed switching items; Formula (7) is the loss function of the optimization problem. Its optimization goal is to make the state vector υ(t) as close to 0 as possible while satisfying the constraints, and the loss function It can be calculated by the integral of the constraints; the constraints (9) are the control inputs of the system, which are the state vector υ(t) and the parameters Equation; Formulas (10), (11) and (12) are the constraints of the sliding surface arrival stage, the actual control input constraints and the system asymptotic stability constraints respectively; During the whole parameter optimization process, each time the rolling horizon optimization method is used, the parameters c3 and near-optimal value.

5. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 4, characterized in that: If the optimized state synchronization scale parameter c3 obtained by the human-computer interaction part cannot ensure the stability of the closed-loop system by solving the optimization problem, then the state synchronization scale parameter c3 that can ensure the stable operation of the system at the previous moment is selected. The parameter c3 is derived from the feasible control parameter set, thereby carrying out the deployment process of the space tethered robot.

6. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 4, characterized in that: To ensure the eventual consistent asymptotic stability of the entire system, the following conditions must always be met: Where ξ is the expression of the deployment degree of the space tethered robot.

7. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 6, characterized in that: The process of executing the deployment task of the space tethered robot is as follows: Establish a dynamic model of a space tethered robot; Determining a final deployment goal as deploying the space tethered robot to a desired length toward the center of the Earth; According to the state synchronization scale between the fully actuated state and the under-actuated state, a state synchronization term is constructed and a sliding surface is designed; Designing a control law based on the sliding surface and the dynamic model; The control law is then introduced into the dynamic model of the system to obtain the under-driven part and the fully driven part in the dynamics.

8. The method for deploying a human-machine collaborative space tethered robot based on parameter optimization according to claim 7, characterized in that: The under-driven portion is Where θ is the opening and closing angle of the tethered robot, The full drive part is Where u is the force between the ropes, 9. A human-machine collaborative space tethered robot deployment system based on parameter optimization, used to implement the human-machine collaborative space tethered robot deployment method based on parameter optimization according to any one of claims 1 to 8; characterized in that: Including human-computer interaction system and space tethered robot system; In the human-machine interaction system, the operator's control intention is obtained through the hand controller, and the entire spatial tethered robot deployment process is made to achieve the desired dynamic performance and steady-state performance through the control system; In the space tethered robot system, an optimizer is used to optimize the control parameters determined by the human-machine interaction system, and then a control law is obtained through a sliding mode controller design. The control law and the dynamics model are combined through a control system to obtain the under-actuated part and the fully-actuated part in the dynamics, and the deployment of the space tethered robot is executed.

Images