Research On The Key Technologies Of Hardware-in-the-loop Simulation System For Space Manipulator Grasping Process

Hardware-in-the-loop(HIL)simulation of space manipulator grasping process is an important means to simulate the grasping process of space manipulator on the ground,and to study the grasping characteristics and initial conditions of different forms of manipulator.The HIL simulation system of space manipulator grasping process consists of docking dynamics model and HIL simulation system.Studying the characteristics of HIL simulation system and improving the simulation ability of the system are the key to simulate the grasping process accurately.However,the control form based on the position inner loop control and force outer loop control makes HIL simulation system a complex system with complex model,strong coupling,many uncertainties and unknowns;meanwhile,the grasping process of space manipulator is composed of drag stage,attitude correction stage and buffering stage,which have different characteristics in different stages;the above characteristics make the research of key technologies such as modeling,analysis,control and compensation of HIL simulation system difficult.In this regard,the following aspects are studied in this paper:The HIL simulation system for space manipulator grasping process is designed.Aiming at the research purpose and requirement,using the lumped parameter method,the docking dynamics model of space manipulator is established by equivalent the manipulator to a six-degree-of-freedom mass-spring-damping system.The HIL simulation system model of space manipulator grasping process is established by synthesizing docking dynamics model,parallel robot model and contact model.The characteristics of HIL simulation system are analyzed from three important aspects: system parameters configuration relation,system stability conditions and docking dynamics frequency simulation ability;the key conclusions that the system meets stability and reproduction accuracy are obtained.In order to improve the stability and simulation accuracy of HIL simulation system,parallel robot is required to have fast response speed or large frequency width in the grasping process.Because of the complex force in the grasping process and the time-varying parameters of parallel robot,the conventional model-based control strategy of parallel robots is difficult to design.To solve this problem,a model-free fuzzy incremental control method is proposed,which has universal approximation characteristics.It can organize the speed and position information of the legs of parallel robots by itself,and dynamically adjust the output of the controller with the designed membership function and control rules.Compared with the traditional control method,the proposed control method has better effect on reducing the delay of parallel robot,and the simulation capability of HIL simulation system is improved.Two force distortion problems in HIL simulation system seriously affect the system characteristics.1)The measurement delay of force sensor: a delay compensation method based on Smith prediction is proposed.The method integrates the first-order phase compensation model and Smith prediction model,which effectively compensates the phase and amplitude deviation of the measured force in the docking dynamics frequency range.2)Contact force deviation caused by the dynamic response delay of parallel robot: the dynamic response delay of parallel robot and the limitation of frequency bandwidth and docking dynamics frequency of parallel robot often make the system unstable or produce precision loss.In order to ensure the effective simulation of the drag process,a method of force and moment compensation based on three-degree-of-freedom stiffness-damping online identification is proposed.Considering the time-varying characteristics of multidegree-of-freedom contact parameters,the stiffness and damping of three mutually vertical directions are identified online.The six-degree-of-freedom force and moment compensation model can effectively compensate for the deviation of contact force and moment caused by the delay of dynamic response of parallel robot,which make the forces and torques used in the docking dynamics calculation not affected by the delay,and the accurate simulation of the drag stage is still guaranteed when the delay is large or the frequency width is small.In the attitude correction and buffering stage of grasping,when the contact stiffness is very large,the HIL simulation system with the mathematical model of the manipulator will be unstable because of the high contact stiffness of the docking mechanisms.Aiming at the high stiffness contact,a method of simulating the space manipulator grasping by equivalent the manipulator to a six-dimensional spring mechanism is proposed.By introducing the stiffness of flexible manipulator into physical contact link,the series stiffness of the system is reduced,the stability of the system is improved,and the simulation of grasping process is realized.The model of HIL simulation system is established and the key conclusions satisfying the stability and reproduction accuracy are analyzed.Aiming at the contact characteristics of multi-stiffness,multi-damp and single-mass in grasping process drag stage,a discrete domain force compensation method based on on-line identification of process parameters is proposed to compensate the force deviation caused by the delay of parallel robot,which improves the simulation accuracy in the drag stage.When the stiffness and contact stiffness of the manipulator are relatively large,the HIL simulation system with the six-dimensional spring object of the manipulator will be unstable due to the influence of high stiffness and low damp in the buffering stage of grasping.In this paper,an active compliance force control method is proposed to simulate the buffering stage of grasping process.The time-varying characteristics of parameters in contact environment and the second-order oscillation element in contact environment will seriously affect the stability and frequency bandwidth of force control system.Meanwhile,the noise in the system and the uncertainty of the parallel robot model also affect the force control characteristics.A robust controller based on ? synthesis theory is proposed and applied to the force control system.Compared with the traditional control method,the proposed method effectively improves the robustness,bandwidth and anti-interference characteristics of the force control system,and ensures the effective simulation in the buffering stage.