| With the development of space exploration technology,the demand and variety of space operation tasks are becoming more and more diverse.However,because of the space environment of unstructured and task of high complexity,and the limitation of technical conditions,space robot autonomous operation mission vision is still difficult to achieve in the short term.Therefore,the field control method of force feedback teleoperation is an important technical approach to complete the task of space robot.Model-mediated teleoperation method due to the end of the preset exclude time delay,a simulation model with its good transparency,traceability,robustness advantages,has become the focus of research.In this paper,an in-depth study on environment modeling,parameter identification and force tracking control of model-mediated force feedback teleoperation are carried out.The main content structure of this paper is as follows:First of all,we research on the current teleoperation environmental dynamics model of linear and nonlinear characteristics,application scenarios,and analyzes the problems,according to different environment stiffness is proposed based on threshold switching hybrid model of the environment.Respectively based on this,the linear model,nonlinear model experiment to explore all kinds of parameter identification algorithm,with the convergence of the identification result and in the face of environmental change responsiveness as an index,and cross validation method is validated with the inspection various environmental optimal online parameter identification algorithm of the model.The mixed modeling method is compared with the mean square error of the single modeling method,and the performance of the mixed environment modeling and parameter identification algorithm is verified.Secondly,in view of the teleoperation under the large time delay,nonlinear dynamics environment fluctuation parameter and position information does not match the force error caused by the problem,we set up nonlinear modeling environment based on optimization of BP neural network based on genetic algorithm.The BP neural network structure is established based on the mapping relationship between the motion state information,contact force information and environmental dynamics parameter information.By setting up single degree of freedom force sensing telepresence teleoperation experiment platform and two typical contact environment,the experimental results show that the neural network model constructed in this paper can overcome the delay effect and predict more accurate contact information and environmental dynamics parameter information,confirm nonlinear modeling environment based on optimization of BP neural network based on genetic algorithm proposed in this paper has great performance.Then,the control technology of the remote operating system is studied.Firstly,the reasons of force error in position based impedance control strategy are analyzed.Then through simulating the process of arms force control,we put forward the adaptive tracking algorithm based on variable impedance parameters.On the basis of this,explicit force control based on impedance inner control is proposed.Finally,we set up a single degree of freedom force feedback teleoperation experiment platform and set the parameters of impedance control as the control group,the force tracking performance and position tracking performance of flexible and rigid materials are investigated in the case of 2s communication time delay.Compared to the traditional impedance controller,the proposed force tracking control algorithm will expect force and the desired position as input controller parameters at the same time,make full use of the relationship between contact force and position information in model-mediated teleoperation.Force tracking control algorithm proposed in this paper not only ensure transparency and traceability,and adaptive control without model switching will have strong robustness.Finally,we set up the model-mediated force feedback teleoperation system integrated experimental platform,a comprehensive experimental analysis of environmental modeling,parameter identification and force tracking control technology proposed in this paper are carried out.The experimental results confirm the theoretical feasibility and application value of the proposed algorithm. |
