Research On Force Feedback Teleoperation Control Strategy Based On Model Prediction

With the development of artificial intelligence technology and the widespread development of space scientific research activities,the complexity and precision of space science experiment tasks are increasing.Faced with an unstructured environment and complex and diverse scientific research tasks,it is unrealistic to rely on robots to autonomously perceive the state of the environment and make decisions to plan and execute tasks.Therefore,the force feedback remote operation control technology based on the predictive model has become an important technical means for realizing scientific tasks of complex and delicate operations.This paper focuses on the identification of environmental modeling parameters,online updating of model parameters,and compliance control technology from the remote end in space teleoperation.The main work is as follows:First,the technology of environment model modeling and parameter identification is studied,and a teleoperation system based on the prediction model is built.The experimental environment and system architecture are introduced first,and the functions and tasks to be completed of the teleoperation system are introduced from multiple modules.Then introduced a variety of environmental modeling and parameter identification algorithms,and conducted in-depth analysis and experimental research,completed rotate and contact experimental tasks,provided practical guidance for the application of environmental model modeling and parameter identification technology,and provided online parameter updates and slave compliance control with the foundation.Secondly,the algorithm of environment model parameter update in teleoperation system is studied.Firstly,the problems caused by the direct parameter update in the master environment model were analyzed and verified,and the influence of the feedback force of the parameter update output of the virtual environment model on the system stability was proved.Then perform theoretical derivation and experimental analysis for multiple environmental model update algorithms.Based on the analysis of progressive update,this paper proposes an adaptive variable rate progressive update algorithm,which adaptively adjusts the update rate and updates according to the gap between the current parameter value and the target parameter value.The algorithm can not only ensure the continuity of update parameters,but also take into account the fast response characteristics.The experimental results show that the algorithm improves the update efficiency and the update performance under the premise of ensuring a certain range of mutation values.Combined with the idea of the update algorithm based on rendering power,the operator experience is better and it improves the teleoperation system's stability and robustness.Thirdly,research on the compliance control of the teleoperation system based on the prediction model from the end is conducted.Introduce an impedance control strategy for the scenario where the slave and the environment interact flexibly.Add an impedance parameter input that represents the characteristics of the slave's force position to the slave's control loop to enhance the compliant interaction between the slave and the environment while ensuring position tracking Impedance characteristics reflect the soft motion characteristics of the slave end.The parameter adaptive adjustment impedance controller in the slave remote operation system is introduced,and then an impedance control algorithm based on radial basis neural network is proposed.It is robust to tolerate errors and fit motion characteristics.Finally,a teleoperation system experimental platform based on prediction models and virtual scenes is set up,and verification experiments and analysis of the research content of this paper are conducted.The experimental results further verify the correctness and effective application of the theoretical method.