Master-slave Teleoperation System With Force Feedback

The use of teleoperated robots instead of humans to explore and work in environments that are difficult for humans to reach has become one of the important applications of robots.In order to help the operator to successfully complete the motion planning and decision-making of the robot,it is necessary to introduce the presence technology in the remote operating system to help the operator perceive as much information as possible on the external environment of the robot.Force feedback technology is one of the core contents of the telepresence technology,and it is also the main content of this research.Because there is often a huge physical distance between the slave and the master in the actual remote operating system,there is a large delay in the communication between the master and the slave,which affects the operational performance and stability of the force feedback remote operating system.The controller design problem of force-sensing feedback teleoperation system under large time delay is studied.Firstly,a master-slave force feedback teleoperation system model with one degree of freedom is established.We pointed out that its operational performance mainly includes transparency and tracking,and then we gave the evaluation indicators.Based on the absolute stability theory and the passive stability theory,the theoretical analysis and controller design of the force feedback teleoperation system are carried out respectively,and the simulation experiments and comparisons are carried out under different time delay conditions.In order to make the operator feel the contact force in real time,the model prediction method is introduced to predict the feedback force.The prediction model is established by the offline identification method based on neural network and the online identification method based on least squares method.The theoretical analysis and simulation experiments prove the effect of the prediction model on system performance improvement.In order to verify the effect of the research results in the actual system,two linear motors were used to build a master-slave teleoperation system experimental platform with one degree of freedom and force feedback.The experimental platform is based on the EtherCAT open architecture,and the real-time control of the system is realized by the TwinCAT3 software on the industrial computer.The collision protection function was designed and the motor step response performance was tested.The passive controller is designed and tested.The experimental results show that the passive controller can guarantee the stability of the system under any time delay,and the force feedback and speed tracking error are within the design requirements.The neural network is used to identify and predict the force feedback offline.The offline model can achieve good prediction results with a maximum error of 0.76 N and a root mean square error of 0.17 N.We attempt to establish a system model using the online identification method based on the least squares method,but due to the large static friction of the linear motor and the environment,the linear model can not deal with the nonlinear friction link,which leads to the unpredictable model prediction and needs further improvement.