Controller Design And Performance Analysis For Networked Teleoperation System

Recently, teleoperation system, which is integrated with the capacities of control theory, computer technology and communications, has attracted a considerable amount of interest around the world due to its theory and application. It has widespread applications, such as in nuclear accident rescue, undersea operation, space exploration and robotic telesurgery. Attracted by its tremendous potential, many researchers from the academic and industry communities have already devoted huge efforts to the research of teleoperation system. In order to facilitate the operation of a specific object(such as dangerous goods, etc), a communication network is required to ensure the real-time and efficient transmission of data, and then a networked teleoperation system is generated. However, the traditional controller is difficult to apply to the networked teleoperation system due to the delay and quantization in the network. The dissertation analyzes the stability performance of networked teleoperation system and designs controllers. The main research works are summarized as follows:(1) For stable region problem of constant delayed teleoperation system, an integral-order PDcontroller and a fractional-order PD? controller are proposed. For closed-loop error system, with CTCR method, sufficient and necessary stability conditions are established for proportional derivative controller and position error based controller from the frequency domain point of view. Moreover, by defining the state variables, the closed-loop state equation of teleoperation system can be obtained. Based on the equation, an algorithm is presented to detect the exactly stable region for the teleoperator with the fractional-order controller.(2) For actuator saturated and time-varying delayed teleoperation system, the saturated controller can be represented a convex combination of the actual feedback and an auxiliary counterpart in the space surrounded by two pairs of intersected parallel surfaces. The novel Lyapunov-Krasovskii functional, which is based on position error and velocities, is constructed. The position error based Newton-Leibnitz equation is proposed to relax the restriction. A new delay-dependent stability criteria is proposed for actuator saturated and interval delayed teleoperator. Furthermore, the estimate of domain of attraction is explored, and an optimization problem is porposed to maximize the estimate of DOA. It provides theoretic guide for the application of such system.(3) For linearization problem of nonlinear networked teleoperator, a T-S fuzzy model for-DOF teleoperation system is proposed. The gravity compensation term is involved and the force feedback is added in control law. Tracking performance is analyzed when the model parameters are the same and different, respectively. By linearizing the nonlinear matrix inequalities, a method of designing controller gains is proposed. n(4) For bandwidth constrained problem, stability performance is analyzed for teleoperation system with quantized information feedback. Both proportional position error plus damping injection controller and slave-torque feedback plus damping injection controller are considered. The relationship among the parameters of quantizer, controller and time delay is established to guarantee stability of bilateral teleoperator.(5) For unobservable velocity signal problem, a novel speed observer is constructed via Immersion and Invariance for networked teleoperation system. Define a manifold and a off-the-manifold coordinate, which determines the distance of the state to the manifold, and then demonstrate the manifold is invariant and attractive. Define a scaled state and observe errors, construct subdynamics and demonstrate the scaled state and observe errors converge to zero. The control law is designed based on the estimated velocity signal. The synchronization between master site and slave site can be guaranteed.