Robust Control For High Rise Elevators Based On T-S Fuzzy Model

The elevator vertical motion and control design problem are discussed for high-rise/high-speed elevators in this thesis. The goal of the control is to obtain releveling and tracking performance, and realize multiobjective control and nonlinear control. The performance requirement entails metrics such as flight time, landing accuracy, position overshoot and riding quality, etc. The performance requirements for modern elevators are very stringent. Furthermore these requirements must be achieved in the presence of considerable model uncertainty and unmodeled dynamics. For high rise/high speed elevators, the "effective length" (distance between the car and drive sheave) of hoist rope varies during elevator transition, so in fact the elevator vertical motion is nonlinear. Also the stiffness and damping coefficients of hoist rope and compensation rope are varies during elevator transition, ie. there exists the affect of uncertainties for system. The robustness requirement demands the control system must account for both parametric uncertainty and unmodeled dynamics associated with, for instance, the rope dynamics. When the elevator vertical motion is described, traditional PID control method hasn't meet the control purpose and obtain good performance, at the same time the parameter variation is considered, so the T-S fuzzy model is built for elevator system. Rope dynamic performance and system parameter uncertainties (the rope stiffness and damping) are fully considered when the model is built.Elevator system is a complicated nonlinear system. An uncertain parameter generalized fuzzy model is built in the thesis. Then the local state feedback controller is designed based on parallel distributed compensation theory, and some sufficient conditions meeting certain H_∞performance are obtained to make the closed loop system stabilized. The sufficient conditions are showed by linear matrix inequality(LMI), and the asymptotic stability is realized just once and the desired performance performance is satisfied for the system.The control method proposed in this paper can satisfy elevator performance requirements with the height of 500 meters(about 135 stories), the maximum run speed of 17m/s. The nonlinearity and uncertainty are coupled into elevator system model, and fully reflecting elevator performance through the model. The 1-floor, 67-floor and 135-floor are selected as designed model, and simulating the elevator running state. The simpleness and effectiveness are showed by the simulations under the performance requirements of the controller.