Study On The Application Of Semi-active Neural Network PID Control Of The MR Dampers In Cable Vibration

Cable is the main part of the cable-stayed bridge with the characteristics of small unit mass, low damping and big flexible. It is easy to occur large amplitude vibration under the effect of various external loads, which dose harm to the using life and safety of the bridge. The structural vibration control of the cable, especially the semi-active control can combine with the advantages of active control and passive control fully, control the vibration of the cable effectively and reduce the bridge damage brought by cable vibration. Neural network PID, which combines with the advantages of conventional PID and neural network, is a kind of intelligent PID technology. It achieves the combine of Intelligent algorithm and conventional algorithms and not only plays the self-learning ability, adaptive ability and the ability to solve nonlinear system of neural network but also plays the advantages of simple control algorithm and good control effect. It exploits the new idea of semi-active control algorithm of the cable vibration.This paper introduces neural network PID to cable - damper system, designs the neural network PID controller based on S-function, verifies and analyzes the validity of the algorithm.This paper analyses the cable-damper control system firstly, establishes and solves its vibration equation, then chooses appropriate random signals as the input of the system, and simulinks the vibratory response of the cable which is without control; Secondly, it establishes the simulink model of the MR damper according to its mechanical model, analyzes the voltage and the force of the damper, and at the same time, designs the neural network PID controller, encapsulation it, and make it become module of general simulation; At last, adds the damper's model and the control algorithm model on the cable to build the total simulink diagram of the system, analyzes the effect of the controller and compares the vibration response, and the results show that neural network PID can give full play to the ability of neural network in dealing with nonlinear problem and the ideal effect of PID control, decrease the vibration response rapidly and get obvious effect. The results prove that the neural network PID is an effective kind of semi-active control algorithm and can solve nonlinear system control problem well.