Abstract: Active vibration control is a new subject developed in the last twenty years. It mainly studies the theory, method and measure of active vibration control of structures. After introducing the linear quadratic optimal control algorithm, the numerical simulation of the linear quadratic optimal control for a piezoelectric flexible cantilever beam is carried out. The model of the flexible cantilever beam is established firstly, including the differential equation of motion, the moment equation of the actuator and the output equation of the sensor. Then the optimal feedback gain matrix of the system is obtained by using the linear quadratic optimal control. In this paper, genetic algorithm is used to optimize the above parameters of BP neural network, and the improved BP neural network is applied to the study of nonlinear model of dynamic coal blending. A method based on piezoelectric self-sensing method is proposed as a test method. A piezoelectric wafer of piezoelectric bimorph is used as the sensing element, analyzing the relationship between the acceleration parameters of the piezoelectric bimorph and the induced charge generated by the sensing element, and a test circuit device for driving force is designed. The method uses genetic algorithm to calculate the control force online and uses the neural network to simulate the dynamic characteristics of the plate, thus replacing the cantilever plate for dynamic analysis. The system fully utilizes the characteristics of genetic algorithms and neural networks and is a new type of vibration control system with promising future.
Keywords: Piezoelectric intelligent structure, neural network, driving force
