The Research On Active Vibration Control Of Piezoelectric Structures Based On Neural Network

With the rapid development of aerospace technology,flexibility is an importantdevelopment trend of various types of aerospace structure,flexible components are more and more attention,its elastic deformation and vibration problem has become the focus of the study.In recent years,piezoelectric intelligent materials because of its wide frequency response zone,easy to process and easy to establish a computer automatic control system and so on,has become hot material in control field.With the advancement of science and technology,In addition to the traditional method such as pole assignment method,independent model space control algorithm and linear quadratic optimal control algorithm,neural network has gradually become the research hotspots in control field.Therefore,this paper takes the neural network control algorithm belong to the intelligent control algorithm to control the piezoelectric cantilever beam.After introducing three kinds of traditional control algorithms,the numerical simulation of a linear quadratic optimal control of a piezoelectric flexible cantilever beam is carried out.The mathematical model of the piezoelectric flexible cantilever beam is established,including the derivative derivative Equation,the actuator moment equation and the output equation of the sensor,and then use the linear quadratic optimal control to obtain the optimal feedback gain matrix of the system.Then the neural network control algorithm is introduced in detail,and the numerical simulation of the neural network control of a piezoelectric intelligent flexible cantilever is carried out,and the weights and thresholds of BP neural network training are obtained.Finally,the active control experiment of piezoelectric intelligent cantilever beam based on BP neural network is designed and completed.The experiment is carried out by using NI's PXI automatic measurement platform and NI Lab VIEW real-time system,with the weights and thresholds from the neural Network Numerical Simulation of Piezoelectric Flexible Beam.The experimental results show that the vibration of the piezoelectric cantilever beam after the control of the neural network is effectively suppressed.