Re search on Dynamic Neural Network Identification for Piezoelectric Smart Structures
Neural network identification of piezoelectric vibration system plays an important role in the application of neural network control method for piezoelectric smart structures,so how to improve the effect of identification is a critical problem.In the paper.a kind of BP neural network with delay which has a spring.parallel identification structure is employed to identify the nonlinear dynamic system of piezoelectric smart cantilever beam,which surface is bonded several piezoelectric patches.Considering the effect of piezoelectric patches to the mass and stifiness of beam and the difficulty of extracting the experimental data,the author analyses the modais and the transient response of beam using ANSYS (firIite element software)and obtains the dynamics response data of beam.A BP neural network with delay time is built and trained offiine in basis of the data using MATLAB NeuraI Network Toolbox.Then a case is analyzed and the results show the neural network has the quick convergence,good generalization ability,ideal identification and can be used into the vibration control of piezoelectric sinart structures using neural network control algorithm..
delay time neural networks piezoelectric smart structures system identification
Jingjun Zhang Ercheng Wang Ruizhen Gao Weize Yuan
国际会议
The International Conference Information Computing and Automation(2007国际信息计算与自动化会议)
成都
英文
359-362
2007-12-19（万方平台首次上网日期，不代表论文的发表时间）