Vibration Control using Neural Network with Adaptively Rearranging Structure

DOI IR IR (HANDLE) Web Site Web Site 17 References Open Access
  • KOMATSUZAKI Toshihiko
    Department of Human and Mechanical Systems Engineering, Kanazawa University
  • SATO Hidenori
    Department of Human and Mechanical Systems Engineering, Kanazawa University
  • IWATA Yoshio
    Department of Human and Mechanical Systems Engineering, Kanazawa University
  • MORISHITA Shin
    横浜国立大学大学院環境情報研究院

Bibliographic Information

Other Title
  • 結合構造可変ニューラルネットワークによる振動制御
  • ケツゴウ コウゾウ カヘン ニューラル ネットワーク ニ ヨル シンドウ セイギョ

Search this article

Description

The neural network has been applied to many engineering problems such as pattern recognition, optimization, system identification and control because of its nonlinear mapping capability. Although the network has a great ability to acquire solution for given problems through learning, several problems are known such as the convergence to a local minimum solution and the empirical tuning of network parameter to obtain desired accuracy. Such an instability may be overcome by finding suitable network structure. The optimized network structure with minimum connection can be obtained by employing optimization techniques such as Genetic Algorithms. However, it is hardly applicable to time-varying systems in view of computation efficiency. In the present paper, we deal with the adaptive neural network controller which rearranges its connection structure suitably according to the given problems. The network is applied to vibration control problem of single degree of freedom nonlinear system, and the performance of the controller is discussed by numerical investigation.

Journal

References(17)*help

See more

Keywords

Details 詳細情報について

Report a problem

Back to top