Direct comparison of neural network, fuzzy logic and model predictive variable structure vortex flow controllers Conference Paper uri icon

abstract

  • 1999 by the American Institute of Aeronautics and Astronautics, Inc. All Rights reserved. A neural network based controller is designed for bang-bang type vortex flow control nozzles on a generic X-29A. A full state feedback controller is used for the continuous control effectors. The neural network designed is a three layer network with symmetric hidden layers, which optimizes a given quadratic performance index. This performance index allows the designer to specify appropriate weights for states and control effectors to satisfy given specifications. This paper also directly compares the Neural Network Controller to previously designed Model Predictive Variable Structure, and Fuzzy Logic Controllers for the same benchmark problem. Evaluation criteria consist of closed-loop system performance; activity level of the VFC nozzles, ease of controller synthesis; and time required to synthesize controller.

name of conference

  • Guidance, Navigation, and Control Conference and Exhibit

published proceedings

  • AIAA GUIDANCE, NAVIGATION, AND CONTROL CONFERENCE, VOLS 1-3

author list (cited authors)

  • Joshi, P., & Valasek, J.

citation count

  • 3

complete list of authors

  • Joshi, P||Valasek, J

publication date

  • August 1999