SOFT COMPUTING TECHNOLOGIES IN DESIGN OF FUZZY CONTROLLER FOR ACTIVE VIBRATION ISOLATION SYSTEMS

In the paper, we develop a two-phase design approach for a fuzzy logic controller (FLC) of an active vibration isolation system, which adopts two components of soft computing techniques, namely fuzzy logic control theory and neural networks. First, we design a fuzzy logic controller for an active vibration isolation system using fuzzy logic control theory. We conduct numerical simulations of an active vibration isolation system with three kinds of exciting loads. The results indicate that the performance of the fuzzy logic controller is very good, but its control rule surface does not measure up in practical terms. Secondly in the neural network, trained by Quick-prop with Newton's method, is used to model an approximation of this fuzzy logic controller, termed the neural network fuzzy controller (NNFC). The control performance indicates that the NNFC has the same performance as the FLC designed in the first phase and its control rule surface is improved and more suitable for actual use. The results indicate that the proposed design approach provides a robust, controllable, practical, and low-cost solution for a FLC of an active vibration isolation system.