Self-Optimum Fuzzy Controller Based on Minesweeping Strategy

Hu Jing-Song,Wang Jia-bing,Yang Chuang-Xin

DOI: https://doi.org/10.1109/cis.workshops.2007.85

2007-01-01

Abstract:The partly continuous algorithm (PCA) (Hu Jing-song et al., 2000) is a simple and fine quality fuzzy control method, which can operate at high speed by a normal digital chip computer with few memories. A fast global strategy, called minesweeping strategy, is presented to optimize the fuzzy controller based on the partly continuous algorithm. The minesweeping strategy lets the current search "jump out", from the current local minimum by exploiting a new area that is far away from local minima obtained erenow. Simulating annealing and tabu search lets the current search "climb out", step by step, hardly and wonderingly. Therefore the strategy to solve local minimum problem is more successful and faster than other methods. The self-optimum fuzzy controller based on minesweeping strategy realizes real-time optimization and fine control quality to a nonlinear plant, of which mathematic model is not known.