The Principle of Maximum Entropy-Based Two-Phase Optimization of Fuzzy Controller by Evolutionary Programming

In this paper, a two-phase evolutionary optimization scheme is proposed for obtaining optimal structure of fuzzy control rules and their associated weights, using evolutionary programming (EP) and the principle of maximum entropy (PME) based on the previous research [1]. 1 Two-Phase Evolutionary Optimization A fuzzy logic controller (FLC) with weighted rules, which is equivalent to a conventional fuzzy controller with a weighting factor of each rule, is adopted [2] and a two-phase evolutionary optimization scheme is applied to the FLCs. In the first phase, initial population for rule structures are given as a stable fuzzy rule. Rule structures and scale factors of the error, change of error and input to the FLC are optimized by EP. The variation of the rule structures is done by the adjacent mutation operator and the scale factors are mutated by the Gaussian random variables. The objective function is constituted by the sum of error, sum of input and the number of used rules. Scale fa...

• Type: Conference
• Year: 2003
• Where: GECCO
• Authors: Chi-Ho Lee, Ming Yuchi, Hyun Myung, Jong-Hwan Kim
• Keywords: Fuzzy Rules, GECCO 2003, Rule Structures, Two-phase Evolutionary Optimization