Optimization of fuzzy rules by Muilti-objective genetic algorithm in avionic fault diagnosis system

The fuzzy rule sets, which have been widely used in avionic fault diagnosis system, have considerable redundancy that leads to time-consuming faults location process. In this paper, to reduce the redundant rules, a multiple objective genetic algorithm, MOGAII, is used to optimize a fuzzy rule set. The optimization problem with two objectives, the maximization diagnostic capability of the system and the minimization number of rules, is formulated. The simulation results show that MOGAII can substantially improve the efficiency of avionics fault diagnosis system comparing with the plain aggregation algorithm, a conventional optimization method of fuzzy rule sets.