Radar Fault Diagnosis based on Chaos Genetic Reduction Algorithm

Based on the study of the rough set theory and the chaos genetic algorithm, a chaos genetic reduction algorithm inspired by knowledge dependency, is proposed and applied to diagnosis of the radar. In this algorithm, the property core of initial population generated randomly by binary is restricted, the degree of dependency of the condition attribute to the decision attribute is introduced to in the fitness function, and the crossover and mutation probability are redesign to produce the amend and verify operator for the new generation. This algorithm for radar fault diagnosis makes use of a simple diagnosis rules but can show the relationship between symptom and cause of the failure, which avoiding the poor accuracy and low efficiency by the traditional expert fault diagnosis system based on fault tree.