基于粗糙集和自适应遗传算法的电力变压器故障诊断

采用自适应遗传算法和粗糙集相结合的方法对电力变压器的决策信息表进行属性约简,并进一步采用改进的值约简算法以获取最小决策表,从最终的约简决策表中提取出故障诊断规则。通过引入初始种群预处理、最优L个个体保存法、双亲单子法等操作加速算法的收敛速度,并结合两个实例证明了规则的正确可行性。整个算法简单、快速,可有效地应用于电力变压器的故障诊断中。

Application of Rough Set and Adaptive Genetic Algorithm to Transformer Fault Diagnosis

A method of genetic algorithm combined with rough set is proposed in this paper for attribute reduction of the decision table for power transformers’ fault diagnosis,and a modified method of value reduction is adopted to obtain the minimal decision table.Besides,the fault diagnosis rules are extracted from the final reduced decision table.Pretreatment of initial population,the method of preserving optimal L individuals,and the method of parents-with-only-child are introduced to accelerate the convergence rate of the algorithm.Lastly the validity and the feasibility of the derivative rules are illustrated by two examples.The algorithm is simple and fast and can be applied into the fault diagnosis of power transformers efficiently.