混沌免疫遗传算法在电力系统故障诊断中应用

Chaos immune genetic algorithm in power system fault diagnosis

针对电力系统故障诊断问题,提出了一种新的免疫遗传算法——混沌免疫遗传算法。该算法将免疫算法、混沌与遗传算法相结合,利用混沌运动的遍历性、随机性产生初始种群,加快搜索的速度;利用免疫原理的浓度计算及调整加入新的混沌序列补充种群,增加种群的多样性,避免陷入局部最优;交叉变异结束后在最优解附近再用混沌进行局部寻优提高解的精度。它能从保护和断路器的拒动和误动中,快速找到故障点。实验结果表明,该算法能快速寻找到最优解,保证系统的实时性和准确性,较好地实现了对电力系统的故障诊断。

Combining immune algorithm,chaos algorithm and genetic algorithm together,a chaos immune genetic algorithm is presented for power system fault diagnosis. Using the over-spread character and randomicity of chaos,a cluster is initiated to accelerate the search,which is then recruited with new chaos sequence according to the concentration calculation and regulation of immune principle,enhancing the cluster diversity to avoid local convergence. Chaos is used again after crossover and mutation in local optimization round the optimal solution to improve its precision. It can find the faulty section fleetly from the misoperations of protection and breaker. Test results show that,the approach can find the optimal solution quickly,guaranteeing the real-time performance and accuracy of fault diagnosis.