基于遗传算法和模拟退火算法的电力系统的故障诊断

Fault Section Estimation in Power Systems Using Genetic Algorithm and Simulated Annealing

根据元件故障与保护动作和断路器跳闸之间的逻辑关系，首先把电力系统的故障诊断问题表示为０－１整数规划问题。在此基础上，采用了两种方法求解，第一种方法是遗传算法，第二种方法是模拟退火算法。这两种方法都能以较大的概率收敛到全局最优解，并能够求得多个最优解。提出的方法理论严密，与当前比较流行的以启发式知识为基础的专家系统方法不同。本文的方法可以处理任意复杂的故障，也可以处理有保护和断路器误动作的情况，且计算时间与故障的复杂程度基本无关。此外，提出的方法原则上适合于并行处理。

By use of Genetic Algorithm(GA)and Simulated Annealing(SA),two new methods to fault section estimation in power systems using are presented. The fault section estimation problem is formulated as a 0-1 integer programming problem,and the proposed methods can find multiple optimal solutions directly in a single run,which is very suitable for complex fsult diagnosis problem especially for the situations with protective relays and/or circuit breakers malfunction,because different combinationsof fault sections and malfunction Of proteCtive relays and/or circuit breakers may cause same alarms. In addition,the proposed and SA methods are ideal for parallel processing computer systems or hardwares.