组合型交叉熵算法在电网故障诊断中的应用

Combinatorial Cross Entropy Algorithm for Power Networks Fault Diagnosis

通过分析交叉熵算法的原理和特征,提出一种基于组合型交叉熵算法的电网故障诊断方法。考虑故障元件与保护器和断路器的动作关系,将电网故障诊断问题表示为使目标函数最小化的0-1整数规划问题,利用组合型交叉熵算法求该问题的最优解,依据最优解的情况识别故障元件。经过与粒子群算法和遗传算法的比较分析,得知本文采用的诊断方法具有收敛速度快、稳定性好等优点。对测试系统上多起故障情况进行了模拟测试,测试结果表明故障诊断结论全部正确,进一步证明了组合型交叉熵算法在电网故障诊断中具有一定的有效性、准确性和实用性。

By analyzing the principle and character of cross entropy algorithm,a new method of fault diagnosis in power networks by combinatorial cross entropy algorithm is proposed. The power networks fault diagnosis is represented as a0-1 integer programming problem based on the logic relationship among fault elements,action information of protective relays and trip information of circuit breakers,then it gives the optimal solution by combinatorial cross entropy algo-rithm. Fault element can be identified by the optimal solution. Compared with PSO and GA,it is indicated that the proposed method have faster convergence rate,higher steadiness etc. Various fault conditions are simulated in test system. The simulation results demostrate that conclusions obtained by proposed method are correct,which further con-firms that the combinatorial cross entropy can diagnose fault availably,precisely and practicably in power networks.