基于余弦自适应差分进化算法的输电网故障诊断

Cosine adaptive differential evolution for fault diagnosis of power transmission networks

为了提高输电网故障诊断数学解析模型的求解效率,保证故障元件辨识的准确性,提出了基于余弦自适应差分进化(Cosine Adaptive Differential Evolution, CADE)算法的故障诊断方法。CADE算法针对差分进化算法变异策略和受控参数难于确定的问题,以变异策略DE/rand/1和DE/best/1为基础,采用余弦函数模型自适应融合两种变异策略,并自适应调节算法受控参数,有效平衡算法在不同搜索进程的局部搜索和全局搜索,促进算法有效收敛。结果表明,CADE算法能以较少的迭代次数完全正确诊断不同复杂度的故障情形,迭代鲁棒性强,诊断效率高。

A cosine adaptive differential evolution(CADE) based fault diagnosis method is proposed to improve the efficiency of solving the analytic model of power transmission network fault diagnosis and the identification accuracy of faulted components.CADE uses a cosine function model to adaptively integrate the mutation strategies DE/rand/1 and DE/best/1,and adjust the controlled parameters of differential evolution.Consequently, the local searching and global searching can be balanced well in different evolutionary processes.Simulation results demonstrate that CADE can diagnose fault cases with diverse complexities correctly.It consumes fewer iterations to identify the faulted components with strong robustness and high diagnostic efficiency.