论文部分内容阅读
针对水电机组故障诊断的复杂性和传统算法存在的缺点,提出采用布谷鸟搜索算法优化BP模糊Petri网进行故障诊断。首先利用布谷鸟搜索算法的全局搜索功能对网络参数寻优,将得出的全局最优解作为BP模糊Petri网的最优初始参数,再用选取的故障样本数据对模糊Petri网进行学习训练,建立故障特征集到故障类型集的映射关系以实现故障分类。仿真试验表明,该故障诊断方法收敛速度快、准确率高,可应用于实际水电机组故障诊断。
Aiming at the complexity of fault diagnosis of hydropower units and the shortcomings of the traditional algorithms, a cuckoo search algorithm is proposed to optimize BP fuzzy Petri nets for fault diagnosis. Firstly, the global search function of Cuckoo search algorithm is used to search the network parameters, and the global optimal solution is used as the optimal initial parameter of BP fuzzy Petri net. Then the training of the fuzzy Petri net is conducted by using the selected fault sample data, Establish the mapping relationship between the fault feature set and the fault type set to realize the fault classification. Simulation results show that the method has the advantages of fast convergence and high accuracy, which can be applied to fault diagnosis of actual hydropower units.