摘要
光伏阵列能否正常工作直接关系到整个光伏发电系统运行的安全性和可靠性。对于光伏阵列故障诊断中传统的BP神经网络诊断算法准确率低、收敛速度慢等问题,提出一种基于粒子群优化RBF神经网络的故障诊断算法。建立以光伏阵列的4种故障特征参数为输入、5种情况为输出的故障诊断模型,对基于粒子群算法的网络模型的自适应权重寻优进行仿真实验。最后,将优化算法与BP神经网络算法以及RBF神经网络算法进行对比。实验结果表明,优化算法不仅可以有效地诊断光伏阵列的故障类型,而且还可以提高故障诊断的准确率。
Normal or abnormal operation of PV aray is closely linked to the security and reliability of PV system. BP neural network fault diagnosis algorithm in PV array have some problems convergence speed, and so on. In order to solve these problems, a method of fault diagnosis of PV array usingRBF neural network optimized by particle swarm is put forward. The PV array fault lished , which uses PV array four characteristic parameters as input variables and five normal circumstances as output variables. The method of adaptive network weight optimization based on particle swarm algorithm is sim-ulated. Final ly, the algorithm proposed, the traditional BP neural network algorithm network algorithm are compared. The simulation experiment shows that tlie proposed algoritlim can not only effectively diagnose the fault types of PV aray, but also improve accuracy of fault diagnosis.
出处
《河南理工大学学报（自然科学版）》
CAS
北大核心
2018年第2期93-98,共6页
Journal of Henan Polytechnic University（Natural Science）
基金
国家自然科学基金资助项目(61405055)
河南省产学研基金资助项目(132107000027)
