蜣螂算法优化概率神经网络的变压器故障诊断

Transformer Fault Diagnosis Based on Probabilistic Neural Network Optimized by Dung Beetle Optimizer

针对仅靠人工经验选取平滑因子的概率神经网络(PNN)变压器故障诊断模型存在诊断正确率偏低的问题,提出1种采用蜣螂算法(DBO)优化PNN平滑因子的变压器故障诊断模型。选取测试函数对DBO算法进行寻优测试,并与粒子群算法(PSO)、人工蜂群算法(ABC)、灰狼优化算法(GWO)对比,DBO在寻优精度、收敛速度和避免局部最优方面更具优势;采用DBO对PNN平滑因子寻优以建立DBO-PNN诊断模型,并与PSO-PNN、ABC-PNN和GWO-PNN模型进行诊断对比,结果表明DBO-PNN模型的诊断效果更好,正确率达96%。

Aiming at the problem that the fault diagnosis model of transformer based on probabilistic neural network(PNN)with smooth factor selected by artificial experience is not accurate,a transformer fault diagnosis model using dung beetle optimizer(DBO)to optimize PNN smoothing factor is proposed.The DBO algorithm is tested by selecting test functions.Compared with particle swarm optimization(PSO),artificial bee colony algorithm(ABC)and gray wolf optimization algorithm(GWO),the results show that DBO algorithm has more advantages in searching precision,convergence speed and avoiding local optimum.The DBO is used to optimize the smooth factor of the PNN in order to establish the DBO-PNN diagnosis model,and diagnosis comparisons are made with the PSO-PNN,ABC-PNN,and GWO-PNN models.The results show that the diagnostic performance of the DBO-PNN model is better,and its correct rate is up to 96%.