改进粒子群优化T-S ANFIS算法诊断油浸式变压器故障研究
2023,31(10):33-39
摘要：为了有效提升油浸式变压器故障诊断的精度与速度,提出一种基于改进粒子群算法(IPSO)优化T-S型自适应模糊神经网络(T-S ANFIS)的油浸式变压器故障诊断模型；引入动态惯性权重和学习因子线性调整策略,并利用收敛域和欧式距离判别雷同粒子,以克服粒子群算法易早熟、后期易陷入局部最优的问题；接着通过IPSO对T-S ANFIS的前提参数进行优化,提高网络的收敛速度；最后通过仿真实验验证基于IPSO优化T-S ANFIS的变压器故障诊断模型效果,结果表明所构建模型的故障诊断最优准确率约为98％,与ANFIS及PSO-ANFIS模型相比具有较高的故障诊断精度及效率。
关键词：油浸式变压器;改进粒子群;自适应模糊神经网络;故障诊断;算法优化
Research on Fault Diagnosis of Oil Immersed Transformer Based on Improved Particle Swarm Optimization T-S ANFIS Algorithm
Abstract：A fault diagnosis model for oil-immersed transformers based on an improved particle swarm optimization (IPSO) optimized T-S adaptive neuro fuzzy inference system (T-S ANFIS) is proposed to enhance the accuracy and efficiency of fault diagnosis. The dynamic inertia weight and linear learning factor adjustment strategy are introduced to overcome the problem of premature convergence and local optima in particle swarm optimization. The convergence domain and Euclidean distance are utilized to distinguish identical particles. Furthermore, the T-S ANFIS"s premise parameters are optimized using IPSO to improve the network"s convergence speed. The effectiveness of the IPSO-optimized T-S ANFIS fault diagnosis model is verified through simulation experiments. The results demonstrate that the proposed model achieves an optimal fault diagnosis accuracy of approximately 98%, which is higher than that of ANFIS and PSO-ANFIS models, indicating its high accuracy and efficiency in fault diagnosis.
Key words：Oil immersed transformer; Improved particle swarm optimization; Adaptive fuzzy neural network; Fault diagnosis; Algorithm optimization
收稿日期：2023-05-06
基金项目：江苏省研究生实践创新计划项目(SJCX22_1414)
