论文部分内容阅读
粗糙集挖掘方法进行故障诊断的过程中,训练易陷入局部最优解,导致电力变压器故障挖掘诊断算法不适用于解决实际工程问题。提出基于RBF神经网络与粗糙集的挖掘电力变压器故障诊断方法,利用RBF神经网络收敛速度快、泛化能力强等特点,对故障数据进行训练。依据给出的训练样本特征获取所求概率密度函数的统计值,将改进后的数据发送至粗糙集。在保证系统分类能力的条件下,依据分类规范,实现电力变压器故障挖掘诊断,对采集到的100组电力变压器故障数据进行仿真分析。结果表明,所提方法的局部搜索能力明显优于传统方法,所提方法在变压器故障的诊断准确率上大大高于传统方法,保证了电力变压器运行的安全性与可靠性。
In the process of fault diagnosis based on rough set mining method, the training is apt to fall into the local optimal solution, resulting in the power transformer fault diagnosis algorithm is not suitable for solving practical engineering problems. The fault diagnosis method of power transformer based on RBF neural network and rough set is proposed. The fault data is trained by using RBF neural network with fast convergence rate and extensive generalization ability. Obtain the statistical value of the probability density function according to the characteristics of the training samples, and send the improved data to the rough set. Under the condition of ensuring the capability of system classification, according to the classification standard, fault diagnosis of power transformer can be diagnosed, and the fault data of 100 sets of power transformer collected can be simulated and analyzed. The results show that the proposed method has better local search ability than the traditional method. The proposed method is much superior to the traditional method in the diagnostic accuracy of transformer faults, ensuring the safety and reliability of power transformer operation.