基于改进AIA-BP算法的煤矿供电线路故障类型识别方法

Fault Type Identification Method of Coal Mine Power Supply Line Based on Improved AIA-BP Algorithm

为了提升煤矿电力系统中供电线路故障识别方法的准确率,提出一种基于小波变换与灰色AIA-BP的煤矿供电线路故障类型识别方法。采用小波变换分析对电力暂态信号进行故障特征提取;引入灰色关联度法对人工免疫算法(AIA)进行改进,并对BP神经网络的的权值与阈值进行了优化,建立灰色AIA-BP故障识别算法模型。仿真结果表明,灰色AIA-BP模型的识别准确率高出传统AIA-BP模型11.37%,故障识别误差率相较于单一BP神经网络模型降低了6.47%,验证了所提方案能够有效地提高供电线路的故障诊断精度;提出模型的平均识别准确率为96.81%,相较于GA-SVM模型和RBF模型提升了14.32%和21.99%,可以对供电线路故障类型进行精确识别判断。

In order to enhance the accuracy of the electrical supply line failure recognition method in the coal mine power system,it proposes a method of fault type recognition method based on the wavelars of coal mines based on small wave transformation and gray AIA-BP.Use a wavelet transformation analysis to extract the fault characteristics of the temporary signal of the power.The introduction of the gray association method to improve the artificial immune algorithm(AIA),and optimize the weight and threshold of the BP neural network,establish a gray AIA-BP fault Identify the algorithm model.The simulation results show that the accuracy of the recognition accuracy of the gray AIA-BP model is 11.37%higher than that of the traditional AIA-BP model,and the fault recognition error rate is reduced by 6.47%compared to a single BP neural network model.The diagnostic accuracy of the line;the average recognition accuracy of the model is 96.81%.Compared with the GA-SVM model and the RBF model,it has increased by 14.32%and 21.99%,which can be accurately identified and judged by the type of failure of the power supply line.