基于改进DBN算法的电力故障预测模型与辅助分析系统

Power fault prediction model and auxiliary analysis system based on improved DBN algorithm

针对如何提高停电事件分析能力,提出了改进型神经网络,构建改进DBN算法模型,通过该方法实现如何实现停电事件分析,然后对提取到的停电事件信息进行训练,停电事件信息训练改进DBN算法模型,建立停电事件分析预测模型,实现对停电事件分析的精准预测。研究还设计了停电事件辅助分析系统,通过采用节点误差数据组的方式区分停电事件和异常数据,通过误差补偿装置提高了DBN算法采用数据的精度。实验结果表明,在进行对停电事件分析预测的精确度测试时,停电事件分析预测的准确度可达97%,在可靠性测试时,停电事件分析管理可靠性可达96%。

In order to improve the ability to analyze power outage events,an improved neural network was proposed and an improved DBNalgorithmmodelwasconstructed.Throughthismethod,howtoachievepoweroutageeventanal⁃ysis was achieved,and then the extracted power outage event information was trained.The power outage event informa⁃tion training improved the DBN algorithm model,established a power outage event analysis and prediction model,and achieved accurate prediction of power outage event analysis.This study also designed an auxiliary analysis system for power outage events,which distinguished power outage events and abnormal data by using node error data groups.The accuracy of the data used in the DBN algorithm was improved through error compensation devices.Results showed that the accuracy of power outage event analysis and prediction could reach 97%when the accuracy test was carried out,and the reliability of power outage event analysis and management could reach 96% during the reliability test.