针对煤矿开采中电力系统稳定性低,导致易发生安全事故或导致矿坑停产的问题,研究基于狄利克雷过程高斯混合模式对历史数据进行聚类分析,并采用长短期记忆网络(Long Short-Term Memo-ry,LSTM)捕捉长期依赖关系,设计出一种基于LSTM的新能源电力系统继电保护控制模型(Dirichlet Process Gaussian Mixture Model-LSTM,DPGMM-LSTM).实验结果表明,该模型在召回率、精确率和F1分数上均表现出色,分别达到了97.2％、98.1％和91.6％,显著优于其他4种对比算法.在处理时间方面,在进行继电保护运行故障的识别时,完整的DPGMM-LSTM模型耗时最长,总耗时6ms.因此提出的新能源电力系统继电保护控制模型在系统故障预测和继电保护控制方面具有更高的准确性和可靠性,且耗时较短,有望为煤炭行业的长期安全运行提供一种新的技术支持.
Research on relay protection control of new energy power system based on improved DPGMM
In response to the low stability of the power system in coal mining,which leads to safety accidents or mine shutdowns,this study investigated the clustering analysis of historical data based on the Dirichlet Process Gaussian Mixture Model(DPGMM-LSTM),and used Long Short-Term Memory(LSTM)to capture long-term dependencies.A new energy power system relay protection control model based on LSTM(Dirichlet Process Gaussian Mixture Model-LSTM)was designed.The experimental results show that the model performs well in recall,accuracy,and F1 score,reaching 97.2％,98.1％,and 91.6％,respectively,significantly better than the other four comparative algorithms.In terms of processing time,the complete DPGMM-LSTM model takes the longest,with a total time of 6 ms,when identifying faults in relay protection operation.Therefore,the new energy power system relay protection control model proposed in the study has higher accuracy and reliability in system fault prediction and relay protection control,and shorter time consumption.It is expected to provide a new technical support for the long-term safe operation of the coal industry.
万方数据
维普
