罗欣,张爽,景伟强,朱蕊倩,魏骁雄,陈博,葛岳军
LUO Xin,ZHANG Shuang,JIN Weiqiang,ZHU Ruiqian,WEI Xiaoxiong,CHEN Bo,GE Yuejun

• 1:国网浙江省电力有限公司电力科学研究院
• 2:浙江华云信息科技有限公司

摘要(Abstract)：

以深度学习为代表的人工智能2.0技术为短期95598话务工单异动预警提供了可行的技术手段。为研究LSTM神经网络深度学习算法对话务工单预测的可行性,提出一种基于LSTM深度学习建模的短期95598话务工单异动预警方法。以95598系统实际运行数据为例,分析95598话务工单的异动规律与预测应用场景。实例分析结果表明,该方法可以更好地学习话务工单所具有的动态学习预测和智能异动预警特征,具有较高的预测精度。
The artificial intelligence 2.0 technology represented by deep learning provides a feasible technical means for short-term alteration warning of 95598 telephone orders. In order to study the feasibility of LSTM network deep learning algorithm for telephone orders prediction, a short-term 95598 telephone orders prediction method based on LSTM network deep learning modeling is proposed. Through the application of the actual operation data of 95598 system, this paper analyzes the variation rule of 95598 telephone orders and forecast the application scenario. The results of case analysis show that the method can better learn the dynamic learning prediction and intelligent alteration warning features of telephone orders, and has high prediction accuracy.

关键词(KeyWords)： 95598话务工单预测;置信区间异动;LTSM;深度学习
95598 telephone order prediction;confidence interval alteration;LTSM;deep learning

Abstract:

Keywords:

基金项目(Foundation):

作者(Author): 罗欣,张爽,景伟强,朱蕊倩,魏骁雄,陈博,葛岳军
LUO Xin,ZHANG Shuang,JIN Weiqiang,ZHU Ruiqian,WEI Xiaoxiong,CHEN Bo,GE Yuejun

DOI: 10.19585/j.zjdl.201812007

参考文献(References)：

• [1]田诺,王殊,张明杰.江苏电力话务量分析预测系统的设计与应用[J].电力需求侧管理,2012,14(6):51-53.
• [2]张明杰,迪里达尔·库尔班,余锦河.电力客户服务报修业务量预测分析[J].电力需求侧管理,2015,17(3):57-60.
• [3]刘斌,陶建华.联合长短时记忆递归神经网络和非负矩阵分解的语音混响消除方法[J].信号处理,2017,33(3):268-272.
• [4]周晓莉,张丰,杜震洪,等.基于CRBM算法的时间序列预测模型研究[J].浙江大学学报,2016,43(4):442-451.
• [5]ZHANG S L,WANG Y X,LIU M Q,et al.Data-based line trip fault prediction in power systems using LSTMnetworks and SVM[J].IEEE Access,2017,6(1):1-11.
• [6]陈亮,王震,王刚.深度学习框架下LSTM网络在短期电力负荷预测中的应用[J].电力信息与通信技术,2017,15(5):8-11.
• [7]WANG Q L,GUO Y f,YU L X,et al.Earthquake prediction based on spatio-temporal data mining:An LSTM Network Approach[J].IEEE Transactions on Emerging Topics in Computing,2017,1(1):1-11.
• [8]史佳琪,张建华.基于深度学习的超短期光伏精细化预测模型研究[J].电力建设,2017,38(6):28-35.
• [9]吴润泽,包正睿,宋雪莹,等.基于深度学习的电网短期负荷预测方法研究[J].现代电力,2018,35(2):43-48.
• [10]黎静华,黄乾,韦善阳,等.基于S-BGD和梯度累积策略的改进深度学习方法及其在光伏出力预测中的应用[J].电网技术,2017,41(10):3292-3299.
• [11]周文杰,严建峰,杨璐.基于深度学习的用户投诉预测模型研究[J].计算机应用研究,2017,34(5):1428-1432.
• [12]严宇平,吴广财,江疆.单变量回归确定移动平均权重算法在话务量预测中的应用研究[J].电子设计工程,2015,23(22):44-49.
• [13]杨旭华,钟楠祎.基于深度信念网络的医院门诊量预测[J].计算机科学,2016,43(11):26-30.
• [14]司徒友,李刚强,张水平.基于小波分解和深度信念网络的短期风速预测[J].广东电力,2017,30(4):25-30.
• [15]夏耀杰.最优组合预测方法在中长期电力负荷预测中的应用[J].电力需求侧管理,2017,18(4):18-22.
• [16]AYDIN O,GULDAMLASIOGLU S.Using LSTM networks to predict engine condition on large scale data processing framework[J].International Conference on Electrical&Electronics Engineering,2017,5(1):281-285.
• [17]CORTEZ B,CARRERA B,KIM Y J,et al.An architecture for emergency event prediction using LSTM recurrent neural networks[J].Expert Systems With Applications,2018,97(1):315-324.
• [18]LIU B Y,CHENG J R,CAI K Q,et al.Singular Point Probability Improve LSTM Network Performance for Long-term Traffic Flow Prediction[J].Theoretical Computer Science,2018,32(768):328-340.
• [19]吴亮,张潮,陈琼.用电信息系统运行数据的统计与分析[J].浙江电力,2017,36(4):56-59.
• [20]龙厚印,刘卫东,黄锦华,李黎.基于业扩报装的月度负荷预测[J].浙江电力,2016,35(12):11-14.
• [21]黄宇腾,韩翊,赖尚栋.深度神经网络在配电网公变短期负荷预测中的应用研究[J].浙江电力,2018,37(5):1-6.