基于長短時記憶網(wǎng)絡(luò)的電力系統(tǒng)負(fù)荷預(yù)測方法研究
王鑫琪,李闖,焦晗,李焱飛
(南京工程學(xué)院 電力工程學(xué)院,江蘇 南京 211167)
摘 要:準(zhǔn)確的負(fù)荷預(yù)測對保持電網(wǎng)的穩(wěn)定性和提高當(dāng)?shù)亟?jīng)濟(jì)效益、節(jié)約成本有重大幫助??紤]到負(fù)荷數(shù)據(jù)帶有時序性,以及智能電網(wǎng)的發(fā)展所帶來的數(shù)據(jù)量的增大,建立了長短時記憶網(wǎng)絡(luò)(LSTM)模型來對未來用電量進(jìn)行短期負(fù)荷預(yù)測。針對Adam訓(xùn)練算法可能存在的收斂問題,對其進(jìn)行了改進(jìn),并通過MATLAB軟件對LSTM網(wǎng)絡(luò)進(jìn)行建模,通過與BP神經(jīng)網(wǎng)絡(luò)進(jìn)行對比,結(jié)果表明,LSTM模型具有更高的精確度以及實(shí)用性。
關(guān)鍵詞:短期負(fù)荷預(yù)測;BP神經(jīng)網(wǎng)絡(luò);長短時記憶網(wǎng)絡(luò);Adam算法
中圖分類號:TM715 文獻(xiàn)標(biāo)識碼:A 文章編號:1007-3175(2019)11-0017-04
Research on Power System Load Forecasting Method Based on Long-Term and Short-Term Memory Network
WANG Xin-qi, LI Chuang, JIAO Han, LI Yan-fei
(School of Electric Power Engineering, Nanjing Institute of Technology, Nanjing 2111 67, China)
Abstract: Load forecasting is an important part of power system dispatching. Accurate load forecasting is of great help to maintain grid stability and improve local economic benefits and cost. Considering the sequential nature of load data and the increase in data volume brought about by the development of smart grids, a long-term and short-term memory network (LSTM) model was established to make shortterm predictions of future electricity consumption. On this basis, the convergence problem of Adam training algorithm may be improved. It is simulated by MATLAB software and compared with BP neural network. The results show that the LSTM model has higher accuracy and practicability.
Key words: short-term load forecast; BP neural network; long-term and short-term memory network; Adam algorithm
參考文獻(xiàn)
[1] 徐堯強(qiáng),方樂恒,趙冬華,王凱. 基于LSTM神經(jīng)網(wǎng)絡(luò)的用電量預(yù)測[J]. 電力大數(shù)據(jù),2017,20(8):25-29.
[2] 張宇航,邱才明,賀興,凌澤南,石鑫. 一種基于LSTM神經(jīng)網(wǎng)絡(luò)的短期用電負(fù)荷預(yù)測方法[J]. 電力信息與通信技術(shù),2017,15(9):19-25.
[3] 陳卓,孫龍祥. 基于深度學(xué)習(xí)LSTM網(wǎng)絡(luò)的短期電力負(fù)荷預(yù)測方法[J]. 電子技術(shù),2018,47(1):39-41.
[4] 劉進(jìn)波,周克男,任明會. 基于相關(guān)因子的BP神經(jīng)網(wǎng)絡(luò)電力負(fù)荷短期預(yù)測[J]. 信息記錄材料,2018,19(1):211-214.
[5] 陳英慧,商淼,李婷,高紅艷. 基于GA-BP算法的呼和浩特地區(qū)短期電力負(fù)荷預(yù)測[J]. 山東工業(yè)技術(shù),2018(5):170.
[6] 王德文,孫志偉. 電力用戶側(cè)大數(shù)據(jù)分析與并行負(fù)荷預(yù)測[J]. 中國電機(jī)工程學(xué)報,2015,35(3):527-537.
[7] 劉文博. 基于神經(jīng)網(wǎng)絡(luò)的短期負(fù)荷預(yù)測方法研究[D]. 杭州:浙江大學(xué),2017.
[8] 程宇也. 基于人工神經(jīng)網(wǎng)絡(luò)的短期電力負(fù)荷預(yù)測研究[D]. 杭州:浙江大學(xué),2017.
[9] 朱振偉. 氣象因素對電網(wǎng)負(fù)荷特性影響的研究[D].杭州:浙江大學(xué),2008.
[10] 李萌,程浩忠,楊宗麟,韓新陽,楊鏡非. 采用分形插值的典型日負(fù)荷曲線改進(jìn)預(yù)測方法[J]. 電力系統(tǒng)及其自動化學(xué)報,2015,27(3):36-41.
[11] 陳鯉江,景程,吳姚鑫,鄭水泉. 數(shù)學(xué)表達(dá)式的歸一化方法研究[J]. 浙江工業(yè)大學(xué)學(xué)報,2012,40(2):229-232.