Suzhou Electric Appliance Research Institute
期刊號: CN32-1800/TM| ISSN1007-3175

Article retrieval

文章檢索

首頁 >> 文章檢索 >> 往年索引

基于CEEMDAN-ISSA-BiLSTM的風(fēng)電功率組合預(yù)測模型

來源:電工電氣發(fā)布時間:2023-11-10 10:10 瀏覽次數(shù):161

基于CEEMDAN-ISSA-BiLSTM的風(fēng)電功率組合預(yù)測模型

童宇軒1,金超2,李燦3
(1 新能源電力系統(tǒng)國家重點實驗室(華北電力大學(xué)),北京 102206;
2 河海大學(xué) 能源與電氣學(xué)院,江蘇 南京 211100;
3 南京工程學(xué)院 電力工程學(xué)院,江蘇 南京 211167)
 
    摘 要:針對風(fēng)電功率存在間歇性、非線性和波動性而難以準(zhǔn)確預(yù)測的問題,提出一種遵循“序列分解-網(wǎng)絡(luò)預(yù)測-序列重構(gòu)”的風(fēng)電功率預(yù)測模型。針對風(fēng)電場集群中的不同風(fēng)電機組出力特性曲線,使用迭代自組織數(shù)據(jù)分析聚類算法 (ISODATA) 聚類得到典型出力曲線;利用自適應(yīng)噪聲完全集成經(jīng)驗?zāi)?/span>態(tài)分解 (CEEMDAN) 算法對聚類得到的原始風(fēng)電序列數(shù)據(jù)進(jìn)行模態(tài)分解,減少數(shù)據(jù)波動所帶來的預(yù)測誤差;建立各模態(tài)分量的雙向長短期記憶網(wǎng)絡(luò) (BiLSTM) 預(yù)測模型,并使用改進(jìn)麻雀搜索算法 (ISSA) 優(yōu)化網(wǎng)絡(luò)參數(shù),再將各模態(tài)分量的預(yù)測結(jié)果疊加得到風(fēng)電功率的最終預(yù)測結(jié)果。算例結(jié)果表明,所提預(yù)測模型的預(yù)測精度相比其他對比模型更高,且有著更好的泛化能力。
    關(guān)鍵詞: 風(fēng)電功率預(yù)測;自適應(yīng)噪聲完全集成經(jīng)驗?zāi)B(tài)分解;雙向長短期記憶網(wǎng)絡(luò);改進(jìn)麻雀搜索算法
    中圖分類號:TM614     文獻(xiàn)標(biāo)識碼:A     文章編號:1007-3175(2023)11-0026-07
 
Wind Power Combination Prediction Model Based on
CEEMDAN-ISSA-BiLSTM
 
TONG Yu-xuan1, JIN Chao2, LI Can3
(1 State Key Laboratory of Alternate Electrical Power System with Renewable Energy
Sources (North China Electric Power University), Beijing 102206, China;
2 College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China;
3 School of Electric Power Engineering, Nanjing Institute of Technology, Nanjing 211167, China)
 
    Abstract: To solve the problem that wind power is difficult to predict accurately due to intermittent and nonlinear fluctuations, a wind power prediction model based on“sequence decomposition-network prediction-sequence reconstruction”is proposed. First, according to the output characteristic curves of various wind turbines in wind farm clusters, the typical output curves are obtained through Iterative Self Organizing Data Analysis Techniques Algorithm (ISODATA) clustering algorithm. Then, the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) is used to decompose the original wind power series data obtained by clustering to reduce the prediction error caused by data fluctuation. Third, the Bi-directional Long Short-Term Memory (BiLSTM) prediction model of each modal component is established, the Improved Sparrow Search Algorithm (ISSA) is used to optimize the network parameters, and the final prediction result of wind power is obtained by superimposing the prediction results of each modal component. The numerical results show that the prediction accuracy of the proposed model is higher than that of other models, and it has better generalization ability.
    Key words: wind power prediction; complete ensemble empirical mode decomposition with adaptive noise; bi-directional long short-term memory; improved sparrow search algorithm
 
參考文獻(xiàn)
[1] 張智剛,康重慶. 碳中和目標(biāo)下構(gòu)建新型電力系統(tǒng)的挑戰(zhàn)與展望[J] . 中國電機工程學(xué)報,2022,42(8) :2806-2818.
[2] 魯宗相,林弋莎,喬穎,等. 極高比例可再生能源電力系統(tǒng)的靈活性供需平衡[J] . 電力系統(tǒng)自動化,2022,46(16) :3-16.
[3] 王紹敏,王守相,趙倩宇,等. 基于頻域分解和精度加權(quán)集成的分布式風(fēng)電功率預(yù)測方法[J] . 電力建設(shè),2023,44(5) :84-93.
[4] 王渝紅,史云翔,周旭,等. 基于時間模式注意力機制的 BiLSTM 多風(fēng)電機組超短期功率預(yù)測[J]. 高電壓技術(shù),2022,48(5) :1884-1892.
[5] 陳璽. 基于 ISSA-LSTM 的超短期風(fēng)電功率預(yù)測[D].銀川:寧夏大學(xué),2022.
[6] 王順. 基于物理- 數(shù)據(jù)融合的風(fēng)電功率組合預(yù)測方法研究[D]. 沈陽:沈陽工業(yè)大學(xué),2022.
[7] 冉靖,張智剛,梁志峰,等. 風(fēng)電場風(fēng)速和發(fā)電功率預(yù)測方法綜述[J] . 數(shù)理統(tǒng)計與管理,2020,39(6) :1045-1059.
[8] 王燕. 時間序列分析——基于R [M] . 北京:中國人民大學(xué)出版社,2015.
[9] 陳昊,高山,王玉榮,等. 基于廣義自回歸條件異方差偏度峰度模型的風(fēng)電功率預(yù)測方法[J] . 中國電機工程學(xué)報,2017,37(12) :3456-3461.
[10] 鄒金,朱繼忠,賴旭,等. 基于時空自回歸移動平均模型的風(fēng)電出力序列模擬[J] . 電力系統(tǒng)自動化,2019,43(3) :101-107.
[11] 朱喬木,李弘毅,王子琪,等. 基于長短期記憶網(wǎng)絡(luò)的風(fēng)電場發(fā)電功率超短期預(yù)測[J] . 電網(wǎng)技術(shù),2017,41(12) :3797-3802.
[12] 楊子民,彭小圣,郎建勛,等. 基于集群動態(tài)劃分與 BLSTM 深度學(xué)習(xí)的風(fēng)電集群短期功率預(yù)測[J] .高電壓技術(shù),2021,47(4) :1195-1203.
[13] ORESHKIN B N, CARPOV D, CHAPADOS N, et al.NBEATS: Neural basis expansion analysis for interpretable time series forecasting[C]//International Conference on Learning Representations, Millennium Hall, Addis Ababa, Ethiopia,2020.
[14] 張雲(yún)欽,程起澤,蔣文杰,等. 基于 EMD-PCA-LSTM 的光伏功率預(yù)測模型[J]. 太陽能學(xué)報,2021,42(9) :62-69.
[15] LIU H, TIAN H G, LI Y F.Comparison of new hybrid FEEMD-MLP , FEEMD-ANFIS,Wavelet Packet-MLP and Wavelet Packet-ANFIS for wind speed predictions[J].Energy Conversion and Management,2015,89(1) :1-11.
[16] 彭秀艷,張彪. 基于 EMD-PSO-LSTM 組合模型的船舶運動姿態(tài)預(yù)測[J] . 中國慣性技術(shù)學(xué)報,2019,27(4) :421-426.
[17] 張辰睿. 基于機器學(xué)習(xí)的短期電力負(fù)荷預(yù)測和負(fù)荷曲線聚類研究[D]. 杭州:浙江大學(xué),2021.
[18] 王維高,魏云冰,滕旭東. 基于 VMD-SSA-LSSVM 的短期風(fēng)電預(yù)測[J] . 太陽能學(xué)報,2023,44(3) :204-211.
[19] 王金玉,金宏哲,王海生,等.ISSA 優(yōu)化 Attention 雙向 LSTM 的短期電力負(fù)荷預(yù)測[J]. 電力系統(tǒng)及其自動化學(xué)報,2022,34(5) :111-117.