











摘要:考慮到新型電力負(fù)荷隨機(jī)性增強(qiáng), 傳統(tǒng)的準(zhǔn)確預(yù)測方法已無法滿足要求, 提出一種EMD-BiLSTM-ANFIS(Empirical Mode Decomposition-Bi-directional Long Short-Term Memory-Adaptive Network-based Fuzzy Inference System)分位數(shù)預(yù)測負(fù)荷概率密度的方法, 使用負(fù)荷預(yù)測區(qū)間取代點預(yù)測的準(zhǔn)確數(shù)值, 能為電力系統(tǒng)分析與決策提供更多數(shù)據(jù), 增強(qiáng)預(yù)測的可靠性。首先將原始負(fù)荷序列通過EMD(Empirical Mode Decomposition)分解成若干分量, 并通過計算樣本熵分為3類分量。然后將重構(gòu)后的3類分量與由相關(guān)性篩選的外界因素特征采用BiLSTM、ANFIS模型進(jìn)行訓(xùn)練和分位數(shù)回歸(QR: Quantile Regression), 并將分量的預(yù)測區(qū)間結(jié)果累加得到最終負(fù)荷的預(yù)測區(qū)間。最后利用核密度估計輸出任意時刻用戶負(fù)荷概率密度預(yù)測結(jié)果。通過與CNN-BiLSTM(Convolutional Neural Network-Bidirectional Long Short-Term Memory)、LSTM(Long Short-Term Memory)模型對比點預(yù)測及區(qū)間預(yù)測結(jié)果, 證明了該方法的有效性。
關(guān)鍵詞:經(jīng)驗?zāi)B(tài)分解; 雙向長短期神經(jīng)網(wǎng)絡(luò); 模糊推理系統(tǒng); 分位數(shù)回歸; 概率密度預(yù)測
中圖分類號: TP183; TM715 文獻(xiàn)標(biāo)志碼: A
Load Interval Forecast Based on EMD-BiLSTM-ANFIS
LI Hongyu, PENG Kang, SONG Laixin, LI Tongzhuang
Abstract:Considering that the randomness of the new power load is enhanced, the traditional accurate forecasting methods can not meet the requirements, an EMD-BiLSTM-ANFIS (Empirical Mode Decomposition Bi-directional Long Short Term Memory Adaptive Network is proposed based Fuzzy Inference System) quantile method to predict the load probability density. It replaces the accurate value of point prediction with the load prediction interval, which can provide more data for power System analysis and decision-making, The reliability of prediction is enhanced. First, the original load sequence is decomposed into several components by EMD, and then divided into three types of components by calculating the sample entropy. Then, the reconstructed three types of components and the characteristics of external factors screened by correlation. And they are used together with the Bilstm and ANFIS models for prediction training and QR(Quantile Regression), and accumulate the results of the prediction interval of the components to obtain the prediction interval of the final load. Finally, the kernel density estimation is used to output the user load probability density prediction results at any time. The validity of this method is proved by comparing the point prediction and interval prediction results with CNN-BiLSTM(Convolutional Neural Network-Bidirectional Long Short-Term Memory) and LSTM (Long Short-Term Memory)models.
Key words:empirical mode decomposition; two way long and short term neural network; fuzzy inference system; quantile regression; probability density prediction
0 引 言
雖然強(qiáng)大的能源可以帶動經(jīng)濟(jì)社會的蓬勃發(fā)展, 但能源資源的使用始終是有限的、 不自由的。在距離第1個“雙碳”目標(biāo)僅有8年的今天, 合理地使用能源、 調(diào)度資源, 是讓社會更加和諧運轉(zhuǎn)的關(guān)鍵。因此如何對電力負(fù)荷進(jìn)行準(zhǔn)確預(yù)測, 對龐大且復(fù)雜的電力系統(tǒng)網(wǎng)絡(luò)具有很高的研究價值。
早期的電力負(fù)荷預(yù)測方法是基于數(shù)學(xué)統(tǒng)計理論模型提出的, 如趨勢外推法、 時間序列法、 回歸分析法、 灰色系統(tǒng)法等傳統(tǒng)預(yù)測方法。隨著人工智能技術(shù)的發(fā)展, 機(jī)器學(xué)習(xí)及人工智能等方面的算法被廣泛應(yīng)用于電力負(fù)荷預(yù)測領(lǐng)域, 如專家系統(tǒng)、 支持向量回歸、 人工神經(jīng)網(wǎng)絡(luò)以及近幾年熱門的深度學(xué)習(xí)[1]。……