發(fā)布時間：2019-04-01 06:44

【摘要】：通過建立改進的4層神經(jīng)網(wǎng)絡(luò)群,以歷史負荷電流作為樣本進行訓(xùn)練,實現(xiàn)對于未來負荷電流的預(yù)測。針對傳統(tǒng)BP神經(jīng)網(wǎng)絡(luò)易收斂到局部極值的問題,引入了動態(tài)調(diào)整的動量因子。為增強對于隨月份動態(tài)變化較劇烈的負荷的預(yù)測能力,提出了BP網(wǎng)絡(luò)群結(jié)構(gòu)。數(shù)據(jù)模擬結(jié)果說明該算法具有高精確性,可有效估算出下一階段線路電流負荷變化趨勢值,并且預(yù)測速度滿足實際使用要求。該模型可以用于監(jiān)測重點單位用電負荷變化情況,及早提示供電單位采取相應(yīng)措施,促進智能電網(wǎng)建設(shè)。
[Abstract]:By establishing an improved 4-layer neural network group and training the historical load current as a sample, the prediction of the future load current can be realized. The momentum factor of dynamic adjustment is introduced to solve the problem that the traditional BP neural network is easy to converge to local extremum. In order to enhance the forecasting ability of load with dynamic change with month, the structure of BP network cluster is proposed. The simulation results show that the proposed algorithm has high accuracy and can effectively estimate the trend value of line current load variation in the next stage, and the prediction speed can meet the requirements of practical application. The model can be used to monitor the change of power load in key units and prompt the power supply units to take appropriate measures to promote the construction of smart grid.
【作者單位】： 浙江大學(xué)電氣工程學(xué)院;
【分類號】：TM715;TP183

【相似文獻】

相關(guān)期刊論文 前10條

1 文漢云;;硫化氫燃燒的神經(jīng)網(wǎng)絡(luò)PID控制及其仿真[J];自動化與儀器儀表;2006年01期

2 張U，

本文編號：2451305