广东电力2017，Vol.30Issue(4)：25-30,60,7.DOI:10.3969/j.issn.1007-290X.2017.04.005
基于小波分解和深度信念网络的短期风速预测
Short-term Wind Speed Forecasting Based on Wavelet Decomposition and Deep Belief Network
司徒友 1李刚强 2张水平1
作者信息
- 1. 广东电网有限责任公司东莞供电局,广东 东莞523000
- 2. 深圳大学,广东 深圳518060
- 折叠
摘要
Abstract
In allusion to the defect of a great error in wind speed forecasting by using surface learning model, this paper presents a kind of wind speed forecasting model based on wavelet transform and deep belief network (WDBN) which firstly uses wavelet transform to break up the original wind speed sequence into different frequency sequences, then designs the deep belief network model on account of each frequency sequence and according to training layer-by-layer method, and finally adopts wavelet inverse transform to reconstruct different frequency sequence so as to obtain the ultimate wind speed forecasting result.It selects and uses data about one wind power field respectively in January and July in 2013 for simulating analysis on wind speed forecasting of WDBN model, and compares results with that from autoregressive moving average method, back-propagation neural network method and Morlet wavelet neural network method.Comparison results indicate that this WDBN model can better learn non-linearity and non-stationary characteristic of wind speed and has higher forecasting precision.关键词
短期风速预测/小波变换/深度信念网络/序列重构Key words
short-term wind speed forecasting/wavelet transform/deep belief network/sequence reconstruction分类
信息技术与安全科学引用本文复制引用
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):25-30,60,7.
