摘要
为了较为准确的对并网光伏发电系统的发电量做出预测,提高光伏并网后电网的稳定性及安全性。文章对硅太阳电池组件发电功率进行了理论计算,建立了多元线性回归光伏发电功率及发电量预测模型。通过改进水电、火电和风电现有的发电量预测模型(基于BP神经网络和G(1,1)灰色理论模型),使得改进后模型更适合于并网光伏发电系统发电量的预测。最后,对三种预测模型的优缺点进行了比较,为今后并网光伏发电的预测提供了一种较为准确可行的方法。
To accurately forecast the generation of grid-connected PV power system and improve security and stability of grid-connected PV system,this paper has made calculations on power output of a single component of silicon solar cell,and made a generation forecasting model of grid-connected PV power system basing on Multiple Linear Regression.The BP neural network and G(1,1)Grey Theory model that had be used to predict the generation in hydropower,thermal power and windpower was modified and be used to forecast the generation of grid-connected PV power system.By comparing the advantages and disadvantages of the three forecasting models above,a more accurate and feasible method for the future prediction of grid-connected PV power generation was provided.
出处
《云南师范大学学报(自然科学版)》
2011年第2期33-38,64,共7页
Journal of Yunnan Normal University:Natural Sciences Edition
关键词
并网光伏系统
非平衡性
随机性
预测模型
可行性
Grid-connected PV system
Non-equilibrium
Randomicity
Forecasting model
Feasibility