基于模糊识别的光伏发电短期预测系统  

Short-Term Photovoltaic Generation Forecasting System Based on Fuzzy Recognition

作　　者：陈昌松[1] 段善旭[1] 蔡涛[1] 代倩[1] 

机构地区：[1]华中科技大学电气与电子工程学院,武汉430074

出　　处：《电工技术学报》2011年第7期83-89,共7页Transactions of China Electrotechnical Society

基　　金：国家重点基础研究发展计划(2010CB227206);科技部2010年度公益性行业(气象)科研专项"太阳能预报技术研究"(GYHY201006036);国家自然科学基金(50907027)资助项目

摘　　要：随着光伏发电系统容量的不断扩大,光伏发电预测技术对于减轻光伏发电系统输出电能的随机性对电力系统的影响具有重要意义。根据光伏发电系统的历史发电量数据和气象数据,分析了天气类型、大气温度和太阳辐射强度等因素对预测结果的影响,采用神经网络对数值天气预测数据进行模糊识别,建立了基于模糊识别的神经网络发电预测模型。研究结果表明,神经网络的结构和扩展速度对预测结果有一定的影响;把数值天气预测数据进行模糊识别后作为神经网络的输入有利于提高神经网络的预测精度;设计的神经网络预测模型具有较高的精度,能够解决光伏发电的随机化问题,有利于电力系统的功率平衡和经济运行。With the increase of the capacity of photovoltaic power generation systems, photovoltaic power prediction is important to the operation of power system with comparatively large amount of photovoltaic power generation.A novel neutral network power forecasting model based on fuzzy recognition is proposed to solve the randomness of photovoltaic power generation.According to historical power and weather data provided by experiments,all factors which influence photovoltaic power generation are discussed and neutral network forecasting module is trained and evaluated.The results indicate that the neutral network structure and the spread value have some impact on the prediction precision and the parameter of fuzzy recognition as input will improve the precision. Forecasting results show the high precision and high efficiency of this forecasting model which is applied in stable operation of photovoltaic power generation system.

关 键 词：发电预测 神经网络 天气预报 模糊识别 光伏 

分 类 号：TM615[电气工程—电力系统及自动化]