基于天气类型及改进的小波神经网络的光伏发电短期预测  

Short Term Forecasting of Photovoltaic Power Generation Based on Weather Type and Improved Wavelet Neural Network

作　　者：陈美珍 柳扬 徐胜彬 郭俊锋 张永强 林金阳 CHEN Meizhen;LIU Yang;XU Shengbin;GUO Junfeng;ZHANG Yongqiang;LIN Jinyang(National Demonstration Center for Experimental Electronic Information and Electrical Technology Education,Fujian University of Technology,Fuzhou 350118,China;Research Center for Microelectronics Technology,Fujian University of Technology,Fuzhou 350118,China)

机构地区：[1]福建工程学院电子信息与电气技术国家级实验教学示范中心,福建福州350118 [2]福建工程学院微电子技术研究中心,福建福州350118

出　　处：《贵州大学学报（自然科学版）》2022年第1期70-77,88,共9页Journal of Guizhou University:Natural Sciences

基　　金：福建省自然科学基金资助项目(20120J01879);福建省科技一般资助项目(GY-Z18038)。

摘　　要：本文提出了一种基于天气类型和季节类型,以布谷鸟算法优化小波神经网络的光伏发电短期预测方法。首先,分析气象因子的特征,并利用皮尔逊相关系数计算气象因子与光伏发电之间的相关性,作为预测神经网络的输入向量;其次,为了避免小波网络的结构不稳定以及由于局部极小值容易陷入预测结果误差太大的问题,提出了利用布谷鸟算法优化小波神经网络(CS-WNN)的预测结构;最后,基于天气类型和季节类型构建了布谷鸟算法优化小波神经网络(CS-WNN)预测模型进行仿真实验,并建立了遗传算法优化BP神经网络(GA-BP)、遗传算法优化小波神经网络(GA-WNN)、WNN神经网络、BP神经网络4个模型与本文结构进行对比。仿真实验结果表明,本文描述的预测方法预测精度较高,预测效果好。This paper proposes a short-term prediction method of photovoltaic power generation based on weather type and season type, which uses cuckoo algorithm to optimize wavelet neural network. Firstly, the characteristics of meteorological factors are analyzed. The correlation between meteorological factors and photovoltaic power generation is calculated by Pearson correlation coefficient as the input vector of predictive neural network. In order to avoid the instability of wavelet network structure and the error of prediction result caused by local minimum, cuckoo algorithm is used to optimize the prediction structure of wavelet neural network. Finally, the cuckoo algorithm optimized wavelet neural network(CS-WNN) prediction model is built based on the weather type and season type for simulation experiments, and the genetic algorithm optimized BP neural network(GA-BP), genetic algorithm optimized wavelet neural network(GA-WNN), WNN neural network, Four models of BP neural network are compared with the structure of this paper. The simulation results show that the prediction method described in this paper has high prediction accuracy and good prediction effect.

关 键 词：天气类型 小波神经网络 布谷鸟算法 

分 类 号：TP389.1[自动化与计算机技术—计算机系统结构] TM615[自动化与计算机技术—计算机科学与技术]