基于模糊聚类和组合预测算法的光伏短期预测
|
摘要
光伏出力受气象因素影响大,未计及气象因素的光伏出力预测结果精度不高;将多种气象数据直接作为模型输入增加了预测模型的维度;BP神经网络具有很强的线性映射能力,被广泛应用于光伏出力预测中,但是单一的BP预测模型预测误差较大。针对这些问题,提出一种基于模糊聚类和组合预测算法的方法进行光伏出力短期预测。通过pearson相关系数定量分析光伏出力与太阳辐照度、大气浑浊度、相对湿度等气象因素之间的相关性,将这三类气象数据作为聚类特征向量并通过改进的模糊C均值聚类(IFCM)算法将光伏出力历史数据和待预测日数据聚类成三类,然后选取待预测日所属类别的历史数据和待预测日的温度数据并利用遗传膜优化BP神经网络预测模型(GAPS-BP)进行预测。最后,通过实际案例仿真,验证了所提模型和优化算法的有效性,表明所提方法能够有效降低模型输入维度并提高预测精度。
This paper proposes a short-term prediction method for photovoltaic power output based on fuzzy clustering and combination prediction algorithm. The correlation between the photovoltaic output and the meteorological factors(such as solar irradiance, atmospheric turbidity, relative humidity) is quantitatively analyzed by pearson correlation coefficient. These three types of meteorological data are used as clustering feature vectors. The PV output history data and the data to be predicted are divided three categories by improved fuzzy C-means clustering(IFCM) algorithm. Then the historical data of the category to be predicted and the temperature data are selected and the genetic film to optimize the BP neural network Model(GAPS-BP) is used to predict. Finally, the simulation on an actual case shows that the effectiveness of the proposed model and the optimization algorithm. The proposed method can effectively reduce the model input dimension and improve the prediction accuracy.
This paper proposes a short-term prediction method for photovoltaic power output based on fuzzy clustering and combination prediction algorithm. The correlation between the photovoltaic output and the meteorological factors(such as solar irradiance, atmospheric turbidity, relative humidity) is quantitatively analyzed by pearson correlation coefficient. These three types of meteorological data are used as clustering feature vectors. The PV output history data and the data to be predicted are divided three categories by improved fuzzy C-means clustering(IFCM) algorithm. Then the historical data of the category to be predicted and the temperature data are selected and the genetic film to optimize the BP neural network Model(GAPS-BP) is used to predict. Finally, the simulation on an actual case shows that the effectiveness of the proposed model and the optimization algorithm. The proposed method can effectively reduce the model input dimension and improve the prediction accuracy.
引文
[1] 龚莺飞,等. 光伏功率预测技术[J].电力系统自动化, 2016, 40(4):140-151.
[2] 袁晓玲,施俊华,徐杰彦. 计及天气类型指数的光伏发电短期出力预测[J]. 中国电机工程学报, 2013,33(34):57-64.
[3] L Chong, et al. Short-term PV Generation Forecasting Based On Weather Type Clustering And Improved GPR Model[C]. 2016 China International Conference on Electricity Distribution (CICED 2016). Xi'an, China:IEEE, 2015:1889-1893.
[4] 董雷,等. 基于动态贝叶斯网络的光伏发电短期概率预测[J]. 中国电机工程学报, 2013,33(S1): 38-45.
[5] 王守相,张娜. 基于灰色神经网络组合模型的光伏短期出力预测[J]. 电力系统自动化, 2012,36(19):37-41.
[6] Shen Jinrong, Jie Hui, Ying Ni. Study on the Photovoltaic Power Generation Prediction Methods Based on Comprehensive Regression Analysis of Environmental Factors[C]. 2015 5th International Conference on Electric Utility Deregulation and Restructuring and Power Technologies (DRPT), Changsha, China:IEEE, 2015: 1889-1893.
[7] 张程熠,等. 适用于小样本的神经网络光伏预测方法[J]. 电力自动化设备, 2017,37(1):101-111.
[8] 赵书强,等. 基于不确定理论的光伏出力预测研究[J]. 电工技术学报, 2015,30(16): 213-220.
[9] 申彦波,赵宗慈,石广玉. 地面太阳辐射的变化、影响因子及其可能的气候效应最新研究进展[J]. 地球科学进展, 2008,23(9): 915-923.
[10] 代倩,等. 基于天气类型聚类识别的光伏系统短期无辐照度发电预测模型研究[J]. 中国电机工程学报, 2011,31(34):28-35.
[11] 廖松有,张继福,刘爱琴. 利用模糊熵约束的模糊C 均值聚类算法[J]. 小型微型计算机系统, 2014,35(2):379-382.
[12] 姚仲敏,潘飞,沈玉会. 基于GA-BP和 POS-BP 神经网络的光伏电站出力短期预测[J]. 电力系统保护与控制, 2015,43(20):83-89.
[13] 陈政. 一种改进的遗传算法优化BP网络的研究及应用[D]. 暨南大学, 2010.
[14] 黄亮. 膜计算优化方法研究[D]. 浙江大学, 2007.
[15] 罗娟. 改进的膜计算优化算法及其在微网负荷预测与经济运行中的应用研究[D]. 西华大学, 2015.
[16] 吴泓俭,等. 基于遗传膜算法的含风电机组和电动汽车的配电网分时段动态重构[J]. 电工技术学报, 2016,31(2):196-220.
[17] 齐月,房世波,周文佐. 近50年来中国东、西部地面太阳辐射变化及其与大气环境变化的关系[J]. 物理学报, 2015,64(8): 089201-1-089201-10.
[18] 姚仲敏,等. 基于GA-BP和POS-BP 神经网络的光伏电站出力短期预测[J]. 电力系统保护与控制, 2015,43(20): 83-89.
[2] 袁晓玲,施俊华,徐杰彦. 计及天气类型指数的光伏发电短期出力预测[J]. 中国电机工程学报, 2013,33(34):57-64.
[3] L Chong, et al. Short-term PV Generation Forecasting Based On Weather Type Clustering And Improved GPR Model[C]. 2016 China International Conference on Electricity Distribution (CICED 2016). Xi'an, China:IEEE, 2015:1889-1893.
[4] 董雷,等. 基于动态贝叶斯网络的光伏发电短期概率预测[J]. 中国电机工程学报, 2013,33(S1): 38-45.
[5] 王守相,张娜. 基于灰色神经网络组合模型的光伏短期出力预测[J]. 电力系统自动化, 2012,36(19):37-41.
[6] Shen Jinrong, Jie Hui, Ying Ni. Study on the Photovoltaic Power Generation Prediction Methods Based on Comprehensive Regression Analysis of Environmental Factors[C]. 2015 5th International Conference on Electric Utility Deregulation and Restructuring and Power Technologies (DRPT), Changsha, China:IEEE, 2015: 1889-1893.
[7] 张程熠,等. 适用于小样本的神经网络光伏预测方法[J]. 电力自动化设备, 2017,37(1):101-111.
[8] 赵书强,等. 基于不确定理论的光伏出力预测研究[J]. 电工技术学报, 2015,30(16): 213-220.
[9] 申彦波,赵宗慈,石广玉. 地面太阳辐射的变化、影响因子及其可能的气候效应最新研究进展[J]. 地球科学进展, 2008,23(9): 915-923.
[10] 代倩,等. 基于天气类型聚类识别的光伏系统短期无辐照度发电预测模型研究[J]. 中国电机工程学报, 2011,31(34):28-35.
[11] 廖松有,张继福,刘爱琴. 利用模糊熵约束的模糊C 均值聚类算法[J]. 小型微型计算机系统, 2014,35(2):379-382.
[12] 姚仲敏,潘飞,沈玉会. 基于GA-BP和 POS-BP 神经网络的光伏电站出力短期预测[J]. 电力系统保护与控制, 2015,43(20):83-89.
[13] 陈政. 一种改进的遗传算法优化BP网络的研究及应用[D]. 暨南大学, 2010.
[14] 黄亮. 膜计算优化方法研究[D]. 浙江大学, 2007.
[15] 罗娟. 改进的膜计算优化算法及其在微网负荷预测与经济运行中的应用研究[D]. 西华大学, 2015.
[16] 吴泓俭,等. 基于遗传膜算法的含风电机组和电动汽车的配电网分时段动态重构[J]. 电工技术学报, 2016,31(2):196-220.
[17] 齐月,房世波,周文佐. 近50年来中国东、西部地面太阳辐射变化及其与大气环境变化的关系[J]. 物理学报, 2015,64(8): 089201-1-089201-10.
[18] 姚仲敏,等. 基于GA-BP和POS-BP 神经网络的光伏电站出力短期预测[J]. 电力系统保护与控制, 2015,43(20): 83-89.