基于K均值模式划分改进模糊聚类与BP神经网络的风力发电预测研究
|
摘要
为了克服模糊聚类不能反应数据内部差别的不足,提出了一种基于模式划分改进的模糊聚类与BP神经网络的风电功率预测算法。该算法首先确定最佳的模式划分数,接着将不同的风速模式下的历史风速数据进行模糊聚类以确定关联系数,并对关联系数高的历史风速数据和发电数据进行训练,然后利用BP神经网络得出风电预测功率。以新疆某地区实际风力发电数据作为仿真算例,对比分析了所提算法与未改进模糊聚类与BP神经网络在风力发电预测中的误差,结果表明所提算法克服了模糊聚类的缺点,具有更高的精度,对地区发电计划安排具有较高的价值。
In order to overcome the disadvantage of fuzzy clustering in reflecting the inner difference of data,a wind power prediction algorithm is put forward based on K-meas pattern classification improved fuzzy clustering and BP neural network.The pattern of wind speed is defined by K-means method firstly, the correlation coefficient of the history wind speed data is defined by fuzzy clustering method under different wind speed pattern,the high correlation coefficient historical wind speed and wind power electricity data would be trained,and obtaining the predicting wind electricity power in BP neural network. Taking actual wind power generation data in a certain region of Xinjiang as simulation example,the errors of the proposed algorithm and the unimproved fuzzy clustering and BP neural network in wind power generation prediction are compared and analyzed. The results show that the proposed algorithm can overcome the disadvantage of fuzzy clustering and have higher accuracy,which can provide reference for the regional power generating plan.
In order to overcome the disadvantage of fuzzy clustering in reflecting the inner difference of data,a wind power prediction algorithm is put forward based on K-meas pattern classification improved fuzzy clustering and BP neural network.The pattern of wind speed is defined by K-means method firstly, the correlation coefficient of the history wind speed data is defined by fuzzy clustering method under different wind speed pattern,the high correlation coefficient historical wind speed and wind power electricity data would be trained,and obtaining the predicting wind electricity power in BP neural network. Taking actual wind power generation data in a certain region of Xinjiang as simulation example,the errors of the proposed algorithm and the unimproved fuzzy clustering and BP neural network in wind power generation prediction are compared and analyzed. The results show that the proposed algorithm can overcome the disadvantage of fuzzy clustering and have higher accuracy,which can provide reference for the regional power generating plan.
引文
[1]童光毅.关于当代能源转型方向的探讨[J].智慧电力,2018,46(10):1-3.TONG Guangyi. Probe into Modern Energy Structure Transition[J].Smart Power,2018,46(10):1-3.
[2]舒印彪,张智刚,郭剑波,等.新能源消纳关键因素分析及解决措施研究[J].中国电机工程学报,2017, 37(1):1-8.SHU Yinbiao, ZHANG Zhigang, GUO Jianbo, et al.Study on key factors and solution of renewable energy accommodation[J]. Proceedings of the CSEE, 2017, 37(1):1-8.
[3] YUAN Xiaohui,TAN Qingxiong,LEI Xiaohui et al. Wind power prediction using hybrid autoregressive fractionally integrated moving average and least square support vector machine[J].Energy,2017,129(15)122-137.
[4]丁珩,杨晓坤,邓长虹,等.风电特性及其对电网调峰影响的量化研究[J].湖北电力,2017,41(12):28-32.DING Heng, YANG Xiaokun, DENG Changhong, et al.Wind power characteristics and its quantitative research on power generation peak shaving[J].Hubei Electric Power,2017,41(12):28-32
[5]田书欣,程浩忠,曾平良,等.大型集群风电接人输电系统规划研究综述[J].中国电机工程学报,2014, 34(10):1566-1574.TIAN Shuxin, CHENG Haozhong, ZENG Pingliang, et al.A review of the planning and research of large cluster wind power transmission system[J]. Proceedings of the CSEE,2014, 34(10):1566-1574.
[6]鲁宗相,李海波,乔颖.高比例可再生能源并网的电力系统灵活性评价与平衡机理[J].中国电机工程学报,2017, 37(1):9-20.LU Zongxiang, LI Haibo, QIAO Ying. Flexibility evaluation and supply/demand and balance principle of power system with high penetration renewable electricity[J].Proceedings of the CSEE, 2017, 37(1):9-20.
[7]王中夫,华栋.消纳大规模风电并网的多目标鲁棒调度[J].广东电力,2016,29(6):35-42+49.WANG Zhongfu, HUA Dong. Multi-objective robust dispatching for large-scale wind power grid-connection absorption[J].Guangdong Electric Power, 2016,29(6):35-42.
[8]刘振亚.中国电力与能源[M].北京:中国电力出版社,2012.4北京.
[9] BILAL B, NDONGO M, ADJALLAH K H,et al. Wind turbine power output prediction model design based on artificial neural networks and climatic spatiotemporal data[C].Lyon:IEEE Industrial Technology,2018:1085-1092.
[10]吴钢,徐枫,王冰,等.基于遗传算法和模糊聚类算法的改进BP神经网络风速预测[J].电子设计工程,2016,24(11):120-123.WU Gang, XU Feng, WANG Bing, et al.The improved BP network wind speed forecasting method based on genetic algorithm and fuzzy c-means algorithm[J].Electronic Design Engineering, 2016, 24(11):120-123.
[11] LI Y Q, DAI C B, WANG T X et al.Separate wind power and ramp predictions based on meteorological variables and clustering method[C].New Delhi:IEEE Power Systems Conference,2016:1-6.
[12]王钤,潘险险,陈迎,等.基于实测数据的风电场风速-功率模型的研究[J].电力系统保护与控制,2014,42(2):23-27.WANG Qian,PAN Xianxian, CHEN Ying, et al.Study of wind speed-active power model for wind farm based on measured data[J]. Power System Protection and Control,2014,42(2):23-27.
[13]王尤嘉,鲁宗相,乔颖,等.基于特征聚类的区域风电短期功率统计升尺度预测[J].电网技术,2017,41(5):1383-1389.WANG Youjia, LU Zongxiang, QIAO Ying et al. Shortterm regional wind power statistical upscaling forecasting based on feature clustering[J]. Power System Technology,2017,41(5):1383-1389.
[14]于秋玲,许长清,李珊.基于模糊聚类和支持向量机的短期光伏功率预测[J].电力系统及其自动化学报,2016,(12):115-118.YU Qiuling, XU Changqing, LI Shan.Application of fuzzy clustering algorithm and support vector machine to shortterm forecasting of PV power[J].Proceedings of the CSUEPSA, 2016,(12):115-118.
[15]彭小圣,熊磊,文劲宇,等.风电集群短期及超短期功率预测精度改进方法综述[J].中国电机工程学报,2016, 36(23):6315-6327.PENG Xiaosheng, XIONG Lei, WEN Jinyu, et al. A summary of the state of the art for short-term and ultrashort-term wind power prediction of regions[J].Proceedings of the CSEE,2016,36(23):6315-6327.
[16]陶玉波,陈昊,秦晓辉,孟昭军.短期风电功率预测概念和模型与方法[J].电力工程技术,2018,37(5):7-13.TAO Yubo,CHENG Hao, QIN Xiaohui,et al.A review of the short term wind power forecasting theory,model and approach[J]. Electric Power Engineering Technology,2018,37(5):7-13.
[2]舒印彪,张智刚,郭剑波,等.新能源消纳关键因素分析及解决措施研究[J].中国电机工程学报,2017, 37(1):1-8.SHU Yinbiao, ZHANG Zhigang, GUO Jianbo, et al.Study on key factors and solution of renewable energy accommodation[J]. Proceedings of the CSEE, 2017, 37(1):1-8.
[3] YUAN Xiaohui,TAN Qingxiong,LEI Xiaohui et al. Wind power prediction using hybrid autoregressive fractionally integrated moving average and least square support vector machine[J].Energy,2017,129(15)122-137.
[4]丁珩,杨晓坤,邓长虹,等.风电特性及其对电网调峰影响的量化研究[J].湖北电力,2017,41(12):28-32.DING Heng, YANG Xiaokun, DENG Changhong, et al.Wind power characteristics and its quantitative research on power generation peak shaving[J].Hubei Electric Power,2017,41(12):28-32
[5]田书欣,程浩忠,曾平良,等.大型集群风电接人输电系统规划研究综述[J].中国电机工程学报,2014, 34(10):1566-1574.TIAN Shuxin, CHENG Haozhong, ZENG Pingliang, et al.A review of the planning and research of large cluster wind power transmission system[J]. Proceedings of the CSEE,2014, 34(10):1566-1574.
[6]鲁宗相,李海波,乔颖.高比例可再生能源并网的电力系统灵活性评价与平衡机理[J].中国电机工程学报,2017, 37(1):9-20.LU Zongxiang, LI Haibo, QIAO Ying. Flexibility evaluation and supply/demand and balance principle of power system with high penetration renewable electricity[J].Proceedings of the CSEE, 2017, 37(1):9-20.
[7]王中夫,华栋.消纳大规模风电并网的多目标鲁棒调度[J].广东电力,2016,29(6):35-42+49.WANG Zhongfu, HUA Dong. Multi-objective robust dispatching for large-scale wind power grid-connection absorption[J].Guangdong Electric Power, 2016,29(6):35-42.
[8]刘振亚.中国电力与能源[M].北京:中国电力出版社,2012.4北京.
[9] BILAL B, NDONGO M, ADJALLAH K H,et al. Wind turbine power output prediction model design based on artificial neural networks and climatic spatiotemporal data[C].Lyon:IEEE Industrial Technology,2018:1085-1092.
[10]吴钢,徐枫,王冰,等.基于遗传算法和模糊聚类算法的改进BP神经网络风速预测[J].电子设计工程,2016,24(11):120-123.WU Gang, XU Feng, WANG Bing, et al.The improved BP network wind speed forecasting method based on genetic algorithm and fuzzy c-means algorithm[J].Electronic Design Engineering, 2016, 24(11):120-123.
[11] LI Y Q, DAI C B, WANG T X et al.Separate wind power and ramp predictions based on meteorological variables and clustering method[C].New Delhi:IEEE Power Systems Conference,2016:1-6.
[12]王钤,潘险险,陈迎,等.基于实测数据的风电场风速-功率模型的研究[J].电力系统保护与控制,2014,42(2):23-27.WANG Qian,PAN Xianxian, CHEN Ying, et al.Study of wind speed-active power model for wind farm based on measured data[J]. Power System Protection and Control,2014,42(2):23-27.
[13]王尤嘉,鲁宗相,乔颖,等.基于特征聚类的区域风电短期功率统计升尺度预测[J].电网技术,2017,41(5):1383-1389.WANG Youjia, LU Zongxiang, QIAO Ying et al. Shortterm regional wind power statistical upscaling forecasting based on feature clustering[J]. Power System Technology,2017,41(5):1383-1389.
[14]于秋玲,许长清,李珊.基于模糊聚类和支持向量机的短期光伏功率预测[J].电力系统及其自动化学报,2016,(12):115-118.YU Qiuling, XU Changqing, LI Shan.Application of fuzzy clustering algorithm and support vector machine to shortterm forecasting of PV power[J].Proceedings of the CSUEPSA, 2016,(12):115-118.
[15]彭小圣,熊磊,文劲宇,等.风电集群短期及超短期功率预测精度改进方法综述[J].中国电机工程学报,2016, 36(23):6315-6327.PENG Xiaosheng, XIONG Lei, WEN Jinyu, et al. A summary of the state of the art for short-term and ultrashort-term wind power prediction of regions[J].Proceedings of the CSEE,2016,36(23):6315-6327.
[16]陶玉波,陈昊,秦晓辉,孟昭军.短期风电功率预测概念和模型与方法[J].电力工程技术,2018,37(5):7-13.TAO Yubo,CHENG Hao, QIN Xiaohui,et al.A review of the short term wind power forecasting theory,model and approach[J]. Electric Power Engineering Technology,2018,37(5):7-13.