PV System Power Forecasting Based on Neural Network with Fuzzy Processing of Weather Factors

Abstract:

A short-term PV system power forecasting method is presented in the paper based on neural network considering fuzzy characteristics of weather factors. Weather factors that affect PV system power output mainly include temperature, radiation intensity, rain and relative humidity which are all of strong fuzziness. The paper firstly made use of membership functions to process their fuzziness. Then, the historical power data of a PV system was put into neural network together with fuzzy processed historical weather data to train the network, therefore, neural network that be able to forecast PV power was get. Finally, data of an actual PV system in Colorado was employed to methods with and without fuzzy processing of weather factors, results show that the method with fuzzy processing is more accurate than that without fuzzy processing.

• Periodical: Advanced Materials Research (Volumes 860-863)
• Pages: 172-175
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.860-863.172
• Online since: December 2013
• Authors: Yong Qiang Hu*, Ming Yu Wang
• Keywords: Fuzzy Processing, Neural Network (NN), Power Forecasting, Pv System, Weather Factors

References

[1] Ying-zi Li, Jin-cang Niu. Forecast of Power Generation for Grid-Connected Photovoltaic System Based on Markov Chain[C]. Power and Energy Engineering Conference, (2009).

DOI: 10.1109/appeec.2009.4918386

[2] LAN Hua, LIAO Zhimin, ZHAO Yang. ARMA Model of the Solar Power Station Based on Output Prediction [J]. Electrical Measurement & Instrumentation, 2011, 48(2): 31-35. (In Chinese).

[3] LI Ran, LI Guang-min. Photovoltaic power generation output forecasting based on support vector machine [J]. ELECTRIC POWER, 2008, 41(2): 74-77. (In Chinese).

[4] Fu Meiping，Ma Hongwei，Mao Jianrong．Short-term photovoltaic power forecasting based on similar days and least square support vector machine [J]. Power System Protection and Control，2012, 40(16): 65-69. (In Chinese).

[5] MELLIT A, ARAB A H, KHORISSI N, et al. An ANFIS-based forecasting for solar radiation data from sunshine duration and ambient temperature [C]. IEEE Power Engineering Society General Meeting, Tampa, FL, USA, 2007: 7-12.

DOI: 10.1109/pes.2007.386131

[6] ZHANG Lan, ZHANG Yan-xia, GUO Chang-min, et al. Photovoltaic system power forecasting based on neutral networks [J]. ELECTRIC POWER, 2010, 43(9): 75-78. (In Chinese).

[7] Atsushi Yona, Tomonobu Senjyu, Toshihisa Funabashi. Application of Recurrent Neural Network to Short-Term-Ahead Generating Power Forecasting for Photovoltaic System[C]. Power Engineering Society General Meeting, 2007: 1-6.

DOI: 10.1109/pes.2007.386072

[8] Information on http: /www. nrel. gov/midc.