Power Forecasting Model For Photovoltaic Power Generation System And Application Research

As the situation of the global energy shortage increasing prominently,our future development strategy is looking for renewable new energy to achieving sustainable development.The PV(photovoltaic)power generation is one of the new energy power generation industries which has a commercial development foreground.And the PV generation technology has become an important researching area.No matter where it has a large application value.The fluctuating,intermittent and not scheduling nature of PV power generation,which is influenced by external environmental factors,poses problems for grid operators.The PV power generation will increase the difficulty of the power grid dispatching after paralleling in grid.And it is harmful to the power grid.Performing research into PV array power generation prediction can address such problems,successful forecasting of PV power generation is highly significant for both power grid security and economical grid operation.In view of the output prediction problem of the solar PV power generation system,this article research the applications of different models of the forecasting theory,also the advantages and disadvantages of each model.And then we put forward a combined forecasting method which called ARIMA(Autoregressive Integrated Moving Average)-Wavelet Analysis in this paper.This model fits the characteristics of PV power generation.Processing the computed residual result by wavelet de-noising after first order difference.On the one hand,the integrality of original data can be guaranteed and reflecting the real situation exactly.On the other hand,making the data more smooth and avoiding the harmful effect leaded by multi-difference.At the same time we can achieve a more accurately result by the corrected residual items.The distinguished identification model of weather types and season types have been joined into the ARIMA-Wavelet Analysis model.Then we improve the forecasting accuracy by analyzing the change rule of power generation and making the model more scientific.Finally we make a science error analysis of the prediction results,and it can reflect the advantages and disadvantages of the model more clearly.To improve the accuracy of PV power generation forecasting and to meet the requirements of engineering applications,a combination forecasting method based on a wavelet de-noising principle to optimize the ARIMA model is proposed.Though the analysis of the generation changing rule,we conclude a proper parameters and build a best model.In addition,an experiment was performed using this model in combination with data that was collected at the Wuhan International Expo Center.The experimental error results indicate that the model is simple,and the amount of required calculation is small.Predictions based on this model can meet the requirements of engineering applications and exhibit good feasibility.It has a great application prospect to further solve the practical engineering problems.