Wind Power Modeling And Prediction In The Urat Area,Inner Mongolia

With the continuous consumption of global resources and the emergence of related environmental problems,the development of renewable energy has begun to attract the attention of countries all over the world.Wind energy is a clean,abundant,sustainable,and free renewable energy source.Therefore,wind power has developed vigorously and has become the key to smart grid infrastructure.However,affected by geographical location and natural conditions,wind power has the characteristics of non-stationarity,fluctuation and indirectness,resulting in uncertainty in wind power generation.Improving the utilization rate of wind power has become a huge challenge,and effective modeling and analysis of wind power is particularly important.Inner Mongolia is rich in wind energy resources.This article takes the Urat area of Inner Mongolia as the research object and conducts modeling research on wind power in this area.At present,there have been a lot of researches to improve the utilization rate of wind power,but there are still some problems such as:1)Most of the existing wind power forecasting systems and related research are universal,and the changes in wind power in different regions are different,and the existing research is not necessarily applicable to Inner Mongolia,which is rich in wind energy resources;2)Due to differences in the randomness of wind power data,data processing methods,models and parameter selection,the accuracy of each model is different,and further research is needed to obtain good accuracy;3)Most of the wind power modeling is carried out by using 10m high wind energy data or extrapolation technology,and the wind tower in wind power generation is usually 70 meters high,which will cause certain errors.In response to the above problems,this paper uses the measured 70-100m wind tower data in the Urat area to model and analyze the wind power in the region.In this paper,we first clarify the relationship between wind speed,wind direction and wind power,and clarify that wind speed,wind direction,etc.are used as the the main goal of this research is to establish wind speed Weibull distribution model and wind speed prediction model to achieve statistical modeling analysis and forecast modeling analysis of wind energy in the Urat area.The main tasks are as follows:(1)In the analysis of wind energy statistical modeling,it is proposed to use the cross entropy method(CEM)to optimize the parameters of the wind speed Weibull distribution model,and compared with other three numerical methods and four intelligent optimization methods.According to the Weibull distribution parameters obtained by CEM,six indicators including Wind Power Density(WPD)are calculated to evaluate the wind power status in the Urat area,and further wind direction changes,seasonal changes,wind power economic construction,etc.are analyzed.The results show that the average annual wind speed in this area can reach 9.33m/s,the wind energy density is 941W/m~2,the annual effective wind energy utilization time exceeds 7000h,the electricity price is 0.072 yuan/k W～0.105yuan/k W,and the payback period of wind farm construction investment does not exceed in10 years,it can be used as a large-scale wind farm for development.(2)In the wind speed prediction modeling,in view of the problem that high frequency subsequences still exist after the wind speed data is decomposed once,which affects the final prediction accuracy,a second decomposition method of wind speed data based on EWT-VMD is proposed,and the sample entropy method is used.To determine the parameters of the decomposition algorithm,and then apply to the LSTM network model modeling of the processing time series to predict the wind speed one step in advance.Compared with baseline models,complex models,and some models in published literature,the EWT-VMD-LSTM model proposed in this paper is improved by at least 10.5%,11.4,and 1.5 compared with the comparison model’s MAE,RMSE,and R~2,and obtained a better forecasting effect.