The Power Prediction Of Photovoltaic Power Station Based On Neural Network

The increasingly serious energy crisis and environmental problems Spurred humans to search for clean renewable energy.After wind power,photovoltaic power generation has become a new growth point of the renewable energy generation.However,the output power of photovoltaic power station has randomness due to many factors,such as environmental temperature,humidity,solar irradiance and so on.The output power in the event of fluctuations,will be a threat to the stability of the operation of power grids.It is really a pity that sometimes the photovoltaic power is abandoned to avoid accidents.In order to ensure the safe and stable operation of power grid and Make full use of light energy at the same time,it is necessary to accurately forecast the output power photovoltaic power station.Once the output power being accurately forecasted,it will be convenient for scheduling department to adjusted by the user side and load side of the grid timely and properly.So it makes sense to study the power prediction method of photovoltaic power station whether from the view of theoretical research or practical application.Firstly,this paper introduces the component of photovoltaic power generation,and analyzes the factors that may affect the output power of photovoltaic power station.Through qualitative analysis,the intensity of irradiation,relative humidity,temperature and wind speed are confirmed to be the main influencing factors affecting the output power of photovoltaic power.After that,measuring instruments are installed around the photovoltaic power station to collect corresponding data every quarter hour.There are too many intercoupling and time-varying factors to adopt theoretical modeling methods.So in this paper,deep learning algorithms based on neural network are choose.Using the BP neural network learning algorithms,the output power prediction model with four input factors is established.Then,this paper get rid of one of the four factors in sequence to get predicting model with three input factors.By the simulation,the influence of various factors on the power output prediction is acquired.Finally by comparing the predictive model based on temperature of solar photovoltaic panels and environmental temperature data,conclusion is obtained that the prior model can predict the output power better.