Research On The Applicability And Application Of Electric Vehicle Load Forecasting Methods

The electric vehicle industry has the incomparable advantage over ordinary fuel vehicle as a new clean environmental protection transportation vehicle, rapid development has taken placed under the support of national policy in recent years. It can be expected that in the future electric vehicle’s charging load will become an important part of a grid electricity load which can’t be ignored. However, the electric vehicle charging behavior will cause great impact to grid, with the considered of orderly for electric vehicle scheduling, power management and distribution network planning, electric load forecasting is put forward more higher demand of forecasting accuracy.This paper selects three forecast scenario sample data include:quick changing type electric buses, fast charging electric taxis and quick changing electric taxi, do the research of three kinds of electric vehicles forecasting methods based on grey theory、 the probability model and BP neural network load forecasting theory for applicability of the principle, mathematical model and forecasting accuracy respectively.Do the analysis of the relationship between input data of grey forecasting model and the load forecasting accuracy of short-term forecasting, the first is the relationship between different amounts of input data and prediction accuracy, the second is the relationship between the discrete degree of input data and the prediction precision.Application based on grey theory and BP neural network, analysis and study the applicability of the two models in the electric vehicle ultrashort-term and short-term load forecasting, the results showed that in practical applications, the BP neural network prediction effect is better than grey prediction, especially under the ultrashort-term time scales load forecasting, it can effectively reduce the average prediction error and relative error of maximum load under the consideration of predict a moment after take the data before that moment with BP neural network model.Compare and analysis the applicability of the medium and long-term time scales load forecasting with the models based on the grey theory and probability with electric vehicle. Example calculation shows that, probability model and grey theory prediction method in mid-long term load forecasting field have their own advantages and disadvantages. Load forecasting method based on probability model is more suitable for medium and long-term load forecasting considering with national policies and electric vehicle development scale of a typical day in the future in principle. Load forecasting method based on grey theory is applicable to the electricity consumption and the application of daily maximum load forecasting.With the consideration of demand of distributed grid temporal and spatial load forecasting, I use the characteristic of EV is that the spatial distribution is not related to the time. Build the temporal and spatial load forecasting model with the time model of EV load forecasting and the load distribution model respectively of EV.