Abstract:

In order to solve the problem that V2G charging and discharging system participates in the interaction of power grid,but lacks orderly management,an orderly charging and discharge control strategy considering the weather factors is proposed.The strategy improves the training sample selection method of BP neural network based on the Euclidean distance of weather data.Then,the trained network model can reliably predict the active or reactive power regulation value of the charging and discharging motor.The simulation results in MATLAB show that the strategy can support the active or reactive power in time under the mode of charging or discharging.Finally,one V2G prototype with power of 60 kW is demonstrated and applied in virtual power plant.The measured data shows that the prototype can achieve the expected goal of power dispatching by combinating the current state of power grid.The proposed strategy can effectively avoid the disorderly flow of charging and discharging energy,and enhance the ability of the power grid to deal with abnormal indicators.

Key words: weather factors, vehicle to grid (V2G), neural network, predicte, power regulation