Prediction of remaining mileage of hydrogen energy vehicles based on LSTM

摘要

The research and usage of hydrogen energy vehicles are of great significance in solving environmental pollution and energy shortage problems. Compared with traditional fuel vehicles, which can use the remaining fuel to predict the remaining mileage, while the remaining amount of hydrogen in the hydrogen storage vessel cannot be directly obtained in hydrogen energy vehicles, so cannot determine how long the remaining hydrogen energy can support the vehicles’ operation. This is the "mileage anxiety" that plagues hydrogen energy vehicle drivers. This paper proposed a remaining mileage prediction model based on the operating mode of the vehicles. The operating mode of the hydrogen energy vehicle is divided into four types. According to the current driving data of the vehicles, it is judged which operating mode the vehicles belongs to, combined with the history running data, use the LSTM algorithm to fit the hydrogen remaining mass curve in the hydrogen storage vessel, calculate the remaining driving time of the vehicles, and predict the remaining mileage.