Research On Energy Management Strategy Of Diesel-electric Hybrid Train

With the continuous development of power battery technology,it has become a trend to use power battery to research and develop new hybrid trains which are more energy saving and environmental protection in rail transit.Diesel-electric hybrid train is one of the schemes to transform the traditional diesel locomotive.The diesel-electric hybrid system optimizes the power distribution between the diesel generator set and the battery by formulating a reasonable energy management system control strategy.It can improve fuel efficiency compared to traditional internal combustion engines.And it has lower operating cost and better economy.Firstly,in order to optimize the energy management system strategy of the diesel-electric hybrid system,the paper establishes the diesel-electric hybrid train model.The diesel-electric hybrid train model mainly includes the train dynamics model,fuel consumption model,power battery model and threshold rule energy management strategy model.Then,the energy management strategy is optimized to minimize the diesel consumption of diesel-electric hybrid train according to Pontryagin minimum principle.An off-line energy management strategy is obtained based on Pontryagin minimum principle.And its effectiveness is verified by train history.Then,in order to meet the requirement of reducing the battery life loss of some trains,the Pontryagin minimum principle strategy is improved by establishing the battery loss model and adding penalty factors related to the life loss.Aiming at the online real-time operation of management system strategy,this paper proposes a Pontryagin minimum principle online energy management strategy based on model prediction,which can be applied to online real-time optimization control.Then,in order to predict the working condition more effectively,the two working condition prediction models based on Markov chain and neural network are compared and analyzed.And the more accurate working condition prediction method based on neural network is finally selected.Then,the regulation law of the coordination factorλ_bat in the Pontryagin minimum principle is analyzed,and two online regulation methods are proposed:the coordination factorλ_bat regulation method based on the battery SOC programming and the coordination factorλ_bat adaptive regulation method based on the neural network.The model prediction energy management strategy based on SOC programming and the model prediction energy management strategy based on neural network parameters adaptation are obtained.Finally,based on the real train conditions,the economic performance of each energy management strategy is compared and analyzed.It is concluded that the global optimal solution can be obtained by the off-line energy management strategy based on Pontryagin minimum principle.Compared with the threshold rule strategy,the economy can be improved by 4.89%on average.Among the online energy management control strategies,the model prediction strategy based on the neural network coordination factorλ_bat adaptive adjustment method has the best economic performance.The average economic performance of the model prediction strategy is 4.45%higher compared with the threshold rule strategy.At the same time,it has high reliability and fast calculation speed.And it can realize the online operation of energy management strategy,which has a strong engineering application prospect.