Real-Time Optimization of Energy Management Strategy for Fuel Cell Vehicles Using Inflated 3D Inception Long Short-Term Memory Network-Based Speed Prediction

Abstract

The performance of speed prediction-based energy management strategy (EMS) for fuel cell vehicles (FCVs) highly relies on the accuracy of predicted speed sequences. Therefore, the future speed sequences are estimated by Inflated 3D Inception long short-term memory (LSTM) network, which can use the historical speed and image information to improve the accuracy of speed prediction. Meanwhile, the energy economy and powertrain system durability are the objectives of real-time optimization. For optimizing energy economy and powertrain system durability of FCVs, the real-time optimization of EMS using the Inflated 3D Inception LSTM network-based speed prediction is proposed. To do this, the mathematical models including energy economy and powertrain system durability of FCVs are developed at the beginning. Then, based on the predicted speed sequences, a real-time optimization method with sequential quadratic programming (SQP) algorithm is proposed to minimize the energy consumption and take into consideration powertrain system degradation in the prediction horizon. Simulation results show that the proposed EMS can significantly reduce the total cost of energy consumption and powertrain system degradation.