Differential Function Fitting Neural Based Neural Network Energy Management Scheme for Plug-In Hybrid Electrical Vehicles with Real-Time Speed Profile and Optimal Battery Depth of Discharge

Abstract

This paper proposes the Differential Function Fitting Neural Based Neural Network Energy Management Scheme (DEMS) for PHEVs aimed at controlling the battery state of charge (SOC) and depth of discharge (DoD) based on the integration of the Pontryagin’s Maximum Principle (PMP) and Artificial Neural Network (ANN). Recognizing the market development in the area of energy efficiency in PHEVs, the increase in battery performance is crucial. The proposed DEMS utilizes PMP for accurate energy management, and optimized learning from ANN for making dynamic changes as per dynamic driving environment. The system is also verified through MATLAB/Simulink models employing a rule-based dynamic programming (DP) strategy. Results demonstrate the proposed scheme's ability to maintain optimal SOC levels, reduce battery degradation, and improve overall vehicle efficiency. The DEMS significantly enhances battery performance, stability, and adaptability, offering a reliable solution for efficient PHEV operation under diverse driving scenarios. This approach contributes to the sustainable development of hybrid vehicle technologies by optimizing energy use and extending battery life.