Artificial Intelligence based fuel and energy management system in hybrid electric vehicle

Abstract

ABSTRACT newlineDay by the demand the sustainable transportation as well as stringent emission norms newlinehas been accelerated growing as the development of intelligent hybrid vehicle newlinetechnologies. Among all types of the electric vehicles, the Parallel Hybrid Electric newlineVehicle (PHEV) architecture has more popular and it taking the attention of the newlineresearchers as well as the industrialist due to its ability to balance the contributions of newlinethe power from both the internal combustion engine (ICE) and electric motors. In the newlineHybrid Electric Vehicle an efficient energy management strategy is play a very newlinecrucial role in order to increase the overall efficiency of the HEV. This thesis presents newlinea comprehensive investigation as well as implementation of an Artificial Intelligence newlinebased Energy Management System. It is designed for a parallel hybrid vehicle newlineconfiguration. Mainly two different control strategies namely artificial Neural newlineNetwork (ANN)based Control and Fuzzy Logic Controller (FLC) are developed, and newlineimplemented and further comparatively analyzed under identical simulation newlineconditions is done. Also the performance of the proposed controller is compared with newlinethe conventional HEV system. newlineThe MATLAB/Simulink simulation platform is used for modeling and simulation of newlinethe complete PHEV drive cycle. In order to analyze the performance of the proposed newlinesystem the results of the battery characteristics, torque profiles, and regenerative newlinebraking functionalities are discussed. Each of the controller is integrated with the newlineEMS in order to control the torque distribution, energy flow, and component newlineengagement strategies. The simulation time is taken as 1000 second for the evaluation newlinethe performance of the HEV system. newlineThe result of the conventional control strategy shows the basic operational stability newlinebut it is observed that they are suffered from slow transient response, higher power newlinelosses, and inefficient regenerative braking. The ANN-based EMS provides the newlineimproved adaptability and dynamic learning capabilit