Using Artificial Neural Networks (ANN), a battery health aware energy management strategy (EMS) is created for a power-split hybrid electric vehicle (HEV). To acquire a dataset, three distinct speed profiles are used. The highway speed profile is HWFET, the third speed profile is NEDC, and the dynamic, transient speed profile is WLTP. During these driving cycles, the vehicle is run in charge-sustaining mode utilising the Equivalent Consumption Minimisation Strategy (ECMS). In simulations, three distinct beginning State-of-Charge (SOC) values are employed. There are three distinct starting SOC levels for each cycle. The ICE torque and speed are controlled by two ANN controllers. The vehicle’s torque requirement, the state of charge, and the fading of the battery capacity are chosen as the ANN’s inputs. This research aims to investigate fuel usage and battery deterioration via the use of artificial neural networks. According to WLTP results, with the lowest starting SOC value, capacity fading may be decreased by up to 14.85% and fuel consumption can be lowered by 3.83%. Fuel consumption is lowered by 1.84% and capacity fading is decreased by 13.80% for intermediate starting SOC values. With a 5.75% rise in fuel consumption, capacity fading is decreased by 14.70% for the highest starting SOC value. For the other two driving cycles, the outcomes are the same. In HWFET and NEDC, battery deterioration is also decreased.