Energy Management in Electric Vehicles Using Function Fitting Neural Network based Controller

An improved artificial neural network (ANN) controller is implemented in an electric vehicle battery management. One of the vital characteristics of a battery storage system in electric vehicles is charging and discharging during various operating conditions. Proper Selection of control scheme for a bidirectional converter in the battery energy storage system, is a crucial factor. Here in this paper an electric vehicle with a lithium-ion battery with DC-DC converter is modelled in MATLAB/SIMULINK. The control scheme such as PID and a function fitting neural network (FNN) is applied to the battery of an electric vehicle. The model is simulated with PID and FNN and the results are compared through various disturbance conditions such as variation of cell temperature and reduction of battery capacity. The system parameter has been self-adjusted to reduce noise, distortions, and oscillations. The system parameter disturbances like voltage, current power, and state of charge (SOC) in battery, converter is compensated by the proposed controller. It is depicted that the outcome of the controller supersedes the conventional PID controller.