earticle

영어

Modern day vehicles can hardly mobilize without the electronic control unit (ECU) – the CPU that controls every embedded system of the vehicle. One of the core system of ECU is the electronic stability program (ESP). It intervenes the driving by controlling the engine and the brake when the vehicle is in discord with the original intention of the driver. Current models of ESP set various modes that differentiate the activation rate of the ESP. Yet, several problems exist, as drivers have to manually alter the modes that match an immutable activation rate. Therefore, this thesis introduces a new ESP algorithm that automatically sets the activation rate. Artificial neural network (ANN) is employed, and experiment is designed to train this ANN algorithm with imaginary inputs. Successful results of customizing ESP will prove the possibility of applying artificial intelligence to the rest of the functions of ECU and thereby fully customizing the ECU.