Antilock Braking System (ABS) Based Control Type Regulator Implemented by Neural Network in Various Road Conditions

Abstract

AbstractAntilock Braking System (ABS) is a braking system to avoid the wheel of the vehicle unlocked. This is happened caused by the friction coefficient between tire and road surface will degrade when the brakes are applied on a slippery surface or during panic breaking. The control algorithm has a limited ability to compensate for a wide variety of road conditions. The learning to the controller can take enabling to compensate for adverse road condition. This paper proposes a solution to hand the problem above by applies the controller that able to learn based on an artificial neural network. By using the neural network concept do not need to learn the inverse dynamics of the plant controlled as usual for neural in the control system. The design concept results from the belief that the main objective of the control system design is to determine the controller generating the signal to achieve the best performance output plant. Therefore, the neural network training just takes a little time (about 300–500 periods) the neural network can get the desire error target (0.02). The simulation has been applied and takes good performance on breaking standards, such as maintaining 20% slip value and keeping the maximum point of breaking coefficient for various adverse road conditions. Instead of this, the neuro-regulator can keep the 10–30% slippery. Simulation’s effectiveness compares with the two-form controller i.e., Bang-bang Controller and Fuzzy logic control in various road conditions.KeywordsAntilock braking system (ABS)Neural networkVarious road