Optimal fixed-time sliding mode control for anti-lock braking systems based fuzzy logic and neural network
Najlae Jennan, El Mehdi Mellouli
Abstract
This study addresses the challenge of optimizing the performance of anti-lock braking systems (ABS) to enhance vehicle safety and improve operational efficiency. The research introduces a novel control strategy that combines fixed-time sliding mode control (SMC), artificial neural networks (ANN), Takagi-Sugeno (T-S) fuzzy logic, and particle swarm optimization (PSO). The ABS system is modeled and controlled using a fixed-time SMC approach, with T-S fuzzy logic employed to approximate the friction function of the ABS model. ANN is used to approximate the reaching law, ensuring optimal fixed-time convergence. PSO is then employed to optimize an additive term in the reaching law, with the aim of reducing errors from the ANN approximation. The stability of the overall system has been validated using the Lyapunov approach. The results of simulations demonstrate that the proposed method offers a significant improvement in braking performance compared to existing methods. This approach achieves better system stability, reduced chattering and enhanced braking efficiency. • Studying the mathematical model of ABS system and Controlling it using the fixed-time sliding mode control methodology. • Approximating the friction function using the fuzzy logic technique. • Approximating the reaching law using the neural network method to minimize the chattering problem resulting from the sliding mode control. • Optimizing the supplementary term added to the reaching law using the particle swarm optimization approach in order to alleviate the errors of neural network. • Studying the system stability using Lyapunov approach in order to generalize the control law respecting the stability conditions.