چكيده به لاتين
The antilock braking system (ABS) is one of the most practical active safety systems in vehicles and plays an effective role in vehicle safety during heavy braking. However, the non-linearity and unmodified uncertainties in its structure make its control problematic. That's why designing a robust controller is critical to these systems. Radial basis function neural networks (RBF-NNs) have shown a high ability to model nonlinear functions. Also, type-2 fuzzy systems have always been considered as one of the most well-known solutions to deal with uncertainties. Therefore, in this dissertation, we propose an adaptive and robust control structure based on RBF-NN with respect to the properties of type-2 distance fuzzy systems (IT2FSs). In order to control the ABS, we used the quarter vehicle model along with the modified Dugoff tire nonlinear model. We also used the properties of the gradient descent backpropagation method to adapt and update the parameters of the control system. The update is done online due to momentary changes in road surface conditions and will compensate for these changes. The stability analysis of the proposed method is also expressed with the help of the Lyapunov function. In order to better evaluate its method, it simulates in three different road scenarios and compares it with common control methods such as sliding mode controller (SMC), interval type-2 fuzzy PID controller (IT2F-PID), and interval type-2 fuzzy SMC (IT2F-SMC) were performed and the simulation and comparison results were presented. The results of the simulation showed that the proposed system, in addition to the presence of disturbing factors and various road conditions, is able to detect the slip of the reference wheel and shows better performance than other methods.
