Neural network augmented backstepping control for uncertain nonlinear systems - application to laboratory antilock braking system

A new control approach is proposed to address the tracking problem of a class of uncertain nonlinear systems. In this approach, one relies first on a partially known model of the system to be controlled using a backstepping control strategy. The obtained controller is then augmented by an online artificial neural network (ANN) that serves as an approximator for the neglected dynamics and modelling errors. Thus, the developed method combines backstepping approach and ANN to address the tracking problem for uncertain systems. The proposed approach is systematic, and exploits the known nonlinear dynamics to derive the stepwise virtual stabilising control laws. At the final step, an augmented Lyapunov function is introduced to derive the adaptation laws of the network weights. The suggested control algorithm is tested experimentally on a Laboratory ABS system showing satisfactory results although the system is highly nonlinear and with unknown physical parameters.