A new method for adaptive brake control

We consider the problem of minimizing the longitudinal braking distance for a single wheel rolling along a surface with unknown tyre-road characteristics. The friction coefficient is modelled by a nonlinear function of slip and tyre-road parameter which corresponds to the actual road conditions. Our method is based on recently proposed adaptive control technique that uses adaptation algorithms in integro-differential, or finite form. These algorithms are capable of dealing with nonlinear parametrizations, and they also ensure improved transient performance of the controlled system. We show that, for a class of practically relevant parameterizations of friction curves, it is possible to steer the system adaptively to the desired state without invoking sliding-mode or gain-scheduling control. At the same time we show that it is possible to estimate the optimal value of the tyre slip ensuring maximal braking force. These estimates, produced by essentially the standard PI algorithm, are used in the control loop to enhance efficiency of the brakes.