Optimal Control of Automotive Multivariable Dynamical Systems

Two distinctive features of challenging control engineering problems are commonly taken into consideration in design of dynamical, mechatronics systems, namely operation ranges, of such systems with nonlinear effects, which are not always near to equilibrium states, as well as a fairly high level of uncertainties of their physical description with which controllers have to cope despite a lack of knowledge on the all system parameters although physical modeling allows to identify their particular nonlinear effects. It should be noted that usage of nonlinear physical modeling in real-time control systems can be computationally very demanding. Hence, it seems to be suitable to use robust control methods based on linearized models with adaptive updating algorithms. However, usually strong nonlinearities can reduce the effectiveness of control methods, and thus of adaptive control algorithms. The controller gains can be often updated by using the estimated parameters. In this contribution the adaptive control systems for automotive applications, which are based on indirect (or self-tuning) controller strategies are discussed. The modeling issue of indirect optimal controller strategies is illustrated by the application example.