Diagnosis and Fault-Adaptive Control for Mechatronic Systems using Hybrid Constraint Automata

Many of today’s mechatronic systems – such as automobiles, automated factories or chemical plants – are a complex mixture of hardware components and embedded control software, showing both continuous (vehicle dynamics, robot motion) and discrete (software) behavior. The problems of estimating the internal discrete/continuous state and automatically devising control actions as intelligent reaction are at the heart of self-monitoring and self-control capabilities for such systems. In this paper, we address these problems with a new integrated approach, which combines concepts, techniques and formalisms from AI (constraint optimization, hidden markov model reasoning), fault diagnosis in hybrid systems (stochastic abstraction of continuous behavior), and hybrid systems verification (hybrid automata, reachability analysis). Preliminary experiments with an industrial filling station scenario show promising results, but also indicate current limitations.