Guaranteed cost control of networked systems with stochastic medium-access protocols: methodology and applications

摘要

This study investigates a switching control problem for a set of plants which share communication channels, and for which medium-access control is achieved based on contention-based protocols such as the binary exponential back-off algorithm. The medium-access procedure can be modelled by a Markov chain with completely or partially known transition probabilities. The entire networked system operates as a stochastic switching system (SSS). Based on the average dwell time technique, guaranteed cost control method, and SSS theory, sufficient conditions are derived to establish the relation between the mean-square exponential stability and the property of the Markov chain-driven medium-access control for the SSS. Depending on these conditions, the designed feedback controller can stabilise the SSS using the mean-square exponential and guarantee the desired robust performance. The proposed control framework is extended and applied to the cooperative adaptive cruise control of vehicles within the context of intelligent vehicle highway systems, which can reject the disturbance resulting from the fluctuating acceleration of preceding vehicles. Numerical simulations and experiments using laboratory-scale cars show the effectiveness of the proposed control method.