A self-triggered stochastic model predictive control for uncertain networked control system

ABSTRACT

This paper is concerned with the stochastic model predictive control problem (SMPC) for a class of networked control systems with exogenous disturbances and restrictions on communication frequency. A self-triggered SMPC algorithm with an improved triggering condition is designed which integrates the co-design of both the current control inputs and the maximum sampling interval, with the aim of reducing the amounts of communication transmission while guaranteeing the specific performance. To give consideration to both probabilistic guarantee and computability, chance constraints on both states and inputs are transformed into deterministic ones by leveraging Cantelli's inequality and linearisation techniques, so as to be solvable for the optimisation problem. Meanwhile, the nonconvex terms in dynamics of covariance propagation are averted through introducing the covariance upper bound control approach. The formulated online optimisation problem is convex to all decision variables. The theoretical analysis on recursive feasibility and closed-loop stability is addressed for the proposed self-triggered SMPC scheme. Finally, the efficacy and achievable performance of the proposed algorithm are showcased through numerical simulations.

KEYWORDS:

• Stochastic model predictive control
• self-triggered mechanism
• networked control system
• probabilistic constraints

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by National Natural Science Foundation of China [Grant No. 61533013, 61633019].