Predictive control for networked high-order fully actuated systems subject to communication delays and external disturbances
Da-Wei Zhang,Guo-Ping Liu
DOI: https://doi.org/10.1016/j.isatra.2023.03.041
IF: 7.3
2023-04-12
ISA Transactions
Abstract:Highlights • A HOFA system model is used to describe the dynamics of NCSs, which uses the full actuation feature to simplify the complexities of control design and avoid some shortcomings of model reduction, such as ill-conditioned matrices, unstable numerical solutions,. • An IHOFA prediction model is developed to establish the multi-step ahead predictions, the predictive control protocol is designed and represented in a HOFA form, which enhances the control effects to improve the tracking performance. • A HOFA feedback is introduced to adjust the closed-loop system performances, which establishes the IHOFA prediction model in a closed-loop form and constructs a feedback correction to regulate the IHOFA prediction model, such that the tracking performance can be improved. • The communication delays in sensor to network node and network node to actuator channels can be actively compensated, such that the stability criterion is constructed with less conservatism and is easy to expand in practice. This research is concerned with an output tracking problem for networked high-order fully actuated (NHOFA) systems subject to communication delays and external disturbances, where communication delays occur in sensor to network node and network node to actuator. A HOFA system model, as a novel system representation, is applied to establish the dynamics of networked control systems (NCSs). Accordingly, a disturbance observer based HOFA predictive control approach is proposed to address this problem. In the proposed approach, a disturbance observer is utilized to cope with the external disturbances, and then a local HOFA feedback with disturbance compensation is designed to adjust the closed-loop system performances. Further, a Diophantine Equation is applied to establish an incremental HOFA (IHOFA) prediction model to substitute a reduced-order prediction model, such that multi-step ahead predictions are derived to minimize a cost function involving the optimization of tracking performance and the compensation of network-induced communication delays. A necessary and sufficient criterion is given to discuss the stability and tracking performance of closed-loop NHOFA systems, it is simple to use in system analysis and extend in practice. The availability of the proposed approach is demonstrated via simulated and experimental results for tracking control of air-bearing spacecraft (ABS) simulator.
automation & control systems,instruments & instrumentation,engineering, multidisciplinary