A new set of multivariable predictive control algorithms for time-delayed nonsquare systems of different domains: A minimum-energy examination

• Published in: Applied energy 2025-03, Vol.381, p.125093, Article 125093
• Main Authors: Hunek, Wojciech P., Feliks, Tomasz
• Format: Article
• Language: English
• Subjects: Generalized inverses; Inverse (internal) model control; Minimum-energy design; Multivariable nonsquare objects; Systems with time delays; Transfer-function/state-space domains
• ISSN: 0306-2619
• DOI: 10.1016/j.apenergy.2024.125093

A new approach to the minimum-energy design of stochastic inverse model control-oriented predictive control algorithms dedicated to the multivariable physical systems is proposed in the paper. For this reason, the novel transfer-function-type stochastic solutions in the forms of respective continuous-time minimum variance control (CMVC) and discrete-time minimum variance control (DMVC), both employing generalized inverses, are examined. The theoretical and practical simulation examples confirm high advantages of the original σ and Smith factorization-oriented inverses over the benefits derived from the well-established Moore–Penrose inverse regarding the energy-based robustification of the discussed control procedures. Henceforth, from now on, the industrial real-life systems can be developed toward a minimum-energy consumption at the same time preserving the maximum-speed and maximum-accuracy important maintenance for modern sustainable energy plants. [Display omitted] •A new concept of predictive algorithms devoted to physical MIMO systems is offered.•The idea is occupied by the IMC-based minimum-energy studies addressing time delays.•The generalized inverse-derived finding is already valid for input-output domains.•The original inverses outperform the Moore–Penrose inverse in terms of convergence.