Investigation of the computational speed of Laguerre network-based MPC in the thermal control of energy-efficient buildings

Abstract

Abstract:The design of computationally efficient model predictive control (MPC) systems for the thermal control of buildings is a challenging task since long prediction horizons may be needed, which can take a significant computational time, especially when multizone buildings are considered. In this paper, we investigate the computational performance of a potential approach for this purpose, Laguerre network-based MPC (LN-MPC), for thermal control of buildings, where parameterization of control input over the prediction horizon is used to reduce the number of decision variables. The computational performance of the suggested framework with comparison to the classical MPC framework is investigated through a detailed case study. It was observed that although LN-MPC can produce almost the same results as the classical MPC with a considerably smaller number of decision variables, it has no computational advantage. The potential reasons behind the lack of improvement in the computational performance of LN-MPC are also discussed.

A Non-Cooperative Distributed Model Predictive Control Using Laguerre Functions for Large-Scale Interconnected Systems

“…The performance index corresponding to the tracking error 𝐽 𝑒 𝑖 is formulated using the squared difference between the set-point y 𝑠𝑝 and the predicted plant behavior y ̂. Similar to the tracking error term, the performance index for the control effort 𝐽 𝑢 𝑖 is also formulated as in (45). The control effort is determined by the squared difference between the present and past control actions.…”

“…A method of designing an MPC using orthonormal functions was proposed with the main advantage of reducing the number of tuning parameters used for the description of the control signal trajectory. This makes fewer computations compared to the traditional MPC approach [44][45][46]. The change in the control trajectory was achieved through the adjustment of the scaling factor incorporated in the orthonormal function.…”

Dual Adaptive Model Predictive Controller Application to Vertical Roller Mill Process Used in the Cement Industry

“…With the support of operator assistance, these methods are not feasible for a complete mill. The multi-variable interacting nature, complex dynamics, nonlinear kinetics, and slowly varying feed characteristics make it challenging to implement a control strategy relying only on operator experience [20,21,22]. Different control methods used in VRM process are summarized in Table 2.…”

Optimal partitioning of multi-thermal zone buildings for decentralized control

“…Among building thermal comfort control strategies, model predictive control (MPC) and its variants are the most popular ones [1]- [4], which can provide energy savings of up to 40% compared to an on-off or rule-based controller [5]- [8]. In energy-efficient thermal control of buildings, both centralized MPC (C-MPC) and decentralized MPC (D-MPC) can be used.…”