Energy-efficient hybrid model predictive control of mobile refrigeration systems

Abstract

Efficiency and temperature compliance of last-mile refrigerated transport suffer from more severe disturbances than stationary applications. The significant energy expenditure and diminished quality or loss of goods have been tackled recently by improved hardware design. However, standard operating strategies only partly use accompanying intrinsic potential. This article introduces an enhanced model predictive control scheme for mobile refrigeration systems with secondary loop cooling units, aiming toward energy-efficient operation complying with temperature restrictions. Its real-time capable optimization covers operational flexibilities and dynamics (storage capacities, door openings, permissible temperature range) more extensively than state-of-the-art solutions. Experimental validation on a specifically designed test rig uses a representative test run with expected and unexpected door openings and a pull-down sequence. A proportional–integral and a standard predictive controller serve for comparison. Performance assessment shows energy savings of 16.4% and improvement in temperature compliance of 3.4% of the advanced predictive over the proportional–integral scheme. If door openings are known upfront, energy consumption decreases by even 29.6%. The standard predictive algorithm shows a similar energy performance but substantially declined temperature compliance as it cannot harness the entire flexibility. Thus, the proposed advanced strategy significantly contributes to more efficient refrigerated transport with reduced environmental impact and loss of goods.