Smell Agent Optimization Based Supervisory Model Predictive Control for Energy Efficiency Improvement of a Cold Storage System

Energy efficiency of cold storage systems has remained a challenge to industries over the years. Despite several attempts made by experts to surmount the challenge, there remains a huge potential for improvement. Therefore, a smell agent optimization (SAO) algorithm is proposed to tune a supervisory model predictive control (MPC) scheme to achieve reduction in high cooling load caused by high ambient temperatures. By this reduction, the compressor is saved from operating at saturated state, a state that is highly inefficient and detrimental to product quality. Simulation results show that the SAO-based supervisory model predictive controller achieved better performance than the traditional regulatory model predictive controller by reducing the cooling load by 55.5%. The SAO-based supervisory MPC also performed better in keeping the product temperature below 5
^o
C while the regulatory MPC response settled at 6.4
^o
C, a temperature that is detrimental to food safety and quality. The results imply that SAO-based supervisory MPC achieved a significant improvement in energy efficiency while preserving the product within the safe temperature range recommended by the food authority. All simulations are performed using MATLAB 2020a.