Decentralized temperature indirect adaptive predictive control of a passive air conditioning unit

This paper presents an application of the indirect adaptive generalized predictive control strategy (IAGPC) with a decentralized architecture to the temperature management of a passive air conditioning unit, presenting an ecological advantage (without cycle of compression and absorption, no gas CFC rejection). The objective of the unit studied, which is composed of three subsystems, is to guarantee a microclimate with controlled temperature and relative humidity setpoints for crop growth chambers. As the process involves time-varying and distributed parameters, the use of a recursive estimation approach with a fixed forget factor was adopted to estimate in real time the system parameters, and to adapt simultaneously the GPC controller parameters. In order to achieve a local and global efficient performance, the proposed decentralized IAGPC architecture is applied to two principal subsystems which compose the global unit. A significant real time experimental improvement in the system performance is observed on temperature control for a wide range of operating points.