Multi-Objective Optimisation-Oriented Model of an Air-Conditioned Environment

Abstract

A dynamic modelling of a workplace in an area with high temperature and low humidity in the summer is described. The aim is to subsequently use the model to apply multi-objective predictive supervision of the environment climate. Two main variables are modelled: thermal comfort and energy consumption. First, a model of the different temperatures, humidity levels and the behaviour of the air-conditioning system is created. These variables respond to two control actions: the setpoint of the air-conditioning system and the speed of an airflow that may come from outside, and to a disturbance, the external temperature. The overall model is disintegrated into sub-models that precisely describe the internal influences between the variables. This model provides new insights on the subject in that it correctly describes the dynamic behaviour of the system by using simple Linear Parameter Varying descriptions to model the non-linear relationships, incorporates humidity into the thermal comfort model, includes forced ventilation of outside air as a strategy to reduce electricity consumption, defines a relative power unit that simplifies the measurement of electrical consumption, and its methodology is easy to apply to other environments. Different experiments demonstrate that the model accurately depicts the actual variables.