A Hierarchical Learning System for Ambient Environmental Control of Open Plan Buildings

Advances in team methodologies have resulted in the reconfiguration of older buildings towards physical open plan seating. Many Building Management Systems (BMS) control actuators based on the temperature in the zone they serve. There is limited consideration of the effect on ambient temperature of such actions. This work proposes a hierarchical directed artificial neural network which optimises ambient temperature for open plan areas. The approach uses a multi-phase Artificial Neural Network (ANN). Two architectural components are introduced an Agent ANN (A-ANN) and a Coordinating ANN (C-ANN). The Agent ANNs (A-ANN) are deployed to provide temperature control at the extremities of the open plan area. The A-ANN operates with a degree of autonomy. A Coordinating ANN (C-ANN) considers the optimal ambient temperature of the entire open plan area and influences the decisions of individual A-ANNs in order to achieve a collectively balanced temperature. Results are presented which baseline the operation of A-ANN instances in varying environmental conditions.