Autonomic Management of a Building’s Multi-HVAC System Start-Up

Most studies about the control, automation, optimization and supervision of building HVAC systems concentrate on steady-state regime, i.e., when equipment is already working at its setpoints. The originality current work consists proposing multi-HVAC from start-up until they reach setpoint, making transition to steady state-based strategies smooth. proposed approach works transient regime optimizing contradictory objectives, such as desired comfort energy costs, based “Autonomic Cycle Data Analysis Tasks” concept. In this case, autonomic cycle composed two data analysis tasks: one for determining if system going towards defined operational that not another task reconfiguring mode redirect it. first uses machine learning techniques build detection prediction models, second defines a reconfiguration model using multiobjective evolutionary algorithms. This proposal proven in real case study characterizes particular performance obtained experiments diverse situations impressive since there high level conformity setpoint deliver operation smoothly, avoiding overshooting other non-desirable transitional effects.