Neural control of non-linear HVAC plant

The unsatisfactory performance of heating, ventilating and air-conditioning (HVAC) control systems is frequently due to the inability of conventional controllers to deal with nonlinearities and to adapt to long-term changes in the behaviour of the plant. A neural control scheme is proposed which is capable of compensating for plant nonlinearities, and adapting online to degradation in the plant, but avoids the instability problems that can arise when neural networks are introduced into the feedback loop. Results are presented which have been obtained from a flow-controlled heating coil, on a full-size air conditioning plant at the UK Building Research Establishment. The neural control scheme is shown to produce more consistent control than a conventional PI controller.<>