Predication control for indoor temperature time-delay using Elman neural network in variable air volume system

Aiming at the prediction control for indoor temperature time-delay in variable air volume (VAV) air conditioning system, this paper presents an indoor temperature prediction control method based on Elman neural network multi-step prediction model. Firstly, this paper introduces basic control principles of pressure-dependent and pressure-independent VAV terminal through comparable analysis and points out significance of indoor temperature prediction control based on pressure-dependent VAV terminal. Then, Elman neural network multi-step prediction model and corresponding indoor temperature prediction control method for pressure-dependent VAV terminal are proposed based on the fundamental principle of periodic prediction control for time-delay system. Finally, the effect of proposed prediction control method is validated by the experimental study according to the test data of supply air volume regulating process, provided that the supply air volume control loop adopts constant static pressure control method. Experimental results indicate the proposed indoor temperature prediction control method based on pressure-dependent VAV terminal could change the conventional regulating mode of the VAV air conditioning system, which will be benefit for improving the control stability of indoor temperature control loop and other corresponding control loops. (C) 2017 Published by Elsevier B.V.