인공신경망 입력 변수에 따른 송풍기 풍량 예측모델 개발 및 평가

Abstract

A model for predicting the supply air flow rate in the fan, which plays a important role in HVAC system, is to be developed using artificial neural network. A predictive model has been developed to study with the Levenbarg-Marquardt algorithm through 8760 sets of one-hour resolution. The model of three cases was constructed according to the combination of the input variables constituting the input data of the neural network, and the accuracy of each case model was evaluated through statistical approach using Coefficient of Variation of Root Mean Square Error and the best performance model was determined. The input parameters includes flow rate, pressure, fan power consumption, outdoor air temperature, outdoor air humidity, supply air temperature and zone air temperature. The suggested model including seven input data shows the best performance. The results show that the developed model can provide results sufficiently accurate.