An intelligent approach to access building occupancy for cooling load prediction

Abstract

Building cooling load prediction is one of the key factors in the success of energy-saving measures. Many computational models available in the industry today have been developed from either forward or inverse modeling approaches. However, most of these models require extensive computer resources and involve lengthy computation. This thesis discusses the use of the multi-layer perceptron (MLP) model - an artificial neural network (ANN) model widely adopted in engineering applications that offers unique characteristics such as adaptability, nonlinearity, and arbitrary function mapping ability - to predict the cooling load of a building. Although it is common knowledge that the presence and activity of building occupants have a significant impact on the required cooling load of buildings, practices currently adopted in modeling the presence and activity of people in buildings do not reflect the complexity of the impact occupants have on building cooling load. In contrast to previous ANN models, most of which employ a fixed schedule or historic load data to represent building occupancy in simulating building cooling load, this thesis introduces a stochastic model of occupants’ presence and activity and uses it as one of the input parameters to mimic building cooling load. The training samples used include weather data obtained from the Hong Kong Observatory and building-related data acquired from two existing grade A mega office buildings in Hong Kong with tenants including many multinational financial companies that require 24-hour air conditioning seven days a week. The dynamic changes that occur in the occupancy of these buildings therefore make it very difficult to forecast building cooling load by means of a fixed time schedule. The paper also discusses the practical difficulties and limitations encountered in obtaining building-related data, particularly those faced in acquiring dynamic data. The performance of simulation results for the two mega office buildings examined demonstrate that building occupancy data play a critical role in building cooling load prediction and that their use significantly improves the predictive accuracy of cooling load models. Keywords: Artificial Neural Network, Cooling Load, Building Occupancy.

Date of Award	4 Oct 2010
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Wai Ming LEE (Supervisor)