Search

This paper reports a new knowledge-driven engineered feature learning approach in response to the Global AI Challenge for Building E&M Facilities held by the Electrical and Mechanical Service Department (EMSD) of the Hong Kong SAR. The results were awarded with a Grand Prize by the competition organizer. A dynamically engineered multi-modal feature learning (DEMMFL) method is proposed for predicting the cooling load of two office buildings. The DEMMFL model is estimated with the Lasso-ridge regression and compared with other well-known methods such as the Lasso. The novel approach applies control system knowledge to engineer useful features and explore load patterns for multi-mode modeling. Deep learning methods including LSTM, GRU, and AutoGluon are implemented for automated machine learning and tested in parallel to compare the performance of the proposed model with existing methods. The proposed model is demonstrated to predict long-term cooling load most accurately using engineered features from weather information only. • A new dynamic feature engineering method for office building cooling load prediction. • Slow and fast dynamic features engineered from available weather data only. • Control-knowledge based feature to learn on-hours dynamic responses. • Multi-modal approach with eight day-type features to learn occupancy loads. • Deep recurrent learning and AutoML methods compared to the proposed method. [ABSTRACT FROM AUTHOR]

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 have been developed from either forward or inverse modeling approaches. However, these models usually require extensive computer resources and lengthy computation. This paper discusses the use of the multi-layer perceptron (MLP) model, one of the artificial neural network (ANN) models widely adopted in engineering applications, to estimate the cooling load of a building. The training samples used include weather data obtained from the Hong Kong Observatory and building-related data acquired from an existing prestigious commercial building in Hong Kong that houses a mega complex and operates 24h a day. The paper also discusses the practical difficulties encountered in acquiring building-related data. In contrast to other studies that use ANN models to predict building cooling load, this paper includes the building occupancy rate as one of the input parameters used to determine building cooling load. The results demonstrate that the building occupancy rate plays a critical role in building cooling load prediction and significantly improves predictive accuracy. [Copyright &y& Elsevier]

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 paper discusses the use of data-driven intelligent approaches, a probabilistic entropy-based neural (PENN) model 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 artificial neural network (ANN) models, most of which employ a fixed schedule or historic load data to represent building occupancy in simulating building cooling load, this paper introduces two input parameters, dynamic occupancy area and rate and uses it to mimic building cooling load. The training samples used include weather data obtained from the Hong Kong Observatory and building-related data acquired from an existing grade A mega office buildings in Hong Kong with tenants including many multi-national financial companies that require 24-h 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 performance of simulation results 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. [Copyright &y& Elsevier]

Driven by advances in smart grid technologies, building prosumers have closer interaction in sharing energy and information towards net-zero energy communities with energy flexibility and smart readiness. Many studies have focused on transactive energy systems, whereas limited attention was paid to coordinated prosumer management involving electric and electro-thermal energy storages. In this paper, a hierarchical energy management system was developed for a commercial neighbourhood, where distributed prosumers can share their generation and batteries, and manage cooling loads and thermal storages, in subtropical Hong Kong. Several operational scenarios were formulated considering building-level and neighbourhood-level controls, to explore the synergies through peer-to-peer (P2P) sharing and demand flexibility. Techno-economic-environmental impacts were analysed comprehensively from the perspectives of the network, the utility, and each participant. Results show that by integrating the demand-side flexibility, the P2P system could use electricity more effectively and economically. The formulated P2P operation strategy utilising neighbourhood-level control minimised the operational cost by 24.6% and net equivalent CO 2 emission by 7.1%, compared to the traditional peer-to-grid operation. The profit for each participant can be allocated by regulating the compensating price in the dynamic internal trading. The study provides prosumers energy-flexible and grid-friendly operation to facilitate the sustainability of the smart grid. • Hierarchical energy management for active prosumers with hybrid energy storages. • Performance comparison under individual and coordinated demand-side controls. • Enhancing local self-sufficiency by 22.8% via the power and information sharing. • Analysis of the participants' cost savings with a dynamic internal trading model. • Guidelines for system planners and operators to promote low-carbon communities. [ABSTRACT FROM AUTHOR]

Abstract: In Hong Kong, conventional air-conditioning (AC) systems cause a high energy consumption and poor indoor thermal comfort. This paper investigated the AC load profiles of three typical local commercial buildings, i.e. office, hotel and retail, and examined the energy consumption of liquid desiccant cooling system (LDCS) for these buildings. Simulation results showed that the cooling load contributed to more than 98% of total annual load, with high latent load proportion from March to September. The high reheating ratio, i.e. from 60% to 90%, is found to be the main reason for the poor thermal comfort. Therefore, the LDCS which could handle sensible and latent loads separately was proposed. However, the electricity-driven LDCS was found not suitable for local commercial buildings. significant energy saving could be achieved by the solar-assisted LDCS system with thermal storage, when the installation area for solar collectors is larger than the minimum required one, i.e. 9.7m2/(kWm−2) for office and 12.7m2/(kWm−2) for hotel buildings. Energy saving of different system types with different installation areas was also studied. The solar-assisted LDCS shows more energy saving in winter but less in summer. Results and conclusions can be useful for researchers and local engineers. [Copyright &y& Elsevier]

Abstract: The shading-type building-integrated photovoltaic (BIPV) claddings can act as power generators as well as external shading devices of a building, which reduce the energy consumption of the building. However, there is little information about energy impacts of different tilt angles of the shading-type BIPV claddings. By considering the typical meteorological conditions of Hong Kong, the energy performance of the shading-type BIPV claddings, in terms of the electricity generation and the cooling load reduction, is analyzed in this paper. The optimum tilt angle of PV modules for maximum electricity generation is found to be 20° instead of local latitude. Combining electricity generation and cooling load reduction, it can be concluded that the optimum tilt angles for the first type of the shading-type BIPV claddings vary from 30° to 50°, while the optimum tilt angle for the second type is 0°. [Copyright &y& Elsevier]

• Building Bio-Climatic Charts are proposed for different climate change scenarios. • Time-varying significance of nine design parameters in this century are provided. • Effectiveness of airtightness will increase up to 329% by the end of this century. • Cooling potentials of natural ventilation will significantly decrease up to 69%. • A holistic passive design can reduce the building annual cooling load by 56.7%. The application of passive design strategies is crucial at the early architectural design stage for building energy use minimization. However, the time-varying effectiveness of passive design strategies in responding to future climate change in hot and humid climates are rather limited in the literature. This paper aims to examine the dynamic effectiveness of passive design strategies for residential buildings in Hong Kong under the context of future climate change. Using the newly developed hourly weather data and adaptive comfort standard model, the dynamic effectiveness of viable passive design strategies for residential buildings are evaluated over time in the 21st century by plotting Givoni building bio-climatic charts (BBCC) and simulation-based sensitivity analyses in a validated EnergyPlus model. Results show that solar protection strategies are still the highly sensitive strategies for building energy performance and the effectiveness of external windows' airtightness is expected to increase up to 329% by the end of this century, whereas the cooling potential of ventilation utilization will significantly decrease over time. When the different combination of sensitive passive design parameters is implemented onto the baseline residential building model for different climate scenarios, the annual and peak cooling load can be reduced up to 56.7% and 64.5%, respectively. [ABSTRACT FROM AUTHOR]

• The climate change is identified by the ground measurement of HK over the 37 years. • The building energy needs can be influenced by the climate change. • The TMY is suggested to be updated every 10 years or less for the climate change. • 10 to 12 years of the latest data can be proper for TMY in the changing climate. The weather data is essential to simulation-based building energy studies. The weather data should render the features of the changing climate. The typical meteorological year (TMY) developments using the long-term climate data, however, may weaken the impact of the recent climate change and the corresponding building energy responses. This paper studies the changes of the climate data (by ground measurements) and their impacts on the simulation results of the subtropical Hong Kong over 37 years. The climate changes for solar radiation, temperature and wind speed were summarised, and their daily total, average, extreme, and/or deviation values were presented. The typical weather data developed by the ground measurements of different times and period lengths (within the 37 years) determined the cooling load of a representative building in the Hong Kong, and the outcomes were compared to that of the yearly data. The period (length) of the raw measurements and the update frequency of TMY were optimized for the building energy simulations in the changing climate. [ABSTRACT FROM AUTHOR]