Renewable energy has long been seen as a solution to climate change. However, they introduce uncertainties in power generation as their outputs are greatly affected by weather conditions and hard to be predicted. This makes balancing the power generation and power demand becomes difficult. A more accurate load forecasting model is needed to alleviate the power system stability problem stemming from renewable integration. In this project, we proposed a 24-Parallel-Hybrid Neural Network-Robust Linear Regression ensemble approach in 24-hours ahead point load forecasting. 7-year data taken from ISO New England Inc. were used in this study. Comparing with the best commonly used approach (24-Parallel Neural Networks), the proposed ensemble approach could reduce the mean absolute percentage error by 0.5239% and improve the accuracy by 20.45%. It also demonstrated superior hourly performance in all 24 hours over the single predictor approach and 24 parallel predictors approach.
COVID-19 virus appearing at the end of the year 2019 has caused serious damage worldwide with more than 20 million infected cases and 740,000 deaths. To avoid the virus spread, quarantine has been adopted by many countries and regions as an effective method. Current quarantine policies inherit shortfall and need further improvement in two aspects. Firstly, current quarantine scheme adopts strict restrictive measures, which primarily slows down the virus transmission but meanwhile brings huge negative effects of economic loss. Secondly, the reliabilities of current quarantine methods need further improvement. Quarantine without reliable localization potentially provides loophole which results in spreading of COVID-19. To address the captioned shortcomings, a new directive namely Optimized Dynamic Geofencing (ODG) is proposed. ODG replaces the traditional narrow quarantine confinement in a small room/area by a dynamic geofencing area for quarantine. ODG trades off between COVID-19 control and negative consequences. The developed ODG provides a new directive for COVID-19 control and prevention which saves lives and meanwhile reduces economic loss.
Energy is considered the topmost problem humanity will be facing for the next 50 years. Currently, over 800 million people have no access to electricity and due to the ever-growing population, electricity demand is projected to increase. Environmental concerns and energy security issues has also initiated rapid transition from fossil fuels to renewable energy sources. However, these renewable energy sources are affected by change in weather conditions leading to variation in amount of electricity generated by these systems. These challenges can be addressed by use of an energy storage device such as redox flow batteries (RFBs) to provide off-peak/peak power management. This project focuses on developing a competitive energy storage technology by investigating sustainable biomimetic organic materials to be deployed as redox active materials in RFBs. Organic molecules are synthetically tuneable allowing molecules to be tailored to have a combination of all the required properties essential for advancing this competitive technology. In this work, the organic RFB technology is advanced through design and modification of abundant, cheap, environmental-friendly and nature derived organic materials making them ideal redox active material for RFBs. These facilitates achievement of a safe, low-cost and sustainable energy storage system.
Space cooling consumes significant amounts of energy and is a major driver of peak electricity demand in HK buildings. Passive radiative cooling requires no electricity input and operates without refrigerants, making it a promising solution for smart-green buildings. Generally, passive radiative cooling is the process of heat removal from a sky-facing surface to the universe, where the temperature is -270 ⁰C, through radiation, cooling the surface below ambient temperature. It is a natural phenomenon happening everywhere, inspiring many applications, transforming our lifestyles (e.g. frost forms on surfaces at night even though the ambient temperature remains above freezing point). This project aims to develop a low-cost high performance passive radiative cooling paint (PRCP) that can be practically used as a cooling source in buildings for lowering energy consumption of traditional air-conditioning systems and reducing environmental pollution. Our preliminary results show that the proposed PRCP can reduce the indoor air temperature of a model house by ~2⁰C, providing an energy-saving of ~10% of traditional air-conditioning systems. Notably, the overall cost is only ~HK$15 per m2, leading to have a payback period of less than a few years. Overall, the success of this project will inspire future architectural and engineering design to reach smart-green buildings.
Despite the leaps of advancement in harvesting solar energy, some bottlenecks still limit them from becoming truly sustainable. Two such bottlenecks that need to be overcome are the intermittent energy supply, and effective thermal management of these solar systems. Phase change materials (PCMs) offer a promising solution to both these bottlenecks as they allow for an effective energy storage medium and can serve as excellent thermal management systems. The PCMs employ phases transition (normally solid-liquid) to store large amounts of energy and which can then be extracted when the material solidifies. However, the inherently low thermal conductivity of PCMs restricts their applications as an effective medium for energy storage. Incorporation of thermally conductive nanofillers has been employed for the enhancement of PCM performance but at the expense of high filler loadings and loss in storage capacity. We present the use of hybrid nanofillers with different aspect ratios and their synergistic approach for thermal enhancement of PCMs. Along with a novel system comprising of these enhanced PCMs for passive cooling of solar cells and simultaneous energy production via an Organic Rankine cycle.
COVID-19 has emerged as a severe global epidemic with high morbidity and mortality, while no effective drug targets it. Drug repurposing is a promising drug discovery method by minimizing the time and cost compared to new drug discovery and traditional randomized clinical trials. Big data-driven network analytics provide a novel way for drug repurposing, but currently, no such public available graph tailor-made for COVID-19. Here, we build a comprehensive COVID-19 knowledge graph by incorporating 25 data sources, including information on drugs, genes, viruses, proteins, diseases, and symptoms, which are notable predictors of clinical efficacy. By utilizing the COVID-19 knowledge graph and network-based methodologies, we can rapidly identify candidate repurposable drugs and drug combinations by discovering complex intrinsic linkages.
Fully utilizing wind energy as energy source has been a long challenge in years. Before building wind farms constructed in the body of water, farm site has to be selected through external information, such as weather data. In addition, the instability and intermittent of wind energy is an important factor restricting the predicting and selecting the wind field and energy field, and hence potentially the wind farm building site. In recent years, the various predictive mechanisms and algorithms has been developed, but on short-term wind power forecast. In our project, we plan to develop models to mimic the time-dependent wind field and energy field and study the applicability of using AI technology deep learning methods for predicting the wind field and energy field offshore. And display the visual prediction results on the website.
Defect inspection is one of the most important tasks in industry manufacture. However, only flat products currently can be inspected by robotic camera system. My project wants to address this problem in terms of free-form shiny objects. In order to deal with this kind of shiny objects, we adopt a robotic system, consisting of a high-resolution line scan camera and a co-axis lighting. In order to meet the condition of the system, we need to divide the free-form object into different flat patches. At first, we will sample points from the CAD file and filter out some useless points. Next, we will adopt K-means, an unsupervised learning algorithm, to finish segmentation task. And then we can plan the path in different regions. In addition, we use a RGB-D camera with a novel registration method to localize the part in robot’s frame. Experiment results show that all the defects on the shiny part can be captured. With the gray scale images, we can use canny algorithm in Opencv to detect edges and then reconstruct the defects to original CAD models.
Traveling salesman problem has a wide range of applications in the fields of transportation, circuit board design, and logistics and distribution. Scholars at home and abroad have conducted a lot of research on it. With the rise of the e-commerce industry, how to properly arrange logistics and distribution has become one of the important issues that need to be solved urgently. Bat Algorithm (BA) is a meta-heuristic optimisation algorithm, which proposed by Yang, X. S. (2010), based on the echo localisation behaviour of micro-bats through using different pulse emissivity and loudness. It is an iterative optimisation algorithm like other intelligent optimisation algorithms. The bat algorithm has the advantages of simple implementation and few parameters and has become a research hotpot in recent years. However, there are still a few applications in TSP. Therefore, this project will improve the bat algorithm to solve the optimisation problem of symmetric traveling salesman problems and compare it with other traditional heuristic algorithms to eliminate the local optimisation generated during the iterative calculation.