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]

Unmanned aerial vehicle (UAV)-based commercial services, exemplified by drone delivery, have captured wide interest in tech companies, entrepreneurs, and policymakers. Structured route-based UAV operations have been implemented for traffic management of UAVs in support of commercial delivery services in cities. Yet, its essence, multi-path planning with constraints is not well solved in the existing literature. Centralized planning might result in inefficiencies and unfairness in the allocation of precious urban airspace to individual routes. This paper describes a novel distributed route planning method to support UAV operations in a high-density urban environment. The method allows each origin–destination (OD) pair to compete against other OD pairs for an optimized route (e.g. shortest distance), coordinated by a system-level evaluation, leading to a network design that maximizes the performance of not only the individual routes but also the entire system. The core concept is the introduction of congestion pricing, a soft constraint to coordinate the allocation of airspace. The method is tested in standard 2D scenarios and compared with other state-of-the-art methods. The results show that (1) the method is able to generate routes with short individual distances as well as occupying the least airspace by the route network; (2) in some complex scenarios, the method is able to find a solution in a short period of time while other state-of-the-art method fails. The method has also been applied to a real urban environment (Mong Kok in Hong Kong) to demonstrate its capability. • Presented a multi-path method to design air route networks for urban drone delivery. • Introduced congestion cost as a flexible constraint for airspace allocation. • Employed a distributed approach for resolving conflicts between origin-destination pairs. • Outperformed other leading methods in balancing efficiency and computation speed. • Validated the proposed system through tests in real-world settings. [ABSTRACT FROM AUTHOR]