Search NASA⌕ Search

Engineering topics

You, Fengqi

Publications and source records attributed to You, Fengqi.

Physics-informed machine learning for building performance simulation-A review of a nascent field

Building performance simulation (BPS) is critical for understanding building dynamics and behavior, analyzing the performance of the built environment, optimizing energy efficiency, improving demand flexibility, and enhancing building resilience. However, conducting BPS is not trivial. Traditional BPS relies on accurate building energy models, which are primarily physics-based and heavily dependent on detailed building information, expert knowledge, and case-by-case model calibrations, significantly limiting their scalability. With the development of sensing technology and the increased availability of data, there is growing attention and interest in data-driven BPS. However, purely data-driven models often suffer from limited generalization ability and a lack of physical consistency, resulting in poor performance in real-world applications. To address these limitations, recent studies have begun integrating physics priors into data-driven models, a methodology known as physics-informed machine learning (PIML). PIML is an emerging field where its definitions, methodologies, evaluation criteria, application scenarios, and future directions remain open. To bridge those gaps, this study systematically reviews the state-of-the-art PIML for BPS, offering a comprehensive definition of PIML and comparing it to traditional BPS approaches regarding data requirements, modeling effort, performance, and computational cost. We also summarize the commonly used methodologies, validation approaches, application domains, available data sources, open-source packages, and testbeds. In addition, this study provides a general guideline for selecting appropriate PIML models based on BPS applications. Finally, this study identifies key challenges and outlines future research directions, providing a solid foundation and valuable insights to advance R&D of PIML in BPS.

Jiang, Zixin↗

Quantum computing based hybrid solution strategies for large-scale discrete-continuous optimization problems

Technologies for a quantum/classical hybrid approach to solving optimization problems is disclosed. In the illustrative embodiment, an optimization problem is decomposed into two sub-problems. The first sub-problem is solved on a classical computer, and a result from the first sub-problem is provided to a quantum computer. The quantum computer then solves the second sub-problem based on the result of the first sub-problem from the classical computer. The quantum computer can then provide a result to the classical computer to re-solve the first problem. The iterative calculation is continued until an end condition is met.

You, Fengqi↗

Quantum computing based hybrid deep learning for fault diagnosis in electrical power systems

Quantum computing (QC) and deep learning have shown promise of supporting transformative advances and have recently gained popularity in a wide range of areas. Here, this paper proposes a hybrid QC-based deep learning framework for fault diagnosis of electrical power systems that combine the feature extraction capabilities of conditional restricted Boltzmann machine with an efficient classification of deep networks. Computational challenges stemming from the complexities of such deep learning models are overcome by QC-based training methodologies that effectively leverage the complementary strengths of quantum assisted learning and classical training techniques. The proposed hybrid QC-based deep learning framework is tested on a simulated electrical power system with 30 buses and wide variations of substation and transmission line faults, to demonstrate the framework’s applicability, efficiency, and generalization capabilities. High computational efficiency is enjoyed by the proposed hybrid approach in terms of computational effort required and quality of diagnosis performance over classical training methods. In addition, superior and reliable fault diagnosis performance with faster response time is achieved over state-of-the-art pattern recognition methods based on artificial neural networks (ANN) and decision trees (DT).

42 ENGINEERING↗

Global carbon intensity of crude oil production

Producing, transporting, and refining crude oil into fuels such as gasoline and diesel accounts for ~15 to 40% of the “well-to-wheels” life-cycle greenhouse gas (GHG) emissions of transport fuels. Reducing emissions from petroleum production is of particular importance, as current transport fleets are almost entirely dependent on liquid petroleum products, and many uses of petroleum have limited prospects for near-term substitution (e.g., air travel). Better understanding of crude oil GHG emissions can help to quantify the benefits of alternative fuels and identify the most cost-effective opportunities for oil-sector emissions reductions. Yet, while regulations are beginning to address petroleum sector GHG emissions, and private investors are beginning to consider climate-related risk in oil investments, such efforts have generally struggled with methodological and data challenges. First, no single method exists for measuring the carbon intensity (CI) of oils. Second, there is a lack of comprehensive geographically rich datasets that would allow evaluation and monitoring of life-cycle emissions from oils. We have previously worked to address the first challenge by developing open-source oil-sector CI modeling tools [OPGEE, supplementary materials (SM) 1.1]. Furthermore, we address the second challenge by using these tools to model well-to-refinery CI of all major active oil fields globally—and to identify major drivers of these emissions.

02 PETROLEUM↗

Life-cycle Analysis of Bioproducts and Their Conventional Counterparts in GREET

To further expand upon the literature in this field and to develop a platform for bioproduct LCA, we developed LCA results for ten bioproducts produced either from algal glycerol or from corn stover-derived sugars. We used Argonne National Laboratory’s Greenhouse gases, Regulated Emissions, and Energy use in Transportation (GREETTM) model as the platform for this study. The data and calculations reported herein are available to GREET users in a bioproducts module included in the fall 2015 GREET release. This report documents our approach to this analysis and the results. In Chapter 2, we review the process we underwent to select the bioproducts for analysis based on market and technology readiness criteria. In Chapter 3, we review key parameters for production of the two feedstocks we considered: corn stover and algae. Given the lack of publicly available information about the production of bioproducts, which is caused in large part by the emerging nature of the industry, we developed Aspen Plus® simulations of the processes that could be used to produce each bioproduct. From these simulations, we extracted the energy and material flows of these processes, which were important inputs to the GREET bioproducts module. Chapter 4 provides the details of these Aspen Plus simulations. It is important to compare the LCA results for bioproducts to those for their petroleum counterparts. We therefore also developed material and energy flow data for conventional products based mostly on the literature. These data are described in Chapter 5 and are also included in the GREET bioproducts module. In Chapter 6, we present results from this analysis and examine areas for refinement and future research.

59 BASIC BIOLOGICAL SCIENCES↗

Optimal design of sustainable cellulosic biofuel supply chains: Multiobjective optimization coupled with life cycle assessment and input-output analysis

This article addresses the optimal design and planning of cellulosic ethanol supply chains under economic, environmental, and social objectives. The economic objective is measured by the total annualized cost, the environmental objective is measured by the life cycle greenhouse gas emissions, and the social objective is measured by the number of accrued local jobs. A multiobjective mixed-integer linear programming (mo-MILP) model is developed that accounts for major characteristics of cellulosic ethanol supply chains, including supply seasonality and geographical diversity, biomass degradation, feedstock density, diverse conversion pathways and byproducts, infrastructure compatibility, demand distribution, regional economy, and government incentives. Aspen Plus models for biorefineries with different feedstocks and conversion pathways are built to provide detailed techno-economic and emission analysis results for the mo-MILP model, which simultaneously predicts the optimal network design, facility location, technology selection, capital investment, production planning, inventory control, and logistics management decisions. The mo-MILP problem is solved with an e-constraint method; and the resulting Pareto-optimal curves reveal the tradeoff between the economic, environmental, and social dimensions of the sustainable biofuel supply chains. The proposed approach is illustrated through two case studies for the state of Illinois.

09 BIOMASS FUELS↗