Search NASA⌕ Search

Engineering topics

Li, Baojie

Publications and source records attributed to Li, Baojie.

10 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

A unified large language model–based framework for heterogeneous PV image diagnosis

With advances in imaging technologies, modern photovoltaic (PV) systems generate large volumes of heterogeneous image data, including visible, electroluminescence (EL), and infrared (IR) images. Existing PV image analysis models, particularly deep learning approaches, are typically task-specific and lack cross-modality generalization. To address this limitation, this paper proposes an open-source large language model (LLM)–based unified framework for heterogeneous PV image diagnostics. Through task-aware diagnostic prompting, the framework enables analysis of visible, EL, and IR images within a single pipeline, supporting both zero-shot and few-shot inference and binary and multiclass classification. It is compatible with state-of-the-art multimodal LLMs, including ChatGPT, Gemini, Claude, Qwen, and CLIP. The framework is evaluated on PV module condition classification (clean, soiling, snow, hail, and bird droppings) using visible images, cell crack detection using EL images, and hotspot detection using IR images. GPT-5.1 in few-shot mode achieves the best performance, with classification accuracy exceeding 97.3%. Open-source models such as Qwen and CLIP also deliver competitive results on visible images (around 90% accuracy), though their performance is more limited on EL and IR modalities. On the full ELPV dataset, the framework achieves 83.5% zero-shot accuracy, within 2.8% of the supervised CNN baseline, confirming scalability to larger benchmarks. Practical aspects such as reproducibility, response latency, and confidence estimation are systematically analyzed. The framework operates across PV image modalities without modality- or task-specific training, making it well suited as a rapid pre-screening tool to support downstream detailed diagnostics. A benchmark dataset of diverse labeled PV images is also released.

Li, Baojie↗
DOE OSTI2026

Machine learning-based bias-corrected future projections of ozone concentrations from a chemistry-climate model

Reliable projection of future near-surface ozone is crucial for air quality management and health risk assessment. However, potential biases in spatial distribution, magnitude and trends in ozone concentrations simulated by global chemistry-climate models limit their applicability in regional-scale evaluations. In this study, LightGBM, a machine learning (ML) algorithm is applied to correct biases in CESM2-simulated ozone concentrations over China, the United States and Europe and calibrate future ozone projections under two diverse Shared Socioeconomic Pathways (SSP1-2.6 and SSP5-8.5) scenarios from 2020 to 2060. The ML-based correction significantly improves the spatial distribution and reduces the model bias by 40%–60%. It also reverses the potentially incorrect trend of ozone change under SSP1-2.6 in eastern China. When applying ML-based bias correction to CESM2 future projections, warm season mean ozone concentrations decrease across China, the United States, and Europe by –13.5, –17.9, and –13.7 µg/m³, respectively, between 2020 and 2060 in SSP1-2.6, while they increase by 9.4, 2.0, and 5.2 µg/m³ in SSP5-8.5. Decomposition analysis show that changes in anthropogenic emissions dominate future ozone changes in both scenarios, while strong climate penalty from ozone changes occurs in polluted eastern China and climate benefit is found in western China, the United States and Europe under SSP5-8.5. These findings demonstrate the value of combining ML with chemistry-climate models to produce more accurate air quality projections, thereby informing more effective and region-specific environmental protection strategies.

Chemistry Model↗
DOE OSTI2025

Exploring Capability of Multimodal Foundation Model for Image-based Fault Detection of Photovoltaic Modules

Multimodal Foundation Model (MFM), like ChatGPT and Gemini, have emerged as powerful tools for their exceptional natural language processing capabilities and their emerging potential in image analysis. This paper investigates the application of MFMs for photovoltaic (PV) fault detection through image analysis, focusing on ChatGPT 4.0 and Gemini 1.5 Pro. Three types of PV images and the corresponding common PV faults are detected: bird droppings using visible images, cell cracks via electroluminescence (EL) images, and hotspots using infrared (IR) images. Among the two models, Gemini 1.5 Pro demonstrated superior performance, achieving near-perfect results with an average F1 score of 0.97, consistently outperforming ChatGPT 4.0 in accuracy and reliability. Unlike traditional machine learning (ML) models, MFMs can operate in a zero shot manner that does not require additional training by the user, and the input images are not limited by size, angle, scope, or PV technology. The strong adaptability and user-friendliness make MFM a promising tool for analyzing PV images and advancing health monitoring for PV modules.

Li, Baojie↗
DOE OSTI2024

Power modeling of degraded PV systems: Case studies using a dynamically updated physical model (PV-Pro)

Power modeling, widely applied for health monitoring and power prediction, is crucial for the efficiency and reliability of Photovoltaic (PV) systems. The most common approach for power modeling uses a physical equivalent circuit model, with the core challenge being the estimation of model parameters. Traditional parameter estimation either relies on datasheet information, which does not reflect the system's current health status, especially for degraded PV systems, or requires additional I-V characterization, which is generally unavailable for large-scale PV systems. Thus, we build upon our previously developed tool, PV-Pro (originally proposed for degradation analysis), to enhance its application for power modeling of degraded PV systems. PV-Pro extracts model parameters from production data without requiring I-V characterization. This dynamic model, periodically updated, can closely capture the actual degradation status, enabling precise power modeling. PV-Pro is compared with popular power modeling techniques, including persistence, nominal physical, and various machine learning models. The results indicate that PV-Pro achieves outstanding power modeling performance, with an average nMAE of 1.4 % across four field-degraded PV systems, reducing error by 17.6 % compared to the best alternative technique. Furthermore, PV-Pro demonstrates robustness across different seasons and severities of degradation. The tool is available as a Python package at https://github.com/DuraMAT/pvpro.

14 SOLAR ENERGY↗
DOE OSTI2024

Global source apportionment of aerosols into major emission regions and sectors over 1850–2017

Anthropogenic emissions of aerosols and precursor gases have changed significantly in the past few decades around the world. In this study, the Explicit Aerosol Source Tagging (EAST) system is merged into the Energy Exascale Earth System Model version 1 (E3SMv1) to quantify the variations in anthropogenic aerosol concentrations, source contributions, and their subsequent radiative impact in four major emission regions across the globe during 1850–1980, 1980–2010, and 2010–2017. In North America and Europe, changes in anthropogenic PM 2.5 were mainly caused by changes in emissions from local energy and industrial sectors. The local industrial sector caused the largest increase in PM 2.5 in East Asia during 1980–2010 and decrease during 2010–2017. In South Asia, the increase in energy-related emissions dominated the rise in PM 2.5 levels during 1980–2017. During 1850–1980, the increases in emissions from North America contributed to the increase in the European PM 2.5 burden by 1.7 mg m -2 and the sources from the Europe were also responsible for the PM 2.5 burden increase in East Asia and South Asia by about 1.0 mg m -2 . During 1980–2010, East Asia contributed to an increase of 0.4–0.6 mg m -2 in the PM 2.5 burden in North America and Europe, while South Asia contributed about 0.3 mg m -2 . During 2010–2017, the contributions from East Asia to the PM 2.5 burdens in the North America, Europe, and South Asia declined by 0.3–0.6 mg m -2 due to the clean air actions in China, while the contributions from South Asia still increased due to the continuous increase in emissions in South Asia. The historical changes in aerosols had an impact on effective radiative forcing through aerosol–radiation interactions (ERF ari ). During 1980–2010, a decline in North American aerosols resulted in a positive ERF ari change (warming effect) in Europe and a decline in aerosols in Europe caused a warming effect in Russia and northern China. The changes in ERF ari from the increase and decrease in aerosols in China during 1980–2010 and 2010–2017, respectively, are comparable in magnitude. The continuous aerosol increases in South Asia from 1980 to 2017 resulted in negative ERF ari (cooling) changes in South Asia, Southeast Asia, and southern China.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Climate responses in China to domestic and foreign aerosol changes due to clean air actions during 2013–2019

In recent years, to improve air quality, significant efforts have been made to reduce regional aerosols including China, Europe and North America, which have potential impacts on climate. In this study, fast and slow climate responses in China to changes in domestic and foreign anthropogenic emissions of aerosols and precursors from 2013 to 2019 are investigated using the Community Earth System Model version 1 (CESM1). Aerosol emissions changes, especially reductions in China, North America and Europe, during 2013–2019 resulted in surface air temperature increases across the Northern Hemisphere. Global aerosol changes induce a regional warming of 0.2 °C in China, equally contributed by domestic and foreign emissions changes. China’s domestic emission reductions lead to a regional average temperature rise of 0.1°C, primarily driven by rapid atmospheric adjustments over eastern China. Foreign aerosol changes, particularly reductions in North America and Europe, also contributed to a 0.1°C warming in China through slow oceanic processes. This warming in China induced by foreign aerosol changes is due to a teleconnection between the aerosol-induced anomalous regional warming in Eastern U.S.-North Atlantic Ocean-Europe and the downstream East Asian climate through anomalous wave trains propagation. The comparable influence of domestic and foreign aerosol changes on climate in China underscores the importance of international collaboration in climate mitigation endeavors.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Source attribution of near-surface ozone trends in the United States during 1995–2019

Emissions of ozone (O 3 ) precursors in the United States have decreased in recent decades, and near-surface O 3 concentrations showed a significant decrease in summer but an increase in winter. In this study, an O 3 source tagging technique is utilized in a chemistry–climate model to investigate the source contributions to O 3 mixing ratios in the US from various emitting sectors and regions of nitrogen oxides (NO x ) and reactive carbon species during 1995–2019. We show that domestic emission reductions from energy and surface transportation are primarily responsible for the decrease in summertime O 3 during 1995–2019. However, in winter, the emission control also weakens the NO x titration process, resulting in considerable increases in O 3 levels from natural sources. Additionally, increases in aviation and shipping emissions and transpacific transport of O 3 from Asia largely contribute to the winter O 3 increase. We also found that changes in large-scale circulation favoring O 3 transport from the upper atmosphere and foreign transport from Asia also explain 15 % of the increase in the US near-surface O 3 levels in winter.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Rapid increase in tropospheric ozone over Southeast Asia attributed to changes in precursor emission source regions and sectors

Observations indicate that tropospheric ozone (O 3 ) concentrations over Southeast Asia have been increasing rapidly since the 1990s. Here, we quantify source contributions from geographical regions and emission sectors of the two distinct types of O 3 precursors, i.e., nitrogen oxides (NO x ) and volatile organic compounds (VOCs), to the increase in tropospheric O 3 in Southeast Asia during 1990–2019 using an O 3 source tagging technique implemented in a global chemistry-climate model. In this work, the results show that although local anthropogenic emission of NO x in Southeast Asia only contributes 18% of the annual averaged near-surface O 3 concentration, the increase in local NO x emission dominates the increasing trend of O 3 concentration in Southeast Asia, accounting for 107% of the regional averaged trend of 1.07 ppb decade -1 . Increases in NO x emissions from East Asia and South Asia explain 29% of the increasing trend, but 9% is offset by the emission reduction in North America. Ground transportation is responsible for 79% of the rapid O 3 increase, followed by 39% contribution from international shipping. Because an increase in anthropogenic NO x emissions enhances the O 3 production efficiency by VOCs, the increase in near-surface O 3 concentrations in Southeast Asia is thereby largely contributed by methane and biogenic VOCs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Determining circuit model parameters from operation data for PV system degradation analysis: $\mathrm{PVPRO}$

Physics-based circuit parameters like series and shunt resistance are essential to provide insights into the degradation status of photovoltaic (PV) arrays. However, calculating these parameters typically requires a full current-voltage characteristic (I-V curve), the acquisition of which involves specific measurement devices and costly methods. Thus, I-V curves of the PV system level are often not available. Here this paper proposes a methodology (PVPRO) to estimate these I-V curve parameters using only operation (string-level DC voltage and current) and weather data (irradiance and temperature). PVPRO first performs multi-stage data pre-processing to remove noisy data. Next, the time-series DC data are used to fit an equivalent circuit single-diode model (SDM) to estimate the circuit parameters by minimizing the differences between the measured and estimated values. In this way, the time evolutions of the SDM parameters are obtained. We evaluate PVPRO on synthetic datasets and find an excellent estimation of both SDM and the key I-V parameters (e.g., open-circuit voltage, short-circuit current, maximum power, etc.) with an average relative error of 0.55%. The performance, especially the extracted degradation rate of parameters, is robust to various measurement noises and the presence of faults. In addition, PVPRO is applied to a 271 kW PV field system. The relative error between the real and estimated operation voltage and current is less than 1%, suggesting that degradation trends are well captured. PVPRO represents a promising open-source tool to extract the time-series degradation trends of key PV parameters from routine operation data.

14 SOLAR ENERGY↗