Research We Do

Our research spans SIX focus areas that accelerate an AI-enabled, secure, low-carbon, and people-centered energy transition:

Focus Area 1: Energy System Design for Green and Grid-friendly AI Data Centers

Data centers are no longer passive electricity users. With the rise of AI and high-performance computing, they have become continuous, high-density energy hubs that influence grid planning, emissions, and infrastructure resilience. Our research develops solutions that make data centers low-carbon, flexible, and grid-supportive rather than burdensome. We focus on three core directions:

1. Modeling AI Training and Inference Energy Footprints: We quantify and model the electricity, thermal, and carbon impacts of AI training, finetuning, and inference workloads. This includes integrating compute demand patterns into grid-aware scheduling, emissions accounting, and infrastructure planning to ensure AI growth aligns with both capacity and climate constraints.

2. Carbon-Aware and Net-Zero Operation: We design carbon-oriented scheduling and dispatch models that align data center loads with real-time and day-ahead grid signals (e.g., congestion, electricity prices, grid-level emissions). This includes renewable matching, carbon allocation mechanisms, and 24/7 clean energy integration.

3. Multi-Energy and Multi-Timescale Optimization: Data centers depend on electricity, cooling, thermal storage, and emerging technologies such as small modular reactors (SMRs) or district systems. We develop integrated frameworks that co-optimize power, HVAC, storage, DERs, and grid services across timescales to turn data centers into flexible energy assets.

4. Robust and Learning-Based Scheduling: To manage uncertain renewables, workloads, and grid conditions, we develop distributionally robust optimization, conformal prediction, and chance-constrained methods to ensure reliable, adaptive operation under variability.

Our goal in this topic area is to turn data centers into catalysts for advancing decarbonization, resilience, and intelligent energy management, rather than a strain on the energy infrastructure.

1. Yijie Yang, Dan Wang, Jian Shi, Chenye Wu, and Zhu Han, "Multi-timescale Optimization Scheduling for Green LLM Inference", under review, 2025.

2. Yijie Yang, Dan Wang, Jian Shi, Chenye Wu, and Zhu Han, "A Conformal Prediction-Based Chance-Constrained Programming Approach for 24/7 Carbon-Free Data Center Operation Scheduling", under review, 2025.

3. Yijie Yang, Dan Wang, Jian Shi, Chenye Wu, and Zhu Han, "Uncertainty-aware Day-ahead Datacenter Workload Planning with Load-following Small Modular Reactors", the 4th Workshop on Sustainable Computer Systems (HotCarbon'25).

4. Yijie Yang, Jian Shi, Dan Wang, Chenye Wu, Zhu Han, "Carbon-Aware Scheduling of Thermostatically Controlled Loads: A Bilevel DRCC Approach", in IEEE Transactions on Smart Grid, 2025.

5. Yijie Yang, Jian Shi, Dan Wang, Chenye Wu, and Zhu Han, "Net-Zero Scheduling of Multi-energy Building Energy Systems: A Learning-based Robust Optimization Approach with Statistical Guarantees", in IEEE Transactions on Sustainable Energy, 2024.

Focus Area 2: Harnessing Generative-AI for Energy Transition

Our research focuses on leveraging cutting-edge Gen-AI to drive the transition toward a sustainable, low-carbon future by addressing the intricate challenges within modern energy systems. Recognizing the global demand for resilient and efficient energy solutions, we explore two pioneering applications of Gen-AI that support the broader energy transformation, with a particular emphasis on Texas’s unique energy landscape.

AOI 1 involves the use of large language models (LLMs) to enhance energy load forecasting by incorporating unstructured socio-demographic information about energy consumers. This approach enables grid operators to improve the accuracy of demand predictions, especially among diverse demographic groups, to achieve better resource allocation and reduced risk of outages. For Texas, where the grid is frequently challenged by extreme demand fluctuations, this LLM-driven model provides a strategic advantage in maintaining grid stability and energy efficiency.

AOI 2 is on AI-based generative agents that simulate consumer energy behaviors at scale. These agents, trained with real-world human-sourced data, model the diverse consumption patterns of individuals and communities and provide a realistic foundation for designing interventions that encourage sustainable practices. By predicting how people might respond to various energy initiatives, these simulations support targeted outreach and policymaking efforts that foster low-carbon lifestyle adoption. This approach is particularly valuable in Texas, where values of individualism and energy independence are deeply embedded.

Select Publications

1. Weilong Chen, Xinran Zhang, Ling Zhu, Jian Shi, Zheng Chang, Zhu Han, and Yanru Zhang, "Socially-Aware Load Forecasting Utilizing Large Language Models", in IEEE Transactions on Industrial Informatics, 2025.

2. Dingwen Pan, Weilong Chen, Jian Shi, Chenye Wu, Dan Wang, Choong Seon Hong, and Zhu Han, "Bayesian Inference-Aided Large Language Model Agents in Infinitely Repeated Games: A Dynamic Network View", in IEEE Transactions on Network Science and Engineering, 2025.

3. Dingwen Pan, Weilong Chen, Jian Shi, Chenye Wu, Dan Wang, Choong Seon Hong, and Zhu Han, “Cooperation and Decision-Making of LLM Agents in Bayesian-Informed Infinitely Repeated Games”, 2025 59th Annual Conference on Information Sciences and Systems (CISS).

4. Weilong Chen, Xinru Liu, Xinran Zhang, Jian Shi, Han Yang, Zhu Han, and Yanru Zhang, "A Socio-Aware Diffusion Model for Residential Electricity Consumption Data Generation", in IEEE Transactions on Smart Grid, 2025.

Focus Area 3: Establishing Secure and Competitive Supply Chain of Critical Metals and Materials for Energy Sovereignty and Independence

Modern energy systems and national security hinge on stable access to critical metals and materials—from lithium and cobalt to rare earths and specialty alloys. As global competition intensifies, the ability to recover, redistribute, and domestically cycle these strategic resources is fundamental to national energy independence and industrial resilience.

To this end, we develop game-theoretic optimization frameworks that transform fragmented recycling, refining, and reclamation into coordinated, secure, and intelligence-driven supply chains. Our models integrate dynamic pricing, strategic cost-sharing, and adaptive policy responses. We align the interests of manufacturers, recyclers, and policymakers to ensure resource sovereignty, strengthen resilience against supply disruptions, and turn circularity into a strategic pillar of national security. We focus on four core directions:

1. Strategic Recovery Modeling: Developing hierarchical game models that capture both competition and collaboration between manufacturers and recyclers for securing domestic control over critical material flows.

2. Dynamic Pricing Under Market Volatility: Designing adaptive pricing mechanisms that respond to raw material fluctuations, policy shifts, and strategic demand, which ensures a stable supply/pricing without over-reliance on external markets.

3. Incentive and Cost-Sharing Mechanisms: Creating structured frameworks to distribute collection, logistics, and processing costs across stakeholders, enable durable resource utilization, and prevent breakdowns in recovery networks.

Our long-term vision is to transform critical material recovery and circulation into core national infrastructure. By anchoring secure supply chains and reducing reliance on foreign sources, our research aims to position the United States at the forefront of industrial resilience and energy leadership.

Select Publications

1. Chuyue Wang, Jian Shi, and Kailai Wang, "Modeling and Optimizing Competition and Collaboration in Dual-Channel E-Waste Supply Chains: A Game-Theoretic Approach", under 2nd round review, Scientific Review, 2025.

2. Chuyue Wang, Jian Shi, and Kailai Wang, "From E-Waste to Wildlife: Linking Electronic Device Recycling to Species Conservation", to be submitted to One Earth, 2025.

Focus Area 4: Deep Decarbonization of Power Systems Incorporating the Carbon Perspective

Electric power systems are the backbone of the global energy system and account for 25% of worldwide greenhouse gas (GHG) emissions. With the rapid trend of electrifying final energy use, a low-carbon electric power system will also enable the clean electrification of transportation, buildings, and industry sectors to support full economy-wide decarbonization.

Our research aims to establish a carbon-oriented framework to facilitate power systems' effective decarbonization and deliver reliable, secure, clean, and cost-effective energy services.

Motivated by the tremendous role decarbonization will be playing in the future energy system, we take a different route than the traditional literature, which is primarily scoped from the electricity and energy perspective concentrating on one piece of technological solutions. Instead, we focus on developing a carbon perspective to analyze and exploit the "life cycle" of carbon emissions embodied in the electric energy value chain.

As shown in the figure above, we divide the carbon flows through the multi-energy power system into three stages: i) carbon allowance initiation and allocation, ii) carbon allowance exchange and pricing, and iii) carbon allowance circulation, based on which interdisciplinary research, e.g., optimization, game theory, and machine learning, can be performed to collectively accelerate the carbon neutral transition of the energy sector.

Select Publications

1. Jian Shi, Joseph Powell, Dirk Smit, Zirui Tong, Zhu Han, and Kaushik Rajashekara, "Carbon Sovereignty and the Balance Sheet Revolution: A Triadic Framework for Security-Driven Energy System Decarbonization", under review, 2025.

2. Jian Shi, Zirui Tong, Dan Wang, Chenye Wu, Zhu Han, Yijie Yang, Kaushik Rajashekara, "A Multi-Layered Framework for Advancing Rapid and Balanced Electric Power System Decarbonization", under review, PNAS Nexus, 2025.

3. Jian Shi, Bo Wang, Tek Tjing Lie, Yuhan Zheng, Mercy Chelangat, Chenye Wu, Luo Xu, Guadalupe Gonzalez, Annika Moman, "Engineered Confidence: Aligning Practice and Ambition in the Clean Energy Transition", IEEE Energy Sustainability Magazine, 2025.

4. Jian Shi, Dan Wang, Chenye Wu, and Zhu Han, "Deep Decarbonization of Multi-Energy Systems: A Carbon-Oriented Framework with Cross-Disciplinary Technologies", in IEEE Carbon Neutrality Newsletter, 2022.

5. Wenqian Jiang, Qiushi Huang, Jian Shi, Chenye Wu, Dan Wang, and Zhu Han, “Federated Shift-Invariant Dictionary Learning Enabled Distributed User Profiling”, in IEEE Transactions on Power Systems, 2023.

Focus Area 5: Maritime Energy Nexus for Electrification, Security, and Resilience

Our research aims to develop a Smart Maritime Energy Nexus (SMEN) framework, to advance the synergies among different technical, operational, policy-level, and stakeholder engagement measures that can be implemented by maritime ports and vessels, improve the resilience of the maritime transportation system, and accelerate its electrification and decarbonization. SMEN includes a holistic set of algorithms ranging from long-term capacity building and policymaking, to daily/hourly-ahead scheduling and market operation. Our work relies on the following three key operational measures:

Port Microgrid

When implemented within the port, a microgrid provides an integrated means to manage the high penetration level of distributed energy resources (DERs), such as renewable energy resources, energy storage devices, and controllable loads.
Port microgrid opens up a variety of opportunities for technology integration, capacity expansion, and sustainability enhancement.
The electrified shipboard power system can also be seen as a remote microgrid that is highly efficient, self-manageable, and sustainable, and can be plugged into the port energy infrastructure via OPS.

Onshore Power Supply

OPS, also known as cold ironing, allows a ship to be “plugged” into the port electricity system.
In this way, the ship can completely shut down its auxiliary diesel-burning engine and utilize the onshore electricity, a much cleaner alternative, to support its energy demand at berth without disruption to onboard services.
According to the U.S. environmental protection agency (EPA), the benefits of OPS deployment contribute to 60%-80% reductions of emissions of CO2 and other air pollutants from ships at different ports globally.

Zero/low-Carbon Vessel Fleet

Carbon-neutral fuels are critical to ease the transition of the maritime energy mix.
Cost and maturity of technology: Electric energy storage, renewable energy, ammonia from electrolysis/natural gas, hydrogen, carbon capture and storage, bio-methanol, bio-LNG, bio-marine gas oil
The key is to design cost-effective and robust ship technologies and fuel strategies for existing vessels and newbuilds complying with the environmental regulations and stakeholder demands.

Select Publications

Book Chapters:

1. Jian Shi, Gino Lim, and Anahita Molavi, “Optimization for Power Systems in Maritime Environment”, accepted for publication, in Encyclopedia of Optimization, 2023.

Journal Publications:

1. Chengji Liang, Weiwei Sun, Jian Shi*, Kailai Wang, Yue Zhang, and Gino Lim, "Decarbonizing Maritime Transport through Green Fuel-Powered Vessel Retrofitting: A Game-Theoretic Approach", in Journal of Marine Science and Engineering, 2024.

2. Yue Zhang, Chengji Liang, Jian Shi*, Gino J. Lim, and Yiwei Wu, “Optimal Port Microgrid Scheduling Incorporating Onshore Power Supply and Berth Allocation Under Uncertainty”, in Applied Energy, vol 313, 2022.

3. Lifen Wang, Chengji Liang, Jian Shi*, Anahita Molavi, Gino J. Lim, and Yue Zhang, “A Bilevel Hybrid Economic Approach for Optimal Deployment of Onshore Power Supply”, in Applied Energy, vol. 292, 2021.

4. Wanlu Zhu, Jian Shi*, Pengfei Zhi, Lei Fan, and Gino J. Lim, “Distributed Reconfiguration of a Hybrid Shipboard Power System”, in IEEE Transactions on Power Systems, vol. 36, no. 1, pp. 4-16, Jan. 2021.

5. Anahita Molavi, Gino J. Lim, and Jian Shi*, “Stimulating Sustainable Energy at Maritime Ports by Hybrid Economic Incentives: A Bilevel Optimization Approach”, in Applied Energy, vol. 272, 2020.

6. Anahita Molavi, Jian Shi*, Yiwei Wu, and Gino J. Lim, “Enabling Smart Ports Through the Integration of Microgrids: A Two-Stage Stochastic Programming Approach”, in Applied Energy, vol. 258, 2020.

Focused Area 6: Smart Grid Cyber-Physical Security

Selected Journal Publications

1. Jieyu Lei, Shibin Gao, Jian Shi, Xiaoguang Wei, and Mohammad Shahidehpour, “A Data-Driven Approach for Quantifying and Evaluating Overloading Dependencies Among Power System Branches under Load Redistribution Attacks”, in IEEE Transactions on Smart Grid, 2023.

2. Yigu Liu, Shibin Gao, Jian Shi, Xiaoguang Wei, Xingpeng Li, and Zhu Han, “Critical Branch Identification Based on False Data Injection Flow Under Load Redistribution Attacks”, in IEEE Transactions on Industrial Informatics, 2023.

3. Jieyu Lei, Shibin Gao, Jian Shi, Xiaoguang Wei, and Mohammad Shahidehpour, and Tao Huang, “A Reasoning Approach Based on Pattern Graph for Analyzing the Risk of Power Outage Propagation”, in IEEE Transactions on Power Systems, 2023.

4. Shibin Gao, Jieyu Lei, Jian Shi, Xiaoguang Wei, Ming Dong, and Zhu Han, “Assessment of Overloading Associations Among Transmission Lines Under False Data Injection Attacks”, in IEEE Transactions on Smart Grid, vol. 13, no. 2, pp. 1570-1581, 2022.

5. Jieyu Lei, Shibin Gao, Jian Shi, Xiaoguang Wei, Ming Dong, Wenshuang Wang, and Zhu Han, “A Reinforcement Learning Approach for Defending Against Multi-Scenario Load Redistribution Attacks”, in IEEE Transactions on Smart Grid, vol. 13, no. 5, pp. 3711-3722, Sept. 2022.

6. Yigu Liu, Shibin Gao, Jian Shi, Xiaoguang Wei, Zhu Han, and Tao Huang, “Pre-overload-Graph-based Vulnerable Correlation Identification under Load Redistribution Attacks”, in IEEE Transactions on Smart Grid, vol. 11, no. 6, pp. 5216-5226, Nov. 2020.

7. Yigu Liu, Shibin Gao, Jian Shi, Xiaoguang Wei, and Zhu Han, “Sequential-Mining-Based Vulnerable Branches Identification under Continuous Load Redistribution Attacks”, in IEEE Transactions on Smart Grid, vol. 11, no. 6, pp. 5151-5160, Nov. 2020.

Page updated

Google Sites

Report abuse