Nengling Tai’s research while affiliated with Shanghai Jiao Tong University and other places

Obtaining accurate probabilistic energy forecasts and making effective decisions amid diverse uncertainties are routine challenges in future energy systems. This paper presents the solution of team GEB, which ranked 3rd in trading, 4th in forecasting, and 1st among student teams in the IEEE Hybrid Energy Forecasting and Trading Competition 2024 (HEFTCom2024). The solution provides accurate probabilistic forecasts for a wind-solar hybrid system, and achieves substantial trading revenue in the day-ahead electricity market. Key components include: (1) a stacking-based approach combining sister forecasts from various Numerical Weather Predictions (NWPs) to provide wind power forecasts, (2) an online solar post-processing model to address the distribution shift in the online test set caused by increased solar capacity, (3) a probabilistic aggregation method for accurate quantile forecasts of hybrid generation, and (4) a stochastic trading strategy to maximize expected trading revenue considering uncertainties in electricity prices. This paper also explores the potential of end-to-end learning to further enhance the trading revenue by adjusting the distribution of forecast errors. Detailed case studies are provided to validate the effectiveness of these proposed methods. Code for all mentioned methods is available for reproduction and further research in both industry and academia.

Grid‐interactive efficient buildings (GEBs) have garnered global attention for their ability to achieve flexible, resilient, and environmentally friendly objectives. However, the increasing integration of renewable energy sources (RESs) introduces challenges which can compromise power system stability. Traditional robust energy management approaches fall short as they fail to address the adverse impacts on small‐signal stability. Additionally, the complexity of coordinating diverse devices and their intricate interactions leave the concept of co‐optimization in GEBs in its nascent stages. To address these challenges, this paper proposes a robust optimization model for GEBs that minimizes costs while ensuring system stability. The model integrates adjustable droop gains in inverters connected to distributed energy resources (DERs). First, dynamic models for various GEB devices are developed. Next, an hourly optimal power flow problem is formulated using interval predictions for RESs to ensure robustness against uncertainties. Leveraging a polyhedral uncertainty set, the model is solved via a Benders decomposition‐based method, incorporating analytical stability sensitivity cuts. Simulations on a 33‐bus GEB demonstrate that the proposed model significantly enhances small‐signal stability at a relatively low cost, outperforming benchmark models in handling uncertainties. This approach marks a significant step forward in advancing the co‐optimization of energy management and stability in GEBs.

The decarbonization of marine transport is a global challenge due to the range and capacity limitations of renewable ships. Offshore charging stations have emerged as an innovative solution, despite increased investment and extended voyage durations. Here we develop a route-specific model for the optimal placement and sizing of offshore charging stations to assess their economic, environmental and operational impacts. Analysing 34 global and regional shipping routes, we find that offshore charging stations can reduce the cost for electric ships by US$0.3–1.6 (MW km)⁻¹ and greenhouse gas emissions by 1.04–8.91 kg (MW km)⁻¹ by 2050. The economic cruising range for 6,500 20-foot equivalent unit electric ships can increase from 3,000 km to 9,000 km. Voyage time costs for these enhancements vary between a 0% and 30% grace period of the original delivery time frame. We further investigate power-to-ammonia offshore refuelling stations as a proxy for e-fuels, which could potentially replace heavy fuel oil ships for routes over 9,000 km with only a 5% grace period.

Energy and logistics systems in low-carbon ports are deeply coupled into an integrated energy-logistics system (IELS). Existing energy and logistics management platforms are separated and hinder the centralized operational optimization of IELS. This paper proposes a fast distributed scheduling method to achieve the lowest operational cost for the low-carbon port. The centralized IELS scheduling problem is decomposed into iterative sub-problems for the separated management platforms using the alternating direction method of multipliers (ADMM). A learning-based warm-start strategy is proposed to ensure both fast resolution of the distributed scheduling problem and optimality. A convolutional neural network (CNN) is designed to learn to optimize the logistics sub-problems in the early iterations. A mixed integer quadratic programming model is formulated to solve the sub-problem of each system during the later iterations. Case studies based on a low-carbon port in Shanghai are carried out to validate the effectiveness of the proposed method. Simulation results demonstrate that the proposed method achieves the global lowest operation cost and reduces the computational time by nearly 50% compared to mechanism-based ADMM method.

Logistics-energy coordination significantly enhances energy efficiency in electrified seaports. However, daily changes in environment data necessitate the re-implementation of optimization procedures, causing huge computational burdens. This paper proposes an adaptive multi-agent reinforcement learning (MARL) large model for logistics-energy spatiotemporal coordination of container seaports. The well-trained large model can directly generate optimal policy for each operating day from environment data without re-solving. To achieve this, a comprehensive logistics-energy coordination model is first established considering the spatial and temporal constraints of all-electric ships (AESs), quay cranes (QCs), auto guided vehicles (AGVs), and the seaport power distribution network (SPDN). The model is formulated as a Markov Decision Process (MDP). Then a MARL large model is developed, involving a hypernetwork mapping environment data to optimal policy, and special structures for both hypernetwork and agent policy networks to adapt to any number of daily arrival AESs. Additionally, a cascading action modification layer is designed to ensure correct action outputs within complex spatiotemporal constraints. A tailored training method with two acceleration strategies are developed for the large model. Case studies illustrate that the large model after training can automatically generate optimal policies with little to no fine-tuning, outperforming existing methods that require extensive solution time.

In electric propulsion systems, virtual synchronous motor (VSM) control plays a critical role in buffering large power surges. Its performance is contingent upon the tuning of parameters including virtual inertia and damping, which should be optimised according to the dynamics of the DC capacitor. Conventional tuning methods often neglect these DC capacitor dynamics, leading to either overutilization or underutilization of the capacitor’s buffering capabilities. The complex, high-order nonlinear equations involved make it challenging to analytically derive the relationship between the VSM parameters and DC capacitor dynamics. Inspired by the classical equal-area criterion of stability, this paper proposes a DC-equal-area method to represent the capacitor dynamics as a whole and thus estimate the optimal parameters that maximize the inertial capacity of the DC capacitor. A small-signal model, encompassing both the typical fluctuation power and the controller dynamics, is developed. The proposed method achieves up to a 25% improvement in load fluctuation smoothing compared to existing approaches, as validated through Hardware-in-the-Loop (HIL) experiments conducted on the RT-LAB platform.

Shore-to-ship power (SSP) technology is an effective way for developing sustainable maritime transportation systems. Its implementation requires attractive pricing and incentive policies. This paper proposes a time-of-use (TOU)-based pricing strategy for SSP services considering the ship behaviors. A hierarchical pricing framework is first proposed to characterize the interactions among government regulators (GR), seaport authorities (SA) and ships, which is formulated as a tri-level two-loop Stackelberg game model. On the ship-side, each ship is treated as an independent stakeholder and the queuing process caused by limited number of berths is considered. Correspondingly, a coupled voyaging-berthing-queuing (CVBQ) model is established for each individual ship to accurately formulate their voyage scheduling, berth selection, queuing behaviors and power dispatch. The optimal CVBQ decision of each ship depends on its queuing time, which cannot be known before all ships make their decisions. To this end, a first-come-first-serve (FCFS)-based ship queuing algorithm is developed to chronologically derive the optimal decisions of all ships. The proposed hierarchical pricing model is solved iteratively by combining heuristics and commercial solvers. Case studies demonstrate the effectiveness of the proposed method.