@article{li_song_hu_lee_wu_rehm_lu_2024, title={Load Profile Inpainting for Missing Load Data Restoration and Baseline Estimation}, volume={15}, ISSN={["1949-3061"]}, url={https://doi.org/10.1109/TSG.2023.3293188}, DOI={10.1109/TSG.2023.3293188}, abstractNote={This paper introduces a Generative Adversarial Nets (GAN) based, Load Profile Inpainting Network (Load-PIN) for restoring missing load data segments and estimating the baseline for a demand response event. The inputs are time series load data before and after the inpainting period together with explanatory variables (e.g., weather data). We propose a Generator structure consisting of a coarse network and a fine-tuning network. The coarse network provides an initial estimation of the data segment in the inpainting period. The fine-tuning network consists of self-attention blocks and gated convolution layers for adjusting the initial estimations. Loss functions are specially designed for the fine-tuning and the discriminator networks to enhance both the point-to-point accuracy and realisticness of the results. We test the Load-PIN on three real-world data sets for two applications: patching missing data and deriving baselines of conservation voltage reduction (CVR) events. We benchmark the performance of Load-PIN with five existing deep-learning methods. Our simulation results show that, compared with the state-of-the-art methods, Load-PIN can handle varying-length missing data events and achieve 15%-30% accuracy improvement.}, number={2}, journal={IEEE TRANSACTIONS ON SMART GRID}, author={Li, Yiyan and Song, Lidong and Hu, Yi and Lee, Hanpyo and Wu, Di and Rehm, P. J. and Lu, Ning}, year={2024}, month={Mar}, pages={2251–2260} }
 @article{hu_li_song_lee_rehm_makdad_miller_lu_2024, title={MultiLoad-GAN: A GAN-Based Synthetic Load Group Generation Method Considering Spatial-Temporal Correlations}, volume={15}, ISSN={["1949-3061"]}, url={https://doi.org/10.1109/TSG.2023.3302192}, DOI={10.1109/TSG.2023.3302192}, abstractNote={This paper presents a deep-learning framework, Multi-load Generative Adversarial Network (MultiLoad-GAN), for generating a group of synthetic load profiles (SLPs) simultaneously. The main contribution of MultiLoad-GAN is the capture of spatial-temporal correlations among a group of loads that are served by the same distribution transformer. This enables the generation of a large amount of correlated SLPs required for microgrid and distribution system studies. The novelty and uniqueness of the MultiLoad-GAN framework are three-fold. First, to the best of our knowledge, this is the first method for generating a group of load profiles bearing realistic spatial-temporal correlations simultaneously. Second, two complementary realisticness metrics for evaluating generated load profiles are developed: computing statistics based on domain knowledge and comparing high-level features via a deep-learning classifier. Third, to tackle data scarcity, a novel iterative data augmentation mechanism is developed to generate training samples for enhancing the training of both the classifier and the MultiLoad-GAN model. Simulation results show that MultiLoad-GAN can generate more realistic load profiles than existing approaches, especially in group level characteristics. With little finetuning, MultiLoad-GAN can be readily extended to generate a group of load or PV profiles for a feeder or a service area.}, number={2}, journal={IEEE TRANSACTIONS ON SMART GRID}, author={Hu, Yi and Li, Yiyan and Song, Lidong and Lee, Han Pyo and Rehm, P. J. and Makdad, Matthew and Miller, Edmond and Lu, Ning}, year={2024}, month={Mar}, pages={2309–2320} }
 @article{shirsat_muthukaruppan_hu_paduani_xu_song_li_lu_baran_lubkeman_et al._2023, title={A Secure and Adaptive Hierarchical Multi-Timescale Framework for Resilient Load Restoration Using a Community Microgrid}, volume={14}, ISSN={["1949-3037"]}, url={https://doi.org/10.1109/TSTE.2023.3251099}, DOI={10.1109/TSTE.2023.3251099}, abstractNote={Distribution system integrated community microgrids (CMGs) can partake in restoring loads during extended duration outages. At such times, the CMGs are challenged with limited resource availability, absence of robust grid support, and heightened demand-supply uncertainty. This paper proposes a secure and adaptive three-stage hierarchical multi-timescale framework for scheduling and real-time (RT) dispatch of CMGs with hybrid PV systems to address these challenges. The framework enables the CMG to dynamically expand its boundary to support the neighboring grid sections and is adaptive to the changing forecast error impacts. The first stage solves a stochastic extended duration scheduling (EDS) problem to obtain referral plans for optimal resource rationing. The intermediate near-real-time (NRT) scheduling stage updates the EDS schedule closer to the dispatch time using new obtained forecasts, followed by the RT dispatch stage. To make the decisions more secure and robust against forecast errors, a novel concept called delayed recourse is designed. The approach is evaluated via numerical simulations on a modified IEEE 123-bus system and validated using OpenDSS and hardware-in-loop simulations. The results show superior performance in maximizing load supply and continuous secure distribution network operation under different operating scenarios.}, number={2}, journal={IEEE TRANSACTIONS ON SUSTAINABLE ENERGY}, author={Shirsat, Ashwin and Muthukaruppan, Valliappan and Hu, Rongxing and Paduani, Victor Daldegan and Xu, Bei and Song, Lidong and Li, Yiyan and Lu, Ning and Baran, Mesut and Lubkeman, David and et al.}, year={2023}, month={Apr}, pages={1057–1075} }
 @article{song_li_lu_2022, title={ProfileSR-GAN: A GAN Based Super-Resolution Method for Generating High-Resolution Load Profiles}, volume={13}, ISSN={["1949-3061"]}, url={https://doi.org/10.1109/TSG.2022.3158235}, DOI={10.1109/TSG.2022.3158235}, abstractNote={This paper presents a novel two-stage load profile super-resolution (LPSR) framework, ProfileSR-GAN, to upsample the low-resolution load profiles (LRLPs) to high-resolution load profiles (HRLPs). The LPSR problem is formulated as a Maximum-a-Posteriori problem. In the first-stage, a GAN-based model is adopted to restore high-frequency components from the LRLPs. To reflect the load-weather dependency, aside from the LRLPs, the weather data is added as an input to the GAN-based model. In the second-stage, a polishing network guided by outline loss and switching loss is novelly introduced to remove the unrealistic power fluctuations in the generated HRLPs and improve the point-to-point matching accuracy. To evaluate the realisticness of the generated HRLPs, a new set of load shape evaluation metrics is developed. Simulation results show that: i) ProfileSR-GAN outperforms the state-of-the-art methods in all shape-based metrics and can achieve comparable performance with those methods in point-to-point matching accuracy, and ii) after applying ProfileSR-GAN to convert LRLPs to HRLPs, the performance of a downstream task, non-intrusive load monitoring, can be significantly improved. This demonstrates that ProfileSR-GAN is an effective new mechanism for restoring high-frequency components in downsampled time-series data sets and improves the performance of downstream tasks that require HR load profiles as inputs.}, number={4}, journal={IEEE TRANSACTIONS ON SMART GRID}, author={Song, Lidong and Li, Yiyan and Lu, Ning}, year={2022}, month={Jul}, pages={3278–3289} }
 @article{paduani_song_xu_lu_2021, title={Maximum Power Reference Tracking Algorithm for Power Curtailment of Photovoltaic Systems}, ISSN={["1944-9925"]}, DOI={10.1109/PESGM46819.2021.9638157}, abstractNote={This paper presents an algorithm for power curtailment of photovoltaic (PV) systems under fast solar irradiance intermittency. Based on the Perturb and Observe (P&O) technique, the method contains an adaptive gain that is compensated in real-time to account for moments of lower power availability. In addition, an accumulator is added to the calculation of the step size to reduce the overshoot caused by large irradiance swings. A testbed of a three-phase single-stage, 500 kVA PV system is developed on the OPAL-RT eMEGAsim real-time simulator. Field irradiance data and a regulation signal from PJM (RTO) are used to compare the performance of the proposed method with other techniques found in the literature. Results indicate an operation with smaller overshoot, less dc-link voltage oscillations, and improved power reference tracking capability.}, journal={2021 IEEE POWER & ENERGY SOCIETY GENERAL MEETING (PESGM)}, author={Paduani, Victor and Song, Lidong and Xu, Bei and Lu, Ning}, year={2021} }