skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Cross-Correlated Scenario Generation for Renewable-Rich Power Systems Using Implicit Generative Models
Generation of realistic scenarios is an important prerequisite for analyzing the reliability of renewable-rich power systems. This paper satisfies this need by presenting an end-to-end model-free approach for creating representative power system scenarios on a seasonal basis. A conditional recurrent generative adversarial network serves as the main engine for scenario generation. Compared to prior scenario generation models that treated the variables independently or focused on short-term forecasting, the proposed implicit generative model effectively captures the cross-correlations that exist between the variables considering long-term planning. The validity of the scenarios generated using the proposed approach is demonstrated through extensive statistical evaluation and investigation of end-application results. It is shown that analysis of abnormal scenarios, which is more critical for power system resource planning, benefits the most from cross-correlated scenario generation.  more » « less
Award ID(s):
2145063
PAR ID:
10486849
Author(s) / Creator(s):
; ; ; ; ;
Publisher / Repository:
MDPI
Date Published:
Journal Name:
Energies
Volume:
16
Issue:
4
ISSN:
1996-1073
Page Range / eLocation ID:
1636
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; (, Energies)
    null (Ed.)
    Contingency Constrained Optimal Power Flow (CCOPF) differs from traditional Optimal Power Flow (OPF) because its generation dispatch is planned to work with state variables between constraint limits, considering a specific contingency. When it is not desired to have changes in the power dispatch after the contingency occurs, the CCOPF is studied with a preventive perspective, whereas when the contingency occurs and the power dispatch needs to change to operate the system between limits in the post-contingency state, the problem is studied with a corrective perspective. As current power system software tools mainly focus on the traditional OPF problem, having the means to solve CCOPF will benefit power systems planning and operation. This paper presents a Quadratically Constrained Quadratic Programming (QCQP) formulation built within the matpower environment as a solution strategy to the preventive CCOPF. Moreover, an extended OPF model that forces the network to meet all constraints under contingency is proposed as a strategy to find the power dispatch solution for the corrective CCOPF. Validation is made on the IEEE 14-bus test system including photovoltaic generation in one simulation case. It was found that in the QCQP formulation, the power dispatch calculated barely differs in both pre- and post-contingency scenarios while in the OPF extended power network, node voltage values in both pre- and post-contingency scenarios are equal in spite of having different power dispatch for each scenario. This suggests that both the QCQP and the extended OPF formulations proposed, could be implemented in power system software tools in order to solve CCOPF problems from a preventive or corrective perspective. 
    more » « less
  2. ; ; (, Proceedings of the AAAI Conference on Artificial Intelligence)
    The study of generative models is a promising branch of deep learning techniques, which has been successfully applied to different scenarios, such as Artificial Intelligence and the Internet of Things. While in most of the existing works, the generative models are realized as a centralized structure, raising the threats of security and privacy and the overburden of communication costs. Rare efforts have been committed to investigating distributed generative models, especially when the training data comes from multiple heterogeneous sources under realistic IoT settings. In this paper, to handle this challenging problem, we design a federated generative model framework that can learn a powerful generator for the hierarchical IoT systems. Particularly, our generative model framework can solve the problem of distributed data generation on multi-source heterogeneous data in two scenarios, i.e., feature related scenario and label related scenario. In addition, in our federated generative models, we develop a synchronous and an asynchronous updating methods to satisfy different application requirements. Extensive experiments on a simulated dataset and multiple real datasets are conducted to evaluate the data generation performance of our proposed generative models through comparison with the state-of-the-arts. 
    more » « less
  3. ; ; ; ; (, ACM Transactions on Spatial Algorithms and Systems)
    Due to the limited availability of actual large-scale datasets, realistic synthetic trajectory data play a crucial role in various research domains, including spatiotemporal data mining and data management, and domain-driven research related to transportation planning and urban analytics. Existing generation methods rely on predefined heuristics and cannot learn the unknown underlying generative mechanisms. This work introduces two end-to-end approaches for trajectory generation. The first approach comprises deep generative VAE-like models that factorize global and local semantics (habits vs. random routing change). We further enhance this approach by developing novel inference strategies based on variational inference and constrained optimization to ensure the validity of spatiotemporal aspects. This novel deep neural network architecture implements generative and inference models with dynamic latent priors. The second approach introduces a language model (LM) inspired generation as another benchmarking and foundational approach. The LM-inspired approach conceptualizes trajectories as sentences with the aim of predicting the likelihood of subsequent locations on a trajectory, given the locations as context. As a result, the LM-inspired approach implicitly learns the inherent spatiotemporal structure and other embedded semantics within the trajectories. These proposed methods demonstrate substantial quantitative and qualitative improvements over existing approaches, as evidenced by extensive experimental evaluations. 
    more » « less
  4. ; ; (, IET Generation, Transmission & Distribution)
    Policy, technology, and economic uncertainties affect the net benefits of grid reinforcements, and should be considered in planning. Stochastic optimisation can improve the robustness and expected performance of transmission plans, but is computationally intensive because model size grows as more scenarios are considered. Therefore, the ability to find a small number of scenarios while still capturing the benefits of stochastic programming is crucial. In this study, the authors evaluate the performance of several promising scenario sampling methods. Criteria for comparison include an index of the economic consequences of simplifying scenarios (the expected cost of naïve solution), changes in first‐stage investment decisions, and maximum regret. The results of an application to multidecadal planning of the Western Electricity Coordinating Council system show that solutions perform well when based on scenarios chosen by either a distance‐based method or the stratified scenario section method with moment‐matched probabilities. In particular, for this application, these methods’ results closely resemble solutions obtained from a much larger model using the full scenario set, and surprisingly have a lower worst case regret. Thus, careful scenario reduction can result in useful models that are more easily solved or, alternatively, can be expanded to accommodate other important features of power systems and markets. 
    more » « less
  5. ; ; ; ; ; ; ; (, Nature Communications)
    Semantic segmentation of medical images is pivotal in applications like disease diagnosis and treatment planning. While deep learning automates this task effectively, it struggles in ultra low-data regimes for the scarcity of annotated segmentation masks. To address this, we propose a generative deep learning framework that produces high-quality image-mask pairs as auxiliary training data. Unlike traditional generative models that separate data generation from model training, ours uses multi-level optimization for end-to-end data generation. This allows segmentation performance to guide the generation process, producing data tailored to improve segmentation outcomes. Our method demonstrates strong generalization across 11 medical image segmentation tasks and 19 datasets, covering various diseases, organs, and modalities. It improves performance by 10–20% (absolute) in both same- and out-of-domain settings and requires 8–20 times less training data than existing approaches. This greatly enhances the feasibility and cost-effectiveness of deep learning in data-limited medical imaging scenarios. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email