arXiv - EE - Systems and Control最新文献_第7页

Quantifying Aleatoric and Epistemic Dynamics Uncertainty via Local Conformal Calibration 通过局部共形校准量化 Aleatoric 和 Epistemic 动力学的不确定性

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08249

Luís Marques, Dmitry Berenson

{"title":"Quantifying Aleatoric and Epistemic Dynamics Uncertainty via Local Conformal Calibration","authors":"Luís Marques, Dmitry Berenson","doi":"arxiv-2409.08249","DOIUrl":"https://doi.org/arxiv-2409.08249","url":null,"abstract":"Whether learned, simulated, or analytical, approximations of a robot's\u0000dynamics can be inaccurate when encountering novel environments. Many\u0000approaches have been proposed to quantify the aleatoric uncertainty of such\u0000methods, i.e. uncertainty resulting from stochasticity, however these estimates\u0000alone are not enough to properly estimate the uncertainty of a model in a novel\u0000environment, where the actual dynamics can change. Such changes can induce\u0000epistemic uncertainty, i.e. uncertainty due to a lack of information/data.\u0000Accounting for both epistemic and aleatoric dynamics uncertainty in a\u0000theoretically-grounded way remains an open problem. We introduce Local\u0000Uncertainty Conformal Calibration (LUCCa), a conformal prediction-based\u0000approach that calibrates the aleatoric uncertainty estimates provided by\u0000dynamics models to generate probabilistically-valid prediction regions of the\u0000system's state. We account for both epistemic and aleatoric uncertainty\u0000non-asymptotically, without strong assumptions about the form of the true\u0000dynamics or how it changes. The calibration is performed locally in the\u0000state-action space, leading to uncertainty estimates that are useful for\u0000planning. We validate our method by constructing probabilistically-safe plans\u0000for a double-integrator under significant changes in dynamics.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"11 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Linear energy storage and flexibility model with ramp rate, ramping, deadline and capacity constraints 具有斜率、斜坡、截止日期和容量限制的线性储能和灵活性模型

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08084

Md Umar Hashmi, Dirk Van Hertem, Aleen van der Meer, Andrew Keane

{"title":"Linear energy storage and flexibility model with ramp rate, ramping, deadline and capacity constraints","authors":"Md Umar Hashmi, Dirk Van Hertem, Aleen van der Meer, Andrew Keane","doi":"arxiv-2409.08084","DOIUrl":"https://doi.org/arxiv-2409.08084","url":null,"abstract":"The power networks are evolving with increased active components such as\u0000energy storage and flexibility derived from loads such as electric vehicles,\u0000heat pumps, industrial processes, etc. Better models are needed to accurately\u0000represent these assets; otherwise, their true capabilities might be over or\u0000under-estimated. In this work, we propose a new energy storage and flexibility\u0000arbitrage model that accounts for both ramp (power) and capacity (energy)\u0000limits, while accurately modelling the ramp rate constraint. The proposed\u0000models are linear in structure and efficiently solved using off-the-shelf\u0000solvers as a linear programming problem. We also provide an online repository\u0000for wider application and benchmarking. Finally, numerical case studies are\u0000performed to quantify the sensitivity of ramp rate constraint on the\u0000operational goal of profit maximization for energy storage and flexibility. The\u0000results are encouraging for assets with a slow ramp rate limit. We observe that\u0000for resources with a ramp rate limit of 10% of the maximum ramp limit, the\u0000marginal value of performing energy arbitrage using such resources exceeds 65%\u0000and up to 90% of the maximum profit compared to the case with no ramp rate\u0000limitations.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"10 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217793","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Resilient Learning-Based Control Under Denial-of-Service Attacks 拒绝服务攻击下基于学习的弹性控制

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.07766

Sayan Chakraborty, Weinan Gao, Kyriakos G. Vamvoudakis, Zhong-Ping Jiang

引用次数: 0

Learning Communities from Equilibria of Nonlinear Opinion Dynamics 从非线性意见动态平衡中学习社群

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08004

Yu Xing, Anastasia Bizyaeva, Karl H. Johansson

{"title":"Learning Communities from Equilibria of Nonlinear Opinion Dynamics","authors":"Yu Xing, Anastasia Bizyaeva, Karl H. Johansson","doi":"arxiv-2409.08004","DOIUrl":"https://doi.org/arxiv-2409.08004","url":null,"abstract":"This paper studies community detection for a nonlinear opinion dynamics model\u0000from its equilibria. It is assumed that the underlying network is generated\u0000from a stochastic block model with two communities, where agents are assigned\u0000with community labels and edges are added independently based on these labels.\u0000Agents update their opinions following a nonlinear rule that incorporates\u0000saturation effects on interactions. It is shown that clustering based on a\u0000single equilibrium can detect most community labels (i.e., achieving almost\u0000exact recovery), if the two communities differ in size and link probabilities.\u0000When the two communities are identical in size and link probabilities, and the\u0000inter-community connections are denser than intra-community ones, the algorithm\u0000can achieve almost exact recovery under negative influence weights but fails\u0000under positive influence weights. Utilizing the fixed point equation and\u0000spectral methods, we also propose a detection algorithm based on multiple\u0000equilibria, which can detect communities with positive influence weights.\u0000Numerical experiments demonstrate the performance of the proposed algorithms.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"5 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Critically Damped Third-Order Langevin Dynamics 临界阻尼三阶朗文动力学

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.07697

Benjamin Sterling, Monica Bugallo

引用次数: 0

Optimal Operation of Distribution System Operator and the Impact of Peer-to-Peer Transactions 配电系统运营商的优化运行及点对点交易的影响

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08191

Hanyang Lin, Ye Guo, Firdous Ul Nazir, Jianguo Zhou, Chi Yung Chung, Nikos Hatziargyriou

{"title":"Optimal Operation of Distribution System Operator and the Impact of Peer-to-Peer Transactions","authors":"Hanyang Lin, Ye Guo, Firdous Ul Nazir, Jianguo Zhou, Chi Yung Chung, Nikos Hatziargyriou","doi":"arxiv-2409.08191","DOIUrl":"https://doi.org/arxiv-2409.08191","url":null,"abstract":"Peer-to-peer (P2P) energy trading, commonly recognized as a decentralized\u0000approach, has emerged as a popular way to better utilize distributed energy\u0000resources (DERs). In order to better manage this user-side decentralized\u0000approach from a system operator's point of view, this paper proposes an optimal\u0000operation approach for distribution system operators (DSO), comprising internal\u0000prosumers who engage in P2P transactions. The DSO is assumed to be a financial\u0000neutral entity, holding the responsibility of aggregating the surplus energy\u0000and deficit demand of prosumers after their P2P transactions while dispatching\u0000DERs and considering network integrity. Impacts of P2P transactions on DSO's\u0000optimal operation have been studied. Results indicate that energy matching P2P\u0000trading where only the total amount of energy over a given period of time is\u0000defined may affect quantities of energy exchanged between the DSO and the\u0000wholesale market, but not internal dispatch decisions of the DSO. Different\u0000levels of real-time power consistency may lead to different total surpluses in\u0000the distribution network. For the real-time power matching P2P trading, as a\u0000special case of energy matching P2P trading, the provided energy and total\u0000surplus are not affected. In other words, DSO can safely ignore P2P\u0000transactions if they follow the format defined in this paper. Case studies\u0000verify these conclusions and further demonstrate that P2P trading will not\u0000affect physical power flow of the whole system, but the financial distribution\u0000between the DSO and prosumers.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"299 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Optimizing Falsification for Learning-Based Control Systems: A Multi-Fidelity Bayesian Approach 优化基于学习的控制系统的证伪：多保真度贝叶斯方法

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08097

Zahra Shahrooei, Mykel J. Kochenderfer, Ali Baheri

{"title":"Optimizing Falsification for Learning-Based Control Systems: A Multi-Fidelity Bayesian Approach","authors":"Zahra Shahrooei, Mykel J. Kochenderfer, Ali Baheri","doi":"arxiv-2409.08097","DOIUrl":"https://doi.org/arxiv-2409.08097","url":null,"abstract":"Testing controllers in safety-critical systems is vital for ensuring their\u0000safety and preventing failures. In this paper, we address the falsification\u0000problem within learning-based closed-loop control systems through simulation.\u0000This problem involves the identification of counterexamples that violate system\u0000safety requirements and can be formulated as an optimization task based on\u0000these requirements. Using full-fidelity simulator data in this optimization\u0000problem can be computationally expensive. To improve efficiency, we propose a\u0000multi-fidelity Bayesian optimization falsification framework that harnesses\u0000simulators with varying levels of accuracy. Our proposed framework can\u0000transition between different simulators and establish meaningful relationships\u0000between them. Through multi-fidelity Bayesian optimization, we determine both\u0000the optimal system input likely to be a counterexample and the appropriate\u0000fidelity level for assessment. We evaluated our approach across various Gym\u0000environments, each featuring different levels of fidelity. Our experiments\u0000demonstrate that multi-fidelity Bayesian optimization is more computationally\u0000efficient than full-fidelity Bayesian optimization and other baseline methods\u0000in detecting counterexamples. A Python implementation of the algorithm is\u0000available at https://github.com/SAILRIT/MFBO_Falsification.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"13 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217792","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improving Initial Transients of Online Learning Echo State Network Control System via Feedback Adjustment 通过反馈调整改进在线学习回声状态网络控制系统的初始瞬态

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08228

Junyi Shen

引用次数: 0

Disinfectant Control in Drinking Water Networks: Integrating Advection-Dispersion-Reaction Models and Byproduct Constraints 饮用水网络中的消毒剂控制：平流-分散-反作用模型与副产品约束的整合

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08157

Salma M. Elsherif, Ahmad F. Taha, Ahmed A. Abokifa

{"title":"Disinfectant Control in Drinking Water Networks: Integrating Advection-Dispersion-Reaction Models and Byproduct Constraints","authors":"Salma M. Elsherif, Ahmad F. Taha, Ahmed A. Abokifa","doi":"arxiv-2409.08157","DOIUrl":"https://doi.org/arxiv-2409.08157","url":null,"abstract":"Effective disinfection is essential for maintaining water quality standards\u0000in distribution networks. Chlorination, as the most used technique, ensures\u0000safe water by maintaining sufficient chlorine residuals but also leads to the\u0000formation of disinfection byproducts (DBPs). These DBPs pose health risks,\u0000highlighting the need for chlorine injection control (CIC) by booster stations\u0000to balance safety and DBPs formation. Prior studies have followed various\u0000approaches to address this research problem. However, most of these studies\u0000overlook the changing flow conditions and their influence on the evolution of\u0000the chlorine and DBPs concentrations by integrating simplified\u0000transport-reaction models into CIC. In contrast, this paper proposes a novel\u0000CIC method that: (i) integrates multi-species dynamics, (ii) allows for a more\u0000accurate representation of the reaction dynamics of chlorine, other substances,\u0000and the resulting DBPs formation, and (iii) optimizes for the regulation of\u0000chlorine concentrations subject to EPA mandates thereby mitigating network-wide\u0000DBPs formation. The novelty of this study lies in its incorporation of\u0000time-dependent controllability analysis that captures the control coverage of\u0000each booster station. The effectiveness of the proposed CIC method is\u0000demonstrated through its application and validation via numerical case studies\u0000on different water networks with varying scales, initial conditions, and\u0000parameters.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"37 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217790","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Impacts of EPA Power Plant Emissions Regulations on the US Electricity Sector 美国环保局发电厂排放法规对美国电力行业的影响

arXiv - EE - Systems and Control Pub Date : 2024-09-12 DOI: arxiv-2409.08093

Qian Luo, Jesse Jenkins

{"title":"Impacts of EPA Power Plant Emissions Regulations on the US Electricity Sector","authors":"Qian Luo, Jesse Jenkins","doi":"arxiv-2409.08093","DOIUrl":"https://doi.org/arxiv-2409.08093","url":null,"abstract":"Taking aim at one of the largest greenhouse gas emitting sectors, the US\u0000Environmental Protection Agency (EPA) finalized new regulations on power plant\u0000greenhouse gas emissions in May 2024. These rules take the form of different\u0000emissions performance standards for different classes of power plant\u0000technologies, creating a complex set of regulations that make it difficult to\u0000understand their consequential impacts on power system capacity, operations,\u0000and emissions without dedicated and sophisticated modeling. Here, we enhance a\u0000state-of-the-art power system capacity expansion model by incorporating new\u0000detailed operational constraints tailored to different technologies to\u0000represent the EPA's rules. Our results show that adopting these new regulations\u0000could reduce US power sector emissions in 2040 to 51% below the 2022 level (vs\u000026% without the rules). Regulations on coal-fired power plants drive the\u0000largest share of reductions. Regulations on new gas turbines incrementally\u0000reduce emissions but lower overall efficiency of the gas fleet, increasing the\u0000average cost of carbon mitigation. Therefore, we explore several alternative\u0000emission mitigation strategies. By comparing these alternatives with\u0000regulations finalized by EPA, we highlight the importance of accelerating the\u0000retirement of inefficient fossil fuel-fired generators and applying consistent\u0000and strict emissions regulations to all gas generators, regardless of their\u0000vintage, to cost-effectively achieve deep decarbonization and avoid biasing\u0000investment decisions towards less efficient generators.","PeriodicalId":501175,"journal":{"name":"arXiv - EE - Systems and Control","volume":"60 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0