Article
Mathematics, Applied
Christian Biefel, Frauke Liers, Jan Rolfes, Martin Schmidt
Summary: Linear complementarity problems are a powerful tool for modeling practical situations and connecting various areas of mathematics. Robust optimization for LCP is still in its early stages, and this paper introduces the concept of affinely adjustable robust LCPs to establish strong characterizations and existence results for uncertain LCP vectors. Additionally, a mixed-integer programming formulation is derived to solve the corresponding robust counterpart.
SIAM JOURNAL ON OPTIMIZATION
(2022)
Article
Automation & Control Systems
Xuanyu Cao, Junshan Zhang, H. Vincent Poor, Zhi Tian
Summary: This article introduces a new variant of ADMM that can preserve agents' differential privacy in consensus optimization. The study shows that to achieve the best convergence performance at a certain privacy level, the magnitude of injected noise should decrease as the algorithm progresses.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2021)
Article
Computer Science, Information Systems
Bahareh Bagheri, Nima Amjady, Shahab Dehghan
Summary: This article introduces a multi-scale multi-resolution uncertainty model for GMS problem in power system, addressing midterm and short-term uncertainties through plausible scenarios and polyhedral uncertainty sets. Affine policies are incorporated to make the approach tractable, and a stochastic affinely adjustable robust optimization (SAARO) problem is formulated to consider both midterm and short-term uncertainties. A new solution methodology involving stochastic optimization and probabilistic dual cut is presented, with numerical results confirming the effectiveness of the proposed model and approach.
IEEE SYSTEMS JOURNAL
(2021)
Article
Engineering, Electrical & Electronic
Qiongxiu Li, Richard Heusdens, Mads Graesboll Christensen
Summary: This paper proposes a quantization-based approach to address the privacy and communication cost issues in distributed optimization. By deploying an adaptive differential quantization scheme, it achieves a low-cost communication and privacy-preserving solution. The approach is applicable to different distributed optimization methods and demonstrates superior performance in terms of accuracy and communication cost.
Article
Energy & Fuels
Khalid Umer, Qi Huang, Mohsen Khorasany, Muhammad Afzal, Waqas Amin
Summary: Peer-to-peer energy trading relies on the participation of numerous prosumers. The proposed ETD-ADMM algorithm, utilizing node coloring, achieves efficient communication and rapid convergence, requiring fewer communications and iterations for optimal outcomes compared to other distributed approaches.
Article
Operations Research & Management Science
T. D. Chuong, V Jeyakumar, G. Li, D. Woolnough
Summary: This paper presents exact dual semi-definite programs for robust SOS-convex polynomial optimization problems with affinely adjustable variables, ensuring equality between optimal values of the robust problem and its associated dual SDP. The method covers commonly used uncertainty sets and is applicable to various robust optimization models.
Article
Engineering, Civil
Gunasekaran Raja, Sudha Anbalagan, Geetha Vijayaraghavan, Sudhakar Theerthagiri, Saran Vaitangarukav Suryanarayan, Xin-Wen Wu
Summary: VANETs are crucial for ITS, but vulnerable to individual or distributed attackers. The SP-CIDS system utilizes DML model with vehicle collaboration to enhance storage efficiency, accuracy, and scalability of IDS. By employing DP technique, a private ensemble classifier secures training data and achieves 96.94% accuracy.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Operations Research & Management Science
V. Jeyakumar, G. Li, D. Woolnough, H. Wu
Summary: In this paper, an improved robust optimization model is proposed for addressing data uncertainty in hypoxia-based radiation treatment planning. The model demonstrates effective handling of uncertainties in the dose influence matrix and re-oxygenation data in numerical experiments.
Article
Engineering, Civil
Yuhong Shuai, Liming Yao
Summary: This study develops a robust water withdrawal scheme for drought regions that can balance the trade-offs between sub-areas and water use participants, ensuring sustainable regional system development under uncertainties of water availability and demand.
WATER RESOURCES MANAGEMENT
(2021)
Article
Engineering, Electrical & Electronic
Hongli Liu, Yawei Shi, Zheng Wang, Liang Ran, Qingguo Lu, Huaqing Li
Summary: Distributed optimization algorithms are crucial for optimal coordination of distributed energy resources. This paper proposes a distributed algorithm based on the relaxed ADMM method to tackle the DER coordination problem, showing effectiveness through simulations on DER coordination.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2022)
Article
Energy & Fuels
Hangbo Yang, Pengcheng You, Ce Shang
Summary: The centralized planning assumption in the electricity network, natural gas network, and energy hubs neglects the independent ownership of different systems. A more practical approach is to establish a distributed planning framework that differentiates the electric system, gas system, and each EH as separate stakeholders. This paper proposes a distributed planning framework based on the ADMM, which decomposes the joint planning problem into multiple sub-problems for each system and EH, with unit commitment embedded for more accurate operation reflection.
Article
Energy & Fuels
Meysam Khojasteh, Pedro Faria, Zita Vale
Summary: This paper proposes a distributed model for determining the optimal energy trading strategy of community participants in energy communities. The model considers local day-ahead energy market, peer-to-peer contracts, and the power grid for trading energy and compensating for power shortages/surpluses. The robust optimization approach is used to model uncertainty, and the augmented Lagrangian relaxation and alternating direction method of multipliers methods are employed to decrease the solution time. A case study demonstrates that the proposed model significantly reduces the solution time of energy management problem in communities.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2023)
Article
Engineering, Electrical & Electronic
Yulu Jin, Lifeng Lai
Summary: This paper proposes a general framework to balance inference accuracy and privacy protection in the inference as service scenario. By preprocessing data and using a privacy-preserving mapping, privacy can be protected while reducing inference accuracy. An optimization problem is formulated to find the privacy-preserving mapping, and an iterative algorithm is developed to solve the problem.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2022)
Article
Automation & Control Systems
Alessandro Falsone, Maria Prandini
Summary: This paper proposes a novel Augmented Lagrangian Tracking distributed optimization algorithm for solving multi-agent optimization problems. The algorithm features a constant penalty parameter, the ability to cope with unbounded local constraint sets, and the ability to handle both affine equality and nonlinear inequality coupling constraints.
Article
Computer Science, Artificial Intelligence
Zhen Zhang, Shaofu Yang, Wenying Xu, Kai Di
Summary: This article presents a privacy-preserving and communication-efficient algorithm for distributed optimization problems. The proposed algorithm incorporates an event-triggered mechanism and a Hessian approximation technique for privacy preservation and computation efficiency. Theoretical analysis shows the algorithm's convergence and accuracy. Numerical simulations demonstrate the effectiveness and efficiency of the algorithm.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Engineering, Electrical & Electronic
Ahmad Attarha, Paul Scott, Sylvie Thiebaux
ELECTRIC POWER SYSTEMS RESEARCH
(2020)
Article
Engineering, Electrical & Electronic
Ahmad Attarha, Paul Scott, Jose Iria, Sylvie Thiebaux
Summary: This study proposes a price-elastic aggregator bidding approach combined with a network conforming layer to overcome issues such as inflexible bidding and bids not conforming to the network boundaries. By decomposing the problem into aggregator and network subproblems and enabling aggregators to provide reactive power support, the feasibility of the network is ensured and disruption to existing wholesale market structures is minimized. The results show a 19% increase in benefits for aggregators compared to the inflexible bidding approach.
IEEE TRANSACTIONS ON SMART GRID
(2021)
Article
Engineering, Electrical & Electronic
Archie Chapman, Andrew Fraser, Laura Jones, Heather Lovell, Paul Scott, Sylvie Thiebaux, Gregor Verbic
Summary: Bruny Island is a small island community and tourist hot spot off the southern coast of Tasmania, Australia, known for its beautiful white sandy beaches, rugged coastline, and delicious produce. The island population fluctuates with holiday peaks, putting strain on the undersea electricity cable connecting the island to mainland Tasmania, which is already reaching its limits.
IEEE POWER & ENERGY MAGAZINE
(2021)
Article
Engineering, Electrical & Electronic
Seyyed Mahdi Noori Rahim Abadi, Ahmad Attarha, Paul Scott, Sylvie Thiebaux
Summary: Two novel two-stage Volt/Var control schemes based on the AARC methodology were proposed to mitigate over-voltage issues caused by the integration of photovoltaic panels into distribution systems. Central measurements and local controllers were used to determine linear mapping functions for controlling reactive power and maintaining voltages within safe limits. The approaches were compared with existing techniques using Monte-Carlo simulation, showing decreased real power loss, reactive power usage, and line congestion.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2021)
Article
Engineering, Electrical & Electronic
Ahmad Attarha, S. Mahdi R. A. Noori, Paul Scott, Sylvie Thiebaux
Summary: This article focuses on the safe and reliable operation of the power system as it embraces consumer-owned distributed energy resources (DER) and enables consumer participation in the wholesale market. The alternating direction method of multipliers (ADMM) is used to optimize operating envelopes, considering the distributed nature of the problem. The introduction of a piecewise affinely adjustable robust bidding approach and a piecewise affine Q-P controller helps compensate for uncertainty variations and optimize network capacity.
IEEE TRANSACTIONS ON SMART GRID
(2022)
Article
Engineering, Electrical & Electronic
S. R. A. Mahdi Noori R. A., Mark Burgess, Masoume Mahmoodi, Ahmad Attarha, Paul Scott
Summary: In this paper, an adjustable scenario optimisation-based solution approach is proposed to maximize the usage of distributed energy resources (DER) generation while ensuring network security. Extensive comparisons and simulations show that the proposed approach outperforms conventional scenario optimisation and adjustable robust approaches in terms of performance.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Alban Grastien, Claire Benn, Sylvie Thiebaux
Summary: The paper emphasizes the importance of agents not only performing permissible actions, but also explicitly permissible actions to signal normative compliance. It proposes an algorithm to compute compliance signalling plans and discusses the significance of solving the computational problem of finding plans that signal compliance in AI planning.
AIES '21: PROCEEDINGS OF THE 2021 AAAI/ACM CONFERENCE ON AI, ETHICS, AND SOCIETY
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Ian Mallett, Sylvie Thiebaux, Felipe Trevizan
Summary: This study introduces novel admissible heuristics to guide the search for cost-optimal policies under multi-objective probabilistic temporal logic constraints. These heuristics estimate the probability that a partial policy extension satisfies the constraints by projecting and decomposing LTL formulae obtained through progression. Experimental results show that these heuristics further enhance the scalability gap between heuristic search and verification approaches to these planning problems.
THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Enrico Scala, Patrik Haslum, Sylvie Thiebaux, Miquel Ramirez
JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH
(2020)
Article
Computer Science, Artificial Intelligence
Sam Toyer, Sylvie Thiebaux, Felipe Trevizan, Lexing Xie
JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH
(2020)
