We introduce a framework for optimal resource allocation in smart grids. We consider two components of the smart grid;the power distribution network and the data communication network. By defining suitable utility fun...
详细信息
ISBN:
(纸本)9781457717024
We introduce a framework for optimal resource allocation in smart grids. We consider two components of the smart grid;the power distribution network and the data communication network. By defining suitable utility functions, the power and bandwidth resources are optimally allocated. This requires the solution of the, so called, local public goods problem, in mathematical economics terminology. We propose an iterative, distributed algorithm for its solution. The algorithm is scalable for deployment in large networks since it requires only O(N) messages per network user per iteration, where N is the number of users. Moreover, it is guaranteed to converge, does not require revelation of private information from each user and all algorithm actions can be realized by programmable smart devices of the smart grid.
In handling massive-scale signal processing problems arising from 'big-data' applications, key technologies could come from the development of decentralized algorithms. In this context, consensus-based methods...
详细信息
In this paper, we investigate a decentralized approach to timestamping transactions in a replicated database, under partial replication in Peer-To-Peer (P2P) environments. In order to solve problems of concurrent upda...
详细信息
The uncertainties in renewable power generators and the proliferation of price-responsive load aggregators make it a challenge for independent system operators (ISOs) to manage the energy trading in the power markets....
详细信息
The uncertainties in renewable power generators and the proliferation of price-responsive load aggregators make it a challenge for independent system operators (ISOs) to manage the energy trading in the power markets. Hence, a centralized framework for the energy trading market may not be remained practical for the ISOs mainly due to violating the privacy of different entities, i.e., load aggregators and generators. It can also suffer from the high computational burden in a market with a large number of entities. Instead, in this paper, we focus on proposing a decentralized energy trading framework enabling the ISO to incentivize the entities toward an operating point that jointly optimize the cost of load aggregators and profit of the generators, as well as the risk of shortage in the renewable generation. To address the uncertainties in the renewable resources, we apply a risk measure called the conditional value-at-risk (CVaR) with the goal of limiting the likelihood of high renewable generation shortage with a certain confidence level. Then by considering the risk attitude of the ISO and the generators, we develop a decentralized energy trading algorithm with some control signals that properly coordinate the entities toward the market operating point of the ISO's centralized approach. Simulation results on the IEEE 30-bus test system show that the proposed decentralized algorithm converges to the solution of the ISO's centralized problem in a timely fashion. Furthermore, the load aggregators can help their consumers reduce their electricity cost by 18% on average through managing their loads using locally available information. Meanwhile, the generators can benefit from 17.1% increase in their total profit through decreasing their generation cost.
In handling massive-scale signal processing problems arising from 'big-data' applications, key technologies could come from the development of decentralized algorithms. In this context, consensus-based methods...
详细信息
ISBN:
(纸本)9781467369985
In handling massive-scale signal processing problems arising from 'big-data' applications, key technologies could come from the development of decentralized algorithms. In this context, consensus-based methods have been advocated because of their simplicity, fault tolerance and versatility. This paper presents a new consensus-based decentralized algorithm for a class of non-convex optimization problems that arises often in inference and learning problems, including 'sparse dictionary learning' as a special case. For the proposed algorithm, we provide sufficient conditions for convergence to a stationary point. Numerical results demonstrate the efficacy of the proposed algorithm and provide evidence that validates our convergence claim.
The current Transmission Expansion Planning (TEP) incentive mechanisms are inadequate. They either fail to ensure revenue sufficiency or achieve socially optimal investment. The non-negligible coordination between TEP...
详细信息
The current Transmission Expansion Planning (TEP) incentive mechanisms are inadequate. They either fail to ensure revenue sufficiency or achieve socially optimal investment. The non-negligible coordination between TEP and Generation Expansion Planning (GEP) in the deregulated environment introduces more computational challenges to the TEP problem. This paper proposes a novel negotiation mechanism that enables Generation Companies (GenCos) and Load-Serving-Entities (LSEs) to negotiate TEP strategies with Transmission Companies (TransCo) directly. The negotiation process is modeled based on the Nash Bargaining theory. We explore the intertwined relationship between TEP and GEP through a bi-level, single-leader-multi-follower model. We transform the upper-level problem for better tractability and introduce a modified Proximal-Message-Passing (PMP) decentralized algorithm to achieve generation investment equilibrium at the lower level. We then utilize an iterative solving approach to coordinate the two levels. The feasibility and efficiency of this mechanism and methodologies are demonstrated using an IEEE 24-bus test system. The numerical results verify that our mechanism ensures revenue sufficiency and achieves socially optimal TEP strategies comparable to state-ofthe-art mechanisms. Additionally, our mechanism maintains transmission network user privacy, aligns the benefits of TransCo with those of transmission network users, and ensures a fair allocation of TEP costs and risks. The proactive participation of market players enabled by the negotiation mechanism can promote the transformation towards new market systems by mitigating the stranded cost issue. Moreover, our decentralized algorithm effectively addresses the non-cooperative nature of GEP, and the computational efficiency analysis justifies the model's scalability and practicality.
In the resource-constrained Internet of Things (IoT)-edge computing environment, split federated (SplitFed) learning is implemented to enhance training efficiency. This method involves each terminal device dividing it...
详细信息
In the resource-constrained Internet of Things (IoT)-edge computing environment, split federated (SplitFed) learning is implemented to enhance training efficiency. This method involves each terminal device dividing its full deep neural network (DNN) model at a designated layer into a device-side model and a server-side model, then offloading the latter to the edge server. However, existing research overlooks four critical issues as follows: 1) the heterogeneity of end devices' resource capacities and the sizes of their local data samples impact training efficiency;2) the influence of the edge server's computation and network resource allocation on training efficiency;3) the data leakage risk associated with the offloaded server-side submodel;and 4) the privacy drawbacks of current centralized algorithms. Consequently, proactively identifying the optimal cut layer and server resource requirements for each end device to minimize training latency while adhering to data leakage risk rate constraint remains a challenging issue. To address these problems, this article first formulates the latency and data leakage risk of training DNN models using SplitFed learning. Next, we frame the SplitFed learning problem as a mixed-integer nonlinear programming challenge. To tackle this, we propose a decentralized proactive model offloading and resource allocation (DP-MORA) scheme, empowering each end device to determine its cut layer and resource requirements based on its local multidimensional training configuration, without knowledge of other devices' configurations. Extensive experiments on two real-world datasets demonstrate that the DP-MORA scheme effectively reduces DNN model training latency, enhances training efficiency, and complies with data leakage risk constraints compared to several baseline algorithms across various experimental settings.
This article delves into the economic dispatch problem (EDP) within smart grids, specifically exploring it in time-varying directed networks. The objective is to allocate generation power efficiently among generators ...
详细信息
This article delves into the economic dispatch problem (EDP) within smart grids, specifically exploring it in time-varying directed networks. The objective is to allocate generation power efficiently among generators to fulfill load demands while minimizing the total generation cost, adhering to local capacity constraints. Each generator carries its unique local generation cost, and the total cost is calculated by summing these individual costs. To this aim, a novel algorithm (ADED-TVD) Accelerated decentralized Economic Dispatch algorithm is introduced, which is suitable for Time-Varying Directed networks well. ADED-TVD takes inspiration from the parameter momentum accelerated technique to improve the convergence with different parameters resulting in different momentum (Nesterov or heavy-ball) methods. In addition, ADED-TVD lies in time-varying directed communication networks, where theoretical evidence of linear convergence towards the optimal dispatch is offered. Also, explicit bounds for the step-size and momentum parameters are obtained. Finally, simulations that delve into various aspects of EDP in smart grids are presented.
In this work, we consider the decentralized non-convex online optimization problem over an undirected network. To solve the problem over a communication-efficient manner, we propose a novel quantized decentralized ada...
详细信息
In this work, we consider the decentralized non-convex online optimization problem over an undirected network. To solve the problem over a communication-efficient manner, we propose a novel quantized decentralized adaptive momentum gradient descent algorithm based on the adaptive momentum estimation methods, where quantified information is exchanged between agents. The proposed algorithm not only can effectively reduce the data transmission volume but also contribute to improved convergence. Theoretical analysis proves that the proposed algorithm can achieve sublinear dynamic regret under appropriate step-size and quantization level, which matches the convergence of the decentralized online algorithm with exact-communication. Extensive simulations are given to demonstrate the efficacy of the algorithm.
In this paper, we present a spatiotemporal decomposition solution approach to the fully decentralized dynamic economic dispatch (DED) problem in a microgrid. Our approach divides the centralized DED problem into a ser...
详细信息
In this paper, we present a spatiotemporal decomposition solution approach to the fully decentralized dynamic economic dispatch (DED) problem in a microgrid. Our approach divides the centralized DED problem into a series of sub-problems in the spatiotemporal dimensions and relies on multiple agents to solve those sub-problems. The proposed method requires no central operator intervention, preserving the decision independence and information privacy of each unit. Approximate value functions are used to describe the interaction among those sub-problems. With the approximate value functions, one agent not only knows the impact of its decision on the decisions of other agents in the same period, but also knows the impact of this decision on the decisions of its subsequent periods. Unlike the existing value function update strategy, which updates the state variables and value functions in one direction, we update the state variables and value functions in two directions based on a forward-push-back strategy. In this manner, the time-delayed problem can be solved, and the iteration speed of the algorithm is greatly improved. Moreover, the proposed algorithm does not require parameter tuning and has good accuracy and adaptability. Numerical simulations for multiple cases demonstrate the effectiveness of the proposed algorithm.
暂无评论