Many existing distributed optimization algorithms are applicable to time-varying networks, whereas their convergence results are established under the standard $B$ -connectivity condition. In this letter, we establish...
详细信息
Many existing distributed optimization algorithms are applicable to time-varying networks, whereas their convergence results are established under the standard $B$ -connectivity condition. In this letter, we establish the convergence of the Fenchel dual gradient methods, proposed in our prior work, under a less restrictive and indeed minimal connectivity condition on undirected networks, which, referred to as joint connectivity, requires the infinitely occurring agent interactions to form a connected graph. Compared to the existing distributed optimization algorithms that are guaranteed to converge under joint connectivity, the Fenchel dual gradient methods are able to handle nonlinear local cost functions and nonidentical local constraints. We also demonstrate the effectiveness of the Fenchel dual gradient methods over time-varying networks satisfying joint connectivity via simulations.
In-Network Processing (INP) is an effective way to aggregate and process data from different sources and forward the aggregated data to other nodes for further processing until it reaches the end user. There is a trad...
详细信息
In-Network Processing (INP) is an effective way to aggregate and process data from different sources and forward the aggregated data to other nodes for further processing until it reaches the end user. There is a trade-off between energy consumption for processing data and communication energy spent on transferring the data. An essential requirement in the INP process is to ensure that the user expectation of quality of information (QoI) is delivered during the process. Using wireless sensor networks for illustration and with the aim of minimizing the total energy consumption of the system, we study and formulate the trade-off problem as a nonlinear optimization problem where the goal is to determine the optimal data reduction rate, while satisfying the QoI required by the user. The formulated problem is a Signomial Programming (SP) problem, which is a non-convex optimization problem. We propose two solution frameworks. First, we introduce an equivalent problem which is still SP and non-convex as the original one, but we prove that the strong duality property holds, and propose an efficient distributed algorithm to obtain the optimal data reduction rates, while delivering the required QoI. The second framework applies to the system with identical nodes and parameter settings. In such cases, we prove that the complexity of the problem can be reduced logarithmically. We evaluate our proposed frameworks under different parameter settings and illustrate the validity and performance of the proposed techniques through extensive simulation.
The increasing penetration of renewable energy and the further coupling of the electricity and carbon markets have hindered the realization of efficient and low-carbon transformation processes in new power systems. Th...
详细信息
The increasing penetration of renewable energy and the further coupling of the electricity and carbon markets have hindered the realization of efficient and low-carbon transformation processes in new power systems. This study addresses the optimization problems of joint peer-to-peer (P2P) electricity and carbon trading in multi-energy microgrids (MEMGs), taking into account the risks associated with renewable generation in a distributed manner. First, a coordinated operation model is developed to describe the joint P2P electricity and carbon trading issues among MEMGs, aiming to minimize operating costs, mitigate potential risk losses, and reduce renewable energy wastage. Second, the conditional value-at-risk technique, paired with stochastic programming, is employed to quantify potential risk losses arising from uncertainties. Finally, a distributed optimization approach is developed based on the alternating direction method of multipliers to maintain the privacy and independence of decision-making in individual MEMGs. During the trading processes, the Lagrangian multipliers are used as price signals to ensure fairness in optimal trading schemes among MEMGs. Moreover, a parallel solution mechanism is implemented to improve overall operational efficiency with minimal calculation expenditure. The simulation results demonstrate that the proposed method can reduce operation costs and carbon emissions while also preventing a significant amount of renewable energy abandonment.
In this paper, we propose the primal-dual method of multipliers (PDMM) for distributed optimization over a graph. In particular, we optimize a sum of convex functions defined over a graph, where every edge in the grap...
详细信息
In this paper, we propose the primal-dual method of multipliers (PDMM) for distributed optimization over a graph. In particular, we optimize a sum of convex functions defined over a graph, where every edge in the graph carries a linear equality constraint. In designing the new algorithm, an augmented primal-dual Lagrangian function is constructed which smoothly captures the graph topology. It is shown that a saddle point of the constructed function provides an optimal solution of the original problem. Further under both the synchronous and asynchronous updating schemes, PDMM has the convergence rate of O(1/K) (where K denotes the iteration index) for general closed, proper, and convex functions. Other properties of PDMM such as convergence speeds versus different parameter-settings and resilience to transmission failure are also investigated through the experiments of distributed averaging.
This paper proposes a distributed optimization operation method for multi-microgrids (MMGs) with electricity-carbon trading. First, a bi-level cooperative dispatch model of distribution networks and MMGs is establishe...
详细信息
ISBN:
(纸本)9798350339345
This paper proposes a distributed optimization operation method for multi-microgrids (MMGs) with electricity-carbon trading. First, a bi-level cooperative dispatch model of distribution networks and MMGs is established. The distribution locational marginal price (DLMP) is used to guide MMGs participating in electricity trading. Second, the ladder carbon trading is introduced in MMGs to achieve low carbon production. The alternating directional multiplier method (ADMM) is applied to calculate the proposed lower model and coordinate peer-to-peer (P2P) electricity trading among MMGs. Finally, the upper and lower models are solved iteratively based on the information of pass-through trading quantities and DLMP between the distribution network and MMGs, which realizes the coupling electricity-carbon trading of the MMGs as well as reflects the win-win situation of both low-carbon and economic aspects.
This paper proposes a distributed discrete-time algorithm for optimization problem with global multiple constraints. Based on the projected primal-dual method, a fully distributed algorithm is proposed, and is proved ...
详细信息
ISBN:
(纸本)9781728101057
This paper proposes a distributed discrete-time algorithm for optimization problem with global multiple constraints. Based on the projected primal-dual method, a fully distributed algorithm is proposed, and is proved to converge to the optimal solution of the optimization problem under certain sufficient condition with respect to the step-size. Numerical example is given to verify the efficacy of the proposed algorithm.
As we have seen, today machine learning and big data technologies are transforming both our daily life and economies fundamentally. An important factor that fuels the progress of learning algorithms is the abundance o...
详细信息
As we have seen, today machine learning and big data technologies are transforming both our daily life and economies fundamentally. An important factor that fuels the progress of learning algorithms is the abundance of data generated everyday. In many scenarios, including internet of things, intelligent transportation systems, mobile/edge computing, and smart grids, the datasets are often generated and stored locally in different locations. Traditional centralized (concentrated) algorithms, however, are facing challenges in these settings because they usually require much higher computation cost on a single machine, more communications for collecting raw local data, and are more vulnerable to possible failure of the host. Therefore the distributed learning and optimization algorithms, which are essentially exempted from those problems, are becoming promising alternatives that attract growing interest in recent years. Generally speaking, distributed algorithms describe the approaches that solve problems in a collaborative manner over multiple agents (machines, nodes, computation units or cores) based on communications among them. The main theme of this work is the identification of efficient and effective ways to exploit distributed procedures and communication structures in this type of settings and applications. The first part of this work contains the discussions of a communication-efficient distributed optimization framework for general nonconvex nonsmooth signal processing and machine learning problems under an asynchronous protocol. As we know, an important criterion for preferable distributed algorithms in latency and communication sensitive applications is that they can complete tasks fast with as less communication resources as possible. Thus in this part we present an asynchronous efficient distributed algorithm with reduced waiting time based on the updates utilizing local higher-order information and investigate the theoretical guarantee for the convergen
Microgrids will make fuller consumption of renew able energy by connecting with each other. On that scenario, the generation schedule is supposed to be optimized on view of the total interconnected microgrids (IMGs), ...
详细信息
ISBN:
(纸本)9781467381321
Microgrids will make fuller consumption of renew able energy by connecting with each other. On that scenario, the generation schedule is supposed to be optimized on view of the total interconnected microgrids (IMGs), which may infringe the privacy of MG. A distributed optimization method for generation scheduling of MIGs using the alternating direction method of multipliers (ADMM) is proposed in this paper, so that the MC won't need to share their own information of sources and loads and so on, but only the expected exchanging power information. The case confirms the efficiency of the distributed optimization method compared to the centralized one.
Wireless sensor networks are capable of collecting an enormous amount of data over space and time. Often, the ultimate objective is to derive an estimate of a parameter or function from these data. This paper investig...
详细信息
ISBN:
(纸本)1581138466
Wireless sensor networks are capable of collecting an enormous amount of data over space and time. Often, the ultimate objective is to derive an estimate of a parameter or function from these data. This paper investigates a general class of distributed algorithms for "in-network" data processing, eliminating the need to transmit raw data to a central point. This can provide significant reductions in the amount of communication and energy required to obtain an accurate estimate. The estimation problems we consider are expressed as the optimization of a cost function involving data from all sensor nodes. The distributed algorithms are based on an incremental optimization process. A parameter estimate is circulated through the network, and along the way each node makes a small adjustment to the estimate based on its local data. Applying results from the theory of incremental subgradient optimization, we show that for a broad class of estimation problems the distributed algorithms converge to within an epsilon-ball around the globally optimal value. Furthermore, bounds on the number incremental steps required for a particular level of accuracy provide insight into the trade-off between estimation performance and communication overhead. In many realistic scenarios, the distributed algorithms are much more efficient, in terms of energy and communications, than centralized estimation schemes. The theory is verified through simulated applications in robust estimation, source localization, cluster analysis and density estimation.
This paper discusses the distributed output optimization problem for a group of linear multi-agent systems via event-triggered control. In this problem, each agent has an individual objective function and is required ...
详细信息
ISBN:
(纸本)9789881563804
This paper discusses the distributed output optimization problem for a group of linear multi-agent systems via event-triggered control. In this problem, each agent has an individual objective function and is required to reach the minimal solution of the sum of these objective functions. By inserting a dynamic compensator to generate this optimal point, we develop a novel distributed event-triggered controller to solve this problem, where the control inputs only update at some time instants specified by us. An example is given to verify the effectiveness of our designs.
暂无评论