A microgrid is an eco-friendly power system because renewable sources such as solar and wind power are used as the main power sources. For this reason, much research, development, and demonstration projects have recently taken place in many countries. Operation is one of the important research topics for microgrids. For efficient and economical microgrid operation, a human operator is required as in other power systems, but it is difficult because there are some restrictions related to operation costs and privacy issues. To overcome the restriction, autonomous operation for microgrids is required. Recently, an intelligent agent system for autonomous microgrid operation has been studied as a potential solution. This paper proposes a multiagent system for autonomous microgrid operation. To build the multiagent system, the functionalities of agents, interactions among agents, and an effective agent protocol have been designed. The proposed system has been implemented by using an ADIPS/DASH framework as an agent platform. The intelligent multiagent system for microgrid operation based on the proposed scheme is tested to show the functionality and feasibility on a distributed environment through the Internet.
A microgrid is an eco-friendly power system because renewable sources are used as main power sources. In the islanded operation mode of a microgrid, maintaining the balance between power supply and power demand is a very important problem. In the case of surplus supply, decreased generation output and/or charge of distributed storages can be applied to solve the imbalance between power supply and demand. In the case of supply shortages, increased generation output and/or discharge of distributed storages can be applied. Especially in the case of critical supply shortages, load shedding should be applied. In a distributed load-shedding approach, microgrid components need to make decisions autonomously. For autonomous microgrid operation, a multi-agent system has been investigated. In this paper, a distributed load-shedding system for agent-based autonomous operation of a microgrid is designed. The designed system is implemented and tested to show the functionality and feasibility of the proposed system.
1 March 2016 in International Journal of Distributed Sensor Networks
Combinatorial auction is a popular approach for resource allocation in cloud computing. One of the challenges in resource allocation is that QoS (Quality of Service) constraints are satisfied and provider's profit is maximized. In order to increase the profit, the penalty cost for SLA (Service Level Agreement) violations needs to be reduced. We consider execution time constraint as SLA constraint in combinatorial auction system. In the system, we determine winners at each bidding round according to the job's urgency based on execution time deadline, in order to efficiently allocate resources and reduce the penalty cost. To analyze the performance of our mechanism, we compare the provider's profit and success rate of job completion with conventional mechanism using real workload data.
Vehicular edge computing (VEC) is a promising technology for supporting computation-intensive vehicular applications with low latency at the network edges. Vehicles offload their tasks to VEC servers (VECSs) to improve the quality of service (QoS) of the applications. However, the high density of vehicles and VECSs and the mobility of vehicles increase channel interference and deteriorate the channel condition, resulting in increased power consumption and latency. Therefore, we proposed a task offloading method with the power control considering dynamic channel interference and conditions in a vehicular environment. The objective is to maximize the throughput of a VEC system under the power constraints of a vehicle. We leverage deep reinforcement learning (DRL) to achieve superior performance in complex environments and high-dimensional inputs. However, most conventional methods adopted the multi-agent DRL approach that makes decisions using only local information, which can result in poor performance, while single-agent DRL approaches require excessive data exchanges because data needs to be concentrated in an agent. To address these challenges, we adopt a federated deep reinforcement learning (FL) method that combines centralized and distributed approaches to the deep deterministic policy gradient (DDPG) framework. The experimental results demonstrated the effectiveness and performance of the proposed method in terms of the throughput and queueing delay of vehicles in dynamic vehicular networks.
Increased data traffic resulting from the increase in the deployment of connected vehicles has become relevant in vehicular social networks (VSNs). To provide efficient communication between connected vehicles, researchers have studied device-to-device (D2D) communication. D2D communication not only reduces the energy consumption and loads of the system but also increases the system capacity by reusing cellular resources. However, D2D communication is highly affected by interference and therefore requires interference-management techniques, such as mode selection and power control. To make an optimal mode selection and power control, it is necessary to apply reinforcement learning that considers a variety of factors. In this paper, we propose a reinforcement-learning technique for energy optimization with fifth-generation communication in VSNs. To achieve energy optimization, we use centralized Q-learning in the system and distributed Q-learning in the vehicles. The proposed algorithm learns to maximize the energy efficiency of the system by adjusting the minimum signal-to-interference plus noise ratio to guarantee the outage probability. Simulations were performed to compare the performance of the proposed algorithm with that of the existing mode-selection and power-control algorithms. The proposed algorithm performed the best in terms of system energy efficiency and achievable data rate.
A microgrid is composed of distributed power generation systems (DGs), distributed energy storage devices (DSs), and loads. To maintain a specific frequency in the islanded mode as an important requirement, the control of DGs’ output and charge action of DSs are used in supply surplus conditions and load-shedding and discharge action of DSs are used in supply shortage conditions. Recently, multiagent systems for autonomous microgrid operation have been studied. Especially, load-shedding, which is intentional reduction of electricity use, is a critical problem in islanded microgrid operation based on the multiagent system. Therefore, effective schemes for load-shedding are required. Meanwhile, the bankruptcy problem deals with dividing short resources among multiple agents. In order to solve the bankruptcy problem, division rules, such as the constrained equal awards rule (CEA), the constrained equal losses rule (CEL), and the random arrival rule (RA), have been used. In this paper, we approach load-shedding as a bankruptcy problem. We compare load-shedding results by above-mentioned rules in islanded microgrid operation based on wireless sensor network (WSN) as the communication link for an agent’s interactions.