核心期刊网首页> 外文会议> 计算机 & 自动化

Agent Technology Applied to Monitoring, Security and Diagnosis of Energy Systems


Large scale energy conversion, distribution, and dissipation systems, for example deregulated power gridsrnwith a heterogeneous mix of local generation and linear and nonlinear loads, are extremely complex with poorly understoodrnpathologies. Such complex systems require a large and often unobtainable number of state variables and couplings. Hence,rnoperators find it hard to effectively grasp an adverse situation and take appropriate response; especially in cases whenrnoperators have yet to identify that an adverse event has occurred. The lessons of robotics and distributed automatic systemsrnneed to be applied to this class of system. “Intelligent” systems can help human operators in such scenarios by decouplingrnresponses, reconfiguring dependencies, and otherwise making the system robust to uncertainty and its derivatives. Forrnexample, the northeast power grid is a distributed system, in which sensors, generation, distribution, and dissipation arerndispersed throughout, and the event of Thursday, August 14, 2003, may very well be partially due to its overly centralizedrncontrol structure. While the event is barely understood at present, it required expansive amounts of data, processing, andrnaction during several increasingly dangerous states. Its manual and automatic system optimizers and fail-safes failed tornprevent a massive power outage in North America, affecting 50 million people in seven states in the United States andrnCanada. In a situation when rapid changes are taking place bottlenecks and misunderstanding can occur. Combine this with arntendency for control actions to be narrowly defined and conservative, and the unacceptable becomes a reality. Thus globallyrnsufficient action leads to locally certain catastrophe. Hence, it is advantageous to have more significant autonomous localrncontrol, which is cooperating with the system, to unburden central control centers of distracting and highly dimensionedrntasks. Some important factors which involve safe operation of an energy system are – data acquisition, monitoring, tracking,rnsecurity and vulnerability assessment. Each of these factors can have their separate control mechanism. Intelligent MultiplernAgents (IMA) basically is “distributed artificial intelligence.” As a community, IMA are able to evolve, self-reproduce,rnlearn, cooperate, and adapt. They can be used and applied to complex systems, which are decomposed into smaller, andrnsmaller, subsystems. IMA can be designed to adapt, learn, and address various aspects for safe operation of a system, such asrnmonitoring, diagnosis, prognosis, security, safety, and vulnerability assessment of various subsystems. IMA technology isrnwell suited for such kind of distributed control operations. It is true that IMA are poor candidates for large generalizedrnproblems and have often been oversold, much as intelligent systems and “artificial intelligence” where in the 1980’s.rnHowever, a small, but vital, subset of a system wide control structure is clearly a set of agent specific tasks and is best servedrnby IMA. (e.g. tasks requiring cooperation and negotiation for learning and adaptation. Reliable energy flow, of acceptablernquality and quantity is necessary for nearly every future human consideration, from the ways in which we consume electricityrnin our homes and businesses to the manner in which we convert energy in our societies. Peoples of the earth are intolerant ofrnenergy system failures and provide maximum political pressure on our leaders to keep energy systems as reliable as possible.rnVery little change occurs in energy technology development or insertion as long as this reliability is maintained. Thus, whenrnthe system is no longer reliable it is usually in catastrophe. IMA based control structures have the potential to change this sadrnbut true axiom. IMA, when applied to well defined problems, which require cooperation and negotiation for learning andrnadaptation, may provide both reliability and consistent improvement (change) in energy systems, such as our national grid,rnnuclear and conventional power plants, distributed generation, local consumption. This paper presents a survey on thernapplication of agent technology in the area of monitoring, diagnosis, security, safety, and vulnerability assessment of energyrnsystems. It points out that agent technology is well suited for such kind of distributed control operations.......

【作者名称】: S. K. Srivastava, D. A. Cartes
【作者单位】: Center for Advanced Power Systems, 2000 Levy Ave, Tallahassee, FL 32310, Center for Advanced Power Systems, 2000 Levy Ave, Tallahassee, FL 32310, dave@eng.fsu.edu
【关 键 词】: Agent Technology Applied to Monitoring, Security and Diagnosis of Energy Systems
【会议名称】: 10th International conference on robotics and remote systems for hazardous environments
【期刊论文数据库】: [DBS_Articles_01]
【期刊论文编号】: 101,419,172
【摘要长度】: 4,429
【会议地点】: Gainsville, FL(US)
【会议组织】: Center for Advanced Power Systems, 2000 Levy Ave, Tallahassee, FL 32310;Center for Advanced Power Systems, 2000 Levy Ave, Tallahassee, FL 32310, dave@eng.fsu.edu;
【会议时间】: 2004
【上篇论文】: 外文会议 - Towards a Bio-inspired Cognitive Architecture for Short-Term Memory in Humanoid Robots
【下篇论文】: 外文会议 - Process Algebra for Event-Driven Runtime Verification: A Case Study of Wireless Network Management

【论文下载】: 免费获取 该期刊&论文全文内容