Agent Technology Applied to the Protection of Power Systems

Introduction Power systems have undergone a great deal of change in the past decade. The United States electric power grid was deregulated in the late 1990’s. As a result, it is increasingly operated with smaller generation reserves and more transmission congestion. If deregulation continues, it seems likely that power flows will be larger and will fluctuate to a greater extent in future years. Faster, more reliable, and better coordinated protection and stability control are even more critical under this new environment than they have been in the past. New methods are needed to meet this challenge. Traditional protection and control systems rely upon standalone units that use local measurements as the basis for most decision making. Communication plays a very limited role in these legacy systems. A new era of decentralized control and efficiency demands has led to an environment demanding efficiency and reliability that pushes these legacy methods to their limits. The power system industry is beginning to recognize the benefits that communication could contribute towards greater system coordination, more rapid action, and increased correctness. The pervasiveness of the Internet has, at the same time, led to greater interest in using communication networks for improved protection and control systems. In particular, there has been an increasing interest within the power community in the use of networked agents to improve the reliability and efficiency of the electric power grid. Despite the interest level in the electric power community, there have been limited opportunities to investigate the effects of agents on realistic systems due to the lack of a platform that could simultaneously simulate both electric power and communication elements. This paper introduces agent studies utilizing EPOCHS, the Electric Power and Communication Synchronizing Simulator. EPOCHS is a new environment allowing users to create and test agent-based communications scenarios involving electric power transients. We present simulation results of practical agents in realistic protection and control scenarios and show that they make improvements on traditional methods. Three scenarios were investigated: a backup protection system, a transmission line differential protection scheme, and a SPS control system. In all cases, agent-based approaches proved to be superior to more traditional alternatives. Agent Technology in Power System Protection Definition The keyword ‘agent’ has been popularized both in and out of the artificial intelligence research community, but there is still no single universally accepted definition about what it means to be an agent. In this chapter, we define a software agent to be a computer program that has the properties of autonomy and interaction. In addition, an agent may exhibit the properties of mobility, intelligence, adaptivity, and communication.