智能代理模型在电力市场模拟系统中的运用

电力市场机制设计及运行是相当复杂的课题，传统的统计学和理论分析工具均很难精确研究电力市场的经济特性与物理特性。针对此类复杂的社会系统，采用智能代理模型仿真研究是种新颖有效的途径，因而基于智能代理模型的电力市场研究平台随之成为深入分析电力市场的重要工具。一般而言，在电力市场智能代理模型的框架中，代理就是代替发电厂商决策报价的程序模块，为了体现发电厂商报价决策的理性，不同的学习算法被用来构建代理的决策模块。本文着重分析并探讨了电力市场模拟系统中的智能代理模型及其学习算法的运用与实现。通过算例表明，智能代理模型及学习算法在电力市场中的应用具有重要意义。

Application of the Agent-Based Modeling In Electricity Market Simulation System

The mechanism design and operation of electricity market are very complex issues, It is extremely difficult to undertake economic and physical reliability research using traditional statistical and other analytical tools. Powerful and novel agent-based modeling(ABM) have been developed to cope with this degree of complexity, so electricity market ABM frameworks are constructed to improve the ability so as to analyze restructured electricity market. In the ABM frameworks, an agent is generally a decision-making module which represent a generation company. In order to ensure the rationality of agent acts, different learing algorithms are used to the construction of the module. The paper gives an introduction of practical Electricity Market Integrated Simulation System, and then discusses mainly the realization and application of the agent-based moding and its learning algorithms. An example shows that the agent-based modeling has great significance in the electricity market.