Application Research Of Multi-Agent Based Fuzzy Decision-making Method In Energy System

Energy is the important material basis of social development and improvement, which relates with country security and country economy vitals. It is the key factor related to the implementation of sustainable development strategy, playing a decisive role in the economy-oriented modernization of China. Hence, the energy system security issue could not be neglected. Research on fuzzy decision-making method, utilizing to realize the decision-making and allocation as well as to make balance between energy supply and consumption, is beneficial to assure energy system safety and stable, which is the key issue that the government has paid high attention at present. However, energy system is a big complicated system with strong non-linearity and dynamic characteristics, in which Multi- Agent method has great advantage. Thus, Multi-Agent method is important theoretical basis of the research of complicated system, and also a significant method to develop in the modeling of energy system.Based on the research on energy projects from Shanghai Development and Reform Research Institute, this research is given in the paper. According to the sustainable development of energy and human society, this paper analyzes the importance of energy security problems. And founded on domestic and foreign research situation of energy system prewarning and decision-making, a multi-agent based fuzzy decision-making method for energy system is proposed in this paper, which offers a new method for security state decision-making issue of energy system. For the up-to-down and down-to-up hierarchical structure characteristics, this paper constructs hierarchical energy system based on Multi-Agent method. According to the different tasks and functions of the agents, Energy System Management Agent (ESMA) and Fuzzy Decision-making Agent (FDA) which have hybrid structures, and Energy System Consumer Agent (ESCA) and Energy Supply Agent (ESA) which have reactive structures, are constructed respectively. The paper emphasizes on the FDA, including fuzzy method, fuzzy inference and the simulation of FDA in energy system. Furthermore, energy supply and demand difference normalization method and calculation formula of approach extent between time node and break nodes were designed in this paper. A learning mechanism can improve fuzzy rules in real time, which maintains the precision of fuzzy rules. Collaboration mechanism is an absolutely necessarily part of Multi-Agent system, which is also a hot domain for multi-agent studying. A collaboration model for multi-agent system with flexible property is designed. At last, simulation model of Multi-Agent based energy system is developed in Netlogo.According to the energy-relevant data offered by Shanghai Development and Reform Research Institute, the simulation result demonstrates that the research on Multi-Agent based fuzzy decision-making method could assure the balance between energy supply and consumption and security of energy system, provide the theoretical reference for energy-relevant department, which plays a positive role in development of domestic economic and society.