基于多智能体仿真的生态工业园建模与优化——以广东省云安生态工业园为例

Modeling and Optimization of Eco-industrial Park based on Multi-Agent Simulation: A Case Study of Yun’an Eco-Industrial Park in Guangdong Province

摘要:

为辅助生态工业园（EIP）的规划设计并缩短验证时间，可利用建模仿真来模拟园区的演化历程并为其提出优化建议。基于多智能体建模法，为EIP各成员建立符合真实企业运作特点的智能体模型，开发一个扩展性较强的EIP动态仿真平台，并提出一种副产品动态交易的多目标优化模型。最后以广东省云安生态工业园为仿真对象，对其产业链协作过程及运营数据进行预测，通过迭代优化及仿真为企业提供决策参考，并利用仿真评估引入新企业、政府补贴对园区内副产品交易的影响。

Abstract:

In order to assist the design process of eco-industrial park (EIP) and shorten the time needed for verification, modeling and simulation can be adopted to predict the evolution of an EIP as well as provide suggestions on optimization. Based on multi-agent modeling, this paper establishes basic models of EIP members, develops a dynamic simulation platform which is highly scalable and applicable, and also proposes a multi-objective optimization model of by-product transaction. Finally, this paper takes Yun’an Eco-industrial Park as the simulation object, predicts the time needed to achieve dynamic stability in the park and accumulated all the operating data during this period. By iterative optimization and simulation, this paper provides a reference for an enterprise to decide its proper production plan. The impact of the introducing new enterprises and government subsidies on the by-product trade is also estimated in this paper.