园区循环经济系统有序度研究：基于复杂网络结构熵视角

The Complex Network Structure Entropy Based Research for Degree of Order of Circular Economy Park System

摘要:

为支持建立系统性的园区循环化改造评价方法，参考天然生态系统，在系统复杂网络建模基础上，选用度分布熵、吴氏结构熵和蔡氏结构熵研究循环化改造中系统的网络结构熵、项目平均熵贡献率、结构熵与网络拓扑参数的相关性，主要结论如下：（1）伴随网络规模增加和结构演化，系统网络结构熵上升，消除规模因素后项目平均熵贡献率下降，系统演化的有序度上升；（2）从点、边异构性同时评价系统有序度的吴氏结构熵更优；（3）园区循环经济系统复杂网络的度分布能起到与结构熵近似的有序度评价效果且形式简洁；（4）园区循环经济系统能量流、物质流网络兼有随机网络和星型网络拓扑特征。

Abstract:

In order to support the establishment of systematic evaluation method of the recycling transformation in industrial park, this study analyzed the structure entropy and degree of order of the park circular economy system with reference to the natural ecosystem. Therefore, this research takes the complex network as a basic model, and studies the network structure entropy, the average entropy contribution rate of project, the correlation of entropy and network topology parameter during the recycling transformation in an industrial park in Jiangsu Province based on the definition of the degree distribution entropy, the Wu structure entropy and the Chua structure entropy. The results show that: ① The network structure entropy increases with the network expands and structure evolution. The average entropy contribution rate of project deceases with the system degree of order increases when the factor of network scale is excluded; ② The Chua entropy which considers both the nodes and edges heterogeneous and the Wu entropy which considers only nodes heterogeneous show a correlation with r=0.990 at the 0.01 significance level bilateral. Since the Chua entropy is more complex, the Wu entropy is a better index; ③ The degree distribution parameter λ in complex network in park circular economy system and the Wu entropy show a correlation with r=0.917 at the 0.01 significance level bilateral. The degree distribution with a simple form is close to the Wu entropy in evaluation effectiveness; ④The complex network of a park circular economy system material and energy flow has both random network and star network topological features.