Industrial clusters are complex networks formed by numerous agents who continuously imitate, learn from each other and make optimal choice accordingly. The paper uses random learning game and multi-agent system models to construct a Chinese traditional industrial clusters low carbon evolution model and introduce an algorithm based on the network external effect and characteristics of agents adaptive behavior. Then the simulation of low-carbon competition, emergence and evolution was conducted, which produced some valuable conclusions.

The low-carbon evolution of traditional industry cluster is the key to a low-carbon economy, and also a frontier of industry cluster theory research. The paper uses evolutionary game theory to construct a low-carbon evolutionary model of Chinese traditional industrial clusters, which considers uncertain factors such as political, economic, cultural, etc. Through the analysis of the cluster low-carbon evolutionary paths and stable equilibrium strategies, the model reflects the inherent law of clusters low-carbon evolution. Finally, the paper gives advices to promote industrial cluster agents to adopt the low-carbon cooperation strategy.