Abstract: Service system as a public service platform to support cluster supply chain operations, its good operation pattern is the precondition of the successful implementation of supply chain. In view of the problem that the optimal service selection strategy easily leads to low utilization rate of enterprise service resources to induce vicious competition of cluster, this paper proposes a service operation strategy based on the sub-optimal service selection, and builds an evolving simulation model of service system operation and assessment method by adopting computational experiment and Agent modeling technology combined with the application background. Analyzing and comparing the cluster development status under these two kinds of service operation strategy on Repast platform, the experimental results show that the sub-optimal service selection strategy is more advantageous to cluster keeping healthy and sound development.

Key words: computational experiment, service operation strategy, Agent, second-best theory, simulation

摘要： 服务系统作为支撑集群式供应链运营的公共服务平台（服务的发现，选择，匹配以及组合），其良好的运营模式是供应链成功实施的前提。针对服务系统运营中最优服务选择策略易导致集群中企业服务资源利用率低而引发集群产生恶性竞争问题，提出了一种基于次优服务选择的服务系统运营策略，并结合应用背景知识，通过采用计算实验、Agent建模方法建立了集群式供应链服务系统运营演化仿真模型和评估方法，在repast仿真软件对最优和次优两种服务系统运营策略进行仿真评估比较，结果表明，基于次优服务运营策略更有利于集群和服务系统保持健康良性的发展。

关键词: 计算实验, 服务运营策略, Agent, 次优理论, 仿真