在云计算与大数据环境下，网络开放性、服务分布性与跨域性、用户需求多变性等因素使服务系统变得日趋复杂。服务选择、服务组合、服务资源调度的优化成为服务系统优化设计与运行的重要科学问题。服务领域特性对于服务优化问题求解效果影响甚大。本项目基于群体智能优化算法理论和领域特性对服务优化问题求解的影响规律，提出了面向服务领域的人工蜂群算法 (S-ABC) 范型及优化理论，可以高效求解服务优化问题；针对服务选择、服务组合和服务资源调度等优化问题，提出了一系列改进型人工蜂群算法簇(S-ABCx)；并提出了算法性能评价模型分析相关算法性能；还在海运物流服务、智慧家庭服务等应用领域验证S-ABC优化理论。本项目将加强人工蜂群算法与服务领域优化问题的结合，探索服务领域优化问题求解效果与效率更佳的新方法，深化群体智能优化理论。

Modern service systems become more and more complicated with the factors such as open network environment, cross-domain distribution and aggregation of massive services, dynamics of service resources, variability of customer requirements. How to solve the problems of service selection, service composition and service resource scheduling effectively and efficiently is a key scientific challenge for optimized service system design and operation. The specific characteristics of service domains have strong influences on problem-solving of service optimization. Based on the theory of swarm intelligence optimization, this project proposes a paradigm and optimization theory of service domain oriented artificial bee colony algorithm (S-ABC), in order to develop a new approach for solving service optimization problems effectively and efficiently. Further, the algorithm clusters of the improved artificial bee colony (S-ABCx) will be presented to solve the problems of service selection, service composition and service resource scheduling respectively, which are based on different influences of service domain characteristics on service optimization problems. Then, the algorithm performance evaluation models will be developed for analysis and comparison of the above algorithms. Finally, the S-ABC algorithm optimization theory will be validated and improved through the applications in specific service domains, such as sea-transportation services and smart home services. This project is expected to strengthen the combination of artificial bee colony algorithms and the optimization problems of service domain, to explore a new and better method for solving service optimization problems, and to extend the theory of swarm intelligence optimization.