Abstract: The method of combining service robot and intelligent space can extend robot’s sensing and decision-making ability, reduce robot’s burden. This paper completes the design of service robot intelligent space software-platform based on UPnP middleware, adopts the service-oriented architecture plus multi-agent application model, implements various intelligent devices and robots by using service as basic unit, and designs a central-node to schedule and manage service robot intelligent space from a global perspective. Experiments show that the software-platform had many merits such as robust performance, high degree of dynamic adaption, loose-coupled device, good expandability, et al. The research has laid a sound foundation for the further study and practicality of service robot intelligent space.

Key words: service robot intelligent space, middleware, UPnP, global scheduling, dynamic adaptation

摘要： 服务机器人与智能空间技术相结合的方法可以扩展机器人的感知和决策能力，减轻其负担。基于UPnP中间件技术搭建了服务机器人智能空间软件平台，采用面向服务架构设计结合多智能体的应用构造模式，以服务为基本单元完成多种智能设备和机器人的实现，并设计中心管理节点对服务机器人智能空间进行全局调度和操作。实验表明，服务机器人智能空间软件平台具有性能稳定、动态自适应性高、设备间耦合度低、可扩展性强等特点，为服务机器人智能空间更深入的研究和推广实用化打下较好的基础。

关键词: 服务机器人智能空间, 中间件, UPnP, 全局调度, 动态自适应性