부산대학교 도서관

Cloud robotics leverages the ubiquitous cloud infrastructure including networks, storage, servers, and services to enable robots to access unlimited resources. Most of the recent reported research concentrates on resource allocation for robotic application, yet the interoperability among multiple resources is a critical issue to address. This paper proposes a loosely coupled cloud robotic framework based on web service composition, which can organize multiple resources including robot nodes and cloud nodes to exchange messages through the abstract interface and fulfill complex robotic applications. Furthermore, the resource-deployment method is designed for the proposed framework to organize resources, so that the user overall quality of service requirements can be satisfied. Besides, we propose the concept of “user sensitivity” and cloud-priority strategy. Finally, we propose novel algorithms called GICA-CP and GICA motivated by imperialist competition for the resource-deployment method under the proposed framework. The simulation results demonstrate that the proposed methods can effectively allocate resources and enhance the cloud resource ratio in cloud robotic workflows. The experimental validation is provided to verify the efficiency of our framework further.