Research On Cyber-Physical System For Optimization Scheduling Algorithm

Cyber-Physical Systems (CPS) is a multi-dimensional and complex system which integrates computing, network and physical environment. It implements the organic integration and the depth of the collaboration of computing, communications and control technology. So it can perform real-time sensing, dynamic control and information services of large complex systems. And it can meet the coordination process optimization and energy saving of complex systems. So this technology has great research value and application value in military, medical, security, industrial and agriculture.Relative to conventional systems, generally CPS has the features about mixed with multi-dimension, massive data and good collaboration. During the operation of the system, it must meet the function such as real-time perception, dynamic topology, reliable transmission, rapid calculation etc. And these functions require the system to reflect the good coordination and feedback capabilities. Among them, the internal task scheduling policy has an important impact on the collaboration and efficiency of the system. Because the traditional scheduling policy can’t meet the current needs of CPS, so we need to design efficient and feasible task scheduling strategy for specific application scenarios.This paper focuses on the environment under the CPS how to schedule the task which need a lot of complex computation, the main work as follows:(1) First we make a comprehensive analysis of the characteristics and structure about CPS. And then we also introduce the current study about task scheduling algorithms of complex systems, and summarize their advantages and disadvantages.(2) In order to meet the rapid mapping of tasks and resources, and reduce the system’s energy consumption, this paper presents an ant colony algorithm and Cuckoo search hybrid algorithm(ACOCS) for task scheduling based on CPS environment. This method is effective to retain the characteristics of ant colony algorithm, which has parallelism and good robustness. And it combines the advantage about fast global search capability of the cuckoo search. The mixed algorithm can quickly establish task-resource mapping and efficiently get the best solution. At the same time it can not only reduce the response time required for scheduling, but also to some extent to improve the utilization rate of system resources.(3) In order to make CPS can meet the different QoS requirements of user in the practical application, this paper give the multi-dimensional QoS weighted single objective function based on the analysis of the current variety of QoS. And we make use of elite clusters and mutation operator to improve the standard PSO algorithm, to enhance its global search ability and solution accuracy. And then we propose a task scheduling strategy which combines the improved particle swarm algorithm and multiple QoS constrains. This strategy can ensure the premise of the user experience, and effectively improve the task-resource mapping efficiency.Finally, we use the CloudSim simulation tools for the above two algorithms performance simulation and analysis. And the simulation results show that compared to the classical algorithms, the proposed has better performance in terms of response time and resource utilization, and ensure the QoS requirements of users to complete the task.