Design And Implementation Of Cloud Service Scheduling Middleware For Aircraft-oriented Collaborative Design

Cloud computing has changed enterprise' operation modes and exerted great influence on people's life. As a comprehensive application platform, the cloud computing platform can provide users with a variety of services. However, as the cloud computing develops, researchers are more concerned about the energy consumption and resource allocation. These problems affect not only user experience, but also resource utilization of cloud computing centers, increasing the operating cost of cloud computing operators. Therefore, it is important to study the task load and resource allocation of cloud computing, which can improve the resource utilization of cloud computing centers and reduce operating costs of cloud computing centers.In the background of aircraft collaborative design, this paper studies the task load and resource allocation of cloud computing centers. The paper designs the cloud service middleware for aircraft collaborative design and studies two key technologies in depth. To handle the dramatic change and prediction difficulty of the cloud computing load, this paper proposes the PSO-WWSVM algorithm. The period and frequency of the input load can be analyzed with the wavelet transformation and the nonlinear regression of the support vector machine can be analyzed. This algorithm can model the task load in the cloud computing centers and the particle swarm optimization algorithm is used to search for the optimal combination of the parameters of the prediction model. The prediction results are verified by the Google cluster trace. To improve the resource utilization and reduce energy consumption of cloud computing centers, this paper studies the resource allocation for cloud computing centers. On the precondition of considering heterogeneity, this paper puts forward a multi-level dynamic heterogeneity resource allocation model for cloud computing centers, chooses the K-means clustering algorithm to divide the workload, and selects the first-fit algorithms for resource allocation. Google cluster trace is used in the simulation environment for verification.In this paper, the cloud service scheduling middleware for aircraft collaborative design is implemented. I completed two parts, including the load forecasting module and dynamic resource allocation module. The scheduling middleware can forecast the load, dynamically adjust the number of machines in cloud computing centers and dynamically allocate resources for task load. Tests show that the cloud service scheduling middleware can dynamically allocate resources and improve resource utilization.