Research On The Cloud Service Quality Evaluation Based On PSO-BP Neural Network In Cloud Manufacturing

As cloud computing is booming nowadays, cloud manufacturing have emerged, and the rapid development of manufacturing cloud services is also showed. From a number of similar features and different service quality of manufacturing cloud services, how to select the best cloud services have become the most important issue for business users. The key question is how to evaluate the overall quality of cloud service, which is the field of manufacturing cloud services to be solved.In this paper, the needs of machinery manufacturing business and characteristics of cloud services are regarded as the starting point, and it raised a cloud service quality evaluation system by the time validity, cost validity, credibility validity, characteristics validity, safety validity of five dimensions constituted. Then the results of gray correlation analysis are served as the desired output, and the PSO-BP neural network model, a sound and efficient cloud service quality evaluation model, is established. In allusion to the drawbacks of slow convergence speed and large training error of BP neural network, when the particle swarm optimization is implemented, the BP neural network model is improved, which makes the model optimization speed fast and gains good generalization ability for the network. Finally the network is trained and validated using MATLAB,the case study shows that this evaluation model has high feasibility and rationality. This study can provide decision-making references and constructive comments for business users to select the best cloud services.