Combination Optimization Of Cloud Manufacturing Services Based On CNN And Improved Genetic Algorithm

Cloud manufacturing is a new manufacturing mode based on the network platform.It is oriented to the needs of users and uses modern interconnected cloud technology to realize the reasonable allocation of highly utilized resources.As the demand for cloud manufacturing services becomes more and more complex,more and more resources are available for deployment,and the data to be processed by the optimization model of cloud manufacturing service portfolio is also more and more huge,which puts forward higher requirements on the adaptability of cloud manufacturing service portfolio model and the convergence speed of intelligent optimization algorithm.To study the cloud manufacturing service portfolio problem,the traditional cloud manufacturing service portfolio model mostly adopts a single linear weighted model,which can not dynamically reflect the user satisfaction problem.To solve this problem,a dynamic satisfaction model based on convolutional neural network(CNN)is proposed.CNN learns historical data and adjusts weights dynamically to build cloud manufacturing service satisfaction models that meet different needs of users.The results show that the customer satisfaction of cloud manufacturing service portfolio based on CNN dynamic model is increased by 17.33% compared with the linear weighted model and 9.34%compared with the nonlinear model constructed by BP neural network,and the mean square error(MSE)is reduced to 0.00226.The essence of cloud manufacturing service portfolio optimization problem is a multi-objective optimization problem.In order to further improve the convergence rate of multi-objective optimization algorithm and ensure the diversity of solutions,the genetic algorithm is improved by introducing dynamic elite strategy and Pareto domination.By combining the improved genetic algorithm with the dynamic satisfaction model based on CNN,a cloud manufacturing service combination optimization algorithm(hereinafter referred to as the algorithm in this paper)is proposed.Finally,the proposed algorithm is compared with four intelligent optimization algorithms including basic genetic algorithm,multi-objective particle swarm optimization algorithm,artificial bee colony algorithm and differential evolution algorithm.The proposed algorithm is superior in terms of effectiveness,robustness and convergence speed.The feasibility and effectiveness of the proposed model construction method and algorithm strategy are verified in the cloud manufacturing service platform.The convergence rate is still fast in the high-dimensional model,which overcomes the disadvantages of the previous model and improves the satisfaction.