Parallel Optimization Of Product Configuration And Supply Chain Design Based On Uncertainty

Product configuration is an effective way to realize rapid product development.After the product configuration scheme is determined,the best supplier should be selected for the purchased parts.But now traditional new product development and supplier selection are independent of each other,and the multiple uncertainty factors in every link in the process may lead to the increase of production cost and delivery delay.In view of the above problems,this paper mainly focuses on the problems of product configuration and supplier selection parallel optimization based on uncertain environment,mainly contents of the parallel optimization modeling and solving in the uncertain environment,the key performance parameter prediction of the product configuration scheme,and the key technology of the statistical analysis of the total cost and the delivery time of the product.The main research contents of this paper are as follows:1.The uncertainty parameters of each link are analyzed and processed scientifically,and the expression is standardized in the same way.Based on the analysis of the uncertainty,the parallel optimization model of product configuration and the supplier selection are established.2.The evolutionary optimization algorithm is used to solve the parallel optimization model of product configuration and supplier selection in uncertain environment,a new interval multi-objective evolutionary optimization algorithm is proposed which can solve the parallel optimization problem of product configuration and supplier selection in uncertain environment based on the new dominance relation.3.The abstract description and mathematical modeling of the performance prediction problem are carried out.The non-linear relationship between the product performance parameters and the configurable component parameters is excavated by the machine learning method to predict the performance of the new configuration scheme in real time.4.Based on the identified configuration options,the mathematical model of the total product cost and the mathematical model of the delivery time are established respectively.The sampling points of uncertain information are sampled,and the statistical distribution of total cost and delivery time of the configuration product is analyzed and calculated.5.The parallel optimization calculation system for product configuration and supplier selection is designed and developed,and has been preliminarily applied in a factory of Beijing Aerospace Science and Engineering Group.