The Research Of Foreign Trade Competitiveness Modeling Method Based On Particle Swarm Optimization Algorithm

In the world of economic globalization today, the enhancement of foreign trade competitiveness is of great importance to a country or a region's economic development. The promotion of foreign trade competitiveness needs to firstly understand the correlation between foreign trade competitiveness and its main affecting factors. In the past researches of foreign trade competitiveness, most of them are evaluations of foreign trade competitiveness, while relatively few of them made in-depth researches in quantitative methods. In this paper, the author conducted an in-depth study of regional foreign trade competitiveness of Chongqing city based on the previous researches, and built a regressive model between foreign trade competitiveness and its main affecting factors by the application of regressive methods. It provides reference in policy-making in the scientific management of foreign trade.There exists a non-linear relation between foreign trade competitiveness and its main affecting factors. And it is difficult to achieve the modeling by the application of multidimensional non-linear regressive methods. Single variant was established traditionally and the difficulty of modeling was reduced through the combination of single regressive modeling. This article sets up weights on single regressive modeling and uses LS criterion to construct object function and minimize it under constraints, so that we could get the best weights of single regression modeling and multidimensional non-linear regressive modeling through the combination. Because the construction of objective functions by adoption of LS criterion is a non-linear program, the problem of a non-linear program can't be solved by a universal and effective method. Particle Swarm Optimization Algorithm mainly deals with continuous optimization problem. This paper applied the improved form of PSOA (WPSOA) to solving weights of single regressive modeling and achieved a better result.This paper uses relatively accurate LS criterion to construct object function and applied WPSOA to solving weights of single regressive modeling. The weights confirmed in this way could best integrate the information of single regressive modeling so that the final regressive modeling of foreign trade competitiveness could be more precise and more reasonable, which plays a theoretical and instructive role in the researches of foreign trade competitiveness. The practical study of Chongqing's foreign trade competitiveness proved that the methods used in this paper are feasible and of practical value.