关键词: 产业链协同, 核心竞争力, BP神经网络, 粒子群算法, 遗传算法

Abstract: To enhance the competitive advantage of industrial chain effectively,and to allocate resource belonging to industrial chain quickly and reasonably,the related mathematical model and index system were established through researching the core competitiveness of industrial chain and analyzing the components of core competitiveness.According to some redundant factors with high correlation and low impact in the high-dimensional index system of industrial chain's core competitiveness,a Self-Organizing Input/Output Mapping Neural Network (SIOM-NN) based on integration of BP neural network,particle swarm optimization algorithm and genetic algorithm was proposed.On this basis,the various complex factors in index system were screened to filter out redundant index factors,and the satisfactory factors that reflected the mapping relationship between input and output were identified.An identification model of key factor was constructed,and the key factors which constituted core competitiveness of industrial chain were identified.

Key words: industrial chains collaboration, core competitiveness, BP neural network, particle swarm optimization algorithm, genetic algorithms