Complex & Intelligent Systems (Sep 2021)
Evaluation of regional industrial cluster innovation capability based on particle swarm clustering algorithm and multi-objective optimization
Abstract
Abstract With the progress of the times and the development of science, industrial clusters have been regarded by all countries in the world as one of the important ways to enhance regional competitiveness, and become an inevitable trend of industrial development. The research on the innovation ability of industrial clusters can not only maintain sustainable development of industrial clusters and obtain sustained competitive advantages, but also provide reference for the government's policy formulation of industrial clusters. This paper aims to study the evaluation of regional industrial clusters' innovation capability based on particle swarm clustering and multi-objective optimization. This paper uses the theory of industrial cluster innovation and takes regional industrial system as the empirical research object to establish a regional industrial system capability evaluation system, which is based on the selection of indicators, combined with analytic hierarchy process and factor analysis to evaluate industrial innovation capability. On this basis, the particle swarm clustering theory is used to verify the innovation ability and evaluation index system of industrial clusters, and provide a reference for the evaluation of the innovation ability of industrial clusters. This paper divides the regional cluster innovation capability into four aspects: innovation input capability, environment support capability, self-development capability and innovation output capability, and systematically analyzes the key elements and in the composition of innovation elements and their relationships. It then constructs the evaluation index system of regional cluster innovation capability. At the same time, this paper introduces clustering analysis algorithm and swarm intelligence algorithm into regional innovation evaluation, combines particle swarm optimization algorithm and K-means clustering algorithm, and optimizes particle swarm clustering algorithm by adjusting adaptive parameters and adding fitness variance. The experimental results of this paper show that from the results of the tested innovation potential of the three industrial clusters, industrial cluster F has the strongest innovation ability, with an evaluation coefficient of 0.851, followed by industrial cluster F, which has a value of 0.623. This result is consistent with the actual innovation status of the selected industry. From this point of view, the established particle swarm clustering model for evaluating the innovation capability of regional industrial clusters is reliable and can be used to evaluate the innovation capability of different industrial clusters.
Keywords
