Study On Software Quality Comprehensive Evaluation Method Based On Improved Particle Swarm Algorithm

With the rapid expansion of software technology, software products are exposedmore and more quality problem, so it is pretty important to evaluate software quality.The thesis studies the software quality, software quality metric model and indexsystem of comprehensive evaluation method and the software quality and relatedtheory and technology. The software quality comprehensive evaluation methods areanalyzed, and I find that there is a strong subjectivity in the present study. In order tosolve the problem, the thesis choose the projection pursuit model to do the softwarequality comprehensive evaluation.The thesis use improved particle swarm algorithmto optimize the projection vector in order to improve the accuracy of the softwarequality comprehensive evaluation. In the process of research on particle swarmoptimization algorithm, in order to improve the performance of particle swarmoptimization algorithm, I do the improvement from aspects such as inertia weight,learning factor, mixed strategy. The thesis does the specific research on the particleswarm improvement process and use improved particle swarm algorithm to optimizethe projection vector of projection pursuit model. The thesis designs a new calculationmethod of the software quality comprehensive evaluation, and uses the instance datato validate the calculation method.The thesis studies the theory of software quality and combines with the softwareselection or acceptance requirements to constructs a solution of comprehensiveevaluation of software quality index. The thesis summarizes the characteristics anddeficiencies of traditional software quality evaluation model, and constructs a newsoftware quality evaluation model based on projection pursuit model and particleswarm optimization. The thesis sets up an improved particle swarm algorithm withhigher performance.