PSO优化灰色Verhulst模型用于研制经费投资决策

Grey Verhulst model with particle swarm optimization in development cost investment decision

针对新型装备研制经费投资分析问题,基于研制经费的数据特点和变化趋势特征以及灰色系统理论的独特优势,提出采用灰色Verhulst模型用于新型装备研制经费的投资分析。考虑到传统Verhulst模型中基于最小二乘的参数估计方法对年度投资变化较大时所出现的不适应性,提出采用粒子群优化(particle swarm optimization,PSO)算法优化模型参数,进而确定对新项目的年度经费投资额。应用分析表明,灰色Verhulst模型能较好适应数据的变化规律,同时基于PSO算法的参数优化方法较最小二乘法能得到更好地结果,可以用于指导研制经费的进度付款和合同签订。

To the problem of the development cost investment analysis of the new equipment,based on the data characteristics and trends of the development cost and the unique advantages of the grey system theory,the grey Verhulst model was adopted to make the investment analysis of the new equipment development cost.Because the traditional least squares parameter estimation method is not adaptive to the large changes in the annual investment,the particle swarm optimization（PSO）algorithm was adopted to optimize the model parameters and to determine the annual funding for the new investment projects.Case analysis shows that the grey Verhulst model can adapt the changes of the data,while the parameter optimization method based on PSO algorithm can obtain better results than the least squares method,which can be used to direct the progress payments of the development cost and the subscription of the development contract.