基于多元线性回归和多目标规划模型的决策问题研究  

Research on Decision Problems Based on Model of Multiple Linear Regression and Multi-objective Programming

作　　者：韩丽娜[1] 李红[1] 韩改宁[1] 

机构地区：[1]咸阳师范学院图形图像处理研究所,陕西咸阳712000

出　　处：《自动化与仪器仪表》2016年第9期177-178,共2页Automation & Instrumentation

基　　金：陕西省教育厅项目(14JK1802);校级引进人才项目(13XSYK053)

摘　　要：为了解决复杂的决策问题,需要组合多个模型来完成,文章以橡胶配方决策问题为例,采用线性回归和多目标规划组合模型得到该问题的决策建议。文章利用给定的13组配方实验数据,首先建立原料与性能关系的多元线性回归模型并求解,其次利用得到回归系数和目标约束要求,建立优先级的多目标规划模型,然后采用适当的权重对多目标规划模型进行求解。通过对该决策问题的研究,证明使用线性回归模型和多目标规划模型的可行性和有效性。In order to solve the complex problem,It needs to combine multiple models.The author,as the rubber formulation of decision problem an example,uses a combination of linear regression and multi-objective programming model for making a decision suggestion.By making use of 13 groups of rubber formula test data,at first this paper established the multivariate linear regression model on the relationship between materials and properties and solved.Second it established the priority of multi-objective programming model based on the regression coefficient and constraint goal,and used the appropriate weight for solving the model.Through the research of the rubber recipe decision problems,using linear regression model and multiple objective programming models are feasible and effective.

关 键 词：多元线性回归模型 多目标规划模型 决策问题 

分 类 号：O639[理学—高分子化学]