摘要
针对灰色关联分析法存在的分辨系数的选择主观性较大,用关联系数的平均值得到的关联度降低了评价准确性等缺点,提出用经济学中的偏相关分析和不变替代弹性(CES)生产函数分析方法,建立安全投入与事故函数分析模型。应用SPSS和EViews软件,找出安全投入和事故之间偏相关系数绝对值最大的要素和标准化回归系数绝对值最大的要素,作为优先投入方向,并按照系数绝对值大小进行排序,为以后企业进行安全投入提供理论参考,避免企业资源的浪费。这两种方法可操作性强,主观随意性小,得到的结果可信性高。
In view of the deficiencies of gray correlation analysis such as great subjectivity in the choice of resolution ration and the reduction of appraisal accuracy caused by gray correlation degree gotten by the average of correlation coefficient, the partial correlation analysis in economics and CES （ constant elasticity of substitution） production functions are put forward and used to construct the analysis models between safety investment and accident function. Then, the factors with the maximum absolute value partial corre- lation coefficients and the maximum absolute value standardized regression coefficients are found out by use of SPSS and Eviews software, which are regarded as the preferred investment directions. Meanwhile, the importance of all the factors can be obtained by sequencing their absolute values, which provides theoreti- cal reference for enterprise to decide the investment direction and to avoid the waste of resources. The two methods can get a high credibility of the results due to their high operability and low subjectivity.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2010年第7期128-133,共6页
China Safety Science Journal
关键词
安全投入
安全产出
生产函数
偏相关分析
优化模型
回归分析
safety investment
safety output
production function
partial correlation analysis
optimized model
regression analysis
作者简介
作者简介：梁瑞（1968-），男，甘肃靖远人，兰州理工大学石油化工学院教授，博士研究生，安全工程本科教学指导委员会学科建设分委会委员。1992年本科毕业于甘肃工业大学化工设备与机械专业，同年留校任教，目前主要从事锅炉压力容器、压力管道及其他特种设备等方面的安全可靠性评价、结构完整性评价、失效、风险及经济性分析的教学及科研工作。E—mail：liangr@lut．cn．
副教授
