摘要
混凝土正交试验中经常出现多指标分析的问题,为了合理分配指标权重,尽可能的缩小配合比的优选范围,以陶粒混凝土正交试验为例,全面分析阐述了目前常用的优选方法:综合平衡法、综合评分法、功效系数法。为合理确定各指标权重,引入矩阵分析法对混凝土多指标正交试验设计进行优化,建立了指标层、因素层和水平层的三层结构模型和三层矩阵,给出指标值权矩阵的求解过程,得出各因素和各水平对试验结果的影响权重。实例分析表明,综合平衡法和综合评分法受主观影响较大;功效系数法受主观影响较小,但适用于各指标重要程度大致相当的试验;矩阵分析法优选过程客观,适用于指标重要性不同的试验。在此基础上确定了符合试验要求的混凝土配合比。
The multi-index optimization problem is often encountered in the concrete orthogonal experiment. In order to narrow down the optimal selection of the optimal mix proportion by mathematical methods, several optimal selection methods including integrated balance method, synthetical scored method and efficiency coefficient method for orthogonal test are mainly analyzed. The matrix analysis method by which index weights can be confirmed objectively is used for the orthogonal experiment, and three-layer model and three-layer structure matrix are set up firstly, then the solution process of weight matrix is shown and the weights of every factor and level are figured out. The analysis of the example indicates the integrated balance method and the synthetical scored method are influenced by subjective judgment, and the efficiency coefficient method not influenced by subjective judgment but for roughly equivalent indices. However, the matrix analysis method is fairly reasonable and for indices of varying importance. The optimum mixture is determined through different methods combined with the concrete example.
出处
《科技通报》
北大核心
2017年第11期47-51,99,共6页
Bulletin of Science and Technology
基金
河南省高等学校深部矿井建设重点学科开放实验室基金课题(2014KF-02)
中铁隧道集团有限公司科技创新计划项目(H14-233)
关键词
正交试验
多指标
矩阵分析法
陶粒混凝土
orthogonal experiment
multi-index
matrix analysis method
ceramsite concrete
