您好, 访客   登录/注册

基于灰色关联度的教职工健康状况分析

来源:用户上传      作者: 蒲艳,臧成丽,石慧

  摘 要 灰色关联分析(也称灰色关联度分析)不仅是灰色系统理论的重要组成部分,也是灰色系统分析、预测和决策的基石。本文首先对灰色关联分析理论进行了分析研究。续而将系统评价方法用于实际的计算,即采用灰色关联度的计算工具对教职工的健康状况做分析,得出教职工的健康比较。最后,通过实际的操作,提出对结果及灰色关联分析法的评价分析。
  关键词 灰色系统;灰色关联度;系统评价;数据无量纲化处理
  中图分类号TP39 文献标识码A 文章编号 1674-6708(2012)64-0190-02
  Based on Gray Relational Degree Analysis of the Staff's Health Condition
  
  PU Yan ,ZANG Cheng-li,SHI Hui
  Chengdu University of Technology, Chengdu 610059,Sichuan Province
  
  Abstract Grey Relational Analysis―GRA is not only the important factor of the grey relational theory but also the basis of the grey system analysis, forecast and decision. This paper primarily analyzes the theory of the GRA and then try to put the method of system evaluating into practice, which using the calculating tools of GRA to analyzing the health situation of teaching faculty which finally gets the results of comparison of teaching faculty’s health.
  Keywords Gray system; gray relation; system evaluation; data dimensionless;
  
  用数学方法定量分析、评价事物的方法越来越受到人们的重视,本文选取的数据杂乱无章,没有一定规律,且样本数据少,基于灰色关联分析“样本要求低、计算量小”的优点,所以采用灰色关联度分析。
  灰色关联度分析主要是通过对灰色系统中不同事物间的相关分析,根据因素之间发展趋势的相似或相异成都,衡量因素间关联程度的一种方法。
  1 原理
  灰色系统关联分析实质上是关联系数的分析。先是求各个方案与由最佳指标组成的理想方案的关联系数,由关联系数得到关联度,再按关联度的大小进行排序、分析,得出结论。灰色关联分析具有总体性、非对称性、非唯一性和有序性。
  它根据评价因素间发展态势的相似和相异程度确定评价因素的关联程度。
  式中: 为第k个时刻比较曲线与参考曲线的相对差值,即xi对x0在k时刻的关联系数,其中称为分辨系数,(0,1),常取0.5.实数。
  根据可求出各个时刻关联度的平均值即关联度:
  式中:为曲线对参考曲线的关联度。
  2 教职工健康状况分析实例研究
  1)在教职工的健康状况分析的具体问题中教职工健康指标由七个(包括血压,心率,血红蛋白,血量,葡萄糖,胆固醇,尿素)指标综合表示的;
  2)计算身体指标对健康的关联度。
  第一步:取表中的第33号个体的身体指标为标准序列的各个指标;因为本文要解决的问题是教职工健康状况的比较评价,而原始数据给了一千多组数据,为了说明问题的客观性,所以算选的比较序列为随即抽取10个个体(分别为第11,127,349,402,677,981,1067,1348,1399,1562号)的身体指标作为比较序列加以分析。
  第二步:数据无纲化处理。
  
  得出结果:
  (1.2432,0.7956,2.0388, 0.6713,0.8702,1.1313,0.2496);
  (1.2006,0.8368 ,1.8919,0.6549,0.970,1.1764,0.2691);
  (1.5525,0.8490,1.8921,0.5943,0.8490,1.0188,0.2442);
  (1.2408,0.7167,1.8612,0.8557,1.0055,1.0376,0.2826);
  (1.4390,0.9802,1.6893,0.7195,0.9281,0.9698,0.2741);
  (1.6257,0.9209,1.6939,0.6821,0.8413,1.0004,0.2357);
  (1.1815,0.7089,1.8341,0.8102,1.0127,1.1477,0.3049);
  (1.3251,0.8358,1.7430,0.8868,0.8970,1.0091,0.3032);
  (1.4413,0.9462,1.8264,0.7041,0.8252,1.0012,0.2556);
  (1.1418,0.8946,2.0718,0.7534,0.8475,1.0477,0.2432);
  (1.4536 ,0.9283,1.8567,0.6230,0.8184,1.0505,0.2696)。
  第三步:求差序列。
   。
  得到结果为:
  (0.0425,0.0412,0.1469,0.0164,0.1000,0.0451,0.0195);
  (0.3093,0.0534,0.1467,0.0770,0.0212,0.1124,0.0055);
  (0.0024,0.0790,0.1776,0.1844,0.1353,0.0937,0.0330);
  (0.1959,0.1846,0.3495,0.0482,0.0579,0.1615,0.0245);
  (0.3825,0.1252,0.3449,0.0108,0.0289,0.1308,0.0140);
  (0.0617,0.0867,0.2047,0.1389,0.1425,0.0164,0.0553);
  (0.0819,0.0402,0.2958,0.2155,0.0268,0.1222,0.0536);
  (0.1981,0.1506,0.2124,0.0328,0.0450,0.1301,0.0060);
  (0.1013,0.0990,0.0330,0.0821,0.0227,0.0836,0.0064);
  (0.2104,0.1327,0.1821,0.0483,0.0518,0.0808,0.0200);
  (0.2104,0.1327,0.1821,0.0483,0.0518,0.0808,0.0200)。
  (0.3868,0.7914,0.5730,0.7218,0.9114,0.6376, 0.9842);
  (1.0000,0.7166,0.5250,0.5154,0.5930,0.6795 ,0.8635);
  (0.5002,0.5152,0.3581,0.8086,0.7773,0.5489,0.8974);

  (0.3375,0.6118,0.3612,0.9582,0.8794,0.6012,0.9435);
  (0.7655,0.6965,0.4891,0.5866,0.5802,0.9322,0.7853);
  (0.7088,0.8366,0.3976,0.4761,0.8881,0.6178,0.7907);
  (0.4973,0.5665,0.4797,0.8640,0.8194,0.6026,0.9818);
  (0.6618,0.6671,0.8635,0.7084,0.9051,0.7044,0.9795);
  (0.4821,0.5977,0.5186,0.8081,0.7966,0.7117,0.9167);
  (0.4821,0.5977,0.5186,0.8081,0.7966,0.7117,0.9167);
  (0.4821,0.5977,0.5186,0.8081,0.7966,0.7117,0.9167)
  第六步:计算关联度。计算结果为:
  由此得到关联序为:
  即:x6的身体健康状况要优于x7,或者说至少x6的身体健康不会差于x7;同理,的身体健康状况要优于x3,或者说至少x7的身体健康不会差于x3;以此类推,可以的得出x4的身体健康状况在所选取的个体中相对最差。
  3)关联度与图形的结合分析
  接下来画出xi对应的折线,然后结合关联度的计算结果对教职工健康状况做分析(xi对应的折线方程为:)通过程序得到xi,的折线,由于选取的十个个体过多,如果在一张图中将所有折线表示出来,可能会使得折线难以分辨,所以作者将折线图分为3个部分(x1-x3,x4-x6,x7-x10)加以分析,其中,x0是标准序列,x1至x10是比较序列。
  从图2中可以看出,x3与x0的图形走向最为接近,所以x3的健康状况在此组序列中相对最优;从图3中可以看出,x6与x0的图形走向最为接近,所以x6的健康状况在此组序列中相对最优;从图4中可以看出,x7与x0的图形走向最为接近,所以x7的健康状况在此组序列中相对最优;从图5中可以看出,x6与x0最接近,其次是x7,与x0最不接近的是x4。由此可以看出,通过折线分析的教职工健康状况与通过关联度分析的教职工健康状况结果一致。
  3 结论
  灰关联分析(亦称关联度分析)是一种因素分析方法,是各因素间发展态势的量化比较分析,它通过对系统统计数列几何关系的比较,分析系统中多因素间的关联程度。
  本文对对灰色关联分析的基本步骤进行了运用,利用邓氏关联度分析了教职工的健康状况,通过对关联的计算,量化了随机抽取的十位教职工的健康状况。
  另外,还通过灰色关联矩阵分析了各个因素对健康状况的影响程度的大小。在该实践的过程中所有的计算均通过程序实现,大大简化了计算的过程。
  但是,由于这种方法只是灰色系统理论在教职工健康评价忠的一个尝试,因此还需要在今后的工作中不断加以检验、改进和完善。
  
  参考文献
  [1]赵建.基于灰色关联度取权重的土地估计模式[J]科技创业,2007(8):46-47.
  [2]邓聚龙.灰色理论与方法[M].北京:石油工业出版社,1992.
  [3]郭娜,郭科,吴金炉,何勇.灰色关联度分析法在土地评价中的应用[J].成都理工大学学报,2007,34(6).
  [4]邓聚龙.灰色系统理论教程[M].武汉:华中理工大学出版社,1990.
  [5]刘思峰,党耀国,方志耕.灰色系统理论及其应用[M].北京:科学出版社,2004.
  [6]傅立.灰色系统及其应用[M].北京:科学技术文献出版社,1999.
  [7]易德生,郭萍.灰色理论与方法:提要,解题,程序,应用[M].北京:石油工业出版社,1992.
  [8]黄现代,王丰效.多变量灰色预测模型算法的Matlab实现[J].四川理工学院学报:自然科学版,2008,21(1):44-46


转载注明来源:https://www.xzbu.com/8/view-1684158.htm