东篱科研大数据发现系统（DRDS）

位置：成果数据库 > 期刊 > 期刊详情页

统计学方法在生牛奶卫生质量评价中的应用

ISSN号：1004-390X
期刊名称：《云南农业大学学报：自然科学版》
时间：0
分类：S879.1[农业科学—畜牧兽医]
作者机构：[1]华南理工大学轻工与食品学院,广东广州510640, [2]广东省燕塘乳业有限公司,广东广州510507, [3]广州市出入境检验检疫局,广东广州510623
相关基金：国家自然科学基金重点项目（20436020）;国家质检总局科技项目（2002IK004-02）.Acknowledgement： We are grateful to the National Natural Science Fund （ Project 20436020）, and the General Administration of Quality Supervision, Inspection and Quarantine of the People＇s Republic of China （Project 2002IK004 -02） for supporting our research. Thanks are also given to all the dairy workers and farmers for their cooperation.

作者：阮征[1], 曾庆孝[1], 李汴生[1], 潘眆[1], 杨爱君[2], 李小丽[3]

关键词：统计学方法, 好氧细菌总数(TABC), 贮缸生牛奶, statistical methods, total aerobic bacterial count （TABC） , bulk-collected raw milk

中文摘要：

就应用统计学方法处理生牛奶卫生检验数据的优势和细节进行了阐述。在2006年全年的观测期内，选择贮罐生牛奶的好氧细菌总数（total aerobic bacterial count，TABC）作为评价生牛奶卫生质量的重要指标进行检测和分析。首先，将原始数据进行基本的统计学整理，获取的初步信息包括时间序列图，柱状图，以及每个月数据的平均值，方差，中位数和变动范围等。其次，使用单因素方差分析和Tukey’s多重比较对各个月数据之间的差异性进行分析。分析前数据经幂转换（幂指数为0．45）以满足正态分布的统计学要求。结果显示，在1-3月，4-5月，6-9月，10月，11-12月这5个时间段内，各组数据内部没有统计学上的显著性差异，因此将数据分别合并以便寻求更大范围内的规律。最后，引入质量分析和控制图来评估各阶段生牛奶TABC的水平和稳定性。并对异常原因进行判定。在因果关系鱼骨图的帮助下，质量发生异常的相关因素可以从人员，物料，设备，操作，检验和环境等几方面进行探讨。在TABC不得高于200000 cfu／mL的限制下，只有在实施乳品安全计划的10月份，生牛奶卫生质量高，生产稳定。而其他时期都不同程度出现质量波动不稳定以致“失控”的现象。因此有必要利用统计控制手段帮助查找原因，修正错误，改善条件，稳定生产。本研究结果表明恰当的统计学方法有助于检验者和管理者把握TABC数据的变化趋势，防患于未然，而不仅仅是做一个产品质量指标合格与否的判断。

英文摘要：

In this paper, the benefits and details of applying Statistics methods on managing of the testing data of bulk-collected raw milk are presented. During a year investigating period of 2006, the total aerobic bacterial count （TABC） of bulk-collected raw milk as an important hygiene quality characteristic were collected from a dairy factory lab and analyzed in retrospect. Firstly, the basic statistics （e. g. charts, mean, median, range） of the raw data were done preliminarily. Secondly, one-way ANOVA and Tukey＇s multiple comparison were used to determine the significant difference among the groups of treated data that were normalized by idemponent transformation （ λ = 0. 45）. The comparing results showed that the difference of the data from the following periods were not statistically different respectively： [ January, March ], [ April, May], [ June-September], [ October] and [ November, December]. Then based on the 5 periods, the data no significant differences were merged to form 5 new groups for the next statistical control analysis. Finally, some control and quality charts for individuals were introduced to evaluate the TABC levels and stability of the bulk-collected raw milk, and to detect the presence of special causes. Under the upper limit TABC level of 200 000 cfu/mL, only in October when a dairy sanitation plan was practiced, the production of high-hygiene-quality raw milk was stable and effective. However, the raw milk quality during other periods showed a trend of ＂out-of-control＂. Possible factors affecting the quality, such as the material, the men （personnel）, the methods, the measurements, the machine and the environment （5M1E） were also discussed through a cause-and-effect diagram. It is necessary to apply statistical control methods to help improving and stabilizing the process. It is concluded that appropriate statistics method would help inspectors and managers to establish a trend analysis of the TABC data, and not just make a list of a Pass/Fail classification for t

同期刊论文项目