棉 花 学 报    Cotton Science  2006,18(6):367-371

 

The Principle and Method of Analysis of Similarity-diffrence Connection Trend for Cultivar Regional Test in Cotton
GUO Rui-lin1,Lü You-jun 1,LIU Ya-fei1,CHEN Chuan-yin2
(1.Department of Boilogical and Food EngineeringAnyang Engineering CollegeAnyangHenan 455000,China; 2.Henan Shangqiu Vocational and Technical College 476000,China) 

Abstract:Based on the analysis method of fuzzy comprehensive evaluation and grey multi-dimension comprehensive evaluation,a kind of new comprehensive analysis method,similarity-difference connection trend analysis was put forward in cultivar regional test by using the principle and method of contact mathematics. Firstly,asertained the ideal value of various characters;secondly,calculated the similarity-difference degree of each character χlk to ideal character value χok in varieties to be evaluated and formed similarity-difference degree matrix P;thirdly,according to speciality knowledges and experiences of experts ,confirmed the weight matrix W of various characters using layer analysis method or grey relational degree method;fourthly,calculated character connection matrix U of varieties to be evaluated with ideal character set,and worked out the formula of connection degree for various varieties;fifthly,calculated connection trend value,fixd on connection trend of various varieties,and carried through the test of connection trend ;sixthly,stipulated remark set E=(Excellence,good,better,common,worse).In term of the principle of golden section,the comprehensive similarity degree was divided into 5 types. Hereby,various varieties were evaluated. Compared with fuzzy comprehensive evaluation and grey multi-dimension comprehensive evaluation method,this method had many advantages such as simple operation,easy control,obtaining more indifferently ,considering many characters comprehensively and analyzing the data in cultivar regional test more indifferently etc.
Key words:cotton;variety;regional test;analysis of similarity-difference connection trend;
           difference test    [Full Text, 1976KB]