Optimization Of Key Elements Of Comprehensive Evaluation And Its Application

In recent years,comprehensive evaluation theory has developed rapidly and has been widely used in various practical fields.In this context,the accuracy of evaluation is particularly important.Comprehensive evaluation is a complex and systematic process,which consists of multiple evaluation elements.In the evaluation,there are often multiple options to choose from for the same evaluation element,and selecting the option with high excellence can effectively improve the accuracy of the evaluation results.Therefore,it is necessary to measure the excellence of these evaluation elements,and then to select them.Based on this,according to the sequence of comprehensive evaluation,this thesis will study the optimization of three key elements:dimensionless method,index system and comprehensive evaluation method.Firstly,from the perspective of original information retention of index,the information quantity measure and the relative difference information measure are constructed,and the preferential analysis of common dimensionless methods is conducted through theoretical derivation.It is concluded that the extension class of linear dimensionless method does not change either the amount of information or the relative difference information of the original index data under the condition that the original index valuex_ij≥0.The conclusion was also verified by generating random numbers to simulate the index data under different distributions.Secondly,in the process of constructing a comprehensive evaluation index system,the differences in the final index system are usually caused by two parts:one is the difference in the evaluator’s subjective understanding of the evaluation problem;the other is the difference in the index system caused by different index screening methods based on the index data.The research object of this thesis is the latter,and the optimization measures are constructed from three aspects:information contribution rate,redundancy and differentiation,whose measure values can reflect the excellence of the index system in different aspects after after index screening.Finally,the relative excellence measure of the comprehensive evaluation method was constructed based on the evaluation value deviation and Sperman rank correlation coefficient.It considered both the evaluation value and the evaluation rank order,and formed a optimization strategy steps of comprehensive evaluation method together with the compatibility test.The green development index system of coal mines was also constructed to evaluate the green development level of six coal mines under China Pingmei Shenma Group（CPSG）.In the evaluation process,the research results of this thesis are used to verify its feasibility and effectiveness.The research in this thesis enriches the comprehensive evaluation theory to a certain extent and has some reference significance for the application of comprehensive evaluation.There are 5 figures,32 tables and 80 references.