Research On Optimization Of Comprehensive Evaluation Method Based On Generalization Theory

Adopting the bidding system in the field of construction engineering can select the construction party that best meets the project requirements for the construction unit,fully embodying the principles of openness,fairness and justice.Due to the huge characteristics of construction projects,a project usually costs a lot of money.When selecting a construction unit,it is necessary to comprehensively investigate the comprehensive strength of the bidding unit.Therefore,the most widely used bid evaluation method in China is the comprehensive evaluation method.As the core link in the bidding process,the bid evaluation method will have a key impact on the bid evaluation results.In order to give full play to the advantages of the bidding system and improve the credibility of the bid evaluation results,this article From the perspective of optimization and improvement of bid evaluation methods.This article first analyzes the three commonly used bid evaluation methods in China,selects the subjective comprehensive evaluation method as the research object,and then introduces the content and role of the post-bid evaluation system.The post-bid evaluation system has just emerged in recent years,and currently only Some cities have enacted relevant laws and regulations,and the content of post-tender evaluation is not perfect,more is to analyze the problems existing in pre-tender work.In order to not only find pre-bid problems,but also provide guidance for subsequent similar projects through data analysis,this article proposes to increase the analysis of the credibility of the bid evaluation results during the post-bid evaluation process,from the perspective of reliability analysis.In-depth optimization and improvement of the comprehensive evaluation method.In order to dig deeper into the value behind the scoring data,this paper introduces the generalization theory in the field of psychometrics,and performs a reliability analysis on the subjective technical standard part.Regarding the optimization of the number of evaluation experts,first establish a random two-sided cross design model,analyze the credibility of the entire evaluation result,and then change the number of experts to infer the generalization coefficients of different generalizations.In the case of limited evaluation costs,Two methods are used to optimize the number of experts.One is based on the specific generalization coefficient value and seeks to maximize the credibility of the evaluation results.The other is based on the growth rate of the generalization coefficient.From the perspective of establishing an expert review fee subsidy system,the bidders should actively and reasonably increase the number of experts when forming a bid evaluation committee.Regarding the optimization of evaluation indicators,first,a random one-sided cross design model is established,and the credibility of the scores in each evaluation indicator dimension is analyzed one by one to find the evaluation indicators with lower generalization coefficients.The sources of error,according to different sources and sizes of errors,can be used to further refine the scoring standard of the indicator,strengthen the training of evaluation experts on the indicator,or replace it with other more appropriate evaluation indicators.Reliability analysis of existing bid evaluation cases can not only find problems before and during the bid evaluation,but more importantly,it has high practicality in the selection of the number of evaluation experts and the establishment of the evaluation index system for subsequent similar projects.Reference value.For the government,it is possible to collect the evaluation data of construction projects of the same type and scale,and on the basis of big data,on the one hand,it can analyze and summarize the number of experts and evaluation index systems suitable for different types of projects,on the other hand,it can be collated Summarize the main problems and minor problems in the bid evaluation activities,and take targeted measures to deal with the related problems through information feedback.