Research On Comprehensive Evaluation Method Of Steel Strip Quality

Steel strip is the most important steel products, accounted for about 45% of the total steel output, which plays an important role in the iron and steel industry as the main raw material of aerospace, shipbuilding, automobiles, home appliances and many other industries. Its quality directly affects the final product quality and performance. With higher and higher requirements of user and the more competitive market, the quality of steel strip is increasingly becoming one of the determinants of the market share. In order to increasing the steel strip market share of the high value-added, high-quality products, the major steel enterprise in the world are spending huge outlays on research of the quality of the steel strip detection, evaluation and control technologies. Many major fund projects in China are supporting of such kind of research, such as National Natural Science Foundation of China (on-line detection and evaluation technologies of strip quality), the National High-Tech R&D Program of China (evaluation and reasoning technologies of strip quality based on information integration). This paper is partly supported by these project. This paper's work is a part of that research project.Whether the product quality is good or bad is related to the economic benefit, existence and development of enterprises. The scientific accurate evaluation of quality will be a directly effect on the control process to continuously improve the product quality. The most comprehensive grasp of the level of overall quality of the product is a comprehensive evaluation of the quality. It is also related to the user's needs and interests of enterprises directly.Overall quality of Steel strip consist of the mechanical properties of steel plate, shape, surface quality, size and the accuracy of certain products, such as the special requirements of different aspects of the quality together. However, the different aspects of the overall quality have different importance, and it may be significant changing because of different applications. The overall quality of this multi-faceted and the features related applications make it being more complicated for comprehensive evaluation. And the expression of testing data for evaluation index of different quality aspects is not uniform, combining the different accuracy. Such multi-character, multi-state accuracy status of quality evaluation takes a higher demand for the comprehensive evaluation to determine the weight.Taking all these problems above, the investigation results of production with expertise knowledge and quality standards, this paper will focus on the work of building a comprehensive evaluation of the quality of steel strip for some of the related issues research. The main contents of the study and the progress results are as follows:(1)For evaluation of the quality of the steel strip with more quality, coupled, nonlinear, multi-level features, the construction of the steel strip quality evaluation system is established in this study, in this system, a new type of non -linear multi-level fuzzy evaluation method adopted by the introduction of non-linear operator to reflect the evaluation of the quality of some of the prominent influence; by constructing multi-level structure of the model to reflect the impact on the quality of the multilevel nature of the factors; by the fuzzy evaluation to calculate the membership degree of the quality evaluation index. So it is possible to make a comprehensive, systematic, stability, scientific evaluation of the quality of the steel strip. The simulation results show it is feasible and effective for the system to make the comprehensive evaluation for steel strip.(2) A number of quality steel strip model is established, each sub-model representative of one aspect impact on the overall quality, these evaluation results will be important as the data of comprehensive evaluation. For different types of characteristics of quality index, corresponding sub-structure model is established. In this paper, the most complex surface quality evaluation, the more representative process performance, and mechanical properties are the sampling model for verification, which links to multi-attribute group decision-making on the surface quality evaluation model based on the defects information, the process performance evaluation model based on the synthesis of non-linear operator, the mechanical properties evaluation of the model based on improved fuzzy evaluation.(3) Research on the weight distribution related to quality index in sub-model of evaluation. As the core elements being a liaison with all levels of heterogeneous characteristics of the various sub-model or indicators, the weight directly impact on the comprehensive evaluation results. According to the diversiform problem, the weight determination method would be different. In this paper, the judgement matrix method and the maximizing deviations methods are adopted to determine the weight of the secondary evaluation sub-models.(4) Dynamic weights for first level evaluation model are established, which represents reasonable and practical logic and experience. It is the solution to the problem which previous use of a fixed weight might faced, regardless any method used to determine the fix weight, there might be some case which the result deviates actual demand. As a result of dynamic weights, the output is in line with the needs of users, thus can improve the outputted products using value and the economic benefits of. Enterprise.(5) In consideration of the problem that some standard which the quality index based on are variance and imperfect, the study of the standard which related to the strip grading evaluation of the quality is carried out. The original uniform disperse standards has been uniformed to be better serve of the comprehensive evaluation of the quality system.