Research On Optimal Method Of Autobody Front End Matching Based Fractional Hausdorff Distance

The matching of auto-body outer panels is important not only because of its aesthetic properties but also will cause the problems of increasing wind resistance, leaking air, leaking rain and so on to largely affect the vehicle performance. There are two important reasons for this thesis. On hand, some of non-closing panels (such as headlight) in auto-body weren't adjustable before, and front end had the stationary adjusting order. So, non-closing panels were easy to adjust while the matching of closing panels (such as door and side body) is difficult to adjust, which caused most of the worldwide research focusing on the closing panels rather than non-closing panels. On the other hand, along with the need of improvement in quality of auto-body matching, most of non-closing panels is adjustable, without the fixed order leading to the adjustment process complicated and inaccuracy. Based on the reasons mentioned above, two methods, based on fractional Hausdorff distance and curvature Hausdorff distance along with the genetic algorithm, are proposed to acquire the optimal position and adjusted value. The research gets great progress in the field of the matching of non-closing panels and lays some evidences for application of tolerance design of the matching panels.First of all, the paper introduces a convenient method to acquire the contour of the matching panels. The process is scanning, de-noising, filtering, and pruning. During this process, the Comet 400 scanner, the software of Imageware and Ug are used to get the information for the measuring points.Secondly,An optimal model based on fractional Hausdorff distance is proposed to improve the gap and flushness of front end, The weighted sum of fractional Hausdorff distances among the corresponding regions is employed as the optimal matching function. In order to controlling gap and conveniently setting search band, another optimizing model based on curvature Hausdorff distance is studied The weighted sum of curvature Hausdorff distances among the corresponding regions of single scaling panel and corresponding two panels is employed as the optimal matching function. The feasibility of two methods are demonstrated by correlation analysis of two optimal matching functions and matching result.Finally, it could find the global optimal position by using the genetic algorithm. The effectiveness of the method was demonstrated by practical examples. The application scopes of two methods are introduced for practical application. At the same, the method of integral adjusting is presented because of the actual demand of the both sides adjusting.