Multi-deviation source online diagnosis system and method for multi-station assembly process of car body

Abstract

The invention relates to a multi-position car body assembly process-oriented online multiple-deviation source diagnosing system and method. The system comprises a clamp failure mode space calculation module, a multisource heterogeneous data processing module, and a recognition and diagnosis module for deviation sources at a plurality of positions. The method is based on the deviation mode definition of CAD positioning information and assembly process information and a calculation method, and realizes real-time diagnosis on the clamp deviation sources at a plurality of positions by use of a deviation space and by projecting measurement point data to the deviation space and drawing an MR control chart, wherein the deviation space is obtained by use of a vector orthogonalization method and comprises a deviation mode and a complementary space; besides, by use of the heterogeneous detection data of the key measurement point of KPC and a multisource mathematical statistic method, a data failure mode is extracted through the covariance matrix of the data, and furthermore, the correlation analysis of the mode is performed to realize diagnosis on out-of-tolerance of parts.

Description

technical field [0001] The invention relates to a deviation diagnosis system in the technical field of automobile engineering, in particular to an online diagnosis system and method for multi-deviation sources facing the multi-station assembly process of a vehicle body. Background technique [0002] Body assembly deviation source diagnosis technology has always been a difficult problem for automobile manufacturers in the process of improving dimensional accuracy. In the complex assembly process of the body, multiple error sources such as part manufacturing deviation, fixture positioning deviation, welding error, and operational defects will affect the assembly deviation of the body in white. The assembly deviation of the body-in-white has a great impact on the quality of the vehicle, the appearance quality of the body, engineering matching and market share. [0003] The traditional SPC quality control method does not detect the source of the deviation. When the out-of-tolera...

AI Technical Summary

Problems solved by technology

However, when multiple faults exist at the same time, the influence of multiple deviation sources is easily mixed in the same principal vector, so that the matching failure mode cannot be obtained, which makes it difficult to identify the deviation source, and the identification of the deviation mode largely depends on the existing The definition and calculation of the deviation mode requires high personnel experience and assembly process knowledge, which also limits its practical application in the assembly workshop

Images