Research On Grey System Theory And Its Application In Data Acquisition And Process Of Reverse Engineering

Reverse Engineering,based on measured data, is a technology trying to reconstruct curves and surfaces in consistent with the measured objects in a requested precision for the practice engineering. The measured data is partly information of original curves and surfaces so that there is always some uncertainty and unknown information. The interpolation and approximation methods should be used to find these unknowns when reconstructing them. It is a method of using discrete data to describe the continuousness of curves and surfaces. If a modeling process deals with the data contains some uncertain properties, we call it the grey reverse process of the original system and relatedly the model is Grey Model(GM). GM is a differential equation model which expresses the continual behavior in the system. The modeling of GM and Reverse Engineering have their similarities in inherent relations. Based on the discipline overlapping, this paper studies the two different theories. After supplementing and enriching new connotations of grey system theory, which is originally proposed by a Chinese scientist, this paper applies the theory to the data measurement and processing technology in Reverse Engineering and furthers the theory's development which is rarely applied in Mechanical Engineering. The followings are main achievements of this paper.1. After analyzing and studying the characteristics of the sequence generation of grey system theory, this paper proposes the concept of three-dimensional data sequence, extends the grey system theory from one-dimensional data sequence to three-dimensional data sequence and enlarges the application range of the theory in Engineering. This paper also proposes a decomposition coefficient selection method and a partial concave-convex characteristic decision method to the preference generation, which can solve the problem that the internal point generated by the original method was inconsistent with the actual system. Considering the errors existed in ends generation with step ratio and smooth ratio methods, the author proposes a tendency average generation and improves the quality of the generation greatly.2. Based on the practical engineering, this paper studies the grey modeling system and proposes a SGM(1,1) for the measured data of reverse engineering. To the least data modeling of GM, this paper presents an idea of"3+1"to build the grey model and demonstrates the validity of it.3. With the study and summarization of the CMM measuring methods of free-form surface, this paper proposes a grey prediction method and an optical measured data based method for the CMM to control the tracing and measuring. The methods proposed can help the CMM to acquire data from the object fast and precisely and overcome the difficulty of high precision and rapid measurement.4. After the orderly reduction methods of massive scattered data being studied, this paper proposes a partial tangent plane slicing method and a virtual second-scan line method after giving a new K-nearest algorithm to re-organize the massive data. The new proposed methods of data reduction and grey theory based unusual noisy data process can be used to generate the scan line type data and it can be directly used to reconstruct curves and surfaces. The research lays a good foundation for reconstructing the CAD model in a point- line - surface manner.