Accumulated error control technique applied in small mass production of piece;

This paper introduces the principle of the 3D Optix laser scanner briefly,discusses the method of improving the data accuracy in single piece measurement and reducing the accumulated error of spliced joint in multi-pieces measurement with an applied example.

Firstly,the nonlinear problem and the cause of accumulated error for inertial strapdown attitude solution were analyzed.

With the purpose to reduce influence of accumulative error from the recurrence formula of residual stress measured with magnetic stress meter on the calculation result,based on the consideration that the stress at a same measuring point has no relation to the measuring order,the author proposed an improved calculation method for the work piece with free boundaries on both ends.

The problem of the error accumulation,three different kinds of linear size marking form is analysed here.

Image alignment error is classified as type I error and type II error in this paper for the error accumulation and image mismatch of microscopic image mosaicing,and a global image alignment model of image mosaicing is proposed to deal with these two types of error.

(1) Study on video mosaic algorithm for 1D scan patternTo overcome the problem of registration error accumulation in traditional registration algorithms, this paper presents a fast and accurate video mosaic approach for 1D scan pattern.

The simulated results show that when muzzle velocity is zero the cumulative error induced by acceleration of 0.

Through the study, it is found that the equation is fine and simultaneously the step used in usual flight simulation can meet the command of step determined by the absolutely stable domain of the equation, and the main source of the error lies in cumulative errors.

This paper makes further research on the error,caused by use of a single step or a variable step of Runge Kutta to solve the Euler equation of quaternion method, find that the large error results from cumulated error with single step, and when using the variable step,the cumulative error is less.