Experimental Result Comparisons of Curve Fitting Algorithms on Fluid Path Lines Modeling in Strengthen Grinding Flow Field

Abstract

Stirred turbulence with abrasive particles were widely employed for strengthen grinding in mechanical industries, therefore, the dynamic characteristic of turbulent flow comes up as a key element influencing grinding mechanism and operation performance in a long term. In this experimental research, we use flow path lines (turbulent path lines) to demonstrate the specific moving process of turbulent flow during strengthen grinding operation, which can be clearly identified by the real-time spatial movements of those abrasive particles, therefore a visual experimental approach can be provided to analyze the three-dimensional motion characteristics of strengthen grinding turbulence. The free movements of abrasive particles in three-dimensional grinding flow field exert an important influence on the geometrical properties of fluid path lines to be fitted, and the influence mechanism caused by those fluid path modeling algorithms should be studied quantitatively in detail. In this article, after extracting the image coordinates of those selected abrasive particles denoted as physical control points, and determining the three-dimensional moving path of turbulent flow by using imaging head array mounted on turbulence monitoring platform, several typical algorithms of curve fitting were sequentially employed for constructing the spatial path lines of turbulent flow, using coordinate transformation and flow motion vector computations. On the basis of computing the newly proposed curve mathematical properties, we made an experimental result comparison of those curve fitting algorithms on the established fluid path models. With several proposed experimental suggestions concerning turbulence inspection, flow path modeling processes and their correlated experimental mechanisms, the performance inspection of strengthen grinding turbulence experiments will be highly facilitated and remarkably optimized. Thereafter, new ideas for studying strengthen grinding process in future were also provided.

Video tracking for high-similarity drug tablets based on reflective lightness intensity and fuzzy recognition system

“…Fuzzy recognition system is an effective tool to analyse the correlation between those targeted objectives with less data and identify many uncertain factors in practice. 27,28 The idea of fuzzy set allows imprecise and unqualified information to be expressed in an exact way and has a membership function taking all values between 0 and 1. Since tablet tracking was influenced by many unpredictable factors or interferential signals come from practical manufacturing environment, which weaken its accuracy and reliability when traditional searching methods were used, fuzzy recognition system paves its own way with higher efficiency and accuracy.…”