1402054378

# A minimum radius of 40mm a maximum radius of 70mm and

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: al and prismatic surfaces. C1 - Spherical surface. The test on spherical surface provides robust results with just a limited number of points. The numerical results from a sampled set of 10 points demonstrate that the spherical class wins the ranking with a log-likelihood value of 6,07, a large advantage over the axial class which has a log-likelihood value of -13.44. The PDF representing the spherical radius ro is roughly centered on the value of 50mm while the set of reference parameters identifying the centre of the sphere is Ω1 ={19,99985, 30,00002, 40.00023}. C2 - Cylindrical surface. The test on the cylindrical surface provides similar good results when applied to a sampled set of 10 points. Algorithms identify the cylindrical class as the optimal description of the sampled points with a log-likelihood value of 11,00. The prismatic class holds the second position with a log-likelihood of -13,23. Statistical Modelling 175 The PDFs representing cylinder radius and height are compatible with the values of the nominal cylinder. The set of reference parameters identifying the cylinder axis shows some discrepancy, because Ω2={α=26,79582°, β=0,00002°, a=-0,00282 b=0,00024} while the nominal values should be {α=30°, β=0°, a=0, b=0}. C7 - Complex surface. The complex surface is easily identified by the proposed algorithms when the sampled set has more than 15 points. The log-likelihood of the complex model is -13,69 and slowly decreases when the number of...
View Full Document

## This note was uploaded on 08/03/2010 for the course DD 1234 taught by Professor Zczxc during the Spring '10 term at Magnolia Bible.

Ask a homework question - tutors are online