Abstract
With manufacturing shifting from traditional products to high value products, the complexity and accuracy of the products are increasing in order to reduce energy costs, create friendly environment and better health care. Structured surfaces, freeform surfaces, and other functional engineering surfaces are becoming the core part of high value manufacturing products. However, measurement of these surfaces is becoming very difficult due to instrumental limitations including measurement range, speed, resolution and accuracy. Multi-instruments/sensors measurement are now being developed for freeform and structured surface assessment, which requires the fusion of the data into a unified system to achieve larger dynamic measurements with greater reliability. This paper discusses the process of combining data from several information sources (instruments/sensors) into a common representational format and the surface topography can be reconstructed using Gaussian processes and B-spline techniques. In this paper the Gaussian process model is extended in order to take into account the uncertainty propagation and a new data fusion model based on least squares B-splines that drastically reduce the computational time are presented. The results are validated by two for freeform surface measurements.
Highlights
In modern manufacturing technology, manufactured surfaces are characterized by complex features, designed to meet functional specifications
A formula to compute the prediction variance for the Gaussian Process (GP)-based fusion model proposed in Ref. [7] was proposed
The advantages of the proposed least squares B-splines approximation (LSBA)-based fusion model were firstly explored with artificially generated data, the prediction performances of the proposed model were compared to the GPbased fusion model with a real test case
Summary
In modern manufacturing technology, manufactured surfaces are characterized by complex features, designed to meet functional specifications. Precision Engineering 60 (2019) 570–586 technique in order to align point clouds coming from two different sensors In this case, the main idea is to reconstruct the information provided by the Lo-Fi sensor with a Bayesian Gaussian Process (GP) model. The authors proposed to use different covariance functions to estimate different features composing the surface: smooth or rough In their following work Yin et al [10] proposed and intelligent sampling plan for the data fusion model based on the uncertainty of the prediction. If the number of measured point is high, it may be not possible to estimate the model parameters’, due to the limited CPU memory To overcome these issues a data fusion model based on B-splines approximation is proposed. The paper is structured as follow: in Section 2 the data fusion models are described, in Section 3 the models are applied to real test cases and in Section 5 conclusions and future developments are given
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
