Politecnico di Torino - Corso Duca degli Abruzzi, 24 - 10129 Torino, ITALY

+39 011 090 6100 info@tech-share.it

Global calibration of an LVM sensor network

Dimensional metrologyGeneralized least squaresInformatica Tsd EnLarge Volume MetrologyLarge Volume MetrologyLarge-sized objectsSensor network

Introduction

Method for the determination of extrinsic parameters (i.e., spatial position and orientation) of various Large-Volume Metrology (LVM) instruments (e.g., laser trackers, rotary-laser automatic theodolites, photogrammetric systems, etc.). This method (denominated “global calibration”) can be adapted to the available intruments and specific applications, simplifying the set-up stage significantly.

Technical features

Typical LVM instruments include a network of sensors, whose spatial positions/orientations can be used to lacate targets in the measurement volume. The location of the sensors of a specific instrument is generally performed in a dedicated set-up process, which is based on repeated acquisitions using specific artefacts. Unfortunately, when utilizing multiple LVM instruments, multiple instrument-dedicated calibration processes are required, making set-up time/cost inevitably increase.

The proposed global-calibration method tackles this problem, including a single acquisition stage, in which an innovative hand-held probe is re-positioned in different areas of the measurement volume. The acquisition stage is followed by a data-processing stage, in which a novel mathematical/statistical model – based on the Generalized Least Squares (GLS) method – is adopted.

Possible Applications

  • Aerospace, railway and marine industry (e.g., in the dimensional verification and assembly stages);
  • Construction of large-sized technological structures (e.g., wind turbines, tanks, telescopes, etc.);
  • Large-sized machine tools (e.g., in the volumetric-calibration and assembly stages).

Advantages

  • Encouraging the combined use of various LVM instruments;
  • Adaptable to any combination of LVM instruments;
  • Able to locate network sensors through a single sequence of acquisitions;
  • Quick convergence to the solution and determination of the relevant uncertainty.