3D Laser Scanning Pointcloud Processing Uncertainty Estimation for Fuel Tank Volume Calibration

Abstract

The implementation of modern 3D scanning method for horizontal fuel tanks’ calibration applications proposes versatility and advantages over the currently used volumetric method which requires using of liquid filler (in most cases water, and in some countries—gasoline). However, the data scatter in the 3D pointcloud deteriorates the accuracy of the volume calculation algorithm output and therefore increases the fuel tank volume measurement uncertainty. In this paper, the preprocessing method of the 3D pointcloud data is defined and the application for calibration of the horizontal fuel tanks is discussed. The uncertainty analysis and estimation of the horizontal fuel tank calibration is performed in this work implementing 3D pointcloud FIR filtering and regression preprocessing techniques. The mathematical model of the volume measurement and graduation was estimated, and the uncertainty sources were identified. The horizontal fuel tank graduation table is developed, and fragments of the calibration results using the scanning data from four fuel tanks with recommendations for the 3D pointcloud data processing are presented in this work.