Mathematical modeling of pipeline features for robotic inspection.

摘要

Underground pipeline systems play an indispensable role in transporting liquids in both developed and developing countries. The associated social and economic cost to repair a pipe upon abrupt failure is often unacceptable. Regular inspection is a preventative action that aims to monitor pipe conditions, catch abnormalities and reduce the chance of undesirable surprises. Robots with CCTV video cameras have been used for decades to inspect pipelines, yielding only qualitative information. It is becoming necessary and preferable for municipalities, project managers and engineers to also quantify the 3-D geometry of underground pipe networks. Existing robots equipped specialized hardware and software algorithms are capable of scanning the interior geometry of pipelines. Improvement in the 3-D models created from the collected data is a prerequisite for true, quantitative assessment of underground pipelines to take hold.;Many issues regarding pipeline scanning and geometry modeling remain unaddressed or unsolved. The ultimate goal of this research is to target several prominent topics related to the robotic inspection and parametric modeling of pipe geometry, filling gaps in the literature needed for more quantitative pipeline assessment.;First, parametric models of a circular cylindrical pipe undergoing deformation are developed. Different shape patterns that develop for typical pipe deformation pathways can be mathematically expressed using a single parameter. This technique offers convenience in generating or fitting 3-D models of pipes whose cross sections vary along the pipe length, where cross sections can consist of combinations of continuous and discontinuous circular and/or elliptical arcs. The parametric model is applied to the ASTM F1216 pipe liner design standard to improve the estimation of pipe ovality.;Second, the impact of robot length, wheel span and wheel radius on the offset between the pipe origin and the origin of the robotic measurement hardware is quantified; this is important because the interpretation of data collected from camera, radar systems and ultrasonic sensors depends on the location of the hardware inside the pipe. Geometry distortions resulting from the passage of a robot through a pipe bend are simulated to demonstrate errors that can arise in cross sectional pipe measurements.;Third, an algorithm is proposed to compute the pitch, yaw and roll of a robot as well as the major and minor axis of a pipe based on laser ring measurements taken from a single end of the robot. An enhanced version of an existing double-ended measurement algorithm is presented to reduce error when pitch, yaw and roll are large.;Fourth, the relationship between geometry measurement and image processing is explored. A template-guided lateral detection paradigm using homogeneous geometric transformations and the Discrete Fourier Transform is proposed and evaluated according to error arising from lateral size, camera position and camera orientation. Relatively large laterals resembling an eclipse are easier to detect than small ones.;Fifth, a new parametric model of the shape assumed by a flexible pipe liner encased in an elliptical host pipe is presented. This model overcomes deficiencies in existing models by correctly accounting for continuity in the slope and curvature of the liner profile.