Autonomous external pipe crawling robot for offshore inspection
Translated title
Autonom external pipe crawling robot til offshore inspektioner
Authors
Uth, Esben Thomsen ; Edling, Mikkel ; Larsen, Jannic Schurmann
Term
4. semester
Education
Publication year
2024
Submitted on
2024-05-31
Pages
189
Abstract
Denne afhandling har til formål at gøre offshore-arbejde med SubC Partners crawler mere driftssikkert og lettere at betjene ved at reducere hjulspin, forenkle operationerne og øge robotens autonomi. Crawleren blev delvist opgraderet med en elektrisk drivlinje med tre elmotorer. Der blev designet to LQR-regulatorer (linear quadratic regulator, en metode der afvejer sporingsnøjagtighed over for styreindsats) til at følge enten position eller hastighed, med gain scheduling så regulatorindstillingerne tilpasses driftsforholdene. For at begrænse hjulspin blev der udviklet en overordnet slip-regulator baseret på en sliding mode-forstyrrelsesobservatør, som estimerer forstyrrelser og justerer trækket. For at passere forhindringer blev en simpel logisk styring udviklet til at reducere crawlerens klemmetryk (den kraft, der bruges til at gribe fast i strukturen). Regulatorerne blev evalueret i simuleringer på en ikke-lineær model under varierende forhold. Resultaterne viser, at den elektriske drivlinje kan gennemføre en fuld omdrejning på en offshore-struktur med en radius på 0,54 m, selv hvis ét hjul spinner. I fysiske tests havde regulatorerne dog udfordringer ved meget lave referencehastigheder på 0,016 rad/s. Studiet konkluderer, at en yderligere gearreduktion på 3 vil give et lavere hastighedsområde; sammen med gain-scheduled regulering kan dette opfylde SubC Partners krav til hastighed og position. Det øgede tilgængelige moment vil også reducere antallet af nødvendige motorer til to.
This thesis aims to make offshore work with the SubC Partner crawler more reliable and easier to operate by reducing wheelspin, simplifying operations, and increasing the robot’s autonomy. The crawler was partially upgraded with an electric drivetrain using three motors. Two LQR controllers (a method that balances tracking accuracy against control effort) were designed to follow either position or speed, with gain scheduling so the control settings adapt to operating conditions. To limit wheel slip, a supervisory controller based on a sliding mode disturbance observer estimates disturbances and adjusts traction. To pass obstacles, a simple logic controller reduces the crawler’s clamp pressure (the force used to grip the structure). The controllers were evaluated in simulations on a nonlinear model under varied conditions. Results indicate the electric drive can complete one full revolution on an offshore member with a 0.54 m radius even if one wheel slips. In physical tests, however, the controllers struggled at very low reference speeds of 0.016 rad/s. The study concludes that adding a 3:1 gear reduction would provide a lower speed range; together with gain-scheduled control, this can meet SubC Partner’s speed and position requirements. The higher available torque would also reduce the number of motors needed to two.
[This summary has been rewritten with the help of AI based on the project's original abstract]
Documents