Gantry Crane Path Planning and Control
Authors
Jensen, Jonathan Midtgaard ; Mignon, Malthe Lang ; Monroy Bjørn, Alberto Martin
Term
4. semester
Education
Publication year
2023
Submitted on
2023-10-02
Abstract
Rapporten adresserer automatisering af portalkraner i containerhavne og undersøger, hvordan en 2D-overheadkran kan planlægge sikre, effektive baner og styres præcist med begrænset sving under praktiske hardwarebegrænsninger. Arbejdet omfatter en gennemgang af state-of-the-art inden for baneplanlægning og kranstyring, opbygning af en matematisk model og 3D-repræsentation af den tilgængelige kran samt en detaljeret opsætning af elektronik (styreboks, Arduino, motorstyringer, sensorer). Der foreslås en samlet løsning, der integrerer en Model Predictive Controller (MPC) med en potentialefelt-baseret baneplanlægger, suppleret af målenormalisering, linearisering og validering af bl.a. friktionskoefficient og model. Løsningen afprøves gennem kravspecifikation og acceptance tests, og to baneplanlægningsalgoritmer sammenlignes; der gives en diskussion og anbefalinger til videre forbedringer. Uddraget angiver ikke kvantitative resultater, men dokumenterer metode og testramme.
The thesis addresses the automation of yard gantry cranes and asks how a 2D overhead crane can plan safe, efficient trajectories and be controlled accurately with limited sway under real hardware constraints. The work includes a review of state-of-the-art path planning and control for cranes, derivation of a mathematical model and 3D representation of the available crane, and a detailed hardware setup (control box, Arduino, motor drivers, sensors). A complete solution is proposed that integrates a Model Predictive Controller (MPC) with a potential-field-based path planner, supported by measurement normalization, linearization, and validation of the friction coefficient and model. The solution is tested against a set of requirements via acceptance tests, and two path-planning algorithms are compared; a discussion and recommendations for further improvements are provided. The excerpt does not report quantitative performance results but documents the methods and testing framework.
[This summary has been generated with the help of AI directly from the project (PDF)]
Documents