Swing-Up and Stabilization of a Cart Pendulum and Twin Pendulum System: Using Nonlinear Control Strategies
Author
Vestergaard, Niels Skov
Term
4. term
Education
Publication year
2019
Submitted on
2019-01-16
Pages
76
Abstract
Dette projekt udvikler ikke-lineære reguleringsstrategier til en vogn-pendel-opstilling i Control and Automation Lab ved Aalborg Universitet (AAU). Målet er at svinge pendulet fra nedhængende til opret stilling, fange det og holde balancen. Del 1 omhandler et enkelt pendul på vognen. Her designes tre energibaserede swing-up-regulatorer (metoder, der udnytter systemets energi til at svinge pendulet op). En sliding mode-regulator (en robust styringsmetode) udvikles til at fange og stabilisere pendulet omkring den oprette position. Én af swing-up-metoderne implementeres på forsøgsopstillingen sammen med sliding mode-regulatoren. Del 2 udvider til et tvillingependel (to koblede penduler). Erfaringerne fra del 1 bruges til at udvikle en swing-up-strategi, og en lineær-kvadratisk regulator (LQR, en optimal stabiliseringsmetode) designes til at holde systemet stabilt. Regulatorerne implementeres, og der konstrueres et Kalman-filter til at estimere de umålte tilstande (fx vinkler og hastigheder) i tvillingependel-systemet.
This project develops nonlinear control strategies for a cart-and-pendulum setup in the Control and Automation Lab at Aalborg University (AAU). The goal is to swing the pendulum from hanging down to upright, catch it, and keep it balanced. Part 1 addresses a single pendulum on the cart. Three energy-based swing-up controllers (methods that use the system’s energy to drive the pendulum upward) are designed. A sliding mode controller (a robust control technique) is developed to catch and stabilize the pendulum near the upright position. One of the swing-up strategies is implemented on the test rig together with the sliding mode controller. Part 2 extends to a twin pendulum (two coupled pendulums). Insights from Part 1 are used to develop a swing-up strategy, and a Linear Quadratic Regulator (LQR, an optimal stabilizing controller) is designed to keep the system balanced. The controllers are implemented, and a Kalman filter is designed to estimate unmeasured states (such as angles and velocities) of the twin pendulum system.
[This abstract was generated with the help of AI]
Keywords
Documents