Vis enkel innførsel

dc.contributor.advisorMossige, Morten
dc.contributor.authorThorkildsen, Erlend Ettersti
dc.date.accessioned2023-09-13T15:51:14Z
dc.date.available2023-09-13T15:51:14Z
dc.date.issued2023
dc.identifierno.uis:inspera:129730556:110961028
dc.identifier.urihttps://hdl.handle.net/11250/3089211
dc.description.abstract
dc.description.abstractThis master thesis investigates the feasibility of using ABB's Externally Guided Motion (EGM) with path correction to adjust a pre-programmed robot path in real-time. The research explores whether the EGM controller, along with a displacement sensor, can improve the relative accuracy of robot movements with respect to an object. The primary objective is to maintain a constant distance from the surface and correct for any skew or unevenness during the robot's path execution. The research considers the potential of this system to be used in ABB's PixelPaint applications, where high accuracy with respect to objects is crucial. The evaluation of the EGM system's performance considered two primary measures: relative accuracy and repeatability. Settling time, robot speed, and oscillations were also considered important to the performance. The performance was tested by subjecting the laser sensor to a step. Through an analysis of the results, the thesis demonstrates that the EGM system achieved high levels of repeatability for different robot speeds. However, the relative accuracy was insufficient. When using the sensor's analogue signal to correct the path, the correction suffered from oscillations. The thesis also highlights that when using the sensor as input, the correction stabilized at around half the measured displacement. The oscillations introduced a level of instability that affected the smoothness and accuracy of the robot's path. The fastest sampling rate for EGM Path Correction is 24ms and was used to maximize the robot's speed. However, when the speed exceeded 100mm/s the sensor struggled to measure a step and the robot's step response did not have time to settle before the trajectory ended. In summary, this thesis contributes to the understanding of the use of external sensors for real-time path correction in robotics by evaluating ABB's EGM Path Correction. It highlights the constant displacement the controller did not correct, and it considers the presence of oscillations and the limitations imposed by the maximum sampling rate. These challenges require more research to improve the performance before EGM Path Correction could be applied to applications such as PixelPaint.
dc.languageeng
dc.publisheruis
dc.titleBruk av ekstern sensor for å korrigere robotbane i sanntid
dc.typeMaster thesis


Tilhørende fil(er)

Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel