Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

An Embedded Foundation for HIL-Based Load Emulation and Validation of Automatic Doors

Bille, Felix (2026)
Department of Automatic Control
Abstract
This thesis presents the development and evaluation of an embedded hardware and software platform for future Hardware-in-the-Loop (HIL) based motor testing at ASSA ABLOY Entrance Systems. The motivation for the work is to reduce the number of physical door setups required for motor testing by complementing them with a configurable HIL rig capable of simulating different door types and configurations. Rather than delivering a complete HIL system, the thesis focuses on the embedded platform required for reliable realtime current control, current sensing and controller evaluation, as a foundation for future HIL development.
A custom four-layer printed circuit board (PCB) was designed and evaluated together with a dual-core STM32H745XI... (More)
This thesis presents the development and evaluation of an embedded hardware and software platform for future Hardware-in-the-Loop (HIL) based motor testing at ASSA ABLOY Entrance Systems. The motivation for the work is to reduce the number of physical door setups required for motor testing by complementing them with a configurable HIL rig capable of simulating different door types and configurations. Rather than delivering a complete HIL system, the thesis focuses on the embedded platform required for reliable realtime current control, current sensing and controller evaluation, as a foundation for future HIL development.
A custom four-layer printed circuit board (PCB) was designed and evaluated together with a dual-core STM32H745XI microcontroller. Two current sensing solutions were compared and the INA240 current sense amplifier was found to perform better than the TMCS1108 Hall-effect sensor, with a standard deviation that was roughly an order of magnitude lower for the usable ADC configurations. The ADC configuration was required to take controller timings into account, as measurement accuracy and real-time control signal calculations were constrained by the controller frequency.
Three controller structures were implemented and evaluated: a PI controller tuned with the Ziegler-Nichols frequency method combined with manual tuning, a state-feedback controller designed through pole placement and a state-feedback controller with integral action. Both controllers with integral action performed well for the locked motor case. A significant system nonlinearity was identified during the evaluation of the observer. The results suggest that both controllers compensated for this nonlinearity by relying heavily on the integral action. The nonlinearity caused the Luenberger observer to fail across all tested pole placements, as the underlying linear model assumption was not satisfied.
The platform satisfies the core requirements for a first HIL prototype, including bidirectional torque control, data logging and reference tracking using prerecorded door current data. The most critical remaining step is to investigate and resolve the system nonlinearity, likely originating from the motor driver configuration. Once this has been resolved, the platform can be used for accurate load emulation. (Less)
Please use this url to cite or link to this publication:
author
Bille, Felix
supervisor
organization
year
type
H3 - Professional qualifications (4 Years - )
subject
report number
TFRT-6320
other publication id
0280-5316
language
English
id
9251180
date added to LUP
2026-09-18 11:38:13
date last changed
2026-09-18 11:38:13
@misc{9251180,
  abstract     = {{This thesis presents the development and evaluation of an embedded hardware and software platform for future Hardware-in-the-Loop (HIL) based motor testing at ASSA ABLOY Entrance Systems. The motivation for the work is to reduce the number of physical door setups required for motor testing by complementing them with a configurable HIL rig capable of simulating different door types and configurations. Rather than delivering a complete HIL system, the thesis focuses on the embedded platform required for reliable realtime current control, current sensing and controller evaluation, as a foundation for future HIL development.
 A custom four-layer printed circuit board (PCB) was designed and evaluated together with a dual-core STM32H745XI microcontroller. Two current sensing solutions were compared and the INA240 current sense amplifier was found to perform better than the TMCS1108 Hall-effect sensor, with a standard deviation that was roughly an order of magnitude lower for the usable ADC configurations. The ADC configuration was required to take controller timings into account, as measurement accuracy and real-time control signal calculations were constrained by the controller frequency.
 Three controller structures were implemented and evaluated: a PI controller tuned with the Ziegler-Nichols frequency method combined with manual tuning, a state-feedback controller designed through pole placement and a state-feedback controller with integral action. Both controllers with integral action performed well for the locked motor case. A significant system nonlinearity was identified during the evaluation of the observer. The results suggest that both controllers compensated for this nonlinearity by relying heavily on the integral action. The nonlinearity caused the Luenberger observer to fail across all tested pole placements, as the underlying linear model assumption was not satisfied.
 The platform satisfies the core requirements for a first HIL prototype, including bidirectional torque control, data logging and reference tracking using prerecorded door current data. The most critical remaining step is to investigate and resolve the system nonlinearity, likely originating from the motor driver configuration. Once this has been resolved, the platform can be used for accurate load emulation.}},
  author       = {{Bille, Felix}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{An Embedded Foundation for HIL-Based Load Emulation and Validation of Automatic Doors}},
  year         = {{2026}},
}