Bogdan Løw-Hansen
Om
Main project
Zero Ice Shield
Ice protection solution for green aviation in collaboration with Ubiq Aerospace
Research interest
Operation of small fixed-wing UAVs and light aircraft in icing conditions, in particular:
- Modeling and identification of the effect of icing on propeller-based propulsion systems.
- System identification of aerodynamic models in nominal and iced conditions.
- Development of model-based change/fault detection algorithms.
Research Group
Project proposals for students (2026/2027)
- Modeling Stall Dynamics for Small Fixed-Wing UAVs in Icing Conditions
- Sensitivity Analysis of a Propeller Ice Detection Algorithm for Fixed-Wing UAVs
- Neural Network-Based Control and Icing-Aware GNC for Small Fixed-Wing UAVs
Supervised by Bogdan Løw-Hansen and Dr. Richard Hann
============================================================
Modeling Stall Dynamics for Small Fixed-Wing UAVs in Icing Conditions
============================================================
THE PROBLEM
Atmospheric icing heavily reduces the interoperability of unscrewed aerial vehicles UAVs in winter conditions, especially in countries like Noway. When ice accumulates on the wings of a fixed-wing UAV, the maximum lift coefficient decreases and the drag surges. To maintain altitude, the flight controller is forced to command a higher angle of attack, pushing the aircraft dangerously close to its stall limits. We current have identified rigid-body models that provide an excellent recreation of the linear flight regime but fail to capture the chaotic, unsteady aerodynamic responses that occur near and during a stall (critical angle of attack). To design safe mitigation strategies for extreme environments, such as icing condition, the industry needs simulation tools that can accurately predict how an iced UAV behaves at the edge of its flight envelope.
THE MISSION
This project bridges the gap between steady-state aerodynamics and stall flight dynamics. The student will be tasked with extending an existing 6-DOF fixed-wing UAV simulation environment to incorporate unsteady aerodynamic phenomena. You will analyze existing flight data and relevant literature to develop and implement stall dynamic models that can be used to replicate ice accretion related aerodynamic changes.
PROJECT OBJECTIVES
1. Conduct a literature review on unsteady aerodynamic modeling.
2. Use the the available flight data to perform system identification and develop an extension model to the current 6-DOF flight model.
3. Implement the unsteady model into a simulation environment and validate the time-domain response against theoretical expectations and the available flight data.
THE TOOLKIT (Skills you will acquire)
* Aerodynamics & Flight Dynamics
* 6-DOF Mathematical Modeling and Simulation
* System Identification concepts
* Proficiency in MATLAB
============================================================
Sensitivity Analysis of a Propeller Ice Detection Algorithm for Fixed-Wing UAVs
============================================================
THE PROBLEM
Atmospheric icing heavily reduces the operability of uncrewed aerial vehicles (UAVs) in winter conditions, especially in countries like Norway. When ice accumulates on the propellers of a UAV, they quickly lose their efficiency due to decreased thrust and increased drag. Small UAVs are particularly exposed to these conditions, and due to their strict payload limits, they typically do not have the space or weight capacity for dedicated ice detection hardware. In such cases the best solutions is to use model- and data-driven methods. State-of-the-art in-flight ice detection relies on monitoring a UAV's performance degradation; by running a mathematical model of a "clean" aircraft in parallel with the real flight, we can evaluate the residuals to detect the onset of propeller or wing icing. However, in the real world, no mathematical model is perfect. Varying operational condition and sensor noise introduce baseline inaccuracies. For industry deployment, algorithms must be robust. If the underlying clean model is off by 5%, does the detection framework still work, or does it trigger false alarms? We need to find the critical threshold of model fidelity required to keep drones safe without demanding computationally expensive, perfectly accurate models.
THE MISSION
This project focuses on the intersection of statistical signal processing, state estimation, and fault detection. First, the student will leverage experimental icing wind tunnel data to mathematically model and simulate propeller icing. By artificially extending this dataset, you will generate a diverse range of synthetic icing scenarios. Using these simulated cases, you will then stress-test an existing statistical change detection framework by systematically injecting inaccuracies into the nominal flight model. You will quantify the detection algorithm's robustness, establish the minimum model fidelity required for reliable operation, and propose adaptive thresholds to prevent false positives in turbulent or off-nominal conditions.
PROJECT OBJECTIVES
* Perform a literature review on statistical modeling and fault detection.
* Implement propeller icing models in a simulation environment based on available propeller icing experiments.
* Design a Monte Carlo simulation architecture to systematically vary the physical and aerodynamic parameters of the reference model.
* Evaluate the sensitivity of the torque and aerodynamic residuals to these underlying model inaccuracies.
* Validate the propeller icing models and the detection method against the available dataset.
THE TOOLKIT (Skills you will acquire)
* Statistical Fault Detection and Isolation (FDI)
* UAV propulsion system modeling and simulation
* Monte Carlo Simulation and Sensitivity Analysis
============================================================
Neural Network-Based Control and Icing-Aware GNC for Small Fixed-Wing UAVs
============================================================
THE PROBLEM
Atmospheric icing heavily reduces the operability of uncrewed aerial vehicles (UAVs) in winter conditions, especially in countries like Norway. When ice accretes on the wings, the UAV's flight envelope shrinks drastically: it cannot climb as steeply, maneuver as sharply, or fly as slowly without stalling. Traditional flight controllers are designed around a nominal, "clean" aircraft model. If the autopilot's Guidance, Navigation, and Control (GNC) modules are unaware of these new physical limitations, they will command maneuvers the degraded aircraft can no longer execute, leading to instability or loss of control. While we have successfully identified accurate mathematical models for both clean and iced UAVs, transitioning the control strategy between these states in real-time remains a major challenge.
THE MISSION
This project aims to close the loop between ice detection and safe flight operations. The student will be tasked with designing a Neural Network-based architecture that integrates directly into the UAV's GNC loop. Utilizing the previously identified clean and iced UAV mathematical models, you will design and train a neural network to estimate the real-time aerodynamic penalties and adapt the low-level controller. Your work will ensure that the UAV can autonomously reconfigure its control strategy, dynamically restrict aggressive commands (envelope protection), and safely maintain stability during a severe icing encounter.
PROJECT OBJECTIVES
* Conduct a literature review on Neural Network-based icing-aware GNC systems.
* Develop a neural network architecture capable of mapping the nonlinear dynamics between the established clean and iced 6-DOF UAV models.
* Design an active system that dynamically adjusts control allocation and flight envelope limits based on the neural network's real-time outputs.
* Implement the NN framework into a simulation environment and demonstrate a safe recovery and trajectory correction during an in-flight icing scenario.
THE TOOLKIT (Skills you will acquire)
* Machine Learning and Neural Networks for Control Systems
* 6-DOF rigid-body aerodyanmic modeling and simulation
* Proficiency in MATLAB and Python (PyTorch/TensorFlow)
PhD Supervisors
Dr. Richard Hann
Prof. Tor Arne Johansen
Dr. Bård Stovner (UBIQ Aerospace)
Dr. Christoph Deiler (DLR)
Publikasjoner
2025
-
Løw-Hansen, Bogdan;
Hann, Richard;
Gryte, Kristoffer;
Johansen, Tor Arne;
Deiler, Christoph.
(2025)
Modeling and identification of a small fixed-wing UAV using estimated aerodynamic angles.
CEAS Aeronautical Journal
Vitenskapelig artikkel
2024
-
Løw-Hansen, Bogdan;
Hann, Richard;
Johansen, Tor Arne;
Deiler, Christoph.
(2024)
Time-Domain System Identification and Validation of Small Fixed-Wing UAV Dynamics.
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
2023
-
Løw-Hansen, Bogdan;
Müller, Nicolas Carlo;
Coates, Erlend Magnus Lervik;
Johansen, Tor Arne;
Hann, Richard.
(2023)
Identification of an Electric UAV Propulsion System in Icing Conditions.
SAE technical paper series
Vitenskapelig artikkel
-
Løw-Hansen, Bogdan;
Hann, Richard;
Stovner, Bård Nagy;
Johansen, Tor Arne.
(2023)
UAV Icing: A Survey of Recent Developments in Ice Detection Methods.
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
-
Müller, Nicolas Carlo;
Løw-Hansen, Bogdan;
Borup, Kasper Trolle;
Hann, Richard.
(2023)
UAV icing: Development of an ice protection system for the propeller of a small UAV.
Cold Regions Science and Technology
Vitenskapelig artikkel
2022
-
Hansen, Bogdan Løw;
Hann, Richard;
Johansen, Tor Arne.
(2022)
UAV Icing: Ice Shedding Detection Methods for an Electrothermal De-Icing System.
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
Tidsskriftspublikasjoner
-
Løw-Hansen, Bogdan;
Hann, Richard;
Gryte, Kristoffer;
Johansen, Tor Arne;
Deiler, Christoph.
(2025)
Modeling and identification of a small fixed-wing UAV using estimated aerodynamic angles.
CEAS Aeronautical Journal
Vitenskapelig artikkel
-
Løw-Hansen, Bogdan;
Müller, Nicolas Carlo;
Coates, Erlend Magnus Lervik;
Johansen, Tor Arne;
Hann, Richard.
(2023)
Identification of an Electric UAV Propulsion System in Icing Conditions.
SAE technical paper series
Vitenskapelig artikkel
-
Müller, Nicolas Carlo;
Løw-Hansen, Bogdan;
Borup, Kasper Trolle;
Hann, Richard.
(2023)
UAV icing: Development of an ice protection system for the propeller of a small UAV.
Cold Regions Science and Technology
Vitenskapelig artikkel
Del av bok/rapport
-
Løw-Hansen, Bogdan;
Hann, Richard;
Johansen, Tor Arne;
Deiler, Christoph.
(2024)
Time-Domain System Identification and Validation of Small Fixed-Wing UAV Dynamics.
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
-
Løw-Hansen, Bogdan;
Hann, Richard;
Stovner, Bård Nagy;
Johansen, Tor Arne.
(2023)
UAV Icing: A Survey of Recent Developments in Ice Detection Methods.
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
-
Hansen, Bogdan Løw;
Hann, Richard;
Johansen, Tor Arne.
(2022)
UAV Icing: Ice Shedding Detection Methods for an Electrothermal De-Icing System.
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
Formidling
2024
-
Faglig foredragLøw-Hansen, Bogdan. (2024) UAV Icing: A Minimal Icing Module for Dynamic Point Mass Simulations. 2nd International UAV Icing Workshop , Trondheim 03.12.2024 - 04.12.2024
-
Vitenskapelig foredragLøw-Hansen, Bogdan. (2024) Time-Domain System Identification and Validation of Small Fixed-Wing UAV Dynamics. AIAA Aviation Forum 2024 , Las Vegas, Nevada, USA 29.07.2024 - 02.08.2024
-
PosterLøw-Hansen, Bogdan; Müller, Nicolas Carlo. (2024) Aerodynamic Modeling of Fixed-Wing UAV. Clean Aviation Evening with Rolls-Royce, Avinor & Widerøe , Trondheim 24.01.2024 - 24.01.2024
2023
-
Vitenskapelig foredragLøw-Hansen, Bogdan. (2023) UAV Icing: A Survey of Recent Developments in Ice Detection Methods. 22nd IFAC World Congress , Yokohama, Japan 09.07.2023 - 14.07.2023
-
Vitenskapelig foredragLøw-Hansen, Bogdan. (2023) Identification of an Electric UAV Propulsion System in Icing Conditions. International Conference on Icing of Aircraft, Engines, and Structures , Vienna, Austria 20.06.2023 - 22.06.2023
2022
-
Faglig foredragHansen, Bogdan Løw. (2022) Cybernetics Approach to Ice Detection on Small Fixed-wing UAVs. UAV Icing Workshop , Trondheim 29.11.2022 - 30.11.2022
-
Vitenskapelig foredragLøw-Hansen, Bogdan. (2022) UAV Icing: Ice Shedding Detection Methods for an Electrothermal De-Icing System. AIAA AVIATION 2022 Forum , Chicago, USA 27.06.2022 - 01.07.2022