Icing detection and identification for unmanned aerial vehicles using adaptive nested multiple models

Cristofaro, Andrea; Johansen, Tor Arne; Aguiar, A. Pedro

doi:10.1002/acs.2787

Cited by 13 publications

(5 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A path planner and ice detection system can help optimize IPS usage by improving efficiency and operation. As Wei et al [12] reviewed, ice detection is possible through sensor technology or a combination of sensors and models [13][14][15][16][17]. The paper by Wei et al applies to wind turbines, but the ice detection concepts can be adapted for UAVs.…”

Section: Introductionmentioning

confidence: 99%

UAV Icing: Icing Cases for Validation of Path Planning Method

Cheung¹,

Hann²,

Johansen³

2023

SAE Technical Paper Series

View full text Add to dashboard Cite

<div class="section abstract"><div class="htmlview paragraph">As part of the complete solution to deal with atmospheric in-flight icing on unmanned aerial vehicles (UAV), a path planner is a valuable tool for finding an optimal path for accomplishing UAV missions. When considering icing conditions, the planner manages areas with icing risk. Together with an electro-thermal ice protection system (IPS), the path planner can optimize energy consumption by comparing energy consumed flying through the cloud or around it, as the UAV can now more safely pass through the ice. The UAV’s aerodynamic stability is also considered by meeting lift requirements, producing enough thrust, and having battery capacity left. These are constraints in the planner to ensure that the UAV can complete its mission. Benchmark icing cases are constructed to validate that the path planner performs as intended. A particle swarm optimization (PSO) is used as a method in the planner due to its ability to handle highly nonlinear problems and to be able to explore the solution space effectively. Weather conditions are chosen following Federal Aviation Administration 14 CFR Part 25 Appendix C icing design envelopes required for certification. A cloud with ice is designed within a defined mission area. The constructed cloud will change size between simulations, where, by a specific cloud size, the PSO predicts that less energy is consumed flying through the iced area with an IPS than flying around it. With these icing cases, a baseline is set for future validation. The work performed in this paper will be used to validate the PSO algorithm. This paper can also benefit any UAV users that require a robust path planner by using the icing cases to identify any inconsistencies in their code. The results show that one version of the PSO handles most of these icing cases well. Inconsistencies were identified when using these icing cases, but this study makes an excellent example of how they can be used.</div></div>

show abstract

Section: Introductionmentioning

confidence: 99%

UAV Icing: Icing Cases for Validation of Path Planning Method

Cheung¹,

Hann²,

Johansen³

2023

SAE Technical Paper Series

View full text Add to dashboard Cite

show abstract

“…This approach has some similarities to methods that have been used for detection of airfoil icing, which has been studied during the last years using various methods such as model-based estimation [16], multiple-model estimation [13,14], and statistical fault diagnosis methods [15].…”

Section: Introductionmentioning

confidence: 99%

Detection and Isolation of Propeller Icing and Electric Propulsion System Faults in Fixed-Wing UAVs

Haaland

Wenz

Gryte

et al. 2021

2021 International Conference on Unmanned Aircraft Systems (ICUAS)

View full text Add to dashboard Cite

This thesis presents a fault detection and isolation (FDI) framework for detecting propeller icing, and other propulsion faults of unmanned aerial vehicles (UAVs). Such faults are among the main causes for incidents and loss of equipment.A theoretical framework for the proposed FDI is covered extensively. A tuning methodology and an implementation guide are also covered in detail. The method has been tested extensively using a software-in-the-loop simulator. The simulation results have proven to be very successful and this motivates future testing on real data sets. Relation to previous work This is a disclaimer about what is new in this thesis and what has been borrowed or reused. Project workThere will be some content-overlap between this thesis and the project work from the fall of 2020. The project work will be quoted in the relevant parts.Many of the core FDI ideas were developed during the project work. The main contribution of the project work was an in-depth analysis of those ideas. The analysis was aimed at limitations and possibilities. The project left off with a non-functioning method, with many open ends. This thesis presents a refined solution.Software documentation is also covered in this thesis. The software project was started during the project work. This was not documented in the project work. By then most of the code has been refactored many times over. Further, the main data structures were developed after the project work. The project work is therefore not quoted in the software section (7). Previous work by othersThis project builds on previous projects at NTNU. Most prominent is the use of the Ardupilot-Simulink simulator. Much of the simulator was developed by A. W. Wenz, K. Gryte, A. Winter and T. A. Johansen. The simulator is described in Section 5. My contribution to the simulator was the fault dynamics and the aircraft speed control. Some of the core ideas were originally from A. W. Wenz. Wenz never published these ideas, but the ideas are present in the project work. Thus, Wenz is credited through the project work citations.

show abstract

“…There are some applications of MMAE in aerospace engineering. Cristofaro et al [7,8] used an adaptive multiple model approach for aircraft icing detection and identifica-tion. Marschke et al [9,10] presented a generalized multiple model adaptive estimator, which uses a window of previous data via the autocorrelation matrix to perform the adaptive update, to estimate the state without rate gyros information or process noise covariance.…”

Section: Introductionmentioning

confidence: 99%

A Bayesian Adaptive Unscented Kalman Filter for Aircraft Parameter and Noise Estimation

Ding

2021

Journal of Sensors

View full text Add to dashboard Cite

This paper proposes a new algorithm for the aerodynamic parameter and noise estimation for aircraft dynamical systems. The Bayesian inference method is combined with an unscented Kalman filter to estimate the augmented states and the unknown noise covariance parameters jointly. A Gauss-Newton method is utilized to sequentially maximize the posterior likelihood function for the noise unknown parameter estimation. The performance of the proposed algorithm is evaluated and compared with two other UKFs via a flight scenario of a given aircraft. The results indicate that the proposed algorithm has equivalent performance to the simplified UKF with prior noise information and slightly outperforms the parallel UKF on precision and efficiency in this flight scenario assessment. Then, the consistency and accuracy of the algorithm are further validated by a Monte Carlo simulation with random process noise covariance. This adaptive algorithm provides another feasible and effective way for estimating aerodynamic parameters from the aircraft real flight data with unknown noise characteristics.

show abstract

Icing detection and identification for unmanned aerial vehicles using adaptive nested multiple models

Cited by 13 publications

References 32 publications

UAV Icing: Icing Cases for Validation of Path Planning Method

UAV Icing: Icing Cases for Validation of Path Planning Method

Detection and Isolation of Propeller Icing and Electric Propulsion System Faults in Fixed-Wing UAVs

A Bayesian Adaptive Unscented Kalman Filter for Aircraft Parameter and Noise Estimation

Contact Info

Product

Resources

About