Sensor integration for inflight icing characterization using neural networks

Melody, James W.; Pokhariyal, Devesh; Merret, Jason M.; Basar, M Tamer; Perkins, W.R.; Bragg, Michael B.

doi:10.2514/6.2001-542

Cited by 16 publications

(17 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This process uses estimates of the flight dynamic parameters and expected values to calculate an estimate of the ice severity [9]. Figure 4 shows the block diagram of the neuralnetwork-based icing characterization routine integrated into the flight simulator.…”

Section: Ice Management System Icing Characterizationmentioning

confidence: 99%

“…To do this an ice management system (IMS) was devised that would sense and characterize the presence of ice, notify the pilot, and ensure the safety of the aircraft. To test and demonstrate the IMS, the IEFS was designed to integrate the different aspects of the SIS project such as the flight dynamics model, autopilot [6], aircraft icing model [7,8], icing characterization routine [9], envelope protection system (EPS) [10], and human factors [11].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Icing Encounter Flight Simulator

Deters

Dimock

Selig

2006

Journal of Aircraft

View full text Add to dashboard Cite

The Icing Encounter Flight Simulator is one part of the Smart Icing Systems project at the University of Illinois at Urbana-Champaign. From the Smart Icing Systems project, an ice management system was designed that would sense and characterize ice, notify the pilot, and if necessary take measures to ensure the safety of the aircraft. The icing simulator was used as a platform to integrate and test different components of the Smart Icing System. To create an Icing Encounter Flight Simulator, functionality and Smart Icing System components were added to the FlightGear flight simulator. Functionality added to FlightGear include a reconfigurable aircraft model and an icing model, and Smart Icing System components added include an envelope protection system and a glass cockpit enhanced with ice management system features. To ensure a real-time simulation, computationally intensive processes were distributed over several desktop computers linked together through a local network. In order to demonstrate the effectiveness of the Smart Icing System components, two fictional but historically motivated icing scenarios were developed and tested with a simulated DeHavilland DHC-6 Twin Otter, specifically, a tailplane stall event during a steep descent and a roll upset event during an emergency approach. Introduction A S part of the Smart Icing Systems (SIS) project [1,2] at the University of Illinois at Urbana-Champaign, the Icing Encounter Flight Simulator (IEFS) [3][4][5] integrates various SIS components in a simulated aircraft icing environment. The SIS project was started to investigate measures that would help keep the aircraft safe during an icing encounter. To do this an ice management system (IMS) was devised that would sense and characterize the presence of ice, notify the pilot, and ensure the safety of the aircraft. To test and demonstrate the IMS, the IEFS was designed to integrate the different aspects of the SIS project such as the flight dynamics model, autopilot [6], aircraft icing model [7,8], icing characterization routine [9], envelope protection system (EPS) [10], and human factors [11].Instead of creating a new flight simulator for this project, it was decided to adapt an existing simulator. The simulator chosen was the FlightGear flight simulator (FGFS) x for its open source and modular code. Added to FlightGear were a reconfigurable aircraft model, autopilot, and an icing model. Other aspects of the SIS such as the ice weather model, EPS, ice protection system (IPS), neural-networkbased icing characterization, and the IMS-enhanced glass cockpit were integrated into the IEFS. Several components of the IEFS became too computationally intensive to run on a single desktop PC, and so to ensure real-time simulations, the IEFS was divided into different modules to be run over a local area network. The resulting distributed simulator contains six modules: flight dynamics model (FlightGear with the autopilot and icing model), SIS support code (ice weather model, EPS, and IPS), neural-network-based icing charac...

show abstract

Section: Ice Management System Icing Characterizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Icing Encounter Flight Simulator

Deters

Dimock

Selig

2006

Journal of Aircraft

View full text Add to dashboard Cite

show abstract

“…NNs also have the capability of being trained online using real data or off-line with recorded or simulated data. Several icing applications have been implemented using NNs, such as icing management system (Melody et al, 2001) and ice shape prediction (Ogretim et al, 2006). This paper, using NN Toolbox of the MATLAB software to conduct a second development program, establishes backpropagation (B-P) NNs with Levenberg-Marquardt (L-M) learning algorithm to catch the relationships of ice geometry (i.e.…”

Section: Introductionmentioning

confidence: 99%

Effects of ice geometry on airfoil performance using neural networks prediction

Cao

Yuan²,

Li³

2011

Aircraft Engineering and Aerospace Technology

View full text Add to dashboard Cite

Purpose -The purpose of this paper is to describe a methodology for predicting the effects of glaze ice geometry on airfoil aerodynamic coefficients by using neural network (NN) prediction. Effects of icing on angle of attack stall are also discussed. Design/methodology/approach -The typical glaze ice geometry covers ice horn leading-edge radius, ice height, and ice horn position on airfoil surface. By using artificial NN technique, several NNs are developed to study the correlations between ice geometry parameters and airfoil aerodynamic coefficients. Effects of ice geometry on airfoil hinge moment coefficient are also obtained to predict the angle of attack stall. Findings -NN prediction is feasible and effective to study the effects of ice geometry on airfoil performance. The ice horn location and height, which have a more evident and serious effect on airfoil performance than ice horn leading-edge radius, are inversely proportional to the maximum lift coefficient. Ice accretions on the after-location of the upper surface of the airfoil leading edges have the most critical effects on the airfoil performance degradation. The catastrophe of hinge moment and unstable hinge moment coefficient can be used to predict the stall effectively. Practical implications -Since the simulation results of NNs are shown to have high coherence with the tunnel test data, it can be further used to predict coefficients at non-experimental conditions. Originality/value -The simulation method by using NNs here can lay the foundation of the further research about the airfoil performance in different ice cloud conditions through predicting the relations between the ice cloud conditions and ice geometry.

show abstract

“…Bragg et al ( , 2001a used hinge moment sensors in order to detect icing on control surfaces. They improved a neural network model to estimate stability and control derivatives.…”

Section: Introductionmentioning

confidence: 99%

Kalman filter and neural network‐based icing identification applied to A340 aircraft dynamics

Aykan¹,

Hajiyev

Çalışkan

2005

Aircraft Engineering and Aerospace Technology

View full text Add to dashboard Cite

Purpose -The purpose of this paper is to maintain safe flight and to improve existing deicing (in-flight removal of ice) and anti-icing (prevention of ice accretion) systems under in-flight icing conditions. Design/methodology/approach -A recent academic research on aircraft icing phenomenon is presented. Several wind tunnel tests of an experimental aircraft provided by NASA are used in the neural network training. Five ice-affected parameters are chosen in the light of these experiments and researches. An offline artificial neural network is used as an identification technique. The Kalman filter is used to increase the state measurement's accuracy such that neural network training performance gets better. A linear A340 dynamic model is selected in cruise conditions. This linear model is simulated in time varying manner in terms of changing icing parameters in a system dynamic matrix. The obtained data are used in neural network training and testing. Findings -Airframe icing can grow in many ways and many points on aircraft. In this research, wing leading edge ice occurrence is only considered at the same level in both left and right wings. During ice growth other faults or anomalies are ignored. Originality/value -Existing icing sensors can only provide an indication about possible ice presence. They cannot give information of the exact level of ice. However, the efficiency of current control system of changed model decreases. The proposed technique offers a method to find out the model changes under icing conditions.

show abstract

Sensor integration for inflight icing characterization using neural networks

Cited by 16 publications

References 6 publications

Icing Encounter Flight Simulator

Icing Encounter Flight Simulator

Effects of ice geometry on airfoil performance using neural networks prediction

Kalman filter and neural network‐based icing identification applied to A340 aircraft dynamics

Contact Info

Product

Resources

About