Neural Network based Sensor Fault Detection for Flight Control Systems

Singh, Seema; Murthy, T. V. Rama

doi:10.5120/9605-4232

Cited by 6 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The advantage of using relevant methods when compared to model-based ones in the literature such as Kalman filter and Luenberger observer is that relevant methods are model-free and it has error tolerance. Therefore, this machine learning methods can easily be used for non-linear systems such as aircrafts without the obligation of having certain assumptions (Singh and Vasudeviah, 2012). In addition, the proposed approach presents more preciseness, fast fault detection and reconstruction without increasing system complexity, computational power and cost and it is suitable fast real time operations (Turkmen, 2018; Taimoor and Aijun, 2019).…”

Section: Methodsmentioning

confidence: 99%

Aircraft air data system fault detection and reconstruction scheme design

Kilic

Unal

2021

AEAT

View full text Add to dashboard Cite

Purpose The purpose of this study is to detect and reconstruct a fault in pitot probe and static ports, which are components of the air data system in commercial aircrafts, without false alarm and no need for pitot-static measurements. In this way, flight crew will be prevented from flying according to incorrect data and aircraft accidents that may occur will be prevented. Design/methodology/approach Real flight data collected from a local airline was used to design the relevant system. Correlation analysis was performed to select the data related to the airspeed and altitude. Fault detection and reconstruction were carried out by using adaptive neural fuzzy inference system and artificial neural networks, which are machine learning methods. MATLAB software was used for all the calculations. Findings No false alarm was detected when the fault test following the fault modeling was carried out at 0–2 s range by filtering the residual signal. When the fault was detected, fault reconstruction process was initiated so that system output could be achieved according to estimated sensor data. Practical implications The presented alternative analytical redundant airspeed and altitude calculation scheme could be used when the pitot-static system contains any fault condition. Originality/value Instead of using the methods based on hardware redundancy, the authors designed a new system within the scope of this study. Fault situations that may occur in pitot probes and static ports are modeled and different fault scenarios that can be encountered in all flight phases have been examined.

show abstract

Section: Methodsmentioning

confidence: 99%

Aircraft air data system fault detection and reconstruction scheme design

Kilic

Unal

2021

AEAT

View full text Add to dashboard Cite

show abstract

“…The advantage of using relevant methods when compared to model-based ones in the literature such as Kalman filter and Luenberger observer is that relevant methods are model-free and it has error tolerance. Therefore, this machine learning methods can easily be used for non-linear systems such as aircrafts without the obligation of having certain assumptions [37]. In addition, the proposed approach presents more preciseness, fast fault detection and reconstruction without increasing system complexity, computational power and cost and it is suitable fast real time operations.…”

Section: Training By Adaptive Neural Fuzzy Inference Systemmentioning

confidence: 99%

Sensor fault detection and reconstruction system for commercial aircrafts

Kilic

Unal

2021

Aeronaut. j.

View full text Add to dashboard Cite

The aim morphing of this study is to detect and reconstruct a fault in angle-of-attack sensor and pitot probes that are components in commercial aircrafts, without false alarm and no need for additional measurements. Real flight data collected from a local airline was used to design the relevant system. Correlation analysis was performed to select the data related to the angle-of-attack and airspeed. Fault detection and reconstruction were carried out by using Adaptive Neural Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), which are machine-learning methods. No false alarm was detected when the fault test following the fault modeling was carried out at 0–1 s range by filtering the residual signal. When the fault was detected, fault reconstruction process was initiated so that system output could be achieved according to estimated sensor data. Instead of using the methods based on hardware redundancy, we designed a new system within the scope of this study.

show abstract

“…The problem with the analytical approach is that most industrial systems cannot easily be modelled due to their sheer size, complexity, unavailability of component data of the design, measurements being corrupted by noise and unreliable sensors within the control system. Owing to this, a number of researchers have focussed their research on the use of neural networks to produce models of industrial processes [2], [4], [5], [6], [8], [13]- [16], [19]- [23]. This is due to the fact that neural networks have the ability to filter out noise and disturbances, thus providing a stable and highly sensitive model of an industrial system without the use of a mathematical model.…”

Section: Introductionmentioning

confidence: 99%

Neural Network Fault Diagnosis System for a Diesel-Electric Locomotive's Closed Loop Excitation Control System

Barnard¹,

Niekerk²

2018

SAIEE Afr. Res. J.

View full text Add to dashboard Cite

In closed loop control systems fault isolation becomes extremely difficult in the case of feedbacks being oscillatory due to corrupted signals or malfunctions in actuators. This paper investigates and highlights the development of an off-line fault detection and isolation system for the isolation of faults, which cause oscillatory conditions on a General Electric (GE) Diesel-Electric Locomotive's excitation control system. The paper illustrates the use of artificial neural networks as a replacement to classical analytical models used for residual generation. The artificial neural network model's design is based on model-based dedicated observer theory to isolate sensor, as well as component faults, where observer theory is utilised to effectively select input-output data configurations for detection of sensor and component faults causing oscillations. Residual Evaluation is done with the use of a moving average filter incorporated with the simple thresholding technique. The results indicated 100% accuracy for the detection and isolation of the component or sensor responsible for causing excessive oscillation in the excitation control system.

show abstract

Neural Network based Sensor Fault Detection for Flight Control Systems

Cited by 6 publications

References 16 publications

Aircraft air data system fault detection and reconstruction scheme design

Aircraft air data system fault detection and reconstruction scheme design

Sensor fault detection and reconstruction system for commercial aircrafts

Neural Network Fault Diagnosis System for a Diesel-Electric Locomotive's Closed Loop Excitation Control System

Contact Info

Product

Resources

About