Category: Fault Tolerant Control

Fault tolerant control systems, an integrated approach to fault detection, identification, isolation and accommodation of sensor faults

Research

Fault Tolerant Air Data System

An integrated fault detection, identification, isolation, and accommodation scheme is proposed for an Air Data System with airspeed and angle of attack sensors. The system uses information from the inertial measurement unit, available air data sensors, and an aircraft digital twin model that provides virtual measurements of its aerodynamic and propulsion forces to feed a nonlinear estimator capable of detecting air sensor failure and suppressing sensor fault effects on the aircraft air data prediction.

UAV for research in Flight Mechanics and Control

A UAV was developed for research in Flight Mechanics and Flight Controls. The project involves integrating sensors with an embedded processor in a small fixed-wing aircraft and creating a flexible software infrastructure for implementing and testing flight control algorithms. The sensor suite consists of an IMU, a magnetometer, a GPS, a pitot tube with a differential pressure sensor, alpha and beta vanes, control surface position sensors, and a sensor for engine RPM. An actuator interface and RF communication link are also included for telecommand and telemetry.

Flight Data Recorder for acquisition of flight test data in a ultra light aircraft

A flight data recording system was developed for an ultralight aircraft. The aircraft was instrumented for recording data from certification flight tests and mathematical aircraft modeling.

Structure Manager, a software infrastructure for the implementation of Real Time Control Systems

Structure manager software in a Eclipse IDE

A software infrastructure for implementing real-time control systems has been developed in C language. The infrastructure, combined with a sound programming methodology (object-oriented), enables some powerful features in the real-time application, accelerating the production cycle of the software. The Structure Manager comprises a library that implements all the infrastructure functionality and a tool to automatically generate code at compilation time that connects the infrastructure with the application.

Andean Condor UAV

The Andean Condor unmanned aerial system is a tactical UAV for surveillance, exploration, reconnaissance, and intelligence. Possible applications in photogrammetry and precision agriculture are also considered. This development resulted from the project “Development of a fixed-wing unmanned aerial system for short-range missions – Aura Jr” (2007-2009).

Software for modeling, analysis and design of a fixed wing aircraft

The software has been developed to model, analyze, and design a fixed-wing aircraft. The software comprises a set of functions written in the Matlab® language and is fully compatible with Octave and a collection of simulation models developed in Simulink® that include the simulation of the nonlinear aircraft model with the flight control system. Several versions of the simulation model are more refined than others.

VISOR 3 ROV

VISOR 3 is an ROV developed by the Excuela Naval Almirante Padilla (ENAP) and Universidad Pontificia Bolivariana (UPB), through the research groups in Naval Engineering (GIIN), Automation and Design A+D and the Institute of Energy and Thermodynamics (IET), with the support of Colciencias. The ROV is an underwater robotic platform for inspecting port structures and hulls of transport vessels to comply with the ISPS code.

Aura Jr UAV

Aura Jr was a small UAV for short-range missions, the first UAV developed at the Universidad Pontificia Bolivariana. The prototype was a technology demonstrator for the project “Research and Design of an Automatic Remote Inspection System for Electric Power Transmission Lines” (2004-2005) sponsored by Interconectado Eléctrica SA (ISA), Colciencias and Universidad Pontificia Bolivariana. It led to another project, “Development of a fixed-wing unmanned aerial system for short-range missions – Aura Jr” (2007-2009).

Preliminary design of Aura UAV for monitoring of Power Transmission Lines

The Aura UAV’s preliminary design resulted from the project “Research and Design of an Automatic Remote Inspection System for Electric Power Transmission Lines” (2004-2005). Its primary objective was to investigate and design an autonomous inspection system prototype for high and extra-high-voltage transmission lines.

Adaptive Mode Transition Control Architecture With an Application to Unmanned Aerial Vehicles

A new approach to unmanned aerial vehicles’ adaptive mode transition control was proposed. The proposed architecture consists of three levels: mission planning routines are at the highest level, where the information about waypoints the vehicle must follow is processed; the mid-level controller uses a trajectory-planning component to coordinate the task execution and provides set points for low-level stabilizing controllers. The adaptive mode-transitioning control algorithm resides at the lowest level of the hierarchy consisting of a mode-transitioning controller and the accompanying adaptation mechanism. A flight demonstration was done as part of a Software Enabled Control research program (sponsored by DARPA) to validate the control algorithms using the GTmax, the Georgia Tech UAV testbed.

VISOR 2 ROV

VISOR 2 was an ROV developed as an underwater vehicle with dual control (1998-2000). Julio Cesar Correa Rodríguez, Luis Benigno Gutiérrez Zea, and Laszlo Jurko led the project.

Neural Network Control of a Flexible Link

A neural network (NN) tracking controller was implemented on a single flexible link, and the performance results of the neural network controller were compared to that of proportional derivative (PD) and proportional integral derivative (PID) standard controllers. The NN controller comprises an outer PD tracking loop, a singular perturbation inner loop to stabilize the fast flexible-mode dynamics, and an NN inner loop to feedback linearize the slow pointing dynamics. It is shown that the tracking performance of the NN controller is far better than that of the PD or PID standard controllers. No offline training or learning was needed for the NN. An extra friction term was added to the tests to demonstrate the ability of the NN to learn unmodeled nonlinear dynamics.

Aircraft Flight Control Aircraft Flight Dynamics Aircraft Flight Mechanics Control Theory Fault Tolerant Control Intelligent Control Systems

Fault Tolerant Air Data System

Summary

An integrated airspeed and angle of attack sensor failure detection identification, isolation and accommodation scheme is proposed. The system uses information from the inertial measurement unit, available air data sensors, and an aircraft digital twin that provides virtual measurements of the aircraft’s aerodynamic and propulsion forces to feed a nonlinear estimator capable of detecting air sensor failure and suppress its effect on the aircraft air data prediction. The novelty of the proposed approach is that sensor fault detection, identification, isolation, and accommodation are integrated into a feedback scheme where the information produced by fault detection is used to modulate the noise covariance of faulty sensors so that the nonlinear estimator is able to maintain the air data estimate with a small error despite the presence of various failures in the air data sensors.
The system was developed and tested in simulation. A Matlab/Simulink Ryan Navion aircraft simulation model was developed using flight test and wind tunnel data from Princeton University Flight Research Laboratory. Matlab/Simulink sensor models were developed using actual measured sensor data. A Dryden wind turbulence model was used to test the system against atmospheric perturbations. Flight simulations included climb, cruise, turns and descent maneuvers.
Independent and joint Pitot tube and angle of attack vane sensor failures were simulated. Simulation results showed that the fault tolerant estimation air data scheme is very accurate and robust against undetected or false alarm failures.

Short presentation

Related publications

[1] O. Hazbón, L. Gutiérrez, C. Bil, M. Napolitano, and M. L. Fravolini, Advances in Transdisciplinary Engineering, vol. 10, ch. Digital Twin Concept for Aircraft Sensor Failure, pp. 370–379. IOS Press, 2019. Available: http://ebooks.iospress.nl/volumearticle/52930, ISBN: 978-164368-020-0 (print) — 978-1-64368-021-7 (online), doi: https://doi.org/10.3233/ATDE190143. pdf

[2] O. Hazbón, L. Gutiérrez, C. Bil, M. Napolitano, and M. Fravolini, “Digital twin concept for aircraft system failure detection and correction,” in AIAA Aviation 2019 Forum, (Dallas, TX, USA), June 17-21, 2019. Available: https://arc.aiaa.org/doi/abs/10.2514/6.2019-2887, doi: https://doi.org/10.2514/6.2019-2887. pdf

[3] O. Hazbón, L. B. Gutiérrez, C. Bil, M. Napolitano, and M. L. Fravolini, “Review of methodologies for aircraft sensors fault detection and correction,” in AIAC18: 18th Australian International Aerospace Congress (2019), (Melbourne, Australia), pp. 259–264, Engineers Australia, Royal Aeronautical Society., Engineers Australia, Royal Aeronautical Society., February 24-26, 2019. Available: https://search.informit.com.au/documentSummary;dn=321908916273250;res=IELENG;type=pdf, ISBN: 9781925627213. pdf

Projects

Fault Tolerant Air Data System

UAV for research in Flight Mechanics and Control

A UAV was developed for research in Flight Mechanics and Flight Controls. The project consists of the integration of a set of sensors with an embedded processor in a small fixed-wing aircraft and the development of a flexible software infrastructure for the implementation and testing of flight control algorithms. The sensor suite consists of an IMU, a magnetometer, a GPS, a pitot tube with a differential pressure sensor, alpha and beta vanes, control surface position sensors, and a sensor for engine RPM. An actuator interface and RF communication link are also included to allow telecommand and telemetry.

Flight Data Recorder for acquisition of flight test data in a ultra light aircraft

A flight data recording system was developed for an ultralight aircraft. The aircraft was instrumented to record data from certification flight tests and for mathematical modeling of the aircraft.

Structure Manager, a software infrastructure for the implementation of Real Time Control Systems

A software infrastructure for the implementation of real time control systems have been developed in C language. The Structure Manager is composed of a library that implements all the infrastructure functionality and a tool to generate code automatically at compilation time, that connects the infrastructure with the application at hand. The infrastructure, combined with a sound programming methodology (object oriented like) enables some powerful features in the real time application accelerating the production cycle of the software.

Andean Condor UAV

The Andean Condor unmanned aerial system is a tactical UAV for surveillance, exploration, reconnaissance, intelligence. Possible applications in photogrammetry and precision agriculture are also considered. This development is the result of the project “Development of a fixed-wing unmanned aerial system for short-range missions – Aura Jr” (2007-2009).

Software for modeling, analysis and design of a fixed wing aircraft

Software has been developed for the modeling, analysis and design of a fixed wing aircraft. The software is composed of a set of functions written in the Matlab® m language and is fully compatible with Octave. There is also a set of simulation models developed in Simulink® that include the simulation of the aircraft non-linear model with the flight control system. There are several versions of the simulation model, some more refined than others.

VISOR 3 ROV

VISOR 3 is an ROV developed by the Almirante Padilla Naval Cadet School (ENAP) and Universidad Pontificia Bolivariana (UPB), through the research groups in Naval Engineering (GIIN), Automation and Design A+D and the Institute of Energy and Thermodynamics (IET), with the support of Colciencias. The ROV is an underwater robotic platform for inspecting port structures and hulls of transport vessels, in order to comply with the ISPS code.

Aura Jr UAV

Aura Jr was a small UAV for short-range missions. This was the first UAV developed at the Universidad Pontificia Bolivariana. This was a prototype built as a technology demonstrator for the project “Research and Design of an Automatic Remote Inspection System for Electric Power Transmission Lines” (2004-2005) sponsored by Interconectado Eléctrica SA (ISA), Colciencias and Universidad Pontificia Bolivariana. It was the starting point of the project “Development of a fixed-wing unmanned aerial system for short-range missions – Aura Jr” (2007-2009).

Preliminary design of Aura UAV for monitoring of Power Transmission Lines

The preliminary design of the Aura UAV was the result of the project “Research and Design of an Automatic Remote Inspection System for Electric Power Transmission Lines” (2004-2005) had as its main objective to investigate and design the prototype of an autonomous inspection system for high and extra high voltage transmission lines.

Adaptive Mode Transition Control Architecture With an Application to Unmanned Aerial Vehicles

A new approach to the adaptive mode transition control of unmanned aerial vehicles was proposed. The proposed architecture consists of three levels: the highest level is occupied by mission planning routines where information about way points the vehicle must follow is processed, The mid-level controller uses a trajectory-planning component to coordinate the task execution and provides set points for low-level stabilizing controllers. The adaptive mode transitioning control algorithm resides at the lowest level of the hierarchy consisting of a mode transitioning controller and the accompanying adaptation mechanism. A flight demonstration was done as part of a DARPA sponsored research program to validate the control algorithms using the GTmax, the Georgia Tech UAV testbed.

VISOR 2 ROV

VISOR 2 was an ROV developed as an underwater vehicle with dual control (1998-2000). The project was led by Julio Cesar Correa Rodríguez, Luis Benigno Gutiérrez Zea and Laszlo Jurko.

Neural Network Control of a Flexible Link

A neural network (NN) tracking controller was implemented on a single flexible link and the performance results of the neural network controller were compared to that of proportional derivative (PD) and proportional integral derivative (PID) standard controllers. The NN controller is composed of an outer PD tracking loop, a singular perturbation inner loop for stabilization of the fast flexible-mode dynamics, and an NN inner loop used to feedback linearize the slow pointing dynamics. No off-line training or learning is needed for the NN. It is shown that the tracking performance of the NN controller is far better than that of the PD or PID standard controllers. An extra friction term was added in the tests to demonstrate the ability of the NN to learn unmodeled nonlinear dynamics.