Brad Seanor

A fault tolerant flight control system for sensor and actuator failures using neural networks

Abstract In recent years neural networks have been proposed for identification and control of linear and non-linear dynamic systems. This paper describes the performance of a neural network-based fault-tolerant system within a flight control system. This fault-tolerant flight control system integrates sensor and actuator failure detection, identification, and accommodation (SFDIA and AFDIA). The first task is achieved by incorporating a main neural network (MNN) and a set of n decentralized neural networks (DNNs) to create a system with n sensors which has the ability to detect a wide variety of sensor failures. The second scheme implements the same main neural network integrated with three neural network controllers. The contribution of this paper focuses on enhancements of the SFDIA scheme to allow the handling of soft failures as well as addressing the issue of integrating the SFDIA and the AFDIA schemes without degradation of performance in terms of false alarm rates and incorrect failure identification. The results of the simulation with different actuator and sensor failures with a non-linear aircraft model are presented and discussed.

AUTONOMOUS AERIAL REFUELING FOR UAVS USING A COMBINED GPS-MACHINE VISION GUIDANCE

The most important factors affecting the performance of a control scheme for Autonomous Aerial Refueling (AAR) for UAVs are the magnitude of the wake effects from the Tanker and the accuracy of the measurements of the UAV-Tanker distance and attitude leading to the docking. The main objective of the effort described in this paper is the implementation of a detailed modeling and simulation environment for evaluating the AAR problem. In particular, a specific control scheme based on a sensor fusion between GPS- based and Machine Vision-based measurements is proposed. Furthermore, the iterative algorithm used for estimating the position of the optical markers has been modified to be robust to a loss of visibility by one or more optical markers during the docking sequence. The paper presents the results of a detailed analysis of the AAR under different scenarios.

Online Parameter Estimation Techniques Comparison Within a Fault Tolerant Flight Control System

The results of a study where two online parameter identie cation (PID) methods are compared for application within a fault tolerant e ight control system are described. One of the PID techniques is time-domain based, whereas the second is featured in the frequency domain. The time-domain method was directly suitable for the online estimates of the dimensionless aircraft stability derivatives. The frequency-domain method was modie ed fromitsoriginalformulationtoprovidedirectestimatesofthestabilityderivatives.Thiseffortwasconductedwithin the research activities of the NASA Intelligent Flight Control System F-15 program. The comparison is performed throughdynamicsimulationswithaspecie cproceduretomodeltheaircraftaerodynamicsfollowingtheoccurrence of a battle damage/failure on a primary control surface. The two PID methods show similar performance in terms of accuracy of the estimates, convergence time, and robustness to noise. However, the frequency-domain-based method outperforms the time-domain-based method in terms of computational requirements for online real-time applications. The study has also emphasized the advantages of using ad hoc short preprogrammed maneuvers to provideenoughexcitationfollowingtheoccurrenceoftheactuatorfailuretoallowtheparameterestimationprocess.

An adaptive threshold approach for the design of an actuator failure detection and identification scheme

Typical logic schemes associated with failure detection and identification algorithms rely on a set of constant thresholds. The selection of the values for these thresholds is generally a tradeoff between the goals of maximizing failure detectability while minimizing false alarm rates. The main purpose of this brief is to propose an alternative to this conventional approach for defining the thresholds of a specific aircraft actuator failure detection and identification scheme. A specific set of detection and identification criteria for failures of the decoupled stabilators, canards, ailerons, and rudders of the NASA Advanced Control Technology for Integrated Vehicle F-15 aircraft have been formulated in terms of neural network estimates and correlation functions of the angular rates. The proposed scheme is based on the use of adaptive thresholds through the floating limiter concept. This new approach eliminates the need for parameter scheduling and has shown to be able to reduce the delays associated with the constant threshold method. The functionality of the approach has been illustrated through numerical simulations on the West Virginia University NASA Intelligent Flight Control System F-15 simulator.

On-line parameter estimation for restructurable flight control systems

Abstract This paper describes the results of a study where an on-line parameter identification (PID) technique is used for determining on-line the mathematical model of an aircraft that has sustained damage to a primary control surface. The mathematical model at post-failure conditions can then be used by a failure accommodation scheme to compute on-line the compensating control signal to command the remaining healthy control surfaces for a safe continuation and/or termination of the flight. Specific criteria for the use of an on-line PID for these critical flight conditions are first discussed. The methodology is illustrated through simulations of a fighter jet at subsonic flight conditions featuring a novel modeling procedure to characterize the post-failure/damage aerodynamic conditions. The simulations have shown the potential of this on-line PID within a fault tolerant flight control system. The results have also highlighted the importance of conducting an ‘ad hoc’ small amplitude and short-duration PID maneuver immediately following a positive failure detection to enhance the reliability of the on-line estimated parameters used in the accommodation scheme.

Neural networks-based sensor validation for the flight control system of a B777 research model

Shows the results of the analysis of a scheme for sensor failure, detection, identification and accommodation (SFDIA) using experimental flight data of a research aircraft model. Conventional approaches to the problem are based on observers and Kalman filters while more recent methods are based on neural approximators. The work described in the paper is based on the use of neural networks (NNs) as online learning nonlinear approximators. The performances of two different neural architectures are compared. The first architecture is based on a multi layer perceptron NN trained with the extended backpropagation algorithm. The second architecture is based on a radial basis function (RBF) NN trained with the extended minimal resource allocating networks (EMRAN) algorithms. The experimental data for this study are acquired from the flight-testing of a 1/24th semi-scale B777 research model designed, built, and flown at West Virginia University.

NLDI guidance control laws for close formation flight

This paper presents identification, control synthesis and simulation results for an YF-22 aircraft model designed, built, and instrumented at West Virginia University. The goal of the project was the experimental demonstration of formation flight for a set of 3 of the above models. In the planned flight configuration, a pilot on the ground maintained controls of the leader aircraft while a follower aircraft was required to maintain a pre-defined position and orientation with respect to the leader. In this paper, the identification of both a linear model and a nonlinear model of the aircraft from (light data is shown first. Then, the control laws, that feature a linear inner loop controller and a NLDI (nonlinear dynamic inversion) based outer loop guidance controller, are discussed in detail. Finally, both simulation and flight test results are presented.

Integration of sensor and actuator failure detection, identification, and accommodation schemes within fault tolerant control laws

This paper describes the design of a fault tolerant scheme for coping with both sensor and actuator failures within a flight control system. The failure detection and identification scheme is based on the use of neural estimators interfaced with correlation functions of the aircraft angular rates. Particular emphasis is placed on the differentiation between sensor and actuator failures. The failure types considered are actuator blockage along with partial/total loss of aerodynamic efficiency of the control surface and angular rate sensor step-type failure. The design of the accommodating control laws for actuator failures is based upon a non-linear dynamic inversion approach with neural network augmentation. The accommodation of sensor failures is performed by replacing the failed sensor output with neural estimates computed as part of the detection and identification process. The performance of the scheme is evaluated using the non-linear simulator for the NASA Intelligent Flight Control System F-15 aircraft developed at West Virginia University. The simulation results confirm the capabilities of the scheme to handle both sensor and actuator failures of different types over a large range of magnitudes.

Formation Flight Test Results for UAV Research Aircraft Models

This paper presents experimental results for a research program using research UAVs built and developed at West Virginia University. The main objective of this effort was to provide a flight demonstration of a formation control scheme using three YF-22 research aircraft models. For several years formation flight research has been an important topic for the aerospace community. The benefits of formation flight and development of formation control problems have been investigated and well documented. A detailed mathematical model was obtained using parameter identification techniques from previously recorded flight data of the WVU YF-22 aircraft. From this data, a Simulink® based simulation environment was developed and used to test the formation control laws. Simulation results used this mathematical model to design and refine a set of control laws to maintain desired formation geometry during maneuvered flight. The formation control law consists of a set of inner and outer loop control schemes executed by an on-board computer system on the “Follower” aircraft. Performance of the “Follower” inner-loop control laws has been assessed through a series of flight tests. Flight-testing activities focusing on the formation control laws initially used a “virtual leader” scenario. Results of the “virtual leader” testing validated the overall design of the formation controller and confirmed the performance of the on-board computer system. Formation flight with two aircraft was then performed during the 2004 flight season. This paper will first describe the UAV models along with their customized electronic computer systems. Next, the aircraft mathematical model as well as the formation control schemes will be presented. The overall control design process, with emphasis on the controller implementation on the on-board computer, will also be described. Finally, the flight-testing operations and experimental results obtained to date will be presented and discussed.

A Comparison of Optical Flow algorithms for Real Time Aircraft Guidance and Navigation

This paper focuses on the analysis of the performance of several optical flow algorithms for application within aeronautic applications such as obstacle detection and collision avoidance for lightweight unmanned aerial vehicles (UAVs). The study involved the comparison of nine different optical flow algorithms. Some of these algorithms were developed at West Virginia University (WVU); some were included in the Video and Image Processing Blockset ® of Matlab ® while others were available for download from different sources. The comparative study was performed using both a number of different real-world videos with images rotating or translating at known speeds and a Virtual Reality Environment (VRE) where an aircraft performs complex maneuvers into a detailed virtual world. The comparison of the Optical Flow (OF) algorithms is provided in terms of accuracy of estimation, and an analysis of the computational effort required by the different algorithms was also performed. As expected, algorithms belonging to common conceptual classes turn out to have similar strengths and weaknesses. However, when the computational requirements are taken into account, there is no clear “winning” algorithm, therefore suggesting that ultimately the selection of the “best” algorithm is has to be driven by the particular application.