Daniel Ossmann

Diagnosis of actuator faults using LPV-gain scheduling techniques

We present a new approach for the synthesis of robust fault detection lters for the model based diagnosis of actuator faults. The underlying synthesis model is a linear parameter varying (LPV) description obtained using a combination of analytic and numerical parameter tting techniques. The actuator LPV model contains uncertain parameters which are partly measurable, thus can be used for gain scheduling, and partly not measurable, for which robustness must be ensured. The detector synthesis approach is basically a parametric nullspace method combined with min-max parameter tting. The synthesis technique is applied to the diagnosis of two classes of actuator faults for a civil aircraft: elevator runaway and elevator jamming at null position.

Energetic Approach for Control Surface Disconnection Fault Detection in Hydraulic Aircraft Actuators

This work proposes a new approach to detect the disconnected mechanical load in electro-hydraulically actuated control systems. The detector applies a Fault Model of the hydraulic actuator in which the load force is neglected. Instead of system output based analytical redundancy the model compares the energy intake of the real actuator and of the Fault Model. With this approach the disconnection between the load and the actuator can be detected, even when the difference between the real system output and Fault Model output is in order of the measurement noise. The proposed method is applicable for aircraft control surface disconnection detection with unmeasurable aerodynamic force.

A fault diagnosis based reconfigurable longitudinal control system for managing loss of air data sensors for a civil aircraft

Abstract An integrated fault diagnosis based fault tolerant longitudinal control system architecture is proposed for civil aircraft which can accommodate partial or total losses of angle of attack and/or calibrated airspeed sensors. A triplex sensor redundancy is assumed for the normal operation of the aircraft using a gain scheduled longitudinal normal control law. The fault isolation functionality is provided by a bank of 6 fault detection filters, which individually monitor each of the 6 sensors using robust low order LPV residual generators. In the case of losses of up to 5 sensors, a fault estimation technique based on LPV estimators can be employed to reconstruct the missing sensor information necessary for gain scheduling. In the worst case of a total failure of all 6 sensors, a robust constant longitudinal control law is employed which ensures a basic longitudinal control performance. The proposed control architecture fulfils the basic requirements formulated in the Benchmark Problem in the RECONFIGURE project.

LPV-Model Based Identification Approach of Oscillatory Failure Cases

The reliable early detection of oscillatory failure cases (OFC) for modern fly-by-wire controlled civil aircraft is an important aspect in optimizing the structural design objectives for reducing the environmental footprint of the aircraft. We propose a complete methodology for the design of a model based fault detection and diagnosis (FDD) system, which allows the reliable early detection of OFC characterized by low amplitude oscillations in a given frequency range. The main factors for the achieved enhanced performance are: the accurate modelling of the surface actuator via linear parameter-varying (LPV) models, an improved oscillation detection by employing the recursive discrete Fourier transform, and the integrated tuning of the free parameters of the FDD system using multi-objective optimization techniques.

Optimization based tuning of fault detection and diagnosis systems for safety critical systems

Abstract The reliable detection of faults located in the control loop of safety critical systems is an important aspect in reducing potential hazards induced by possible faults. Any fault detection and diagnosis system for a safety critical system has to fulfill strong safety specifications which can be expressed in terms of different performance criteria as detection time performance, missed detection rate, or false alarm rate. To satisfy all design requirements in the presence of unknown external signals and parametric uncertainties a tuning of the free parameters in the fault detection and diagnosis system often becomes necessary after the design of residual filters. In this paper an advanced approach to tune these free parameters based on a multi objective parameter optimization setup is presented.

Enhanced detection and isolation of angle of attack sensor faults

An enhanced detection and isolation method to monitor the state-of-practice triplex redundant angle of attack measurements on modern civil transport aircraft is presented. The developed fault detection and diagnosis architecture relies on advanced model and signal based techniques monitoring each of the three sensors individually. This allows the correct isolation of erroneous sensors also in case of multiple sensor faults. The gathered isolation information is used in an advanced sensor fusion scheme, allowing the propagation of an adequate angle of attack value to the flight control computer in case of failure. The fault detection and diagnosis system is validated using a high �delity benchmark model of a large commercial transport aircraft using different wind excitations together with challenging pilot and auto-pilot scenarios.

Advanced sensor fault detection and isolation for electro-mechanical flight actuators

Moving towards the more electric aircraft to be able to replace mechanic, hydraulic and pneumatic components of an aircraft, the aircraft industry calls for new technologies able to support this trend. One of these technologies is the development of advanced electro-mechanical actuators for aircraft control surfaces. Step by step hydraulic actuators are replaced by their electro-mechanical alternatives featuring weight and cost savings. As hydraulic actuators are used for decades by the aircraft industry, they are augmented with advanced signal and model based fault detection and diagnosis systems able to monitor the actuator and initiate adaptations in case of failures. For electro-mechanical actuators such advanced monitoring systems are still in the initial stages. In this paper, fault detection and isolation filters are designed by applying advanced residual filter synthesis algorithms to be able to monitor the sensor of electro-mechanical actuators. This paves the way for possible adaptations in electro-mechanical actuator systems in case of failures.

Fault detection and isolation of vehicle dynamics sensors and actuators for an overactuated X-by-wire vehicle

Model-based fault detection and isolation (FDI) for an overactuated mechatronic vehicle is presented. The linear single-track model is extend to reflect the layout of the overactuated vehicle as well as its longitudinal dynamics, and sensor and actuator faults are added into the model. The DLR Fault Detection Toolbox, which makes use of rational nullspace bases computation to design residual generators, is used for the systematic design of structured residuals. A minimum set of residuals is selected according to their robustness and ability to isolate faults. The fault detection and isolation system is validated in a simulation in connection with a double track model using the parameters of the ROboMObil prototype vehicle.

Enhancement of the Nonlinear OLOP-PIO-Criterion Regarding Phase-Compensated Rate Limiters

The unique dynamic coupling effects between airframe, flight control system, and the pilot of modern, highly or super-augmented aircraft introduce new stability and handling qualities problems which do not occur on conventional airplanes. In particular, phenomena leading to a destabilization of the closed-loop system consisting of the airframe, the control and stability augmentation system, and the pilot, triggered by an undesired and unexpected interaction between pilot and augmented aircraft dynamics are known to be very dangerous for the aircraft and crew. They are commonly referred to as Pilot Involved or Pilot-In-the-Loop Oscillations. This paper focuses on the enhancement of the so called Open-Loop-Onset-Point-criterion, originally developed to predict PIO-susceptibility due to nonlinear effects such as position and non-phase-compensated rate limiting. The criterion is modified to also cover phase-compensated rate limiters, which are now commonly found in modern flight control systems.

A New Flight Test Technique for Pilot Model Identification

For today’s highly augmented fighter aircraft, the aircraft dynamics are specifically tailored to provide Level 1 handling qualities over a wide regime of the service flight envelope. This requires a profound understanding of the human pilot to assure that stability margins of the airframe plus controller dynamics are sufficient to accommodate the additional pilot dynamics introduced into the system during closed loop tasks. Whereas the mathematical formulations of the airframe and controller dynamics are reasonably exact, the human pilot remains to be the most unpredictable element in the Pilot Vehicle System. In the past decades various pilot models have been developed in conjunction with analytical handling qualities and Pilot Involved Oscillations prediction criteria, mainly focusing on air-to-air tracking tasks. This paper focuses on the development of a novel flight test technique, which allows the identification of the pilot dynamics during air-to-surface aiming tasks. During an extensive flight test campaign, data was gathered and processed, using state of the art systemidentification techniques to derive a mathematical model of the human pilot during air-tosurface tracking tasks. Flight test and model-based data are compared with each other to support the validity of the developed pilot models.