A sliding mode approach for control and estimation in active aircraft

Lead Research Organisation: University of Leicester
Department Name: Engineering

Abstract

Monitoring the effectiveness of any drag reduction scheme is vitally important to instil a level of confidence in its effectiveness. The active or passive drag reduction schemes will represent a 'hidden' technology on the aircraft in the sense that their effects and proper functioning will not be readily visible to the naked eye. Monitoring the efficiency of any flow control or drag reduction scheme over long periods of operation allows the pilot to estimate more accurately the required fuel reserve and assists the ground crews by indicating issues requiring inspection/maintenance which may have arisen during a flight. The objective is to utilize sliding mode observers to monitor changes in the performance of the aircraft which can be correlated to changes in the overall drag performance. Changes in drag will be formulated as unknown (and unmeasurable) changes in the aircraft system and will be estimated utilizing the equivalent output error injection signal necessary to maintain sliding. Direct development of nonlinear observers based on the nonlinear longitudinal equations of motion eliminate the errors associated with the approximations arising from linearizing - which appear to be significant when considering the level of accuracy required to estimate 1% changes in the overall drag. Sliding mode observer schemes are ideally suited to cope with such scenarios since the design process is not limited to linear system representations. All the different strategies that will be developed can be viewed as monitoring systems working independently of each other. The 'central nervous system' can then adopt a higher level monitoring objective, confirming genuine drag reduction, detecting sensor faults, and raising alarms when confronted with issues of inaccurate measurements or faults.

Publications

10 25 50
publication icon
Chee Pin Tan (2010) Robust Fault Reconstruction in Uncertain Linear Systems Using Multiple Sliding Mode Observers in Cascade in IEEE Transactions on Automatic Control

publication icon
Chee Pin Tan (2010) Sliding Mode Methods for Fault Detection and Fault Tolerant Control

publication icon
Jose Manuel Andrade Da-Silva (2010) Sliding Mode Observers for Systems With Mismatched Uncertainties

 
Description The second year of the project sought to develop an on-line parameter estimation scheme operating over a wireless network, based on sliding modes. The research was based around a high fidelity simulation of a four engine transport aircraft. An approach using a sliding mode observer involving a second order sliding mode 'super-twisting' injection signal to estimate changes in the drag coefficient was proposed. A peer-to-peer wireless network was considered in which one of the computers generates data from the high fidelity four engine transport aircraft model in real-time, which represents a 'virtual transport aircraft', and the other computer processes each data package to carry out on-line estimation of the drag coefficient. Results show that good estimation of drag (reduction) is achieved.
Exploitation Route Global warming and CO2 emission levels are issues high on both the public and political agendas. In response to the ACARE 2020 (Advisory Council for Aeronautics Research in Europe) target for a 50% reduction in emissions by the year 2020, Airbus is looking to develop technology to lower in-flight drag, thereby increasing fuel efficiency. The work in this proposal aims to develop novel control and estimation schemes, which are underpinned by rigorous theory, but are practical enough to provide implementable on-line drag estimation and fault tolerant flow control, to meet these technological challenges. This work has continued under the auspices of the ADDSAFE FP7 project. The aim of the project is to study and develop advanced model-based FDD schemes for aircraft in order to help remove technological barriers that prevent the use of innovative 'green' solutions. The idea is by improving the FDD performance in the flight control systems, it allows better optimization of aircraft structural design, weight saving and drag reduction, which in turn lowers fuel consumption and carbon emission. One example is the development of sensor fault FDD schemes to detect small incipient faults that can result in non-optimized control surface positions and therefore an increase in drag. This increases (unnecessarily) the fuel consumption over a long period of time. The work has lead to the development of adaptive super-twisting structures which can be used (usually in a bespoke way) for other estimation problems. These ideas have been applied to so-called Oscillatory Failure Case (OFC) problems in aerospace actuators. This situation occurs as a result of the generation of erroneous sinusoidal signals from faulty electronic components which propagate through the actuator control loop between the Flight Control Computer (FCC) and the control surface. When coupled with the flexible modes of the structure, the oscillations can become unacceptable and cause high vibrations and loads due to resonance phenomenon. The proposed estimation scheme is based on an adaptive sliding mode supertwist differentiator which allows the gains to adapt based on the 'quality' of the sliding motion. The FDI scheme requires an estimate of the rod speed which is provided by the adaptive supertwist differentiator. Due to the conditions in which the actuator operates, normally the differentiator gains are initialized at low values to ensure good rod speed estimation in fault free conditions. However for large amplitude/frequency OFCs, the gains must adapt in order to maintain sliding and provide a good estimation. Simulations on a high fidelity nonlinear aircraft benchmark model have been carried out for both liquid and solid OFCs in the presence of sensor and process noise. Good estimation of both actuator rod speed and the OFC is obtained, allowing detection within the specified time frame.
Sectors Aerospace/ Defence and Marine
 
Description This work has continued under the auspices of the ADDSAFE and RECONFIGURE projects. The aim of the ADDSAFE project was to study and develop advanced model-based FDD schemes for aircraft in order to help remove technological barriers that prevent the use of innovative 'green' solutions. The idea is by improving the FDD performance in the flight control systems, it allows better optimization of aircraft structural design, weight saving and drag reduction, which in turn lowers fuel consumption and carbon emission. One specific example has been the development of sensor fault FDD schemes to detect small incipient faults that can result in non-optimized control surface positions and therefore an increase in drag. This increases (unnecessarily) the fuel consumption over a long period of time. These ideas have been applied to so-called Oscillatory Failure Case (OFC) problems in aerospace actuators and tested on industrial simulators at AIRBUS' facilities in Toulouse. In the next few months one of Exeter's fault tolerant controllers from RECONFIGURE will be evaluated at AIRBUS Toulouse to address a fault scenario involving the multiple simultaneous loss of sensor information. These simulator tests represent the initial AIRBUS validation process of any new control/monitoring systems. These tests are intended to demonstrate that the controllers have the potential to solve problems of industrial relevance and allow full-time all event availability of performance optimized on-board guidance & control functions. This has impact on civil aviation safety, as well as less obvious benefits such as increased fuel efficiency.
First Year Of Impact 2015
Sector Aerospace, Defence and Marine
Impact Types Societal,Economic
 
Description RECONFIGURE
Amount € 500,000 (EUR)
Funding ID 314544 
Organisation European Commission 
Department Seventh Framework Programme (FP7)
Sector Public
Country European Union (EU)
Start 01/2013 
End 07/2016