Tentative workshop outline

(List below is not in the final sequence of presentation)

8:30am - 6:00pm - Renaissance Washington, DC Downtown Hotel, USA

Workshop Abstracts

(List below is not in the final sequence of presentation)

Introduction to Fault-tolerant Control and Cooperative Control: Motivation, Concept, History, Existing and Future Developments, and Applications to a Multiple Quadrotor UAVs Testbed(Dr. Zhang)

Unmanned systems including Unmanned Aerial Vehicles (UAVs) are gaining more and more attention during the last few years due to their important contribution and cost effective application in several tasks such as surveillance, search, rescue, military and security applications. A team of researchers at the Department of Mechanical and Industrial Engineering of Concordia University, with the support from three Canadian-based industrial partners (Quanser Inc., Opal-RT Technologies Inc., and Numerica Technologies Inc.), have been working on a Networked Fault-Tolerant Cooperative Autonomous Vehicles (NFTCAV) research project as well as for "Flight Control Systems" and "Fault Diagnosis and Fault Tolerant Control Systems" courses teaching using multiple quadrotor helicopter UAVs. The main objective of the project is to provide theoretical and experimental results on on-line and on-line UAV modeling, cooperative decision-making and tasks assignment, trajectory and path planning, formation flight, fault diagnosis and fault-tolerant control, and at the same time to transfer quickly the research outcomes to the undergraduate and graduate courses teaching. A set of unmanned vehicles testbeds with several quadrotor UAVs have been built at the Department of Mechanical and Industrial Engineering of Concordia University based on the financial support of NSERC (Natural Sciences and Engineering Research Council of Canada) since 2007, with the help of Quanser Inc. for the testbed development.

In this presentation, brief introduction to the concept on fault-tolerant control and cooperative control will be given first. Historical development and new challenges in this active research area will be outlined. An overview of our past, current and future research activities and research outcomes on fault diagnosis, fault-tolerant control, path and trajectory planning/re-planning and cooperative control with applications to unmanned systems including the quadrotor helicopter UAV, NASA's GTM fixed-wing UAV and an Airbus A380 model UAV, will be presented.

Health Management for Persistent Surveillance: Theory and Practical Results (Dr. Rabbath)

We present the problem of coverage of an area with a team of aerial robotic drones. The drones self-position themselves to maintain coverage, thus removing the burden of multi-vehicle control and management from the human operator. The drones have the ability to adapt their position in case adverse events take place during the course of an operation. Adverse events include the loss of one or more robots, inter-vehicle communication problems, intruders entering the area being monitored, and loss of effectiveness of one or more robots. We present the step-by-step design of such intelligent system, and importantly illustrate the performances obtained by means of indoor, controlled experiments with a small team of quadrotor drones. Videos are an integral part of the presentation. Hardcopy notes will be given to the audience. In this presentation, the concept of health management is defined, and current systems enabling team coordination in case of health problems are discussed. Part of the presentation relates to the first book on safety and reliability for teams of robotic drones recently published by the speaker, and entitled "Safety and Reliability in Cooperating Unmanned Aerial Systems".

Iterative Design Towards Improved Fault Tolerance: A Framework for Improved SUAS Airworthiness(Dr. Chen)

In order to guarantee the airworthiness of a SUAS, there are some redundancies that need to be implemented in the design of UAS. But too many redundancies place a hard condition on the payload of UAS. This presentation aims at providing recommendation on what kind of faults in actuators are forbidden that we should make a backup in the design of UAS and what kind of faults are allowed without affecting the performance of UAS. It is common that when design a feedback controller the physical property of system are often overlooked. In this presentation, we put the 'physics' of UAV back in the design of fault tolerant controller for a fixed-wing test-bed and we try to find what are the maximum faults that can be tolerated in this kind of UAS. The results presented are intended to support the ongoing discussion on airworthiness and SUAS integration into the National Airspace System. Simulation results with different faults are also presented to validate the effectiveness of the presented fault tolerant controllers and other related airworthiness techniques.

Sliding Mode Schemes for Fault Detection and Fault Tolerant Control (Dr. Edwards)

Sliding mode methods have been historically studied because of their strong robustness properties to a certain class of uncertainty. This is achieved by employing nonlinear control/injection signals to force the system trajectories to attain in finite time a motion along a surface in the state-space. The associated reduced order dynamics, whilst constrained to the surface is called the sliding motion, and possess strong robustness properties. This talk will consider how these ideas can be exploited for fault detection (specifically fault signal estimation) and subsequently fault tolerant control. The talk will also describe an application of these ideas to aerospace systems. It will describe flight simulator results associated with the EL-AL 1862 Bijlmermeer scenario studied as part of the GARTEUR AG16 action group on fault tolerant control. The controller design was carried out without any knowledge of the types of faults/failures occurring on the aircraft, and employs sliding mode methods. The results demonstrate the successful real-time implementation of the proposed fault tolerant control scheme on a motion flight simulator configured to represent the EL-AL aircraft.

Tools for Teaching Autonomous Unmanned Vehicle Systems (Mr. Fulford)

Unmanned Vehicle Systems (UVS) are growing in popularity across a broad spectrum of applications such as search and rescue, military, mining, and environmental surveillance. Likewise, the UVS research community is growing and there is an increasing demand for novel hardware and software platforms on which to develop and test UVS algorithms and controllers.

To meet the growing demand for new technologies to teach and develop the next-generation unmanned systems, this workshop presents the latest technologies for UVS teaching and research. As part of this workshop, we will review how leading universities have integrated autonomous unmanned systems into their teaching and research programs using this state-of-the-art rapid controls prototyping framework and open-architecture data acquisition hardware designed for unmanned systems. This workshop will also demonstrate how innovative hardware-in-the-loop systems can be used to augment virtual 3D UVS missions in order to teach fundamental control concepts while motivating students with exciting, real-world UVS applications. More advanced concepts will be introduced with specific focus on tools for autonomous unmanned vehicle systems. Demonstrations will show autonomous unmanned vehicle missions planned out and executed in simulations with rendered 3D visualization. Topics and Target Audience: - Autonomous unmanned systems for teaching and research; - Rapid control development tools; - Real-time control; - Tools for simulation and operation; - Curriculum for unmanned vehicle systems.

A Passive Fault Tolerant Flight Control for Maximum Allowable Vertical Tail Damaged Aircraft (Dr. Liu)

It investigates a passive fault tolerant control to aircraft that suffers from vertical tail damage. A novel notion of damage degree is introduced to parameterize the damaged ßight dynamics model. It is applied to seek the maximum allowable damage degree (tolerance capacity) stabilizable by the proposed passive fault tolerant and backup control under a linearized model. The design algorithms are presented and illustrated through numerical simulations on one aircraft model. Furthermore, the impact of potential control saturation is taken into account in the proposed design and a set of design parameters are tuned such that the maximum allowable damage degree is bounded, represented as the so-called critical damage degree.

Fault Diagnosis and Tolerant Control of Aerospace Systems using LPV Techniques (Dr. Puig)

The problem of robust fault detection is addressed using an adaptive threshold generation for non-linear systems described by means of LPV models. Adaptive thresholds are generated using an interval LPV observer that generates a band of predicted outputs taking into account the parameter uncertainties bounded using zonotopes. The interval LPV observer is designed via pole placement using Linear Matrix Inequalities (LMI). LPV fault sensitivity analysis is used to characterize the minimum detectable fault as well as to determine the limitations of proposed FDI strategy. The isolation task uses the fault estimation to isolate the faults. Fault estimation relies on the knowledge about the faulty system behavior using the fault sensitivity concept.

On the other hand, the FTC problem is addressed using three approaches. The first one is a LPV FTC design based on Admissible Model Matching (AMM), where a set of admissible models is used, which provide stability/performance guarantees. The main contribution of this approach is to accommodate the controller guarantying that the system closed-loop behavior is in the set of admissible behaviors. This accommodation involves the on-line controller reconfiguration in presence of parametric faults and the fault estimation. This estimation is considered as a scheduling variable that allows the reconfiguration of the controller. The second approach consists in adapting the faulty plant to the nominal controller instead of adapting the controller. That is, the faulty plant together with the reconfiguration block allows to the controller to see the same plant as before the fault. When a sensor fault is considered, an observer is used to calculate a replacement value. This approach is known as virtual sensor. By duality, the results can also be applied to derive a virtual actuator when the actuator fault is considered.

Finally, an integrated FTC design procedure for LPV systems that considers the fault estimation using an unknown input observer and a virtual actuator is proposed. The FTC controller is implemented as a state feedback controller. This controller is designed such that it can stabilize the faulty plant using LPV techniques and LMIs.

The effectiveness and performances of the FDI/FTC methods will be illustrated with several examples with special emphasis in a two-degree of freedom helicopter.

Design of Fault-tolerant Control Methods Based on Reliability (Dr. Theilliol & Dr. Zhang)

Faults or failures such as defects in components, instruments, controllers and/or control loop can cause undesired reactions and consequences such as damages to technical parts of the plant, to human life or to the environment. Traditionally, the objective of Fault Tolerant Control System (FTCS) is to maintain its current performance close to the desired one and preserve its stability conditions despite of component and/or instrument faults; in some circumstances a reduced performances may have to be accepted as a trade-off leading to a sub-optimal outcome. Design of control systems to achieve fault-tolerance for closed-loop control of safety-critical systems has been an active area of investigation for many years. It becomes more and more clear that there are certain trades-offs between achievable normal performance and fault-tolerance capability. However, despite of the many efforts in control community, most of the contributions did not consider or take into account the reliability of components, algorithms or soft computing structures to guarantee such performance and to reduce the gap between nominal and faulty case. This contribution aims at presenting new and innovative research results on how to design Fault Tolerant Control Systems with particular attention to consider and combine reliability analysis in the design procedure and/or real-time control synthesis. Current and future research is presented in order to solve the above challenging research problems devoted to safety-critical systems such as flying vehicles, unmanned aerial vehicles (UAVs), missiles, airships etc.

A Weighted H_infinity Consensus Achievement for Networked Unmanned Systems with Directed and Switching Topologies (Dr. Khorasani)

In the past few years, the consensus problem has attracted a lot of attention in the multi-agent and unmanned systems cooperative control community. The consensus problem is concerned with the convergence of the outputs or states of all agents to a common value. This implies that each agent has access to other agents'state, known as neighboring agents, by using either a communication network or sensing devices. Most of the work in this area consider agents dynamics as first-order or second-order integrators and have focused on issued such as communication delays, communication network's graph connectivity preservation and reference signals. However, research and development on the H_infinity consensus problem of unmanned systems under switching topologies is less investigated.

In contrast to the existing literature, in this work the communication network is directed, the network topology can switch arbitrarily, the proposed algorithm can solve a weighted H_infinity consensus problem, and finally there is no need to solve any set of LMIs and instead the controller can be designed by solving an algebraic Riccati equation. It is worth noting that the existence of a solution for an LMI cannot always be guaranteed and the time complexity of solving an LMI is O(n^2 p^4 ), where n denotes the number of agents in the team and p denotes the number of states of each agent. Therefore, it is not always computationally feasible to design a consensus algorithm for teams with large number of agents or agents with large number of states by using the existing techniques. On the other hand algebraic Riccati equations can be solved with time complexity of O(p^6 ) and the existence of the solution can be guaranteed.

In this presentation, a design procedure is presented for the consensus algorithm for a team of homogeneous unmanned systems/multi-agent systems by utilizing an algebraic Riccati equation (ARE). The key feature and novelty that is developed here is that the control design is achieved in two steps. First, a local state feedback controller is designed to ensure that the agents are marginally stable, and in the second step the control gains for the relative state feedback controllers are designed by solving an ARE. The required necessary conditions are that the communication graphs of the team subject to all switching topologies are strongly connected and the agents are stabilizable.

Workshop materials:

To be delivered to participants during and before workshop with the presentation slides, notes and other necessary supporting documents.

Workshop References (Author with bold face is one of the speakers at this workshop):

- Halim Alwi, Christopher Edwards, and Chee Pin Tan, Fault Detection and Fault Tolerant Control using Sliding Modes, Springer, 2011.

- Nader Meskin Khashayar KhorasaniFault Detection and Isolation: Multi-Vehicle Unmanned Systems, Springer, 2011.

- Christopher Edwards, Thomas Lombaerts, and Hafid Smaili (Eds.) Fault Tolerant Flight Control: A Benchmark Challenge, Springer, 2010.

- Camille-Alain Rabbath and Nicolas Lechevin, Safety and Reliability in Cooperating Unmanned Aerial Systems, World Scientific Publishing, 2010.

- Hassan Noura , Didier Theilliol, Jean-Christophe Ponsart, and Abbas Chamseddine, Fault-tolerant Control Systems: Design and Practical Applications, Springer, 2009.

- Mufeed Mahmoud, Jin Jiang, and - Youmin Zhang-, Active Fault Tolerant Control Systems: Stochastic Analysis and Synthesis, Springer, 2003.

- Youmin Zhang and Jin Jiang, Bibliographical Review on Reconfigurable Fault-tolerant Control Systems, Annual Reviews in Control, vol. 32, no. 2, Dec. 2008, pp. 229-252 (Ranked No. 1 in the "Top 10 Cited" and No. 6 in "Most Downloaded" articles published in the last five years at the journal).

- Other references will be provided during the workshop to participants.