CE48 - Fondements du numérique: informatique, automatique, traitement du signal

Intelligent Estimation Algorithms for Smart Mobility – ArtISMo

ArtISMo: Intelligent estimation algorithms for smart mobility

The development of controllers with high performance and reliability for connected and autonomous vehicles (CAVs) will require real-time measurements or estimates of many variables on each vehicle. Examples of variables that are needed for feedback include: longitudinal distances, velocities and accelerations of other nearby vehicles; lateral position of the vehicle in its own lane; vehicle yaw angle; slip angle; yaw rate; steering angle; lateral acceleration; and roll angle.

Safety and autonomy of the autonomous and connected vehicle

In this project, we will propose original ideas and we will develop efficient estimation algorithms reconstructing the state variables necessary for the control and diagnosis of the autonomous and connected vehicle. The problems considered are the tracking of vehicles, the estimation of sensor and actuator faults, as well as the detection of cyber-attacks. Our objective is to propose a new approach for vehicle tracking in highways as well as in urban roads. The idea we will explore is the development of learning-based nonlinear observers. Several elements in a vehicle (such as tires) have very complex patterns whose parameters vary over time. This project will use a modeling approach combining physical differential equations and adaptive neural networks based on e-learning. In particular, well-known phenomena such as force ratios, Newton's mechanical motion, aerodynamic drag, rolling resistance, slope, combined acceleration terms for lateral and roll accelerations and the influence of the The angle of inclination of the road will be modeled by differential equations. Tire models for lateral and longitudinal forces, friction, motors, and suspension stiffness and damping characteristics will be modeled by neural networks whose weights can initially be obtained by the backpropagation method. The parameters of the model based on neural networks and part of the parameters of the physical differential equations will also be updated online during regular use of the vehicle.

- We have used estimation methods from classical Theoretical Automation.
- We have developed techniques based on online learning, combined with the classical synthesis of estimation theory.
- Different cases of applications to the autonomous and connected vehicle have been treated.

- Significant fundamental research work has made it possible to overcome the obstacles and limitations encountered by current estimation techniques by proposing effective and useful solutions for a large class of nonlinear models in general, and representing the autonomous and connected vehicle, especially. We have published two papers in the IEEE-CDC 2022 international conference and in the IEEE Control Systems Letters journal, on how to deal with nonlinear systems that do not satisfy the Lipschitz property globally. We have introduced important mathematical techniques on function extensions, namely the use of the Hibert projection on closed-bounded sets; the development of a Kirszbraun-Valentine type extension. These results are striking and make it possible to deal with all autonomous and connected vehicle models, which represent non-linearities that do not obey the Lipschitz condition globally. Hence the need for these results in all future work on estimation within the ArtISMo project.
- A complete state of the art has been carried out on the security of cyber-physical systems, from the point of view of estimation, control and identification. A book, containing several applications including the autonomous and connected vehicle, was published in Springer, and co-edited by Ali Zemouche, coordinator of the ArtISMo project.

- Several theoretical estimation methods have been developed and different application cases to the autonomous and connected vehicle have been discussed.

We are in the development phase of estimation algorithms using computer vision. This phase is theoretically the most difficult of the project; we are carrying out interesting reflections.
Then there are the experimental validation stages on real vehicles using the experimental platforms of the various partners of the ArtISMo project.

1. Observer design for non-globally Lipschitz nonlinear systems using Hilbert projection theorem. IEEE Control Systems Letters, IEEE, 2022, 6, pp.2581-2586. ?10.1109/LCSYS.2022.3170534?.
2. LMI-based observer design for non-globally Lipschitz systems using Kirszbraun-Valentine extension theorem. IEEE Control Systems Letters, IEEE, 2022, 6, pp.2617-2622.
3. Simultaneous state estimation and tire model learning for autonomous vehicle applications. IEEE/ASME Transactions on Mechatronics, Institute of Electrical and Electronics Engineers, 2021, 26 (4), pp.1941-1950.
4. A resilient nonlinear observer for light-emitting diode optical wireless communication under actuator fault and noise jamming. Security and Resilience in Cyber-Physical Systems: Detection, Estimation and Control, Chapter 12, Springer Nature Switzerland AG, In press, 978-3030971656.
5. Simultaneous state estimation and tire model learning for autonomous vehicle applications. IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2021, Jul 2021, Delft (virtual), Netherlands.
6. A new discrete-time interval estimator for vehicle side-slip angle estimation. Joint 8th IFAC Symposium on System Structure and Control, 17th IFAC Workshop on Time Delay Systems, 5th IFAC Workshop on Linear Parameter Varying Systems, Sep 2022, Montreal, Canada.
7. On high-gain observer design for nonlinear systems with delayed output measurements. Automatica, Elsevier, 2022, 141, pp.110281. ?10.1016/j.automatica.2022.110281?.
8. Security and Resilience in Cyber-Physical Systems: Detection, Estimation and Control. Springer Nature Switzerland AG, In press, 978-3030971656.
9. Introduction to cyber-physical security and resilience. Security and Resilience in Cyber-Physical Systems: Detection, Estimation and Control, Chapter 1, Springer Nature Switzerland AG, In press, 978-3030971656.
10. Resilient cooperative control of input constrained networked cyber-physical systems. Masoud Abbaszadeh and Ali Zemouche. Security and Resilience in Cyber-Physical Systems: Detection, Estimation and Control, Chapter 9, Springer Nature Switzerland AG, In press, 978-3030971656.
11. Filtered High-Gain Observer Design fora Class of Nonlinear Systems Subject to Delayed Measurements: Application to a Quadrotor UAVs. In proceedings of American Control Conference (ACC 22), pp 1186-1186,2022.
12. Continuous Discrete Time High Gain Observer Design for State and Unknown Inputs Estimations Of Quadrotor UAV. In proceedings of 21st the European Control Conference (ECC 21), pp 1186-1186,2021.
13. An integrated design of PI interval observer-based FTC for LTI systems. 30th Mediterranean Conference on Control and Automation (MED), 2022.

The development of controllers with high performance and reliability for connected and autonomous vehicles (CAVs) will require real-time measurements or estimates of many variables on each vehicle. Examples of variables that are needed for feedback include: longitudinal distances, velocities and accelerations of other nearby vehicles; lateral position of the vehicle in its own lane; vehicle yaw angle; slip angle; yaw rate; steering angle; lateral acceleration; and roll angle. There are also environmental variables which need to be measured such as tire-road friction coefficient, snow cover on road, and the presence of unexpected obstacles.
Measurement of all of the above variables requires significant expense. Indeed, some of the sensors above, such as slip angle and roll angle, can be extremely expensive to measure, requiring sensors that cost thousands of dollars. For example the Datron optical sensor for measurement of slip angle has a price over 10k€. In addition, several variables cannot be measured due to unavailability of sensors (at any cost).
Furthermore, a CAV requires highly reliable sensors and actuators. Failure of any one sensor or actuator, due to faults, cyber-attacks or denial of service, can cause a disastrous accident. Hence reliable fault diagnostic and fault handling systems are also needed. Such systems cannot be based on hardware redundancy which requires many extra copies of the same sensors. Instead, they need to rely on estimation algorithms and analytical redundancy. For all the above reasons, the development of intelligent estimation algorithms is highly important for autonomous vehicles.

Throughout this project we propose original ideas on estimation, which is a necessary and crucial step for reliability, resilience, and safety of CAVs. The overall objectives of the proposal consist in developing efficient estimation algorithms to reconstruct the unmeasurable state variables, which are required to design controllers and fault diagnostic schemes for CAVs. More specifically, the considered issues are safe and stable trajectory, estimation of faults in sensors and actuators, and cyber-attacks detection. We aim to propose a novel approach to tracking vehicles in a platoon and urban roads. The idea we will explore in this project is the development and use of learning-based nonlinear observers. Several components on a vehicle (e.g. tires) have highly complex models whose parameters are difficult to obtain and also vary significantly with time. This proposal will therefore use a modeling approach consisting of a combination of physically meaningful differential equations and adaptive online-learning-based neural networks to represent the vehicle dynamics. In particular, well understood phenomena such as force balances, mechanical motion per Newton's laws, aerodynamic drag, rolling resistance, road grade, combined acceleration terms for lateral and roll accelerations and road bank angle influence will be modeled using analytical differential equations. Tire models for both lateral and longitudinal forces, the friction circle, engine maps, and suspension stiffness and damping characteristics will be modeled using neural networks whose weights can be initially obtained using training via back-propagation. In addition to initial training, model parameters for the neural networks and a subset of parameters for the physically meaningful differential equations will also be updated automatically online during regular vehicle use.

More sophisticated and intelligent algorithms will be developed to face sensor faults and disturbances, cyber-attacks, and data-loss. All possible complex architectures of cyber-attacks and data-loss will be investigated.

Although this project belongs to fundamental research, experimental developments will be considered through the industrial partner. The partner FAAR will put at the disposal of the project, an innovation platform for the validation of the project’s developments.

Project coordination

Ali ZEMOUCHE (Centre de recherche en automatique de Nancy (CRAN))

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partner

FAAR FAAR SAS
CRAN Centre de recherche en automatique de Nancy (CRAN)
IBISC Informatique, BioInformatique, Systèmes Complexes
LISEC Laboratory for Innovations in Sensing, Estimation and Control (LISEC)
IRSEEM ESIGELEC

Help of the ANR 514,283 euros
Beginning and duration of the scientific project: February 2021 - 48 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter