Research topics

 

Scientific program

This project is a basic research one that aims at proposing an integrated approach to design new methodologies for the improvement of the vehicle dynamical behaviour. As said above, the modelling/identification, observation and control steps will be linked to achieve the same objective: avoid that some critical situations lead to injuries with tragic consequence.

Since all cases cannot be considered, the contribution and interest of the project will be illustrated through some representative critical situations such as:

  • Braking on uneven (wet) roads in a curve (or not)
  • Driving in banked curve (uneven or not, with or without braking)
  • Driving in a curve at high-speed, close to roll-over.
  • Car following under various road conditions (dry/wet) and at different speeds (low/high)

The overall organization is quite natural and shown below. Indeed the proposed approach needs some models the different partners will share. This could be different kind of models for design but also the simulation (Matlab/Simulink) models for the validation of the observation and control methods. The observer and control steps can be handled separately, while they need some coordination for the compatibility and integration of the results (through the detection of critical situations). The last technical task concerns the integration and validation of the methods.

Finally a coordination task is introduced for the project management and to ensure that all the contributions can be incorporated in the overall structure.

The proposed structure is based on the methodological development and needs some interactions between the following different task.

 

TASK 1. Modelling and Identification for vehicle dynamics

The main developments in this task will be the following:

  • Development, order reduction and validation of generic models, easily reusable, in particular for observer and control synthesis
  • Elaboration of a model library for Matlab/Simulink to be used by all the partners for control validation purpose
  • Models identification for other tasks (observers and controllers)

The figure below illustrates the organisation of Task 1, separated in 3 subtasks in order to achieve the objectives.

 

TASK 2. Observation and data fusion for detection of critical situations

The main developments in this task will be the following:

  • Classification of driving situations, of trajectories types, of uncertainties sources to characterize their criticity.
  • Development of new observers (reliable, robust, efficient, and easily implementable) to detect critical situations and phenomena such as loss of adherence, too important lateral accelerations and yaw rates, too small inter vehicles distances, roll over, fast variations of driving and road conditions, actuators and/or sensors failures.
  • Development of algebraic estimators both for estimation and compensation of unknown dynamics in the control algorithms.

The figure below illustrates the organisation of Task 2, separated in 4 subtasks in order to achieve the objectives.

 

TASK 3. Robust, adaptive and fault-tolerant control for vehicle dynamics

This task aims at dealing with:

  • Design of integrated control laws, requiring the cooperation of several sub-systems and actuators (breaking, suspension, steering, differentials ...) that guaranty the safety, while maintaining the comfort of the passengers.
  • Reconfiguration and adaptation of the controllers to potential critical situations (trajectories sensors, actuators failures, driving conditions variations), detected by the mean of observers.

This task is also divided in different subtasks, in order to reach a single control structure, as shown below. It is woth noting that the global chassis controller and the high-level one have some interaction. As well it should be noticed that the control structure depends on the estimated variables given by Task 2.

 

TASK 4. Integration, validation and tests

This task will follow the steps below:

  • Integration and validation in simulation for a first step.
  • Integration and validation on test-benches (as quarter-car or 4 posters bench) for preliminary validation of observers and controllers efficiency on some specific critical situations.
  • Integration and test on vehicle for validation of unified procedures for detection and correction of critical situations

The figure below illustrates the organisation of Task 4, separated in 3 subtasks in order to achieve the objectives.