To validate control systems, we build a comprehensive model-based validation process to ensure a complete management of the system and the control. Our competences in modeling and control systems allow us to develop field-proven processes, methods and tools to achieve this goal.
Since 2009 we have carried out over 120 projects related to testing and validation. Thanks to our methodology and know-how we have proved that we can reduce validation time and increase system reliability.
We are deeply involved in validation activities with regard to different aspects encompassing development of a validation platform, analysis of results and operational assessment:
- Development of system environment models and adaptation to real time constraints
- Define model parameters with the required abstraction level and the necessary accuracy
- Test definition and encoding to reach the coverage rate target
- Design and implementation of an interface to monitor tests (HMI) using solutions such as dSPACE or NI
- Hardware specification of HIL validation test beds and RP (Rapid Prototyping): calculator, electronic interface, sensors conditioning, power supply, etc.
- Control system validation by rapid prototyping on the real system: control system is embedded in a prototype calculator (dSPACE or NI)
- MIL-SIL-HIL validation: test automation, requirements verification, results analysis and automatic generation of test reports
- Final validation by connecting the calculator to the real system
Methods and tools
Control system validation process is stepped according to the requirements to be verified and to the system to be developed. In our model-based approach the system is represented in an « environment model » which must evolve and assimilate all process steps.
Phases of a validation process are:
Systemic modeling: This model is used to design the system and the tests required for validation by integrating all customer’s requirements
Model In the Loop (MIL): the control model is connected directly to the model of the physical system. Test plan is executed to verify requirements. Results will be used as reference in next steps.
Software In the Loop (SIL): the production software code is incorporated into the system simulation that contains the model of the physical system. It allows verifying that code generated from the model still verify the requirements.
Hardware In the Loop (HIL): the code is integrated in the real-time embedded system which will be combined with the model of the physical system. The embedded system to be tested interacts with this plant simulation.
The validation process requires specialized tools and therefore we use the following products for such purpose:
- Modeling tools based on SysML to generate the systemic model : our PhiSystem solution developed over Papyrus
- Matlab/Simulink and our control libraries, PhiControl, to create the control model
- Simulink, PhiSim, AMESim, Dymola for the environment model
- Test generation tools: MathWorks, MaTeLo (All4Tec), etc
- Solutions to adapt the model to the real-time constraints: MathWorks, dSPACE, etc
- Electronic test bench for HIL validation and rapid prototyping: dSPACE or NI
Key commercial references
|2008-14||Volvo Trucks||HIL validation of a powertrain control||2 persons|
|2007-14||PSA||MIL/HIL validation of a powertrain control||1 person|
|2011-14||Renault||SIL/HIL validation of a calculator for an air conditioning system||1 person|