Since Sherpa Engineering’s creation in 1997, we have achieved many modeling and simulation projects with our industrial customers. Our tools and methodologies have allowed our customers to significantly reduce development cost and time, risks and operational costs.
Sherpa Engineering has carried out 250 modeling and simulation projects in several industrial sectors during the last 5 years.
Challenges of modeling a system do not depend on a given component but on the system as a whole. That is why we need to develop what we call « system models » that aim at providing a comprehensive view of all phenomena, the design and the validation of the system.
The « system models » associated pre and post treatment utilities, result in a simulation tool that can be used in all phases of development:
- Sensitivity study and physical organs sizing
- Validation and multi-criteria optimization of a system architecture
- Control system design
- MIL/SIL/HIL validation of an embedded calculator
Our main activities are:
- Development of polymorphic and multilevel models: physical models, control system models, sensors and actuators
- Data analysis, model identification and parameter estimation
- Development of simulators and other specific tools: user interfaces, pre-treatment and post-treatment data, static and dynamic simulations
- Model reengineering
- Model adaption and compilation for real time application
Methods and tools
We use the Bond-Graph methodology for multiport modeling in order to graphically represent the physical parts of the system. Bond-Graph is particularly adapted to multi-physic systems and to manage the causality of the model.
A simulator is built according to the canonic form of a controlled system, including:
- Multi-domain physical model: thermal fluid, mechanical, electrical and hydraulic
- Digital control model for controllers: block diagram with multi-rate signal processing
- Sequential model for supervisory control, fault and working modes management (state diagram and state transition)
- Actuator and sensor models
- Test scenario codes
- Simulation result analysis and visualization
We use the most representative simulation tools in the market: Matlab-Simulink, AMESim and Dymola.
We have also developed our own modeling and simulation product, PhiSim which is composed of a set of scientific application libraries.
We propose the following training modules (for more information go to training):
|BG1||Dynamic modeling based on Bond-Graph method||2 days|
|BG2||Bond-Graph – Practice : Mechatronic applications||2 days|
|BG3||Bond-Graph – Practice : Energy applications||1 day|
|MSNL1||Non-linear system modeling||3 days|
|SEI1||Tests specifications and non-linear system identification||3 days|
|PhiSim||PHISIM module for system modeling||3 days|
|SIM||SIMULINK module for system modeling||3 days|
|SimulationX||SimulationX module for mechatronic applications||3 days|
Key commercial references
|2012-13||Alstom Power||Dynamic simulator for HIL validation||12 person.month|