Testing control algorithms is both expensive and time-consuming. This project aims to reduce these costs, shorten the time-to-market and improve quality.
Given the current industrial trend towards mechatronic machines and vehicles, an ever increasing number of control algorithms is being used in these systems. However, before being able to integrate these algorithms in commercial products, they must be subjected to extensive tests so as to make sure that the required controller performance is achieved in all possible situations. These tests are both time-consuming and expensive.
On top of that, some situations occur only very rarely and can be difficult to emulate. It is also possible that the same controller must be tested during different design iterations and using different product variants. As such, experimental testing for controller validation is not only time-consuming and expensive but also complex.
Considering the drawbacks of experimental testing and the increasing availability of already developed system models, this project aims to
- explore the potential of virtual controller testing;
- develop a methodology to define an optimal iterative virtual/experimental testing strategy, combining both experimental and virtual performance tests.
The overall objective is developing a methodology to establish the most optimal testing strategy for the control algorithm of a system or system family by optimally and iteratively combining both experimental and virtual tests. This methodology will be validated on practical applications.
To be able to reach this objective, the following steps have been defined:
- Developing a methodology and supporting software tools for defining representative virtual tests on the basis of experimental tests;
- Developing a methodology and supporting software tools for defining the most informative experimental tests on the basis of virtual tests;
- Extending the above-mentioned methodologies and software tools so as to make them suited for the efficient testing of re-tuned controllers and different product variants;
- Validation of methodologies and supporting software tools on industrial use cases.
By applying the model-based experimental/virtual testing methodology, the industry can profit from the following benefits:
- Reduction of experimental field validation costs by selecting the most optimal experimental tests only;
- Reduced time-to-market as the time needed for experimental field validation is shortened;
- Improved product quality thanks to a better and more efficient product validation process.
Want additional information on this project? Or do you have a specific question? Contact project leader Suzanne Van Poppel.