Den här sidan är utskriven från Högskolan i Halmstads webbplats (www.hh.se). Texten uppdaterades senast den 2015-05-07. Besök webbplatsen om du vill vara säker på att läsa den senaste versionen.
To develop new, smart and safe technologies, we must rely on correct computer modelling and simulation. New research on so-called hybrid modelling languages will make simulations smarter, faster and much more accurate – enabling the development of future intelligent innovations.
If two vehicles are about to collide, an automatic sensor system can activate the brakes of one vehicle to avoid the collision. A system like this must work without fault, or the result would be devastating. When simulating such a scenario, some factors can be allowed to affect the outcome, but others not. For example, it should not matter if the tires are warm or at what direction the wind blows. The exact position of the car after stopping is not important either – the only thing that matters is that the collision is avoided.
This video clip shows an Acumen simulation the where a truck avoids a collision with a car by an automatic braking system. The unique feature of Acumen, which is a hybrid hodelling hanguage, is that it can handle uncertainties and take variations in to account, making it more realistic than the simulation programs used today. Simulation by Adam Duracz, PhD student at Halmstad University.
Walid Taha, Professor in Computer Science at Halmstad University, leads research aiming to develop simulation methods that verify such claims, even when there is a variation in, or uncertainty about, the system’s parameters.
– A problem with the simulation technologies used today is that they assume a perfect environment – but reality is never perfect. Making this assumption produces simulations that are fragile and can sometimes be completely misleading. We are developing a new type of modelling language that can handle uncertainties, and take variations in simulation parameters and inputs in to account. This ability make simulations much more realistic – like the example of the braking car where the focus is avoiding an accident, says Walid Taha.
Hybrid modelling languages (HyMLs) make simulation and modelling fast, robust and easy to use. Since they efficiently imitate reality, HyML simulations are very useful when developing new technologies, for example “cyber-physical systems”, like robots, autonomous vehicles or other embedded systems. Professor Taha is chairing a meeting about HyMLs at Rice University in Houston, USA, May 7–8, 2015.
– The meeting is a forum for presenting state-of-the-art research related to hybrid systems modelling and simulation. We want to inspire researchers working in this area and speed up the development process by stimulating collaboration across different areas within the field. In the end, our goal is to secure a safe and intelligent design of cyber-physical systems that can be a useful part of everyday life, says Walid Taha.
Text: LOUISE WANDEL
The HyML meeting will bring internationally renowned academic and industry experts together to facilitate collaborations on research and education related to the hybrid systems modelling and simulation, validated numerics and formal semantics.
Speakers will include:
Walid Taha is a Professor of Computer Science at Halmstad University and holds a joint appointment as a part-time full professor at Rice University. He is interested in the design, semantics, and implementation of programming and hardware description languages. His current research focus is on modelling, rigorous simulation, and verification of cyber-physical systems. Taha is leading the development of the Acumen modelling language and is the director of the Effective Modelling Research Group, a multi-university initiative sponsored by the National Science Foundation in USA and the Swedish KK Foundation.