| Key Words |
GPU-in-the-loop; computation parallelization; state-space controller, GPGPU |
| Abstract |
Many practical computational tasks require processingof big data sets. When calculation cannot be performed fast enough multithread parallelization is most obvious solution. However interdependence of intermediate results can limit the parallelization of the computational problem. In modern control theory a key mathematical model is the state-space representation. The aim of the present paper is to demonstrate how such a model can be implemented on a conventional graphical processing unit without disturbing the response of the respective closed-loop system where the model is participating. |
| Article PDF | Download article (PDF) |
