The concept of a “controller grid”, which makes effective use of computational resources distributed on a network while guaranteeing real-time operation, is proposed and applied to realize highly advanced control. It facilitates the total optimization of a plant control and achieves the high efficiency that is not acquired by individual plant optimization. To realize this concept, migration of a control task customized to be executed on one particular microcontroller to another microcontroller is necessary while strictly observing the required response time. Two techniques to meet this requirement are proposed: “task migration” for a control system and “real-time guaranteed scheduling of task migration and execution”. The effectiveness of the controller grid is assessed by applying it in experiments with electronic-throttle-body (ETB) advanced control.
Controller Grid: Real-Time Load Balancing of Distributed Embedded Systems
Sae Technical Papers
SAE World Congress & Exhibition ; 2007
16.04.2007
Aufsatz (Konferenz)
Englisch
Controller Grid: Real-Time Load Balancing of Distributed Embedded Systems
British Library Conference Proceedings | 2007
|Controller grid: real-time load balancing of distributed embedded systems
Kraftfahrwesen | 2007
|2007-01-1615 Controller Grid: Real-Time Load Balancing of Distributed Embedded Systems
British Library Conference Proceedings | 2007
|Towards dynamic load balancing for distributed embedded automotive systems
Tema Archiv | 2007
|