Optimal Digital Valve Control Using Embedded GPU

Digital hydraulics is a technology where simple valve components are used together with intelligent automation to achieve efficient control of hydraulic systems. A major challenge with digital valve systems is that efficient control requires the combination of reliable real-time control and computationally expensive optimization. In this paper we present a control system which is extended with an embedded graphics processor which provides massive parallelism to perform optimizations while critical control is performed by safety critical components. The focus in this paper is on how well a complex graphics processor can function as part of a reliable real-time control system.