Skip to Main Content
This paper describes a concept for a task scheduling mechanism in factory simulation systems for tasks with differing cycle times. The goal is to partition tasks in a way that makes parallel computation possible, while maintaining a deterministic behavior. To achieve this behavior, there has to be a strict separation of read and write operations to the shared database from the actual computation. All read and write operations have to happen in a linear way, while all the calculations can happen simultaneously. The proposed concept uses a single master thread to perform all read and write operations, while all the calculation is done simultaneously by multiple slave threads. This offers the opportunity to use the whole potential of modern multi-core processors and thus to increase the speed of calculation.