Cyclic task scheduling in manufacturing production line using neural networks

Scheduling and sequencing tasks arise frequently in many manufacturing production operations, allocation and planning of periodic resources and, for instance, in parallel processing. In some of these applications the tasks are executed repeatedly, e.g. the execution of production or maintenance tasks on a manufacturing production line. The scheduling of these tasks has a direct impact on the system throughput and one wants to optimise the production on the manufacturing line. In this paper we propose a new approach for the cyclic task scheduling problem. This approach is based on a neural network technique. We model the system and optimise it by applying our approach. The technique is validated by simulating it on a largescale manufacturing production problem.