Adaptive job-shop scheduling in discrete event dynamic systems
Head of the project: Prof. Baltazár Frankovič, DrSc., dr.h.c.
A research of a scheduling problem solution is focused on development of scheduling algorithms, which may be based on a heuristic approach, operational research (linear programming, branch-n-bound, tabu-search) and artificial intelligence AI (hill climbing, genetic algorithms, neuron networks, expert systems, constraint logic programming, etc.). The scheduling can be adaptively adjusted to a current situation or rescheduled in case of flexible systems.
The goal of scheduling problems may be formulated as follows:
To find optimal strategies of producing devices for job/resource scheduling.
The aim of the solution is to determine such configurations job set performance with verification of all constraints, that the total time of performance or the sum of the job costs would reach the minimum value.
One of the possible method for optimal scheduling is using multi-agents systems. The aim is to create a multi-agent scheduling system and the system of rules for mutual coordination, negotiation and assessment of conflict situations in the process of the scheduling in production systems. The possible use of such a system is in energetic, in market driven production planning, in job-shop scheduling of a flexible production system for small and medium enterprises, in telecommunication systems, and there are many more applications.
Publications