A two-level Particle Swarm Optimisation algorithm on Job-Shop Scheduling Problems

The paper first presents an efficient PSO algorithm called GLN-PSOc. It is a variant of the standard PSO algorithm that requires a mapping procedure in order to solve a specific problem. Two mapping procedures are proposed in this paper for GLN-PSOc. The first mapping combined with GLN-PSOc is a heuristics for JSP and is called JSP-PSO. The second mapping combined with GLN-PSOc resulted in a parameter-tuning algorithm called PT-PSO. The combination of these two algorithms JSP-PSO and PT-PSO forms a two-level PSO. It is efficient and able to generate very good solutions on the benchmark instances.