Particle Swarm Optimization with Controllable Diversity of Flexible Job-shop Scheduling with Lot-splitting Production

Focusing on the complicated processes permutation of sub-lots in multi-objective flexible jobshop scheduling with lot-splitting model, the matrix encoding and decoding methods based on workpiece, processes and lots is designed and then a particle swarm optimization with controllable diversity is proposed.It constructs a non-dominated solution set by means of pseudo binary tree's rule. The population entropy is applied to calculate particle swarm diversity.According to the scope of its changes, the particle neighborhoods are adjusted by cellular automata with local self-adaptive adjustment to balance accuracy and speed of the algorithm.Finally, compared with relevant examples, the algorithm is proved to be effective.