A novel discrete Particle Swarm Optimization for optimal assignment problem

Optimal Assignment Problem (OAP), a category of special integer planning problem, belongs to linear programming problem. For the Hungary Algorithm (HA) is difficult to program and isn't convergent in some cases, a novel discrete Particle Swarm Optimization (PSO) algorithm for OAP is proposed, which is based on the model of velocity and displacement. But, with a view to the particularity of OAP, the particle's velocity and displacement are endowed with new connotation, and relevant update mechanisms of velocity and displacement are designed. The algorithm is applied in general OAPs (including the problems for which HA is not convergent), and compared with GA on performance by testing at OAP with different scales. Experimental results show the algorithm is effective, well adaptability, and better at convergence and performance than GA.