This paper proposed a novel particle swarm optimization (NPSO) algorithm and applied it to short-term scheduling of single-stage batch plants with parallel units to minimization of earliness. The model is formulated as a mixedinteger linear programming (MILP) problem using the continuous-time domain representation. PSO is an optimization technique in real-number spaces, however scheduling of batch plants is a problem in discrete space. So, the improvement of NPSO includes introducing GAs operators for generating particle's flying velocity and position, and some heuristic rules for generating better initialization population. They all have no effect on the optimality of the scheduling problem. Computational experiments show that NPSO are more clearly appropriate than standard GA for scheduling of batch plants with due date constraints, and NPSO becomes more effective after involving heuristic rules.
