The uncorrelated parallel machine schedulNE.Cms_Inserting problem is a typical problem in the workshop scheduling, and the single piece small batch production mode leads to frequent job switching and a large number of setup times, which reduces equipment utilization and production efficiency. This dissertation presents a research on the scheduling of uncorrelated parallel machines based on the grouping technique, which is dependent on the setup time. According to the similarity of the resources required for workpiece processing, the workpieces are clustered and grouped, and with machine constraints condition met, the allocation of all the workpiece groups on the machines as well as the order of the workpiece groups and that within each group on the same machine is determined. In this paper, a mathematical model is constructed with the minimization of total delay time as the optimization goal and genetic tabu search (GATS) algorithm is applied to solve it. Artificial bee colony (ABC) algorithms and genetic simulated annealing (GASA) algorithms are used for case studies. The comparison results show that the proposed algorithm has better searching ability.