Abstract: Aiming at optimizing the quality of welding production under constraint time, an iterative reduction optimization algorithm (IRO) based on the balance of time and quality is proposed. The algorithm adopts a workflow model, makes use of the serial path for easy implementation, standardizes the range of welding each task's activities, so as to select an appropriate service; combines a dynamic combination strategy to achieve a virtual reduction of the parallel task set, simplifies the original model, and finds the final optimization path by iterating through layers. The results of the test case show that IRO can achieve the optimization of the welding quality under constraint time. Finally, the influence of the deadline and the number of welding tasks on the performance of the algorithm is also analyzed. It is concluded that the optimization performance can be improved by changing the corresponding parameters.