English：

One⁃dimensional cutting problem is an important field of operations research,which is widelyused in manufacturing industry. How to reduce cost, increase efficiency and improve material utilizationis the key of cost control. Aiming at the problems of low utilization rate of raw materials and low workingefficiency in one⁃dimensional steel bar cutting, a two⁃stage steel bar cutting optimization technology basedon dynamic adjustment of threshold value according to the actual engineering demand for steel bars isproposed. The threshold was determined by pre⁃processing and analysis of the demand data of steels barwhich are needed, and then an optimization model with the minimum remaining material as the objectivefunction was established for different types of demand steel bars. The improved particle swarmoptimization algorithm is used to solve the problem, in which the threshold can be dynamically adjusted tominimize the residual material. Engineering examples show that the model effectively improves theutilization rate of raw materials and the optimization algorithm has good stability.