Abstract: With the continuous development of science and technology, laser cutting technology has become one of the indispensable technologies in modern industrial production. Carbon steel laser cutting parts processing technology is also getting more and more applications, with the quality of the processed workpiece increasingly demanding, the optimization of the laser processing process in order to obtain a higher quality of cutting and cutting accuracy of the steel plate is of great significance, in order to optimize the laser processing of carbon steel, this paper to 3 mm carbon steel as the object, using a 1 500 W laser systematic study of the laser power, cutting speed, the amount of out-of-focus, Auxiliary gas pressure and other cutting parameters on the cutting quality. The cutting width, surface roughness and slag height are selected as the standards for evaluating the cutting quality, and the influence law of process parameters on cutting quality is comprehensively analyzed and the influence mechanism is studied. At the same time, the data are predicted and analyzed by convolutional neural network, and the genetic algorithm with the non-dominated sorting of the elite strategy is adopted to optimize the process parameters, so as to get the optimal process parameters for laser cutting of carbon steel under the condition of the auxiliary gas is oxygen. The optimal process parameters for laser cutting carbon steel under the condition of oxygen as auxiliary gas are P=1 500 W, v=64 mm/s, F=2.6 bar, and h=2.9 mm. It is found that the cut surface of the optimized piece is smooth without slagging phenomenon and the width of the slit is small, which meets the expected standard.