Abstract: Taking the gear skiving process as the object, the changing law of cutting force is studied, and predictive optimization of cutting forces, and a parameter control method for optimizing the process is proposed. Firstly, according to the kinematic theory of gear skiving, the solid models of workpiece and cutter are constructed, and the response data of cutting force in the process of single-tooth cutting are obtained through simulation experiments; secondly, the main cutting force prediction model is established based on multiple linear regression method, and the influence law of three process parameters on the main cutting force is analyzed: cutting speed, feed amount and axis intersection angle between tool and workpiece; finally, the main cutting force is used as Finally, the process parameters were optimized based on genetic algorithm with the main cutting force as the constraint condition and the highest machining efficiency as the target. The results show that the single and multiple process parameters with the greatest influence on the main cutting force are the feed and feed-axis intersection angle, respectively, and the error between the optimized predicted value and the simulation result is within a reasonable range, which verifies the validity of the prediction model. The research method and results provide an important reference for improving the machining quality and tool life of gear turning process.