Abstract: The cutting path planning of boom-type roadheader is the basis of realizing autonomous cutting. Aiming at the problem that the planning trajectory is too simple to support the automatic cutting control effectively due to insufficient consideration of cutting head and target section constraints, a cutting trajectory planning method based on improved NSGA-II was proposed. Firstly, the kinematics model of boom-type roadheader is established to associate roadheader, target section and roadway to achieve a unified coordinate conversion benchmark. Secondly, a simplified model of the cutting head and a target section constraint model are constructed, based on which the cutting space driven by the fuselage pose is designed, and the feasible cutting space point set is obtained. Then, the bi-objective function with the shortest total cut path and the smallest total Angle is designed and transformed into a bi-objective traveler problem with adjacency constraints. The characteristics of the section forming cutting process were summarized, the adjacency constraint processing method was designed, depth-first search was used to improve the NSGA-II method in terms of population initialization and genetic manipulation, and the solution to the double-objective traveler problem was optimized to generate the executable cutting trajectory. The effectiveness of the proposed method is verified by numerical simulation. Compared with the original algorithm, the convergence speed of the improved NSGA-II is increased by 68.13%, and the total time is reduced by 91%. At the same time, the total distance and total Angle are improved to some extent. At the same time, the total distance and total Angle are improved to some extent. The experiment of non-full section forming cutting is carried out by virtual simulation experiment, which verifies the effectiveness of cutting trajectory driven by different fuselage positions and has higher forming quality. Finally, an experimental verification platform was built to compare the planned trajectory and tracking trajectory, and it was found that the tracking effect of the cutting trajectory was good, and the maximum deviation was 23.879 mm, which could effectively support the autonomous cutting control. This method provides a new solution for further developing cutting trajectory planning and autonomous control of section forming.