Abstract: Intelligent plugging technique in high-pressure pipeline is a new plugging technique developed by PSI Company in 1990s. Chucking capability which is determined by locking fluke is the key technique for pipe plugging. In order to optimize the design of locking fluke for more effective chucking force, the structure optimization is desired. The paper presents the function relationship by utilizing the high nonlinear mapping capability of neural network regarded to the locking fluke angle, joint thickness, gap, inclination angel and curvature radius. Utilizing the global optimization feature of genetic algorithm to solve the abovementioned function, the optimized combination of the factors for affecting the contact force between fluke and pipe is obtained. The optimization result shows that the contact force is raised up to 87.81% compared with the structure before optimized design and the overall performance is significantly enhanced.