| The most commonly used printing method in 3D printing is layered printing.By layering the target object model and accumulating printing from bottom to top,the target entity is finally obtained.Path planning is an important step in 3D printing.Choosing an appropriate printing path can significantly reduce the number of interruptions in the printing process,thus improving the printing efficiency.However,traditional path planning methods usually print according to a pre-planned regular path,which is prone to produce concentrated stress,which will reduce the heat dissipation of materials,lead to deformation and cracking,or lead to frequent head jump and "wire drawing" phenomenon,resulting in excessive curing edges.Based on the above problems,this paper proposes a path planning algorithm for 3D printing based on reinforcement learning.This paper innovatively applies reinforcement learning to 3D printing path planning.We transform the shortest path planning problem in Q learning into ergodic problem.This paper adds two constraints to set the reward value in Q learning.One is to reduce the number of start and stop of the print head,the other is to reduce the number of rotation.The experimental results show that compared with the traditional path planning method,this method can reduce the number of interruptions and turns in printing,and the forming effect is better.On this basis,this paper also proposes a path planning method combining intelligent printing with traditional printing.Firstly,the regular printing area and complex graphics area are divided by using the partition algorithm,and the suitable traditional printing path planning is selected in the regular printing area,and the reinforcement learning algorithm is used in the complex area.The experimental results show that the method of combining traditional printing with intelligent printing is adopted.It can greatly improve the efficiency of training and printing. |
