Laser Molten Pool Temperature Detecting Based on Recursive Estimation in 3D Printing

Forming accuracy of the product is directly affected by the laser molten pool temperature. Since platform is in a high temperature environment, the pool temperature is difficult to detect. At present, the high-range infrared imager is expensive. It cannot be used widely. Therefore, in this paper, a detection method based on recursive estimation is proposed. In this method, the low-range infrared imager is used to detect the temperature of the laser molten poor, and the CNN is used to implement recursive estimation. The overrange detection of the temperature is achieved in this method. Through actual data verifying, the error of recursive estimation is satisfactory.