Optimization of dilution rate of laser cladding repair based on deep learning

During the laser cladding repair, the heat accumulation effect is produced due to the continuous input of energy, which leads to the increase of dilution rate under the same process parameters. In order to get the appropriate dilution rate, an accurate finite element model was established based on the experiment, and the variation of dilution rates under different power functions was analyzed. According to the results of finite element analysis, different power functions and dilution rates at different times were collected, and the dilution rate was optimized through deep learning, and the ideal power function was obtained. The results show that the dilution rate of the whole cladding layer fluctuates around the ideal value all the time by adjusting the laser power in real time, and the results are satisfactory. The maximum error of dilution rate is 21.75%, the minimum error is 0, and the average error is 5.79%.