Millisecond–nanosecond-combined pulse laser drilling technology optimized by a BP neural network

In laser drilling, issues such as irregular hole shapes, insufficient hole accuracy, low drilling efficiency, and high energy consumption are commonly encountered. To address these challenges, this paper presents a laser perforation system based on a BP neural network combined with millisecond and nanosecond pulsed lasers. The integration of millisecond and nanosecond pulsed lasers enhances perforation speed and reduces energy consumption compared to the use of a single laser, thereby offering improved performance in the drilling process. To address the issue of irregular hole drilling caused by the combined laser process, a BP neural network-based predictive model for hole shape is developed. The model optimizes the energy density and pulse delay parameters of the pulsed laser in real time based on the cone angle size of the hole shape. The goal is to achieve higher drilling quality while maintaining faster drilling speed and lower energy consumption. Experimental results indicate that, even with a limited dataset, the predictive model is highly reliable in predicting the cone angle characteristics of aluminum alloys.