This invention discloses a
laser cleaning device and method for surfaces made of multiple materials, belonging to the field of industrial
manufacturing technology. Addressing the problems of poor adaptability, imprecise
endpoint determination, narrow material compatibility, and high
process development costs in existing
laser cleaning methods, the device comprises a central control module, a multi-
wavelength tunable laser, a
laser-induced breakdown
spectroscopy device, an AI spectral model, a process digital twin
system, and a multi-
physics field
coupling simulation module, forming a closed-
loop control system. The cleaning method achieves cleaning by constructing a digital twin model, using LIBS-AI
coupling to identify material states, simulating and optimizing parameters, monitoring cleaning across
multiple modes, intelligently determining the endpoint, and iteratively optimizing the process. In this invention, the device effectively improves adaptability and cleaning accuracy through real-time component identification, dynamic
wavelength adjustment, multi-source
signal fusion criteria, and
simulation prediction. It can adapt to a wide range of spectral materials, shorten the process cycle, and is suitable for high-precision laser cleaning of multiple materials and complex surfaces.