Multi-objective process parameters optimization of laser repairing steel plate in pressure vessels using machine learning

Laser cladding (LC) is an additive manufacturing technology, and it is often used to repair damaged parts. This study uses machine learning to optimize the process parameters for LC repair of Q345R pressure plates. The BP neural network improved by the SSA algorithm (SSA-BP) was trained using data obtained from experiments. Then the process parameters were optimized by NSGA-II. Compared with other process parameters, the sample prepared with optimal process parameters has a higher impact absorption energy, reflecting the effectiveness of the optimization. This study provides guidance for the optimization of multi-objective process parameters in laser cladding repairing.