Simulation and Optimization of Temperature Field for Laser‐Cladded CoCrFeNiMn High‐Entropy Alloy Coating on 316L Stainless Steel Surface Based on Newton–Raphson‐Based Optimization‐Extreme Learning Machine and Classification and Pareto Dominance‐Based Multiobjective Evolutionary Algorithm

A laser cladding heat source parameter optimization method based on the Newton–Raphson‐based optimization‐extreme learning machine (NRBO‐ELM) and classification and Pareto dominance‐based multiobjective evolutionary algorithm (CPS‐MOEA) frameworks is proposed to improve the simulation accuracy of the laser cladding temperature field for high‐entropy alloy powder. CoCrFeNiMn high‐entropy alloy coating is prepared on 316 L stainless steel surfaces by laser cladding technology. An L25(5 3 ) orthogonal test is designed, and the transient temperature field in the laser cladding process is simulated by ANSYS. A mapping model between heat source parameters and molten pool quality characteristics is established based on NRBO‐ELM. The heat source parameters are optimized via the CPS‐MOEA algorithm to generate the Pareto solution set. A comprehensive evaluation and decision‐making method based on weighted rank sum ratio is developed to rank the Pareto solution set and determine the optimal combination of heat source parameters. Results demonstrate the existence of a critical point near shape parameters b and c = 2.5 mm, where the material's low thermal conductivity coefficient, coupled with the nonuniform distribution of heat source energy, results in localized heat accumulation. The NRBO‐ELM mapping model accurately predicts the transient temperature field with less than 2% error. The optimized transient temperature field in laser cladding shows consistency with actual experimental measurements.