Optimization of process parameters of Ni–Co–WC composite coating based on random forest prediction algorithm

Abstract In this study, magnetic field-assisted jet electrodeposition (MFAJED) was employed to prepare composite coatings using two sets of optimal process parameters, determined respectively by the orthogonal experimental method and an artificial intelligence algorithm. The microstructure, cross-sectional thickness, adhesion, and corrosion resistance of the coatings prepared under these two distinct processes were compared and analyzed. The results demonstrate that, under identical substrate conditions (304 stainless steel) and deposition technique (MFAJED), the coating prepared using the algorithm-predicted theoretical optimal parameters exhibited an average thickness of approximately 45.265 μm—a 27.4% increase compared to that obtained with the orthogonal-experiment-derived optimal parameters. Additionally, its surface friction coefficient was as low as 0.191 (a 21.4% reduction), microhardness reached 786.4 HV (a 17.1% improvement), and adhesion strength attained 29.43 N (a 34.81% enhancement). The self-corrosion current density was 5.413 × 10 −3 μA cm −2 (a 32.77% improvement), and the polarization resistance increased by nearly twofold. Both coatings exhibited favorable microstructural and performance characteristics, providing a valuable reference for surface treatments aimed at enhancing wear and corrosion resistance in applications such as marine vessels and mechanical gears.