Data-driven Decision-making Approach of Laser Cladding Parameters for Low Carbon

Abstract

Abstract The machining performance, such as carbon emission, dilution ratio rate, and powder utilization rate, can be improved by selecting appropriate process parameters in laser cladding. However, the importance evaluation of machining performance by users is often fuzzy. How to make a reasonable decision about laser cladding parameters under this condition? To resolve the problem, a multi-objective fuzzy decision-making approach of laser cladding parameters is proposed using support vector regression (SVR), multi-objective dragonfly algorithm (MODA), and fuzzy TOPSIS. Firstly, SVR is used to build the correlation model among laser cladding dilution rate, powder utilization rate, carbon emission, and process parameters. Secondly, MODA is applied to search non-dominated process parameters (non-dominated solution) based on the model. Finally, the non-dominated process parameters are sorted by fuzzy TOPSIS according to the user’s evaluation, the process parameter ranking first is the laser cladding process parameters that best meet the requirements of users, and the feasibility and effectiveness of the approach are verified via the experiments. The low-carbon decision-making problem of laser cladding parameters under the fuzzy evaluation can be well solved by the proposed approach.