首页|基于CWOA算法的激光熔覆工艺参数优化方法

基于CWOA算法的激光熔覆工艺参数优化方法

扫码查看
激光熔覆层的质量与激光熔覆工艺参数直接相关.为了获得合理的工艺参数以提高熔覆层质量,提出一种基于混沌搜索策略的鲸鱼优化算法(CWOA)激光熔覆工艺参数优化方法.针对工艺参数对熔覆层稀释率、宽高比、显微硬度的影响,以12Cr13不锈钢表面激光熔覆15-5PH粉末为例,设计L25(53)正交实验;基于实验结果构建工艺参数与熔覆层质量的拟合回归数学模型,通过层次分析法与变异系数法组合权重(AHP-CV)将多目标优化问题转化为单目标问题,并采用CWOA算法实现工艺参数优化;通过与鲸鱼优化算法(WOA)的对比,验证CWOA算法的优越性与准确性.结果表明,所提方法可以获得理想的稀释率、宽高比和显微硬度,具有较高的准确性,对生产实践具有很好的指导作用.
Optimization method of laser cladding process parameters based on CWOA algorithm
The quality of laser cladding layer is directly related to the laser cladding process parameters.To obtain reasonable process parameters for improving the quality of the cladding layer,a Whale Optimization Algorithm based on Chaotic search strategy(CWOA)laser cladding process parameter optimization method was proposed.For the influence of process parameters on the dilution rate,aspect ratio and microhardness of the molten layer,L25(53)orthogonal experiments were designed for laser melting of 15-5PH powder on the surface of 12Cr13 stainless steel as an example.Based on the experimental results,a fitted regression mathematical model of process parameters and molten layer quality was constructed,and the multi-objective optimization problem was transformed into a sin-gle-objective problem by combining weights of Analytic Hierarchical Process and Coefficient of Variation(AHP-CV),and the CWOA algorithm was used to realize the process parameters optimization.The superiority and accura-cy of the CWOA algorithm were verified by comparison with the WO A algorithm.The results showed that the pro-posed method could obtain ideal dilution rate,aspect ratio and microhardness with high accuracy,which was a good guide for production practice.

laser claddingwhale optimization algorithm based on chaotic search strategyprocess parameters optimi-zationanalytic hierarchical process and coefficient of variation weights

杜彦斌、张文平、梁强、李志强

展开 >

重庆工商大学制造装备机构设计与控制重庆市重点实验室,重庆 400067

激光熔覆 基于混沌搜索策略的鲸鱼优化算法 工艺参数优化 层次分析法与变异系数法组合权重

2024

计算机集成制造系统
中国兵器工业集团第210研究所

计算机集成制造系统

CSTPCD北大核心
影响因子:1.092
ISSN:1006-5911
年,卷(期):2024.30(12)
  • 1