Abstract—Plasma spray is a widespread thermal spraying technology due to its unique advantages, but control of various defects in the coating is still the major trouble currently faced in the field. The key problem left to be solved is how to determine the optimum combination of input process parameters to achieve the required quality of coating. This research work integrates Orthogonal Test Design Method, Support Vector Machine (SVM), and Particle Swarm Optimization (PSO) algorithm to ascertain the optimal process parameter settings of the plasma spray. Orthogonal Test Design method is used to design a set of representative experiments for reducing the number of tests, simultaneously, accessing to the most valuable sample data. The data obtained from orthogonal test is used as samples for training and testing SVM so as to establish relation model between process parameters and the coating quality. Then, SVM is combined with Improved PSO to find the optimal combination of spraying parameters. In the end, the verification test is implemented to verify the effectiveness of the optimal process parameters.
Index Terms—Plasma spray, orthogonal test design method, support vector machine, particle swarm optimization.
The authors are with the School of Reliability and Systems Engineering, Beihang University, China (e-mail: xjing@huaa.edu.cn, buaahuang@163.com).
[PDF]
Cite: Jing Xue and Min Huang, "Optimization of Plasma Spray Process VIA Orthogonal Test Design Method, SVM, and Improved PSO," International Journal of Materials, Mechanics and Manufacturing vol. 5, no. 3, pp. 153-158, 2017.