Modeling and prediction of paint film deposition rate for robotic spray painting

Abstract

Paint deposition rate model is a key factor for determining process parameters in automatic trajectory programming of robotic spray painting. In order to establish the paint deposition rate model according with actual operating condition, firstly, the experimental data needs to be obtained through spraying an elliptic fog cone on a part using a spray painting robot. Then, the paint deposition rate model is fitted by using the Bayesian normalization algorithm and genetic algorithm respectively. In contrast with the experimental data, the result shows that the two models have high precision. However, compared with Bayesian normalization algorithm, the genetic algorithm converges faster and can obtain a concrete function expression of the paint deposition rate model. Thus genetic algorithm is better than Bayesian normalization algorithm in modeling the paint deposition rate.