基于高斯和模型的喷涂机器人涂层生长率建模

Gaussian sum based coat growth rate modeling of spray painting robots

针对具有圆锥形喷炬的空气喷枪的涂层生长率建模问题进行研究,通过分析喷涂系统和涂层形成过程,给出了有关喷涂建模的合理假设.提出了一种新的能够适应涂层不同分布情况的高斯和静态喷涂模型,以获得比一般喷涂建模方法更好的建模精度和更广的应用范围.并根据涂层分布数据采用Levenberg-Maquart算法对模型进行最小二乘拟合,在此基础上构建了用于实际喷枪轨迹规划的动态涂层厚度分布模型.现场喷涂实验结果及和其他常用模型的比较表明所提出的高斯和模型具有更高的拟合精度和更好的实用性.

The growth rate modeling problem for air spray gun with circular spray cone was studied in this paper.The painting system and coat depositing process was analyzed in-depth,and several reasonable assumptions were given on painting modeling problem.A new Gaussian sum based static painting model that can be adapted to different kinds of coat distributions was proposed,which can achieve better accuracy and broader application fields.The Levenberg-Marquart algorithm was used to fit the model with coating distribution data in the meaning of least square.On the basis of that the dynamic painting model for real-time trajectory planning of spray gun was constructed.Results of field spraying experiments and comparisons to other common models show that the proposed Gaussian sum model has higher fitting accuracy and better practicability.