摘要
首先分析了影响空气喷枪涂层厚度及喷涂机器人作业过程中的可调和常变因素,然后利用BP神经网络方法对平板直行喷涂实验获得的实验数据加以拟合,建立以喷涂距离、喷枪移动速度、喷枪流量和测量点与喷枪轴线距离作为输入的喷枪喷涂模型。与传统模型相比,该模型用相对少量的实验数据就可以预测不同喷涂距离、喷枪移动速度和喷枪流量下的涂层厚度分布。实验数据表明,该模型准确、有效。
The factors affecting the coating thickness and the adjustable and changeable parameters during the working process of the spray painting robot were classified and analyzed.The experimental data acquired by the straight spraying experiment on the flat surface were fitted using a BP neural network.A spray gun model was built using the spray distance,spray gun moving velocity,spray flow rate,and the distance from the measure point to the axis of spray gun as inputs.Compared with the traditional model,the proposed model can predict the distribution of the coating thickness under different spraying distances,spray gun moving velocities,and spray flow rates with the relatively few experimental data.The validation experiments show that the proposed model is accurate and effective.
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2012年第1期188-192,共5页
Journal of Jilin University:Engineering and Technology Edition
基金
'863'国家高技术研究发展计划项目(2009AA043701)
摩擦学国家重点实验室项目(SKLT09A03)
国家自然科学基金项目(50975148
51005126)
关键词
自动控制技术
喷涂机器人
空气喷枪
涂层厚度分布
喷枪喷涂模型
BP神经网络
automatic control technology
spray painting robot
air spray gun
film thickness distribution
spray gun spray model
BP neural network
