小样本数据的喷墨打印墨滴状态预测方法

Prediction Method for Droplet Status in Inkjet Printing with Small Sample Data

在高精密电子器件喷墨打印制造过程中,墨滴状态的精确控制对器件的性能至关重要。然而,获取大量的墨滴状态数据费时费力,因此,本研究提出一种基于小样本数据的喷墨打印墨滴状态预测方法。首先采用单级梯形波驱动喷头,通过改变驱动电压幅值与持续时间产生不同状态的墨滴作为样本数据;然后对样本数据进行预处理后采用支持向量机(Support Vector Machine,SVM)、反向传播神经网络(Back Propagation Neural Network,BPNN)、决策树(Decision Tree,DT)、随机森林(Random Forest,RF)和极限梯度提升(Extreme Gradient Boosting,XGBoost)等五种方法对墨滴状态进行预测。实验结果表明XGBoost方法在小样本情况下具有较大优势,其平均绝对误差、均方根误差和平均绝对百分比误差分别为0.088、0.123和1.85%,均为五种方法中最优,可将其作为一种小样本数据的喷墨打印墨滴状态预测方法。

In the manufacturing process of high-precision electronic devices through inkjet printing,precise control of droplet status is crucial for device performance.However,acquiring a large amount of droplet status data can be time-consuming and labor-intensive.Therefore,a prediction method for droplet status based on small sample data was proposesd in this study.Firstly,a single-pole trapezoidal wave was used to drive the nozzle to produce droplets in different states as sample data by changing the amplitude and duration of the driving voltage.Then,five methods including Support Vector Machine(SVM),Back Propagation Neural Network(BPNN),Decision Tree(DT),Random Forest(RF)and Extreme Gradient Boosting(XGBoost)were used to predict the droplet status after preprocessing.Experimental results showed that the XGBoost method has a significant advantage in small sample situations,with an average absolute error,root mean square error,and average absolute percentage error of 0.088,0.123,and 1.85%,respectively,all being the best among the five methods.The XGBoost method can be used as a prediction method for droplet status of inkjet printing based on small sample data.