压电喷射点胶过程的神经网络控制与参数优化

Neural Network Control and Parameter Optimization of Piezoelectric Jetting Dispensing Process

摘要: 针对微电子封装中小直径胶点喷射的需求，本文基于压电驱动点胶阀的工作原理，建立了撞针与喷嘴碰撞结构的二维几何模型，并采用COMSOL多物理场软件对点胶过程中的胶液流动与喷射特性进行了数值仿真。研究系统分析了驱动气压、喷嘴锥角、出口内径及内侧间隙等参数对胶液流速变化的作用机理，并进行了对比分析。仿真结果表明，喷嘴结构参数对胶水喷射速度及胶点直径具有显著影响。为进一步提高点胶精度，本文引入前馈神经网络算法，对仿真结果进行了优化验证。该研究为压电式点胶阀设计优化及微小胶点制造提供了理论依据和技术参考。

Abstract: To address the demand for small-diameter adhesive dot jetting in microelectronic packaging, this paper establishes a two-dimensional geometric model of the collision structure between the striker pin and the nozzle based on the working principle of a piezoelectric-driven dispensing valve. COMSOL multiphysics software is employed to perform numerical simulations of the adhesive flow and jetting characteristics during the dispensing process. The study systematically analyzes the mechanisms by which parameters such as driving air pressure, nozzle cone angle, outlet inner diameter, and inner clearance affect the variation in adhesive flow velocity, along with a comparative analysis. Simulation results indicate that nozzle structural parameters have a significant influence on the adhesive jetting velocity and dot diameter. To further improve dispensing accuracy, a feedforward neural network algorithm is introduced to optimize and validate the simulation results. This research provides a theoretical basis and technical reference for the design optimization of piezoelectric dispensing valves and the fabrication of micro-scale adhesive dots.