论文部分内容阅读
在晶圆/液晶面板等批次加工过程中,产品质量的及时估计与品质管制是提高产能和降低成本的有效途径.针对“少量多样”的混合制程,利用逐步回归算法挑选该制程的关毽变量,引入产品的效益因子,建立混合制程的虚拟测量模型;为克服系统扰动对模型精度的影响,以产品效益因子为状态量建立该制程的状态方程,利用Kalman滤波器递归估计模型参数得到动态的MANCOVA模型;最后通过某湿式蚀刻制程的工程应用验证了该算法的有效性.
In the batch processing of wafers / liquid crystal panels, the timely estimation of product quality and quality control are effective ways to increase productivity and reduce costs. For the “small and diverse” hybrid manufacturing process, the stepwise regression algorithm is used to select the process In order to overcome the influence of system disturbance on the accuracy of the model, the state equation of the process is established with the product benefit factor as the state quantity, and the model parameters are recursively estimated by using the Kalman filter The dynamic MANCOVA model is obtained. Finally, the effectiveness of the algorithm is verified by engineering application of a wet etching process.