本发明提供基于现场数据的半导体生产质量预测方法及系统，包括以下步骤：S100、获取半导体生产设备生产过程中的现场数据，并存储；数据至少包括机台字段、时间字符串值、多种工况参数及产品质量数据；S200、对获取的现场数据进行处理；S300、回归建模，使用不同算法建立两个模型，然后对两个模型进行融合；使用融合后的模型对测试集预测质量，利用预测结果与实际质检结果的偏差，反馈优化回归模型，直至得到目标模型；S400、预测质量，利用目标模型预测实时生产数据的质量。与现有技术相比，该方法能适应半导体生产工艺复杂的特点；利用生产过程中数据对产品质量进行预测，预测结果较准确，能快速发现不良问题，协助调整生产工艺，并有效节约检测资源。

The present invention provides a semiconductor production quality prediction method and system based on field data, including the following steps: S100, acquiring field data in the production process of semiconductor production equipment, and storing; condition parameters and product quality data; S200, process the acquired field data; S300, regression modeling, use different algorithms to establish two models, and then fuse the two models; use the fused model to predict the quality of the test set, Using the deviation between the predicted result and the actual quality inspection result, feedback the optimized regression model until the target model is obtained; S400 , predict the quality, and use the target model to predict the quality of the real-time production data. Compared with the prior art, the method can adapt to the complex characteristics of the semiconductor production process; use the data in the production process to predict the product quality, the prediction results are more accurate, the bad problems can be quickly found, the production process can be adjusted, and the detection resources can be effectively saved. .