论文部分内容阅读
首先提出了一种在并行工程环境下用于小批量生产的质量反馈两层模型，能够准确、客观、在线动态地评价现行加工工序的状况和进行相应的加工过程质量仿真，以便对新零件的公差带设计提供指导。接着针对多批次小批量问题，提出了一种抑制数据中系统噪声和观测噪声的方法，可以有效地提高建模的精度。然后利用多组实验数据对提出的模型进行研究，充分地说明了以上方法的有效性和实用性。最后，针对广东省惠阳机械厂生产的ＺＪＴ－４０型全自动胶囊充填机的充填杆外圆磨削加工过程，运用提出的多批次数据建模方法，有效地弥补了因小批量数据所固有的信息不完整性而无法精确建模的缺陷。
First of all, a two-layer model of quality feedback for small-batch production under concurrent engineering environment is put forward, which can accurately, objectively and on-line evaluate the status of the current machining process and perform the corresponding quality simulation of the machining process so that the new parts Tolerance zone design provides guidance. Aiming at the problem of batch and small batch, a method of suppressing system noise and observation noise in data is proposed, which can effectively improve the accuracy of modeling. Then using multiple sets of experimental data to study the proposed model, fully illustrates the effectiveness and practicality of the above method. Finally, according to the cylindrical grinding process of the filling rod of the ZJT-40 automatic capsule filling machine produced by Huiyang Machinery Factory of Guangdong Province, the proposed method of multi-batch data modeling effectively compensates for the problem that the small- The imperfect information can not be accurately modeled defects.