为有效剔除工业现场采集数据中的显著误差,降低随机误差的影响,使校正后的数据更好的满足物料平衡和能量平衡,最大程度提高软测量模型的精度,提出一种基于同步算法的数据协调方法,并对双酚A生产工艺现场采集到的数据进行校正.仿真表明,校正后的数据误差率得到显著降低,提高了数据源的精度和模型的泛化能力.%In order to effectively remove gross error of collected data from the industrial field and reduce the impact of random error, to let that the coordination of data to better meet the material and energy balance and maximize the accuracy of the soft sensor model, the gross error detection based on the clustering ideological and the data reconciliation based on the synchronization algorithm were proposed. The data collected from bisphenol-A production process field was corrected. The simulation results show that after correction the data error rate has been reduced, thereby improving the accuracy of the data source and the generalization ability.
展开▼
