【作者】 蒋丽英； 李一波； 李成博；

【Author】 Jiang Liying1, Li Yibo1, Li Chengbo2 (1 Shenyang Institute of Aeronautical Engineering, Shenyang 110136,China;2 PLA 65113 Unit, Shenyang 110163,China )

【机构】 沈阳航空工业学院； 中国人民解放军65113部队；

【摘要】 对于间歇生产过程,只有当一个生产批次结束之后才能得到该批次完整的历史轨迹以及最终产品质量。针对这一问题,提出了多模型PLS软测量建模方法。该方法根据过去的历史信息分别建立多个PLS软测量模型预测产品的最终质量。所以,进行在线预测产品质量时,该方法不需要估计或填充为测量的数据。通过对实际工业链霉素发酵过程数据的分析表明了提出的算法可以有效地预测链霉素发酵过程的最终产品质量。更多还原

【Abstract】 For batch processes, a complete batch trajectory and the final production quality can be obtained until the end of its operation. In order to solve this problem, a new soft sensor model, a multi-model PLS method, is proposed. Base on the current and past measured variables, the multiple PLS models are built to predict the final production quality. Therefore, the proposed method overcomes the need of estimated or filled up the future unmeasured values in online product quality prediction. The method is proved to effectively predict the final production quality by the application in diagnosing of a multi-stage streptomycin fermentation process.更多还原