论文部分内容阅读
针对间歇过程批次与批次之间,操作条件缓慢变化的特性,提出一种基于自适应多向独立成分分析(MICA)的监控算法。该方法首先用MICA法建模,然后在历史数据集中加入新的正常批次并剔除最早批次,逐渐更新模型,同时引入遗忘因子,提高对新过程特性的适应性。青霉素发酵过程的仿真结果表明,自适应MICA比MICA更准确地描述过程行为,并有效减少检测故障时的误报。
In view of the characteristics of batch process and batches, the operating conditions are slowly changing, a monitoring algorithm based on adaptive multi-directional independent component analysis (MICA) is proposed. The method is first modeled by the MICA method, then the new normal batch is added into the historical data set and the earliest batch is removed, and the model is gradually updated. At the same time, the forgetting factor is introduced to improve the adaptability to the new process characteristics. The simulation results of penicillin fermentation process show that adaptive MICA can describe process behavior more accurately than MICA, and can effectively reduce the false positives when detecting faults.