青霉素生产过程的在线统计监测与产品质量控制

将多向偏最小二乘(MPLS)方法应用于青霉素间歇生产过程的建模与故障诊断中。从青霉素反应过程的特点来看,数据具有多维性,应用传统的偏最小二乘方法会使过程的统计建模与故障诊断难以实现。MPLS可对间歇过程的多维数据沿变量方向进行分割,使得多批量的数据可以在过程的各操作阶段建立相应的PLS模型,从而完成对该反应过程的实时监视与故障诊断。运用T2统计、Q统计方法,结合贡献图对过程进行了仿真分析,从理论分析和仿真实验结果的一致性,证明了该方法在青霉素生产过程的故障检测与诊断方面是可行的。

On - line statistical monitoring and production quality control of penicillin cultivation process

On-line monitoring and fault diagnosis of a fed-batch penicillin fermentation process is presented using multi-way partial least squares method (MPLS). Since data from penicillin cultivation process are multidimensional, modeling and fault diagnosis based on the statistical data cannot be realized using traditional partial least squares (PLS) method. Multidimensional data will be cut along the variable axis, PLS model be built for every phase using MPLS method, and on-line monitoring and fault diagnosis of batch process be attained. The simulation is performed by means of statistical method such as Hotelling T2, Q and contribution plots. From the consistency of theoretical analysis and simulation results, the implementation of the methodology is illustrated.