局部近邻标准化偏最小二乘的多模态间歇过程故障检测

本文针对多模态间歇过程数据多中心和模态方差差异明显的问题,提出了一种基于局部近邻标准化偏最小二乘方法.首先,采用统计模量方法处理间歇过程数据,再利用局部近邻标准化方法将统计模量后的训练数据进行高斯化处理,建立偏最小二乘监控模型,确定控制限;然后,同样对统计模量后的测试数据进行局部近邻标准化处理,再计算测试数据的高斯偏最小二乘监控指标,进行过程监视及故障检测.最后,通过数值实例和青霉素发酵过程验证方法有效性.实验结果表明所提方法解决了故障样本近邻集跨模态问题,对多模态数据具有更好的故障检测能力.

Fault detection for multi-modal batch process based on the local neighborhood standardization partial least squares

In this paper, a local neighborhood standardization partial least squares (LNS-PLS) method is proposed to solve the problem of multi center and the distinctly different modal variance in multi-modal batch process data. Firstly, statistical pattern method is used for batch process data, and the local nearest neighbor standardization (LNS) method is used to transform the training data after statistical pattern into Gaussian distribution. The partial least squares (PLS) model is established and the control limits of T2 and squared prediction error (SPE) are determined. Next, the LNS standardized is performed on the test data of statistical pattern, and the new Gaussian PLS monitoring indexes are calculated for process monitoring and fault detection. Finally, the effectiveness of the algorithm is verified by the simulation experiment of numerical example and penicillin fermentation process. The results show that the proposed method solves the problem of the neighbor set of fault samples spanning two modes and has better fault detection ability for multi-modal data.