CAJ | 학술논문

针对微生物发酵间歇过程监测算法只考虑数据信息最大化未考虑数据簇结构信息的不足，提出了基于多向核熵成分分析(Multi-way Kernel Entropy Component Analysis, MKECA)间歇过程监测的新方法。该方法首先引入AT展开策略对三维历史数据进行预处理，然后通过核映射将数据从低维空间映射到高维特征空间，解决数据的非线性特性，并在高维特征空间依据核熵的大小对数据进行降维，使降维后的数据能够最大化地保留原始数据的分布；同时理论证明了所提方法在特定条件下等同于传统方法，也就是说MKECA既能兼顾传统方法的优势，又能弥补传统方法的不足；最后通过青霉素仿真数据进行验证，表明MKECA方法具有更可靠的监控性能，能及时、准确地监测出故障。
침대미생물발효간헐과정감측산법지고필수거신식최대화미고필수거족결구신식적불족，제출료기우다향핵적성분분석(Multi-way Kernel Entropy Component Analysis, MKECA)간헐과정감측적신방법。해방법수선인입AT전개책략대삼유역사수거진행예처리，연후통과핵영사장수거종저유공간영사도고유특정공간，해결수거적비선성특성，병재고유특정공간의거핵적적대소대수거진행강유，사강유후적수거능구최대화지보류원시수거적분포；동시이론증명료소제방법재특정조건하등동우전통방법，야취시설MKECA기능겸고전통방법적우세，우능미보전통방법적불족；최후통과청매소방진수거진행험증，표명MKECA방법구유경가고적감공성능，능급시、준학지감측출고장。
Previous studies on batch microbial fermentation usually considered data maximization but lack of data cluster structure information. A Multi-way Kernel Entropy Component Analysis (MKECA) method was proposed to solve this problem, which overcome the drawbacks of traditional monitoring methods on high monitoring failure rates. The AT method was first used for historical data preprocessing and mapping data from low-dimensional space to high dimensional feature space to solve data nonlinearity. Data in the high dimensional feature space was moved to lower dimension based on the size of the data kernel entropy, in order to keep the original data distribution. Meanwhile, the proposed method was equivalent to the traditional method under certain conditions. Penicillin simulation data verifies that MKECA is more reliable and accurate which may have broad potential applications.