CLASSIFICATION OF PROCESS TRENDS BASED ON FUZZIFIED SYMBOLIC REPRESENTATION AND HIDDEN MARKOV-MODELS

This paper presents a strategy to represent and classify process data for detection of abnormal operating conditions. In representing the da ta, a wavelet-based smoothing algorithm is used to filter the high fre quency noise. A shape analysis technique called triangular episodes th en converts the smoothed data into a semi-qualitative form. Two member ship functions are implemented to transform the quantitative informati on in the triangular episodes to a purely symbolic representation. The symbolic data is classified with a set of sequence matching hidden Ma rkov models (HMMs), and the classification is improved by utilizing a time correlated HMM after the sequence matching HMM. The method is tes ted on simulations with a non-isothermal CSTR and compared with method s that use a back-propagation neural network with and without an ARX m odel. (C) 1998 Elsevier Science Ltd. All rights reserved.