Search

This paper illustrates two strategies for the detection and classification of abnormal process operating conditions in which multiple process variable trends are available. The first strategy uses a hidden Markov model (HMM) for overall process classification while the second method uses a back-propagation neural network (BPNN) to determine the overall process classification. The methods are compared in terms of their ability to detect and correctly diagnose a variety of abnormal operating conditions for a non-isothermal CSTR simulation. For the case study problem, the BPNN method resulted in better classification accuracy with a moderate increase in training time compared with the HMM approach.

This paper presents a strategy to represent and classify process data for detection of abnormal operating conditions. In representing the data, a wavelet-based smoothing algorithm is used to filter the high frequency noise. A shape analysis technique called triangular episodes then converts the smoothed data into a semi-qualitative form. Two membership functions are implemented to transform the quantitative information in the triangular episodes to a purely symbolic representation. The symbolic data is classified with a set of sequence matching hidden Markov models (HMMs), and the classification is improved by utilizing a time correlated HMM after the sequence matching HMM. The method is tested on simulations with a non-isothermal CSTR and compared with methods that use a back-propagation neural network with and without an ARX model.