Condition prediction of chemical complex systems based on Multifractal and Mahalanobis-Taguchi system

Abnormal condition is hazardous which may lead to accidents in chemical industry and effective condition prediction methods are imperative for chemical complex system. Comparing with traditional techniques of condition prediction without concerning nonlinearity of complex system, multifractal analysis elaborately reveals scale-invariance or self-similarity properties of observed data, which is one of the intrinsic characteristics of complex system. To predict the condition of chemical complex system, nonlinear features are extracted from the monitoring variable through multifractal analysis by using Multifractal Detrended Fluctuation Analysis (MF-DFA) algorithm, and multiple variables are investigated through Mahalanobis-Taguchi system (MTS) as a multidimensional analysis method to discover significant patterns. The effectiveness of the approach is illustrated using both experiment data and real data collected from chemical industry.