Operating optimality assessment and nonoptimal cause identification for multimode industrial process with transitions

Due to inappropriate operating or other uncertainties, process operating performance may deteriorate from the optimal state, which leads to a disappointing comprehensive economic index. However, sufficient attention has not yet been paid and few studies have been reported in this area so far. In this study, a novel operating optimality assessment and nonoptimal cause identification method is proposed for multimode processes with transitions. In the offline part, the operating optimality assessment models are formulated for both stable and transitional modes because of their different process characteristics. In the online part, the online mode identification strategy is used based on Bayesian inference, and then the process operating performance is evaluated by the proposed operating optimality assessment method. When the process operating performance is evaluated as nonoptimal, the cause variables can be determined based on the variable contribution rates. Finally, the effectiveness of the proposed method is verified though the Tennessee Eastman (TE) process.