Abstract Disturbance and uncertainty in a production environment may result in a decline in product quality or even the occurrence of abnormal conditions in the coal slurry flotation process. Bayesian network (BN) has been widely applied in the safe operation control and product quality control of complex industrial processes due to its ability to effectively integrate expert knowledge with data information. Nevertheless, the low inference accuracy of previous discrete BN has hindered the further improvement of product quality in coal slurry flotation process, even after restoring normal conditions. To address this issue, this paper proposes an integrated safe operation and product quality control method for the coal slurry flotation process based on hybrid BN. This method establishes a hybrid BN model that includes both discrete nodes and continuous nodes, aiming to enhance the inference accuracy, achieve effective integration of safe operation control and product quality control, and further improve the coal product quality based on the rapid restoration of normal conditions in the coal slurry flotation process. The abnormal conditions data are input into the hybrid BN as evidence information for reasoning to obtain safe operation control decision. Additionally, the quality control decision that can optimize coal product quality is obtained by combining the simulated annealing (SA) algorithm on this basis. The simulation results show that the proposed method can quickly and effectively eliminate the abnormities in coal slurry flotation process and improve the coal product quality.
Chu et al. (Thu,) studied this question.