基于深度模糊推理模型的烟气脱硫设备运行状态预测OA

Flue-gas desulfurization treatment is an important measure to focus on pollution prevention and achieve en-ergy conservation and emission reduction.The safe and stable operation of flue-gas desulfurization equipment directly affects the sustainable development of the economy and society as well as the life and health of residents.However,un-der long-term operation,faults often inevitably occur in this equipment,which are difficult to be predicted and thus lead to a chain reaction and even accidents.In view of these problems,this paper proposes a deep fuzzy inference mod-el(DFIM)-based method to predict the operating status of flue-gas desulfurization equipment.First,a data-driven deep fuzzy belief network is constructed and trained using a fuzzy feature learning algorithm,which can map the raw data obtained from the sensor into anti-interference fuzzy semantic features and further give predicted outputs of fuzzy features.Second,an inference gating layer is designed to recognize the output fuzzy features,focusing on the fault modes of flue-gas desulfurization equipment.Third,according to the operational experience and operating environ-ment as well as fault modes of the equipment,the disposal plan is then provided.Finally,taking the pH of the slurry in the circulating pump and the temperature of the circulating pump bearing as the variables of the equipment operat-ing status,the effectiveness of DFIM in the prediction of the operating status of flue-gas desulfurization equipment is tested.The experimental results show that the proposed DFIM outperforms the existing methods in terms of predic-tion accuracy,stability,rapidity and other aspects,which has important guiding significance for the operation and maintenance as well as management of the operating equipment.