Research And Applicatino Of Chemical Time Series Data Modeling Analysis Method

Complex chemical production data is characterized by high dimensionality,uncertainty and noise,so it is difficult to directly control raw material consumption and measure product quality in the process of complex chemical production.In recent years,the complex chemical industry has developed rapidly,and chemical production facilities have a direct impact on the goal of achieving energy saving and emission reduction and low carbon economy.However,with the rapid increase of chemical production equipment,the capacity utilization rate has been reduced.At the same time,complex chemical production data has the characteristics of dynamics,noise,uncertainty and noise,which makes it difficult to directly predict the consumption of raw materials and measure the product quality in the production process.Therefore,it is of great research significance and practical application significance to continuously optimize and improve the energy efficiency level of chemical production equipment.To this end,this paper selected the purified terephthalic acid(PTA)solvent system and ethylene production equipment for modeling analysis and optimization.The main content of this article,1.Aiming at the complexity of the chemical process,the collected data will have abnormal problems.This paper proposed a 3Sigma anomaly detection method combined with density clustering to reduce the false recognition rate of the original 3 Sigma detection method and verified the effectiveness of the method by PTA and ethylene data.2.In view of the dynamic and real-time characteristics of chemical processes,it is difficult to accurately predict some variables in the internal production process.This paper proposes a method combining long-short term memory networks with improved attention mechanism and logistic regression to effectively extract and utilize the temporal characteristics of chemical data,and obtain the dependence of chemical production data at different time points,thus obtaining more accurate forecast result.Finally,the effectiveness of the proposed method is verified by the standard UCI data set.3.The PTA solvent system and ethylene production equipment were modeled and analyzed by the method proposed in this paper.The results of different comparison models were compared and analyzed to verify the effectiveness,practicability and superiority of the proposed method.Finally,according to the method proposed in this paper,the production process of PTA and ethylene is optimized to provide operational guidance for industrial production,so as to achieve the goal of increasing productivity and reducing energy consumption.