Optimization Of Ethylene Cracking Furnace Operation Based On Improved NNIA

Ethylene is one of the important raw materials produced in the chemical field and can be used to synthesize basic chemical raw materials such as fibers,rubber and plastics.Ethylene decomposition reactor is the central equipment of ethylene production equipment.chamber,single radiation chamber and millisecond furnace.The production capacity and technology level of the ethylene cracking furnace directly determine the production scale,output and product quality of the entire ethylene plant.Therefore,optimizing the operation of ethylene cracking furnace has important theoretical and practical significance.In view of the complexity of ethylene cracking furnace process,starting with the analysis of the internal mechanism of the process,this paper studies and applies the cracking furnace process modeling based on neural network,and uses multi-objective optimization algorithm to solve the operation optimization problem of cracking furnace.The main research work of this paper is as follows:(1)An improved nnia algorithm based on chaotic algorithm is proposed.The adjacent distance is introduced to replace the congestion distance of the original algorithm,and the chaos optimization algorithm combines with the multi group cloning strategy and improves the convergence and distribution of the algorithm.(2)Considering the complexity of the cracking process,the internal mechanism of the cracking furnace manufacturing process is studied,and the important parameters affecting the cracking furnace work are discussed,and the auxiliary variables necessary for the modeling are determined,and the theoretical basis for subsequent model establishment is given.(3)In view of the many uncertain interference factors in the chemical process,the mechanism model does not have self-adaptive ability,and the mechanism model structure is complex and requires long-term iterative calculation,this paper will apply neural network to the ethylene cracking furnace.In the modeling research,models of ethylene cracking furnaces based on BP and DBN are established respectively and simulation comparison is made.As the result,it was shown that the ethylene decomposition reactor model based on DBN satisfies the demand of the universality and the real time requirement of the ethylene decomposition reactor model,and it can be applied.To accommodate errors in parameter testing,it can replace the traditional mechanism model.(4)In order to increase the output of ethylene propylene,at the same time effectively reduce the input of raw materials,and improve the input-output ratio of the entire production process.In this paper,the improved nnia algorithm is used for multi-objective optimization on the established DBN-based ethylene cracking furnace model.At the same time,the satisfaction method is used to select a compromise solution for the enterprise based on the fuzzy function,which is convenient for the enterprise to make subsequent decisions or automate parameters.Adjustment.