Research On Optimization Of Pressure Set Value Of Coke Over

Coke is the main product of coking production process,with the rapid development of coking industry,the quality requirements of coke are getting higher and higher,and the demand for coke production is also increasing,at the same time,the “13th Five-Year Plan” clearly points out that promoting energy conservation and consumption reduction,adhering to green development and improving the ecological environment.For the sustainable development of enterprises,it is necessary to comprehensively consider the impact of coke quality,production and coking energy consumption.Coke oven coking is a complex production process,the temperature system and pressure system are the two major systems for coking production,the set value of the collector pipe pressure is mainly set by the on-site manual experience for the coking production process,and it is difficult to adjust in real time according to the actual production conditions,resulting in unstable coke production and quality,and the current situation of excessive energy consumption in coking production,based on the massive historical data of 3# and 4# coke ovens in a coking plant,combined with the coking process and expert experience,the correlation model between the collector pressure and the production target is established,the optimization of the set value of the collector pressure and conduct the following exploration and research:(1)Analysis of production status based on coking process and mass production data: including statistical analysis of gas collecting pipe pressure,coking energy consumption,coke production and quality,it can be seen that the energy consumption and output of the coking plant are highly fluctuating,there is great room for optimization.Energy consumption,coke production and quality are the optimization goals.(2)A large amount of historical data was processed in the coking process and a production target model was established.The collected data were processed,and then based on expert experience and correlation analysis,seven variables related to coking energy consumption,five variables related to coke production,and eight variables related to quality were selected as model inputs.The energy consumption,yield and quality models are established by using Radial Basis Neural Network and Deep Belief Network.The simulation results show that both the radial basis neural network and the deep belief network have predictive ability,but the deep belief network has higher prediction accuracy and smaller error,which can better reflect the actual production situation,it lays a good foundation for the optimization of the collector pressure setting.(3)Combining the advantages of both differential evolution and particle swarm optimization,according to the production situation,the single target and multi-objective optimization of the coke collector pressure setting were carried out.When coke production is satisfied,priority is given to the effect of pressure on coking energy consumption,and differential evolution particle swarm optimization is used to optimize energy consumption single target.The simulation results show that although the energy consumption is reduced,the coke production is correspondingly reduced;Considering the energy consumption,production and quality objectives,the differential evolutionary particle swarm optimization is used to optimize the radial basis neural network model and the deep belief network model.The simulation results show that the optimized pressure setting can reduce coking energy consumption and increase coke production,provide theoretical support and operational guidance for the actual site,and lay the foundation for gas collection pressure control.(4)Optimize the resulting pressure setting for verification.Based on the set value verification of the target model,the high-precision deep belief network model is used as the verification model,and the radial basis neural network is optimized to obtain the pressure set value,and the single target and multi-target set value verification is performed.The results show that the optimized pressure setting can reduce energy consumption and increase production;According to the actual site control method,the coke oven collecting pipe is taken as the controlled object,the PID control algorithm is used,the feed forward compensation decoupling is used,the multi-objective optimization control simulation map is built,and the pressure output is used to establish the production target model in each sampling time,the multi-objective optimized pressure value is simulated as a system reference.The results show that the set value of the collector pipe pressure in the on-site production process should be adjusted in real time according to the optimization results.