Research On Optimization Of Cement Rotary Kiln Control Parameters Based On Improved Firework Algorithm

In recent years,China,which has the largest cement output in the world,has paid more and more attention to environmental pollution and energy waste.How to achieve low pollution and high efficiency in cement preparation has become an urgent problem to be solved in the cement industry.Nowadays,the cement industry adopts methods to reduce nitrogen oxide emissions by using low nitrogen oxide burners,using selective catalytic reduction technology(SCR),using selective non-catalytic reduction technology(SNCR)and other methods,but these technical methods have their own advantages.Shortcomings and shortcomings: the low nitrogen oxide burner structure is more complex and the denitrification efficiency is limited,the SCR investment and operation cost is high and the technology is immature,and the SNCR operation cost is high and the denitrification efficiency is average.Because the nitrogen oxides produced in the process of cement production mainly come from the cement rotary kiln,it is possible to reduce the amount of nitrogen oxides generated from the source by optimizing the control parameters of the cement rotary kiln.This paper studies the use of Improved Fireworks Algorithm(IFWA)combined with Extreme Learning Machine(ELM)to model and optimize the control parameters of cement rotary kiln to achieve the goal of low emissions and high efficiency in the cement production process.First of all,in view of the unreasonable Gaussian mutation operator and selection strategy of the Fireworks Algorithm(FWA),the algorithm’s optimization accuracy is low and the convergence speed is slow.The Cauchy mutation operator is used to replace the Gaussian mutation operator and the elite random The selection strategy replaces the original selection strategy,introduces the difference mutation operator and the reverse learning operator to propose IFWA.Five classical test functions are used to test the convergence speed and optimization ability of IFWA,FWA,quantum particle swarm algorithm,differential evolution algorithm and chicken swarm algorithm to verify the superiority of IFWA.Then,use IFWA to optimize the input weight of ELM and the hidden layer threshold to establish the IFWA-ELM algorithm.Three algorithms of IFWA-ELM,BP,ELM are used to establish the NOx concentration prediction model at the kiln tail and the actual coal consumption prediction model per ton of clinker respectively.The comparison of the performance evaluation indicators of various models shows that the model built by IFWA-ELM has good prediction accuracy.Finally,based on the operating mechanism of the cement rotary kiln combined with the data collected at the site of a cement plant in Tangshan,the IFWA-ELM algorithm was used to establish a prediction model with dual outputs of the NOx concentration at the kiln tail and the actual coal consumption per ton of clinker.Establish an objective function that takes into account the NOx concentration of the kiln tail and the physical coal consumption per ton of clinker,and use IFWA to optimize the selected cement rotary kiln control parameters within the constraints.The test results show that the IFWA algorithm can reduce the NOx concentration in the kiln tail while reducing the physical coal consumption per ton of clinker,achieving the expected goal of low emissions and high efficiency operation of the cement rotary kiln.