Improvement Of Genetic Algorithms And Optimization Of Grate Cooler Parameters

Genetic algorithm is a better optimal solution search technology.How to obtain a more accurate genetic algorithm and how to improve the diversity of population has always been the focus of research.This paper proposed an Adaptive Multi-Population NSGA-II genetic algorithm(AMP-NSGA2)and a Improved genetic algorithm(IGA)to improve the accuracy and the populations diversity.The IGA genetic algorithm is used to optimize the MKLSSVM parameters and an IGA-MKLSSVM model identification algorithm is established.Using the IGA-MKLSSVM algorithm and the adaptive multi-population NSGA-II genetic algorithm respectively to establish the key parameters of the cement grate cooler model and optimize the parametesr.The specific research work is as follows:First of all,aiming at the limitation of search performance of NSGA-II genetic algorithm,an AMP-NSGA2 genetic algorithm is proposed.Based on the theory of discontinuity equilibrium,this algorithm constructs a multi-population and multi-crossover operator to improve the population diversity.According to the contribution of sub-population to the EXS solution set,Logistic model is adopted to realize the adaptive adjustment of sub-population.At the same time,the algorithm also combined with the local search method to improve the search ability.The experimental simulation shows that the algorithm has better convergence and optimization capabilities.The algorithm provides the algorithm basis for grate cooler parameter optimization.Secondly,this paper proposes an IGA algorithm to solve the problems of poor population diversity and poor local optimization ability.According to the contribution of the period of multi-population and multi-crossover operators to EXS sets,determining the most adaptive crossover operator type in solving problems.Based on the different performance of the EXS different regions,seting different crossing probability and mutation probability.While maintaining the superiority of individuality,it can also improve the diversity of solution sets.And the IGA genetic algorithm was used to optimize the parameters of MKLSSVM to establish the IGA-MKLSSVM algorithm.This recognition algorithm is superior to other comparison algorithms in terms of recognitionaccuracy and generalization ability,and provides a model basis for the establishment of a grate cooler prediction model.Finally,IGA-MKLSSVM model identification algorithm is used to identify the key parameters of grate cooler: secondary air temperature and two-chamber grate under pressure.The AMP-NSGA2 algorithm be used to optimize the parameters,and get the value of the control variable that maximizes the controlled variable.By analyzing the energy consumption and heat exchange efficiency of the cement cooler,the necessity of the optimization of the key parameters of the grate cooler and the accuracy of the algorithm are verified.