Air Quality Evaluation And Prediction Based On Humanoid Intelligent Optimization Algorithm

"The ecological environment is an important political issue,related to the party's mission." This is an important speech delivered by general secretary xi jinping,so we must treat the ecological environment as we treat life.To create a good ecological environment as a priority for people's livelihood,one of the most important is to win the defense of blue skies,and based on the obvious improvement of air quality.Therefore,air quality prediction is of great significance to ecological environment management and environmental protection.Air quality pollutant data is a kind of nonlinear time series with high dynamics,complexity and spatiotemporal variability.It is impossible to establish an accurate mathematical model,and traditional prediction methods such as multiple linear regression and time series have poor prediction effects.In recent years,a variety of humanoid intelligent optimization algorithms have been proposed,which provide a feasible method for the fast optimization of solutions through the deterministic algorithm and heuristic random search.Humanoid intelligent algorithm is an intelligent optimization algorithm which involves human learning,competition and simulates human brain thinking,human organs and cells.In this paper,humanoid intelligence optimization algorithm,BP neural network and support vector machine are used to study the air quality pollutant data of taiyuan city from 2014 to 2019.The main research contents are as follows:(1)MEA_SVM,a new model combining mind-evolution-algorithm(MEA)and support vector machine(SVM),was applied to predict the air quality index(AQI).Experimental results show that MEA_SVM algorithm not only guarantees the accuracy of SVM prediction,but also significantly improves the prediction speed.In terms of prediction reliability and prediction accuracy,MEA_SVM algorithm is better than the combination method of genetic algorithm and SVM and the combination method of particle swarm optimization algorithm and SVM.Therefore,MEA_SVM algorithm has certain practical value in urban air quality prediction.(2)Heuristic search was conducted through the evolution and fusion of individuals in the brain storming optimization,and optimization was conducted for the penalty function c and kernel function g of support vector machine(SVM).A model combining BSO and SVM(BSO_SVM)is proposed.The proposed BSO_SVM hybrid model was used to predict the AQI of taiyuan city.Experimental results show that BSO_SVM is better in prediction accuracy,prediction speed,error rate and reliability,and more suitable for air quality prediction.(3)The BSO_TLBO hybrid algorithm was proposed by using the operator improvement brainstorm algorithm(BSO)in the teaching and learning stages of TLBO,the BP neural network was optimized by using the BSO_TLBO hybrid algorithm,and the BSO_TLBO_BP model was constructed to classify the air quality grade of taiyuan city.The experimental results show that the air quality classification results of the model are ideal,the accuracy is high and the error is small.The three prediction models proposed in this paper: MEA_SVM,BSO_SVM and BSO_TLBO_BP are used for air quality evaluation and prediction,which can provide new ideas for air quality prediction and theoretical support for the prevention and control of air pollution and the early warning of emergencies caused by air pollution.