Research On Improved Seagull Optimization Algorithm And Ozone Concentration Predictio

Seagull Optimization Algorithm(SOA)is a swarm intelligence optimization algorithm,which is inspired by seagull migration and attack behavior.It has the advantages of simple structure and easy to understand and operate,and is widely used in many fields such as scientific research and engineering practice.However,SOA has poor performance in finding optimal values for high-dimensional complex optimization problems.Therefore,this paper proposes an Improved Seagull Optimization Algorithm(ISOA)and use ISOA to predict the problem of ozone concentration.The specific research work is as follows:(1)To solve the shortcomings of SOA,which is prone to fall into local optimization and slow convergence speed during the iteration process,three strategies are proposed to improve SOA.Firstly,in order to balance global search and local search,nonlinear control factor is proposed,which enables the algorithm to quickly target the vicinity of the optimal solution at the beginning of the iteration,and carefully search the vicinity of the optimal solution at the later stage of the iteration to improve the convergence speed of the algorithm;Secondly,using levy flight strategy to change the attack behavior of SOA,levy flight is a random walk mechanism that can prevent the algorithm from falling into a local optimal state and expand the search range of the algorithm and improve the convergence accuracy of the algorithm.Finally,parallel search strategy is proposed to enable individuals of seagulls to conduct multiple searches to obtain better attack locations,thereby expanding the search range of the algorithm and improving the convergence accuracy of the algorithm.Further testing with 12 benchmark functions shows that the ISOA has faster optimization speed and higher convergence accuracy.(2)To solve the problem of ozone concentration prediction,this paper selects the ozone concentration data of Chengdu from January 1,2021 to December 31,2021.Firstly,the correlation analysis of factors affecting ozone concentration is conducted,and 10 important factors are obtained.Secondly,a support vector regression(SVR)model based on 5-fold cross validation and grid search was used to predict the ozone concentration in Chengdu;Finally,ISOA was used to optimize the penalty coefficient and kernel function parameters in the SVR model,and an ISOA-SVR hybrid model was established to predict the ozone concentration in Chengdu.Using the values of MAPE,RMSE and MAE to compare the prediction results,the results show that the ISOA-SVR hybrid model has higher prediction accuracy and better effect.