Application And Research Of Intelligent Optimization Algorithm In Environmental Data

With the advancement of China’s industrialization process,environmental pollution has become more and more serious,especially in the aspect of air pollution,which has caused great harm to people’s health and property,and has affected the sustainable development of the national economy.At the same time,people’s awareness protecting atmospheric environment is also increasing,and the desire for governance of atmospheric pollution is very strong.Therefore,it is important prerequisites for the protection of the atmosphere that real-time understanding the changes in atmospheric pollutants and providing more reliable and accurate prediction methods.First of all,in this paper,there are some shortcomings in whale’s global exploration and local optimization,the self-adaptive weighting and Cauchy mutation is proposed in whale algorithm.It balances whale’s global exploration ability and local optimization ability.Then,based on self-adaptive weights and Cauchy mutation’s whale algorithm,a non-linear convergence factor was introduced to further change whale’s walking patterns and improve the speed and accuracy of the whale algorithm.Finally,the self-adaptive weights and the Cauchy mutation’s whale algorithm and the whale algorithm used nonlinear convergence factor are applied to initialize the back propagation neural network weights and thresholds to predict the concentration of atmospheric pollutants.First of all,this paper uses two improved whale algorithms for function test experiments.For the self-adaptive weights and Cauchy mutation’s whale algorithm,the optimization effect is improved in the function.In addition,the whale algorithm that used a non-linear convergence factor has a better effect in a unimodal function than a multimodal function.Then,this paper used two improved whale algorithms to optimize the weights and thresholds of the back propagation neural network,and predicts the concentrations of atmospheric pollutants at multiple sites and Xiaohe Yan in Shenyang.Finally,the experimental results show that the prediction results are more accurate based on improving whale algorithm optimized back propagation neural network.