微型电脑应用2025，Vol.41Issue(8)：26-29,4.
基于改进海洋捕食者算法和优化极限学习机的空气质量指数预测研究
Study on Prediction of Air Quality Index Based on Improved Marine Predator Algorithm and Optimized Extreme Learning Machine
摘要
Abstract
In view of the problem of low prediction accuracy of the air quality index traditional(AQI)prediction method,an AQI prediction method based on the optimized extreme learning machine(ELM)and improved marine predator algorithm(MPA)is proposed.Based on the solution principle of MPA algorithm,the quasi-reflection learning strategy,Cauchy variation strategy and horizontal crossover strategy are introduced.The improved MPA algorithm is used to optimize the ELM model,and an AQI prediction model is built based on the improved MPA-ELM.The results show that the improved MPA algorithm achieves fast convergence with 76 iterations,which is more stable convergence curve than MPA algorithm.In the test set,root mean square error(RMSE),mean absolute error(MAE)and mean absolute percentage error(MAPE)prediction errors are 3.118％,2.936％and 3.447％,respectively,which are all lower than the traditional salp swarm algorithm-CNN-long short-term memory(SSA-CNN-LSTM model,ELM-auto encoder(AE)-BP model and complete ensemble empirical mode decomposi-tion with adaptive noise-LSTM(CEEMDAN-LSTM)model.According to the results,the prediction error of the proposed model for AQI is smaller,and the prediction accuracy is higher,which meets the actual requirements of air quality monitoring.关键词
人工智能/海洋捕食者算法/柯西变异/极限学习机/空气质量指数预测Key words
artificial intelligence/marine predator algorithm/Cauchy variation/extreme learning machine/AQI prediction分类
信息技术与安全科学引用本文复制引用
马楠,王强..基于改进海洋捕食者算法和优化极限学习机的空气质量指数预测研究[J].微型电脑应用,2025,41(8):26-29,4.基金项目
山西省社科联大健康产业高质量发展科研专项课题(DJKZXKT2023081) （DJKZXKT2023081）
