融合协同进化离散型人工鱼群算法和多重分形的雾霾预测方法

发布时间：2018-03-09 23:01

本文选题：人工鱼群算法　切入点：协同进化　出处：《系统工程理论与实践》2017年04期 　论文类型：期刊论文

【摘要】：鉴于目前日益严重的雾霾污染,导致空气质量水平大幅下降,通过采用协同进化离散型人工鱼群算法,多重分形维数,并结合极限学习机,提出了融合协同进化离散型人工鱼群算法和多重分形的雾霾预测方法.首先使用佳点集理论初始化种群,通过引入人工鱼游速,改进人工鱼群算法聚群,追尾和觅食行为,及对其进行离散化,并引入竞争和合作机制;其次将协同进化离散型人工鱼群算法结合多重分形维数,对雾霾数据集进行约简;最后运用极限学习机建立雾霾预测模型.通过对北京,上海和广州三地区近两年的雾霾数据集进行实验及参数分析,实验结果表明,较其他方法,预测性能更优,具有良好的稳定性和可信性.
[Abstract]:In view of the serious pollution of haze at present, the level of air quality is greatly reduced. By using co-evolution discrete artificial fish swarm algorithm, multifractal dimension and extreme learning machine are used. In this paper, a haze prediction method based on co-evolution discrete artificial fish swarm algorithm and multifractal algorithm is proposed. Firstly, the population is initialized by using the theory of good point set. By introducing artificial fish swimming speed, the artificial fish swarm algorithm is improved, and the behavior of the artificial fish swarm clustering, rear-end and foraging is improved. Secondly, the co-evolution discrete artificial fish swarm algorithm is combined with multifractal dimension to reduce the haze data set. Finally, the haze prediction model is established by using the extreme learning machine. Through the experiment and parameter analysis of the haze data sets in Beijing, Shanghai and Guangzhou in the past two years, the experimental results show that the prediction performance is better than other methods. With good stability and credibility.
【作者单位】：合肥工业大学管理学院;过程优化与智能决策教育部重点实验室;美国俄亥俄大学工程学院工业与系统工程系;
【基金】：国家自然科学基金重大研究计划培育项目(91546108);国家自然科学基金(71271071);国家自然科学基金重大项目(71490725);国家自然科学基金青年项目(71301041)~~
【分类号】：TP18;X513

【相似文献】