论文部分内容阅读
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.
In view of the current increasingly serious haze pollution, resulting in a sharp decline in the level of air quality, through the use of co-evolutionary discrete artificial fish swarm algorithm, multifractal dimension, and limit learning machine, the proposed fusion of co-evolutionary discrete artificial fish swarm algorithm and Multifractal haze prediction method.Firstly, using good point set theory to initialize the population, introducing artificial fish speed, improving clustering, rear-end and foraging behavior of artificial fish swarm algorithm and discretization, and introducing competition and cooperation Mechanism.Secondly, the co-evolutionary discrete artificial fish swarm algorithm is combined with the multifractal dimension to reduce the haze data set.At last, the limit learning machine is used to establish the haze forecast model.Through the analysis of the past two years in Beijing, Shanghai and Guangzhou The experimental data show that the proposed method has better prediction performance and better stability and reliability than other methods.