基于极限学习机回归的海水Chla浓度预测方法

Prediction method of Chlorophyll-a concentration in seawater based on extreme learning machine regression

有效监测海水Chl a浓度状况对近海赤潮等海洋灾害的预警预报有着重要意义。运用灰色关联分析法确定预测模型的输入变量,可有效降低预测模型系统维数。采用极限学习机回归方法建立海水Chl a浓度预测模型,通过与广义回归神经网络、支持向量机回归二种模型的预测效果进行对比,表明极限学习机回归预测模型具有较好的预测精度、预测效率和泛化能力,能够实现针对研究水域环境下Chl a浓度的有效预测。

Effective monitoring the state of chlorophyll-a concentration in seawater plays an important role for the early warning of marine disasters,such as coastal red tides. Grey correlation analysis method is used to determine the input variables of the prediction model. It can effectively reduce the dimension of the model system. Extreme learning machine regression（ ELMR） method was used to build the prediction model of chlorophyll-a concentration in seawater.Comparing with the generalized regression neural network and support vector machine regression model,it indicates that extreme learning machine regression has better accuracy,efficiency and generalization ability of prediction than other methods. It adapts to be used for predicting the concentration of chlorophyll-a in this researched seawater area.