基于Adaboost算法的水质组合预测方法研究

Research on Water Quality Combination Forecasting Method Based on Adaboost Algorithm

摘要:

水质预测是水环境污染防治的重要内容,针对传统水质预测方法精度低、收敛速度慢等问题,研究首先选取Symlets和Daubechies小波系作为小波函数,对原始数据进行去噪处理并对比,再结合RBF、Elman神经网络以及支持向量机各自优点,通过不同算法优化三种预测模型,提出基于Adaboost算法将优化后的RBF、Elman神经网络以及支持向量机相结合的组合预测方法。以北海为对象进行仿真实验,验证基于Adaboost的溶解氧组合预测方法的有效性,并分别与单一模型的预测结果进行对比,结果表明该方法相比于传统的单一模型预测精度得到了提高,为水质精准预测提供了一种新思路。

Abstract:

Water quality prediction is an important content of water pollution prevention and control, aiming at the problems such as low accuracy and slow convergence of traditional water quality prediction methods. In this paper, the Symlets and Daubechies wavelet systems are selected as the wavelet function, and the original data are denoised and compared. Combined with the advantages of RBF, Elman neural network and support vector machines, three different prediction models are optimized by different algorithms, a combined forecasting method based on Adaboost algorithm is proposed to optimize RBF, Elman neural network and support vector machine. Taking Beihai as the object to carry on the simulation experiment, the effectiveness of the method based on Adaboost"s dissolved oxygen combination is verified and compared with the prediction results of the single model. The results show that the proposed method is improved compared with the traditional single model, which provides a new idea for the accurate prediction of water quality.