摘要
针对水质指标数据所具有的非平稳季节性特征,在基于IGA-BP网络的水质预测模型基础上,提出考虑季节因素的SIGA-BP网络水质预测方法,通过构建季节性样本来凸现水质监测数据的季节性特征,用遗传算法优化BP网络的结构、隐层神经元阈值和连接权值,将输入层神经元个数加入编码方案和适应度函数,并改进选择算子,以上海青浦急水港2004-2016年逐月水质监测的DO值数据为例与IGA-BP网络和BP神经网络进行的水质预测对比实验,表明考虑季节因素的SIGA-BP神经网络模型进行水质预测更为有效。
Aiming at the non-stationary seasonal characteristics in the water quality data, proposes the SIGA-BP neural network water quality predic- tion method based on seasonal factors on the IGA-BP network water quality prediction model. The seasonal samples are rebuilt to emerge the seasonal characteristics for the water quality monitoring records, genetic algorithm is used to optimize the structure of BP network, the thresholds and the connection weights of hidden layer neural nodes, the number of input layer neurons is added into the coding scheme and the fitness function, and the selection operator is improved. The experiment results show that the SIGA-BP network water quality predic- tion method based on seasonal factors can predict water quality more effectively than both of the IGA-BP network and the BP neural net- work water quality prediction model do.
作者
李忠波
高茂庭
LI Zhong-bo GAO Mao-ting(College of Information Engineering, Shanghai Maritime University, Shanghai 201306)
出处
《现代计算机》
2017年第14期3-9,共7页
Modern Computer
基金
上海市科委科技创新项目(No.12595810200)
关键词
水质预测
季节因素
BP网络
遗传算法
Water Quality Prediction
Seasonal Factor
BP Network
Genetic Algorithm
