摘要
为解决传统方法水质预测精度低、鲁棒性差等问题,提出了基于小波分析(WA)、人工蜂群(ABC)优化加权最小二乘支持向量回归机(WLSSVR)的工厂化育苗溶解氧组合预测模型(WA-ABC-WLSSVR模型).该模型采用小波分析对原始非平稳溶解氧时间序列数据进行多尺度特征提取,通过加权最小二乘支持向量回归机对不同尺度下的溶解氧数据子序列分别建模,利用改进人工蜂群优化算法(ABC)对各分量序列WLSSVR模型参数进行组合优化,最后叠加各尺度下的预测结果.运用该模型对工厂化育苗溶解氧进行预测,并与BPNN、标准LSSVR、WAACO-LSSVR、WA-PSO-LSSVR等模型对比分析,结果表明,该溶解氧预测模型具有较高的预测精度和泛化能力.
In order to solve the low prediction accuracy and bad robustness of the traditional predicting methods in water quality, the paper puts forward the prediction model of dissolved oxygen content based on wavelet analysis （WA） and weighted least squares support vector machine （WLSSVM）, in which the hyper-parameters is optimized by improved artificial bee colony optimization algorithm（ABC）. This paper uses the advantage of multiscale wavelet analysis in dealing with the non-stationary signal to decompose and reconstruct the dissolved oxygen content monitoring data sequences. Then different WLSSVM models are built to predict the data sequence mining of different scales; the improved artificial bee colony optimiza- tion algorithm （ABC） is adopted to optimize and choose parameters combinations for WLSSVR model of each data sequence. The predicting results of each scale are accumulated finally. The proposed hybrid model is applied to predict dissolved oxygen in industrialized vannamei breeding ponds. The results show that： compared with BPNN, LSSVR, WA-ACO-LSSVR, WA-PSO-LSSVR model, the proposed dissolved oxygen content predicting model in this paper has high predicting accuracy and generalization ability.
作者
徐龙琴
陈跃霞
张军
刘双印
李道亮
XU Longqin CHEN Yuexia ZHANG Jun LIU Shuangyin LI Daoliang(School of Information Science and Technology, Zhongkai University of Agriculture and Engineering, Guangzhou 510225, China R＆D Center Flight Control Hydraulic Institute, AVIC Aircraft Co. ,Ltd, Hanzhong 723213, China AVIC Aircraft Aerostructure Manufacturing（Hanzhong） Co. ,Ltd, Hanzhong 723213, China Beijing ERC for Internet of Things in Agriculture, China Agricultural University, Beijing 100083, China)
出处
《武汉大学学报（工学版）》
CAS
CSCD
北大核心
2017年第4期608-617,共10页
Engineering Journal of Wuhan University
基金
国家自然科学基金项目(编号:61471133
61473331)
国家科技支撑计划项目(编号:2012BAD35B07)
广东省科技计划项目(编号:2013B090500127
2013B021600014
2015A070709015
2015A020209171
2016A040402043)
广东省自然基金项目(编号:S2013010014629
2014A030307049)
广东海洋大学创新强校工程项目(编号:GDOU2014050227)
关键词
溶解氧预测
加权最小二乘支持向量回归机
人工蜂群算法
小波分析
参数组合优化
dissolved oxygen prediction
weighted least squares support vector regression
artificial bee col- ony algorithm
wavelet analysis
parameter composite optimization
