[1]冯国富,卢胜涛,陈明,等.基于自注意力机制和改进的K-BiLSTM的水产养殖水体溶解氧含量预测模型[J].江苏农业学报,2024,(03):490-499.[doi:doi:10.3969/j.issn.1000-4440.2024.03.011]
　FENG Guo-fu,LU Sheng-tao,CHEN Ming,et al.Prediction model of dissolved oxygen content in aquaculture water based on self-attention mechanism and improved K-BiLSTM[J].,2024,(03):490-499.[doi:doi:10.3969/j.issn.1000-4440.2024.03.011]
点击复制

基于自注意力机制和改进的K-BiLSTM的水产养殖水体溶解氧含量预测模型()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2024年03期
页码:
490-499
栏目:
农业信息工程
出版日期:
2024-03-30

文章信息/Info

Title:
Prediction model of dissolved oxygen content in aquaculture water based on self-attention mechanism and improved K-BiLSTM
作者:
冯国富1; 2; 卢胜涛1; 2; 陈明1; 2; 王耀辉3
(1.上海海洋大学信息学院,上海201306；2.农业农村部渔业信息重点实验室,上海201306；3.南通龙洋水产有限公司,江苏南通226634)
Author(s):
FENG Guo-fu1; 2; LU Sheng-tao1; 2;  CHEN Ming1; 2; WANG Yao-hui3
(1.College of Information Technology, Shanghai Ocean University, Shanghai 201306, China;2.Key Laboratory of Fisheries Information, Ministry of Agriculture and Rural Affairs, Shanghai 201306, China;3.Nantong Longyang Aquatic Products Co., Ltd., Nantong 226634, China)
关键词:
水产养殖; 溶解氧预测; K-means聚类; 双向长短期记忆网络（BiLSTM）; 自注意力机制
Keywords:
aquaculture; dissolved oxygen prediction; K-means clustering; bidirectional long-term and short-term memory network (BiLSTM); self-attention mechanism
分类号:
TP391
DOI:
doi:10.3969/j.issn.1000-4440.2024.03.011
摘要:
为精确预测水产养殖水体溶解氧含量,本研究提出一种基于自注意力机制(ATTN)和改进的K-means聚类-基于残差和批标准化(BN)的双向长短期记忆网络(BiLSTM)的水产养殖水体溶解氧含量预测模型。首先,根据环境数据的相似性,使用改进的K-means算法将数据划分成若干个类别；然后,在BiLSTM基础上构建残差连接和加入BN完成高层次特征提取,利用BiLSTM的长期记忆能力保存特征信息；最后,引入自注意力机制突出不同时间节点数据特征的重要性,进一步提升模型的性能。试验结果表明,本研究提出的基于自注意力机制和改进的K-BiLSTM模型的平均绝对误差为0.238、均方根误差为0.322、平均绝对百分比误差为0.035,与单一的BP模型、CNN-LSTM模型、传统的K-means-基于残差和BN的BiLSTM-ATTN等模型相比具有更优的预测性能和泛化能力。
Abstract:
In order to accurately predict the content of dissolved oxygen (DO) in aquaculture water, a prediction model of dissolved oxygen content in aquaculture water based on self-attention mechanism (ATTN) and improved K-means clustering-bidirectional long-term and short-term memory network (BiLSTM) was proposed. Firstly, according to the similarity of environmental data, the improved K-means algorithm was used to divide environmental data into several categories. Then, based on BiLSTM, residual connection was constructed and batch normalization (BN) was added to complete high-level feature extraction, and the feature information was saved by the long-term memory ability of BiLSTM. Finally, the self-attention mechanism was introduced to highlight the importance of data characteristics at different time nodes, which further improved the performance of the model. The experimental results showed that the mean absolute error (MAE), root mean square error (RMSE) and average absolute percentage error (MAPE) of the hybrid model based on self-attention mechanism and improved K-BiLSTM were 0.238, 0.322 and 0.035, respectively. Compared with single BP model, CNN-LSTM model and traditional K-means-BiLSTM-ATTN model based on residual and BN, the model constructed in this study had better prediction performance and generalization ability.

参考文献/References:

[1]LIPIZER M, PARTESCANO E, RABITTI A, et al. Qualified temperature, salinity and dissolved oxygen climatologies in a changing Adriatic Sea[J]. Ocean Science,2014,10(5):771-797.
[2]陈英义,程倩倩,方晓敏,等. 主成分分析和长短时记忆神经网络预测水产养殖水体溶解氧[J]. 农业工程学报,2018,34(17):183-191.
[3]金光炎. 水文统计理论与实践[M]. 南京:东南大学出版社,2012.
[4]刘明,李由明,王平,等. 基于小波分解的凡纳滨对虾养殖水体水质的仿真研究[J]. 广东农业科学,2013,40(17):170-172.
[5]徐梅,晏福,刘振忠,等. 灰色GM(1,1)-小波变换-GARCH组合模型预测松花江流域水质[J]. 农业工程学报,2016,32(10):137-142.
[6]岳遥,李天宏. 基于模糊集理论的马尔可夫模型在水质定量预测中的应用[J]. 应用基础与工程科学学报,2011,19(2):231-242.
[7]黄廷林,韩晓刚,卢金锁. 基于Lyapunov指数的混沌预测方法及在水质预测中的应用[J]. 西安建筑科技大学学报(自然科学版),2008,40(6):846-851.
[8]ALVAREZ MEZA A M, DAZA SANTACOLOMA G. Parameter selection in least squares support vector machines regression oriented,using generalized cross-validation[J]. Dyna-Colombia,2012,79(171):23-30.
[9]邹志红,王学良. BP模型在河流水质预测中的误差分析[J].环境科学学报,2007,27(6):1038-1042.
[10]AMID S, GUNDOSHMIAN T M. Prediction of output energies for broiler production using linear regression, ANN (MLP, RBF), and ANFIS models[J]. Environmental Progress & Sustainable Energy,2017,36(2):577-585.
[11]刘东君,邹志红. 最优加权组合预测法在水质预测中的应用研究[J]. 环境科学学报,2012,32(12):3128-3132.
[12]刘双印,徐龙琴,李振波,等. 基于PCA-MCAFA-LSSVM的养殖水质pH值预测模型[J]. 农业机械学报,2014,45(5):239-246.
[13]龚怀瑾,毛力,杨弘. 基于变尺度混沌QPSO-LSSVM的水质溶氧预测建模[J]. 计算机与应用化学,2013,30(3):315-318.
[14]孙伯寅,董国庆,张荣. 支持向量机在水源水化学耗氧量预测中的应用[J]. 环境与健康杂志,2016,33(6):544-547.
[15]罗华军,黄应平,刘德富. 基于WA-SVM的水库溶解氧预测[J]. 西北农林科技大学学报(自然科学版),2009,37(3):181-186.
[16]宦娟,刘星桥. 基于K-means聚类和ELM神经网络的养殖水质溶解氧预测[J]. 农业工程学报,2016,32(17):174-181.
[17]陈英义,方晓敏,梅思远,等. 基于WT-CNN-LSTM的溶解氧含量预测模型[J]. 农业机械学报,2020,51(10):284-291.
[18]曹守启,周礼馨,张铮. 采用改进长短时记忆神经网络的水产养殖溶解氧预测模型[J]. 农业工程学报,2021,37(14):235-242.
[19]WU Y H, SUN L Q, SUN X B, et al. A hybrid XGBoost-ISSA-LSTM model for accurate short-term and long-term dissolved oxygen prediction in ponds[J]. Environ Sci Pollut Res Int,2021,29(12):18142-18159.
[20]YANG H H, LIU S E. Water quality prediction in sea cucumber farming based on a GRU neural network optimized by an improved whale optimization algorithm[J]. PeerJ Comput Sci,2022,8:e1000.
[21]ZOU Q H, XIONG Q Y, LI Q D, et al. A water quality prediction method based on the multi-time scale bidirectional long short-term memory network[J]. Environmental Science and Pollution Research,2020,27(9):16853-16864.
[22]YANG W B, LIU W, GAO Q. Prediction of dissolved oxygen concentration in aquaculture based on attention mechanism and combined neural network[J]. Math Biosci Eng,2023,20(1):998-1017.
[23]ZHANG Q, WANG R Q, QI Y, et al. A watershed water quality prediction model based on attention mechanism and Bi-LSTM[J]. Environmental Science and Pollution Research,2022,29(50):75664-75680.
[24]LI Y T, LI R. Predicting ammonia nitrogen in surface water by a new attention-based deep learning hybrid model[J]. Environmental Research,2023,216:114723.
[25]CAO X K, LIU Y R, WANG J P, et al. Prediction of dissolved oxygen in pond culture water based on K-means clustering and gated recurrent unit neural network[J]. Aquacultural Engineering, 2020,91:102122.
[26]何津民,张丽珍. 基于自注意力机制和CNN-LSTM深度学习的对虾投饵量预测模型[J]. 大连海洋大学学报,2022,37(2):304-311.

相似文献/References:

[1]刘国锋,徐跑,吴霆,等.中国水产养殖环境氮磷污染现状及未来发展思路[J].江苏农业学报,2018,(01):225.[doi:doi:10.3969/j.issn.1000-4440.2018.01.033]
　LIU Guo-feng,XU Pao,WU Ting,et al.Present condition of aquaculture nitrogen and phosphorus environmental pollution and future development strategy[J].,2018,(03):225.[doi:doi:10.3969/j.issn.1000-4440.2018.01.033]
[2]任妮,鲍彤,刘杨,等.基于粒子群优化算法和长短时记忆神经网络的蟹塘溶解氧预测[J].江苏农业学报,2021,(02):426.[doi:doi:10.3969/j.issn.1000-4440.2021.02.020]
　REN Ni,BAO Tong,LIU Yang,et al.Prediction model of dissolved oxygen in Chinese mitten crab ponds based on particle swarm optimization algorithm and long short-term memory neural networks[J].,2021,(03):426.[doi:doi:10.3969/j.issn.1000-4440.2021.02.020]
[3]唐毅,徐全,杜彬,等.基于SARIMA-VMD-LSSVM的水产养殖溶解氧质量浓度预测[J].江苏农业学报,2024,(08):1473.[doi:doi:10.3969/j.issn.1000-4440.2024.08.012]
　TANG Yi,XU Quan,DU Bin,et al.Prediction of dissolved oxygen mass concentration in aquaculture based on SARIMA-VMD-LSSVM[J].,2024,(03):1473.[doi:doi:10.3969/j.issn.1000-4440.2024.08.012]

备注/Memo

备注/Memo:
收稿日期：2023-01-28基金项目：江苏现代农业产业关键技术创新项目[CX(20)2028]；广东省重点领域研发计划项目(2021B0202070001)作者简介：冯国富(1971-),男,河南鹤壁人，博士,副教授,研究方向为嵌入式技术研究。(E-mail)gffeng202212@163.com通讯作者：陈明, (E-mail)chengmm202212@163.com
更新日期/Last Update: 2024-05-20