Agricultural greenhouse production has to require a stable and acceptable environment,it is therefore essential for future greenhouse production to obtain full and precisely internal dynamic environment parameters.Dyn...Agricultural greenhouse production has to require a stable and acceptable environment,it is therefore essential for future greenhouse production to obtain full and precisely internal dynamic environment parameters.Dynamic modeling based on machine learning methods,e.g.,intelligent time series prediction modeling,is a popular and suitable way to solve the above issue.In this article,a systematic literature review on applying advanced time series models has been systematically conducted via a detailed analysis and evaluation of 61 pieces selected from 221 articles.The historical process of time series model application from the use of data and information strategies was first discussed.Subsequently,the accuracy and generalization of the model from the selection of model parameters and time steps,providing a new perspective for model development in this field,were compared and analyzed.Finally,the systematic review results demonstrate that,compared with traditional models,deep neural networks could increase data structure mining capabilities and overall information simulation capabilities through innovative and effective structures,thereby it could also broaden the selection range of environmental parameters for agricultural facilities and achieve environmental prediction end-to-end optimization via intelligent time series model based on deep neural networks.展开更多
Environmental parameter data collected by sensors for monitoring the environment of agricultural facility operations are usually incomplete due to external environmental disturbances and device failures.And the missin...Environmental parameter data collected by sensors for monitoring the environment of agricultural facility operations are usually incomplete due to external environmental disturbances and device failures.And the missing of collected data is completely at random.In practice,missing data could create biased estimations and make multivariate time series predictions of environmental parameters difficult,leading to imprecise environmental control.A multivariate time series imputation model based on generative adversarial networks and multi-head attention(ATTN-GAN)is proposed in this work to reducing the negative consequence of missing data.ATTN-GAN can capture the temporal and spatial correlation of time series,and has a good capacity to learn data distribution.In the downstream experiments,we used ATTN-GAN and baseline models for data imputation,and predicted the imputed data,respectively.For the imputation of missing data,over the 20%,50%and 80%missing rate,ATTN-GAN had the lowest RMSE,0.1593,0.2012 and 0.2688 respectively.For water temperature prediction,data processed with ATTN-GAN over MLP,LSTM,DA-RNN prediction methods had the lowest MSE,0.6816,0.8375 and 0.3736 respectively.Those results revealed that ATTN-GAN outperformed all baseline models in terms of data imputation accuracy.The data processed by ATTN-GAN is the best for time series prediction.展开更多
基金Overseas High-level Youth Talents Program(China Agricultural University,China,Grant No.62339001)Science and Technology Cooperation-Sino-Malta Fund 2019:Research and Demonstration of Real-time Accurate Monitoring System for Early-stage Fish in Recirculating Aquaculture System(AquaDetector,Grant No.2019YFE0103700)+1 种基金China Agricultural University Excellent Talents Plan(Grant No.31051015)Major Science and Technology Innovation Fund 2019 of Shandong Province(Grant No.2019JZZY010703),National Innovation Center for Digital Fishery,and Beijing Engineering and Technology Research Center for Internet of Things in Agriculture.The authors also appreciate constructive and valuable comments provided by reviewers.
文摘Agricultural greenhouse production has to require a stable and acceptable environment,it is therefore essential for future greenhouse production to obtain full and precisely internal dynamic environment parameters.Dynamic modeling based on machine learning methods,e.g.,intelligent time series prediction modeling,is a popular and suitable way to solve the above issue.In this article,a systematic literature review on applying advanced time series models has been systematically conducted via a detailed analysis and evaluation of 61 pieces selected from 221 articles.The historical process of time series model application from the use of data and information strategies was first discussed.Subsequently,the accuracy and generalization of the model from the selection of model parameters and time steps,providing a new perspective for model development in this field,were compared and analyzed.Finally,the systematic review results demonstrate that,compared with traditional models,deep neural networks could increase data structure mining capabilities and overall information simulation capabilities through innovative and effective structures,thereby it could also broaden the selection range of environmental parameters for agricultural facilities and achieve environmental prediction end-to-end optimization via intelligent time series model based on deep neural networks.
基金supported by the National Natural Science Foundation of China:“Regularity and prediction model of juvenile fish growth under synergistic effect of water temperature and flow fields in recirculating aquaculture”(Grant No.32373185)2115 Talent Development Program of China Agricultural University,Overseas High-level Youth Talents Program(China Agricultural University,China,Grant No.62339001)+2 种基金China Agricultural University Excellent Talents Plan(Grant No.31051015)Major Science and Technology Innovation Fund 2019 of Shandong Province(Grant No.2019JZZY010703)National Innovation Center for Digital Fishery,and Beijing Engineering and Technology Research Center for Internet of Things in Agriculture.
文摘Environmental parameter data collected by sensors for monitoring the environment of agricultural facility operations are usually incomplete due to external environmental disturbances and device failures.And the missing of collected data is completely at random.In practice,missing data could create biased estimations and make multivariate time series predictions of environmental parameters difficult,leading to imprecise environmental control.A multivariate time series imputation model based on generative adversarial networks and multi-head attention(ATTN-GAN)is proposed in this work to reducing the negative consequence of missing data.ATTN-GAN can capture the temporal and spatial correlation of time series,and has a good capacity to learn data distribution.In the downstream experiments,we used ATTN-GAN and baseline models for data imputation,and predicted the imputed data,respectively.For the imputation of missing data,over the 20%,50%and 80%missing rate,ATTN-GAN had the lowest RMSE,0.1593,0.2012 and 0.2688 respectively.For water temperature prediction,data processed with ATTN-GAN over MLP,LSTM,DA-RNN prediction methods had the lowest MSE,0.6816,0.8375 and 0.3736 respectively.Those results revealed that ATTN-GAN outperformed all baseline models in terms of data imputation accuracy.The data processed by ATTN-GAN is the best for time series prediction.
