Since the early 1980 s, the multi-cropping index for rice has decreased significantly in main double-cropping rice area in China, which is the primary double-cropping rice(DCR) production area. This decline may bring ...Since the early 1980 s, the multi-cropping index for rice has decreased significantly in main double-cropping rice area in China, which is the primary double-cropping rice(DCR) production area. This decline may bring challenges to food security in China because rice is the staple food for more than 60% of the Chinese population. It has been generally recognized that rapidly rising labor costs due to economic growth and urbanization in China is the key driving force of the ‘double-to-single' rice cropping system adaption. However, not all provinces have shown a dramatic decline in DCR area, and labor costs alone cannot explain this difference. To elucidate the reasons for these inter-provincial distinctions and the dynamics of rice cropping system adaption, we evaluated the influencing factors using provincial panel data from 1980 to 2015. We also used household survey data for empirical analysis to explore the mechanisms driving differences in rice multi-cropping changes. Our results indicated that the eight provinces in the study can be divided into three spatial groups based on the extent of DCR area decline, the rapidly-declining marginal, core, and stable zones. Increasing labor cost due to rapid urbanization was the key driving force of rice cropping system adaption, but the land use dynamic vary hugely among different provinces. These differences between zones were due to the interaction between labor price and accumulated temperature conditions. Therefore, increasing labor costs had the greatest impact in Zhejiang, Anhui, and Hubei, where the accumulated temperature is relatively low and rice multi-cropping index declined dramaticly. However, labor costs had little impact in Guangdong and Guangxi. Differences in accumulated temperature conditions resulted in spatially different labor demands and pressure on households during the busy season. As a result, there have been different profits and rice multi-cropping changes between provinces and zones. Because of these spatial differences, regionally appropriate policies that provide appropriate subsidies for early rice in rapidly-declining marginal zone such as Zhejiang and Hubei should be implemented. In addition, agricultural mechanization and the number of agricultural workers have facilitated double-cropping; therefore, small machinery and agricultural infrastructure construction should be further supported.展开更多
In recent times,the use of artificial intelligence(AI)in agriculture has become the most important.The technology adoption in agriculture if creatively approached.Controlling on the diseased leaves during the growing ...In recent times,the use of artificial intelligence(AI)in agriculture has become the most important.The technology adoption in agriculture if creatively approached.Controlling on the diseased leaves during the growing stages of crops is a crucial step.The disease detection,classification,and analysis of diseased leaves at an early stage,as well as possible solutions,are always helpful in agricultural progress.The disease detection and classification of different crops,especially tomatoes and grapes,is a major emphasis of our proposed research.The important objective is to forecast the sort of illness that would affect grapes and tomato leaves at an early stage.The Convolutional Neural Network(CNN)methods are used for detecting Multi-Crops Leaf Disease(MCLD).The features extraction of images using a deep learning-based model classified the sick and healthy leaves.The CNN based Visual Geometry Group(VGG)model is used for improved performance measures.The crops leaves images dataset is considered for training and testing the model.The performance measure parameters,i.e.,accuracy,sensitivity,specificity precision,recall and F1-score were calculated and monitored.The main objective of research with the proposed model is to make on-going improvements in the performance.The designed model classifies disease-affected leaves with greater accuracy.In the experiment proposed research has achieved an accuracy of 98.40%of grapes and 95.71%of tomatoes.The proposed research directly supports increasing food production in agriculture.展开更多
[目的/意义]为解决当前作物管理中个性化需求难以捕捉、决策过程缺乏灵活性难题,本研究提出了一种基于大语言模型的个性化作物生产智能决策方法[方法]通过自然语言对话收集用户在蔬菜作物管理过程中的个性化需求,涵盖产量、人力资源消...[目的/意义]为解决当前作物管理中个性化需求难以捕捉、决策过程缺乏灵活性难题,本研究提出了一种基于大语言模型的个性化作物生产智能决策方法[方法]通过自然语言对话收集用户在蔬菜作物管理过程中的个性化需求,涵盖产量、人力资源消耗和水肥消耗等方面。随后,将作物管理过程建模为多目标优化问题,同时考虑用户个性化偏好和作物产量,并采用强化学习算法来学习作物管理策略。水肥管理策略的训练通过与环境的交互持续更新,学习在不同条件下采取何种行动以实现最优决策,从而实现个性化的作物管理。[结果和讨论]在gym-DSSAT(Gym-Decision Support System for Agrotechnology Transfer)仿真平台上进行的实验,结果表明,所提出的个性化作物生产智能决策方法能够有效地根据用户的个性化偏好调整作物管理策略。[结论]通过精准捕捉用户的个性化需求,该方法在保证作物产量的同时,优化了人力资源与水肥资源的消耗。展开更多
基金National Program on Key Basic Research Project(No.2015CB452706)
文摘Since the early 1980 s, the multi-cropping index for rice has decreased significantly in main double-cropping rice area in China, which is the primary double-cropping rice(DCR) production area. This decline may bring challenges to food security in China because rice is the staple food for more than 60% of the Chinese population. It has been generally recognized that rapidly rising labor costs due to economic growth and urbanization in China is the key driving force of the ‘double-to-single' rice cropping system adaption. However, not all provinces have shown a dramatic decline in DCR area, and labor costs alone cannot explain this difference. To elucidate the reasons for these inter-provincial distinctions and the dynamics of rice cropping system adaption, we evaluated the influencing factors using provincial panel data from 1980 to 2015. We also used household survey data for empirical analysis to explore the mechanisms driving differences in rice multi-cropping changes. Our results indicated that the eight provinces in the study can be divided into three spatial groups based on the extent of DCR area decline, the rapidly-declining marginal, core, and stable zones. Increasing labor cost due to rapid urbanization was the key driving force of rice cropping system adaption, but the land use dynamic vary hugely among different provinces. These differences between zones were due to the interaction between labor price and accumulated temperature conditions. Therefore, increasing labor costs had the greatest impact in Zhejiang, Anhui, and Hubei, where the accumulated temperature is relatively low and rice multi-cropping index declined dramaticly. However, labor costs had little impact in Guangdong and Guangxi. Differences in accumulated temperature conditions resulted in spatially different labor demands and pressure on households during the busy season. As a result, there have been different profits and rice multi-cropping changes between provinces and zones. Because of these spatial differences, regionally appropriate policies that provide appropriate subsidies for early rice in rapidly-declining marginal zone such as Zhejiang and Hubei should be implemented. In addition, agricultural mechanization and the number of agricultural workers have facilitated double-cropping; therefore, small machinery and agricultural infrastructure construction should be further supported.
文摘In recent times,the use of artificial intelligence(AI)in agriculture has become the most important.The technology adoption in agriculture if creatively approached.Controlling on the diseased leaves during the growing stages of crops is a crucial step.The disease detection,classification,and analysis of diseased leaves at an early stage,as well as possible solutions,are always helpful in agricultural progress.The disease detection and classification of different crops,especially tomatoes and grapes,is a major emphasis of our proposed research.The important objective is to forecast the sort of illness that would affect grapes and tomato leaves at an early stage.The Convolutional Neural Network(CNN)methods are used for detecting Multi-Crops Leaf Disease(MCLD).The features extraction of images using a deep learning-based model classified the sick and healthy leaves.The CNN based Visual Geometry Group(VGG)model is used for improved performance measures.The crops leaves images dataset is considered for training and testing the model.The performance measure parameters,i.e.,accuracy,sensitivity,specificity precision,recall and F1-score were calculated and monitored.The main objective of research with the proposed model is to make on-going improvements in the performance.The designed model classifies disease-affected leaves with greater accuracy.In the experiment proposed research has achieved an accuracy of 98.40%of grapes and 95.71%of tomatoes.The proposed research directly supports increasing food production in agriculture.
文摘[目的/意义]为解决当前作物管理中个性化需求难以捕捉、决策过程缺乏灵活性难题,本研究提出了一种基于大语言模型的个性化作物生产智能决策方法[方法]通过自然语言对话收集用户在蔬菜作物管理过程中的个性化需求,涵盖产量、人力资源消耗和水肥消耗等方面。随后,将作物管理过程建模为多目标优化问题,同时考虑用户个性化偏好和作物产量,并采用强化学习算法来学习作物管理策略。水肥管理策略的训练通过与环境的交互持续更新,学习在不同条件下采取何种行动以实现最优决策,从而实现个性化的作物管理。[结果和讨论]在gym-DSSAT(Gym-Decision Support System for Agrotechnology Transfer)仿真平台上进行的实验,结果表明,所提出的个性化作物生产智能决策方法能够有效地根据用户的个性化偏好调整作物管理策略。[结论]通过精准捕捉用户的个性化需求,该方法在保证作物产量的同时,优化了人力资源与水肥资源的消耗。
