This study aimed to address the challenge of accurately and reliably detecting tomatoes in dense planting environments,a critical prerequisite for the automation implementation of robotic harvesting.However,the heavy ...This study aimed to address the challenge of accurately and reliably detecting tomatoes in dense planting environments,a critical prerequisite for the automation implementation of robotic harvesting.However,the heavy reliance on extensive manually annotated datasets for training deep learning models still poses significant limitations to their application in real-world agricultural production environments.To overcome these limitations,we employed domain adaptive learning approach combined with the YOLOv5 model to develop a novel tomato detection model called as TDA-YOLO(tomato detection domain adaptation).We designated the normal illumination scenes in dense planting environments as the source domain and utilized various other illumination scenes as the target domain.To construct bridge mechanism between source and target domains,neural preset for color style transfer is introduced to generate a pseudo-dataset,which served to deal with domain discrepancy.Furthermore,this study combines the semi-supervised learning method to enable the model to extract domain-invariant features more fully,and uses knowledge distillation to improve the model's ability to adapt to the target domain.Additionally,for purpose of promoting inference speed and low computational demand,the lightweight FasterNet network was integrated into the YOLOv5's C3 module,creating a modified C3_Faster module.The experimental results demonstrated that the proposed TDA-YOLO model significantly outperformed original YOLOv5s model,achieving a mAP(mean average precision)of 96.80%for tomato detection across diverse scenarios in dense planting environments,increasing by 7.19 percentage points;Compared with the latest YOLOv8 and YOLOv9,it is also 2.17 and 1.19 percentage points higher,respectively.The model's average detection time per image was an impressive 15 milliseconds,with a FLOPs(floating point operations per second)count of 13.8 G.After acceleration processing,the detection accuracy of the TDA-YOLO model on the Jetson Xavier NX development board is 90.95%,the mAP value is 91.35%,and the detection time of each image is 21 ms,which can still meet the requirements of real-time detection of tomatoes in dense planting environment.The experimental results show that the proposed TDA-YOLO model can accurately and quickly detect tomatoes in dense planting environment,and at the same time avoid the use of a large number of annotated data,which provides technical support for the development of automatic harvesting systems for tomatoes and other fruits.展开更多
Different irrigation schemes have different effects on water consumption in rice production.However,few studies have been conducted on the water consumption processes between dry direct seeding rice and transplanting ...Different irrigation schemes have different effects on water consumption in rice production.However,few studies have been conducted on the water consumption processes between dry direct seeding rice and transplanting rice under different irrigation schemes.Water consumption process,water use efficiency and correlation effect of water consumption on yield under different planting models in rice production were investigated in northeast China in 2018.Seven treatments were implemented:drip irrigation dry direct seeding rice(DDSR),wet irrigation dry direct seeding rice(WDSR),flooded irrigation dry direct seeding rice(FDSR),transplanting flooded rice(TFR),controlled irrigation transplanting rice(CTR),intermittent irrigation transplanting rice(ITR)and wet irrigation transplanting rice(WTR).Among them,TFR was the control.The results showed that the peaks of the water consumption amount,intensity and its modulus coefficient of the seven treatments all appeared in the middle tillering and the jointing booting stages.The total water consumption amount(ET)and average water consumption intensity of DDSR,WDSR,FDSR and WTR were lower than those of TFR,CTR and ITR.The maximum water use efficiency of yield(WUEy)occurred in DDSR with a value of 3.8 kg·m^(-3).WUEy of DDSR,WDSR and FDSR were significantly higher than those of TFR,CTR and ITR.In the middle tillering and the heading and flowering stages,the water consumption amount of each treatment had a positive effect on yield formation,and the water consumption amount in the late tillering stage had a negative effect on yield formation.The relationship between ET and yield(Y)of dry direct seeding and transplanting planting models showed a quadratic function curve.ET of transplanting planting model had a significant positive impact on Y,and ET of dry direct seeding planting model had no impact on Y.DDSR had the least total water consumption of 199.8 mm·m^(-2),the lowest water consumption intensity of 2.0 mm·d^(-1) and the greatest water use efficiency of 3.8 kg·m^(-3),which suggested that DDSR had the most significant water saving effect.The combination of dry direct seeding planting model and drip irrigation scheme would be a good option for determining a water-saving rice planting model in northeast China.展开更多
虚拟电厂聚合分布式能源作为第三方主体参与市场,其交易过程存有多种不确定性风险因素,准确识别并有效评估其交易风险尤为重要。该文首先基于文本挖掘技术辨识风险因素,并使用失效模式与影响分析法确定关键风险因素,进而设计风险评估指...虚拟电厂聚合分布式能源作为第三方主体参与市场,其交易过程存有多种不确定性风险因素,准确识别并有效评估其交易风险尤为重要。该文首先基于文本挖掘技术辨识风险因素,并使用失效模式与影响分析法确定关键风险因素,进而设计风险评估指标体系。其次,结合博弈论思想,对关键风险因素主客观组合赋权。再次,构建风险评估的二维云模型以描述风险发生概率的随机性和风险产生后果的模糊性问题。最后,采用所提评估方法计算多场景虚拟电厂参与市场交易情况的总体风险水平并排序,且与优劣解距离法(technique for order preference by similarity to ideal solution,TOPSIS)、秩和比综合评价法(rank sum ratio,RSR)及折衷排序方法(multi-criteria optimization and compromise solution,VIKOR)对比分析,验证了模型及方法的可行性和有效性。所做研究为VPP交易管理和风险防范提供了有益的参考,具有工程应用价值。展开更多
基金The National Natural Science Foundation of China (32371993)The Natural Science Research Key Project of Anhui Provincial University(2022AH040125&2023AH040135)The Key Research and Development Plan of Anhui Province (202204c06020022&2023n06020057)。
文摘This study aimed to address the challenge of accurately and reliably detecting tomatoes in dense planting environments,a critical prerequisite for the automation implementation of robotic harvesting.However,the heavy reliance on extensive manually annotated datasets for training deep learning models still poses significant limitations to their application in real-world agricultural production environments.To overcome these limitations,we employed domain adaptive learning approach combined with the YOLOv5 model to develop a novel tomato detection model called as TDA-YOLO(tomato detection domain adaptation).We designated the normal illumination scenes in dense planting environments as the source domain and utilized various other illumination scenes as the target domain.To construct bridge mechanism between source and target domains,neural preset for color style transfer is introduced to generate a pseudo-dataset,which served to deal with domain discrepancy.Furthermore,this study combines the semi-supervised learning method to enable the model to extract domain-invariant features more fully,and uses knowledge distillation to improve the model's ability to adapt to the target domain.Additionally,for purpose of promoting inference speed and low computational demand,the lightweight FasterNet network was integrated into the YOLOv5's C3 module,creating a modified C3_Faster module.The experimental results demonstrated that the proposed TDA-YOLO model significantly outperformed original YOLOv5s model,achieving a mAP(mean average precision)of 96.80%for tomato detection across diverse scenarios in dense planting environments,increasing by 7.19 percentage points;Compared with the latest YOLOv8 and YOLOv9,it is also 2.17 and 1.19 percentage points higher,respectively.The model's average detection time per image was an impressive 15 milliseconds,with a FLOPs(floating point operations per second)count of 13.8 G.After acceleration processing,the detection accuracy of the TDA-YOLO model on the Jetson Xavier NX development board is 90.95%,the mAP value is 91.35%,and the detection time of each image is 21 ms,which can still meet the requirements of real-time detection of tomatoes in dense planting environment.The experimental results show that the proposed TDA-YOLO model can accurately and quickly detect tomatoes in dense planting environment,and at the same time avoid the use of a large number of annotated data,which provides technical support for the development of automatic harvesting systems for tomatoes and other fruits.
基金Supported by the National Key Research and Development Program of China(2016YFC040010101)。
文摘Different irrigation schemes have different effects on water consumption in rice production.However,few studies have been conducted on the water consumption processes between dry direct seeding rice and transplanting rice under different irrigation schemes.Water consumption process,water use efficiency and correlation effect of water consumption on yield under different planting models in rice production were investigated in northeast China in 2018.Seven treatments were implemented:drip irrigation dry direct seeding rice(DDSR),wet irrigation dry direct seeding rice(WDSR),flooded irrigation dry direct seeding rice(FDSR),transplanting flooded rice(TFR),controlled irrigation transplanting rice(CTR),intermittent irrigation transplanting rice(ITR)and wet irrigation transplanting rice(WTR).Among them,TFR was the control.The results showed that the peaks of the water consumption amount,intensity and its modulus coefficient of the seven treatments all appeared in the middle tillering and the jointing booting stages.The total water consumption amount(ET)and average water consumption intensity of DDSR,WDSR,FDSR and WTR were lower than those of TFR,CTR and ITR.The maximum water use efficiency of yield(WUEy)occurred in DDSR with a value of 3.8 kg·m^(-3).WUEy of DDSR,WDSR and FDSR were significantly higher than those of TFR,CTR and ITR.In the middle tillering and the heading and flowering stages,the water consumption amount of each treatment had a positive effect on yield formation,and the water consumption amount in the late tillering stage had a negative effect on yield formation.The relationship between ET and yield(Y)of dry direct seeding and transplanting planting models showed a quadratic function curve.ET of transplanting planting model had a significant positive impact on Y,and ET of dry direct seeding planting model had no impact on Y.DDSR had the least total water consumption of 199.8 mm·m^(-2),the lowest water consumption intensity of 2.0 mm·d^(-1) and the greatest water use efficiency of 3.8 kg·m^(-3),which suggested that DDSR had the most significant water saving effect.The combination of dry direct seeding planting model and drip irrigation scheme would be a good option for determining a water-saving rice planting model in northeast China.
文摘虚拟电厂聚合分布式能源作为第三方主体参与市场,其交易过程存有多种不确定性风险因素,准确识别并有效评估其交易风险尤为重要。该文首先基于文本挖掘技术辨识风险因素,并使用失效模式与影响分析法确定关键风险因素,进而设计风险评估指标体系。其次,结合博弈论思想,对关键风险因素主客观组合赋权。再次,构建风险评估的二维云模型以描述风险发生概率的随机性和风险产生后果的模糊性问题。最后,采用所提评估方法计算多场景虚拟电厂参与市场交易情况的总体风险水平并排序,且与优劣解距离法(technique for order preference by similarity to ideal solution,TOPSIS)、秩和比综合评价法(rank sum ratio,RSR)及折衷排序方法(multi-criteria optimization and compromise solution,VIKOR)对比分析,验证了模型及方法的可行性和有效性。所做研究为VPP交易管理和风险防范提供了有益的参考,具有工程应用价值。
