Climate change can aff ect fi sh individuals or schools,and consequently the fi sheries.Studying future changes of fi sh distribution and abundance helps the scientifi c management of fi sheries.The dynamic bioclimate...Climate change can aff ect fi sh individuals or schools,and consequently the fi sheries.Studying future changes of fi sh distribution and abundance helps the scientifi c management of fi sheries.The dynamic bioclimate envelope model(DBEM)was used to identify the“environmental preference profi les”of the studied species based on outputs from three Earth system models(ESMs).Changes in ocean conditions in climate change scenarios could be transformed by the model into those in relative abundance and distribution of species.Therefore,the distributional response of 17 demersal fi shes to climate change in the Yellow Sea could be projected from 1970 to 2060.Indices of latitudinal centroid(LC)and mean temperature of relative abundance(MTRA)were used to represent the results conducted by model.Results present that 17 demersal fi sh species in the Yellow Sea show a trend of anti-poleward shift under both low-emission scenario(RCP 2.6)and high-emission scenario(RCP 8.5)from 1970 to 2060,with the projected average LC in three ESMs shifting at a rate of-1.17±4.55 and-2.76±3.82 km/decade,respectively,which is contrary to the previous projecting studies of fi shes suggesting that fi shes tend to move toward higher latitudes under increased temperature scenarios.The Yellow Sea Cold Water Mass could be the major driver resulting in the shift,which shows a potential signifi cance to fi shery resources management and marine conservation,and provides a new perspective in fi sh migration under climate change.展开更多
By establishing a distribution and environmental factor database of 21 typical harmful dinoflagellates in global waters,the MaxEnt model was used to predict shifts in the habitat of harmful dinoflagellates in Chinese ...By establishing a distribution and environmental factor database of 21 typical harmful dinoflagellates in global waters,the MaxEnt model was used to predict shifts in the habitat of harmful dinoflagellates in Chinese waters under global climate change.The results revealed that offshore distance was the most important predictive factor and that surface seawater temperature(SST),primary productivity,and nitrate concentration were the key ecological factors influencing the distribution of harmful dinoflagellates.Under the low greenhouse gas emission scenario defined by the Intergovernmental Panel on Climate Change(IPCC),by approximately 2050,17 of the 21harmful dinoflagellate species in high-suitability areas(HSA)will migrate northward,six species will migrate eastward,and six species will expand their HSA.By 2100,approximately 18 of the 21 harmful dinoflagellate species in HSA will have migrated northward,seven species will have migrated eastward,and four species will have expanded their HSA.Notably,the HSA content of highly toxic Alexandrium minutum is expected to increase by 13.4%and 9.4%by 2050 and 2100,respectively.Under the high greenhouse gas emissions,there will be 17species migrating northward,6 species migrating eastward,and 4 species increasing in their size in HSA by 2050;moreover,there will be 16 species migrating northward,2 migrating eastward,and 4 species according to their size of HSA by 2100.Specifically,the HSA of A.minutum is predicted to increase by 7.0%and 25.9%by 2050 and 2100,respectively.Notably,A.ostenfeldii,which is currently seldom present in the China seas,is predicted to exhibit an HSA in most coastal areas of the Yellow Sea,the Bohai Sea,the Hangzhou Bay,the Zhejiang Coast,and the Beibu Gulf of the South China Sea.Conversely,the HSA of Noctiluca scintillans,a typical red-tide species,will be reduced by 7%–90%.The northward migration of Karenia mikimotoi exceeded 100 km and 300 km under low and high greenhouse gas emission scenarios,respectively.These changes underscore the significant impact of climate change on the distribution and habitat suitability of harmful dinoflagellates,thus indicating a potential shift in their ecological dynamics and consequent effects on marine ecosystems.展开更多
Unsupervised Domain Adaptation(UDA)intends to achieve excellent results by transferring knowledge from labeled source domains to unlabeled target domains in which the data or label distribution changes.Previous UDA me...Unsupervised Domain Adaptation(UDA)intends to achieve excellent results by transferring knowledge from labeled source domains to unlabeled target domains in which the data or label distribution changes.Previous UDA methods have acquired great success when labels in the source domain are pure.However,even the acquisition of scare clean labels in the source domain needs plenty of costs as well.In the presence of label noise in the source domain,the traditional UDA methods will be seriously degraded as they do not deal with the label noise.In this paper,we propose an approach named Robust Self-training with Label Refinement(RSLR)to address the above issue.RSLR adopts the self-training framework by maintaining a Labeling Network(LNet)on the source domain,which is used to provide confident pseudo-labels to target samples,and a Target-specific Network(TNet)trained by using the pseudo-labeled samples.To combat the effect of label noise,LNet progressively distinguishes and refines the mislabeled source samples.In combination with class rebalancing to combat the label distribution shift issue,RSLR achieves effective performance on extensive benchmark datasets.展开更多
Road traffic safety can decrease when drivers drive in a low-visibility environment.The application of visual perception technology to detect vehicles and pedestrians in infrared images proves to be an effective means...Road traffic safety can decrease when drivers drive in a low-visibility environment.The application of visual perception technology to detect vehicles and pedestrians in infrared images proves to be an effective means of reducing the risk of accidents.To tackle the challenges posed by the low recognition accuracy and the substan-tial computational burden associated with current infrared pedestrian-vehicle detection methods,an infrared pedestrian-vehicle detection method A proposal is presented,based on an enhanced version of You Only Look Once version 5(YOLOv5).First,A head specifically designed for detecting small targets has been integrated into the model to make full use of shallow feature information to enhance the accuracy in detecting small targets.Second,the Focal Generalized Intersection over Union(GIoU)is employed as an alternative to the original loss function to address issues related to target overlap and category imbalance.Third,the distribution shift convolution optimization feature extraction operator is used to alleviate the computational burden of the model without significantly compromising detection accuracy.The test results of the improved algorithm show that its average accuracy(mAP)reaches 90.1%.Specifically,the Giga Floating Point Operations Per second(GFLOPs)of the improved algorithm is only 9.1.In contrast,the improved algorithms outperformed the other algorithms on similar GFLOPs,such as YOLOv6n(11.9),YOLOv8n(8.7),YOLOv7t(13.2)and YOLOv5s(16.0).The mAPs that are 4.4%,3%,3.5%,and 1.7%greater than those of these algorithms show that the improved algorithm achieves higher accuracy in target detection tasks under similar computational resource overhead.On the other hand,compared with other algorithms such as YOLOv8l(91.1%),YOLOv6l(89.5%),YOLOv7(90.8%),and YOLOv3(90.1%),the improved algorithm needs only 5.5%,2.3%,8.6%,and 2.3%,respectively,of the GFLOPs.The improved algorithm has shown significant advancements in balancing accuracy and computational efficiency,making it promising for practical use in resource-limited scenarios.展开更多
Aiming at solving pressing parking issues in the urban environment, a residential parking spaces sharing model was proposed in this study. In this model, firstly, the residential community pattern, the status of idle ...Aiming at solving pressing parking issues in the urban environment, a residential parking spaces sharing model was proposed in this study. In this model, firstly, the residential community pattern, the status of idle parking spaces, and the temporal and spatial characteristics of sharing parking had been analyzed. Next,in the convenience of modeling,medical institutions that have the most prominent parking problems were selected as the subject of study. Based on the K-S statistical analysis results and the actual parking sharing situation,it was observed that the residential parking sharing time satisfied the shifted negative exponential distribution( SNED). Finally,a probability model of shared service capacity based on the SNED and critical time condition was established. By applying the statistical analysis method,the time of vehicles passing in and out of parking spaces, the idle time of parking spaces, the shifted distribution parameters, and other important model parameters had been calibrated,which was leading to the algorithm of model. In addition,considering the feasibility of model without sufficient data,the vehicle travel probability,the stagnation rate of parking space,and the status of parking spaces were defined and the reference data were also provided. The results of case studies indicate that it is very promising to solve urban parking issues if the residential community shares its rich parking resources with adjacent commercial buildings.展开更多
Dear Editor, The increasing frequencies and intensities of extreme heat resulting from climate change are likely to coerce distribution shifts of many species(Wernberg et al., 2024) and may increase extinction risk(Ma...Dear Editor, The increasing frequencies and intensities of extreme heat resulting from climate change are likely to coerce distribution shifts of many species(Wernberg et al., 2024) and may increase extinction risk(Malanoski et al., 2024). Elevated temperatures will raise the energy demands for thermoregulation in organisms, possibly destabilizing intracellular homeostasis and resulting in mortality(Scheffers et al., 2014).展开更多
This paper presents a methodology which determines the allocation of power demand among the committed generating units while minimizes number of objectives as well as meets physical and technological system constraint...This paper presents a methodology which determines the allocation of power demand among the committed generating units while minimizes number of objectives as well as meets physical and technological system constraints. The procedure considers two decoupled problems based upon the dependency of their goals on either active power or reactive power generation. Both the problems have been solved sequentially to achieve optimal allocation of active and reactive power generation while minimizes operating cost, gaseous pollutants emission objectives and active power transmission loss with consideration of system operating constraints along with generators prohibited operating zones and transmission line flow limits. The active and reactive power line flows are obtained with the help of generalized generation shift distribution factors （GGDF） and generalized Z-bus distribution factors （GZBDF）, respectively. First problem is solved in multi-objective framework in which the best weights assigned to objectives are determined while employing weighting method and in second problem, active power loss of the system is minimized subject to system constraints. The validity of the proposed method is demonstrated on 30-bus IEEE power system.展开更多
基金Supported by the National Natural Science Foundation of China(No.42176234)the Chinese Arctic and Antarctic Creative Program(No.JDXT2018-01)the Key Special Project for Introduced Talents Team of Southern Marine Science and Engineering Guangdong Laboratory(Guangzhou)(No.GML2019ZD0402)。
文摘Climate change can aff ect fi sh individuals or schools,and consequently the fi sheries.Studying future changes of fi sh distribution and abundance helps the scientifi c management of fi sheries.The dynamic bioclimate envelope model(DBEM)was used to identify the“environmental preference profi les”of the studied species based on outputs from three Earth system models(ESMs).Changes in ocean conditions in climate change scenarios could be transformed by the model into those in relative abundance and distribution of species.Therefore,the distributional response of 17 demersal fi shes to climate change in the Yellow Sea could be projected from 1970 to 2060.Indices of latitudinal centroid(LC)and mean temperature of relative abundance(MTRA)were used to represent the results conducted by model.Results present that 17 demersal fi sh species in the Yellow Sea show a trend of anti-poleward shift under both low-emission scenario(RCP 2.6)and high-emission scenario(RCP 8.5)from 1970 to 2060,with the projected average LC in three ESMs shifting at a rate of-1.17±4.55 and-2.76±3.82 km/decade,respectively,which is contrary to the previous projecting studies of fi shes suggesting that fi shes tend to move toward higher latitudes under increased temperature scenarios.The Yellow Sea Cold Water Mass could be the major driver resulting in the shift,which shows a potential signifi cance to fi shery resources management and marine conservation,and provides a new perspective in fi sh migration under climate change.
基金The National Key Research and Development Program of China under contract Nos 2019YFE0124700 and 2022YFC3106002。
文摘By establishing a distribution and environmental factor database of 21 typical harmful dinoflagellates in global waters,the MaxEnt model was used to predict shifts in the habitat of harmful dinoflagellates in Chinese waters under global climate change.The results revealed that offshore distance was the most important predictive factor and that surface seawater temperature(SST),primary productivity,and nitrate concentration were the key ecological factors influencing the distribution of harmful dinoflagellates.Under the low greenhouse gas emission scenario defined by the Intergovernmental Panel on Climate Change(IPCC),by approximately 2050,17 of the 21harmful dinoflagellate species in high-suitability areas(HSA)will migrate northward,six species will migrate eastward,and six species will expand their HSA.By 2100,approximately 18 of the 21 harmful dinoflagellate species in HSA will have migrated northward,seven species will have migrated eastward,and four species will have expanded their HSA.Notably,the HSA content of highly toxic Alexandrium minutum is expected to increase by 13.4%and 9.4%by 2050 and 2100,respectively.Under the high greenhouse gas emissions,there will be 17species migrating northward,6 species migrating eastward,and 4 species increasing in their size in HSA by 2050;moreover,there will be 16 species migrating northward,2 migrating eastward,and 4 species according to their size of HSA by 2100.Specifically,the HSA of A.minutum is predicted to increase by 7.0%and 25.9%by 2050 and 2100,respectively.Notably,A.ostenfeldii,which is currently seldom present in the China seas,is predicted to exhibit an HSA in most coastal areas of the Yellow Sea,the Bohai Sea,the Hangzhou Bay,the Zhejiang Coast,and the Beibu Gulf of the South China Sea.Conversely,the HSA of Noctiluca scintillans,a typical red-tide species,will be reduced by 7%–90%.The northward migration of Karenia mikimotoi exceeded 100 km and 300 km under low and high greenhouse gas emission scenarios,respectively.These changes underscore the significant impact of climate change on the distribution and habitat suitability of harmful dinoflagellates,thus indicating a potential shift in their ecological dynamics and consequent effects on marine ecosystems.
基金supported by the National Key R&D Program of China(2022ZD0114801)the National Natural Science Foundation of China(Grant No.61906089)the Jiangsu Province Basic Research Program(BK20190408).
文摘Unsupervised Domain Adaptation(UDA)intends to achieve excellent results by transferring knowledge from labeled source domains to unlabeled target domains in which the data or label distribution changes.Previous UDA methods have acquired great success when labels in the source domain are pure.However,even the acquisition of scare clean labels in the source domain needs plenty of costs as well.In the presence of label noise in the source domain,the traditional UDA methods will be seriously degraded as they do not deal with the label noise.In this paper,we propose an approach named Robust Self-training with Label Refinement(RSLR)to address the above issue.RSLR adopts the self-training framework by maintaining a Labeling Network(LNet)on the source domain,which is used to provide confident pseudo-labels to target samples,and a Target-specific Network(TNet)trained by using the pseudo-labeled samples.To combat the effect of label noise,LNet progressively distinguishes and refines the mislabeled source samples.In combination with class rebalancing to combat the label distribution shift issue,RSLR achieves effective performance on extensive benchmark datasets.
文摘Road traffic safety can decrease when drivers drive in a low-visibility environment.The application of visual perception technology to detect vehicles and pedestrians in infrared images proves to be an effective means of reducing the risk of accidents.To tackle the challenges posed by the low recognition accuracy and the substan-tial computational burden associated with current infrared pedestrian-vehicle detection methods,an infrared pedestrian-vehicle detection method A proposal is presented,based on an enhanced version of You Only Look Once version 5(YOLOv5).First,A head specifically designed for detecting small targets has been integrated into the model to make full use of shallow feature information to enhance the accuracy in detecting small targets.Second,the Focal Generalized Intersection over Union(GIoU)is employed as an alternative to the original loss function to address issues related to target overlap and category imbalance.Third,the distribution shift convolution optimization feature extraction operator is used to alleviate the computational burden of the model without significantly compromising detection accuracy.The test results of the improved algorithm show that its average accuracy(mAP)reaches 90.1%.Specifically,the Giga Floating Point Operations Per second(GFLOPs)of the improved algorithm is only 9.1.In contrast,the improved algorithms outperformed the other algorithms on similar GFLOPs,such as YOLOv6n(11.9),YOLOv8n(8.7),YOLOv7t(13.2)and YOLOv5s(16.0).The mAPs that are 4.4%,3%,3.5%,and 1.7%greater than those of these algorithms show that the improved algorithm achieves higher accuracy in target detection tasks under similar computational resource overhead.On the other hand,compared with other algorithms such as YOLOv8l(91.1%),YOLOv6l(89.5%),YOLOv7(90.8%),and YOLOv3(90.1%),the improved algorithm needs only 5.5%,2.3%,8.6%,and 2.3%,respectively,of the GFLOPs.The improved algorithm has shown significant advancements in balancing accuracy and computational efficiency,making it promising for practical use in resource-limited scenarios.
基金National High Technology Research and Development Plan Project,China(No.2014BAG03B03)National Natural Science Fundation,China(No.51378171)
文摘Aiming at solving pressing parking issues in the urban environment, a residential parking spaces sharing model was proposed in this study. In this model, firstly, the residential community pattern, the status of idle parking spaces, and the temporal and spatial characteristics of sharing parking had been analyzed. Next,in the convenience of modeling,medical institutions that have the most prominent parking problems were selected as the subject of study. Based on the K-S statistical analysis results and the actual parking sharing situation,it was observed that the residential parking sharing time satisfied the shifted negative exponential distribution( SNED). Finally,a probability model of shared service capacity based on the SNED and critical time condition was established. By applying the statistical analysis method,the time of vehicles passing in and out of parking spaces, the idle time of parking spaces, the shifted distribution parameters, and other important model parameters had been calibrated,which was leading to the algorithm of model. In addition,considering the feasibility of model without sufficient data,the vehicle travel probability,the stagnation rate of parking space,and the status of parking spaces were defined and the reference data were also provided. The results of case studies indicate that it is very promising to solve urban parking issues if the residential community shares its rich parking resources with adjacent commercial buildings.
基金jointly supported by Guangdong Natural Science Foundation (2022A1515010626)Key-Area Research and Development Program of Guangdong Province (2022B1111230001)the Forestry Science and Technology Innovation Project of Guangdong (2023KJCX023)。
文摘Dear Editor, The increasing frequencies and intensities of extreme heat resulting from climate change are likely to coerce distribution shifts of many species(Wernberg et al., 2024) and may increase extinction risk(Malanoski et al., 2024). Elevated temperatures will raise the energy demands for thermoregulation in organisms, possibly destabilizing intracellular homeostasis and resulting in mortality(Scheffers et al., 2014).
文摘This paper presents a methodology which determines the allocation of power demand among the committed generating units while minimizes number of objectives as well as meets physical and technological system constraints. The procedure considers two decoupled problems based upon the dependency of their goals on either active power or reactive power generation. Both the problems have been solved sequentially to achieve optimal allocation of active and reactive power generation while minimizes operating cost, gaseous pollutants emission objectives and active power transmission loss with consideration of system operating constraints along with generators prohibited operating zones and transmission line flow limits. The active and reactive power line flows are obtained with the help of generalized generation shift distribution factors （GGDF） and generalized Z-bus distribution factors （GZBDF）, respectively. First problem is solved in multi-objective framework in which the best weights assigned to objectives are determined while employing weighting method and in second problem, active power loss of the system is minimized subject to system constraints. The validity of the proposed method is demonstrated on 30-bus IEEE power system.
