The correct identification of traffic signs plays an important role in automatic driving technology and road safety driving.Therefore,to address the problems of misdetection and omission in traffic sign detection due ...The correct identification of traffic signs plays an important role in automatic driving technology and road safety driving.Therefore,to address the problems of misdetection and omission in traffic sign detection due to the variety of sign types,significant size differences and complex background information,an improved traffic sign detection model for RT-DETR was proposed in this study.Firstly,the HiLo attention mechanism was added to the Attention-based Intra-scale Feature Interaction,which further enhanced the feature extraction capability of the network and improved the detection efficiency on high-resolution images.Secondly,the CAFMFusion feature fusion mechanism was designed,which enabled the network to pay attention to the features in different regions in each channel.Based on this,the model could better capture the remote dependencies and neighborhood feature correlation,improving the feature fusion capability of the model.Finally,the MPDIoU was used as the loss function of the improved model to achieve faster convergence and more accurate regression results.The experimental results on the TT100k-2021 traffic sign dataset showed that the improved model achieves the performance with a precision value of 90.2%,recall value of 88.1%and mAP@0.5 value of 91.6%,which are 4.6%,5.8%,and 4.4%better than the original RT-DETR model respectively.The model effectively improves the problem of poor traffic sign detection and has greater practical value.展开更多
With the rapid development of urban rail transit,there have been an urgent problem of excessive stray current.Because the stray current distribution is random and difficult to verify in the field,we designed an improv...With the rapid development of urban rail transit,there have been an urgent problem of excessive stray current.Because the stray current distribution is random and difficult to verify in the field,we designed an improved stray current experimental platform by replacing the simulated aqueous solution with a real soil environment and by calculating the transition resistance by measuring the soil resistivity,which makes up for the defects in the previous references.Firstly,the mathematical models of rail-drainage net and rail-drainage netground were established,and the analytical expressions of current and voltage of rail,drainage net and other structures were derived.In addition,the simulation model was built,and the mathematical analysis results were compared with the simulation results.Secondly,the accuracy of the improved stray current experimental platform was verified by comparing the measured and simulation results.Finally,based on the experimental results,the influence factors of stray current were analyzed.The relevant conclusions provide experimental data and theoretical reference for the study of stray current in urban rail transit.展开更多
Aiming at prevalent violations of non-motorists at urban intersections in China, this paper intends to clarify the characteristics and risks of non-motorist violations at signalized intersections through questionnaire...Aiming at prevalent violations of non-motorists at urban intersections in China, this paper intends to clarify the characteristics and risks of non-motorist violations at signalized intersections through questionnaires and video recordings, which may serve as a basis for non-motorized vehicle management. It can help improve the traffic order and enhance the degree of safety at signalized intersections. To obtain the perception information, a questionaire survey on the Internet was conducted and 972 valid questionnaires were returned. It is found that academic degree contributes little to non-motorist violations, while electrical bicyclists have a relatively higher frequency of violations compared with bicyclists. The video data of 18 228 non-motorist behaviors indicate that the violation rate of all non-motorists is 26.5%; the number of conflicts reaches 1 938, among which violation conflicts account for 66.8%. The study shows that the violation rates and the violation behavior at three types of surveyed intersections are markedly different. It is also concluded that the conflict rates and the violation rates are positively correlated. Furthermore, signal violation, traveling in the wrong direction, and overspeeding to cross the intersection are the most dangerous among traffic violation behaviors.展开更多
In order to avoid the noise and over fitting and further improve the limited classification performance of the real decision tree, a traffic incident detection method based on the random forest algorithm is presented....In order to avoid the noise and over fitting and further improve the limited classification performance of the real decision tree, a traffic incident detection method based on the random forest algorithm is presented. From the perspective of classification strength and correlation, three experiments are performed to investigate the potential application of random forest to traffic incident detection： comparison with a different number of decision trees; comparison with different decision trees; comparison with the neural network. The real traffic data of the 1-880 database is used in the experiments. The detection performance is evaluated by the common criteria including the detection rate, the false alarm rate, the mean time to detection, the classification rate and the area under the curve of the receiver operating characteristic （ROC）. The experimental results indicate that the model based on random forest can improve the decision rate, reduce the testing time, and obtain a higher classification rate. Meanwhile, it is competitive compared with multi-layer feed forward neural networks （MLF）.展开更多
文摘The correct identification of traffic signs plays an important role in automatic driving technology and road safety driving.Therefore,to address the problems of misdetection and omission in traffic sign detection due to the variety of sign types,significant size differences and complex background information,an improved traffic sign detection model for RT-DETR was proposed in this study.Firstly,the HiLo attention mechanism was added to the Attention-based Intra-scale Feature Interaction,which further enhanced the feature extraction capability of the network and improved the detection efficiency on high-resolution images.Secondly,the CAFMFusion feature fusion mechanism was designed,which enabled the network to pay attention to the features in different regions in each channel.Based on this,the model could better capture the remote dependencies and neighborhood feature correlation,improving the feature fusion capability of the model.Finally,the MPDIoU was used as the loss function of the improved model to achieve faster convergence and more accurate regression results.The experimental results on the TT100k-2021 traffic sign dataset showed that the improved model achieves the performance with a precision value of 90.2%,recall value of 88.1%and mAP@0.5 value of 91.6%,which are 4.6%,5.8%,and 4.4%better than the original RT-DETR model respectively.The model effectively improves the problem of poor traffic sign detection and has greater practical value.
基金supported by National Natural Science Foundation of China(Nos.51476073,51266004)Natural Science Foundation of Gansu Province(No.138RJZA199).
文摘With the rapid development of urban rail transit,there have been an urgent problem of excessive stray current.Because the stray current distribution is random and difficult to verify in the field,we designed an improved stray current experimental platform by replacing the simulated aqueous solution with a real soil environment and by calculating the transition resistance by measuring the soil resistivity,which makes up for the defects in the previous references.Firstly,the mathematical models of rail-drainage net and rail-drainage netground were established,and the analytical expressions of current and voltage of rail,drainage net and other structures were derived.In addition,the simulation model was built,and the mathematical analysis results were compared with the simulation results.Secondly,the accuracy of the improved stray current experimental platform was verified by comparing the measured and simulation results.Finally,based on the experimental results,the influence factors of stray current were analyzed.The relevant conclusions provide experimental data and theoretical reference for the study of stray current in urban rail transit.
基金The National Key Technology R&D Program during the 11th Five-Year Plan Period(No.2009BAG13A05)the National Natural Science Foundation of China(No.51078086)
文摘Aiming at prevalent violations of non-motorists at urban intersections in China, this paper intends to clarify the characteristics and risks of non-motorist violations at signalized intersections through questionnaires and video recordings, which may serve as a basis for non-motorized vehicle management. It can help improve the traffic order and enhance the degree of safety at signalized intersections. To obtain the perception information, a questionaire survey on the Internet was conducted and 972 valid questionnaires were returned. It is found that academic degree contributes little to non-motorist violations, while electrical bicyclists have a relatively higher frequency of violations compared with bicyclists. The video data of 18 228 non-motorist behaviors indicate that the violation rate of all non-motorists is 26.5%; the number of conflicts reaches 1 938, among which violation conflicts account for 66.8%. The study shows that the violation rates and the violation behavior at three types of surveyed intersections are markedly different. It is also concluded that the conflict rates and the violation rates are positively correlated. Furthermore, signal violation, traveling in the wrong direction, and overspeeding to cross the intersection are the most dangerous among traffic violation behaviors.
基金The National High Technology Research and Development Program of China(863 Program)(No.2012AA112304)the Scientific Innovation Research of College Graduates in Jiangsu Province(No.CXZZ13-0119)
文摘In order to avoid the noise and over fitting and further improve the limited classification performance of the real decision tree, a traffic incident detection method based on the random forest algorithm is presented. From the perspective of classification strength and correlation, three experiments are performed to investigate the potential application of random forest to traffic incident detection： comparison with a different number of decision trees; comparison with different decision trees; comparison with the neural network. The real traffic data of the 1-880 database is used in the experiments. The detection performance is evaluated by the common criteria including the detection rate, the false alarm rate, the mean time to detection, the classification rate and the area under the curve of the receiver operating characteristic （ROC）. The experimental results indicate that the model based on random forest can improve the decision rate, reduce the testing time, and obtain a higher classification rate. Meanwhile, it is competitive compared with multi-layer feed forward neural networks （MLF）.
