Predicting short-term traffic crashes is chal-lenging due to an imbalanced data set characterized by excessive zeros in noncrash counts,random crash occur-rences,spatiotemporal correlation in crash counts,and inherent...Predicting short-term traffic crashes is chal-lenging due to an imbalanced data set characterized by excessive zeros in noncrash counts,random crash occur-rences,spatiotemporal correlation in crash counts,and inherent heterogeneity.Existing models struggle to effec-tively address these distinct characteristics in crash data.This paper proposes a new joint model by combining the time-series generalized regression neural network(TGRNN)model and the binomially weighted convolutional neural network(BWCNN)model.The joint model aims to capture all these characteristics in short-term crash predic-tion.The model was trained and tested using real-world,highly disaggregated traffic data collected with inductive loop detectors on the M1 motorway in the UK in 2019,along with crash data extracted from the UK National Accident Database for the same year.The short-term is defined as a 30-min interval,providing sufficient time for a traffic control center to implement interventions and mitigate potential hazards.The year was segmented into 30-min intervals,resulting in a highly imbalanced data set with over 99.99%noncrash samples.The joint model was applied to predict the probability of a crash occurrence by updating both the crash and traffic data every 30 min.The findings revealed that 75.3%of crashes and 81.6%of noncrash events were correctly predicted in the southbound direction.In the northbound direction,78.1%of crashes and 80.2%of noncrash events were accurately captured.Causal analysis and model-based interpretation were used to analyze the relative importance of explanatory variables regarding their contribution to crashes.The results reveal that speed variance and speed are the most influential factors contributing to crash occurrence.展开更多
Accurate real-time traffic crash prediction is crucial for proactive traffic safety manage-ment.Currently,the majority of real-time models forecast crashes every 5 min to support different intelligent transportation s...Accurate real-time traffic crash prediction is crucial for proactive traffic safety manage-ment.Currently,the majority of real-time models forecast crashes every 5 min to support different intelligent transportation systems.However,these intervals might be too short for practical use in manually implementing proactive traffic safety measures such as deploying traffic law enforcement and emergency rescue resources.Therefore,this study develops hourly crash prediction models to provide network operators with sufficient time to take measures in advance.A section of a mountainous freeway in Guizhou province is divided into homogeneous segments,with crash data,traffic operations data,and meteo-rological data being collected hourly.As the result is an imbalanced dataset of crash and non-crash instances,the training dataset is resampled using synthetic minority over-sampling technique(SMOTE)to address the issue.To fully capture the complex spatiotem-poral relationships in the data and achieve high crash prediction accuracy,a graph convo-lutional network-long short-term memory(GCN-LSTM)model is constructed for the first time,combining a graph convolutional network(GCN)and long short-term memory(LSTM)neural network.For comparison purposes,LSTM,extreme gradient boosting(XGBoost),and logistic regression(LR)models are developed.The results show that the GCN-LSTM model outperforms other models in hourly traffic crash prediction,and the optimal prediction performance is achieved with the crash-to-non-crash ratio of 1:4.The GCN-LSTM method is found to effectively capture the complex spatiotemporal relation-ships in prediction data and to handle imbalanced traffic crash data.展开更多
Traffic crashes in Riyadh city cause losses in the form of deaths,injuries and property damages,in addition to the pain and social tragedy affecting families of the victims.In 2005,there were a total of 47,341 injury ...Traffic crashes in Riyadh city cause losses in the form of deaths,injuries and property damages,in addition to the pain and social tragedy affecting families of the victims.In 2005,there were a total of 47,341 injury traffic crashes occurred in Riyadh city(19%of the total KSA crashes)and 9%of those crashes were severe.Road safety in Riyadh city may have been adversely affected by:high car ownership,migration of people to Riyadh city,high daily trips reached about 6 million,high rate of income,low-cost of petrol,drivers from different nationalities,young drivers and tremendous growth in population which creates a high level of mobility and transport activities in the city.The primary objective of this paper is therefore to explore factors affecting the severity and frequency of road crashes in Riyadh city using appropriate statistical models aiming to establish effective safety policies ready to be implemented to reduce the severity and frequency of road crashes in Riyadh city.Crash data for Riyadh city were collected from the Higher Commission for the Development of Riyadh(HCDR)for a period of five years from 1425H to 1429H(roughly corresponding to 2004-2008).Crash data were classified into three categories:fatal,serious-injury and slight-injury.Two nominal response models have been developed:a standard multinomial logit model(MNL)and a mixed logit model to injury-related crash data.Due to a severe underreporting problem on the slight injury crashes binary and mixed binary logistic regression models were also estimated for two categories of severity:fatal and serious crashes.For frequency,two count models such as Negative Binomial(NB)models were employed and the unit of analysis was 168 HAIs(wards)in Riyadh city.Ward-level crash data are disaggregated by severity of the crash(such as fatal and serious injury crashes).The results from both multinomial and binary response models are found to be fairly consistent but the results from the random parameters model seem more reasonable.Age and nationality of the driver,excessive speed,wet road surface and dark lighting conditions and single vehicle crashes are associated with increased probability of fatal crashes.More specifically,the probability of having a fatal crash increases with the age of the driver and Saudi drivers(relative to non-Saudi drivers)are associated with the probability of fatal crashes(relative to serious injury crashes).A crash involving a single vehicle is found to be more severe than a crash involving a multiple vehicles.The results from the frequency models suggest that percentage of non-Saudi found positively associated with serious injury crashes;percentage of illiterate people and the income per capita found to be positively significant with the frequency of fatal and serious injury crashes;and the increased residential,transport,and educational areas of land use is associated with the decreased level of fatal and serious injury crashes occurrences.Based on the findings,a range of countermeasures are proposed to reduce the severity and frequency of traffic crashes in Riyadh city.展开更多
文摘Predicting short-term traffic crashes is chal-lenging due to an imbalanced data set characterized by excessive zeros in noncrash counts,random crash occur-rences,spatiotemporal correlation in crash counts,and inherent heterogeneity.Existing models struggle to effec-tively address these distinct characteristics in crash data.This paper proposes a new joint model by combining the time-series generalized regression neural network(TGRNN)model and the binomially weighted convolutional neural network(BWCNN)model.The joint model aims to capture all these characteristics in short-term crash predic-tion.The model was trained and tested using real-world,highly disaggregated traffic data collected with inductive loop detectors on the M1 motorway in the UK in 2019,along with crash data extracted from the UK National Accident Database for the same year.The short-term is defined as a 30-min interval,providing sufficient time for a traffic control center to implement interventions and mitigate potential hazards.The year was segmented into 30-min intervals,resulting in a highly imbalanced data set with over 99.99%noncrash samples.The joint model was applied to predict the probability of a crash occurrence by updating both the crash and traffic data every 30 min.The findings revealed that 75.3%of crashes and 81.6%of noncrash events were correctly predicted in the southbound direction.In the northbound direction,78.1%of crashes and 80.2%of noncrash events were accurately captured.Causal analysis and model-based interpretation were used to analyze the relative importance of explanatory variables regarding their contribution to crashes.The results reveal that speed variance and speed are the most influential factors contributing to crash occurrence.
文摘Accurate real-time traffic crash prediction is crucial for proactive traffic safety manage-ment.Currently,the majority of real-time models forecast crashes every 5 min to support different intelligent transportation systems.However,these intervals might be too short for practical use in manually implementing proactive traffic safety measures such as deploying traffic law enforcement and emergency rescue resources.Therefore,this study develops hourly crash prediction models to provide network operators with sufficient time to take measures in advance.A section of a mountainous freeway in Guizhou province is divided into homogeneous segments,with crash data,traffic operations data,and meteo-rological data being collected hourly.As the result is an imbalanced dataset of crash and non-crash instances,the training dataset is resampled using synthetic minority over-sampling technique(SMOTE)to address the issue.To fully capture the complex spatiotem-poral relationships in the data and achieve high crash prediction accuracy,a graph convo-lutional network-long short-term memory(GCN-LSTM)model is constructed for the first time,combining a graph convolutional network(GCN)and long short-term memory(LSTM)neural network.For comparison purposes,LSTM,extreme gradient boosting(XGBoost),and logistic regression(LR)models are developed.The results show that the GCN-LSTM model outperforms other models in hourly traffic crash prediction,and the optimal prediction performance is achieved with the crash-to-non-crash ratio of 1:4.The GCN-LSTM method is found to effectively capture the complex spatiotemporal relation-ships in prediction data and to handle imbalanced traffic crash data.
文摘Traffic crashes in Riyadh city cause losses in the form of deaths,injuries and property damages,in addition to the pain and social tragedy affecting families of the victims.In 2005,there were a total of 47,341 injury traffic crashes occurred in Riyadh city(19%of the total KSA crashes)and 9%of those crashes were severe.Road safety in Riyadh city may have been adversely affected by:high car ownership,migration of people to Riyadh city,high daily trips reached about 6 million,high rate of income,low-cost of petrol,drivers from different nationalities,young drivers and tremendous growth in population which creates a high level of mobility and transport activities in the city.The primary objective of this paper is therefore to explore factors affecting the severity and frequency of road crashes in Riyadh city using appropriate statistical models aiming to establish effective safety policies ready to be implemented to reduce the severity and frequency of road crashes in Riyadh city.Crash data for Riyadh city were collected from the Higher Commission for the Development of Riyadh(HCDR)for a period of five years from 1425H to 1429H(roughly corresponding to 2004-2008).Crash data were classified into three categories:fatal,serious-injury and slight-injury.Two nominal response models have been developed:a standard multinomial logit model(MNL)and a mixed logit model to injury-related crash data.Due to a severe underreporting problem on the slight injury crashes binary and mixed binary logistic regression models were also estimated for two categories of severity:fatal and serious crashes.For frequency,two count models such as Negative Binomial(NB)models were employed and the unit of analysis was 168 HAIs(wards)in Riyadh city.Ward-level crash data are disaggregated by severity of the crash(such as fatal and serious injury crashes).The results from both multinomial and binary response models are found to be fairly consistent but the results from the random parameters model seem more reasonable.Age and nationality of the driver,excessive speed,wet road surface and dark lighting conditions and single vehicle crashes are associated with increased probability of fatal crashes.More specifically,the probability of having a fatal crash increases with the age of the driver and Saudi drivers(relative to non-Saudi drivers)are associated with the probability of fatal crashes(relative to serious injury crashes).A crash involving a single vehicle is found to be more severe than a crash involving a multiple vehicles.The results from the frequency models suggest that percentage of non-Saudi found positively associated with serious injury crashes;percentage of illiterate people and the income per capita found to be positively significant with the frequency of fatal and serious injury crashes;and the increased residential,transport,and educational areas of land use is associated with the decreased level of fatal and serious injury crashes occurrences.Based on the findings,a range of countermeasures are proposed to reduce the severity and frequency of traffic crashes in Riyadh city.
