ABSTRACT
In the last years, it is evident that cycling is becoming an alternative transportation mode to driving and has gained more popularity among all age groups, particularly in metropolitan cities, due to COVID-19. Although cycling is beneficial to individuals and urban cities (i.e., reduction of traffic congestion and promotion of a healthy lifestyle), it could also expose cyclists to risky situations, resulting in serious consequences. Therefore, this research aims at conducting a comprehensive analysis of the key contributory factors by using data derived from cycling accident and literature reports. More specifically, the accident data are first used to prioritize contributory factors contributing to a high level of cycling risk, and then the results guide the development of the literature review. The literature review analysis emphasized the characteristics, relationships, and control measures against different selected contributory factors identified from cycling accident reports. The in-depth analysis aids to figure out and better understand what the characteristics and relationships of these factors are, how they affect the safety of cyclists individually and jointly, and what to do to control their negative effects. The findings will not only provide practical insights for transport authorities to control contributory factors influencing cycling safety, but also engage more research for the improvement of cycling popularity, prevention of cycling risks, and enhancement of cycling safety in future. © 2023 American Society of Civil Engineers.
ABSTRACT
People's travel has changed greatly under the impact of COVID-19. However, it is controversial that whether traffic restrictions of COVID-19 have a positive or negative impact on traffic accidents. At present, there are few studies on the variations of traffic accidents under the impact of COVID-19 in China, and quantitative analysis is rare. Therefore, this study explores the traffic accidents characteristics of W city seriously affected COVID-19. Based on wavelet transform, traffic accident prediction model is established using property damage only accidents data to predict accident frequency without the impact of COVID-19. Compared with the actual traffic accidents frequency, this paper quantitatively analyzes the impact of COVID-19 on traffic accident. The results show that traffic accidents show a trend of decline-bottom-recovery;the frequency of accidents after the recovery is more than the previous year's level;compared with other periods in 2020, the proportion of injury accidents increased sharply during the period when traffic restrictions were gradually loose. The result of accident prediction shows that BP neural network has the best prediction effect. After the implementation of traffic restrictions, the frequency of accidents shows three stages: rapid decline, bottom and continuous rise. In the three stages, the frequency of property damage only accidents decreased by 379.06, 654.72 and 288.19 per day on average. © 2022 IEEE.
ABSTRACT
Speaking of the increase in traffic accidents, transportation safety plays an in important role. This study aims to analyze the number of traffic accidents in Makassar City. On the other hand, the implementation of Large-Scale Social Restrictions (LSSB) to prevent the spread of Covid-19 has an impact on transportation movements in Makassar City. According to this, the traffic accident data used is accident data for 2016 to 2020, which the data for 2020 is data on traffic accidents affected by the LSSB policy. From the results of data analysis, it is known that the accident rate trend tends to follow seasonal patterns so that the model used is the SARIMA Model. Sarima's model with a period (1,1,1) (1,1,1)6 is assumed to be the best model with a MAPE value of 81.6%. Based on the parameters generated by this model, it shows that the number of accidents in 2021 has decreased significantly. This is due to the government policy against the spread of the Covid-19 virus, which is implementing the PSBB. This model cannot be utilized in forecasting for a long period. This is because a long period can cause large estimates that fluctuate in value and even have a tendency for the accident rate to be negative.