Cyclist trajectory prediction is of great significance for both active collision avoidance and path planning of intelligent vehicles. This paper presents a trajectory prediction method for the motion intention of cyclists in real traffic scenarios. This method is based on dynamic Bayesian network (DBN) and long short-term memory (LSTM). The motion intention of cyclists is hard to predict owing to potential large uncertainties. The DBN is used to infer the distribution of cyclists' intentions at intersections to improve the prediction time. The LSTM with encoder-decoder is used to predict the cyclists' trajectories to improve the accuracy of prediction. Therefore, the DBN and LSTM are adopted to guarantee prediction accuracy and improve the prediction time. The experiment results are presented to show the effectiveness of the predict strategies.
Pedestrian–vehicle collision is an important component of traffic accidents. Over the past decades, it has become the focus of academic and industrial research and presents an important challenge. This study proposes a modified Driving Safety Field (DSF) model for pedestrian–vehicle risk assessment at an unsignalized road section, in which predicted positions are considered. A Dynamic Bayesian Network (DBN) model is employed for pedestrian intention inference, and a particle filtering model is conducted to simulate pedestrian motion. Driving data collection was conducted and pedestrian–vehicle scenarios were extracted. The effectiveness of the proposed model was evaluated by Monte Carlo simulations running 1000 times. Results show that the proposed risk assessment approach reduces braking times by 18.73%. Besides this, the average value of TTC−1 (the reciprocal of time-to-collision) and the maximum TTC−1 were decreased by 28.83% and 33.91%, respectively.
With the rapid development of urban rail transit, passenger traffic is increasing, and obstacle violations are more frequent, and the safety of train operation under high-density traffic conditions is becoming more and more thought-provoking. In order to monitor the train operating environment in real time, this paper first adopts multi-sensing technology based on machine vision and lidar, which is used to collect video images and ranging data of the track area in real time, and then it performs image preprocessing and division of regions of interest on the collected video. Then, the obstacles in the region of interest are detected to obtain the geometric characteristics and position information of the obstacles. Finally, according to the danger level of the obstacles, determine the degree of impact on train operation , the automatic response mode and manual response mode of the signal system are used to transmit the detection results to the corresponding train to control train operation. Through simulation analysis and experimental verification, the detection accuracy and control performance of the detection method are confirmed, which provides safety guarantee for the train operation.
With the development of the economy and the increase in passenger flow, the contradiction between urban rail transit demand and capacity is becoming more and more prominent. Increasing the number of on-line vehicles can ease this situation. Because of no increase in investment, the total number of subway vehicles is fixed. And the total vehicles include the maintenance vehicles and the on-line vehicles. Therefore, this article aims to optimize advanced maintenance cycle, so that the maintenance vehicles reduce and the on-line vehicles increase. First, the minimum value of the key components safe and reliable operating mileage is determined. Then, the Queuing Theory is used to obtain the optimized advanced maintenance cycle. Finally, reasonable maintenance plans are arranged based on the optimized maintenance cycle. The on-line vehicles increase by the optimized advanced maintenance cycle, which can relieve passenger flow pressure and meet urban rail transit demand. In addition, the reasonable maintenance plans can ensure that vehicles at the same level of reliability and within the specified mileage to complete the advanced maintenance and ensure the safe and reliable operation of vehicles.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.