Jiyuan Tan

The characteristics on commuting travel mode split and origin destination (OD) distribution in Shenzhen, China

With the rapid development of urbanization in China, the governments usually face the problem of traffic congestion, mainly due to the rapid increase in traffic demand. Group disparity between commuting mode choice and origin destination space distribution has enabled us to explore the cause of these differences and basic causes of traffic congestion. In recent years, commuting mode and behavior-based spatial analysis have sprung up. However, it is difficult to study the relationship between commuting mode choice and origin destination space distribution since the commuting characteristics data are invisible in a specific space. In this paper, the classical origin destination (OD) theory is usually employed to calculate data when studying commuting mode choice and we take Shenzhen as an example to study the daily travel mode of residents. Meanwhile, the traffic volume distribution of taxis, buses and subways in rush hours, which was analyzed as the basis for commuting travel split. Moreover, the OD data of taxi, bus and subway trips were used to explain the distribution of job-housing and traffic demand in different urban area. The results will be useful for optimizing the structure of the Shenzhen municipal traffic system in the final, which has important reference value.

The comparison of travel patterns between taxi and private car at Beijing Capital International Airport area

In last two years, the rapid development of Internet based ride-sharing has brought great changes to travel pattern of residents. However, few studies has been made to find out the unique travel patterns of Internet based ride-sharing. In this paper, License plate data at the Beijing Capital International Airport area are used to study the travel patterns of taxis and private cars. By comparing the trip records of taxis and private cars, two indicators, the distribution characteristics of vehicle volume and the balanced patterns of time, are selected to identify the ride-sharing cars from private cars. The results showed that there are more taxis than private cars arrival the airport in the evening and the arriving time of private cars is more concentrated. Finally, we used these indicators identify the ride-sharing vehicles from private cars. All these findings will be useful for further research and policy making.

Review and Implementation of Driving Fatigue Evaluation Methods Based on RR Interval

The evaluation methods of driving fatigue have been the hot topics in the study of traffic. This paper studied the main existing driving fatigue evaluation methods based on RR intervals and verified the effectiveness of these methods by experiments. There were 15 drivers selected in the simulation driving experiment, and the pulse signal of them was collected and recorded during the test period. Average heart rate, time and frequency domain features of HRV, nonlinear features of HRV, and statistic of MSPC were chosen as the indexes to verify. The experimental results show that “T2” of HRV time domain indexes, including MSPC, SDNN, RMSSD and total power, HRV frequency domain indexes like LF/HF, and the longer axis of Poincare scatter plot can distinguish the fatigue state and the waking state of drivers with a favorable effect.

Continuous and Discrete-Time Optimal Controls for an Isolated Signalized Intersection

A classical control problem for an isolated oversaturated intersection is revisited with a focus on the optimal control policy to minimize total delay. The difference and connection between existing continuous-time planning models and recently proposed discrete-time planning models are studied. A gradient descent algorithm is proposed to convert the optimal control plan of the continuous-time model to the plan of the discrete-time model in many cases. Analytic proof and numerical tests for the algorithm are also presented. The findings shed light on the links between two kinds of models.

Driver's mental workload estimation based on empirical physiological indicators

In order to explore drivers mental workload under different traffic situations, the field experiment aimed at giving quantitative assessments of drivers mental workload within different maneuvers by physiological indicators, such as ECG and EDA. Moreover, the driving safety assessments in different mental workload were analyzed. Firstly, the HRV was extracted from the ECG signal de-noised and filtered. And then the AVHR, SDNN, RMSSD, PNN50 and LF/HF of HRV in time-frequency were analyzed respectively. Consequently, the results showed that the drivers AVHR, SDNN, RMSSD, PNN50 and LF/HF of HRV increased obviously in left-turning maneuver when compared with those in straight-going maneuver, and there was no significant physiological difference between those in right-turning maneuver and straight-going maneuver. The inference was drivers mental workload was the highest when taking a left-turn. Furthermore, the physiological index indicated the mental workload was higher in night driving than that in daytime driving. Conclusion is helpful for the development of driving assistance based on physiological information detection.

Driving Physiological Indicator Recognition Based on Smart Bracelet Data

Various research on heart rate variability (HRV) have focused on the characteristics of the peaks and the valleys of a heart rate graph and most have been analyzed using medical devices. This paper developed a driving physiological indicator extraction method. The method is based on a pulse signal detection with a feature point identification model and an abnormal signal detection combined with a revamp model. According to physiological theory and the relationship between pulse and ECG readings, the pulse feature points can be detected. This method is easier and more accurate than the existing wavelet pulse signal detection method, which can be used in the wearable devices, such as the smart band and the smart watch. The method was also validated efficiently and accurately in the experiments.

Distribution Models for Effective Flow Rate at Signalized Intersections in Beijing

Lots of research has been carried out to study lost time and departure headway of certain phase duration in urban intersections, due to its importance to signal timing. Most previous research focused on the situation of static queuing vehicles or constant vehicle arrival rate. However, the departure process is much more complex than an ideal scenario. In this paper, the authors propose a distribution model for effective flow rate at signalized intersections, including both the saturated situation and unsaturated situation. Finally, the Monte-Carlo simulation shows that the model is simple and efficient in different situations.

Modeling of Stop or Run Behaviors at Urban Signalized Intersections in Beijing

In China, over the last decade, Red Light Running (RLR) has been one of the most important reasons for the accidents at signalized intersections. Although several methods such as dilemma zone protection and signal violence enforcement have been used to reduce the probability of RLR behaviors, how the drivers make decisions when the signal changed from green to red is seldom discussed. Based on the video data collected in Beijing, a discrete choice model is proposed in this paper to analyze drivers’ behaviors. Further, the authors will analyze cases under different traffic situations. And finally, how to improve road safety and efficiency will be discussed.