This paper presents a new methodology to calculate unblocked reliability of a road network. Unblocked reliability analysis is used to assess the operation performance of a road network at peak hour in order to support the decision-making involved in road network planning. A four-level model of road network unblocked reliability (link, path, Origin-destination pair and whole network) is shown to assess the operation performance of each level of the road network. A method of sensitivity analysis is provided to determine the priority of the road section in the planning scheme. However, the most difficult task in unblocked reliability analysis is how to find a route choice pattern, namely, which routes are selected and how proportions of Origin-destination flows are loaded on these used routes between every Origin-destination pair. In order to resolve this problem, a new traffic assignment method is proposed in this paper based on the Frank-Wolf algorithm. This algorithm not only can find the link flows but can also solve the route choice pattern, which satisfy the User Equilibrium criterion. The proposed methodology is applied to solve the goal of traffic demand control and the necessary road capacity when a road network is improved to the expected operation performance. The proposed methodology provides a useful tool to assess the operation performance of a road network at several levels of road capacity and traffic demand, and can be used to evaluate various road network planning and to design a reliable road network system.

View full abstract

In this paper, an integrated driving environment recognition system is established on THASV-II (Tsinghua Advanced Safety Vehicle) test platform. The system collects environment information and vehicular motion information based on the fusion of machine vision, laser radar and other vehicular sensors. An adaptive bi-threshold method and steerable filters are adopted in the image pre-processing to enhance the lane marker and vehicle features. A multi-level fusion structure is designed to recognize the front vehicles and a high-order statistic method is proposed to recognize the side and rear vehicles. The anti-interference ability of lane marker recognition is improved by a point set optimization method and the target vehicle is determined through a multi-feature fusion algorithm. Field tests demonstrate that the integrated system can meet the demands of environment recognition for various lateral and longitudinal active safety systems, achieving good adaptability to environmental variation and good anti-interference ability under heavy traffic situation.

View full abstract

To modify aerodynamic characteristics of pantographs for high-speed trains, we propose to blow air from the surface of a panhead. The outlets for air blowing are bored in rows near the trailing edge of the panhead. To estimate the effect of the blowing, we first conduct a wind tunnel test using a two-dimensional panhead model. The results reveal that the pressure distribution on the model can be changed by the blowing. Then we examine a full-scale pantograph model by a wind tunnel test to prove that the proposed method has a great potential for practical applications.

View full abstract

The aerodynamic characteristics of pantographs for high-speed trains are a critical factor towards steady current collection. We introduce a new panhead in which air is taken in at the leading edge and ejected around the trailing edge to control lift force. The aerodynamic characteristics of this panhead are estimated using a numerical simulation and a wind tunnel experiment, with the results showing that the system is effective in controlling lift force.

View full abstract

Vibration of high-speed trains increases in tunnels caused by aerodynamic force whose mechanism is unknown. To investigate this aerodynamic force, the authors conduct an analysis of the running test data and a numerical simulation. The running test data indicates that the aerodynamic force acts to vibrate the train in the tunnel and it originates from large-scale coherent structures in the space between the tunnel wall and the train. These flow structures develop from the head toward the 6th to 8th cars, and become steady thereafter to the tail of the train set. The computation reveals that the flow becomes unstable under the train. The resulted vortices are spread on the train side by the tunnel wall, and then the unsteady aerodynamic force arises when the vortices pass.

View full abstract

A traction slip ratio control method using fuzzy dynamical sliding mode strategy (Fuzzy DSMC) is proposed to reduce the slip ratio oscillations in the independent AWD EV traction control. The slip ratios are also accurately estimated in a new way to support this control process. Firstly in this control method, the fuzzy logic method is applied respectively to regulate the switching surface and the reaching law of DSMC with the estimated slip ratios, which are used to weaken the chattering and improve the convergence rate to some extent. Furthermore the control structure of DSMC is designed to obtain the smooth torque outputs from all independent traction motors, which are implemented in the anti-skid control for EV in the end. The mathematics analysis for the controller parameters choosing and simulation experiments show that the method can greatly avoid the drawback of control chattering occurred in the classical sliding mode control. Moreover, the robustness of systems for parameter uncertainties is also guaranteed.

View full abstract

A direct yaw-moment control (DYC) method based on the dynamic regulation of the control target is proposed to achieve integrated optimization between maneuverability and stability for the independent 4WD EV. Firstly, the yaw rate responses are calculated from the modified bicycle model, which respectively represent the maneuverability and stability of EV. With these responses deduced, the integrated control target for the maneuverability and stability is determined for all steering situations. Furthermore based on the “feedforward+feedback” control structure, the DYC controller is designed which combines the dynamical sliding mode control (DSMC) and LQ control. DSMC avoids the drawbacks of the oscillations by chattering happening in the classical SMC and allows the smoothness of the direct yaw-moment. The simulation experiments show that this DYC system can restrain the side slip angle effectively and keep higher yaw rate, which guarantees the EV maneuverability and stability. Moreover, the robustness of systems for road adhesion conditions variation and vehicle parameters uncertainties is also guaranteed in simulation validation.

View full abstract

A traffic accident prediction method using a priori knowledge based on accident data is proposed for safe driving support. Implementation is achieved by an algorithm using particle filtering and fuzzy inference to estimate accident risk factors. With this method, the distance between the host vehicle and a vehicle ahead and their relative velocity and relative acceleration are obtained from the results of particle filtering of driving data and are used as attributes to build the relative driving state space. The attributes are evaluated as likelihoods and then consolidated as a risk level using fuzzy inference. Experimental validation was done using videos of general driving situations obtained with an on-vehicle CCD camera and one simulated accident situation created based on the video data. The results show that high risk levels were calculated with the proposed method in the early stages of the accident situations.

View full abstract

Determining appropriate alarm timing for Forward Collision Warning Systems (FCWS) may play an important role in enhancing system acceptance by drivers. It is not always true that a common alarm trigger logic is suitable for all drivers, because presented alarms may be differently viewed for each driver, i.e., paying attention or requiring appropriate actions. The current study focused on adaptive alarm timing which was adjusted in response to braking behaviour for collision avoidance for the individual. In Experiment I, the braking performance of individual driver was measured repeatedly to assess the variation of each performance. We utilised the following two indices: elapsed time from the deceleration of the lead car to release of the accelerator (accelerator release time) and elapsed time to application of the brakes (braking response time). Two alarm timings were then determined based on these two indices: (i) the median of the accelerator release time of the driver and (ii) the median of the braking response time of the driver. Experiment II compared the two alarm timings for each driver in order to investigate which timing is more appropriate for enhancing driver trust in the driver-adaptive FCWS and the system effectiveness. The results showed that the timing of the accelerator release time increased the trust ratings more than the timing of braking response. The timing of the braking response time induced a longer response time to application of the brakes. Moreover, the degree to which the response time was longer depended on alarm timing preference of the driver. The possible benefit and drawback of driver-adaptive alarm timing are discussed.

View full abstract