Katsuya Tanifuji

A Study of Driver Behavior Inference Model at Time of Lane Change using Bayesian Networks

Recent years have brought hope that driving support systems tailored to the characteristics of each driver can be developed. To accomplish this, a driver model must be constructed that considers the drivers psychological function when inferring driver behavior. This paper thus proposes a method to infer driver behavior by capturing time-series steering angle data at the time of lane change. The proposed method uses a static type conditional Gaussian model on Bayesian networks. By using this method, if the driver behavior of the subject and learned data nearness of features (norms) are below a certain level, it is possible to infer driver behavior with nearly 100% probability. Moreover, compared to the HMM models, this method reduces the rate of incorrect inference inclusion.

Active Steering of a Rail Vehicle with Two-Axle Bogies Based on Wheelset Motion

Summary This paper deals with a basic study on the actively steered rail vehicle with two-axle bogie trucks, which employs a control law based on the self-steering ability of wheelset. However, this control law not only helps the steering performance, but also tends to lower the running stability. Two methods are proposed to improve the stability. The first method is to add a feedback of wheelset lateral velocity, the second one is to give some time lag to the control force, and both are proved to be affective. Here, the latter is more practical method and the delay of control due to the time lag does not deteriorate the steering performance.

Effect of Crosswind on Derailment of Railway Vehicles Running on Curved Track at Low Speed

Owing to the lightening of railway vehicles and increased operation speeds, the reduction of running safety in the presence of crosswind is becoming an important problem. In particular, the running safety tends to decrease when vehicles run on curved track. When a crosswind acts on a vehicle negotiating a curve from the outer side, flange climbing can occur. In this study, a full-vehicle model was constructed using the multi-body simulation software SIMPACK, and a simulation of a bogie vehicle with two-axle trucks negotiating a curve was carried out to examine the running safety under the condition where a crosswind acts on the vehicle from the outer side of the curve. As a result, it was verified that the derailment coefficient of the first wheelset becomes large in the exit transition curve and the coefficient of the third wheelset does in the entrance transition curve, and this trend becomes pronounced at low operation speeds in the presence of a stronger crosswind. It was also shown that the critical derailment coefficients obtained by modified Nadal’s formula considering the effect of attack angle become close to the actual derailment coefficients at the timing that flange climbing occurs.

Modelling of aerodynamic force acting on high speed train in tunnel and a measure to improve the riding comfort utilising restriction between cars

For a train running at high speeds in a tunnel, the car body vibration considerably increases due to the aerodynamic force generated around the car body, thereby deteriorating riding comfort. To improve riding comfort of a train running at high speeds in a tunnel, the effects of stiffness and damping between adjoining cars are examined. In case the restriction between cars is too large, the curving performance may deteriorate around sharp curves. In this paper, analytical models of the aerodynamic force and a train are developed, and the riding comfort at high speed in a tunnel and the curving performance around a sharp curve are simulated. As a result, a solution, which improves the riding comfort at high speeds without increasing the wheel lateral force around sharp curves, is found by increasing inter-vehicle damping in the longitudinal and lateral directions.

Mechatronics in Japanese rail vehicles: active and semi-active suspensions

Abstract This paper introduces the present state of mechatronics application to the railway vehicle in Japan. The objective mechanisms in the vehicle are classified into five categories such as drive and braking, car body tilting, steering, pantograph and suspension. First, the outline of research and development in each category is described briefly. Next, the situation in the category of suspension is explained mainly on active control system and semi-active one. Finally, the prospect of research studies relating to high-speed curve negotiation is described.

A Prediction of Wheel/Rail Lateral Force Induced by Actively Controlled Suspension for High Speed Railway Vehicles

Abstract Instead of good effects on the ride comfort, the active suspension may have unfavourable effects on the wheel/rail interaction such as the increase of wheel lateral force. This paper deals with an analytical study about the lateral dynamics of two types of railway vehicle, in which an active suspension mechanism controls the running vibration. One vehicle type has trucks with bolster and side bearers, and the other type does bolster-less trucks with yaw dampers. The analytical models for the vehicles consider non-linear characteristics such as creep forces and wheel/rail side collision. In this study the active suspension has a controller based on the optimal control theory and assumes ideal actuators to be available. Their operations on a track with alignment irregularities are simulated up to the speed of 350 km/h. The results show that the active suspension can improve the ride quality without any increase of wheel lateral force, and that the bolster-less truck is favourable for active suspens...

A VERTICAL VIBRATION ANALYSIS OF COUPLED BOGIE CARS IN A TRAIN FOR EVALUATION OF RIDING COMFORT

SUMMARY In the recent trend of speed-up on railways, it appears that the coupling of cars as a train has an important effect on the car vibration. The effect is remarkable at each coupled car on the Shinkansen, in the speed range over 210 km/h. Especially, the vibration on the rear side of the last car grows large and its riding comfort is harmed with the speed-up of the train operation. This paper describes analysis of the vertical vibration and its effect on the riding comfort of a train composed of several bogie cars. In this study, a calculation model of 5 cars coupled in a train is used. A simplified restrictive force is considered between neighboring cars at each car end. The simulation model is validated with field data of Shinkansen, and some calculation results are shown for the improvement of the riding comfort.

On-Track Test and Dynamic Simulation of Back Gauge of Gauge Changing Train

Gauge changing train is developed to realize through operation between standard gauge and narrow gauge. The wheelset of gauge changing train has small lateral clearance and torsion stiffness in wheelset assembly. Thus the amount of change of back gauge is larger than that of solid wheelset. To prevent the wheelset from misguiding (running into wrong track), it must be verified that the wheelset always keep back gauge in the predetermined range. This paper describes the relation between lateral force and back gauge by on-track test. Especially, while the wheelset passes turnouts, the back side of wheel contacts with a guardrail and large back lateral force widens back gauge. The amount of change of back gauge increases according to increase in lateral force, but is saturated at a certain value. The safety margin against misguiding is secured enough even when the maximum lateral force is loaded at turnout passing. Also, in this paper, numerical simulations using the simple wheelset model are performed. The characteristic of gauge changing wheelset is modeled by clearance, torsion stiffness and rolling friction. Authors confirmed that the results of on-track test are well reproduced by the wheelset model.

Basic Study of Wheel Flange Climbing Using Model Wheelset

This paper deals with an experimental study on the wheel flange climbing of railway vehicles, which is a major factor leading to derailment. An experiment is carried out on a 1/5-scale model wheelset of a truck used on a standard-gauge track, which is placed on a roller rig. The lateral external force acting on the wheelset is ramped up until derailment occurs under the condition of a fixed attack angle and wheel-load unbalance ratio. Three parameters, the height of wheel lift, the lateral force, and the wheel load acting on the outer rail, are measured until derailment occurs. From these measurements, it is possible to observe the behavior of the wheelset and to elucidate how the attack angle, the wheel-load unbalance ratio and the lateral external force affect flange-climb derailment. Then, a numerical simulation is carried out using an analytical model based on a single wheelset. As a result, the flange-climb behavior observed in the experiment can be explained theoretically on the bases of the analytical results, although further improvement of the model is desired.

Quantification of Riding Comfort for High-speed Trains on Curved Track Based on Experiments Using Vehicle Dynamic Simulator

Riding comfort of a train on curved track is generally evaluated with an index on stationary lateral acceleration of vehicle body. However, it does not consider the influence of running vibration throughout the curve and also it can not be applied on transition curve. This paper suggests a new method that quantifies riding comfort on whole curved section including transition curves. Experiments with subject passengers are carried out using a simulator to estimate the correlation between the riding comfort and combinations of related three parameters on vehicle body, the conditions of which are made based on lateral accelerations measured on a vehicle body. The parameters are as follows : riding quality level calculated from the running vibration, stationary lateral acceleration due to a centrifugal force and lateral jerk on a transition curve. As a result, the following matters are clarified. (1) Riding comfort on whole curved section including transition curves can be quantified by the index combined with above mentioned three parameters. (2) The influence of each parameter on riding comfort is estimated quantitatively. (3) The deterioration of riding comfort due to the increase of stationary lateral acceleration caused by speed-up can be compensated by decreasing the vibration level of vehicle body.