No abstract

To study the vibration of a passenger's head and internal organs at different locations of a high-speed train, a 9-degrees-of-freedom (DOF) model of seated passengers is proposed in this paper, and its parameters of the damping coefficients and stiffnesses are identified. Next, the response of the head and internal organs is simulated by applying the vibrational stimulation generated by a 27-DOF vehicle model under track irregularity. Moreover, by applying the measured vibration signal, the following conclusions can be drawn: (1) the weakest response is detected at the centre of the compartment of the wagon, and a stronger response is detected at the centre of the bogie, with the rolling motion having a greater effect 1 m away from the centre of the bogie; (2) the response of the human internal organs is stronger than that of the head under stimulation with a lower frequency of less than 3 Hz, and a similar conclusion can be drawn in the range of 5 to 8 Hz. However, if the frequency is in the range between 8 and 15 Hz, the situation is entirely different. The responses of both the head and internal organs are reduced at frequencies over 20 Hz; (3) from the real application, it can be inferred that the greatest response can be detected at approximately 3 Hz for internal organs and at 8 Hz or higher for the head.

The precise location ofa moving vehicle is the one of kyfactors in intelligent transpnotation system JTS) The Kalman filtering technique, which is often applied directly to Vehicle A/avigation ysktem (ES) operated by UPS, can give optimal estimation of movirng vehi'cles. The navigation accuracy using Kalman filter depends on a reliable function model, stochastic model and proper esrtmation method. In order to gain good navigation precision, an adaptively robust Kalman filereing algohithnm based on the current statistical model is presented in this paper. Tis filter is a corbination of adaptive LTD decomposition Kalman filter with Quasi-Accurate Detection of gross errors QUAD) method. It uses QUAD method to detect anid correct gross errors i.n UPS observations, applies LD decompositionl technique to improve computation precision and stability of filter and employ Sage adaptive filter to avoid divergence. Features ofgood stabifity strong adalptabiity and well elimination of gross errors of this new conmprehensive filter have been demonstrated with two simulation examples at last. I. INTRODUCTION Intelligent Transportation Systein (ITS) technologies are generally considered to be the incorporation of advanced technology into all parts of the transportation system to improve the movement of people, goods, services and information, as in [1]. They can include more immediately recognizable features such as emergency call boxes, changeable message signs, signal synchronization and preemption, and Highway Advisory Radio, to more advanced technologies such as Traffic Operations Centers, Automatic Vehicle Location devices, information kiosks, and electromnc payment services for tolls. GPS-based Vehicle Navigation System (VNS), as one of the most important part of ITS, can tell you exactly xxhere yon are and where yotn are heading, because the VNS operates by GPS, combined with GIS technology holds all the information the system needs to show you the way. In general Kalman filter is often desiged to give an accurate estimate of future state of a moving vehicle. More comprehensive description of the Kalman filtering system appears in [2] to [11]. So we will use data fusion algonffiths specifically like GPS and Kalman filters to improve the safety and efficiency of transportation systems particularly with respect to GPS signal being lost in this paper.When using Kalman filter algorithm for kinematic positing and nMaigation of a moving vehicle, reliable function model, stochastic model and reasonable estimation method are necessary. First of all, the selected physical equation of motion and observation should precisely characterize physical and geometric facts of a moving vehicle. Then the variancecovariance matrix of model error and observation error should properly describe confidence level of function model and observation information. Meanwile, the principle and algorithm of parameter estimation might solve the precise ad reliable parameters of motion state employinig valuable information provided by motion equation and o...

No abstract

No abstract