Browsing by Author "Jou, YJ"

Now showing 1 - 3 of 3

Estimation of dynamic origin-destination by Gaussian state space model with unknown transition matrix
(2003)
Jou, YJ
;
Hwang, MC
;
Wang, YH
;
Chang, CH
;
統計學研究所
Dynamic origin-destination (O-D) pattern representing time-dependent trip demands from one place (origin) to another (destination) is one of the most essential input data for most traffic operational analyses. Historical studies assumed that the transition matrix is known or at least approximately known, which is unrealistic for a real world network. And due to the fact that the number of trips to a specific destination, y, is easy to obtain and the O-D variable, x (path flow based in this research), is not directly observable, a Gaussian state space model is formulated to describe the relationships of x and y, observation equations, and the dynamics of x, state equations with unknown transition matrix. Under the assumption of Gaussian noise terms in state space model, the distribution of random transition matrix F is derived. A solution algorithm combining Gibbs sampler and Kalman filter to tackle the problem of simultaneous estimation of F and x, based on the latest available information is proposed. Real O-D data from Taipei Rapid Transit is used to verify the presented model and solution method. Preliminary results are generally satisfactory, showing that also in the unknown transition matrix case, significant estimates are achieved.
Modeling of nonlinear stochastic dynamic traffic flow
(2001)
Jou, YJ
;
Lo, SC
;
運輸與物流管理系註：原交通所+運管所
;
Department of Transportation and Logistics Management
Most dynamic flow models are developed under deterministic assumptions or simple linear models. Although these models can describe dynamic phenomena, they cannot adapt a variance of the real world. However, in the developing trend of intelligent transportation systems, operators have to understand and accurately predict traffic flow to predict, evaluate, and manage the performance of present and future systems. Thus, a model capable of describing variant traffic phenomena, which encompasses both nonlinearity and stochasticity, is necessary. This study formulates nonlinear stochastic dynamic traffic flow models based on conventional macroscopic models; the nonlinear terms are decomposed by polynomials to reduce the complexity of the models. Then, the 116 equation is introduced to convert the deterministic model to a stochastic one. Also considered here is the traffic flow model with a diffusion effect.
The implementation of Markov bias corrected grey system in freeway travel time prediction
(2003)
Jou, YJ
;
Lee, TT
;
Lan, CL
;
Hsu, CH
;
統計學研究所
The purpose of this study is to develop a freeway travel time prediction. model and, combined with real-time database, to obtain the most recise travel time prediction. First, a Grey System Theory for travel time prediction is developed, and then Markov Bias Correction mechanism is used for bias correction. Prediction is divided into "data-of-week" in time category because of that, travel time is corresponding to travel pattern and,travel pattern itself is highly related to whether it is weekday or weekend. In the link section, this study divided the path into pairs of interchanges because of the information obtained. And then, construct a real-time database, and connect to Taiwan Area National Freeway Bureau via internet for real-time travel speed.-and uses this information on freeway travel time prediction. Besides prediction, this study also. uses this information for calibration.