| Summary: | 碩士 === 國立臺灣大學 === 土木工程學研究所 === 91 === The investigation of characteristics of motorcycle traffic flow at mixed traffic is very important for understanding the spatiotemporal traffic condition for evaluating whether the traffic condition is safe or not. However, it is difficult to model a motorcycle traffic flow at mixed traffic. Therefore, this study will try to create a model of motorcycle traffic flow at mixed traffic and them to analysis the spatiotemporal changing behavior of motorcycle traffic flow at mixed traffic.
For constructing the model, chaos theory is employed. By using algorithm and developing a supervised spatiotemporal learning network, the chaos artificial neural network for motorcycle traffic flow at mixed traffic is derived in this study. Furthermore, by using the field observed traffic data, the experiential formulation of the model ism obtained. The research results show that the traffic flow can be predicted accurately using the model created in this study. In addition, the findings of the study depict the spatiotemporal changing behavior of motorcycle traffic flow at mixed traffic can be explained reasonably.
Finally, through a comparative analysis, it is found that the main advantages of the method developed in this study include improvement of the predicting accuracy and the enhancement of the executing efficiency. The research results indicate that the model in this study is better feasible and more practical usable.
|
|---|
