Link to Content Area
:::

Institute of Transportation, MOTC

:::
  • small size
  • medium size
  • large size
  • print
  • facebook
  • plurk
  • twitter

Transportation Dissertation

Title Survivable Network Design Model for Earthquake Disaster
Year 2006
Summary

Peng-Hsi Hou, 2006.09

Institute of Traffic and Transportation National Chiao Tung University

  Disaster-prevention network is a crucial system in a disaster prevention city. Therefore, how to establish design principles for the network becomes an important issue. In this paper, we proposed the method to assess the spatial performance and the importance against disasters in individual road sections. In the viewpoint of comprehensive network, we proposed the network design model with respect to the characteristics of earthquake disaster based on survival network theories. The design goals are to achieve features of rapidity, connectivity, and efficiency. Three network models: Shortest Path Forest Network (SPFN), Systematic Detour Network (SDN), and Mutual Assistance Network (MAN), are developed and integrated in the survivable network model. The proposed survivable network for earthquake disaster has significant advantages. The SPFN promises the shortest path within the responsible territories. The structure of SDN assures the network connectivity wherever any road link fails. The bridges of MAN let the supply units in different territories connect to each other hence insufficient self-capacity of one supply may be allowed. Finally the survivable network model was implemented with Java technology. And a case in a test network of grid pattern was given to show the features in the scenario of evenly scattered demand areas with two supply resources. It was found that where the two resources were increasingly separating from each other, the average travel cost steadily decreased; detour cost dropped dramatically and then declined slowly; mutual assistance cost gradually increased; network cost seriously varied in the initial stage, but converged to a gently decreasing trend afterwards. 

Count Views:301
Top