PSAM 16 Conference Paper Overview
Welcome to the PSAM 16 Conference paper and speaker overview page.
Lead Author: Shunichi Tada Co-author(s): Taro Kanno, kanno@sys.t.u-tokyo.ac.jp
| Applying the Genetic Algorithm for Finding the Worst Scenario for the Post-Disaster Recovery of Water Distribution Network |
|---|
| Because water is an essential resource for numerous activities, a water distribution network (WDN) is one of the most important lifelines; therefore, considerations must be made to prepare for the restoration of WDNs during post-disaster periods. To evaluate the restoration plan of damaged pipes, we are developing the agent-based simulation that can reproduce the restoration processes of the WDNs and how the restoration plan affects to the performance of each subsystem of a city during the post-disaster periods. In many researches, the damage scenario was usually manually generated; the number of damaged pipes was estimated by an empirical equation considering the magnitude of earthquake and the properties of pipes, while a geographical distribution of the damaged pipes was randomly selected. In our previous research, it was found that the performance of the restoration plan is highly dependent on the geographical distribution of the damaged pipes. Therefore, it is difficult to appropriately evaluate the resilience of WDNs using such randomly generated scenarios, rather it is necessary to find the most difficult scenario under the given number of damaged pipes. In addition, scenarios for training and exercise should be designed and prepared by their difficulties for training objectives, not randomly or in an ad hoc manner. However, it was difficult to classify the scenarios according to their difficulties or characteristics since there are a huge number of possible damage distributions. In this research, we applied the genetic algorithm to explore the most difficult scenario of damage distribution to repair. In this GA, an individual represents one disaster scenario that describes a geographical distribution of the damaged pipes, and the population represents a set of various scenarios. As an objective function, we used a resilience triangle obtained from the simulation results. Through the evolutional operations, such as selection, crossover and mutation, the individual with the largest resilience triangle, the worst scenario, will be obtained. We can also obtain a set of scenarios with different resilience triangles, that is a set of scenarios ordered by their difficulties. We conducted a test search with one hundred damaged pipes out of more than 4000 pipes and confirmed our proposed method reaches to conversion after 100 generations. Analyzing the population after conversion, we found there are some common pipes in difficult scenarios, which suggests these pipes were critical for the WDNs resilience. |
Paper TA160 Preview
Author and Presentation Info
Lead Author Name: Shunichi Tada (tada-shunichi513@g.ecc.u-tokyo.ac.jp)
Bio: Shunichi Tada is a Ph.D. student in the School of Engineering at the University of Tokyo. His research focuses on cognitive aspects of humans from the perspective of engineering and currently he is working on developing a system enhancing the interaction between human and computer-based technologies especialy the computational simulation.
Country: Japan
Company: School of Engineering, University of Tokyo
Job Title: Ph.D. student
Bio: Shunichi Tada is a Ph.D. student in the School of Engineering at the University of Tokyo. His research focuses on cognitive aspects of humans from the perspective of engineering and currently he is working on developing a system enhancing the interaction between human and computer-based technologies especialy the computational simulation.
Country: Japan
Company: School of Engineering, University of Tokyo
Job Title: Ph.D. student