A new objective function to build seismic networks using differential evolution

Abstract

Natural phenomena such as earthquakes have caused devastating effects in different cities around the word. To prevent a great disaster, it is necessary to construct seismic stations at strategical locations to warn population. Many Disaster Alert Systems (DAS), such as the Seismic Alert System of Mexico City (SAS) [4] or the Deep-ocean Assessment and Reporting of Tsunamis (DART II) [11], were located not based in earthquake or tsunami data, but simply by spacing the sensors more or less evenly around the contour of the Pacific Ocean. The objective of a DAS is simple: to emit an alert as fast as possible, in order to warn the population as early as possible. According to a new location of its seismic stations, the SAS could issue a longer warning time. This research focuses on designing the locations of seismic sensing stations maximizing the “warning time”; that is, the gap between the time when an earthquake is detected and the alert is launched, and the arrival time of the disaster. Since locating these stations is basically a numerical problem, in this research, the authors propose a new objective function to maximize the warning time using a differential evolution algorithm. In order to perform the experiments and validate the efficiency of the algorithm, it was considered the epicenters of recorded earthquakes located in the State of Guerrero, México. This data is used in the objective function to set the fitness value of a candidate solution. The main disasters targeted in this paper are earthquakes, but this research can be extended easily to tsunamis or volcanic eruptions alert systems, locating telecommunications antennas, etc.