Reducing Maintenance Cost of Mud Slide Prediction System with Categorization of Sensor Data in Sensor Network

摘要

Reduction of the human burden is necessary for the efficient use of the disasters prevention system predicing mud slide disaster with a wireless sensor network. This paper proposes a method to reduce the human burden by categorizing sensor data based on similarity of the characteristics of the sensor data. The user can determine the categorization interactively. Once categorization is determined by the proposed method, we can reduce the number of the data to be analyzed with the selection of a representative sensor data from each category. In addition, we can reduce the frequency of repairing damaged sensors by using another sensor data in the same category instead of the damaged ones. The experimental result shows that the proposed method can derive the categorization useful to reduce the human burden.