A WSN-based landslide prediction model using fuzzy logic inference system

Abstract

This paper proposes a new WSN-based Landslide Prediction Algorithm, developed using Fuzzy Logic Inference System. Three factors conditioning landslides are considered, namely: slope angle, soil moisture and topographical elevation. These conditioning factors are sensed using WSN and analysed using proposed algorithm at sink node. A Mamdani-type Fuzzy Inference System (FIS) is used to develop the algorithm. Triangular membership functions are considered for all FIS parameters. A total of 45 rules have been developed in this FIS, which holds capability to generate a three-level alarm to warn residents of area about any impeding danger due to landslide. For results, surface plots are generated which show us the variation of landslide susceptibility with the change in three parameters considered.