Deploying Firefly DNN and Spider Monkey Model in Flood Disaster Surveillance

Authors

  • Mr. Raghukumar K. S, Dr. Rajashree V Biradar

Abstract

In numerous Asian countries, says India, flooding has turned out to be a common trait that has attained the disasters’ status causing devastations with reference to the loss of lives as well as big economic losses each year because of inundation in lower regions down river channels and also localized unprecedented precipitation. Research on Flood Disaster (FD) produces ideas, in addition, provokes the best solution for Disaster Management (DM). Here, an IoT and Big Data (BD) centered flood disaster monitoring system is proposed utilizing Firefly algorithm based Deep Neural Network (FFA-DNN). The proposed Flood Monitoring System (FMS) comes under ‘2’ stages: i) training and ii) testing. In training, ‘3’ processes: a) preprocessing, b) Feature Selection (FS), and c) classification, is carried out. Initially, the flood BD dataset is utilized to gather the flood data. Then, these amassed flood data is preprocessed, and after that, the pertinent features are selected as of the pre-processed data utilizing the Crossover and Mutation centered Spider Monkey Optimization Algorithm (CM-SMOA). Next, these selected features are inputted to the FFA-DNN for classification. The classification outcomes of FFA-DNN contain ‘2’ classes of data: i) chances for the occurrences of flood and ii) no chances for the occurrences of a flood. After training, the testing operations test the real-time sensed IoT data by comparing the training results. The proposed technique’s outcomes are examined and contrasted with the other existing techniques to confirm that the proposed FFA-DNN identifies the FD centered on IoT - BD more efficiently.

Published

2020-12-04

Issue

Section

Articles