Smart Disaster Management and Prevention using Reinforcement Learning in IoT Environment

Abstract

At starting of the Internet of Things (IoT), it is passing around a world, in which diverse kinds of different objects are there connected to the Internet. It contains the use of smart phones, sensors, cameras, and other devices to make over the actions of people and things into data and link it to the Internet. With its capability to model the real world in digital form and accomplish scrutiny and replication in cyberspace, the IoT is able to reveal new value at an unparalleled rate and deliver it as response to the real world. This is set to convey main changes that will lengthen to the structure of industry in addition to the infrastructure of society itself. Therefore although the occurrence of the IoT contributes rise to new value, it besides means the occurrence of new threats. The proposed work covenant with disaster management as well as prevention to manufacturing industry using IoT. System first investigates the threat scenario during general execution of work, and finds the critical situations. The system processes learning approach for identifying such critical situations and execute the output appliances. System utilized multiple input along with output sensor for experiment. The Q-Learning approach has used for updating the policy which can provide the best result with high accuracy.