Machine Learning-Based IOT Air Quality and Pollution Detection

Abstract

In India, gas leakage from the different factories harmful to human surveying in the last fifty years is very low. However, there is a lack of prior detection of the chemical gases detection system in the situation raised. So, In this regard, there is a gap identification of chemical gases intensity detection needed. In this work, the main objective is to identify chemical gases intensity and maintain the stream data in the database from different locations. To fill this gap, that is identifying the high-intensity chemical gases from the chemical gas disaster areas. The first step needs to identify the different chemical gases and natural gas compositions. In this regard in this work for design internet-based gases in the air system. So, the sensors MQ2(Ethanol i-Butane Methane Alcohol Gas Sensor Sensor), MQ3(Sensitivity Alcohol Detector), MQ4 (Methane and Natural Gas (CNG)), MQ-5 ( LPG GAS SENSOR), MQ-7 (CO Gas Sensor Module Test Carbon Monoxide Detector), MQ-8 (hydrogen Gas Sensor), MQ-9 (carbon monoxide), MQ-135 Sensor(Air Quality Sensor Hazardous Gas Detector) and DHT11 Digital Temperature Humidity Sensors. These sensors are interfacing with the micro-control STM 32 board. It is also called one Pollution identification terminal by using it to pull the sensor stream data from location to centralized data. This stream data transportation is a service to pull the data. For this Data pulling, design an algorithm store into a cloud database. In this research work, design the electronic terminal with a wifi circuit using IoT technologies.Moreover, getting these attributes as data. Data need to apply the preprocessing techniques and extracted feature techniques also. This paper discusses mainly designing the terminal for pollution attributes , cleaning the data, and applying the Machine Learning based Feature extraction techniques.