DEVELOPMENT OF A HARDWARE-SOFTWARE SYSTEM USING ARTIFICIAL INTELLIGENCE FOR CALIBRATION OF THE MQ-2 GAS SENSOR

Abstract

This study examines the possibility of calibrating the MQ-2 gas sensor based on the ratio of the sensor's resistance in clean air to its resistance in the presence of gas, using the approximation of the dependence of these two values on gas concentration in PPM. Additionally, it explores the use of artificial intelligence for calculations. The article focuses on analyzing the calibration problem, reviewing existing software or hardware-software solutions, identifying challenges, and introducing innovative methods in the form of artificial intelligence to address the calibration issue. The study also describes the development of a hardware-software system based on the NodeMCU V3 ESP8266 Wi-Fi module, which enables reading data from the MQ-2 sensor, connecting to the global internet, and transmitting the sensor readings to artificial intelligence hosted on the network. Given the use of the NodeMCU module, the programming language chosen for the software is C++, with Arduino IDE as the development environment. Furthermore, the article details methods for optimizing data transmission to artificial intelligence to speed up the calibration process. An analysis of the data obtained from the developed hardware-software system will be conducted and compared to a reference value. The reference values will be obtained using the professional gas concentration measuring device HABOTEST HT601B+, which is certified for quality and, therefore, provides reliable measurements. All measurements from the developed hardware-software system and the reference device will be carried out in an isolated hermetic environment to avoid erroneous readings. Based on the research results, conclusions will be drawn to determine whether the proposed methods are indeed effective and whether they can be scaled for larger, more complex tasks.