Research On Intelligent Control Strategy And Control System Of Air Conditioner Based On Infrared Vision

With the rapid economic development and the substantial improvement of the quality of life,how to create a healthy and comfortable indoor thermal environment has become a problem that researchers continue to explore.Because the adjustment of the room temperature environment has the characteristics of hysteresis,non-linearity,etc.,and requires the active adjustment of the occupants,it often happens that the occupants feel too cold or too hot in real life.In addition,this adjustment method not only causes a certain degree of energy waste,but also cannot solve the problem of health damage caused by the inability to adjust the room temperature in time in a sleep or disability state.Therefore,it is of great significance to study air-conditioning intelligent control strategies and control systems that meet human comfort.Firstly,the paper introduces the framework of the air conditioning control system based on infrared vision and human comfort.The software and hardware design with the Raspberry Pi as the processor and the MLX90640 array thermal infrared sensor as the main components are elaborated in detail.It also includes components such as the camera,laser ranging sensor and digital temperature sensor.The system uses infrared temperature sensor to measure the human body temperature,calibrates the human body temperature and uses fuzzy logic algorithm to judge the comfort level of the human body.After defuzzification,the precise airconditioning control quantity is obtained to control room temperature and body temperature.Secondly,the thesis carried out temperature compensation correction to the infrared temperature measurement module of the system through experiments.First,the measurement errors of the infrared temperature sensor MLX90640 at different measurement distances and ambient temperatures are collected as sample data,and then three machine learning algorithms:feedforward neural network,extreme learning machine and support vector machine are used to train and predict the samples.Experimental results show that these three machine learning algorithms can greatly improve the measurement accuracy of infrared temperature sensors.Finally,the paper proposes an air-conditioning control strategy based on fuzzy control algorithm,and explores the feasibility of this strategy to meet human comfort in room temperature control through experiments.The experiment first takes room temperature as the control target to achieve accurate control of the room temperature in different areas,and then conducts the experiment with the human body surface temperature as the control target.Experimental results show that the proposed control strategy can effectively adjust the air conditioner and meet the requirements of human comfort.After two stages of experiments,it is concluded that the air-conditioning control strategy and system studied in this paper are expected to improve the defects of traditional air conditioners that require manual operation and solve the problem of physical discomfort caused by the inability to adjust the temperature in time,and the air conditioning can be reasonably and intelligently controlled to meet the goal of human comfort based on real-time measured body temperature.