A Machine Learning Approach to Self-Adjust Optimal Air Conditioning Temperature using Differential Air Velocity

Temperature prediction in cooling systems is a significant theme within the application domain of IoT and Machine Learning. It is tough to vary the temperature during sleeping hours, meeting times, and for the most part, wherever an arbitrarily large number of people occupy a room. There are several parameters that play a pivotal role in deciding the accuracy of temperature predictions made by automatic systems, which include the quality and quantity of the data used to train the machine learning model, the complexity of the model, and the variability of the environment in which the air conditioner is used. Automatic temperature prediction also improves comfort levels for building occupants, as the air conditioner is set to maintain the predicted comfortable temperature without requiring manual adjustments. The proposed Intelligent Air Conditioning system in the paper can predict the desired temperature of the user and self-adjust the predicted temperature of the room automatically using IoT and Machine Learning techniques. The algorithm deployed in the system inputs various physical parameters such as humidity, AQI, altitude, atmospheric pressure, number of occupants, room dimensions, ambient light, light intensity, and air velocity to ensure accurate temperature predictions. If the predicted temperature does not match the comfort temperature recommended by ASHRAE, the system will cool/condition the room with a temperature within the range of comfort temperature recommended by ASHRAE and close to the predicted temperature. The ceiling fan unit is employed to compensate for the temperature difference without decreasing the room temperature using differential air velocity. This promising technology will satisfy the complete level of automation through Image processing, Machine Learning, and IoT.