Reliable sleep temperature regulation based on PMV model and dynamic thermal parameters

Abstract

Quality of nighttime sleep is crucial. The common practice of air conditioners' sleep mode is to recommend a fixed sleep curve or suggest a pre-trained temperature curve based on geologic location. However, these models neglect the user preference and dynamic changes in the room's thermal environment, making it difficult to satisfy people of different ages and genders and resulting in discomfort during sleep. Demanding user to input their preferences or private data causes privacy issues, which leads to a bottleneck in solving this problem. Therefore, the paper proposes a new dynamic thermal parameter model for sleep, which learns the users' metabolic rate and thermal resistance of clothing and bedding through the user's undetected setting before sleep and simple feedback after waking up. These parameters fully reflect the inherent preferences of users and the thermal environment of rooms. After 1–2 rounds of feedback, the model can simulate the needs of users pretty well. In addition, we have optimized the PMV sleep model based on the dynamic thermal parameter model to regulate the temperature of air conditioner. In the controlled experiment, 84% of the participants expressed satisfaction with the regulated sleep temperature environment using the new model, which is 17% higher than that of the fixed curve. This model establishes a personalized model for users without obtaining their privacy information. In future research, wind speed control and humidity control can be introduced, and other sleep monitoring devices can be added to achieve more precise air conditioning temperature control.