| For bedrooms and hospital patient rooms,the bed is the area where people are exposed for the longest time and where the human body is one of the main source of contamination(biological effluents from the body and exhaled air).At present,such rooms usually use total volume ventilation to improve indoor air quality,which not only consumes more energy and has poor pollution control ability.Personalized ventilation systems should be integrated with the bed environment to solve pollution problems at the source and provide a personalized and optimal bed microenvironment.Therefore,a ventilated mattress was designed to remove bioeffluents from the bed before mixing with room air.The ventilated mattress is filled with a textile material made of acid-treated activated carbon fibers,which are drawn into the mattress through an opening located near the end of the bed at the feet,and simultaneously sucked away.The contaminated air is exhausted to the outside or locally cleaned by a filter placed in the mattress.Ventilated mattresses have been shown in previous studies to be energy efficient in removing contaminants and meeting user requirements for respirable air quality,followed by manikin experiments to investigate the effects of ventilated mattress use on the thermal environment and found that the application of ventilated mattresses can disrupt the original thermal environment and increase the risk of local discomfort.However,the interaction between the human body and the ventilated mattress microenvironment is unknown,the heat transfer mechanism is unclear,and there is a lack of experimental basis,which restricts the design of personalized intelligent bed microenvironment.Therefore,this paper investigates the optimal design strategy of the bed microenvironment control system based on the real-time regulation experiment of the bed microenvironment based on the thermal sensory feedback of real subjects.The thermal sensation of subjects was assessed in a comprehensive manner using various measures,such as body temperature,skin temperature of different body parts,and a thermal sensation questionnaire.The micro-environment of the bed was customized based on participants’acceptability of thermal sensation.The experiment also included a personalized intelligent control strategy for micro-environment based on physical and physiological signals.Heart rate and variability were also measured,and participants wore sleepwear during the experiment.The main research work and findings of the paper are as follows:(1)The experiments of bed thermal microenvironment control based on the real-time response of the subjects reveal the mechanism of the ventilated mattress on the bed thermal microenvironment and propose a personalized control strategy for the bed microenvironment based on the physical parameters of the microenvironment.The experimental results show that the distribution of bed thermal microenvironment is not uniform,and the highest temperature difference is up to 13℃.The use of ventilated mattresses brought local cooling effects for 78%of women and 75%of men.This improved the acceptability of the environment for the subjects at high room temperatures(28°C).At 19°C and 23°C,the use of a ventilated mattress alone may cause local cold discomfort and some subjects required local heating.Subjects’acceptability of the thermal microenvironment in bed increases with continuous adjustment of airflow rate and local heating,and the risk of local discomfort(hot torso,cold extremities)decreases.The airflow rate of the ventilated mattress can be set to a constant value of 6 L/s,and the temperature of the measurement point on the mattress surface near the feet correlates with the thermal sensation of the feet and can be used to control local heating.(3)By analyzing the correlation between physiological parameters and thermal sensation,the control strategy based on the skin temperature of subjects is investigated.It was found that the correlation between local skin temperature and local thermal sensation was strong,among which,the correlation between feet skin temperature and feet thermal sensation was the strongest with the highest correlation coefficient r~2 reaching 0.9.The difference between the overall thermal sensation and the skin temperature of the back of the foot showed a linear correlation.The individual optimal skin temperature can be calculated based on the correlation parameters,so that individual control strategies can be developed for individual foot and back skin temperatures.For more advanced control requirements of other parts,the skin temperature corresponding to the acceptable degree can be calculated by Gaussian regression to calculate the optimal skin temperature interval,so as to design control strategies for different parts.The method can continuously record iterative user control habits and update the optimal skin temperature interval on the basis of automatic control,so as to achieve the purpose of intelligent control of personalized bed thermal microenvironment for different users.(4)Six different machine learning models are established,which can be used to achieve the prediction function of the overall thermal sensation in the bed thermal microenvironment,and the data dimensionality reduction is performed by classification database,principal component analysis to improve the performance of the machine learning models.The advantages and disadvantages of several machine learning algorithms are compared,and it is found that the data after principal component analysis makes the quality of the machine learning algorithm significantly improved,and the SVM algorithm performs better compared to other machine learning model algorithms when dealing with poor quality data sets.The method enables the prediction of overall thermal sensation in the practical application of ventilated mattresses for the purpose of intelligent control. |
PDF Full Text Request