Method for testing influence of indoor CO2 concentration and temperature interaction on human cognition

Abstract

The invention discloses a method for testing influence of indoor CO2 concentration and temperature interaction on human cognition, which belongs to the field of indoor environment control and mainly comprises the following steps: screening testers participating in an experiment; exposure environment construction: preparing an environment chamber and setting experiment conditions, wherein the experiment conditions comprise setting different CO2 concentration levels and setting different temperature levels; measurement of cognitive performance: performing different types of cognitive task tests in an exposure environment, and measuring subjective response, physiological parameters and an electroencephalogram; and measurement data analysis: processing the measurement data under different types of cognitive tasks, and comparing the sensitivities of different types of cognitive tasks to different components in the environment. According to the invention, by combining psychological, physiological and neural tests and comparing different cognitive tasks, the existing evidence about the influence of carbon dioxide concentration and temperature increase on human beings is supplemented. Meanwhile, a deep learning cognitive comfort model is constructed based on time domain electroencephalogram features, a new insight is provided for the influence of environmental factors on cognition, and potential application of optimizing cognitive comfort through environmental control is prompted.

Description

Technical Field

[0001] The present invention relates to the field of indoor environment control, in particular to an indoor CO 2 Methods for testing the effects of concentration and temperature interactions on human cognition. Background Art

[0002] Carbon dioxide (CO 2 ) is a normal atmospheric component, with a content of about 400 ppm. According to ASHRAE Standard 62.1, CO 2 Concentration is considered to be an indoor air pollutant caused by human metabolism. Indoors, due to human activities, CO 2 The Occupational Safety and Health Administration (OSHA) and the American Conference of Governmental Industrial Hygienists (ACGIH) have set a CO2 limit of 5,000 ppm for an 8-hour workday. 2 The World Health Organization (WHO) and ASHRAE recommend keeping indoor CO2 levels below 5000ppm. In addition, temperature also significantly affects human cognitive ability and is widely discussed as an indoor factor. Among noise, CO2 concentration and temperature, temperature is the most important factor affecting cognitive ability.

[0003] Studies have shown that high CO 2 Concentrations or high temperatures significantly reduce decision-making performance on complex cognitive tasks. Test subjects reported more headaches, fatigue, agitation, and depression, along with thermal discomfort and symptoms such as drowsiness, decreased weight, respiratory rate, heart rate, skin temperature, eardrum temperature, and arterial oxygen saturation. There was also a significant correlation between EEG signals and thermal comfort. Cognitive performance in traditional environments was significantly lower than in green buildings, with test subjects experiencing more symptoms and increased heart rate. However, these results are controversial, as some studies have found no significant changes in acute health symptoms, perceived air quality, or performance on tasks similar to typical office work, and no significant effects on physiological responses or performance on simple cognitive tasks.

[0004] Furthermore, previous studies have focused on building comfort models using physiological signals by changing environmental conditions, mainly modeling thermal comfort, anxiety, and cognitive comfort. Features such as skin temperature (ST), heart rate (HR), electrodermal activity (EDA), and EEG signals have been shown to be significantly correlated with comfort. Among them, modeling methods include traditional machine learning methods such as k-nearest neighbor (KNN), random forest (RF), and support vector machine (SVM), as well as advanced deep learning models such as deep neural network (DNN) and transformer-based networks. These techniques have been used to model comfort.

[0005] Most studies have been univariate, examining CO using a single high-level cognitive task or similar office tasks. 2 Effects of increasing CO2 concentrations and temperatures on cognitive performance. Some literature has used multi-channel functional near-infrared spectroscopy (fNIRS) to investigate cortical activation during the Stroop task under these conditions. However, gaps remain in understanding how different types of high-level cognitive tasks respond to the same exposure conditions, and the physiological drivers remain underexplored. Electroencephalography, a non-invasive method for studying brain function, has not been used to examine the interaction of CO2 concentration and temperature on cognitive performance, and there is a lack of cognitive comfort models specifically for indoor environments. Summary of the invention

[0006] The purpose of the present invention is to overcome the technical problems existing in the prior art and provide an indoor CO 2 The approach to testing the interactive effects of concentration and temperature on human cognition aims to complement existing evidence on the effects of increasing CO2 concentrations and temperature on humans by combining the use of psychological, physiological, and neural tests and comparing different cognitive tasks.

[0007] The objective of the present invention is achieved through the following technical solutions:

[0008] Provide an indoor CO 2 A method for testing the effect of concentration and temperature interaction on human cognition, characterized by comprising:

[0009] Screening of testers to participate in the experiment;

[0010] Exposure environment construction: prepare the environmental chamber and set the experimental conditions, including: setting different CO 2 Concentration level, setting different temperature levels;

[0011] Measurement of cognitive performance: Different types of cognitive tasks were performed in the exposure environment and subjective responses, physiological parameters, and EEG were measured;

[0012] Measurement data analysis: Process the measurement data under different types of cognitive tasks and compare the sensitivity of different types of cognitive tasks to different components in the environment.

[0013] In some embodiments, the screening of test subjects to participate in the experiment includes:

[0014] 24 healthy test subjects were selected, including 12 females and 12 males, aged 18-31 years old, with a body mass index of 18.3-28.8 kg / m 2 .

[0015] In some embodiments, the environmental chamber uses a cylinder to provide CO to the subject's breathing zone. 2 , to maintain uniform CO 2 Concentration field, determination of CO by non-dispersive infrared carbon dioxide sensor 2 concentration and temperature.

[0016] In some embodiments, the different CO 2 The concentration levels include 500 ppm and 5000 ppm; the different temperature levels include 22°C and 28°C.

[0017] In some embodiments, the different types of cognitive tasks include reactive control tasks and proactive control tasks.

[0018] In some embodiments, the measurement of the subjective response includes: collecting subjective responses through an electronic questionnaire and scaling the thermal sensation under different conditions; the physiological parameters include blood pressure, body temperature, heart rate and cortisol.

[0019] In some embodiments, the measurement of the EEG comprises:

[0020] During the task, EEG activity was recorded continuously from 32 electrodes at a sampling rate of 512 Hz using a Grael 45ch amplifier compatible with the Green Panel and Curry 9 software.

[0021] In some embodiments, the processing of measurement data under different types of cognitive tasks includes:

[0022] The acquired raw data were filtered; the noise in the EEG was removed by independent component analysis; for each task, the slice range was from -1s to 2s, the baseline correction was from -1 to 0, and the trials with positive and negative amplitudes greater than 100 were deleted; the frequency bands were divided into δ (0.5-4Hz), θ (4-8Hz), α (8-13Hz) and β (14-27Hz) according to fast Fourier transform, and the power spectral density of each frequency band was calculated.

[0023] In some embodiments, the method further comprises:

[0024] The cognitive comfort model is constructed using the sliced ​​time domain signals.

[0025] In some embodiments, the cognitive comfort model includes EEG-TCNet64, SVM and RandomForest.

[0026] It should be further explained that the technical features corresponding to the above options can be combined or replaced with each other to form a new technical solution if there is no conflict.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] This paper supplements the existing evidence on the effects of increased carbon dioxide concentration and temperature on humans by combining psychological, physiological and neurological tests and comparing different cognitive tasks. At the same time, a deep learning cognitive comfort model based on time-domain EEG features is constructed, which provides new insights into the effects of environmental factors on cognition and suggests potential applications for optimizing cognitive comfort through environmental control, providing important empirical evidence and support for healthy living in indoor environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Flow chart of the experimental procedure of Stroop and DMS tasks shown in the embodiment of the present invention;

[0030] Figure 2 A frequency domain signal comparison diagram of an EEG topographic map shown in an embodiment of the present invention;

[0031] Figure 3 The embodiment of the present invention shows the different CO 2 and reaction time and accuracy of the Stroop and DMS tasks under temperature exposure conditions;

[0032] Figure 4 Physiological indicators after task testing under different exposure conditions shown in the embodiments of the present invention;

[0033] Figure 5 ERP waveforms under four different environmental conditions during STROOP and DMS tasks shown in an embodiment of the present invention;

[0034] Figure 6 The results of the interaction of five scalp regions under different exposure conditions are shown in an embodiment of the present invention;

[0035] Figure 7 The classification results of the embodiments of the present invention using three machine learning models under four environmental conditions are shown.

[0036] Note: C1T1: 500ppm, 22℃; C2T1: 5000ppm, 22℃; C1T2: 500ppm, 28℃; C2T2: 5000ppm, 28℃. DETAILED DESCRIPTION

[0037] The technical solution of the present invention is clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present application described and shown in the accompanying drawings can be arranged and designed in various configurations. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0038] It should be noted that the defects existing in the solutions in the above-mentioned prior art are the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above-mentioned problems and the solutions proposed in the embodiments of the present application for the above-mentioned problems below should all be the contributions made by the inventor to the present application in the process of invention and creation, and should not be understood as technical contents known to technical personnel in this field.

[0039] The technology of the present invention originates from a national key research and development project (approval number: 2022YFF1202900). In view of the technical problems pointed out in the background technology, the embodiments provided by the present invention are as follows:

[0040] In an exemplary embodiment, the specific effects of elevated CO2 concentrations (500 ppm and 5000 ppm) and temperatures (22°C and 28°C) on cognitive performance and neural responses were explored, and a cognitive comfort model was constructed using EEG features. 24 test subjects were divided into four groups, each exposed to four conditions in two 2.5-hour sessions within a day. They completed computer-based cognitive tests and questionnaires in each condition. CO2 was found on noise acceptability, thermal comfort, and EEG, theta, and delta bands. 2 A significant interaction between concentration and temperature was observed. EEG features from these tests were used to train a machine learning model to distinguish cognitive states, providing new insights into the impact of environmental factors on cognition.

[0041] The specific test method includes the following parts:

[0042] 1. Tester

[0043] Twenty-four (N = 24) healthy college students were recruited through social media in Tianjin, including 12 females and 12 males, aged 18-31 years old, with a body mass index of 18.3-28.8 kg / m 2. They had no self-reported history of smoking, drinking, or drug abuse. Exclusion criteria included self-reported or suspected: 1) history of claustrophobia, anxiety, or depression; 2) neurological diseases or brain trauma that may affect brain function and structure; 3) cognitive dysfunction; 4) cardiovascular and cerebrovascular diseases, diabetes, hypertension, or other chronic diseases; 5) asthma, allergic rhinitis, bronchitis, or other respiratory diseases; 6) metal products or dentures in the body. The subjects were told not to drink caffeine or alcohol and to sleep in 24 hours before the experiment. The women participating in the experiment were not in their menstrual cycle during the exposure period. During preparation, the subjects adjusted their clothes until they felt thermoneutral. Then, they were asked to use the same clothing parts throughout the exposure experiment.

[0044] 2. Exposure environment construction

[0045] Environmental chamber. The experiment was conducted in a small environmental chamber (3.24m*3.17m*2.55m) at Tianjin University. Figure 1 A shows that CO 2 The structure of the concentration and temperature exposure chamber and the subjects involved in the task. The room is equipped with two tables, each with a computer connected to the Internet. The cylinder provides CO to the subject's breathing zone. 2 , to maintain uniform CO 2 Concentration field. CO 2 The concentration and temperature are determined by non-dispersive infrared CO2 sensors. The sensors are located in three different directions for averaging.

[0046] Experimental conditions. Two CO 2 Levels: 500ppm and 5000ppm, and two temperature levels: 22°C and 28°C. The rationale for choosing these conditions is briefly described below:

[0047] Human-based ventilation rates and CO 2 Generation rate, measured indoor CO 2 Levels rarely exceed 5000 ppm, the current 8-hour occupational exposure limit set by OSHA and ACGIH. 2 The reference exposure condition was set at 500 ppm from outdoor air and when the test person was breathing. ASHRAE-55 (2004) shows a thermal comfort range of 20.5-25.5°C in winter and 24.5-28.0°C in summer. Therefore, 28°C was used as the high temperature condition and 22°C was selected as the neutral temperature condition. In a preliminary study on cancer patients, researchers also used a temperature comparison of 22°C and 28°C.

[0048] 3. Measurement

[0049] Measurement of cognitive performance. Neurobehavioral methods are often used to assess responses to external environmental stimuli. Human behavior itself is mainly classified according to emotion, cognition, and execution. Among them, cognition can be subdivided into four main functions: learning and memory, thinking, perception, and expression. According to Braver's (2012) dual control mechanism (DMC) framework, reactive control is a post-correction process that occurs after the detection of a conflict or error event. It allows the selection of an appropriate response by inhibiting interfering responses. In contrast, proactive control is a top-down selection process that occurs before cognitively demanding events. It maintains task goals and biased sensory processing to facilitate the processing of task-related information. Task selection is based on the above two theories. The Stroop task is a reactive control task that belongs to the category of perception and thinking. The DMS task is a proactive control task that belongs to the category of perception and memory. All cognitive tasks were presented using E-Prime 2.0 software (Psychology Software Tools, Pittsburgh, USA).

[0050] Furthermore, in the Stroop task, participants named ink colors under congruent and incongruent conditions, while ignoring reading the color words; in the DMS task, participants simply maintained the stimulus in the maintenance condition, while mirroring the colored squares perpendicular to the stimulus midline in the manipulation condition; Questionnaire i included participants’ demographic information, perceived environment, and acute health symptoms; Questionnaire ii included information on perceived environment, acute health symptoms, and task evaluation; the Stroop task consisted of congruent and incongruent trials ( Figure 1 B). Stroop trials were randomly presented in 10 blocks, each consisting of 16 trials. Subjects were allowed to familiarize themselves with the task in 16 simulation trials before the 160 task trials. All stimuli were presented on a white background until the subject pressed a key and were indicated by a central black fixation cross for 500ms. In congruent trials, the Chinese color words for red, yellow, blue, and green were indicated by the corresponding ink colors. In contrast, incongruent trials consisted of color words and colors that differed in their meaning. Subjects were asked to indicate their respective ink colors as accurately and quickly as possible while ignoring the meaning of the color words. Responses were made by pressing the keys “J” and “K” on the keyboard with the right index finger and “F” and “D” with the left index finger.

[0051] The DMS task consists of two conditions, namely, maintenance trials and manipulation trials, each of which includes 54 task trials and 18 practice trials ( Figure 1B). Stimuli consisted of 1 or 4 randomly presented colored squares integrated into a 6 × 6 matrix. On hold trials, subjects were asked to maintain the position of a green square during a delay period. The stimulus lasted for 500 ms, followed by a 2000 ms delay period, during which afterimage effects were prevented by presenting a mask consisting of gradient gray squares. On manipulation trials, subjects mentally mirrored the positions of four red squares perpendicular to the stimulus midline and maintained the mirrored positions. The stimulus lasted for 1000 ms, followed by a 2000 ms delay period, during which afterimage effects were prevented by presenting a mask consisting of gradient gray squares. Then, a probe matrix was presented whose gray squares either matched or did not match the positions of the colored squares of the encoding stimulus. Subjects were instructed to determine whether the gray square of the probe was in the same position (match) or different (mismatch) than the maintained colored square by pressing the key “F” on the keyboard with the left index finger or the key “J” with the right index finger. On non-match trials of the manipulation trials, only one of the four squares did not match the correct mirrored position. The correct response on half of the trials was a match and on the other half was a mismatch. During the probe presentation, the subjects were asked to respond as accurately and quickly as possible. During the adjacent jittered inter-trial interval (ITI), a central fixation cross with a duration of 500-1500ms appeared. All stimuli appeared in the center of a white background.

[0052] Subjective measurements. An electronic questionnaire was used to collect subjective responses. Questions on thermal acceptance (TA), thermal sensation (TS), and thermal comfort (TC); perceived air quality (PAQ), odor intensity (OI), noise, lighting; and sick building syndrome symptoms (SBS). Considering that there were only two completely different temperature conditions in the experiment, the 7-point scale of ASHRAE38 was chosen to determine thermal sensation.

[0053] Measurement of physiological parameters. Physiological parameters include blood pressure, body temperature, heart rate and cortisol. The instruments used for measurement have been calibrated. Blood pressure and heart rate were measured with an Omron-U724J sphygmomanometer (40-180 beats / min ± 5% accuracy). A glass mercury thermometer was used to measure body temperature. Measure axillary temperature. Before and after contact, saliva samples were collected for subsequent analysis of changes in the stress biomarker cortisol. The testers were asked to drool into a container to collect samples. The collected saliva samples were immediately stored in a refrigerator (-80°C) until analyzed using a human salivary cortisol ELISA kit (Wuhan Adandi Biotechnology Co., Ltd.). Before using the ELISA kit, all samples were centrifuged at 2000rpm for 20 minutes to obtain the supernatant. The ELISA kit was then used for a series of operations. Finally, the enzyme-labeled coated plate was placed at a wavelength of 450nm (Multi-function microplate reader-H1M, Guangzhou Darui Kangti Engineering Technology Co., Ltd.) to measure the OD value of each well.

[0054] EEG measurements. During the tasks, EEG activities were continuously recorded from 32 electrodes at a sampling rate of 512 Hz using a Grael 45ch amplifier compatible with the Green Panel (Wuhan Greentech Technology Co., Ltd., China) and Curry 9 software. The scalp electrodes were mounted on a fitted cap in view of the extended 10-20 system. The electrode impedance was mostly kept below 5 kΩ during the continuous EEG recording. The acquired signals were referenced by M1 and M2. The EEGLAB toolkit in MATLAB was used to remove EMG artifacts and eye movements from the raw recordings.

[0055] 4. Experimental steps

[0056] During the experiment, individual differences needed to be offset, and a balanced Latin square design was used to control the effects of practice and fatigue, in which 24 testers were divided into 4 groups, with 6 testers in each group. Only one person in each group participated in the experiment at a time. In addition, each group was exposed to two temperature conditions for one day. Each temperature exposure condition lasted 2.5 hours from 9:00 to 11:30 in the morning or from 2:00 to 4:30 in the afternoon. Table 1 shows the standardized two-by-two balanced Latin square design, in which factor T represents the temperature variable (T1-22°C, T2-28°C), factor D refers to the four-day period, and factor C represents the carbon dioxide concentration variable (C1-500ppm, C2-5000ppm).

[0057] Table 1 Standardized two-by-two balanced Latin square design

[0058]

[0059] Figure 1 C shows the progress of the experiment. In the preparation phase, the subjects entered the cabin, wore the EEG equipment, filled out questionnaire i, and took baseline physiological measurements within 30 minutes. The experiment was carried out in the order of exposure. The detailed experimental plan for the first half day is as follows Figure 1 C. Each CO 2 The concentration condition lasted for 60 minutes. First, CO 2 The exposure level of the concentration is adjusted to a specific concentration. Then, the test subjects adapt to the indoor environment within 20 minutes. After exposure, the test subjects are tested within 30 minutes to evaluate human cognitive performance and neural responses. After the test is completed, the test subjects' physiological parameters are continuously measured for 5 minutes. Finally, the test subjects need to complete Questionnaire II for about 5 minutes to evaluate their self-test efforts and acute health symptoms. During the 2.5-hour experiment, the test subjects were not allowed to leave the environmental chamber. At the same time, all measurements were carried out indoors.

[0060] 5. Data processing

[0061] Cognitive task performance. Accuracy (ACC) and reaction time (RT) were used as measurement indicators, which are important indicators of cognitive performance. The 36 data for each person, condition, and task were averaged, and extreme data were excluded.

[0062] Neural responses of EEG were used. First, direct current (DC) offset was removed by applying a 0.5-40 Hz bandpass filter to the raw data collected from each channel. Secondly, concave filtering was used to remove 50 Hz power frequency interference. In addition, noise in the EEG was removed by independent component analysis (ICA). Then, for each task, the range of the slices was from -1s to 2s, and the baseline correction was from -1 to 0. Trials with positive and negative amplitudes greater than 100 were deleted. Subsequently, the frequency bands were divided into δ (0.5-4 Hz), θ (4-8 Hz), α (8-13 Hz), or β (14-27 Hz) according to the fast Fourier transform (FFT). Finally, the power spectral density (PSD) of each frequency band was calculated. Specifically, the transformation formula of the time domain signal is as follows:

[0063]

[0064] The power spectral density is calculated as follows:

[0065]

[0066] Where s(t) is the function of the time domain signal; T is time; is a discrete signal in time domain.

[0067] The cognitive comfort model was constructed using the sliced ​​time domain signals, with 80% of the data used as the training set and 20% as the test set. Cross-validation and grid search were used to adjust the parameters of the SVM and random forest models. By experimenting with different parameter values, the best cross-validation accuracy was selected as the final parameter of the training model.

[0068] 6. Statistical analysis

[0069] The results were analyzed using SPSS 22.0 software (SPSS, Chicago, USA). Two-way repeated measures ANOVA was used to test CO 2 level, temperature level and CO 2 The interaction between the main effect and the temperature level had an impact on the data. For significant results from the interaction ANOVA, simple effects analysis was used to compare differences between different factor levels. When the main effect was significant, multiple testing correction (Bonferroni test) was used to compare differences between different factor levels.

[0070] To measure how test takers responded to different types of tasks, this study analyzed the differences between the two different conditions for each task using paired t-tests. The results for the congruent and incongruent trials, including the means and standard deviations, t-scores, p-values, and effect sizes d, were then compared for both retention and manipulation trials. Results were considered statistically significant when the p-value was equal to or lower than 0.05.

[0071] In an example, based on the above test method, specific test results are given, including the following parts:

[0072] 1. Physical parameters

[0073] Table 2 shows the measured CO 2 concentration and temperature. The expected level was reached, the temperature deviation did not exceed 0.5℃, CO 2 Concentrations fluctuated within a range not far from expected levels.

[0074] Table 2 Measurement conditions during exposure in the test chamber (mean ± SD)

[0075]

[0076] 2. Cognitive performance

[0077] Figure 3 Repeated measures ANOVA showed that the 2 Results of the Stroop and DMS tasks under the exposure conditions of temperature, where the blue line indicates congruent trials and the red line indicates incongruent trials in the Stroop task; the blue line indicates retention trials and the red line indicates manipulation trials in the DMS task. For the Stroop task, there were no significant differences in accuracy and reaction time between the exposure conditions in both congruent and incongruent trials (all p values ​​> 0.138). For the DMS task, the analysis of manipulation trials showed that accuracy was marginally significant under the interaction (F(1,23) = 3.34, p = 0.081), while reaction time was marginally significant under the effect of temperature (F(1,23) = 3.20, p = 0.087), but not in retention trials (all p values ​​> 0.181).

[0078] The effects of performance in the Stroop and DMS tasks were analyzed using paired t tests to compare performance on congruent and incongruent trials (accuracy: t(95) = 6.65, p < 0.001, d = 0.68, reaction time: t(95) = 11.48, p < 0.001, d = 1.17), as well as on hold and manipulation trials (accuracy: t(95) = 21.90, p < 0.001, d = 2.23 reaction time: t(95) = 9.80, p < 0.001, d = 1.00). Taken together, these results indicate that congruent trials consistently resulted in faster reaction times and higher accuracy than incongruent trials, and hold trials consistently resulted in faster reaction times and higher accuracy than manipulation trials, regardless of exposure condition.

[0079] 3. Subjective reaction

[0080] Table 3 shows the results of the subjective assessment. There was a significant main effect of temperature on nasal dryness (F(1,23)=5.90, p=0.023), with lower temperatures associated with less dryness. 2 There was no significant interaction between level and temperature. Other sick building syndrome symptoms showed no significant effects of CO2 or temperature during exposure (all p>0.096). 2 There were significant main effects of air quality on air quality (F(1,23)=10.90, p=0.003) and temperature (F(1,23)=15.58, p=0.001), with lower CO 2 The air quality was perceived as fresher with increasing CO concentration and temperature. The acceptability of air quality was marginal with a main effect of temperature (F(1,23)=4.18, p=0.053). For noise, there was a significant main effect with an interaction (F(1,23)=4.39, p=0.047). Simple effects analysis showed that at low temperatures, CO 2 The higher the concentration, the more acceptable the noise (p = 0.057); 2 The lower the temperature, the more acceptable the noise (p = 0.056), but neither reached a significant level. This may be because low CO 2Concentrations required ventilation, and external sounds interfered with participants' perception of noise. For odor, there was a main effect of CO2 (F(1,23)=8.10, p=0.009), with higher concentrations being unacceptable. CO2 (F(1,23)=12.55, p=0.002) and temperature (F(1,23)=32.30, p<0.001) also affected thermal sensation, with participants feeling warmer at higher levels. Thermal comfort showed a significant interaction effect (F(1,23)=8.42, p=0.008); higher CO2 concentrations at low temperatures made participants more comfortable (p=0.007). For temperature acceptability, a significant interaction effect was found (F(1,23)=5.24, p=0.032). At high temperatures, lower CO 2 level made the temperature more acceptable (p = 0.036); 2 When the temperature was lower, the subjects were more likely to accept it (p=0.012). There was no significant effect on lighting (all p>0.103).

[0081] Exposure condition did not significantly affect task difficulty, effort, perceived stress, or self-perceived performance (all p>0.183).

[0082] Table 3 Subjective evaluation of acute health symptoms under different exposure conditions

[0083]

[0084]

[0085]

[0086] 4. Physiological parameters

[0087] A series of physiological parameters were measured during the experiment. The results of repeated measures analysis of variance showed that temperature had a significant effect on body temperature and heart rate. The results of blood pressure showed that under different CO 2 and temperature exposure conditions. CO 2 There was no significant interaction between CO and temperature, but there was a slight significant effect on cortisol concentration. For body temperature, there was a main effect of temperature (F(1,23)=40.75, p<0.001). 2 In the case of CO, the body temperature under high temperature was significantly higher than that under low temperature. For heart rate, temperature had a main effect (F(1,23)=8.26, p=0.009). The heart rate under high temperature was significantly higher than that under low temperature, regardless of CO. 2 . Blood pressure is not affected by carbon dioxide and temperature and remains stable over time ( Figure 4A). The results showed that under the interaction, cortisol concentration was marginally significant (F(1,23)=3.57, p=0.072). Figure 4 B). In addition, no significant differences were found between the pre-test and post-test (all p-values ​​were greater than 0.146).

[0088] 5. Electroencephalogram

[0089] Event-related potentials (ERPs) are non-invasive measures of human brain activity reflecting a range of sensory, cognitive, emotional, and motor processes. Figure 5 Figure ERP waveforms for four different context conditions during the STROOP and DMS tasks. Mean values ​​of scalp potentials measured at electrodes FZ, CZ, and PZ during the four exposure conditions are given. Stroop task. The orange line indicates congruent trials and the blue line indicates incongruent trials. Destination Management System Task. The orange line indicates retention trials and the blue line indicates manipulation trials. A two-way repeated measures ANOVA on the mean amplitudes in the 0 to 1 second interval did not reveal any significant differences. Figure 5 A focuses on the ERP differences between the incongruent and congruent conditions, which in part quantifies the cognitive effort required in the incongruent trials. Differences were observed in the 0.3-0.6 s interval. Therefore, mean amplitude comparisons were performed in this interval, revealing marginal main effects of temperature for electrodes FZ (F(1,23)=3.33, p=0.081) and CZ (F(1,23)=2.68, p=0.061). Figure 5 B illustrates the difference in ERP waveforms between the hold and manipulation conditions in the DMS task, thereby quantifying the effect of memory processes under the manipulation condition. No significant differences were found in the 0.3–0.6 s interval.

[0090] In the Stroop and DMS tasks, the time-frequency plots of the four exposure conditions are shown in Figure 2. Figure 5As shown. The amplitudes of the four spectral bands (δ, θ, α, and β) in the FZ channel under different exposure conditions were compared. For the Stroop task, significant temperature main effects were found only in the θ (F(1,23)=5.12, p=0.033) and β bands (F(1,23)=9.98, p=0.004) of incongruent trials. For the DMS task, a significant α wave central main effect was found in the 0-1s manipulation trials (F(1,23)=4.55, p=.044), and a significant α wave interaction was found in the 1-2s manipulation trials (F(1,23)=6.84, p=.015). Under the manipulation condition of the DMS task, the theta power spectral density was continuously elevated in the 1-2s time interval. This suggests that ongoing cognitive processing involves updating or manipulating working memory content, reflecting the dynamic demands for cognitive control and manipulation strategies in this stage of the task. Simple effect analysis showed that under low temperature, CO 2 The lower the concentration, the higher the amplitude of α (p = 0.003); 2 The amplitude of α was higher at lower temperatures (p = .016) under different concentrations. Significant main effects of concentration in the β band were found in both the 0-1 s (F(1,23) = 5.99, p = 0.022) and 1-2 s (F(1,23) = 7.44, p = 0.012) manipulation trials.

[0091] Topographic maps of the three spectral bands (theta, alpha, and beta) in the four exposure conditions during the Stroop and DMS tasks are shown in Figure 2. Figure 2 A. The power spectral density was compared between the different exposure conditions, and the significance of the FZ channel was marked in the topographic map. For the Stroop task, the FZ channel was significantly increased in the beta band for congruent trials (F(1,23)=4.02, p=0.057) and in the alpha and beta bands for incongruent trials (α: F(1,23)=3.57, p=0.071β: F(1,23)=3.44, p=0.076). In the beta band, the CO was significantly increased in congruent trials (F(1,23)=4.83, p=0.038) and incongruent trials (F(1,23)=4.43, p=0.047). 2 The main effect of CO 2 The lower the concentration, the higher the PSD in the beta band. The other frequency bands were not affected by the exposure conditions (all p values ​​> 0.135). For the DMS task, the interaction and main effects were not significant in each frequency band (all p values ​​were > 0.128).

[0092] To determine whether these effects were scalp region-dependent, the relative power of four frequency bands (alpha, beta, theta, and delta) was calculated for five scalp regions (frontal, central, parietal, occipital, and temporal). During the task test, the relative frequency band power was calculated for the frontal (FP1, FP2, F3, FZ, F4), central (C3, CZ, C4), parietal (P7, P3, PZ, P4, P8), occipital (O1, OZ, O2), and temporal (T7, T8) electrodes for each EEG recording. Figure 6 The significance of the interaction of relative optical power in five scalp regions under different exposure conditions is shown. The black line indicates the frontal lobe, the red line indicates the central lobe, the blue line indicates the parietal lobe, the pink line indicates the occipital lobe, and the green line indicates the temporal lobe. Figure 2 As shown in A and 2B, the relative power changes in the α and β bands were more pronounced than those in other frequency bands. 2 The interaction between α and temperature was not particularly sensitive. In the alpha band, the differences were significant in the central (incongruent: F(1,23)=4.17, p=0.053), parietal (congruent: F(1,23)=3.60, p=0.070 incongruent: F(1,23)=6.17, p=0.021), and occipital (congruent: F(1,23)=3.45, p=0.076 incongruent: F(1,23)=4.88, p=0.037). In the theta and delta bands, only the temporal lobe showed significant differences in the incongruent trials (theta: F(1,23)=4.89, p=0.037; delta: F(1,23)=5.43, p=0.029). Figure 6 As shown in C and 6D, no interaction was found to be affected by exposure conditions (all p values ​​> 0.083).

[0093] Figure 6 B shows the significant main effects of relative power in the five scalp regions under different exposure conditions. For the Stroop task, the main effects were reflected in the beta band (central: F(1,23) = 8.24, p = 0.009 parietal: F(1,23) = 6.01, p = 0.022 occipital: F(1,23) = 3.46, p = 0.076 temporal: F(1,23) = 6.57, p = 0.017) and beta (central: F(1,23) = 10.77, p = 0.003 parietal: F(1,23) = 6.42, p = 0.019 temporal: F(1,23) = 7.49, p = 0.012) and delta band (parietal: F(1,23) = 3.49, p = 0.075), as shown in Table 1. Figure 3 B. Among them, only the β band of the occipital lobe and the δ band of the parietal lobe were related to the main effect of temperature, while the rest were related to CO 2 The main effect of concentration is related to Figure 6As shown in B, the DMS task was significantly affected by the main effect. For the retention test, the main effect was reflected in the alpha band (occipital band: F(1,23) = 3.26, p = 0.084), beta band (central band: F(1,23) = 3.42, p = 0.078 parietal: F(1,23) = 3.18, p = 0.088 occipital: F(1,23) = 3.79, p = 0.064), theta band (occipital: F(1,23) = 3.15, p = 0.089), and delta band (central: F(1,23) = 4.26, p = 0.051 parietal: F(1,23) = 3.69, p = 0.067 time: F(1,23) = 4.29, p = 0.050). Among them, only the delta band of the central and parietal lobes were related to the main effect of temperature, while the rest were related to CO 2 For the manipulation test, the main effect was found in the alpha band (occipital: F(1,23)=3.17, p=0.088), beta band (frontal: F(1,23)=4.02, p=0.057 occipital: F(1,23)=4.02, p=0.057), and delta band (frontal: F(1,23)=7.89, p=0.010 central: F(1,23)=3.88, p=0.061 parietal: F(1,23)=4.62, p=0.042 temporal: F(1,23)=3.56, p=0.072). Of these, only the occipital alpha band was associated with CO 2 The main effect of concentration was related to the main effect of temperature. No other main effects were found to be affected by exposure conditions (all p values ​​> 0.105).

[0094] 6. Cognitive comfort model. In this invention, the effects of four environmental conditions on EEG activity were investigated. Using three machine learning models (EEG-TCNet64, SVM, and RandomForest), we classified the EEG data to identify the effects of these conditions and cognitive states. Figure 7 The classification results are shown. Figure 7 A shows the classification accuracy for each subject, indicating the significant impact of different environments on EEG activity. The high accuracy across subjects suggests that environmental changes lead to clear and measurable differences in the EEG signal. Figure 7 B shows the overall classification accuracy when data from all subjects were combined, demonstrating that environmental conditions consistently affected EEG responses at the group level. Figure 7 C provides a cumulative confusion matrix that highlights the correct and incorrect classifications in various conditions. The matrix shows that while most classifications are correct, some confusion occurs under similar conditions. For example, under the DMS task condition, CO 2 The effects of concentration are mixed. The effects of temperature and CO2 on the STROOP task are clearly categorized.

[0095] 7. Conclusion

[0096] This study sought to investigate the effects of the interaction of CO2 concentration and temperature on human cognitive performance and neural responses. No significant interaction effects of exposure conditions were found on behavioral performance. Increased CO2 concentrations caused participants to perceive the air and nose as drier. Increased temperature caused participants to perceive more thermal discomfort. Drier air and noses led to a significant increase in body temperature and heart rate. At the neural level, CO 2 The interaction of concentration and temperature had a greater adverse effect on the Stroop task, while temperature had a greater adverse effect on the DMS task. Finally, the confusion matrix of the cognitive comfort model reached a consistent conclusion and provided a basis for determining the cognitive comfort level of the participants. In summary, cognitive tasks of different control types have different sensitivities to different components of the environment. Exposure to high CO 2 Adverse conditions of concentration and temperature can affect physiological indicators and comfort, which in turn affect cognitive performance and neural responses. Therefore, it is reasonable to construct a cognitive comfort model based on neural responses.

[0097] The above specific implementation methods are detailed descriptions of the present invention. It cannot be determined that the specific implementation methods of the present invention are limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions and substitutions can be made without departing from the concept of the present invention, which should be regarded as belonging to the protection scope of the present invention.

Claims

1. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition, characterized in that: include: Screening of testers to participate in the experiment; Exposure environment construction: prepare the environmental chamber and set the experimental conditions, including: setting different CO2 concentration levels and setting different temperature levels; Measurement of cognitive performance: Different types of cognitive tasks were performed in the exposure environment and subjective responses, physiological parameters, and EEG were measured; Measurement data analysis: Process the measurement data under different types of cognitive tasks and compare the sensitivity of different types of cognitive tasks to different components in the environment.

2. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: The testers who are screened to participate in the experiment include: 24 healthy test subjects were selected, including 12 females and 12 males, aged 18-31 years old, with a body mass index of 18.3-28.8 kg / m 2 .

3. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: A cylinder is used in the environmental chamber to provide CO2 to the breathing zone of the test subject to maintain a uniform CO2 concentration field, and the CO2 concentration and temperature are determined by a non-dispersive infrared carbon dioxide sensor.

4. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: The different CO2 concentration levels include 500 ppm and 5000 ppm; the different temperature levels include 22°C and 28°C.

5. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: The different types of cognitive tasks include reactive control tasks and proactive control tasks.

6. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: The measurement of the subjective response includes: collecting subjective answers through an electronic questionnaire and scaling the thermal sensation under different conditions; the physiological parameters include blood pressure, body temperature, heart rate and cortisol.

7. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: The EEG measurements include: During the task, EEG activity was recorded continuously from 32 electrodes at a sampling rate of 512 Hz using a Grael 45ch amplifier compatible with the Green Panel and Curry 9 software.

8. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 1, characterized in that: The processing of measurement data under different types of cognitive tasks includes: The acquired raw data were filtered; the noise in the EEG was removed by independent component analysis; for each task, the slice range was from -1s to 2s, the baseline correction was from -1 to 0, and the trials with positive and negative amplitudes greater than 100 were deleted; the frequency bands were divided into δ (0.5-4Hz), θ (4-8Hz), α (8-13Hz) and β (14-27Hz) according to fast Fourier transform, and the power spectral density of each frequency band was calculated.

9. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 8, characterized in that: Also includes: The cognitive comfort model is constructed using the sliced ​​time domain signals.

10. A method for testing the effect of the interaction between indoor CO2 concentration and temperature on human cognition according to claim 9, characterized in that: The cognitive comfort model includes EEG-TCNet64, SVM and RandomForest.

Images