Personalized sleep quality adjusting method and device based on big data analysis

Abstract

The invention discloses a personalized sleep quality adjusting method and device based on big data analysis, and relates to the technical field of intelligent control, and the method comprises the steps: monitoring a sleep user in real time according to a multi-modal monitoring sensor, and obtaining physiological feature, environment feature and behavior feature data; loading the expected sleep duration of the user, fitting sleep quality expectation in combination with big data equipment and a time zone division granularity constraint table, and establishing a sleep quality expectation curve; performing real-time sleep quality evaluation based on the feature data to obtain a real-time sleep quality coefficient; performing negative variation depth evaluation by comparing the expected curve to generate sleep quality negative variation depth; according to the negative variation depth and the feature data, an intervention strategy is decided through a positive intervention channel; and the sleep quality is adjusted according to the intervention strategy, so that the technical problem that the sleep quality adjustment is lack of individuation and real-time performance due to the fact that monitoring and analysis cannot be performed according to the characteristics of different user individuals in the existing sleep quality adjustment technology is solved.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology, and particularly to a personalized sleep quality adjustment method and device based on big data analysis. Background Art

[0002] With the acceleration of the social rhythm and the increase of life pressure, the problem of sleep quality has become an important factor affecting people's health. Research shows that long-term poor sleep quality may lead to health problems such as decreased immune function, metabolic disorders, and cardiovascular diseases. Existing sleep monitoring devices and adjustment methods usually rely on a single type of sensor, for example, by monitoring physiological characteristic data such as heart rate, respiratory rate, or brain waves to evaluate the sleep state. However, these technologies have obvious deficiencies in data collection dimensions, personalized adaptability, and real-time adjustment capabilities, which limit their effectiveness in improving users' sleep quality. Current technologies mostly rely on physiological characteristic data, while ignoring the influence of environmental characteristics (such as noise, light) and behavioral characteristics (such as turning frequency) on the sleep state, resulting in the monitoring results being difficult to comprehensively reflect the actual sleep situation of users. At the same time, due to the differences in physiology, environment, and behavior among individuals, fixed sleep adjustment schemes are difficult to meet the personalized needs of users. There are significant differences in the sleep duration and sleep stage requirements among different users, but existing methods usually cannot dynamically adjust strategies to adapt to these differences. Summary of the Invention

[0003] This application provides a personalized sleep quality adjustment method and device based on big data analysis, which solves the technical problem that existing sleep quality adjustment technologies cannot monitor and analyze the characteristics of different user individuals, resulting in the lack of personalization and real-time nature in sleep quality adjustment. It realizes the technical effect of improving the accuracy and adaptability of sleep quality adjustment by dynamically fitting the sleep quality expectation curve and performing real-time intervention according to the negative variation depth.

[0004] The present application provides a personalized sleep quality regulation method for big data analysis. The method includes: based on multimodal monitoring sensors, monitoring a sleeping user in real time to obtain the user's physiological characteristic data, user environmental characteristic data, and user behavior characteristic data; loading the expected sleep duration of the sleeping user, and combining big data collection devices and a time zone division granularity constraint table to perform sleep quality expectation fitting on the sleeping user to establish a sleep quality expectation curve; performing real-time sleep quality evaluation based on the user's physiological characteristic data, user environmental characteristic data, and user behavior characteristic data to obtain a real-time sleep quality coefficient; performing negative variation depth evaluation on the real-time sleep quality coefficient according to the sleep quality expectation curve to generate a sleep quality negative variation depth; based on the sleep quality negative variation depth, the user's physiological characteristic data, user environmental characteristic data, and user behavior characteristic data, making a positive intervention decision for the sleeping user according to the sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy; and regulating the sleep quality of the sleeping user according to the sleep quality positive intervention strategy.

[0005] The present application also provides a personalized sleep quality regulation device for big data analysis, including: a real-time monitoring module: based on multimodal monitoring sensors, monitoring a sleeping user in real time to obtain the user's physiological characteristic data, user environmental characteristic data, and user behavior characteristic data; a sleep quality expectation fitting module: loading the expected sleep duration of the sleeping user, and combining big data collection devices and a time zone division granularity constraint table to perform sleep quality expectation fitting on the sleeping user to establish a sleep quality expectation curve; a real-time sleep quality evaluation module: performing real-time sleep quality evaluation based on the user's physiological characteristic data, user environmental characteristic data, and user behavior characteristic data to obtain a real-time sleep quality coefficient; a negative variation depth evaluation module: performing negative variation depth evaluation on the real-time sleep quality coefficient according to the sleep quality expectation curve to generate a sleep quality negative variation depth; a positive intervention decision module: based on the sleep quality negative variation depth, the user's physiological characteristic data, user environmental characteristic data, and user behavior characteristic data, making a positive intervention decision for the sleeping user according to the sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy; and a sleep quality regulation module: regulating the sleep quality of the sleeping user according to the sleep quality positive intervention strategy.

[0006] A personalized sleep quality regulation method and device based on big data analysis proposed in this application monitors sleep users in real time according to multi-modal monitoring sensors to obtain user physiological characteristic data, user environmental characteristic data, and user behavior characteristic data; loads the expected sleep duration of the sleep user, combines big data collection devices and a time zone division granularity constraint table to perform sleep quality expectation fitting on the sleep user, and establishes a sleep quality expectation curve; performs real-time sleep quality evaluation based on the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data to obtain a real-time sleep quality coefficient; performs negative variation depth evaluation on the real-time sleep quality coefficient according to the sleep quality expectation curve to generate a sleep quality negative variation depth; based on the sleep quality negative variation depth, the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data, makes a positive intervention decision on the sleep user according to a sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy; and adjusts the sleep quality of the sleep user according to the sleep quality positive intervention strategy. This solves the technical problem in the existing sleep quality regulation technology that it is impossible to monitor and analyze the characteristics of different user individuals, resulting in the lack of personalization and real-time nature in sleep quality regulation, and achieves the technical effect of improving the accuracy and adaptability of sleep quality regulation by dynamically fitting the sleep quality expectation curve multiple times and performing real-time intervention according to the negative variation depth. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments of the present disclosure will be briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the devices according to the embodiments of the present application. It should be understood that the operations before or below do not necessarily need to be executed precisely in sequence. On the contrary, according to needs, they can be executed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or several operations can be removed from these processes.

[0008] Figure 1 Schematic flowchart of a personalized sleep quality regulation method based on big data analysis provided by an embodiment of the present application; Figure 2 Schematic structural diagram of a personalized sleep quality regulation device based on big data analysis provided by an embodiment of the present application.

[0009] Description of reference numerals: real-time monitoring module 1, sleep quality expectation fitting module 2, real-time sleep quality evaluation module 3, negative variation depth evaluation module 4, positive intervention decision module 5, sleep quality regulation module 6. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0010] The above description is only an overview of the technical solution of this application. In order to understand the technical means of this application more clearly, it can be implemented in accordance with the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the specific embodiments of this application are given below.

[0011] In order to make the purpose, technical solution and advantages of this application clearer, the following will further describe this application in detail with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

[0012] In the following description, "some embodiments" are involved, which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. The terms "first / second" involved are only used to distinguish similar objects and do not represent a specific order for the objects. The terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or server including a series of steps or units does not necessarily limit to those steps or units clearly listed, but may include other steps or modules not clearly listed or inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of this application. The terms used herein are only for the purpose of describing the embodiments of this application.

[0013] The embodiments of this application provide a personalized sleep quality adjustment method for big data analysis, as Figure 1 shown, the method includes: According to multimodal monitoring sensors, the sleep user is monitored in real time to obtain user physiological characteristic data, user environmental characteristic data and user behavior characteristic data.

[0014] In the embodiments of the present application, a multi-modal monitoring sensor is used to monitor the sleeping user in real time to comprehensively obtain the user's physiological characteristic data, user environment characteristic data, and user behavior characteristic data. Among them, the multi-modal monitoring sensor includes a heart rate sensor, a respiration monitor, an environment sensor (such as a temperature sensor, a humidity sensor, etc.), a pressure distribution sensor, etc.; in terms of the user's physiological characteristic data, a heart rate sensor, a respiration monitor, an electroencephalogram acquisition device, etc. are used to record the user's heart rate changes, respiration frequency, electroencephalogram characteristics, etc. in real time. These data can reflect the user's sleep depth, sleep stage distribution, and physiological state changes, thereby providing a reliable basis for sleep quality evaluation; in terms of the user's environment characteristic data, parameters such as light intensity, noise level, temperature and humidity, and air quality in the user's sleep environment are collected through the environment sensor. These environmental factors have a direct impact on the user's sleep state. Precise environmental monitoring can provide an important reference for the formulation of adjustment plans; in terms of the user's behavior characteristic data, a mattress pressure distribution sensor is used to record the user's turning times, turning directions, and sleep posture changes (such as supine, side lying, etc.), and the movement amplitude of the user's limbs is monitored through a motion sensor, including micro-movements, static states, and exercise intensity. These behavior data can comprehensively reflect the user's dynamic activities during sleep and provide key support for real-time sleep quality evaluation and adjustment. By comprehensively collecting the above three types of data through the multi-modal sensor, not only can the user's sleep state be accurately evaluated, but also comprehensive data support can be provided for the formulation of personalized sleep adjustment strategies.

[0015] Load the expected sleep duration of the sleeping user, and perform sleep quality expectation fitting on the sleeping user in combination with the big data acquisition device and the time zone division granularity constraint table to establish a sleep quality expectation curve.

[0016] In one embodiment, based on the user's expected sleep duration, the big data acquisition device and the time zone division granularity constraint table are used to perform expectation fitting on the user's sleep quality to establish a personalized sleep quality expectation curve. Specifically, first, according to the expected sleep duration set by the user, analyze the historical sleep data similar to its characteristics, obtain the sleep quality sample distribution information through the big data acquisition device, and then extract the concentrated samples that can represent the user's expected sleep quality according to the central tendency of these data samples; subsequently, by matching with the time zone division granularity constraint table, refine and analyze the characteristics of the time zone where the user is located and its sleep time period, and divide the expected sleep duration into multiple sleep windows; then, based on these sleep windows, combine the user's expected value and the sample data to fit and generate a curve that can dynamically reflect the change of the user's expected sleep quality, forming the user's sleep quality expectation curve, providing a benchmark reference for subsequent sleep quality evaluation and adjustment.

[0017] Furthermore, the present application provides for loading the desired sleep duration of the sleeping user, and performing sleep quality expectation fitting on the sleeping user by combining a big data collection device and a time zone division granularity constraint table to establish a sleep quality expectation curve, including: Based on the desired sleep duration, according to the big data collection device, collect the historical sleep quality coefficients of the same characteristics of the sleeping user to obtain a sleep quality sample distribution; calculate the central value according to the sleep quality sample distribution to obtain a sleep quality central sample; input the sleep quality central sample into the time zone division granularity constraint table to obtain the time zone division granularity; divide the desired sleep duration according to the time zone division granularity to obtain multiple desired sleep windows; based on the big data collection device, perform sleep quality expectation fitting on the sleeping user according to the multiple desired sleep windows to generate the sleep quality expectation curve.

[0018] Preferably, according to the desired sleep duration set by the user, use the big data collection device to collect historical sleep data of users with similar characteristics to this user, extract the sleep quality coefficients of these data to form a sleep quality sample distribution, and these similar characteristics include age, gender, desired sleep duration, etc.; subsequently, through statistical analysis of the sleep quality sample distribution, calculate a representative central value. For samples with a relatively symmetric distribution, use the arithmetic mean to calculate this central value. For sample distributions with more extreme values, use the median to calculate this central value to reduce the influence of extreme values; then, input the calculated central value as the sleep quality central sample into the time zone division granularity constraint table to match a suitable time zone division granularity. This time zone division granularity constraint table records multiple sleep quality coefficients and the corresponding time zone division granularity, which is set through historical experience and expert suggestions. This time zone division granularity is used to divide the desired sleep duration of the sleeping user. For example, if the time zone division granularity is 10 minutes, the entire desired sleep duration will be divided into several 10-minute desired sleep windows. Such a time zone division granularity can be adjusted according to the actual needs of the user to more precisely adapt to individual differences; then, according to the divided desired sleep windows, use the big data collection device to obtain the historical sleep data of users with the same characteristics as the sleeping user, and fit the sleep quality characteristics in each time window to generate a dynamic sleep quality expectation curve. This curve can not only reflect the ideal state of the sleeping user during the entire sleep cycle, but also be refined to each time window to show the expected change trend of sleep quality in different time periods; finally, this sleep quality expectation curve provides an accurate benchmark reference for subsequent real-time monitoring, sleep quality evaluation, and the formulation of intervention strategies, thereby realizing comprehensive support for the personalized sleep needs of users.

[0019] Further, the present application provides a method for fitting the expected sleep quality of a sleeping user based on the big data collection device according to the multiple expected sleep windows to generate the expected sleep quality curve, including: Retrieving users with the same characteristics as the sleeping user based on the big data collection device to determine multiple users with the same characteristics; collecting normal sleep quality samples for the multiple users with the same characteristics based on the multiple expected sleep windows and the big data collection device to obtain the normal sleep quality sample distributions of multiple windows; calculating the central values according to the normal sleep quality sample distributions of multiple windows to obtain the expected sleep quality coefficients of multiple windows; constructing the expected sleep quality curve based on the multiple expected sleep windows and the multiple expected sleep quality coefficients.

[0020] Optionally, using the big data collection device, retrieving a historical user group with similar characteristics to the target user from the database, and screening out multiple users with the same characteristics; subsequently, using the multiple expected sleep windows of the sleeping user as the time benchmark for data collection, screening out the historical records marked as normal sleep among the determined multiple users with the same characteristics, and extracting the sleep quality coefficients of these records as the normal historical sleep quality sample data; then, according to the multiple expected sleep windows, classifying the collected normal historical sleep quality coefficients by time window to form the normal sleep quality sample distribution of each time window. For example, within each 10-minute window, collect the corresponding historical sleep quality coefficients to form the sample set of this window; then, perform the same central value calculation as described above for the normal sleep quality sample distribution of each window to generate the expected sleep quality coefficient of each time window, and these coefficients reflect the sleep quality characteristics of the user group with the same characteristics within this time window; then, based on the multiple expected sleep windows and their corresponding expected sleep quality coefficients, fit these data points into a continuous expected sleep quality curve, which can reflect the ideal sleep quality change trend of the sleeping user during the entire sleep cycle. For example, the high-quality stage during deep sleep and the low-quality stage during light sleep. Through the above process, the constructed expected sleep quality curve can provide an accurate benchmark for the sleep quality evaluation and intervention strategy of the sleeping user, making sleep regulation more personalized and scientific.

[0021] Performing real-time sleep quality evaluation according to the user's physiological characteristic data, the user's environmental characteristic data, and the user's behavioral characteristic data to obtain a real-time sleep quality coefficient.

[0022] In one embodiment, the collected user physiological characteristic data, user environmental characteristic data, and user behavior characteristic data are input into a pre-trained sleep quality evaluator. The evaluator performs real-time evaluation of the user's sleep quality based on this multi-dimensional data, generates a quantified real-time sleep quality coefficient, and the real-time sleep quality coefficient is dynamically updated as the monitoring data changes, which can reflect whether the current sleep state of the sleeping user deviates from the expectation or is abnormal, provides a real-time and quantified evaluation basis for the sleep state of the sleeping user, helps to detect problems in a timely manner, and provides basic support for subsequent sleep quality adjustment and intervention.

[0023] Furthermore, the present application provides a method for performing real-time sleep quality evaluation according to the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data, and obtaining a real-time sleep quality coefficient, including: Performing sleep quality evaluation record learning based on P sleep quality evaluation learning models to construct P sleep quality evaluators that meet the sleep quality evaluation accuracy constraint, where P is a positive integer greater than 1; inputting the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data into the P sleep quality evaluators to obtain P sleep quality evaluation coefficients; calculating the proportion according to the P sleep quality evaluation accuracies corresponding to the P sleep quality evaluators to obtain P evaluation accuracy incentive coefficients; and performing weighted calculation on the P sleep quality evaluation coefficients according to the P evaluation accuracy incentive coefficients to generate the real-time sleep quality coefficient.

[0024] Optionally, based on P sleep quality evaluation learning models, P sleep quality evaluators are constructed by training and learning from sleep quality evaluation records to evaluate the user's real-time sleep state from different dimensions, where P is a positive integer greater than 1; these sleep quality evaluation learning models can be constructed based on algorithms such as neural networks, decision trees, support vector machines, and random forests. The combination method can be the joint use of multiple algorithms or independent construction based on a single algorithm. For example, taking the neural network as an example, randomly sampling with replacement from the sleep quality evaluation records to construct P training sets, P validation sets, and P test sets. Each sample set contains the user's physiological characteristics, environmental characteristics, behavioral characteristics, and the corresponding sample sleep quality evaluation coefficients; for any one sleep quality evaluation learning model, based on the neural network for model design, including the input layer, hidden layer, output layer, etc., through the forward propagation method, the sample sleep data is transmitted from the input layer to the output layer, and after weighted summation and activation function processing, the predicted sleep quality evaluation coefficient is output; calculate the loss between the predicted value and the sample sleep quality evaluation coefficient through the mean square error (MSE), and use an optimizer (such as Adam) to update the network weights and biases to reduce the loss function value; the model is iterated multiple times on the training set until the loss value converges or reaches the predetermined number of training times; after each training batch, use the validation set to evaluate the model performance and determine whether there is an overfitting phenomenon; if the validation set loss does not decrease for several consecutive times, stop training; after training, use the test set to verify the generalization ability of the model, and evaluate the model effect through the accuracy. If the accuracy does not meet the requirements of the sleep quality evaluation accuracy constraint, the model parameters (such as the learning rate, hidden layer structure) can be adjusted to optimize the model. When the accuracy meets the requirements, output the current sleep quality evaluation learning model as a sleep quality evaluator and save its accuracy as the sleep quality evaluation accuracy of this evaluator. Repeat the above steps until P sleep quality evaluators are constructed.

[0025] Optionally, input the user's real-time physiological characteristic data, environmental characteristic data, and behavioral characteristic data into the P sleep quality evaluators. Each evaluator analyzes and evaluates the input data to generate P sleep quality evaluation coefficients. Each evaluator has its corresponding evaluation accuracy, which is used to measure the reliability of the evaluator; according to the evaluation accuracies of the evaluators, calculate P sleep quality evaluation incentive coefficients, indicating the influence degree of each evaluator on the final result. The calculation method is to calculate the ratio of the evaluation accuracy of each evaluator to the total sum of the evaluation accuracies of all evaluators; then, use these sleep quality evaluation incentive coefficients to perform weighted calculation on the P sleep quality evaluation coefficients to output the final real-time sleep quality coefficient, which is used to dynamically quantify the user's sleep state. This sleep quality evaluation method integrating multiple evaluators can comprehensively integrate the characteristics of multi-dimensional data, improve the accuracy and real-time performance of the evaluation, and provide a more scientific and reliable basis for subsequent sleep quality adjustment.

[0026] Perform a negative variation depth evaluation on the real-time sleep quality coefficient according to the sleep quality expectation curve to generate a negative variation depth of sleep quality.

[0027] In one embodiment, compare the real-time sleep quality coefficient with the window expected sleep quality coefficient at the corresponding time point in the sleep quality expectation curve. The sleep quality expectation curve represents the target value of the sleep quality coefficient of the sleep user at different time periods under ideal conditions, and the real-time sleep quality coefficient is the current actual monitored value; for each time point, calculate the difference between the real-time sleep quality coefficient and the corresponding window expected sleep quality coefficient. If the difference is less than 0, that is, the real-time sleep quality is lower than the expected value, it is recorded as a negative deviation. If it is greater than or equal to 0, it does not participate in the subsequent calculation; subsequently, count all the time points of negative deviation, and assign different weights according to the importance of the time point to sleep quality (such as the importance of deep sleep period is higher than that of light sleep period), and then perform a weighted sum of the absolute values of all negative deviations through the weights of each time point to obtain a weighted negative variation depth. This weighted negative variation depth will be output as the negative variation depth of sleep quality to reflect the degree of deviation between the current state and the expected state of the sleep user; as the real-time sleep quality coefficient changes, continuously perform deviation calculation and update of the negative variation depth. If the negative variation depth gradually increases, it indicates that the degree and duration of the deviation of the user's current sleep state from the ideal state are both increasing. If it gradually decreases, it indicates that the sleep state is recovering to the ideal state. This process can dynamically quantify the abnormal degree of the user's sleep state, provide a scientific basis for real-time adjustment and intervention strategies, and make the sleep quality adjustment more accurate and effective.

[0028] Based on the negative variation depth of sleep quality, the user's physiological characteristic data, the user's environmental characteristic data, and the user's behavior characteristic data, make a positive intervention decision for the sleep user according to the positive intervention channel of sleep quality to obtain a positive intervention strategy for sleep quality.

[0029] In one embodiment, the deviation degree of the user's current sleep state is evaluated based on the negative variation depth of the user's sleep quality, the user's physiological characteristic data, the environmental characteristic data, and the behavioral characteristic data. The negative variation depth of sleep quality reflects the difference between the user's sleep quality and the desired state. When this depth reaches a certain threshold, based on these multi-dimensional data inputs and in combination with the positive sleep quality intervention channel, a positive intervention decision is made. The positive sleep quality intervention channel is a comprehensive decision-making model that determines the positive sleep quality intervention strategy based on the user's physiological, environmental, and behavioral data, combined with the negative variation depth of sleep quality. This positive sleep quality intervention strategy includes adjusting environmental conditions (such as temperature, light, or noise), recommending changes in sleep posture or behavior (such as reducing the frequency of turning over or relaxing the body), and even providing physiological adjustment suggestions (such as breathing adjustment or sleep cycle synchronization) to help the user improve sleep quality and restore it to an ideal sleep state.

[0030] Furthermore, the present application provides a positive intervention decision for the sleeping user based on the negative variation depth of the sleep quality, the user's physiological characteristic data, the user's environmental characteristic data, and the user's behavioral characteristic data according to the positive sleep quality intervention channel, to obtain a positive sleep quality intervention strategy, including: The positive sleep quality intervention channel includes a positive sleep intervention gating model, a physiological characteristic sleep intervention decision model, an environmental characteristic sleep intervention decision model, and a behavioral characteristic sleep intervention decision model; determine whether the negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of sleep quality; if the negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of sleep quality, input the negative variation depth of the sleep quality into the positive sleep intervention gating model to obtain a positive sleep intervention gating coefficient; based on the positive sleep intervention gating coefficient and the user's physiological characteristic data, obtain a physiological characteristic sleep intervention decision according to the physiological characteristic sleep intervention decision model; based on the positive sleep intervention gating coefficient and the user's environmental characteristic data, obtain an environmental characteristic sleep intervention decision according to the environmental characteristic sleep intervention decision model; based on the positive sleep intervention gating coefficient and the user's behavioral characteristic data, obtain a behavioral characteristic sleep intervention decision according to the behavioral characteristic sleep intervention decision model; fuse the physiological characteristic sleep intervention decision, the environmental characteristic sleep intervention decision, and the behavioral characteristic sleep intervention decision to obtain the positive sleep quality intervention strategy.

[0031] Preferably, it is determined whether the negative variation depth of sleep quality is greater than or equal to the negative variation depth threshold of sleep quality. If the negative variation depth is greater than or equal to this threshold, it indicates that the user's sleep quality has deviated from the ideal state and intervention is required. At this time, the positive sleep quality intervention channel is activated. This positive sleep quality intervention channel includes a positive sleep intervention gating model, a physiological characteristic sleep intervention decision-making model, an environmental characteristic sleep intervention decision-making model, and a behavioral characteristic sleep intervention decision-making model. Inside the physiological characteristic sleep intervention decision-making model, the environmental characteristic sleep intervention decision-making model, and the behavioral characteristic sleep intervention decision-making model, there are multiple corresponding characteristic sleep intervention decision-making sub-models for generating corresponding characteristic sleep intervention decisions. The positive sleep intervention gating model is used to analyze the negative variation depth of sleep quality and determine the positive sleep intervention gating coefficient. These models can be constructed based on deep learning (such as neural networks, LSTM, CNN), machine learning (such as decision trees, support vector machines, random forests), etc., in combination with corresponding sample data (such as sample negative variation depth sets of sleep quality, sample positive sleep intervention gating coefficient sets, sample user physiological characteristic data, sample physiological characteristic sleep intervention decisions, etc.). The specific construction process is similar to the foregoing, and all are carried out through steps such as forward propagation, loss function calculation, backpropagation, gradient update, and parameter optimization; when the negative variation depth meets the intervention condition, the negative variation depth is input into the positive sleep intervention gating model. The role of this model is to output a positive sleep intervention gating coefficient according to the current sleep quality deviation, which is used to represent the number of activated intervention decision-making sub-models. This coefficient reflects the severity of the user's current sleep problem. The higher the gating coefficient, the more serious the problem and the more intervention decision-making sub-models are needed to intervene; after obtaining the positive sleep intervention gating coefficient, the positive sleep intervention gating coefficient is combined with the user's physiological characteristic data and transmitted to the physiological characteristic sleep intervention decision-making model, the positive sleep intervention gating coefficient is combined with the user's environmental characteristic data and transmitted to the environmental characteristic sleep intervention decision-making model, and the positive sleep intervention gating coefficient is combined with the user's behavioral characteristic data and transmitted to the behavioral characteristic sleep intervention decision-making model, so as to obtain the physiological characteristic sleep intervention decision, the environmental characteristic sleep intervention decision, and the behavioral characteristic sleep intervention decision; among them, the physiological characteristic sleep intervention decision is a decision for sleep intervention based on the user's physiological state data. For example, α-wave sleep-promoting music can be dynamically played based on heart rate and breathing data, and the frequency is adjusted to 8 - 12 Hz to promote deep sleep. When the user's breathing depth is insufficient, a vibration feedback device is enabled to guide the user to take deep breaths; the environmental characteristic sleep intervention decision is a decision for sleep intervention based on the characteristic data of the user's sleep environment. For example, when the noise intensity exceeds the threshold (such as 40 dB), an intelligent white noise device is started for active noise reduction, and the air conditioner temperature or humidifier humidity is adjusted in real time according to the temperature and humidity change trend to restore the environmental parameters to the ideal range;The behavioral feature sleep intervention decision is a decision for sleep intervention based on the behavioral data during the user's sleep. For example, when the number of body turns is excessive, the intelligent vibrating mattress is used to guide the user to adjust the sleeping position. When the user is in a poor sleeping position (such as lying on the stomach which compresses the breathing), a voice prompt is used for correction. Finally, all intervention decisions (including physiological, environmental, and behavioral intervention decisions) are integrated, and the current sleep state of the user is comprehensively considered to output a personalized positive sleep quality intervention strategy. This strategy will adjust the content and intensity of the intervention according to the actual situation and needs of the user to ensure effective regulation of the user's sleep quality and promote the recovery of the sleep state to an ideal level.;

[0032] Furthermore, the present application provides a physiological feature sleep intervention decision obtained based on the sleep positive intervention gating coefficient and the user's physiological feature data according to the physiological feature sleep intervention decision model, including: The physiological feature sleep intervention decision model includes Q physiological feature sleep intervention decision sub-models, where Q is a positive integer greater than 1. According to the sleep positive intervention gating coefficient, the Q physiological feature sleep intervention decision sub-models are feature-activated to obtain a plurality of activated sleep intervention decision sub-models. The user's physiological feature data is input into the plurality of activated sleep intervention decision sub-models to obtain a plurality of physiological feature sleep intervention plans. The plurality of physiological feature sleep intervention plans are fused to obtain the physiological feature sleep intervention decision.

[0033] Optionally, the physiological characteristic sleep intervention decision model is composed of Q physiological characteristic sleep intervention decision sub-models, where Q is a positive integer greater than 1. Each sub-model analyzes the user's physiological characteristic data (such as heart rate, respiratory rate, brain waves, etc.) from different perspectives and formulates corresponding sleep intervention strategies based on this data. The design of the sub-model can be based on different algorithms, such as neural networks, decision trees, etc.; during the decision-making process, the number of activated physiological characteristic sleep intervention decision sub-models is determined by the sleep positive intervention gating coefficient. Assuming the sleep positive intervention gating coefficient is 3, it means that 3 physiological characteristic sleep intervention decision sub-models need to be activated. The magnitude of this coefficient reflects the severity of the current user's sleep quality problem. The larger the gating coefficient, the more serious the user's sleep problem and the more intervention measures are required; the activation process is to select the corresponding number of sub-models from the Q sub-models for activation according to the magnitude of the gating coefficient. Each activated sub-model will generate an intervention plan based on the user's real-time physiological data, thus obtaining multiple physiological characteristic sleep intervention plans; then, the intervention plans generated by multiple activated physiological characteristic sleep intervention decision sub-models are fused. This fusion process is to count the measures in each plan and extract the measures with the highest frequency of occurrence to form the final physiological characteristic sleep intervention decision, ensuring that it can have the greatest positive impact on the user's physiological characteristics. The fused intervention plan can cover a variety of intervention measures, such as playing sleep-aid music, conducting deep breathing guidance, adjusting sleeping positions, etc., to help the user improve sleep quality.

[0034] Further, the present application provides a method for determining whether the negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of the sleep quality, including: If the negative variation depth of the sleep quality is less than the negative variation depth threshold of the sleep quality, based on the future sleep window, perform feature prediction on the user's physiological characteristic data, the user's environmental characteristic data, and the user's behavioral characteristic data to obtain predicted user physiological characteristic data, predicted user environmental characteristic data, and predicted user behavioral characteristic data; perform sleep quality evaluation according to the predicted user physiological characteristic data, the predicted user environmental characteristic data, and the predicted user behavioral characteristic data to obtain a predicted sleep quality coefficient; perform negative variation depth evaluation on the predicted sleep quality coefficient according to the sleep quality expectation curve to generate a predicted negative variation depth of the sleep quality; determine whether the predicted negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of the sleep quality; if the predicted negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of the sleep quality, based on the predicted negative variation depth of the sleep quality, the predicted user physiological characteristic data, the predicted user environmental characteristic data, and the predicted user behavioral characteristic data, perform sleep quality prediction and adjustment on the sleep user according to the sleep quality positive intervention channel.

[0035] Optionally, when the negative variation depth of sleep quality is less than the preset negative variation depth threshold of sleep quality, feature prediction will be performed on the user's physiological feature data, environmental feature data, and behavioral feature data; specifically, based on the historical physiological feature data, historical environmental feature data, and historical behavioral feature data, using the same training method as described above, a sleep quality prediction model will be constructed through an LSTM neural network. This model will predict various future features based on real-time monitoring data and the set future sleep window, including predicting the user's physiological feature data, predicting the user's environmental feature data, and predicting the user's behavioral feature data. This future sleep window refers to the sleep cycle that the user will enter in the future, such as 30 minutes; subsequently, based on the predicted user's physiological feature data, environmental feature data, and behavioral feature data, P sleep quality evaluation learning models will evaluate these data to calculate a comprehensive predicted sleep quality coefficient, which reflects the expected sleep quality of the user in the future for a period of time; after that, this coefficient will be compared with the sleep quality expectation curve to perform the same negative variation depth evaluation as described above to obtain the predicted negative variation depth of sleep quality and determine whether it is greater than or equal to the set negative variation depth threshold of sleep quality; if the predicted negative variation depth meets the threshold requirement, it indicates that there are relatively large problems with the predicted sleep quality. At this time, the sleep quality positive intervention channel will be activated, and the predicted negative variation depth of sleep quality, predicted user's physiological feature data, predicted user's environmental feature data, and predicted user's behavioral feature data will be input into the sleep quality positive intervention channel for sleep quality prediction adjustment to generate a sleep intervention strategy adapted to the current state of the sleeping user; the intervention strategy will be dynamically adjusted according to the real-time monitoring results to ensure that it can accurately respond to the changes of the user during sleep and achieve an ideal sleep quality.

[0036] Adjust the sleep quality of the sleeping user according to the sleep quality positive intervention strategy.

[0037] In one embodiment, after obtaining the sleep quality positive intervention strategy, corresponding adjustments will be made through the physiological feature sleep intervention decision, environmental feature sleep intervention decision, and behavioral feature sleep intervention decision in this intervention strategy, such as playing sleep music with an alpha wave frequency of 8 - 12 Hz, using a smart white noise device for noise reduction, and guiding through a smart mattress to help adjust the sleeping position and reduce the number of turns, etc., to ensure that the sleeping user maintains a high sleep quality throughout the sleep cycle and avoid a decline in sleep quality.

[0038] In the above text, reference is made to Figure 1 A personalized sleep quality adjustment method based on big data analysis according to an embodiment of the present invention is described in detail. Next, reference will be made to Figure 2 Describe a personalized sleep quality adjustment device based on big data analysis according to an embodiment of the present invention.

[0039] A personalized sleep quality regulation device for big data analysis according to an embodiment of the present invention solves the technical problems in the existing sleep quality regulation technology that it is impossible to monitor and analyze the characteristics of different user individuals, resulting in the lack of personalization and real-time nature of sleep quality regulation, and realizes the technical effect of improving the accuracy and adaptability of sleep quality regulation by multi-dynamically fitting the sleep quality expectation curve and performing real-time intervention according to the negative variation depth. A personalized sleep quality regulation device for big data analysis includes: Real-time monitoring module 1: According to multi-modal monitoring sensors, it monitors sleep users in real time to obtain user physiological characteristic data, user environmental characteristic data, and user behavior characteristic data; Sleep quality expectation fitting module 2: Loads the expected sleep duration of the sleep user, combines big data collection devices and time zone division granularity constraint tables to perform sleep quality expectation fitting on the sleep user, and establishes a sleep quality expectation curve; Real-time sleep quality evaluation module 3: Performs real-time sleep quality evaluation according to the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data to obtain a real-time sleep quality coefficient; Negative variation depth evaluation module 4: Performs negative variation depth evaluation on the real-time sleep quality coefficient according to the sleep quality expectation curve to generate a sleep quality negative variation depth; Positive intervention decision-making module 5: Based on the sleep quality negative variation depth, the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data, makes a positive intervention decision for the sleep user according to the sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy; Sleep quality regulation module 6: Regulates the sleep quality of the sleep user according to the sleep quality positive intervention strategy.

[0040] Further, the sleep quality expectation fitting module 2 further includes: Based on the expected sleep duration, according to the big data collection device, collects the same-feature historical sleep quality coefficients of the sleep user to obtain a sleep quality sample distribution; Calculates the central value according to the sleep quality sample distribution to obtain a sleep quality central sample; Inputs the sleep quality central sample into the time zone division granularity constraint table to obtain a time zone division granularity; Divides the expected sleep duration according to the time zone division granularity to obtain multiple expected sleep windows; Based on the big data collection device, performs sleep quality expectation fitting on the sleep user according to the multiple expected sleep windows to generate the sleep quality expectation curve.

[0041] Further, the sleep quality expectation fitting module 2 further includes: Retrieve users with the same characteristics for the sleeping user based on the big data collection device to determine multiple users with the same characteristics; based on the multiple expected sleep windows, collect normal sleep quality samples for the multiple users with the same characteristics according to the big data collection device to obtain the normal sleep quality sample distributions of multiple windows; calculate the central values according to the normal sleep quality sample distributions of multiple windows to obtain the expected sleep quality coefficients of multiple windows; construct the expected sleep quality curve based on the multiple expected sleep windows and the expected sleep quality coefficients of multiple windows.

[0042] Further, the real-time sleep quality evaluation module 3 further includes: Perform sleep quality evaluation record learning based on P sleep quality evaluation learning models to construct P sleep quality evaluators that meet the sleep quality evaluation accuracy constraints, where P is a positive integer greater than 1; input the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data into the P sleep quality evaluators to obtain P sleep quality evaluation coefficients; calculate the proportion according to the P sleep quality evaluation accuracies corresponding to the P sleep quality evaluators to obtain P evaluation accuracy incentive coefficients; perform weighted calculation on the P sleep quality evaluation coefficients according to the P evaluation accuracy incentive coefficients to generate the real-time sleep quality coefficient.

[0043] Further, the positive intervention decision-making module 5 further includes: The sleep quality positive intervention channel includes a sleep positive intervention gating model, a physiological characteristic sleep intervention decision-making model, an environmental characteristic sleep intervention decision-making model, and a behavior characteristic sleep intervention decision-making model; determine whether the negative variation depth of sleep quality is greater than or equal to the negative variation depth threshold of sleep quality; if the negative variation depth of sleep quality is greater than or equal to the negative variation depth threshold of sleep quality, input the negative variation depth of sleep quality into the sleep positive intervention gating model to obtain the sleep positive intervention gating coefficient; based on the sleep positive intervention gating coefficient and the user physiological characteristic data, obtain the physiological characteristic sleep intervention decision according to the physiological characteristic sleep intervention decision-making model; based on the sleep positive intervention gating coefficient and the user environmental characteristic data, obtain the environmental characteristic sleep intervention decision according to the environmental characteristic sleep intervention decision-making model; based on the sleep positive intervention gating coefficient and the user behavior characteristic data, obtain the behavior characteristic sleep intervention decision according to the behavior characteristic sleep intervention decision-making model; fuse the physiological characteristic sleep intervention decision, the environmental characteristic sleep intervention decision, and the behavior characteristic sleep intervention decision to obtain the sleep quality positive intervention strategy.

[0044] Further, the positive intervention decision-making module 5 further includes: The physiological feature sleep intervention decision model includes Q physiological feature sleep intervention decision sub-models, where Q is a positive integer greater than 1; according to the sleep positive intervention gating coefficient, the Q physiological feature sleep intervention decision sub-models are feature-activated to obtain multiple activated sleep intervention decision sub-models; the user physiological feature data is input into the multiple activated sleep intervention decision sub-models to obtain multiple physiological feature sleep intervention schemes; the multiple physiological feature sleep intervention schemes are fused to obtain the physiological feature sleep intervention decision.

[0045] Further, the positive intervention decision module 5 further includes: If the negative variation depth of the sleep quality is less than the negative variation depth threshold of the sleep quality, based on the future sleep window, feature prediction is performed on the user physiological feature data, the user environmental feature data, and the user behavior feature data to obtain predicted user physiological feature data, predicted user environmental feature data, and predicted user behavior feature data; sleep quality evaluation is performed according to the predicted user physiological feature data, the predicted user environmental feature data, and the predicted user behavior feature data to obtain a predicted sleep quality coefficient; negative variation depth evaluation is performed on the predicted sleep quality coefficient according to the sleep quality expectation curve to generate a predicted negative variation depth of the sleep quality; it is determined whether the predicted negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of the sleep quality; if the predicted negative variation depth of the sleep quality is greater than or equal to the negative variation depth threshold of the sleep quality, based on the predicted negative variation depth of the sleep quality, the predicted user physiological feature data, the predicted user environmental feature data, and the predicted user behavior feature data, sleep quality prediction adjustment is performed on the sleep user according to the sleep quality positive intervention channel.

[0046] The personalized sleep quality regulation device for big data analysis provided by the embodiments of the present invention can execute the method for personalized sleep quality regulation for big data analysis provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method.

[0047] Although the present application makes various references to certain modules in the device according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or the server. The included various units and modules are only divided according to the functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0048] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A personalized sleep quality regulation method for big data analysis, characterized in that, The method includes: Based on multi-modal monitoring sensors, the sleeping user is monitored in real time to obtain the user's physiological characteristic data, user environment characteristic data, and user behavior characteristic data; Load the expected sleep duration of the sleeping user, and combine the big data collection device and the time zone division granularity constraint table to perform sleep quality expectation fitting on the sleeping user, and establish a sleep quality expectation curve; Based on the user's physiological characteristic data, the user environment characteristic data, and the user behavior characteristic data, perform real-time sleep quality evaluation to obtain a real-time sleep quality coefficient; Based on the sleep quality expectation curve, perform negative variation depth evaluation on the real-time sleep quality coefficient to generate a sleep quality negative variation depth; Based on the sleep quality negative variation depth, the user's physiological characteristic data, the user environment characteristic data, and the user behavior characteristic data, make a positive intervention decision on the sleeping user according to the sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy; Adjust the sleep quality of the sleeping user according to the sleep quality positive intervention strategy.

2. The personalized sleep quality regulation method for big data analysis according to claim 1, characterized in that, Loading the expected sleep duration of the sleeping user, and combining the big data collection device and the time zone division granularity constraint table to perform sleep quality expectation fitting on the sleeping user, and establish a sleep quality expectation curve, includes: Based on the expected sleep duration, according to the big data collection device, collect the historical sleep quality coefficients of the same characteristics of the sleeping user to obtain a sleep quality sample distribution; Perform a central value calculation according to the sleep quality sample distribution to obtain a sleep quality central sample; Input the sleep quality central sample into the time zone division granularity constraint table to obtain the time zone division granularity; According to the time zone division granularity, divide the expected sleep duration to obtain multiple expected sleep windows; Based on the big data collection device, perform sleep quality expectation fitting on the sleeping user according to the multiple expected sleep windows to generate the sleep quality expectation curve.

3. The personalized sleep quality regulation method for big data analysis according to claim 2, characterized in that, Based on the big data collection device, perform sleep quality expectation fitting on the sleeping user according to the multiple expected sleep windows to generate the sleep quality expectation curve, includes: Based on the big data collection device, perform a same-characteristic user search on the sleeping user to determine multiple same-characteristic users; Based on the multiple expected sleep windows, according to the big data collection device, collect normal sleep quality samples of the multiple same-characteristic users to obtain multiple window normal sleep quality sample distributions; Perform a central value calculation according to the multiple window normal sleep quality sample distributions to obtain multiple window expected sleep quality coefficients; Based on the multiple expected sleep windows and the multiple window expected sleep quality coefficients, construct the sleep quality expectation curve.

4. The personalized sleep quality regulation method for big data analysis according to claim 1, characterized in that Based on the user's physiological characteristic data, the user environment characteristic data, and the user behavior characteristic data, perform real-time sleep quality evaluation to obtain a real-time sleep quality coefficient, includes: Based on P sleep quality evaluation learning models, perform sleep quality evaluation record learning to construct P sleep quality evaluators that meet the sleep quality evaluation accuracy constraint, where P is a positive integer greater than 1; Input the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data into the P sleep quality evaluators to obtain P sleep quality evaluation coefficients; Perform proportion calculation based on the P sleep quality evaluation precisions corresponding to the P sleep quality evaluators to obtain P evaluation precision incentive coefficients; Perform weighted calculation on the P sleep quality evaluation coefficients according to the P evaluation precision incentive coefficients to generate the real-time sleep quality coefficient.

5. The personalized sleep quality regulation method for big data analysis according to claim 1, characterized in that, Based on the sleep quality negative variation depth, the user physiological characteristic data, the user environmental characteristic data, and the user behavior characteristic data, make a positive intervention decision for the sleep user according to the sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy, including: The sleep quality positive intervention channel includes a sleep positive intervention gating model, a physiological characteristic sleep intervention decision model, an environmental characteristic sleep intervention decision model, and a behavior characteristic sleep intervention decision model; Judge whether the sleep quality negative variation depth is greater than or equal to the sleep quality negative variation depth threshold; If the sleep quality negative variation depth is greater than or equal to the sleep quality negative variation depth threshold, input the sleep quality negative variation depth into the sleep positive intervention gating model to obtain a sleep positive intervention gating coefficient; Based on the sleep positive intervention gating coefficient and the user physiological characteristic data, obtain a physiological characteristic sleep intervention decision according to the physiological characteristic sleep intervention decision model; Based on the sleep positive intervention gating coefficient and the user environmental characteristic data, obtain an environmental characteristic sleep intervention decision according to the environmental characteristic sleep intervention decision model; Based on the sleep positive intervention gating coefficient and the user behavior characteristic data, obtain a behavior characteristic sleep intervention decision according to the behavior characteristic sleep intervention decision model; Fuse the physiological characteristic sleep intervention decision, the environmental characteristic sleep intervention decision, and the behavior characteristic sleep intervention decision to obtain the sleep quality positive intervention strategy.

6. The personalized sleep quality regulation method for big data analysis according to claim 5, wherein Based on the sleep positive intervention gating coefficient and the user physiological characteristic data, obtain a physiological characteristic sleep intervention decision according to the physiological characteristic sleep intervention decision model, including: The physiological characteristic sleep intervention decision model includes Q physiological characteristic sleep intervention decision sub-models, where Q is a positive integer greater than 1; According to the sleep positive intervention gating coefficient, perform feature activation on the Q physiological characteristic sleep intervention decision sub-models to obtain multiple activated sleep intervention decision sub-models; Input the user physiological characteristic data into the multiple activated sleep intervention decision sub-models to obtain multiple physiological characteristic sleep intervention schemes; Fuse the multiple physiological characteristic sleep intervention schemes to obtain the physiological characteristic sleep intervention decision.

7. The personalized sleep quality regulation method for big data analysis according to claim 5, characterized in that Judge whether the sleep quality negative variation depth is greater than or equal to the sleep quality negative variation depth threshold, including: If the depth of negative variation of the sleep quality is less than the threshold of the depth of negative variation of the sleep quality, based on the future sleep window, perform feature prediction on the user's physiological characteristic data, the user's environmental characteristic data, and the user's behavioral characteristic data to obtain predicted user physiological characteristic data, predicted user environmental characteristic data, and predicted user behavioral characteristic data; Conduct a sleep quality evaluation based on the predicted user physiological characteristic data, the predicted user environmental characteristic data, and the predicted user behavioral characteristic data to obtain a predicted sleep quality coefficient; Conduct a negative variation depth evaluation on the predicted sleep quality coefficient according to the sleep quality expectation curve to generate a predicted sleep quality negative variation depth; Determine whether the predicted sleep quality negative variation depth is greater than or equal to the threshold of the sleep quality negative variation depth; If the predicted sleep quality negative variation depth is greater than or equal to the threshold of the sleep quality negative variation depth, based on the predicted sleep quality negative variation depth, the predicted user physiological characteristic data, the predicted user environmental characteristic data, and the predicted user behavioral characteristic data, perform sleep quality prediction adjustment on the sleeping user according to the sleep quality positive intervention channel.

8. A personalized sleep quality regulation device for big data analysis, characterized in that, The device is used to implement the personalized sleep quality adjustment method of big data analysis according to any one of claims 1-7, including: Real-time monitoring module: According to multi-modal monitoring sensors, conduct real-time monitoring on the sleeping user to obtain user physiological characteristic data, user environmental characteristic data, and user behavioral characteristic data; Sleep quality expectation fitting module: Load the expected sleep duration of the sleeping user, and combine the big data acquisition device and the time zone division granularity constraint table to perform sleep quality expectation fitting on the sleeping user to establish a sleep quality expectation curve; Real-time sleep quality evaluation module: Conduct a real-time sleep quality evaluation according to the user physiological characteristic data, the user environmental characteristic data, and the user behavioral characteristic data to obtain a real-time sleep quality coefficient; Negative variation depth evaluation module: Conduct a negative variation depth evaluation on the real-time sleep quality coefficient according to the sleep quality expectation curve to generate a sleep quality negative variation depth; Positive intervention decision-making module: Based on the sleep quality negative variation depth, the user physiological characteristic data, the user environmental characteristic data, and the user behavioral characteristic data, perform positive intervention decision-making on the sleeping user according to the sleep quality positive intervention channel to obtain a sleep quality positive intervention strategy; Sleep quality adjustment module: Adjust the sleep quality of the sleeping user according to the sleep quality positive intervention strategy.

Images