WO2020140913A1

WO2020140913A1 - Data processing method and apparatus, electronic device and storage medium

Info

Publication number: WO2020140913A1
Application number: PCT/CN2019/130837
Authority: WO
Inventors: 郑智民
Original assignee: 中国移动通信有限公司研究院; 中国移动通信集团有限公司
Priority date: 2019-01-02
Filing date: 2019-12-31
Publication date: 2020-07-09
Also published as: CN111460850B; CN111460850A

Abstract

A data processing method and apparatus, an electronic device and a storage medium. The method comprises: acquiring environmental data (S110); acquiring current state data and feature data of a target user (S120); selecting, according to the current state data, a regular sleep model or an irregular sleep model as a target model (S130); inputting the environmental data and the feature data into the selected target model, and obtaining light control parameters by means of the regular sleep model and the irregular sleep model (S140); and using the light control parameters to control the light emission of a light-emitting device (S150), wherein the step of controlling the light emission of the light-emitting device comprises controlling the light-emitting device to emit light that improves sleep of the target user, or controlling the light-emitting device to emit light that inhibits sleep of the target user.

Description

Data processing method and device, electronic equipment and storage medium

Cross-reference of related applications

This disclosure is based on a Chinese patent application with an application number of 201910002765.1 and an application date of January 2, 2019, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is incorporated by reference in this disclosure.

Technical field

The present disclosure relates to the field of information technology but is not limited to the field of information technology, in particular to a data processing method and device, electronic equipment, and storage medium.

Background technique

The biological effects of light on people can be divided into visual effects and non-visual effects. The visual effect is mainly composed of cone cells in the retina, which senses the luminosity and color; the non-visual effect is mainly composed of the pineal cells of the pineal gland, which senses the light, generates bioelectricity, affects the sympathetic nerves, and innervates the pineal gland cells to release melatonin In the flowing blood, hormone secretion decreases, metabolism slows down, and natural sleep is induced.

At present, there are some technical solutions for adjusting sleep through light, but it is found that the adjustment effect is not good, especially for users who have irregular sleep.

Summary of the invention

Embodiments of the present disclosure are expected to provide a data processing method and device, electronic equipment, and storage medium. The technical solution of the present disclosure is implemented as follows:

A first aspect of an embodiment of the present disclosure provides a data processing method, including:

Obtain environmental data;

Obtain the current status data and characteristic data of the target user;

According to the current state data, select a regular sleep model or an irregular sleep model as the target model;

Input the environmental data and the characteristic data into the selected target model to obtain lighting control parameters;

Using the light control parameters to control the light emission of the light-emitting device, wherein controlling the light emission of the light-emitting device includes: controlling the light-emitting device to emit light that promotes the sleep of the target user, or controlling the light-emitting device to suppress the emission The light that the target user sleeps.

Based on the above solution, the selection of a regular sleep model or an irregular sleep model as the target model according to the current sleep state of the target user includes at least one of the following:

If the current state data indicates that the current sleep state of the target user does not conform to the sleep rule corresponding to the regular sleep model, the irregular sleep model is selected as the target model;

If the current state data indicates that the current sleep state of the target user conforms to the sleep rule corresponding to the regular sleep model, the regular sleep model is selected as the target model.

Based on the above solution, the method further includes:

If the target model is a non-sleep regular model, acquire irregular sleep data of the target user;

The inputting the environmental data and the characteristic data into the selected target model to obtain the lighting control parameters includes:

The environmental data, the characteristic data and the irregular sleep data are input into the irregular sleep model to obtain the lighting control parameters.

Based on the above solution, the irregular sleep data includes at least one of the following:

Sleep time deviation data into sleep;

Time zone deviation data of the target user's time zone;

Continuous state data of this irregular sleep;

Data on the frequency of irregular sleep.

Based on the above solution, the acquiring environmental data includes at least one of the following:

Get current season data;

Obtain the lighting data of the space where the current target user is located;

Obtain the temperature data of the space where the current target user is located;

and / or,

The acquiring characteristic data of the target user includes:

Obtaining the static characteristic data of the user;

Obtain the dynamic feature data of the user.

Based on the above solution, the acquiring the dynamic feature data of the user includes at least one of the following:

Collecting current motion feature data of the target user;

Collect the current physical characteristic data of the target user.

Based on the above solution, the method further includes:

Performing a first denoising process on the feature data, removing the first noise data outside the preset frequency range and obtaining the first feature data;

Performing correlation analysis filtering on the first feature data to remove second noise data located in the preset frequency range and obtain second feature data;

Parsing the second characteristic data to obtain third characteristic data characterizing the state of the target user;

The third characteristic data and the environment data are input to the target model of the target user to obtain the lighting control parameter.

Based on the above solution, the parsing the second feature data to obtain third feature data characterizing the state of the target user includes:

Parsing the second feature data, and extracting time-domain feature data of the target user from the second feature data, wherein the time-domain feature data includes: a characteristic curve corresponding to the second feature data in the time domain Peak data and/or trough data;

Peak data and/or trough data of the target user's action curve or sign change curve;

and / or,

Parsing the second feature data, extracting the frequency domain feature data of the target user from the second feature data; wherein, the frequency domain feature data includes: a characteristic curve corresponding to the second feature data in the frequency domain Peak data and/or trough data.

Based on the above solution, the first noise data includes: jitter data with a jitter frequency of the acquisition device outside the preset frequency range, electromagnetic interference data with an electromagnetic frequency outside the preset frequency range, and electromagnetic frequency Magnetic field noise outside the set range;

and / or,

The second noise data includes: jitter data whose jitter frequency of the acquisition device is within the preset frequency range.

Based on the above solution, the inputting of the environmental data and the characteristic data into the selected target model to obtain lighting control parameters includes:

According to the dimensionality reduction processing strategy, performing dimensionality reduction processing on the third feature data and environment data to obtain input data of a preset dimension;

Inputting input data of a preset dimension into the target model to obtain the lighting control parameter.

Based on the above solution, performing the dimensionality reduction processing on the third feature data and the environmental data according to the dimensionality reduction processing strategy to obtain input data of a preset dimension includes:

Determine whether the target user is in an overall static state based on the first preset condition and the motion feature data of the target user;

If the target user is in an action state, determine whether the target user is in the first type of partial static state according to the action feature data;

If the target user is not in the first type of local stationary state, determine whether the target user is in the second type of local stationary state according to the motion feature data;

If the target user is in an action state, the target user's action feature data is sampled according to the target user's current action state to obtain sampling feature data as the input data.

Based on the above solution, according to the dimensionality reduction processing strategy, performing dimensionality reduction processing on the third feature data and environment data to obtain input data of a preset dimension, further including:

If the target user is in the overall static state, the determination of whether the target user is in the first type of partial static state and the second type of partial static state is stopped;

and / or,

If the target user is in the first type of local stationary state, the determination of whether the target user is in the second type of local stationary state is stopped.

Based on the above solution, after using the light control parameters to control the light emission of the light emitting device, the method further includes:

Obtain lighting control effect data, wherein the effect data includes: at least one of the target user's sleep effect data, sleep activity data, and non-sleep activity data;

According to the effect data, the regular sleep model and/or the irregular sleep model are optimized.

A second aspect of an embodiment of the present disclosure provides a data processing apparatus, including:

The first obtaining module is configured to obtain environmental data;

The second obtaining module is configured to obtain the current state data and characteristic data of the target user;

A selection module configured to select a regular sleep model or an irregular sleep model as the target model based on the current state data

A third acquisition module configured to input the environmental data and the characteristic data into the selected target model to obtain lighting control parameters;

The control module is configured to use the light control parameters to control the light emission of the light emitting device, wherein the controlling the light emission of the light emitting device includes: controlling the light emitting device to emit light that promotes the sleep of the target user, or controlling the light emission The device emits light that suppresses the sleep of the target user. Regular sleep model and irregular sleep model Regular sleep model and irregular sleep model Regular sleep model and irregular sleep model Regular sleep model and irregular sleep model Regular sleep model and irregular sleep model Regular sleep model and irregular sleep model Model and irregular sleep model Regular sleep model and irregular sleep model Regular sleep model and irregular sleep model

A third aspect of the embodiments of the present disclosure provides an electronic device, including:

Memory

A processor, connected to the memory, is configured to implement the data processing method provided by the foregoing one or more technical solutions by executing computer-executable instructions located on the memory.

A fourth aspect of an embodiment of the present disclosure provides a computer storage medium that stores computer-executable instructions; after the computer-executable instructions are executed, the data processing method provided by the foregoing one or more technical solutions can be provided.

The technical solution provided by the embodiments of the present disclosure will acquire environmental data, current state data and characteristic data of the target user, and first determine whether to use the regular sleep model or the irregular sleep model as the target model according to the current state data to form the light control parameters , In this way, the lights can be controlled according to whether the user is currently sleeping regularly or irregularly, thereby providing better light control that is beneficial to inhibit or promote sleep, regardless of whether the current target user's sleep is regular or irregular , Can achieve accurate light control, when the target user needs to promote sleep to emit light that promotes sleep, and needs to suppress the emission of light that suppresses sleep during sleep, in order to use light to accurately adjust the target user's sleep.

BRIEF DESCRIPTION

FIG. 1 is a schematic flowchart of a first data processing method provided by an embodiment of the present disclosure;

2 is a schematic flowchart of a second data processing method provided by an embodiment of the present disclosure;

3 is a schematic structural diagram of a data processing device according to an embodiment of the present disclosure;

4 is a schematic flowchart of a third data processing method provided by an embodiment of the present disclosure;

FIG. 5 is a schematic flowchart of a method for providing sleep adjustment for irregular sleep according to an embodiment of the present disclosure;

6 is a schematic structural diagram of an electronic device provided by implementation of the present disclosure;

7 is a schematic diagram of an irregular sleep model provided by an embodiment of the present disclosure to control a light emitting device to emit light.

detailed description

The technical solutions of the present disclosure will be further elaborated below with reference to the drawings and specific embodiments of the specification.

The study found that people who fall in jet lag, three shifts in the factory, and medical staff are most likely to cause chronic circadian clock imbalance, which can easily cause stomach upset to ulcers and heart disease. Sympathetic excitability is related to the color and intensity of light reaching the pineal gland. Some manufacturers have introduced blue light irradiation, which can improve irregular sleep. Appropriate illumination of the sleep aid lamp (N1 light sleep stage (1200-7000 Lux, N4 before auto-awakening, 2500-10000lux at the REM stage), will increase the production of melatonin in the body, melatonin (with circadian rhythm, secrete up to midnight Peak) helps to deepen the depth of sleep (the proportion of REM and N4), change the sleep-wake rhythm, adjust the physiological clock, and adjust the quality of sleep. For example: genetic studies have shown that non-visual effects are maximum at 480-485nm, and visible light is different The sensitivity of the wavelength is also inconsistent, it has the highest sensitivity to yellow-green light, but the sensitivity to red light, blue light, and violet light is very low. The three colors of red, green, and blue light respectively suppress the melatonin under the irradiation of 1000 lx The red light is very small, the green light is the largest, and the blue light is slightly lower. The irradiation of light will significantly increase the heart rate. The shorter the wavelength, the more obvious, young people are more obvious than the elderly. Humans are most sensitive to light between 480 and 485 nm. Different people are affected by the yellow pigment in the central macula area of the human eye. With age, the visual crystals turn yellow, which will cause individual differences. However, people with different ages and sleep constitutions may enter different sleep stages at different times. The sensitivity of light is also different, only according to the time when the general population enters each sleep stage, setting the same sleep aid light adjustment mechanism for everyone. It may instead affect sleep, such as elderly people who sleep lightly, and do not enter light sleep at 24:00, may instead Sleep is affected. In addition, the intensity required for irregular sleep and normal sleep is not the same, the individual differences between irregular sleep and normal sleep, such as how many consecutive hours of irregular sleep, the length of irregular sleep, irregular sleep The recent appearance of ratings, etc., is also an important consideration for training differential models.

In view of this, considering the regular sleep and irregular sleep, the light requirements of the user to promote or inhibit sleep are inconsistent, and the data processing method shown in Figure 1 is proposed, including:

Step S110: Acquire environmental data;

Step S120: Obtain the current state data and characteristic data of the target user;

Step S130: Select a regular sleep model or an irregular sleep model as the target model according to the current state data;

Step S140: Input the environmental data and the characteristic data into the selected target model, and the regular sleep model and the irregular sleep model obtain the light control parameters;

Step S150: Using the light control parameters to control the light emission of the light emitting device, wherein the controlling the light emission of the light emitting device includes: controlling the light emitting device to emit light that promotes the sleep of the target user, or controlling the light emitting device to emit Suppressing the light of sleep of the target user.

The data processing method provided in this embodiment can be applied to various electronic devices, for example, user equipment or a home gateway. The user equipment may be: a user's mobile phone, tablet computer or wearable device. After obtaining the light control parameters, these devices control the light-emitting devices to emit light, or control themselves to emit light.

The light-emitting device may be various devices capable of emitting light, and may be the user device itself, or a light-emitting device at a location of a target user such as a bedroom or a living room.

In this embodiment, the electronic device first obtains environmental data, which represents the environmental conditions of the space where the target user is currently located. For example, collecting environmental data through environmental sensors.

Different environments have different effects on users' sleep activities and non-sleep activities. For example, in the case of too bright or too dark, none of the target user's sleep is utilized. In this embodiment, it is expected that lights can be used to assist sleep activities or non-sleep activities. In this embodiment, the characteristic data of the target user is obtained, and the characteristic data may constitute an individual static portrait of the target user, an individual dynamic portrait of the target user, a sign portrait of the target user, an emotional portrait of the target user, and so on.

The user feature data may include: feature data that is collected, and also feature data that is dynamically collected, for example, feature data that is dynamically collected using a physiological sensor, and the sign data includes but is not limited to: the heartbeat data of the target user, the target user’s Brain wave data, pulse data of target users, etc.

The current state parameter may at least represent whether the target user is currently in a sleep state or a non-sleep state.

In some embodiments, the current state parameter can also be used to characterize the sleep stage or sleep depth of the target user in the sleep state; the current state parameter can also be used to characterize the degree of awake of the target user in the non-sleep state, etc. parameter.

In this embodiment, a target user corresponds to two models, namely a regular sleep model and an irregular sleep model.

The regular sleep model may be used to form the light control parameter when the target user's current sleep state is regular sleep; the irregular sleep model may be used to form the light control parameter when the target user's current sleep state is irregular sleep.

In this way, in step S140, corresponding light control parameters can be formed according to whether the current sleep state of the target user is regular sleep, so as to achieve precise control of the light.

Due to the color of the light and/or the brightness of the light, there is a certain promotion or suppression effect on human emotions and signs. In this embodiment, the environmental data and characteristic data are used as the input of the target model, which can be The light control parameters are obtained and used to control the light emission of the light emitting device.

For example, the light control parameter includes at least one of the following:

The color control parameter of the light is used to control the color of the light. For example, the color control parameter may include: a light emitting wavelength, and different wavelengths correspond to different colors of light;

The brightness control parameter of the light is used to control the brightness of the light emitting device. For example, the brightness control parameter may include: a light value.

Control parameters of the color change of the light,

The brightness of the light changes control parameters. In some cases, the brightness of the light changes to promote user sleep or inhibit user sleep. For example, in the morning, the light may be getting brighter and brighter; when falling asleep at night, the brightness of the light is getting darker and darker; and the color of the light may need to change from warm tones when it is necessary to wake up the target user The lights gradually switch to cool-tone lights; and you may need to gradually switch from cool-tone lights to warm-tone lights before falling asleep.

The lighting angle control parameter of the light is used to control the lighting angle of the light; if the light is directly irradiated to the target user's field of view, it may inhibit sleep, and the side illumination may help the user to sleep, so the light control parameter also includes illumination Angle control parameters.

In short, in this embodiment, the environmental data and the characteristic data are input to the regular sleep model and the irregular sleep model set by the target user. In this way, the light control parameters of the target user are obtained; In order to provide lighting control parameters for target users of different ages, different sleeping habits, and different genders, the lighting control is performed for each user, thereby ensuring the lighting control that meets the user's personality.

The step S130 may include at least one of the following:

For example, according to the sleep rule of regular sleep, the target user at the current time should be in a sleep state, but the target user of the current state data is still in a non-sleep state, indicating that it is currently suitable to use an irregular sleep model to form the light control parameter.

As another example, according to the sleep rule of regular sleep, the current time is the user's sleep time but the time zone of the target user has changed. It can also be considered that it does not conform to the sleep rule corresponding to the regular sleep model. It is necessary to select the non-sleep rule as The target model.

In short, if any of the sleep time and sleep time zone (the time zone where the target user is located) does not satisfy the sleep rule corresponding to the regular sleep model, the irregular sleep model is selected as the target model, otherwise the regular sleep model is selected as the target model.

In other embodiments, the method further includes:

The step S140 may include: inputting the environmental data, the characteristic data, and the irregular sleep data into the irregular sleep model to obtain the light control parameter.

The irregular sleep data may include various data representing the difference between the irregular sleep state and the regular sleep state. In this way, the irregular sleep model can combine irregular sleep data to determine whether it is necessary to suppress sleep or promote sleep. At the same time, the irregular sleep model can combine the environmental data and the characteristic data to determine the specific lighting control parameters required to promote sleep or inhibit sleep.

In some embodiments, as shown in FIG. 7, the irregular sleep data includes at least one of the following:

Sleep time deviation data into sleep, for example, the target user usually falls asleep at 10:00 pm, the current time is already 1:15 the next day, the target user has not yet entered sleep, the time difference between the two, 2:15 can be used as One of the time deviation data;

Time zone deviation data of the target user's time zone; for example, the target user is flying from Beijing to London. Due to the difference between Beijing's time zone and London's time zone, the time zone deviation data can also be used as irregular sleep data in China;

Continuous state data of this irregular sleep; for example, the continuous state data may include at least one of the following: duration of the target user's irregular sleep, duration of the target user's irregularly maintained non-sleep state, and target user's irregularly maintained sleep state The length of time

The frequency of occurrence of irregular sleep; for example, the frequency of occurrence of irregular sleep may include at least one of the following: the number and/or frequency of occurrence of irregular sleep within a preset period such as the last half month or a month.

After these irregular sleep data are input to the irregular sleep model, it is convenient for the irregular sleep model to determine whether to promote sleep or suppress sleep. In addition, the non-sleep data can also be used together with environmental data and characteristic data to generate light control parameters that promote sleep or suppress sleep.

As can be seen from FIG. 7, a lot of irregular sleep data (or irregular sleep factors, for example, sleep time deviation data when entering sleep, time zone deviation data of the time zone where the target user is located, irregular One or more of the occurrence frequency data of sleep and the continuous state data of this irregular sleep.

In some embodiments, the step S110 may include at least one of the following:

Get current season data;

Obtain the lighting data of the space where the current target user is located;

Get the temperature data of the space where the current target user is located.

Different seasons have different temperatures, different sunshine, and different sleep durations and/or deeper and lower sleep required by the target user.

In some embodiments, the environmental data may further include: humidity data, and different humidity has different effects on the sleep of the target user.

and / or,

The step S120 may include:

Obtaining the static characteristic data of the user;

Obtain the dynamic feature data of the user.

The static feature data may include: data corresponding to the user's static personal portrait, for example, the age, gender, and/or personal sleep characteristics of the target user.

The dynamic characteristic data includes but is not limited to at least one of the following:

Physical movement data of the target user;

Target user's heart rate data;

Target user's heart rate variability (HRV) data, etc. HRV characterizes the active level of the sympathetic nerve and parasympathetic application of the target user. The more active the sympathetic nerve, the more excited the user's emotions and the less suitable for sleep.

In some embodiments, the acquiring dynamic characteristic data of the user includes at least one of the following:

Collect the current motion feature data of the target user; for example, hand motion data, foot motion data, head motion data and/or torso motion data;

Collect current physical characteristic data of the target user, for example, heart rate data, blood oxygenation data, and/or HRV data, etc.

In some embodiments, as shown in FIG. 2, the method further includes:

Step S101: Perform a first denoising process on the feature data, remove the first noise data outside the preset frequency range, and obtain the first feature data;

Step S102: Perform correlation analysis filtering on the first feature data, remove second noise data located in the preset frequency range, and obtain second feature data;

Step S103: Analyze the second feature data to obtain third feature data characterizing the state of the target user;

The step S130 may include: inputting the third characteristic data and the environment data to the target model of the target user to obtain the lighting control parameter.

In this embodiment, in order to reduce the influence of the noise data input into the regular sleep model and the irregular sleep model on the accuracy of the light control parameters, in this embodiment, the feature data will be first denoised. In this embodiment, in order to remove the noise data in the feature data as much as possible, denoising will be performed twice. For example, through the first denoising process, the first noise data outside the preset frequency range where the feature data is located can be removed, and then through correlation analysis, the second noise data within the preset frequency range where the feature data is located can be removed. The user's movement characteristics and/or signs will show a certain regularity, but the second noise data may not show such a regularity, this regularity may be a variational law in the time domain, and/or, a variational law in the frequency domain, For example, the user's heartbeat data has a certain periodicity, which is a change law in the time domain. The user's brain waves may switch between waves of different frequencies, but they may all be between specific frequency points. This law is the law of change in the frequency domain. The correlation analysis may be: judging whether the separated signals satisfy the change rule of the target user's characteristics in the time and/or frequency domain, and filtering out changes that do not meet the target user's time and/or frequency domain Regular second noise data. Of course, this is only an example, and the specific implementation is not limited to this.

In some embodiments, the step S103 may include:

and / or,

If the target user is moving, the action curve can be drawn in the time or frequency domain by collecting the action feature data. The peak value and/or the trough value of the action curve may be very large data that characterize the user's action, so In this embodiment, peak data and/or trough data of the action curve in the time domain and/or the action curve in the frequency domain may be extracted as the characteristic data and input into the individual's sleep portrait. For example, from the second data, parameters such as the action amplitude, action intensity, maximum action amplitude and/or action intensity at each action frequency when extracting action peaks and troughs are extracted.

The user's sign data can also draw corresponding sign curves in the time domain and/or frequency domain. According to the sign curves, the peak data and/or trough data in the second characteristic parameters can be extracted for processing.

In some embodiments, the first noise data includes: jitter data with a jitter frequency of the acquisition device outside the preset frequency range, electromagnetic interference data with an electromagnetic frequency outside the preset frequency range, and electromagnetic frequency The magnetic field noise outside the preset range;

and / or,

The vibration of the device will affect the user's movement characteristic data, and the electromagnetic interference noise and/or magnetic field noise will affect the user's brain wave data.

In some embodiments, the step S130 may include:

Input the input data of a preset dimension to the target to obtain the light control parameter.

After actually combining the environmental data and the characteristic data according to a predetermined combination method, a high-latitude vector and/or matrix will be formed, but some of these data may not contribute much to the lighting control parameters. Or, some data combination will affect the lighting control parameters. In order to reduce the amount of data processing, in this embodiment, dimensionality reduction processing will be performed on the data using a dimensionality reduction processing strategy, for example, through dimensionality reduction processing of environmental parameters and feature data, only M pieces of data are obtained as the human sleep portrait model Input data; the M can be a value of 6, 9, or 12, and the specific value can be set dynamically according to demand.

In some embodiments, performing dimensionality reduction processing on the third feature data and environment data according to a dimensionality reduction processing strategy to obtain input data of a predetermined dimension includes:

Determine whether the target user is in an overall static state based on the first preset condition in combination with the target user's motion feature data;

For example, the human body posture data in the motion feature data can be used as a judgment as to whether the entire body is in a stationary state. If the entire body is in a stationary state as a whole, it may indicate that the target user is currently in a sleep state or a sleep enter state.

For another example, based on the motion feature data of each part in the motion feature data, it is determined whether the target user is in an overall still state as a whole, for example, the user's hand motion data and foot motion data are used to determine whether the target user is in an overall still state. For example, the user's hand movement data indicates that the target user's hand movements are slight, and the foot movement data indicates that the foot movements are slight. The target user may be considered to be in an overall static state, otherwise, the target user may be considered to be in an action state. If the target user is in an action state, it is necessary to further determine whether the target user is in a partial action state. For example, if the target user may lie down or sit down but still play a mobile phone, the target user is in a hand action state, rather than a stationary hand status. In this embodiment, it is also determined whether the user is in a local static state. If a target user is in a local static state, the local action state data may be removed and not used as input data; or, it is required as data. In this embodiment, the first type of local static state and the second type of local static state are different static states, and the difference may be reflected in at least one of the following:

For example, different types of local stationary states, such as local translational stationary states and local rotational stationary states;

For another example, the part corresponding to the first partial static state is larger than the part corresponding to the second partial static state. For example, the part corresponding to the first partial static state may be the entire upper limb; the part corresponding to the second partial static state may be a finger.

In order to reduce unnecessary data processing, according to the dimensionality reduction processing strategy, performing dimensionality reduction processing on the third feature data and environment data to obtain input data of a preset dimension, further including:

and / or,

By stopping the judgment of different static states in time, the amount of calculation required can be reduced and the processing rate can be improved.

In some embodiments, the method further includes:

Obtain lighting control effect data, wherein the effect data includes: the target user's sleep effect data and/or non-sleep activity data;

According to the effect data, the regular sleep model or the irregular sleep model is optimized.

In this embodiment, the regular sleep model may be trained using regular sleep data and regular sleep effect data of the regular sleep model.

The training data of the irregular sleep model may include:

Regular sleep data, lighting control data and regular sleep effect data;

Irregular sleep data, lighting control data, and irregular sleep effect data.

The specific content of the irregular sleep data here can be referred to the foregoing embodiment, and will not be discussed here.

The regular sleep data, light control data and regular sleep effect data are introduced into the training of irregular sleep models. In this way, the model training can be performed according to the regular sleep data, light control data, and irregular sleep effect data, thereby making irregular sleep The model has further improved the accuracy of lighting control parameters. After controlling the light-emitting device to emit light, the sleep effect data of the user will also be collected, for example, the sleep effect data may represent the promotion and/or inhibition of the light after the controlled adjustment on the user's sleep, which is adopted in this embodiment. Obtain the effect data, you can know the effect after the current light adjustment. Whether the effect achieves the expected effect, and the corresponding light control parameters can be used as regular sleep model and irregular sleep model to further optimize the training sample data, and perform deep optimization of the regular sleep model and irregular sleep model, so that the regular sleep The more the model and the irregular sleep model are used, the more they can characterize the personal characteristics of the target user, so that they can output light control parameters that meet the personal characteristics of the target user to adjust the light; to further improve the sleep stage and/or non-sleep of the user Stage lighting control.

As shown in FIG. 3, this embodiment also provides a data processing apparatus, including:

The first obtaining module 110 is configured to obtain environmental data;

The second obtaining module 120 is configured to obtain the current state data and characteristic data of the target user;

The selection module 130 is configured to select a regular sleep model or an irregular sleep model as the target model according to the current state data;

The third obtaining module 140 is configured to input the environmental data and the characteristic data into the selected target model to obtain lighting control parameters;

The control module 150 is configured to use the light control parameters to control the light emission of the light emitting device, wherein the controlling the light emission of the light emitting device includes: controlling the light emitting device to emit light that promotes sleep of the target user, or, controlling the The light emitting device emits light that suppresses the sleep of the target user.

In some embodiments, the first acquisition module 110, the second acquisition module 120, the selection module 130, the third acquisition module 140, and the control module 150 may all be program modules, which can be implemented after being executed by the processor Acquisition of the aforementioned various data, and lighting control of the lighting device.

In some embodiments, the first acquisition module 110, the second acquisition module 120, the selection module 130, the third acquisition module 140, and the control module 150 may all be soft-hard combination modules. The soft-hard combination module may include: A programmable array, for example, a field programmable array or a complex programmable array.

In still other embodiments, the first acquisition module 110, the second acquisition module 120, the selection module 130, the third acquisition module 140, and the control module 150 may all be pure hardware modules, and the pure hardware modules may include: dedicated integrated circuit.

In some embodiments, the selection module is configured to perform at least one of the following: if the current state data indicates that the current sleep state of the target user does not conform to the sleep rule corresponding to the regular sleep model, select the The irregular sleep model is the target model; if the current state data indicates that the current sleep state of the target user conforms to the sleep rule corresponding to the regular sleep model, the regular sleep model is selected as the target model.

In some embodiments, the device further includes:

An irregular sleep data module configured to obtain irregular sleep data of the target user if the target model is an irregular sleep model;

The third obtaining module is configured to input the environment data, the characteristic data and the irregular sleep data into the irregular sleep model to obtain the light control parameter.

In some embodiments, the irregular sleep data includes at least one of the following:

Sleep time deviation data into sleep;

Time zone deviation data of the target user's time zone;

Continuous state data of this irregular sleep;

Data on the frequency of irregular sleep.

In some embodiments, the first obtaining module 110 is configured to perform at least one of the following:

Get current season data;

Obtain the lighting data of the space where the current target user is located;

and / or,

In some embodiments, the second obtaining module 120 is configured to obtain the static characteristic data of the user; and obtain the dynamic characteristic data of the user.

In some embodiments, the second acquisition module 120 is configured to perform at least one of the following:

Collecting current motion feature data of the target user;

Collect the current physical characteristic data of the target user.

In some embodiments, the device further includes:

A first denoising module, configured to perform a first denoising process on the feature data, remove the first noise data outside the preset frequency range and obtain the first feature data;

A second denoising module configured to perform correlation analysis filtering on the first feature data, remove second noise data located in the preset frequency range, and obtain second feature data;

An analysis module configured to analyze the second characteristic data to obtain third characteristic data characterizing the state of the target user;

The third obtaining module is configured to input the third characteristic data and the environment data to the regular sleep model and irregular sleep model of the target user to obtain the light control parameter.

In some embodiments, the parsing module is configured to parse the second feature data and extract the time-domain feature data of the target user from the second feature data, where the time-domain feature data includes: Peak data and/or trough data of the characteristic curve corresponding to the second feature data in the time domain; peak data and/or trough data of the target user's action curve or sign change curve;

And/or, parse the second feature data, and extract frequency domain feature data of the target user from the second feature data; wherein, the frequency domain feature data includes: the second feature data location in the frequency domain The peak data and/or trough data of the corresponding characteristic curve.

In some embodiments, the first noise data includes: jitter data with a jitter frequency of the acquisition device outside the preset frequency range, electromagnetic interference data with an electromagnetic frequency outside the preset frequency range, and electromagnetic frequency Magnetic field noise outside the preset range; and/or, the second noise data includes: jitter data of a jitter frequency of the acquisition device within the preset frequency range.

In still other embodiments, the third acquiring module 140 is configured to perform dimension reduction processing on the third feature data and environment data according to a dimension reduction processing strategy to obtain input data of a preset dimension; Input data to the regular sleep model and the irregular sleep model to obtain the light control parameters.

In some embodiments, the third acquiring module 140 is configured to determine whether the target user is in an overall still state based on the first preset condition and the action characteristic data of the target user; if the target user is in action State, determine whether the target user is in the first type of local static state according to the action feature data; if the target user is not in the first type of local static state, determine whether the target user is based on the action feature data In the second type of local static state; if the target user is in an action state, the target user's action feature data is sampled according to the target user's current action state to obtain the sampling feature as the input data data.

In some embodiments, the third acquiring module 140 is configured to stop whether the target user is in the first type of partial static state and the second type of partial static state if the target user is in the overall static state Determination; and/or, if the target user is in the first type of local stationary state, stop determining whether the target user is in the second type of local stationary state.

In some embodiments, the device further includes:

The fourth obtaining module is configured to obtain the effect data of the light control after the use of the light control parameter to control the light emission of the light emitting device, wherein the effect data includes: the sleep effect data of the target user and/or Sleep activity data;

The optimization module is used to optimize the regular sleep model and the irregular sleep model according to the effect data.

The following provides some specific examples in combination with any of the above embodiments:

Example 1:

The present disclosure collects the user's movements and heart rate through a smart bracelet, performs color light adjustment projection through a smart projection mobile phone or gateway and other smart devices, and adaptively improves the learning and working environment. Combined with different wavelengths of light, seasons, and personal signs, the corresponding portrait model in the overall crowd is used to predict the most suitable color sleep adjustment model for each person, and the individual sleep effect evaluation model based on the heart rate and movement changes caused by the bracelet detection is supervised. Learning, so as to establish a regular sleep model or an irregular sleep model that is most suitable for the individual's color sleep adjustment, and it is continuously revised according to the genetic algorithm. The opposite direction of promoting sleep at the same time is to suppress sleep, so that this awake state suppresses sleep.

In some embodiments, a test sample of a first-level label is obtained, the test sample is imported into the demand model, and the accuracy rate of the demand model is obtained according to the output result of the demand model; the demand model here may be Including the aforementioned regular sleep model and irregular sleep model.

When the accuracy rate does not meet the requirement of the preset accuracy rate, according to the accuracy rate of the demand model, the influence weight of each of the influence factors in the training sample and the modification of the training sample are modified;

The modified training sample is used to train the demand model until the accuracy rate of the obtained demand model meets the preset accuracy rate requirement.

In some embodiments, the step of obtaining a general model matching a first-level tag from a plurality of general models in a database includes:

Find a general model matching the crowd type of the first-level tag from the multiple general models of the database according to the crowd type of the first-level tag;

A general model consistent with the configuration of the next-level tag of the first-level tag is selected from the found general models according to the configuration of the next-level tag in the first-level tag.

In some embodiments, each of the impact factors includes multiple features, and the step of preprocessing the training sample includes:

Detecting whether the integrated output amplitude of the multiple influence factors in the training sample is within a preset range;

If it is not within the preset range, then calculate the variance values of the comprehensive output amplitude of each of the impact factors relative to the impact factors in the general model;

Detecting whether the variance value corresponding to each of the impact factors is greater than a preset threshold, and if it is greater than the preset threshold, a preset number of impact factors are randomly selected from the multiple impact factors, and among the extracted impact factors The characteristics of each of the impact factors are normalized to simplify the data volume of each of the impact factors.

In some embodiments, the general model is constructed based on a neural network. The neural network includes an input layer, an output layer, and a hidden layer. The input layer, the output layer, and the hidden layer include multiple neurons, respectively. The neurons between the input layer, the output layer and the hidden layer have connection weight values, and the step of importing the pre-processed training samples into the general model to train the general model includes:

Import the pre-processed training samples to the input layer of the neural network, and after the training of the hidden layer, output from the output layer;

Detecting whether the output result of the output layer reaches the expected result, and if the expected result is not reached, an error signal is obtained according to the output result and the expected result, and enters the back propagation stage;

Use the error signal as the input signal in the backward propagation stage to reversely return from the output layer to the input layer, and in the process of reverse return, correct the input layer, output layer, and hidden layer The weight value of the connection between neurons is to gradually reduce the final output error signal.

In some embodiments, the step of modifying the connection weight value of the neurons between the input layer, the output layer, and the hidden layer during the reverse return process to gradually reduce the final output error signal includes: :

In the process of reverse transmission, the following formula is used to modify the connection weight values of the neurons between the input layer, the output layer and the hidden layer to gradually reduce the final output error signal:

Among them, W _ij represents the connection weight value between the i-th neuron of the input layer to the j-th upgrade of the hidden layer, and X _P represents the i-th input value of the P-th training sample in the input layer,

Represents the threshold of the jth neuron in the hidden layer.

In some embodiments, the step of modifying the influence weight of each influence factor in the training sample and modifying the training sample according to the accuracy rate of the demand model includes:

Compare the accuracy rate of the current demand model with the accuracy rate of the demand model obtained in the historical times;

If the accuracy rate of the current demand model is higher than the accuracy rate of more than half of the demand model obtained in the history, the current training sample is retained, and the impact weight of each of the impact factors is modified;

If the accuracy rate of the currently obtained demand model is lower than the accuracy rate of more than half of the demand model obtained in the historical times, part of the training samples in the training samples are deleted, and new training samples are added.

In some embodiments, after the step of modifying the influence weight of each influence factor in the training sample and modifying the training sample according to the accuracy rate of the demand model, the method further includes:

Establish an fitness function to evaluate the fitness value of each of the training samples in the modified training samples, and use the selection mechanism of the genetic algorithm to select the training sample with the highest fitness value;

Use the cross mechanism of genetic algorithm to randomly select any two training samples from multiple training samples to cross to obtain the next generation of training samples;

Using the fitness function to calculate the training sample with the highest fitness value among the next-generation training samples;

Detecting whether the fitness value of the next generation of training samples is lower than the fitness value of the previous generation of training samples, and if it is lower, a mutation factor is introduced using the mutation mechanism of the genetic algorithm to mutate the next generation of training samples Operation, and then calculate the fitness value of the training sample after the mutation operation;

Modify the training sample again according to the fitness value of the training sample.

In some embodiments, after the step of training the demand model using the modified training samples until the accuracy rate of the obtained demand model meets the preset accuracy rate requirement, the method further includes:

Receiving currently input information to be tested, the information to be tested carries a plurality of different influence factors, and each of the influence factors carries corresponding influence weights;

Import the information to be tested into the demand model for prediction to obtain a prediction result;

Receiving feedback information on the prediction result of the demand model input by the user;

Adjust the model parameters of the demand model according to the feedback information.

In some embodiments, after the step until the accuracy rate of the obtained demand model meets the preset requirements, the method further includes:

Storing the obtained demand model that meets the requirements of the preset accuracy rate in the database to update the data in the database.

An embodiment of the present disclosure also provides a data processing apparatus. The data processing apparatus includes:

The general model acquisition module is used to obtain a general model matching a first-level tag from multiple general models in the database;

The training sample acquisition module is used to obtain a training sample of a first-level label, the training sample carries a plurality of different influence factors, each of the influence factors carries a corresponding influence weight, and the training sample includes the Staff characteristics and medical resource deployment information;

A demand model obtaining module, which is used to preprocess the training samples, import the preprocessed training samples into the general model, and train the general model to obtain the demand model corresponding to the first-level label;

An accuracy rate obtaining module, used to obtain test samples of first-level tags, import the test samples into the demand model, and obtain the accuracy rate of the demand model according to the output result of the demand model;

A modification module, configured to modify the influence weight of each influence factor in the training sample and modify the training sample according to the accuracy rate of the demand model when the accuracy rate does not meet the preset accuracy rate requirement;

The training module is used for training the demand model using the modified training samples until the accuracy rate of the obtained demand model meets the preset accuracy rate requirement.

A supervised classification algorithm is used, with environmental data as the input layer and individual sleep quality score as the output layer. By comparison with the model (historical best environmental data) formed by the last environmental data input, the quality of individual evaluation is used as a training supervision factor, preferably 1 and worse 0.

The forward propagation of the working signal. During this period, the weights and thresholds of the neurons in the network remain unchanged. Each layer of neurons only affects the input and state of the next layer of neurons. If the desired output is not obtained at the output The output value, the network is transferred to the back propagation process of the error signal. The back propagation of the error signal, the error signal starts to be transmitted back layer by layer from the output end. In this propagation process, the weights and thresholds of the neurons of the network are adjusted by error feedback according to certain rules. The above two stages are carried out alternately and cyclically, and each time it is completed, it is corrected by genetic algorithm.

At the same time, the individual as a new input factor of the corresponding group portrait population in the overall population, SVM genetic modification of the overall population corresponding portrait crowd environment model, the corresponding sleeping environment user portraits are constantly clear and refined. SVM classifier fitness function f(x _i )=min(1-g(x _i )),

Divide the correct rate of the samples for the SVM classifier. As the sample size increases, if the correct rate is higher than the historical best model, the model replaces the original best model, so as the sample size increases, the model is continuously optimized and perfected. .

The adaptive improvement of the model may include: With the increase of the sample size, the SVM classifier can continuously optimize and improve the adaptive.

SVM classifier fitness function f(x _i )=min(1-g(x _i )),

The accuracy of dividing samples for SVM classifier includes:

3D modeling;

Set the boundary conditions;

Unsteady calculation;

Judging the change of boundary conditions?

If not, the judgment result is steady? If not, set the camera angle, travel path and rendering effect; if yes, return to unsteady calculation;

If yes, return to setting boundary conditions.

The boundary condition here is the classification boundary of the SVM classifier.

Appropriate illumination of the sleep aid lamp (N1 light sleep stage (1200-7000Lux, N4 before auto-awakening, REM 2500-10000lux) will increase the production of melatonin in the body, melatonin (with circadian rhythm, midnight secretion up to Peak) helps to deepen the depth of sleep (the proportion of REM and N4), change the sleep-wake rhythm, adjust the physiological clock, and adjust the quality of sleep.

Genetic research shows that the non-visual effect is the largest at 480-485nm, and the sensitivity of different wavelengths of visible light is also inconsistent. It has the highest sensitivity to yellow-green light, but very low sensitivity to red, blue, and purple light. The three kinds of red, green and blue light under 1000lx irradiation, the inhibition rate of melatonin, red light is very small, green light is the largest, blue light is slightly lower, the light irradiation will significantly increase the heart rate, the shorter the wavelength, the more significant, Young people are more obvious than old people. Humans are most sensitive to light between 480 and 485 nm. Different people are different, affected by the yellow pigment in the central macula area of the human eye's retina. With age, the visual crystals turn yellow, which can cause individual differences.

People of different ages and sleep physiques may enter different sleep stages at different times and have different levels of sensitivity to light. Only according to the time when the general population enters each sleep stage, the same sleep-assisted light adjustment mechanism is set for everyone. It may affect sleep on the contrary, for example, the old man sleeps lightly, and has not entered light sleep at 24:00, but may be affected.

Example 2:

As shown in FIG. 4, this example provides a data processing method for sleep adjustment, especially sleep adjustment for irregular sleep, specifically including:

Detect the current status data of the target user;

Determine whether the current user's sleep is regular sleep according to the current state data;

If it is, then use regular sleep model combined with the collected environmental data and target user's characteristic data to control the light;

If not, the irregular sleep model is used to combine the collected environmental parameters, characteristic data and current state data for lighting control.

Further, as shown in FIG. 5, the data processing method provided in this example may include the following steps when adjusting for irregular sleep:

According to the current state data, determine whether the target user is in a state of lack of sleep;

If not, control the light emitting device to emit light that inhibits sleep;

If yes, determine whether the current time zone of the target user is within the sleep time range according to the location of the target user; the sleep time range may be the insomnia time range of regular sleep of the target user throughout the day;

If so, control the light emitting device to emit lights that promote sleep;

If not, determine whether the level of sleeplessness is greater than the predetermined level;

If it is greater than the predetermined level, the light-emitting device is controlled to emit light that promotes sleep within a predetermined time, and after the light that emits sleep is equal to the predetermined time, the light-emitting device is controlled to emit light to suppress sleep;

If it is not greater than the predetermined level, the light emitting device is controlled to emit a light that suppresses sleep.

Determine that the level of lack of sleep of the target user is greater than the predetermined level, according to whether the duration of the target user's continuous non-sleep has reached the length corresponding to the predetermined level; or, the current state of the target user is characterized by the lack of sleep. The sign status corresponding to the level. In short, there are many ways to determine whether the target user's lack of sleep level is greater than a predetermined level, which is not limited to any of the above. As shown in FIG. 6, this embodiment provides an electronic device, including:

Memory, used for information storage;

A processor, connected to the memory, for implementing the method provided by one or more of the foregoing technical solutions by executing computer-executable instructions stored in the memory, for example, FIG. 1, FIG. 2, FIG. 4, FIG. 5, and FIG. 5 One or more of the methods shown. In some embodiments, the electronic device further includes: a communication interface and/or a human-machine interaction interface. The communication interface may include: a transceiver antenna and/or a network interface, which may be used for information interaction with other electronic devices. The human-computer interaction interface may be used to interact with humans, and the human-computer interaction interface may include: physical buttons and/or touch screens.

This embodiment provides a computer storage medium for storing computer-executable instructions; after the computer-executable instructions are executed by a processor, the method provided by one or more of the foregoing technical solutions can be implemented, for example, One or more of the methods shown in FIGS. 1, 2, 4, and 5.

In the several embodiments provided by the present disclosure, it should be understood that the disclosed device and method may be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a division of logical functions. In actual implementation, there may be other divisions, for example, multiple units or components may be combined, or Can be integrated into another system, or some features can be ignored, or not implemented. In addition, the coupling or direct coupling or communication connection between the displayed or discussed components may be through some interfaces, and the indirect coupling or communication connection of the device or unit may be electrical, mechanical, or other forms of.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units; Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, the functional units in the embodiments of the present disclosure may all be integrated into one processing module, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above integration The unit can be implemented in the form of hardware, or in the form of hardware plus software functional units.

Those of ordinary skill in the art may understand that all or part of the steps to implement the above method embodiments may be completed by program instructions related hardware. The foregoing program may be stored in a computer-readable storage medium, and when the program is executed, Including the steps of the above method embodiments; and the aforementioned storage media include: mobile storage devices, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks or optical disks, etc. A medium that can store program codes.

The above are only specific implementations of the present disclosure, but the scope of protection of the present disclosure is not limited to this, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present disclosure. It should be covered by the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.

Claims

A data processing method, including:

Obtain environmental data;

Obtain the current status data and characteristic data of the target user;

According to the current state data, select a regular sleep model or an irregular sleep model as the target model;

Input the environmental data and the characteristic data into the selected target model to obtain lighting control parameters;

Using the light control parameters to control the light emission of the light-emitting device, wherein controlling the light emission of the light-emitting device includes: controlling the light-emitting device to emit light that promotes the sleep of the target user, or controlling the light-emitting device to suppress the emission The light that the target user sleeps.
The method according to claim 1, wherein the selecting a regular sleep model or an irregular sleep model as the target model according to the current sleep state of the target user includes at least one of the following:

If the current state data indicates that the current sleep state of the target user does not conform to the sleep rule corresponding to the regular sleep model, the irregular sleep model is selected as the target model;

If the current state data indicates that the current sleep state of the target user conforms to the sleep rule corresponding to the regular sleep model, the regular sleep model is selected as the target model.
The method according to claim 1 or 2, wherein the method further comprises:

If the target model is a non-sleep regular model, acquire irregular sleep data of the target user;

The inputting the environmental data and the characteristic data into the selected target model to obtain the lighting control parameters includes:

The environmental data, the characteristic data and the irregular sleep data are input into the irregular sleep model to obtain the lighting control parameters.
The method of claim 3, wherein the irregular sleep data includes at least one of the following:

Sleep time deviation data into sleep;

Time zone deviation data of the target user's time zone;

Continuous state data of this irregular sleep;

Data on the frequency of irregular sleep.
The method according to claim 1, wherein

The acquiring environment data includes at least one of the following:

Get current season data;

Obtain the lighting data of the space where the current target user is located;

Obtain the temperature data of the space where the current target user is located;

and / or,

The acquiring characteristic data of the target user includes:

Obtaining the static characteristic data of the user;

Obtain the dynamic feature data of the user.
The method according to claim 5, wherein the acquiring dynamic characteristic data of the user includes at least one of the following:

Collecting current motion feature data of the target user;

Collect the current physical characteristic data of the target user.
The method according to claim 1 or 2, wherein the method further comprises:

Performing a first denoising process on the feature data, removing the first noise data outside the preset frequency range and obtaining the first feature data;

Performing correlation analysis filtering on the first feature data to remove second noise data located in the preset frequency range and obtain second feature data;

Parsing the second characteristic data to obtain third characteristic data characterizing the state of the target user;

The inputting the environmental data and the characteristic data into the selected target model to obtain the lighting control parameters includes:

The third characteristic data and the environment data are input to the target model of the target user to obtain the lighting control parameter.
The method according to claim 7, wherein the parsing the second feature data to obtain third feature data characterizing the target user state includes:

Parsing the second feature data, and extracting time-domain feature data of the target user from the second feature data, wherein the time-domain feature data includes: a characteristic curve corresponding to the second feature data in the time domain Peak data and/or trough data;

Peak data and/or trough data of the target user's action curve or sign change curve;

and / or,

Parsing the second feature data, extracting the frequency domain feature data of the target user from the second feature data; wherein, the frequency domain feature data includes: a characteristic curve corresponding to the second feature data in the frequency domain Peak data and/or trough data.
The method according to claim 7, wherein

The first noise data includes: jitter data of the acquisition device whose jitter frequency is outside the preset frequency range, electromagnetic interference data whose electromagnetic frequency is outside the preset frequency range, and electromagnetic frequency data outside the preset frequency range Magnetic field noise

and / or,

The second noise data includes: jitter data whose jitter frequency of the acquisition device is within the preset frequency range.
The method according to claim 7, wherein the inputting the environmental data and the characteristic data into the selected target model to obtain the light control parameters includes:

According to the dimensionality reduction processing strategy, performing dimensionality reduction processing on the third feature data and the environmental data to obtain input data of a preset dimension;

Inputting input data of a preset dimension into the target model to obtain the lighting control parameter.
The method according to claim 10, wherein the performing dimension reduction processing on the third feature data and environment data according to the dimension reduction processing strategy to obtain input data of a preset dimension includes:

Determine whether the target user is in an overall static state based on the first preset condition and the motion feature data of the target user;

If the target user is in an action state, determine whether the target user is in the first type of partial static state according to the action feature data;

If the target user is not in the first type of local stationary state, determine whether the target user is in the second type of local stationary state according to the motion feature data;

If the target user is in an action state, the target user's action feature data is sampled according to the target user's current action state to obtain sampling feature data as the input data.
The method according to claim 11, wherein said performing dimension reduction processing on said third feature data and environment data according to a dimension reduction processing strategy to obtain input data of a preset dimension, further comprising:

If the target user is in the overall static state, the determination of whether the target user is in the first type of partial static state and the second type of partial static state is stopped;

and / or,

If the target user is in the first type of local stationary state, the determination of whether the target user is in the second type of local stationary state is stopped.
The method according to claim 1 or 2, wherein after the use of the light control parameter to control the light emission of the light emitting device, the method further comprises:

Obtain lighting control effect data, wherein the effect data includes: at least one of the target user's sleep effect data, sleep activity data, and non-sleep activity data;

According to the effect data, the regular sleep model and/or the irregular sleep model are optimized.
A data processing device, including:

The first obtaining module is configured to obtain environmental data;

The second obtaining module is configured to obtain the current state data and characteristic data of the target user;

A selection module configured to select a regular sleep model or an irregular sleep model as the target model based on the current state data;

A third acquisition module configured to input the environmental data and the characteristic data into the selected target model to obtain lighting control parameters;

The control module is configured to use the light control parameters to control the light emission of the light emitting device, wherein the controlling the light emission of the light emitting device includes: controlling the light emitting device to emit light that promotes the sleep of the target user, or controlling the light emission The device emits light that suppresses the sleep of the target user.
An electronic device, including:

Memory

A processor, connected to the memory, is configured to implement the data processing method provided in any one of claims 1 to 13 by executing computer-executable instructions located on the memory.
A computer storage medium storing computer-executable instructions; after being executed, the computer-executable instructions can implement the data processing method provided in any one of claims 1 to 13.