Device control method, apparatus, device and system based on behavior monitoring

Abstract

This application relates to a device control method, apparatus, device, and system based on behavior monitoring. The device control method based on behavior monitoring includes: extracting features from behavior monitoring data of a target object to obtain behavior features; performing a first behavior evaluation process on the behavior features to obtain a first evaluation result; performing a second behavior evaluation process on the behavior features to obtain a second evaluation result; identifying the target behavior state of the target object based on the first and second evaluation results; and adjusting the device in the space where the target object is located based on the target behavior state. This solution can improve the intelligence level of the device control process.

Description

Technical Field

[0001] This application relates to the field of human behavior monitoring technology, and in particular to a device control method, apparatus, equipment and system based on behavior monitoring. Background Technology

[0002] With the continuous development of technology and the continuous improvement of people's living standards, smart devices such as smart homes are increasingly entering people's lives. These devices can reduce user intervention to a certain extent, thereby providing convenience for users.

[0003] However, the operation of the aforementioned devices usually relies on pre-set fixed programs and still requires some manual operation by the user, so their level of intelligence is insufficient. Summary of the Invention

[0004] Therefore, it is necessary to provide a device control method, apparatus, device, and system based on behavior monitoring that can improve the intelligence level of smart devices, addressing the aforementioned technical problems.

[0005] One embodiment of this application provides a device control method based on behavior monitoring, the method comprising:

[0006] Feature extraction is performed on the behavior monitoring data of the target object to obtain behavioral features;

[0007] The behavioral characteristics are subjected to a first behavioral evaluation process to obtain a first evaluation result;

[0008] The behavioral characteristics are subjected to a second behavioral evaluation process to obtain a second evaluation result;

[0009] Based on the first evaluation result and the second evaluation result, the target behavior state of the target object is identified, and the equipment in the space where the target object is located is adjusted based on the target behavior state.

[0010] Another embodiment of this application provides a device control apparatus based on behavior monitoring, the apparatus comprising:

[0011] The feature extraction module is used to extract features from the behavior monitoring data of the target object to obtain behavioral features;

[0012] The first evaluation module is used to perform a first behavior evaluation process on the behavioral characteristics to obtain a first evaluation result;

[0013] The second evaluation module is used to perform a second behavioral evaluation process on the behavioral characteristics to obtain a second evaluation result.

[0014] The state recognition module is used to identify the target behavior state of the target object based on the first evaluation result and the second evaluation result, and to adjust the device in the space where the target object is located based on the target behavior state.

[0015] In one embodiment, when the first evaluation module performs a first behavior evaluation process on the behavioral features to obtain a first evaluation result, it is specifically used for:

[0016] The behavioral characteristics of multiple dimensions are evaluated using a first-behavior evaluation strategy to obtain the first-behavior evaluation results for each dimension.

[0017] Based on the first processing result and the first self-learning parameter, a first evaluation result corresponding to the first behavioral state is determined; the first self-learning parameter is determined based on historical data corresponding to the first behavioral state.

[0018] In one embodiment, when the first evaluation module determines the first evaluation result corresponding to the first behavioral state based on the first processing result and the first self-learning parameters, it is specifically used for:

[0019] Obtain the first self-learning parameter; the first self-learning parameter is determined based on the first time interval corresponding to the first behavior state in historical data;

[0020] The first processing result, the first behavior weight, and the first self-learning parameter are fused together to obtain the first evaluation result corresponding to the first behavior state.

[0021] In one embodiment, when the second evaluation module performs a second behavior evaluation process on the behavioral features to obtain a second evaluation result, it is specifically used for:

[0022] The behavioral characteristics of multiple dimensions are evaluated using a second behavioral assessment strategy to obtain the second processing results corresponding to each dimension.

[0023] Based on the second processing result and the second self-learning parameter, a second evaluation result corresponding to the second behavioral state is determined; the second self-learning parameter is determined based on the historical data corresponding to the second behavioral state.

[0024] In one embodiment, when the second evaluation module determines the second evaluation result corresponding to the second behavioral state based on the second processing result and the second self-learning parameters, it is specifically used for:

[0025] Obtain the second self-learning parameter; the second self-learning parameter is determined based on the first time interval corresponding to the first behavioral state in historical data;

[0026] The second processing result, the second behavior weight, and the second self-learning parameter are fused together to obtain the second evaluation result corresponding to the second behavior state.

[0027] In one embodiment, the behavioral monitoring data includes heart rate data, respiratory rate data, and body movement data;

[0028] When the feature extraction module extracts features from the behavior monitoring data of the target object to obtain behavioral features, it is specifically used for:

[0029] Acquire heart rate, respiratory rate, and body movement data within the same time interval;

[0030] Feature extraction was performed on heart rate data, respiratory rate data, and body movement data according to time series to obtain behavioral features corresponding to each dimension of heart rate, respiratory rate, and body movement.

[0031] In one embodiment, the target behavior state includes a first behavior state and a second behavior state. When the state recognition module adjusts the device in the space where the target object is located based on the target behavior state, it is specifically used for:

[0032] If the behavior changes from the first behavior state to the second behavior state, the device in the space where the target object is located executes the control command corresponding to the scene mode of the second behavior state;

[0033] If the behavior changes from the second behavior state to the first behavior state, the device in the space where the target object is located executes the control command corresponding to the scene mode of the first behavior state.

[0034] In one embodiment, if the behavior changes from a first behavior state to a second behavior state, when the state recognition module executes control commands corresponding to the scene mode of the second behavior state on the device in the space where the target object is located, it is specifically used for:

[0035] If the behavior changes from the first behavior state to the second behavior state, and the current time is within the first preset time period, the device in the space where the target object is located executes the control command corresponding to the scene mode of the second behavior state.

[0036] If the behavior changes from the second behavior state to the first behavior state, when the state recognition module executes the control command corresponding to the scene mode of the first behavior state in the space where the target object is located, it is specifically used for:

[0037] If the behavior changes from the second behavior state to the first behavior state, and the current time is within the second preset time period, the device in the space where the target object is located executes the control command corresponding to the scene mode of the first behavior state.

[0038] In one embodiment, the first behavioral state includes an awake state, the second behavioral state includes a asleep state, and the device further includes:

[0039] The determination module is used to determine the sleep onset point and the wakefulness point based on the awake state and the sleep state; the sleep onset point is the time point at which the awake state changes to the sleep state, and the wakefulness point is the time point at which the sleep state changes back to the awake state;

[0040] If the behavior changes from the first behavior state to the second behavior state, when the state recognition module executes the control command corresponding to the scene mode of the second behavior state in the space where the target object is located, it is specifically used for:

[0041] If the current behavior state is at the point of falling asleep, then control the devices in the space where the target object is located to execute the control command for the sleep mode;

[0042] If the behavior changes from the second behavior state to the first behavior state, when the state recognition module executes the control command corresponding to the scene mode of the first behavior state in the space where the target object is located, it is specifically used for:

[0043] If the current behavioral state is at the waking point, then control the devices in the space where the target object is located to execute the control command for the wake-up mode.

[0044] Another aspect of this application provides a computer device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it at least implements the steps of the device control method based on behavior monitoring of the various embodiments of this application.

[0045] Another embodiment of this application provides a device control system based on behavior monitoring, the system comprising: a data acquisition device and a computer device; the data acquisition device and the computer device are communicatively connected.

[0046] The data acquisition device is installed in the monitoring area or worn by the target object located in the monitoring area, and is used to collect behavioral monitoring data of the target object;

[0047] The computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements at least the steps of the device control method based on behavior monitoring according to the embodiments of this application.

[0048] Another aspect of this application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, at least implements the steps of the behavior monitoring-based device control method of the various embodiments of this application.

[0049] A computer program product or computer program includes computer instructions stored in a computer-readable storage medium; a processor of a computer device reads the computer instructions from the computer-readable storage medium, and when the processor executes the computer instructions, it implements the steps in the behavior monitoring-based device control method of various embodiments of this application.

[0050] The aforementioned device control method, apparatus, equipment, and system based on behavior monitoring extracts features from the behavior monitoring data of a target object to obtain behavioral characteristics; performs a first behavior evaluation process on the behavioral characteristics to obtain a first evaluation result, and performs a second behavior evaluation process on the behavioral characteristics to obtain a second evaluation result; and identifies the target behavioral state of the target object based on the first and second evaluation results, and adjusts the equipment in the space where the target object is located based on the target behavioral state. This setup, through two different evaluation processing methods (first and second evaluation processes), accurately assesses the target behavioral state of the target object, and then rationally adjusts the equipment in the space based on the target behavioral state of the target object without user intervention, thus improving the intelligence of the device control process. Attached Figure Description

[0051] Figure 1 This is a diagram illustrating the implementation environment of a device control method based on behavior monitoring in one embodiment.

[0052] Figure 2 This is a hardware structure block diagram of a gateway in one embodiment.

[0053] Figure 3 This is a flowchart illustrating a device control method based on behavior monitoring in one embodiment.

[0054] Figure 4 for Figure 3 A flowchart illustrating a specific implementation method of step S302.

[0055] Figure 5 for Figure 3 A flowchart illustrating a specific implementation method of step S303.

[0056] Figure 6 This is a structural block diagram of a device control apparatus based on behavior monitoring in one embodiment.

[0057] Figure 7 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0059] Application Overview

[0060] The development of smart devices has greatly facilitated people's lives. Taking smart home devices as an example, they connect various devices in a user's home, such as audio-visual equipment, lighting systems, curtain controls, and air conditioning controls, through Internet of Things (IoT) technology. This allows them to provide various functions and methods, including appliance control, lighting control, indoor and outdoor remote control, and HVAC control. However, traditional control schemes typically achieve automatic control of smart devices based on factory-preset or user-defined automatic execution programs. User intervention is still required when necessary, and they cannot dynamically adjust the device's operating status and parameters based on changes in the user's behavior, meaning their level of intelligence is insufficient.

[0061] To address the aforementioned problems, this application proposes a device control method, apparatus, device, and system based on behavior monitoring that can improve the intelligence level of smart devices. This method accurately identifies the behavioral state of a target object (i.e., the user), and then accurately adjusts the operating state and parameters of the smart device based on this behavioral state. The following examples provide a non-limiting description of the specific implementation scheme.

[0062] Exemplary Implementation Environment

[0063] The device control method based on behavior monitoring provided in this application can be applied to, for example... Figure 1 In the implementation environment shown, one or more data acquisition devices 110 are connected to a gateway 120 via ZIGBEE or similar methods. The gateway 120 is connected to a router 130 via Wi-Fi or similar methods. The router 130 is connected to a cloud server 140 via Wi-Fi or similar methods. Furthermore, in some scenarios, the router 130 is also connected to a terminal device 150 via Wi-Fi or similar methods. The terminal device 150 is also connected to the cloud server 140 via 2G / 3G / 4G / 5G / Wi-Fi or similar methods. It should be understood that the communication connections between the devices are not limited to the methods listed above and can be adjusted as needed. After the devices are interconnected, communication can be achieved, including but not limited to data exchange and command transmission.

[0064] The data collection device 110 can be a smart wearable device, such as a smart bracelet, or a non-wearable device, such as a smart sleep belt, a smart mattress, or millimeter-wave radar. It is used to collect user behavior monitoring data and send it to the gateway 120 for further processing, or the gateway 120 forwards it to the cloud server 140 via the router 130 for further processing, thereby enabling the monitoring of the user's behavioral state. The user's behavior monitoring data includes, but is not limited to, heartbeat signals, breathing signals, electroencephalogram (EEG) signals, and body movements (referred to as body movements), which enable the identification and differentiation of the user's behavioral state.

[0065] Figure 2 This is a hardware structure block diagram of a gateway according to an exemplary embodiment. This gateway 120 is suitable for... Figure 1 The implementation environment is shown.

[0066] It should be noted that this gateway is merely an example adapted to this invention and should not be construed as providing any limitation on the scope of use of this invention. Furthermore, this gateway should not be interpreted as requiring or depending on any specific feature. Figure 2 One or more components of the exemplary gateway 120 shown.

[0067] The hardware structure of the gateway 120 can vary significantly due to differences in configuration or performance, such as Figure 2 As shown, the gateway 120 includes: a power supply 1210, an interface 1230, at least one memory 1250, and at least one central processing unit (CPU) 1270.

[0068] The power supply 1210 is used to provide operating voltage for the various hardware devices on the gateway 120.

[0069] Interface 1230 includes at least one wired or wireless network interface 1231, at least one serial-to-parallel conversion interface 1233, at least one input / output interface 1235, and at least one USB interface 1237, etc., for communicating with external devices.

[0070] The memory 1250 serves as a resource storage medium and can be a read-only memory, random access memory, disk, or optical disk, etc. The resources stored on it include an operating system 1251, application programs 1253, or data 1255, and the storage method can be temporary or permanent storage. The operating system 1251 manages and controls the various hardware devices on the gateway 120 and the application programs 1253 to enable the central processing unit 1270 to perform calculations and processing on massive amounts of data 1255. It can be Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, FreeRTOS, etc. The application programs 1253 are computer programs that perform at least one specific task based on the operating system 1251, and may include at least one module (…). Figure 2 (Not shown in the image), each module can contain a series of computer-readable instructions for gateway 120. Data 1255 can be photos, images, etc., stored on a disk.

[0071] The central processing unit 1270 may include one or more processors and is configured to communicate with the memory 1250 via a bus for computing and processing massive amounts of data 1255 in the memory 1250.

[0072] As described in detail above, the gateway 120 to which this invention is applied will perform the corresponding processing method by reading a series of computer-readable instructions stored in the memory 1250 by the central processing unit 1270.

[0073] Furthermore, the present invention can also be implemented through hardware circuits or hardware circuits combined with software instructions. Therefore, the implementation of the present invention is not limited to any specific hardware circuit, software, or combination thereof.

[0074] Exemplary methods

[0075] Please see Figure 3 In one exemplary embodiment, a device control method based on behavior monitoring is provided, applicable to... Figure 1 The implementation environment shown includes electronic devices such as gateways or cloud servers. The method is illustrated using an electronic device as an example; the electronic device can be... Figure 1 The device control method based on behavior monitoring can be executed by either the gateway or the cloud server, depending on the requirements for processing speed. The method may include the following steps:

[0076] Step S301: The electronic device extracts features from the behavior monitoring data of the target object to obtain behavior features.

[0077] The target object refers to any object whose behavior needs to be monitored. Specifically, it can be an object with biological characteristics, such as an object with heart rate, respiration, or other biological characteristics. For example, the target object can be a human body, an animal, etc. The following embodiments use the human body as an example for illustration.

[0078] Among these, the behavioral monitoring data of the target object refers to data that can characterize the differences between different behavioral states of the target object. For example, when the target object is running, walking, sitting, lying down, or sleeping, its heart rate, respiratory rate, and number of body movements will vary to varying degrees. Therefore, depending on the actual application scenario, one or more of these data can be collected as the target object's behavioral monitoring data. It is understood that, in addition to heart rate, respiratory rate, and body movement data, behavioral monitoring data can also include other types of data, such as electroencephalogram (EEG) data, which will not be listed here.

[0079] Furthermore, depending on the type of behavior monitoring data collected, the data can be obtained from wearable devices worn by the target individual, or from non-wearable devices that are fixed or mobile. In some embodiments, since some wearable devices may have a certain impact on the user's behavior to be monitored, it is preferable to use non-wearable devices to collect the target individual's behavior monitoring data, taking this factor into consideration.

[0080] It should be understood that the ultimate purpose of this embodiment is to adjust the devices within the space where the target object is located. Therefore, the behavior monitoring data described in this embodiment can typically be behavior monitoring data that occurs when the target object is located in the monitoring area. The monitoring area corresponds to the space where the target object is located, but is not necessarily identical. For example, the monitoring area can be a specific area within the space where the target object is located, or it can be a specific area outside the space. For instance, the space where the target object is located can be the entire area of ​​the target object's home, while the detection area can only include the bedroom area.

[0081] In addition, taking the gateway as an example, the process of acquiring user behavior monitoring data can be real-time, such as the collection device reporting the collected behavior monitoring data to the gateway after each collection, or it can be acquired at certain intervals, in which case the total behavior monitoring data acquired over a period of time is obtained.

[0082] After acquiring the behavior monitoring data of the target object, the electronic device extracts features using appropriate feature extraction methods to obtain the behavioral features corresponding to the behavior monitoring data. These extracted behavioral features highlight certain representative characteristics of the behavior monitoring data and can be represented more easily, directly, or obviously than the behavior monitoring data itself, thus facilitating more effective differentiation of different behavioral states based on the extracted features.

[0083] Specifically, in some embodiments, behavioral features can be characteristics used to represent the stability and trend of behavioral monitoring data. Stability, also known as data volatility, is an indicator that measures the volatility and dispersion of data; the smaller the data volatility and the lower the dispersion, the higher the stability. Trend refers to the overall trend of the data increasing or decreasing. Through behavioral features, the overall changes in user behavioral monitoring data can be determined.

[0084] To facilitate understanding, let's take behavioral monitoring data, including heart rate and respiratory rate data, as an example. When a user is exercising, as the exercise continues, their heart rate and respiration will generally increase, but they will also fluctuate significantly, meaning the stability of heart rate and respiratory rate data is relatively poor. Conversely, when a user is sitting, their heart rate and respiration will be slower and more even, indicating better stability. And when a user changes from an active to a sitting state, their heart rate and respiration will gradually decrease, showing an overall decreasing trend in heart rate and respiratory rate data. Therefore, by observing the stability and trends of behavioral monitoring data, we can at least to some extent determine a user's behavioral state and changes in that state.

[0085] More specifically, in some embodiments, behavioral characteristics may include one or more of the following: first derivative, standard deviation, and range. The first derivative can be used to determine the trend of data change, the standard deviation can be used to determine the dispersion of data, and the range can be used to determine the magnitude of data fluctuation. Of course, behavioral characteristics may also include features of other dimensions; any feature capable of characterizing behavioral monitoring data is acceptable, and there is no limitation on this.

[0086] In addition, in some embodiments, if the behavior monitoring data includes heart rate data, respiratory rate data, and body movement data, then step S301 (extracting features from the behavior monitoring data of the target object to obtain behavior features) may specifically include: acquiring heart rate data, respiratory rate data, and body movement data in the same time interval; extracting features from the heart rate data, respiratory rate data, and body movement data according to the time series to obtain the behavior features corresponding to each dimension of heart rate, respiratory rate, and body movement.

[0087] Specifically, by acquiring behavioral monitoring data such as heart rate, respiratory rate, and body movement data within the same time interval, it can be ensured that the behavioral monitoring data is obtained based on the user's consistent behavioral state. The length of this time interval can be selected based on factors such as the frequency of changes in the user's behavioral state and the ease of subsequent processing by the electronic device; for example, 15 minutes can be chosen. For instance, the more frequently the user's behavioral state changes, the shorter the selected time interval should be. A shorter time interval results in a shorter processing time per operation by the electronic device, but also requires more processing iterations when monitoring behavioral states over a longer period.

[0088] Furthermore, to ensure that the extracted behavioral features accurately represent the target object's behavioral state, feature extraction from behavioral monitoring data such as heart rate, respiratory rate, and body movement data must be performed according to a time series, i.e., in chronological order of the occurrence of the behavioral monitoring data. Taking heart rate data as an example, assuming the selected time interval includes times t1, t2, t3, ..., tn; and the heart rate at time t1 is h1, at time t2 is h2, at time t3 is h3, ..., at time tn is hn, then the time series corresponding to the heart rate data can be represented as (h1, h2, h3, ..., hn). Features can then be extracted based on the time series corresponding to the heart rate data to obtain the corresponding behavioral features, such as calculating the first derivative, standard deviation, and range based on the time series.

[0089] Furthermore, in practical applications, some data acquisition devices collect data approximately every 10 seconds. This means that in the time series corresponding to the behavior monitoring data obtained in the aforementioned steps, the time difference between two adjacent data points is generally around 10 seconds. If we assume the time series corresponding to the behavior monitoring data contains 15 minutes of data, then each time series contains approximately 90 data points. Extracting behavioral features based on these 90 data points would take considerable time, hindering the overall solution's rapid processing. To address this issue, some embodiments can further divide the data in the time series, for example, dividing the entire time series into at least two time intervals, calculating the mean of the behavior monitoring data in each time interval, and then constructing a new time series, which can be called the target time series.

[0090] To facilitate understanding, let's continue using heart rate data as an example. Assume the selected time series corresponding to the heart rate data is (h1, h2, h3, h4, h5, h6, h7, h8), and the time difference between two adjacent data points is 30 seconds. We can then divide the time into 1-minute intervals and calculate the average. The resulting target time series for the heart rate data can be represented as ((h1+h2) / 2, (h3+h4) / 2, (h5+h6) / 2, (h7+h8) / 2). It is evident that the target time series for the heart rate data obtained through this processing contains less data than the original time series. Therefore, fewer operations are required for subsequent behavioral feature extraction, the time required is shorter, and the configuration requirements for processing equipment are reduced.

[0091] Step S302: The electronic device performs a first behavior evaluation process on the behavioral characteristics to obtain a first evaluation result.

[0092] The first behavior assessment process is used to determine whether a behavioral characteristic corresponds to a first behavior, or to determine the probability that a behavioral characteristic corresponds to a first behavior, and so on. Accordingly, the obtained first assessment result can be in the form of numerical values ​​such as scores or probabilities, or it can be in other forms, such as comments. Any reasonable method is acceptable, and this application does not impose any restrictions on it.

[0093] In some feasible embodiments, such as Figure 4 As shown, step S302 (the electronic device performs a first behavior evaluation process on the behavioral characteristics to obtain a first evaluation result) may specifically include:

[0094] Step S3021: The electronic device performs first behavior evaluation processing on the behavioral features of multiple dimensions through the first behavior evaluation strategy, and obtains the first processing results corresponding to each dimension.

[0095] Specifically, the first behavior assessment strategy can include the processing flow or algorithm that can be used when performing the first behavior assessment, and it can be pre-set according to the needs of the actual application scenario. The multiple dimensions of behavioral characteristics include behavioral characteristics such as first derivative, standard deviation, and range. For example, the first behavior assessment strategy can include strategies for calculating the first derivative, standard deviation, and range based on behavior monitoring data, as well as the calculation formulas that can be used in the calculation process.

[0096] Furthermore, since it is usually difficult to comprehensively represent all the features of behavior monitoring data when using only a single dimension of behavioral features, this step performs a first behavior assessment on the behavioral features of multiple dimensions and obtains the first processing results corresponding to each dimension, which facilitates comprehensive consideration in subsequent steps.

[0097] Step S3022: The electronic device determines a first evaluation result corresponding to the first behavioral state based on the first processing result and the first self-learning parameter; the first self-learning parameter is determined based on historical data corresponding to the first behavioral state.

[0098] Specifically, after obtaining the first processing results for each dimension in the previous step, considering the differences in behavioral states among different target objects—for example, when the behavioral states to be identified include falling asleep and waking up—the sleep habits of different users of different ages, genders, occupations, and sleep environments will vary significantly. Therefore, this step introduces a first self-learning parameter, which is a parameter obtained by "learning" the characteristics unique to the behavioral monitoring data of the target object. Based on this, the first self-learning parameter can be determined based on the historical data corresponding to the first behavioral state of the target object.

[0099] In this step, after comprehensively processing the first processing result and the first self-learning parameters, it is determined that the obtained first evaluation result will be more in line with the unique characteristics of the target object itself.

[0100] More specifically, the historical data corresponding to the first state can include, but is not limited to, the start time (segment), duration, end time (segment), frequency of occurrence, and probability of occurrence, etc. There are no restrictions on this, and any data that can reflect the differences between different target objects can be used. Furthermore, since the results of "self-learning" usually require a sufficient amount of data to be relatively accurate, the first self-learning parameter can be obtained by comprehensively considering historical data within a certain time period.

[0101] Furthermore, in some embodiments, the specific implementation method of step S3022 may include: obtaining a first self-learning parameter; the first self-learning parameter is determined based on a first time interval corresponding to the first behavior state in historical data; and the first processing result, the first behavior weight, and the first self-learning parameter are fused to obtain a first evaluation result corresponding to the first behavior state.

[0102] Specifically, in this embodiment, the first self-learning parameter is determined based on the first time interval corresponding to the first behavioral state in historical data. The first time interval can be the time interval where the first behavioral state occurs most frequently, obtained through statistical analysis of historical data; that is, it can be the time interval in which the target object is most likely to be in the first behavioral state. This first time interval can better reflect the characteristics of the target object's first behavioral state. Correspondingly, if the current time is within the first time interval, the target object is more likely to be in the first behavioral state compared to when the current time is not within the first time interval, so the weight of the first self-learning parameter in the first evaluation result will be relatively increased. Conversely, if the current time is not within the first time interval, the target object is less likely to be in the first behavioral state, so the weight of the first self-learning parameter in the first evaluation result will be relatively decreased. Of course, it should be understood that the first time interval can also be other time intervals obtained through other methods, and the first self-learning parameter can be adjusted accordingly; there are no restrictions on this.

[0103] Furthermore, after obtaining the first processing results corresponding to the behavioral features of each dimension in the aforementioned steps, considering that the ability to characterize the features of behavioral monitoring data may differ between different dimensions of behavioral features, in this embodiment, when fusing the first processing results and the first self-learning parameters, a first behavioral weight is also introduced. That is, according to the actual situation, different weights are assigned to the first processing results corresponding to different dimensions of behavioral features, and then the processing results combined with the weights are fused with the first self-learning parameters, so that the final first evaluation result is more accurate.

[0104] For example, the process of calculating the first evaluation result corresponding to the first behavioral state can be expressed by the formula:

[0105] P1 = x1 × a1 + x2 × a2 + x3 × a3 + z1

[0106] In the formula, P1 is the first evaluation result, x1, x2 and x3 are the first processing results corresponding to different dimensions of behavioral features, a1, a2 and a3 are the weights of the first processing results corresponding to different dimensions of behavioral features, a1, a2 and a3 are all first behavioral weights, and a1+a2+a3=1, z1 is the first self-learning parameter.

[0107] Step S303: The electronic device performs a second behavior evaluation process on the behavioral characteristics to obtain a second evaluation result.

[0108] Similar to the first behavior assessment process, the second behavior assessment process is used to determine whether a behavior feature corresponds to a second behavior, or to determine the probability that a behavior feature corresponds to a second behavior, etc. Accordingly, the second assessment result can be in the form of a score, probability, or other numerical values, or it can be in other forms, such as comments. Any reasonable method is acceptable, and this application does not impose any restrictions on it.

[0109] In some feasible embodiments, such as Figure 5 As shown, step S303 (the electronic device performs a second behavior evaluation process on the behavioral characteristics to obtain a second evaluation result) may specifically include:

[0110] Step S3031: The electronic device performs second behavior evaluation processing on the behavioral features of multiple dimensions through the second behavior evaluation strategy, and obtains the second processing results corresponding to each dimension.

[0111] Specifically, the second behavior assessment strategy may include processing procedures or algorithms that can be used when conducting the second behavior assessment, and these can be pre-set according to the needs of the actual application scenario. The multiple dimensions of behavioral characteristics include behavioral features such as first derivative, standard deviation, and range. For example, the second behavior assessment strategy may include strategies for calculating the first derivative, standard deviation, and range based on behavior monitoring data, as well as the calculation formulas that can be used in the calculation process.

[0112] Furthermore, since it is usually difficult to comprehensively represent all the features of behavior monitoring data when using only a single dimension of behavioral features, this step performs a second behavioral evaluation on the behavioral features of multiple dimensions and obtains the second processing results corresponding to each dimension, which facilitates comprehensive consideration in subsequent steps.

[0113] Step S3032: The electronic device determines a second evaluation result corresponding to the second behavioral state based on the second processing result and the second self-learning parameter; the second self-learning parameter is determined based on historical data corresponding to the second behavioral state.

[0114] Specifically, after obtaining the second processing results corresponding to each dimension in the previous step, considering that the behavioral states of different target objects usually differ—for example, when the behavioral states to be identified include falling asleep and waking up—the sleep habits of different users under different age groups, genders, occupations, and sleep environments will vary significantly. Therefore, this step introduces a second self-learning parameter, which is a parameter obtained by "learning" the features unique to the behavioral monitoring data of the target object. Based on this, the second self-learning parameter can be determined based on the historical data corresponding to the second behavioral state of the target object.

[0115] In this step, the second processing result and the second self-learning parameters are combined to determine that the obtained second evaluation result is more consistent with the unique characteristics of the target object.

[0116] More specifically, the historical data corresponding to the second behavioral state can include, but is not limited to, the start time (segment), duration, end time (segment), frequency of occurrence, and probability of occurrence, etc. There are no restrictions on this, and any data that can reflect the differences between different target objects can be used. Furthermore, since the results of "self-learning" usually require a sufficient amount of data to be relatively accurate, the second self-learning parameter can be obtained by comprehensively considering historical data within a certain time period.

[0117] Furthermore, in some embodiments, the specific implementation method of step S3032 may include: obtaining a second self-learning parameter; the second self-learning parameter is determined based on a second time interval corresponding to the second behavior state in historical data; and the second processing result, the second behavior weight, and the second self-learning parameter are fused to obtain a second evaluation result corresponding to the second behavior state.

[0118] Specifically, in this embodiment, the second self-learning parameter is determined based on the second time interval corresponding to the second behavioral state in historical data. The second time interval can be the time interval where the second behavioral state occurs most frequently, obtained through historical data statistics; that is, it can be the time interval where the target object is most likely to be in the second behavioral state. This second time interval can better reflect the characteristics of the target object's second behavioral state. Correspondingly, if the current time is within the second time interval, the target object is more likely to be in the second behavioral state compared to when the current time is not within the second time interval. Therefore, the weight of the second self-learning parameter in the second evaluation result will be relatively increased. Conversely, if the current time is not within the second time interval, the target object is less likely to be in the second behavioral state, so the weight of the second self-learning parameter in the second evaluation result will be relatively decreased. It should be understood that the second time interval can also be other time intervals obtained through other methods, and the second self-learning parameter can be adjusted accordingly; there are no restrictions on this.

[0119] Furthermore, after obtaining the second processing results corresponding to the behavioral features of each dimension in the aforementioned steps, considering that the ability to characterize the features of behavioral monitoring data may differ between different dimensions of behavioral features, in this embodiment, when fusing the second processing results and the second self-learning parameters, a second behavioral weight is also introduced. That is, according to the actual situation, different weights are assigned to the second processing results corresponding to different dimensions of behavioral features, and then the processing results combined with the weights are fused with the second self-learning parameters, so that the final second evaluation result is more accurate.

[0120] For example, the process of calculating the second evaluation result corresponding to the second behavioral state can be expressed by the formula:

[0121] P2 = y1×b1 + y2×b2 + y3×b3 + z2

[0122] In the formula, P2 is the second evaluation result, y1, y2 and y3 are the second processing results corresponding to different dimensions of behavioral features, b1, b2 and b3 are the weights of the second processing results corresponding to different dimensions of behavioral features, b1, b2 and b3 are all second behavioral weights, and b1+b2+b3=1, z2 is the second self-learning parameter.

[0123] Step S304: The electronic device identifies the target behavior state of the target object based on the first evaluation result and the second evaluation result, and adjusts the device in the space where the target object is located based on the target behavior state.

[0124] Specifically, after obtaining the first evaluation result and the second evaluation result in the aforementioned steps, the electronic device can determine the target behavior state corresponding to the target object based on the actual situation of the first evaluation result and the second evaluation result, such as scores or probabilities, and can further adjust the device in the space where the target object is located based on the target behavior state corresponding to the target object.

[0125] In some embodiments, the target behavior state includes a first behavior state and a second behavior state. Accordingly, the step of adjusting the device in the space where the target object is located based on the target behavior state may specifically include: if the behavior state changes from the first behavior state to the second behavior state, controlling the device in the space where the target object is located to execute a control command corresponding to the scene mode of the second behavior state; if the behavior state changes from the second behavior state to the first behavior state, controlling the device in the space where the target object is located to execute a control command corresponding to the scene mode of the first behavior state.

[0126] For example, assuming the first behavior state is a moving state and the second behavior state is a sitting state, when the target object changes from a moving state to a sitting state, the device in the space where the target object is located can be controlled to execute the control command corresponding to the scene mode of the sitting state. For example, since the heat dissipation required by the user will decrease after changing from a moving state to a sitting state, the target cooling temperature of the indoor air conditioner can be controlled to rise to the set value.

[0127] In this way, based on the behavioral state of the target object, the equipment in the space can be reasonably adjusted without user intervention, thus improving the intelligence of the equipment control process.

[0128] Based on the above scheme, in order to further improve the intelligence of control, the step of controlling the device in the space where the target object is located to execute the control command corresponding to the scene mode of the second behavior state if the behavior state changes from the first behavior state to the second behavior state can further include: if the behavior state changes from the first behavior state to the second behavior state and the current time is within a first preset time period, controlling the device in the space where the target object is located to execute the control command corresponding to the scene mode of the second behavior state.

[0129] Specifically, in addition to the prior scheme, this embodiment also judges the time when the behavior state changes. That is, if the current time is within the first preset time period, it indicates that the control conditions are met and the device in the space where the target object is located can be controlled to execute the corresponding control command; otherwise, the device is not controlled.

[0130] For example, if the target object changes from a moving state to a sitting state at noon, the ambient temperature may be very high. In this case, even if the user changes from a moving state to a sitting state, the required heat dissipation will still be high, so the target cooling temperature of the indoor air conditioner will not be increased. In other words, if the target object changes from a moving state to a sitting state at a time that is not within the first preset time period, the device will not be controlled.

[0131] Similarly, in some embodiments, the step of controlling the device in the space where the target object is located to execute the control command corresponding to the scene mode of the first behavior state if the behavior state changes from the second behavior state to the first behavior state may further include: if the behavior state changes from the second behavior state to the first behavior state and the current time is within a second preset time period, controlling the device in the space where the target object is located to execute the control command corresponding to the scene mode of the first behavior state.

[0132] The principle and purpose of this embodiment are similar to those of the previous embodiments, and will not be described in detail here.

[0133] Furthermore, in some embodiments, the first behavioral state includes an awake state, and the second behavioral state includes a sleep state. The control method described above may further include: determining a sleep point and an awake point based on the awake state and the sleep state; the sleep point is the time point at which the awake state changes to the sleep state, and the awake point is the time point at which the sleep state changes back to the awake state; correspondingly, if the behavior changes from the first behavioral state to the second behavioral state, controlling the device in the space where the target object is located to execute the control command corresponding to the scene mode of the second behavioral state may specifically include: if the current behavioral state is at the sleep point, then controlling the device in the space where the target object is located to execute the control command for the sleep mode; if the behavior changes from the second behavioral state to the first behavioral state, controlling the device in the space where the target object is located to execute the control command for the scene mode corresponding to the first behavioral state may specifically include: if the current behavioral state is at the awake point, then controlling the device in the space where the target object is located to execute the control command for the wake-up mode.

[0134] In this embodiment, after determining that the target object's target behavior state is awake or asleep, the sleep point and wake-up point are further determined, and the device in the space where the target object is located is controlled to execute a sleep mode or a wake-up mode based on the sleep point and wake-up point.

[0135] Furthermore, this embodiment can also combine the current time corresponding to the fall asleep point and the wake-up point to determine whether the control device should execute a sleep mode or a wake-up mode. For example, if the current time corresponding to the current wake-up point is between 7:30 and 8:00 (this time period can be set by the target object), it indicates that the current wake-up time is the target's corresponding wake-up time, and the device in the space can be controlled to execute the wake-up mode, such as controlling the lights to turn on, controlling the curtains to open, etc.

[0136] The aforementioned device control method based on behavior monitoring extracts behavioral features from the behavior monitoring data of the target object to obtain behavioral features; performs a first behavior evaluation process on the behavioral features to obtain a first evaluation result, and performs a second behavior evaluation process on the behavioral features to obtain a second evaluation result; and identifies the target behavioral state of the target object based on the first and second evaluation results, and adjusts the device in the space where the target object is located based on the target behavioral state. This setup, through the two different evaluation processing methods of the first and second evaluation processes, can accurately assess the target behavioral state of the target object, and then rationally adjust the device in the space based on the target behavioral state of the target object without user intervention, thus improving the intelligence of the device control process.

[0137] Furthermore, in some embodiments, the above method may also include: sending at least one of the target object's behavior monitoring data, a first evaluation result, and a second evaluation result to a preset device for display. The preset device is a terminal device, such as a user's mobile phone. Specifically, the data can be displayed in the form of text, numbers, images, and tables, allowing users to intuitively view their own behavioral state and changes in that state.

[0138] Exemplary application scenarios

[0139] This application also provides an application scenario in which the above-described device control method based on behavior monitoring is applied. Specifically, in this application scenario, the first behavioral state corresponds to a sleep state, and the second behavioral state corresponds to a wakeful state; that is, in this application scenario, the method used to determine whether the target object's target behavioral state is a sleep state or a wakeful state is used. Based on this, the application of the device control method based on behavior monitoring in this application scenario is as follows:

[0140] Step 1: Obtain the human respiratory rate sequence x, heart rate sequence y, and body movement count sequence move_nums from the sleep monitoring belt within the same cycle of n minutes (one cycle is from when the user lies down to when they get out of bed). Calculate the average heart rate and respiratory rate per minute (AVERAGE), forming a new heart rate sequence, i.e., the target heart rate sequence heart_rate (=AVERAGE(x / n)), and a new respiratory rate sequence, i.e., the target respiratory rate sequence breathalyze (=AVERAGE(y / n)). Body movement count refers to the number of times the body undergoes movement changes, such as turning over. The sleep monitoring belt can be placed on the user's bed to detect human heart rate, respiratory rate, and body movement count data imperceptibly while the user is lying on it during sleep.

[0141] It is understood that, in some embodiments, in addition to directly detecting human respiratory rate, heart rate, and body movement counts via a sleep tracking device, at least one other data detection device with the same function can be used to collect these data, such as fitness trackers or millimeter-wave radar. For example, fitness trackers can be used to collect heart rate and respiratory rate data, and millimeter-wave radar can be combined to detect human body movement data.

[0142] Step 2: Based on the heart rate sequence h_rate (heart_rate), extract the feature data h_diff, h_std, and h_sub representing heart rate changes. Based on the respiratory rate sequence b_rate (breathe_rate), extract the feature data b_diff, b_std, and b_sub representing respiratory rate changes. Calculate the total number of body movements m every 10 minutes.

[0143] Where, diff represents the first derivative, used to determine the trend of the sequence, that is, (the next number minus the previous number) divided by the mean (mean can be used to represent the mean, or simply the average); std represents the standard deviation, used to determine the stability of the sequence; sub represents the range, that is, the maximum value minus the minimum value (can be expressed as max-min), used to determine the stability of the sequence.

[0144] The specific calculation method is as follows:

[0145] Minimum heart rate per minute: h_min = np.min(h_rate) = MIN(h_rate); where np.min() is the function to calculate the minimum value in the Python programming language, and MIN() is the function to calculate the minimum value in other programming languages ​​(such as C++, the same below); h_min represents the minimum heart rate;

[0146] Maximum heart rate per minute: h_max = np.max(h_rate) = MAX(h_rate); where np.max() is the function in the Python programming language to calculate the maximum value, and MAX() is the function in other programming languages ​​to calculate the maximum value; h_max represents the maximum heart rate.

[0147] Minimum respiratory rate per minute: b_min = np.min(b_rate) = MIN(b_rate); where b_min represents the minimum respiratory rate;

[0148] Maximum respiratory rate per minute: b_max = np.max(b_rate) = MAX(b_rate); where b_max represents the maximum respiratory rate.

[0149] Average heart rate per minute: h_mean = np.mean(h_rate) = AVERAGE(h_rate); where np.mean() is a function in the Python programming language to calculate the average value; h_mean represents the average heart rate;

[0150] Average respiratory rate per minute: b_mean = np.mean(b_rate) = AVERAGE(b_rate); where b_mean represents the average heart rate;

[0151] Heart rate variability characteristics:

[0152] h_diff=np.mean(np.abs(h_rate[0:-1]-h_rate[1:]))=AVERAGE{ABS(h_rate_2-h_rate_1)+ABS(h_rate_3-h_rate_2)+…+ABA[h_rate_n-h_rate_(n-1)]}; where np.abs() is the function for calculating absolute value in the Python programming language; [0:-1] and [1:] represent the slicing operation unique to Python programming, also called the cutting operation, where 0 represents the first element from the left and -1 represents the last element; n ..._(n-1)))=AVERAGE{ABS(h_rate_2-h_rate_1)+ABS(h_rate_3-h_rate_2)+…+ABA[h_rate_n-h_rate_(n-1)]}; _rate[1:])) means that starting from the first element in the heart rate sequence h_rate, select one element and the next adjacent element in sequence, take the difference between the two and then take the absolute value, so as to obtain the absolute value of multiple differences. Then, use the average function to calculate the average of multiple absolute values ​​to obtain the required first derivative h_diff; ABS() is a function for calculating absolute values ​​in some other programming languages. Correspondingly, AVERAGE{ABS(h_rate_2-h_rate_1)+ABS(h_rate_3-h_rate_2)+…+ABA[h_rate_n-h_rate_(n-1)] is another function representation for calculating the first derivative h_diff;

[0153] h_std = np.std(h_rate) = sqrt{[(h_rate_1-b_mean)^2+(h_rate_2-h_mean)^2...+(h_rate_n-b_mean)^2] / n}; where np.std() is the function for calculating the square root in the Python programming language, and sqrt() is the function for calculating the square root in some other programming languages; h_std represents the standard deviation of heart rate;

[0154] h_sub = h_max - h_min = MAX(h_rate) - MIN(h_rate); where h_sub represents the range of heart rate.

[0155] Characteristics of respiratory rate changes:

[0156] b_diff=np.mean(np.abs(b_rate[0:-1]-b_rate[1:]))=AVERAGE[ABS(b_rate_2-b_rate_1)+ABS(b_rate_3-b_rate_2)+...+ABS(b_rate_n-b_rate_(n-1))]; where b_diff represents the first derivative of the respiratory rate;

[0157] b_std=np.std(b_rate)=sqrt{[(b_rate_1-b_mean)^2+(b_rate_2-h_mean)^2+...+(b_rate_n-b_mean)^2] / n}; where b_diff represents the standard deviation of the respiratory rate;

[0158] b_sub = b_max – b_min = MAX(b_rate) - MIN(b-rate); where h_sub represents the range of respiratory rates;

[0159] The total number of physical movements per 10 minutes: m = np.sum(move_nums) = SUM(move_nums); where np.sum() is the summation function in the Python programming language, and SUM() is the summation function in some other programming languages.

[0160] Step 3: Based on the feature data calculated in Step 2, score sleep onset and wakefulness in three dimensions: heart rate, respiratory rate, and body movement. The sleep onset score can include assessment scores for heart rate, respiratory rate, and body movement. Similarly, the wakefulness score can also include assessment scores for heart rate, respiratory rate, and body movement.

[0161] Specifically, the sleep onset score can be calculated as follows:

[0162] Assessment of the centrality dimension of the sleep onset score:

[0163] sleep_h_score = [max(100-10×h_std,0)+max(100-5×h_sub,0)+max(100-10×h_diff,0)] / 3; where sleep_h_score represents the heart rate score for falling asleep; the value 100 represents the total score, which can be set manually; max(100-10×h_std,0) determines the maximum value between 100-10×h_std and 0 (the same applies to other terms). This means that, considering the actual situation, for a normal human body, 100-10×h_std cannot be less than 0, that is, h_std cannot be greater than or equal to 10 (the same applies to other terms), otherwise it indicates an abnormality in the human body (or, under normal circumstances, it is impossible for h_std≥10). For the human body, the standard deviation, range, and first derivative of heart rate are relatively stable values, and they are all slightly greater in the awake state than in the sleeping state. Therefore, this can be used to some extent to determine whether a person has fallen asleep.

[0164] Assessment of the respiratory rate dimension in the sleep onset score:

[0165] sleep_b_score = [max(100-10×b_std,0)+max(100-5×b_sub,0)+max(100-10×b_diff,0)] / 3; where sleep_b_score represents the respiratory rate score at which sleep begins; it can be understood that this formula is based on the same principle as the calculation formula for sleep_h_score, so it will not be elaborated further.

[0166] Assessment of the physical movement dimension in the sleep onset score:

[0167] sleep_m_score=max(100-10×m,0); where sleep_m_score represents the body movement score for falling asleep; the value 100 also represents the total score, which can be set manually; max(100-10×m,0) determines the maximum value between 100-10×m and 0. Its meaning is that, in reality, the number of body movements in the human body is less when falling asleep than when awake. Therefore, the smaller the sleep_m_score, the greater the possibility that the monitored subject is in a state of falling asleep, which is reflected in a higher score.

[0168] Specifically, the method for calculating the alertness score can be as follows:

[0169] Assessment of the centrality dimension of the alertness score:

[0170] wake_h_score = min(20×h_std,100)×0.4 + min(10×h_sub,100)×0.3 + min(20×h_diff,100)×0.3; where wake_h_score represents the heart rate score indicating wakefulness. This formula is similar in principle to the calculation of wake_h_score. For the human body, since the standard deviation, range, and first derivative of heart rate are relatively stable values, and they are all slightly greater when the body is awake than when it is asleep, this can be used to determine whether the body is awake to a certain extent.

[0171] Assessment of the respiratory rate dimension in the alertness score:

[0172] wake_b_score = min(20×b_std,100)×0.4 + min(10×b_sub,100)×0.3 + min(20×b_diff,100)×0.3; where wake_b_score represents the score of the respiratory rate dimension indicating wakefulness; this formula is based on the same principle as the calculation formula of wake_h_score, so it will not be repeated here.

[0173] Assessment of the physical movement dimension in the alertness score:

[0174] wake_m_score=min(m×2,100); where wake_m_score represents the score of the body movement dimension indicating wakefulness; in reality, the number of body movements in the human body is relatively less when the human body is asleep than when the human body is awake. Therefore, the larger the wake_m_score is, the greater the possibility that the monitored object is awake, which is reflected in the higher score.

[0175] Step 4: Based on historical records, determine the time periods with the highest frequency of sleep onset and wakefulness detected by the device, along with their corresponding probabilities. The time interval with the highest frequency of sleep onset is denoted as [sleep_time1, sleep_time2] (equivalent to the first time interval in the exemplary method section), and the probability is sleep_rate.

[0176] The time interval with the highest frequency of wakefulness is denoted as [wake_time1, wake_time2] (which is equivalent to the second time interval in the exemplary method section), and the probability is wake_rate.

[0177] Set the self-learning base score as base_score (the specific score can be obtained through experimentation), and denot the minimum probability that sleep_rate and wake_rate must satisfy as rate (rate is the threshold calculated from the user's historical sleep data). Select a time point from the n minutes selected in step one (it can be arbitrarily selected or an intermediate value can be selected), and record this time point as the current time point time.

[0178] When sleep_time1≤time≤sleep_time2 and sleep_rate≥rate, it means that the current time point falls within the time period with the highest frequency of falling asleep. In this case, the calculation expression of the self-learning sleep score (self_learning_sleep_score, which is equivalent to the first self-learning parameter in the exemplary method part) is: self_learning_sleep_score=sleep_rate×base_score; that is, if the current time point falls within the time period with the highest frequency of falling asleep, the self-learning sleep score is obtained through calculation.

[0179] Otherwise, the self-learning sleep score calculation expression is self_learning_sleep_score=0; that is, if the current time does not fall within the time period with the highest frequency of falling asleep, the self-learning sleep score is 0.

[0180] If wake_time1 <= time <= wake_time2 and wake_rate >= rate, it means that the current time point falls within the time period with the highest frequency of wakefulness. In this case, the self-learning wakefulness score is calculated as: self_learning_wake_score = wake_rate × base_score. That is, if the current time point falls within the time period with the highest frequency of wakefulness, the self-learning wakefulness score is obtained through calculation.

[0181] Otherwise, the self-learning wakefulness score calculation expression is: self_learning_wake_score=0; that is, if the current time does not fall within the time period with the highest frequency of wakefulness points, the self-learning wakefulness score is 0.

[0182] Step 5: Combine all scores from Steps 3 and 4 according to the following formula, where a1, b1, and c1 are the weights of the sleep onset score (i.e., the weights of the first row in the exemplary method section), and a1 + b1 + c1 = 1; and a2, b2, and c2 are the weights of the wakefulness score (i.e., the weights of the second row in the exemplary method section), and a2 + b2 + c2 = 1; (the specific weights can be obtained through multiple experiments).

[0183] sleep_score = a1 × sleep_m_score + b1 × sleep_b_score + c1 × sleep_h_score + self_learning_sleep_score; where sleep_score is the sleep assessment score, which is equivalent to the first assessment result in the exemplary method section.

[0184] wake_score = a2 × wake_m_score + b2 × wake_b_score + c2 × wake_h_score + self_learning_wake_score; where wake_score is the wakefulness assessment score, which is equivalent to the second assessment result in the exemplary method section;

[0185] Step Six: Set the sleep threshold (sleep_thresh) and wake-up threshold (wake_thresh). When sleep_score > sleep_thresh, the system is determined to be asleep, and the sleep state is set to the sleep state (equivalent to the target behavior state in the exemplary method section). When wake_score > wake_thresh, the system is determined to be awake, and the sleep state is set to the awake state (equivalent to the target behavior state in the exemplary method section).

[0186] Furthermore, after obtaining the evaluation results, if the user's sleep state is determined to be asleep, then the smart home devices in the current space can be controlled to perform actions corresponding to the sleep mode. For example, if the user is asleep but the room lights or TV lights are not turned off, the lights or TV lights in the room will be automatically turned off.

[0187] If, based on the evaluation results, it is determined that within a specified time period, such as the designated wake-up time in the morning, the user's sleep state changes from falling asleep to waking up, then the smart home devices in the current space can be controlled to perform actions corresponding to the wake-up mode. For example, if the user is awake during a preset time period in the morning, the curtains in the room can be automatically opened, and so on.

[0188] In this embodiment, by combining respiratory rate, heart rate, and body movement characteristics, separate assessment methods are set up for the sleep state and the awake state, which can more accurately determine the user's sleep state. Therefore, by controlling the smart devices in the user's space according to the sleep and awake states, more precise automatic control of the devices can be achieved based on the user's behavior, greatly improving the intelligence and accuracy of device control, and also significantly enhancing the user experience.

[0189] Exemplary device

[0190] In one embodiment, such as Figure 6 As shown, a device control device 600 based on behavior monitoring is provided. This device control device 600 is applied to electronic devices such as gateways or cloud servers, and includes: a feature extraction module 601, a first evaluation module 602, a second evaluation module 603, and a state recognition module 604; wherein:

[0191] Feature extraction module 01 is used to: extract features from the behavior monitoring data of the target object to obtain behavioral features;

[0192] The first evaluation module 602 is used to: perform a first behavior evaluation process on the behavioral characteristics to obtain a first evaluation result;

[0193] The second evaluation module 603 is used to: perform a second behavioral evaluation process on the behavioral characteristics to obtain a second evaluation result;

[0194] The state recognition module 604 is used to: identify the target behavior state of the target object based on the first evaluation result and the second evaluation result, and adjust the device in the space where the target object is located based on the target behavior state.

[0195] In one embodiment, when the first evaluation module 602 performs a first behavior evaluation process on the behavioral features to obtain a first evaluation result, it is specifically used to: perform a first behavior evaluation process on the behavioral features of multiple dimensions through a first behavior evaluation strategy to obtain a first processing result corresponding to each dimension; determine the first evaluation result corresponding to the first behavior state based on the first processing result and the first self-learning parameter; the first self-learning parameter is determined based on the historical data corresponding to the first behavior state.

[0196] In one embodiment, when the first evaluation module 602 determines the first evaluation result corresponding to the first behavior state based on the first processing result and the first self-learning parameter, it is specifically used to: obtain the first self-learning parameter; the first self-learning parameter is determined based on the first time interval corresponding to the first behavior state in historical data; and perform fusion processing on the first processing result, the first behavior weight, and the first self-learning parameter to obtain the first evaluation result corresponding to the first behavior state.

[0197] In one embodiment, when the second evaluation module 603 performs second behavior evaluation processing on the behavioral features to obtain a second evaluation result, it is specifically used to: perform second behavior evaluation processing on behavioral features of multiple dimensions through a second behavior evaluation strategy to obtain a second processing result corresponding to each dimension; determine the second evaluation result corresponding to the second behavior state based on the second processing result and the second self-learning parameter; the second self-learning parameter is determined based on the historical data corresponding to the second behavior state.

[0198] In one embodiment, when the second evaluation module 603 determines the second evaluation result corresponding to the second behavior state based on the second processing result and the second self-learning parameter, it is specifically used to: obtain the second self-learning parameter; the second self-learning parameter is determined based on the first time interval corresponding to the first behavior state in historical data; and perform fusion processing on the second processing result, the second behavior weight, and the second self-learning parameter to obtain the second evaluation result corresponding to the second behavior state.

[0199] In one embodiment, the behavior monitoring data includes heart rate data, respiratory rate data, and body movement data; when the feature extraction module 601 extracts features from the behavior monitoring data of the target object to obtain behavior features, it is specifically used to: acquire heart rate data, respiratory rate data, and body movement data in the same time interval; and extract features from the heart rate data, respiratory rate data, and body movement data according to the time series to obtain the behavior features corresponding to each dimension of heart rate, respiratory rate, and body movement.

[0200] In one embodiment, the target behavior state includes a first behavior state and a second behavior state. When the state recognition module 604 adjusts the device in the space where the target object is located based on the target behavior state, it is specifically used to: if the behavior state changes from the first behavior state to the second behavior state, control the device in the space where the target object is located to execute the control command corresponding to the scene mode of the second behavior state; if the behavior state changes from the second behavior state to the first behavior state, control the device in the space where the target object is located to execute the control command corresponding to the scene mode of the first behavior state.

[0201] In one embodiment, when the state recognition module 604 executes a control command corresponding to the scene mode of the second behavior state on the device in the space where the target object is located, if the state changes from the first behavior state to the second behavior state and the current time is within a first preset time period, the state recognition module 604 executes a control command corresponding to the scene mode of the second behavior state on the device in the space where the target object is located; when the state recognition module 604 executes a control command corresponding to the scene mode of the first behavior state on the device in the space where the target object is located, if the state changes from the second behavior state to the first behavior state and the current time is within a second preset time period, the state recognition module 604 executes a control command corresponding to the scene mode of the first behavior state on the device in the space where the target object is located.

[0202] In one embodiment, the first behavioral state includes an awake state, and the second behavioral state includes a sleep state. The state recognition module 604 is further configured to: determine a sleep point and an awake point based on the awake state and the sleep state; the sleep point is the time point at which the awake state changes to the sleep state, and the awake point is the time point at which the sleep state changes to the awake state; correspondingly, if the behavior changes from the first behavioral state to the second behavioral state, when the state recognition module 604 executes the control command corresponding to the scene mode of the second behavioral state on the device in the space where the target object is located, it is specifically configured to: if the current behavioral state is at the sleep point, then execute the control command of the sleep mode on the device in the space where the target object is located; if the behavior changes from the second behavioral state to the first behavioral state, when the state recognition module 604 executes the control command corresponding to the scene mode of the first behavioral state on the device in the space where the target object is located, it is specifically configured to: if the current behavioral state is at the awake point, then execute the control command of the wake-up mode on the device in the space where the target object is located.

[0203] For specific limitations regarding the device control mechanism based on behavior monitoring, please refer to the limitations of the device control method based on behavior monitoring mentioned above, which will not be repeated here. Each module in the aforementioned sleep monitoring device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0204] Exemplary computer device

[0205] In one embodiment, a computer device is provided, which may be a gateway or a cloud server, and its internal structure diagram may be as follows: Figure 7 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and the database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data such as a first behavior evaluation strategy and a second behavior evaluation strategy. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a device control method based on behavior monitoring.

[0206] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0207] In one embodiment, a computer device is provided, which may be a gateway or a cloud server, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the various steps of the behavior monitoring-based device control method shown in the exemplary method section of the above embodiments.

[0208] Example control system

[0209] In one embodiment, a device control system based on behavior monitoring is provided, comprising: a data acquisition device and a computer device; the data acquisition device and the computer device are communicatively connected.

[0210] The data acquisition device is installed in the monitoring area or worn by the target object located in the monitoring area, and is used to collect behavioral monitoring data of the target object;

[0211] The computer device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the behavior monitoring-based device control method shown in the various embodiments of the exemplary method section above.

[0212] Exemplary computer program products and computer-readable storage media

[0213] In addition to the above-described device control methods, apparatuses, and devices based on behavior monitoring, embodiments of this application also provide a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the device control methods based on behavior monitoring according to various embodiments of this application as described in the exemplary method section of this specification.

[0214] Computer program products can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the device or as a standalone software package.

[0215] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the various steps of the behavior monitoring-based device control method shown in the embodiments of the exemplary method section above.

[0216] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0217] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0218] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method of controlling a device based on behavior monitoring, the method comprising: The method comprises: feature extraction is performed on the behavior monitoring data of the target object to obtain the first derivative, standard deviation and range of the heart rate dimension, the first derivative, standard deviation and range of the respiratory rate dimension, and the number of body movements; The first derivative, standard deviation and range of the heart rate dimension are scored by a first behavior evaluation strategy to obtain a first processing result corresponding to the heart rate dimension; the first derivative, standard deviation and range of the respiratory rate dimension are scored by the first behavior evaluation strategy to obtain a first processing result corresponding to the respiratory rate dimension; and the number of body movements is scored by the first behavior evaluation strategy to obtain a first processing result corresponding to the body movement dimension; A first self-learning parameter is obtained; the first self-learning parameter is determined based on a first time interval corresponding to a first behavior state in historical data; if the current time is located within the first time interval, the first self-learning parameter is increased relative to the case where the current time is not located within the first time interval; and the first time interval is a time interval in which the first behavior state occurs most frequently, which is obtained by statistical analysis of historical data; The first processing results corresponding to the heart rate dimension, respiratory rate dimension and body movement dimension, and the first behavior weights corresponding to each of the first processing results, are weighted and summed, and then added to the first self-learning parameter to obtain a first evaluation result corresponding to the first behavior state; the first evaluation result represents whether the target object is in the first behavior state; The first derivative, standard deviation and range of the heart rate dimension are scored by a second behavior evaluation strategy to obtain a second processing result corresponding to the heart rate dimension; the first derivative, standard deviation and range of the respiratory rate dimension are scored by the second behavior evaluation strategy to obtain a second processing result corresponding to the respiratory rate dimension; and the number of body movements is scored by the second behavior evaluation strategy to obtain a second processing result corresponding to the body movement dimension; A second self-learning parameter is obtained; the second self-learning parameter is determined based on a second time interval corresponding to a second behavior state in historical data; if the current time is located within the second time interval, the second self-learning parameter is increased relative to the case where the current time is not located within the second time interval; and the second time interval is a time interval in which the second behavior state occurs most frequently, which is obtained by statistical analysis of historical data; The second processing results corresponding to the heart rate dimension, respiratory rate dimension and body movement dimension, and the second behavior weights corresponding to each of the second processing results, are weighted and summed, and then added to the second self-learning parameter to obtain a second evaluation result corresponding to the second behavior state; the second evaluation result represents whether the target object is in the second behavior state; wherein the first behavior evaluation strategy and the second behavior evaluation strategy are different; According to the first evaluation result and the second evaluation result, the target behavior state of the target object is identified, and a device in a space where the target object is located is adjusted based on the target behavior state.

2. The method of claim 1, wherein, The behavior monitoring data comprises heart rate data, respiratory rate data and body movement data; The behavior monitoring data of the target object is feature extracted to obtain a first derivative, a standard deviation and a range of a heart rate dimension, a first derivative, a standard deviation and a range of a respiratory rate dimension, and a body movement frequency, including: obtaining heart rate data, respiratory rate data and body movement data in the same time interval; respectively extracting features from the heart rate data, the respiratory rate data and the body movement data according to a time sequence to obtain a first derivative, a standard deviation and a range of a heart rate dimension, a first derivative, a standard deviation and a range of a respiratory rate dimension, and a body movement frequency; the feature extraction according to the time sequence includes feature extraction according to the chronological order of the occurrence time of the behavior monitoring data.

3. The method of claim 1, wherein, The target behavior state includes a first behavior state and a second behavior state, and the adjustment of the device in the space where the target object is located based on the target behavior state includes: if the first behavior state changes to the second behavior state, controlling the device in the space where the target object is located to execute the control instruction of the scene mode corresponding to the second behavior state; if the second behavior state changes to the first behavior state, controlling the device in the space where the target object is located to execute the control instruction of the scene mode corresponding to the first behavior state.

4. The method of claim 3, wherein, If the first behavior state changes to the second behavior state, and the current time is within the first preset time period, the device in the space where the target object is located is controlled to execute the control instruction of the scene mode corresponding to the second behavior state. If the second behavior state changes to the first behavior state, and the current time is within the second preset time period, the device in the space where the target object is located is controlled to execute the control instruction of the scene mode corresponding to the first behavior state. The first behavior state includes a wake state, and the second behavior state includes a sleep state, and the method further includes: determining a sleep point and a wake point according to the wake state and the sleep state; the sleep point is the time point when the wake state changes to the sleep state, and the wake point is the time point when the sleep state changes to the wake state; 5. The method of claim 3, wherein, If the current behavior state is at the sleep point, the device in the space where the target object is located is controlled to execute the control instruction of the sleep mode. If the current behavior state is at the wake point, the device in the space where the target object is located is controlled to execute the control instruction of the wake-up mode. The device includes: ​ ​ ​ 6. An apparatus control device based on behavior monitoring, characterized by, ​ The feature extraction module is configured to perform feature extraction on the behavior monitoring data of the target object to obtain a first derivative, a standard deviation and a range of a heart rate dimension, a first derivative, a standard deviation and a range of a respiratory rate dimension, and a number of body movements; The first evaluation module is configured to score the first derivative, the standard deviation and the range of the heart rate dimension by using a first behavior evaluation strategy to obtain a first processing result corresponding to the heart rate dimension, score the first derivative, the standard deviation and the range of the respiratory rate dimension by using the first behavior evaluation strategy to obtain a first processing result corresponding to the respiratory rate dimension, score the number of body movements by using the first behavior evaluation strategy to obtain a first processing result corresponding to the body movement dimension, and obtain a first self-learning parameter. The first self-learning parameter is determined based on a first time interval corresponding to a first behavior state in historical data. If the current time is located within the first time interval, the first self-learning parameter is increased relative to a case where the current time is not located within the first time interval. The first time interval is a time interval in which the first behavior state occurs most frequently, which is obtained by statistical analysis of the historical data. The first evaluation module is further configured to add the first processing results corresponding to the heart rate dimension, the respiratory rate dimension and the body movement dimension respectively and the first behavior weights corresponding to the respective first processing results, perform weighted summation, and then add the first self-learning parameter to obtain a first evaluation result corresponding to the first behavior state. The first evaluation result represents whether the target object is in the first behavior state. The second evaluation module is configured to score the first derivative, the standard deviation and the range of the heart rate dimension by using a second behavior evaluation strategy to obtain a second processing result corresponding to the heart rate dimension, score the first derivative, the standard deviation and the range of the respiratory rate dimension by using the second behavior evaluation strategy to obtain a second processing result corresponding to the respiratory rate dimension, score the number of body movements by using the second behavior evaluation strategy to obtain a second processing result corresponding to the body movement dimension, and obtain a second self-learning parameter. The second self-learning parameter is determined based on a second time interval corresponding to a second behavior state in historical data. If the current time is located within the second time interval, the second self-learning parameter is increased relative to a case where the current time is not located within the second time interval. The second time interval is a time interval in which the second behavior state occurs most frequently, which is obtained by statistical analysis of the historical data. The second evaluation module is further configured to add the second processing results corresponding to the heart rate dimension, the respiratory rate dimension and the body movement dimension respectively and the second behavior weights corresponding to the respective second processing results, perform weighted summation, and then add the second self-learning parameter to obtain a second evaluation result corresponding to the second behavior state. The second evaluation result represents whether the target object is in the second behavior state. The first behavior evaluation strategy and the second behavior evaluation strategy are different. The state recognition module is configured to recognize a target behavior state of the target object based on the first evaluation result and the second evaluation result, and adjust a device in a space in which the target object is located based on the target behavior state.

7. A computer device, comprising: The system comprises a collection device and a computer device; the collection device is in communication connection with the computer device; 8. A behavior-monitoring-based device control system, characterized by, The collection device is arranged in a monitoring area or is worn by a target object located in the monitoring area, and is used for collecting behavior monitoring data of the target object; The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the behavior monitoring-based device control method according to any one of claims 1 to 5 when executing the computer program. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the behavior monitoring-based device control method according to any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the behavior monitoring-based device control method according to any one of claims 1 to 5.

10. A computer program product, characterised in that, The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the behavior monitoring-based device control method according to any one of claims 1 to 5.

Images