A method for detecting leaving bed

Abstract

The present invention provides a method for detecting when leaving the bed, which relates to the field of smart home technology, including: obtaining a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extracting each parameter feature based on the cardiac impact signal and the pressure signal; determining the first environmental state of the smart mattress at the target time, and performing scene classification on the first environmental state based on each parameter feature to obtain a target scene state; using each parameter threshold, performing bed leaving detection on each parameter feature respectively to obtain multiple detection results; based on the weight associated with the target scene state, performing weighted processing on each detection result to obtain a target bed leaving detection result. The present invention performs scene classification on the environment where the smart mattress is located based on each parameter feature, obtains the target scene state of the smart mattress at the target time, and performs bed leaving judgment in combination with the scene state and the parameter threshold updated by self-learning in different scenes, effectively improving the accuracy of bed leaving detection.

Description

Technical Field

[0001] The present invention relates to the field of smart home technology, and in particular to a bed leaving detection method. Background Art

[0002] At present, the common method of judging whether the user is in bed or out of bed is to collect signals from sensors and then perform bed-in-bed calculations. The specific process is as follows: 1: The sensor collects signals; 2: Signal processing; 3: Feature extraction of the signal to obtain relevant features such as pressure, heart rate and breathing; 4: The extracted features are used for corresponding algorithm calculations to determine whether the user is in bed or out of bed. However, in actual use, the accuracy of the judgment of smart mattresses in and out of bed is often affected by environmental interference. For example, when the user lifts the head of the bed or turns on functions such as vibration massage, the signal collected by the sensor may be erroneous, thereby affecting the accuracy of the judgment result. Summary of the invention

[0003] The present invention provides a method for detecting when a user leaves a bed, aiming to solve at least one of the above-mentioned technical problems.

[0004] The present invention provides a method for detecting when leaving bed, comprising:

[0005] Acquire a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extract various parameter features based on the cardiac impact signal and the pressure signal;

[0006] Determine a first environmental state corresponding to the smart mattress at the target time, wherein the first environmental state refers to the working mode of the smart mattress at the target time and / or the working mode state of the smart device associated with the smart mattress;

[0007] Based on the parameter characteristics, the first environmental state is subjected to scene classification to obtain a target scene state of the smart mattress at a target time, wherein the target scene state is used to characterize scene states of different interference intensities;

[0008] Using the parameter thresholds updated in real time, each parameter feature is detected while leaving the bed to obtain multiple detection results, wherein the parameter threshold is obtained by self-learning and updating based on the power in bed, pressure information and scene status;

[0009] Based on the weight associated with the target scene state, each of the detection results is weighted to obtain the target bed leaving detection result.

[0010] According to a bed leaving detection method provided by the present invention, based on each of the parameter features, the first environmental state is scene classified to obtain a target scene state of the smart mattress at a target time, including:

[0011] Determine whether each environmental variable in the first environmental state is the same as each environmental variable in the second environmental state at a previous moment, wherein the previous moment refers to a moment before the target moment;

[0012] If there are different environmental variables in the first environmental state, determining each first parameter feature in a first preset time window before the target moment and each second parameter feature in a second preset time window after the target moment based on each parameter feature in a preset time period;

[0013] Based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window, updating a power impact factor and a pressure impact factor of any target environmental variable at the target time, wherein the target environmental variable refers to an environmental variable in the first environmental state that is different from the second environmental state;

[0014] Determining a target power influence coefficient and a target pressure influence coefficient at the target time based on the power influence factors and pressure influence factors of all target environmental variables;

[0015] The target scene state is determined based on the target power influence coefficient and the target pressure influence coefficient.

[0016] According to a method for detecting leaving bed provided by the present invention, the first parameter feature includes respiratory energy, heart rate energy, high frequency energy, in-bed power and pressure information within a first preset time window;

[0017] The second parameter characteristics include respiratory energy, heart rate energy, high frequency energy, in-bed power and pressure information within a second preset time window;

[0018] Accordingly, updating the power impact factor and the pressure impact factor of any target environmental variable at the target time based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window includes:

[0019] Determining influencing factors of the respiratory energy, the heart rate energy, the high-frequency energy, the in-bed power, and the pressure information based on each first parameter feature within the first preset time window and each second parameter feature within the second preset time window;

[0020] Adding the influencing factors corresponding to the breathing energy, the heart rate energy, the high-frequency energy, and the in-bed power, respectively, to obtain the parameter influencing factor at the target moment;

[0021] Calculate the power impact factor of any target environmental variable according to the power impact factor at the previous moment and the parameter impact factor at the target moment;

[0022] The pressure influence factor of any target environmental variable is calculated based on the pressure influence factor of the pressure information at the previous moment and the influence factor of the pressure information at the target moment.

[0023] According to a bed leaving detection method provided by the present invention, the determining of the influencing factors of the respiratory energy, the heart rate energy, the high frequency energy, the bed power and the pressure information based on the first parameter characteristics within the first preset time window and the second parameter characteristics within the second preset time window includes:

[0024] Based on the respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the first preset time window, first means corresponding to the respiratory energy, the heart rate energy, the high frequency energy, the bed power and the pressure, respectively, and a sum of first first-order differences corresponding to the pressure information are calculated;

[0025] Based on the respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the second preset time window, second means corresponding to the respiratory energy, the heart rate energy, the high frequency energy, the bed power and the pressure, respectively, and a sum of second first-order differences corresponding to the pressure information are calculated;

[0026] Based on the first mean and the second mean respectively corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power, the mean ratios respectively corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power are calculated;

[0027] Based on the mean ratios of the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power, respectively, the influence factors corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power are calculated;

[0028] An impact factor corresponding to the pressure information is determined based on a first difference between a first mean and a second mean corresponding to the pressure information and a second difference between the sum of the first first-order differences and the sum of the second first-order differences.

[0029] According to a bed leaving detection method provided by the present invention, the power impact factor of any target environmental variable is calculated based on the power impact factor at the previous moment and the parameter impact factor at the target moment, including:

[0030] Determine the number of variables that have different environmental variables;

[0031] Determine the time interval between the target moment and the last update moment of the coefficient, wherein the update coefficient is used to characterize the weight coefficient of the power impact factor at the previous moment and the parameter impact factor at the target moment in the power impact factor update process;

[0032] Determine the power impact factor corresponding to each environmental variable at the previous moment;

[0033] Determine the update coefficient at the target moment based on the time interval, the number of variables, and the power impact factor corresponding to each environmental variable at the previous moment;

[0034] The power impact factor of any of the target environmental variables is calculated based on the update coefficient at the target moment, the power impact factor at the previous moment, and the parameter impact factor at the target moment.

[0035] According to a bed leaving detection method provided by the present invention, the power influence factor and the pressure influence factor of all target environmental variables are used to determine the target power influence coefficient and the target pressure influence coefficient at the target moment, including:

[0036] Determine the power impact factor and the pressure impact factor of the remaining environmental variables in the activated state at the target time;

[0037] Adding the power impact factors of all target environmental variables and the power impact factors of the remaining environmental variables to obtain the target power impact coefficient;

[0038] The pressure influence factors of all target environmental variables and the pressure influence factors of the remaining environmental variables are added together to obtain the target pressure influence coefficient.

[0039] According to a bed leaving detection method provided by the present invention, the target scene state is determined based on the target power influence coefficient and the target pressure influence coefficient, including:

[0040] If the target power influence coefficient is greater than a first power coefficient threshold, or the target pressure influence coefficient is greater than a preset pressure coefficient threshold, determining that the target scene state is in a scene state of a first interference intensity;

[0041] If the target power influence coefficient is greater than the second power coefficient threshold and less than or equal to the first power coefficient threshold, and the target pressure influence coefficient is less than or equal to the pressure coefficient threshold, determining that the target scene state is in a scene state of a second interference intensity;

[0042] If the target power influence coefficient is less than or equal to the second power coefficient threshold, and the target pressure influence coefficient is less than or equal to the pressure coefficient threshold, determining that the target scene state is in a scene state of a third interference intensity;

[0043] The first power coefficient threshold is greater than the second power coefficient threshold, the first interference intensity is greater than the second interference intensity, and the second interference intensity is greater than the third interference intensity.

[0044] According to a bed leaving detection method provided by the present invention, the self-learning update process of the parameter threshold is as follows:

[0045] Acquiring in-bed power and pressure information within a third preset time window;

[0046] Based on the in-bed power and pressure information within a third preset time window, determining whether the in-bed power and the pressure information are in a stable state;

[0047] If it is in a stable state, and the power in bed at the target time is less than the first preset power threshold, then the parameter thresholds corresponding to the various parameter characteristics of the smart mattress when no one is in the bed are updated;

[0048] If it is in a stable state, and the power in bed at the target moment is greater than the second preset power threshold, then the parameter thresholds corresponding to the various parameter characteristics of the smart mattress when it is in a occupied state are updated;

[0049] The first preset power threshold and the second preset power threshold are determined in combination with the scene state at the target moment.

[0050] According to a bed leaving detection method provided by the present invention, judging whether the bed leaving power and the pressure information are in a stable state based on the bed leaving power and the pressure information within a third preset time window includes:

[0051] Determine a power standard deviation and a power mean corresponding to the in-bed power within the third preset time window, and a pressure standard deviation between the pressure information within the third preset time window;

[0052] determining a ratio between the power standard deviation and the power mean;

[0053] If the ratio is greater than a preset ratio threshold, or the pressure standard deviation is greater than a preset standard deviation threshold, it is determined to be in an unstable state.

[0054] According to a bed leaving detection method provided by the present invention, the weighted processing is performed on each of the detection results based on the weight associated with the target scene state to obtain the target bed leaving detection result, including:

[0055] Performing weighted sum processing on the weight associated with the target scene state and each of the detection results to obtain a weighted calculation result;

[0056] If the weighted calculation result is greater than a preset threshold, it is determined that the target is in bed when leaving the bed.

[0057] If the weighted calculation result is not greater than a preset threshold, it is determined that the target is in the out-of-bed state in the out-of-bed detection result.

[0058] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the bed leaving detection method as described in any one of the above is implemented.

[0059] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the bed exit detection method as described in any one of the above is implemented.

[0060] The bed leaving detection method provided by the present invention includes: obtaining a cardiac impact signal and a pressure signal within a preset time period collected by a sensor, and extracting each parameter feature based on the cardiac impact signal and the pressure signal; determining a first environmental state corresponding to a smart mattress at a target time, wherein the first environmental state refers to a working mode of the smart mattress at the target time and / or a working mode state of a smart device associated with the smart mattress; based on each of the parameter features, performing scene classification on the first environmental state to obtain a target scene state of the smart mattress at the target time, wherein the target scene state is used to characterize scene states with different interference intensities; using each parameter threshold value updated in real time, performing bed leaving detection on each of the parameter features respectively to obtain multiple detection results, wherein the parameter threshold value is obtained by self-learning and updating based on the bed power, pressure information and scene state; and based on the weight associated with the target scene state, performing weighted processing on each of the detection results to obtain a target bed leaving detection result. The present invention classifies the environment in which the smart mattress is located based on various parameter characteristics, obtains the target scene state of the smart mattress at the target time, and combines the scene state and the parameter thresholds updated by self-learning in different scenes to make bed leaving judgment on various parameter characteristics at the target time, obtains the final target bed leaving detection result, and effectively improves the accuracy of bed leaving detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced one by one below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 This is one of the flow charts of the bed leaving detection method provided by the present invention;

[0063] Figure 2 is a structural diagram of a device provided by an embodiment of the present invention;

[0064] Figure 3 It is a schematic diagram of extracting a respiratory signal, a heart rate signal and a high-frequency signal provided by an embodiment of the present invention;

[0065] Figure 4 This is the second flow chart of the bed leaving detection method provided by the present invention;

[0066] Figure 5 This is the third flow chart of the bed leaving detection method provided by the present invention;

[0067] Figure 6 This is the fourth flow chart of the bed leaving detection method provided by the present invention;

[0068] Figure 7 It is a structural schematic diagram of the bed leaving detection device provided by the present invention;

[0069] Figure 8 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] It should be understood that, although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present invention, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of the present invention, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "when..." or "when...".

[0072] The embodiment of the present invention can be applied to scenarios such as smart sofas and smart mattresses. The embodiment of the present invention is specifically described using a smart mattress as an example.

[0073] Specifically, Figure 1 As shown, Figure 1 This is one of the flow charts of the bed leaving detection method provided by the present invention. The bed leaving detection method comprises:

[0074] Step S11, acquiring a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extracting each parameter feature based on the cardiac impact signal and the pressure signal;

[0075] In one embodiment, referring to Figure 2 , Figure 2 : is a device structure diagram provided by an embodiment of the present invention, wherein the sensor includes a piezoelectric signal sensor and a pressure signal sensor, wherein the piezoelectric signal sensor is used to collect a cardiac impact signal, and the pressure sensor is used to collect a pressure signal. In a specific example, the sensor is set at the position where the user's shoulder is located, and it should be noted that the sensor can be deployed inside the smart mattress or on the electric bed frame.

[0076] Specifically, the cardiac impact signal and pressure signal collected by the sensor within a preset time period are obtained, and then the cardiac impact signal is band-pass filtered to obtain the physiological signal. For example, the cardiac impact signal is filtered using an infinite impulse response (IIR) type Butterworth band-pass filter, and the passband range is set to [0.1, 20] Hz, so that low-frequency noise and high-frequency noise in the signal can be filtered out; further, the physiological signal is extracted from different frequency bands to obtain a respiratory signal, a heart rate signal and a high-frequency signal, referring to Figure 3 , Figure 3 It is a schematic diagram of extracting a respiratory signal, a heart rate signal and a high-frequency signal provided by an embodiment of the present invention. In a specific example, the passband of the respiratory signal is set to [0.1, 0.6] Hz, and the respiratory signal can be extracted using an IIR filter. The signal envelope of the physiological signal can also be extracted, and then the signal envelope is band-pass filtered. In addition, the passband of the heart rate signal is set to [0.6, 8] Hz, and the passband of the high-frequency signal is set to [20, 30] Hz. Based on the respiratory signal, the heart rate signal and the high-frequency signal, multiple parameter features are extracted. Optionally, the extraction process of multiple parameter features is as follows:

[0077] In one embodiment, for the breathing signal: the breathing characteristics are obtained based on the breathing signal extraction, and the specific steps are: based on the breathing signal within a preset time period, the effective peak points and the effective trough points are determined; the mean between the peak point amplitudes within the preset window time and the mean between the trough point amplitudes are calculated, and then the mean between the peak point amplitudes and the mean between the trough point amplitudes are subtracted to obtain the breathing amplitude characteristics.

[0078] In addition, based on the effective peak point amplitude within the preset window time and the average value between each effective peak point amplitude, the peak point amplitude stability feature is calculated. , the specific calculation formula is as follows:

[0079]

[0080] Where n is the number of valid peak points within the preset window time. To find the absolute value, is the amplitude of the i-th effective peak point within the preset window time, It is the average value between the effective peak point amplitudes within the preset window time. The method for calculating the average value between the effective peak point amplitudes is as follows:

[0081]

[0082] In addition, the time interval sequence between the effective peak points within the preset window time can also be calculated. The time interval sequence refers to the sequence corresponding to the time intervals between all two adjacent effective peak points. Then, based on the time interval sequence, the time interval stability characteristics of the effective peak points within the preset window time are calculated. , the specific calculation method is as follows:

[0083]

[0084] Where n is the number of valid peak points within the preset window time. To find the absolute value, is the i-th time interval within the preset window time, is the time interval mean corresponding to the time interval sequence. The time interval mean is calculated as follows:

[0085]

[0086] It should be noted that the steps for extracting effective peak points are as follows: based on the respiratory signal, determine the minimum point and the maximum point; if the point is a minimum value, and the amplitude is less than the first preset amplitude threshold TH1, then set the minimum value flag to 1, and update the minimum value and TH1 within the preset window time; if the point is a maximum value, and at the same time meets: ① greater than the second preset amplitude threshold TH2; ② a matching effective minimum point appears before the maximum point (the minimum point flag is 1); ③ the time interval between the maximum point and the previous effective maximum point is within the preset time interval, then the point is considered to be a valid peak point, update the maximum value and TH2, and set the effective minimum point flag to 0. In addition, if the above conditions ① and ② are met, and the time interval between the maximum point and the previous effective maximum point is less than the preset time interval, compare the amplitude of the maximum point with the amplitude of the previous effective maximum point, and if the amplitude of the maximum point is greater than the previous effective maximum point, then update the previous effective maximum point as the current maximum point. In addition, if the above conditions ① and ② are met, and the time interval between the maximum point and the previous valid maximum point is greater than the preset time threshold, the maximum point is considered to be a valid maximum point, the valid minimum point flag is set to 0, and the time interval between the maximum point and the previous maximum point is set to an invalid time interval.

[0087] It should be noted that the process of determining the effective trough point is the same as the process of determining the effective peak point, which will not be repeated here.

[0088] In one embodiment, for the heart rate signal: heart rate characteristics are extracted based on the heart rate signal, and the heart rate characteristics include characteristics such as the time interval stability characteristics of the heart rate peaks, the heart rate peak point amplitude stability characteristics and the heart rate amplitude characteristics. The specific calculation method of the heart rate related characteristics is the same as the breathing characteristic calculation method, which will not be repeated here.

[0089] In one embodiment, the in-bed feature is calculated based on the respiratory signal, the heart rate signal and the high-frequency signal. Optionally, in one embodiment, the in-bed feature includes features such as bed power, heart rate energy, respiratory energy, the ratio of heart rate energy to respiratory energy, heart rate energy, high-frequency energy, the ratio of heart rate energy to high-frequency energy, and in-bed power change features, wherein the in-bed power is the sum of the absolute values ​​of the signal energies of the preset passband within the preset window time, and the preset passband can be set according to the situation, for example, the passband is [0.8, 2] Hz; respiratory energy refers to the sum of the absolute values ​​of all respiratory signals within the preset window time; heart rate energy refers to the sum of the absolute values ​​of all heart rate signals within the preset window time; high-frequency energy refers to the sum of all high-frequency energy signals within the preset window; and the in-bed power change feature refers to the ratio of the current in-bed power value to the in-bed power after low-pass filtering.

[0090] In one embodiment, for the pressure signal: median filtering is performed on the pressure signal to extract the pressure feature. In more detail: median filtering is performed on the pressure signal to obtain the target pressure value, and then the first-order difference value is calculated based on the target pressure value within the preset window time. Further, the first pressure mean is calculated based on the target pressure value at the target moment and the target pressure value at the first moment, and the second pressure mean is calculated based on the target pressure value at the second moment and the target pressure value at the third moment. Then, the difference between the pressure mean values ​​is calculated based on the first pressure mean and the second pressure mean, wherein the first moment, the second moment and the third moment are moments before the target moment, for example, the first moment refers to 1 second before the target moment, the second moment refers to 2 seconds before the target moment, and the third moment refers to 3 seconds before the target moment. Based on the first-order difference value and the difference between the pressure mean values ​​and the target pressure value subjected to median filtering, the pressure feature at the target moment is formed.

[0091] Step S12, determining a first environmental state corresponding to the smart mattress at the target time;

[0092] Step S13, based on the parameter characteristics, classifying the first environmental state into scenes to obtain the target scene state of the smart mattress at the target time;

[0093] It should be noted that the target scene state is used to characterize scene states of different interference intensities, and the scene states include scene states of a first interference intensity, a second interference intensity, and a third interference intensity, wherein the first interference intensity is greater than the second interference intensity, and the second interference intensity is greater than the third interference intensity; in other embodiments, more scene states with different interference intensity levels can be divided.

[0094] Specifically, the first environmental state corresponding to the smart mattress at the target time is determined. It should be noted that the first environmental state refers to the working mode of the smart mattress at the target time and / or the working mode state of the smart device associated with the smart mattress, for example, whether the smart mattress turns on vibration massage, turns on the head of the bed to be raised or lowered, turns on the sleep aid mode, and whether there is someone in the bed on the other side in the double bed mode, or the working state of the smart device associated with the smart mattress, such as whether the speaker, dehumidifier, and air conditioner associated with the smart mattress are turned on. Further, based on each of the parameter features, the first environmental state is scene-classified to obtain the target scene state of the smart mattress at the target time. In more detail: each environmental variable in the first environmental state is compared with each environmental variable in the second environmental state at the previous moment. If there are different environmental variables, the power influence factor and pressure influence factor of any changed environmental variable at the target time are updated based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window. Then, based on the power influence factors and pressure influence factors of all environmental variables in the on state at the target moment, the target power influence coefficient and the target pressure influence coefficient at the target moment are calculated; further, based on the target power influence coefficient and the target pressure influence coefficient, the target scene state of the smart mattress at the target moment is obtained respectively. The specific process of scene state classification is specifically described in the following embodiments and will not be repeated here.

[0095] Step S14, using the parameter thresholds updated in real time, respectively performing bed leaving detection on the parameter features to obtain multiple detection results;

[0096] It should be noted that the parameter thresholds include parameter thresholds corresponding to parameter characteristics such as respiratory energy, heart rate energy, high-frequency energy, and pressure information. The parameter thresholds are obtained by self-learning and updating based on bed power, pressure information, and scene status. The specific self-learning and updating process is specifically described in the following embodiments and will not be repeated here.

[0097] Specifically, corresponding bed leaving judgment models can be established for respiratory features, heart rate features, bed-in-bed features, and pressure features, respectively. The parameters in the bed leaving judgment model include respiratory energy, heart rate energy, high-frequency energy, parameter thresholds corresponding to pressure information, etc. It should be noted that the bed leaving judgment model performs bed leaving detection including but not limited to simple threshold comparison, and can also be other binary or multi-classification models for detection, so as to detect the detection results on the four feature dimensions corresponding to the respiratory features, heart rate features, bed-in-bed features, and pressure features. In one embodiment, if the bed-in state is detected, the detection result is set to 1. If the bed leaving state is detected, the detection result is 0.

[0098] Step S15, based on the weight associated with the target scene state, weighted processing is performed on each of the detection results to obtain the target leaving bed detection result.

[0099] It should be noted that different detection results correspond to different weights. Specifically, this embodiment is combined with the target scene state of the smart mattress at the target moment to use the weight associated with the target scene state to perform weighted summation processing on each of the detection results to obtain a weighted calculation result; then the weighted calculation result is compared with a preset threshold, wherein the preset threshold can be set according to actual conditions, for example, set to 0.5. Further, if the weighted calculation result is greater than the preset threshold, it is determined that the target is in the bed when the bed is out of bed detection result; in addition, if the weighted calculation result is not greater than the preset threshold, it is determined that the target is in the bed when the bed is out of bed detection result.

[0100] The embodiment of the present invention, through the above scheme, includes: obtaining a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extracting each parameter feature based on the cardiac impact signal and the pressure signal; determining a first environmental state corresponding to the smart mattress at the target time, wherein the first environmental state refers to the working mode of the smart mattress at the target time and / or the working mode state of the smart device associated with the smart mattress; based on each of the parameter features, performing scene classification on the first environmental state to obtain a target scene state of the smart mattress at the target time, wherein the target scene state is used to characterize scene states with different interference intensities; using each parameter threshold value updated in real time, performing bed leaving detection on each of the parameter features respectively, and obtaining multiple detection results, wherein the parameter threshold value is obtained by self-learning and updating based on the bed power, pressure information and scene state; based on the weight associated with the target scene state, performing weighted processing on each of the detection results to obtain a target bed leaving detection result. The embodiment of the present invention classifies the environment in which the smart mattress is located based on various parameter features, obtains the target scene state of the smart mattress at the target time, and combines the scene state and the parameter thresholds updated by self-learning in different scenes to perform bed leaving judgment on various parameter features at the target time, obtains the final target bed leaving detection result, and effectively improves the accuracy of bed leaving detection.

[0101] like Figure 4 As shown, Figure 4 This is the second flow chart of the bed leaving detection method provided by the present invention. In one embodiment of the present invention, based on each of the parameter characteristics, the first environmental state is scene-classified to obtain the target scene state of the smart mattress at the target time, including:

[0102] Step S21, determining whether each environmental variable in the first environmental state is the same as each environmental variable in the second environmental state at the previous moment;

[0103] It should be noted that the previous moment refers to the moment before the target moment; specifically, the first environmental state corresponding to the smart mattress at the target moment is determined, and then the various environmental variables in the first environmental state are compared to see whether they are the same as the various environmental variables in the second environmental state at the previous moment. For example, the second environmental state at the previous moment is that the smart mattress turns on vibration massage, and the first environmental state at the target moment is that the smart mattress turns on vibration massage and the head of the bed is raised. The environmental variable that is different in the first environmental state is that the head of the bed is raised.

[0104] Step S22, if there are different environmental variables in the first environmental state, determining each first parameter feature in a first preset time window before the target moment and each second parameter feature in a second preset time window after the target moment based on each parameter feature in a preset time period;

[0105] It should be noted that the different environmental variables in the first environmental state refer to environmental variables whose state changes from off to on, and do not include environmental variables whose state changes from on to off. For example, the environmental variables turned on at the previous moment include A and B, and the environmental variables turned on at the target moment include A, C, and D. The different environmental variables existing in the first environmental state are C and D.

[0106] It should be noted that the first parameter feature includes information such as respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the first preset time window; the second parameter feature includes information such as respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the second preset time window. The first preset time window and the second preset time window are set according to actual conditions and are not specifically limited here.

[0107] Step S23, based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window, updating the power influence factor and the pressure influence factor of any target environmental variable at the target time;

[0108] It should be noted that the target environmental variable refers to an environmental variable in the first environmental state that is different from the second environmental state. In the initial stage, the power impact factor and pressure impact factor of each environmental variable are set with corresponding initial values. The initial values ​​of different environmental variables are set according to actual conditions and are not specifically limited here.

[0109] Specifically, based on each first parameter feature within the first preset time window and each second parameter feature within the second preset time window, the influencing factors of the respiratory energy, the heart rate energy, the high-frequency energy, the in-bed power and the pressure information are calculated, and then the influencing factors corresponding to the respiratory energy, the heart rate energy, the high-frequency energy and the in-bed power are added to obtain the parameter influencing factors at the target moment. Further, in combination with the power influencing factors at the previous moment, the parameter influencing factors at the target moment are updated to obtain the power influencing factors of any of the target environmental variables at the target moment, and according to the pressure influencing factors of the pressure information at the previous moment and the influencing factors of the pressure information at the target moment, the pressure influencing factors of any of the target environmental variables at the target moment are calculated. The specific updating process is specifically described in the following embodiments and will not be repeated here.

[0110] Step S24, determining a target power influence coefficient and a target pressure influence coefficient at the target moment based on the power influence factors and pressure influence factors of all target environmental variables;

[0111] Specifically, the power impact factors and pressure impact factors of the remaining environmental variables that are activated at the target moment are determined; the power impact factors of all target environmental variables and the power impact factors of the remaining environmental variables are then added together to obtain the target power impact coefficient, that is, the target power impact coefficient includes all environmental variables that are activated at the target moment; and the pressure impact factors of all target environmental variables and the pressure impact factors of the remaining environmental variables are added together to obtain the target pressure impact coefficient. Continuing with the example of step S23, the environmental variables turned on at the target moment include A, C, and D. Environmental variables C and D are different from the previous moment, and the pressure impact factors of environmental variables A, C, and D need to be added together to obtain the target pressure impact coefficient.

[0112] In addition, if there are no different environmental variables in the first environmental state, that is, the environmental variables in the activated state in the second environmental state already all include the environmental variables in the activated state in the first environmental state, then there is no need to execute the process of updating the power impact factor and the pressure impact factor of any target environmental variable at the target moment, and the power impact factors of the environmental variables in the activated state at the target moment are directly added to obtain the target power impact coefficient. The calculation process of the target pressure impact coefficient is similar and will not be repeated here. For example, the environmental variables turned on at the previous moment include A, B and C, and the environmental variables turned on at the target moment include A and B. There is no need to update the power impact factor and the pressure impact factor, and the power impact factors of the activated A and B environmental variables are directly added to obtain the target power impact coefficient at the target moment.

[0113] Step S25: determining the target scene state based on the target power influence coefficient and the target pressure influence coefficient.

[0114] Specifically, if the target power influence coefficient is greater than the first power coefficient threshold, or the target pressure influence coefficient is greater than the preset pressure coefficient threshold, it is determined that the target scene state is in a scene state of a first interference intensity; wherein the first power coefficient threshold and the pressure coefficient threshold are set according to actual conditions, for example, the first power coefficient threshold is set to 4, and the pressure coefficient threshold is set to 1. In addition, if the target power influence coefficient is greater than the second power coefficient threshold and is less than or equal to the first power coefficient threshold, and the target pressure influence coefficient is less than or equal to the pressure coefficient threshold, it is determined that the target scene state is in a scene state of a second interference intensity, wherein the first power coefficient threshold is greater than the second power coefficient threshold, and the second power coefficient threshold is set according to actual conditions, for example, the second power coefficient threshold is set to 2. In addition, if the target power influence coefficient is less than or equal to the second power coefficient threshold, and the target pressure influence coefficient is less than or equal to the pressure coefficient threshold, it is determined that the target scene state is in a scene state of a third interference intensity.

[0115] The embodiment of the present invention uses the above scheme to compare each environmental variable in the first environmental state at the target moment with each environmental variable in the second environmental state at the previous moment, so as to update the power influence factor and the pressure influence factor of any environmental variable that has changed at the target moment, and then determine the target scene state of the smart mattress at the target moment, so as to make a judgment on leaving the bed by combining the scene state, effectively reduce the influence of environmental interference, and improve the accuracy of leaving the bed detection.

[0116] like Figure 5 As shown, Figure 5 This is the third flow chart of the bed leaving detection method provided by the present invention. In one embodiment of the present invention, based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window, the power influence factor and the pressure influence factor of any target environmental variable at the target time are updated, including:

[0117] Step S31, determining the influencing factors of the respiratory energy, the heart rate energy, the high frequency energy, the in-bed power and the pressure information based on the first parameter characteristics within the first preset time window and the second parameter characteristics within the second preset time window;

[0118] Specifically, based on the respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the first preset time window, the first means corresponding to the respiratory energy, heart rate energy, high frequency energy, bed power and pressure, and the first order difference corresponding to the pressure information are calculated respectively, and then each first order difference is added to obtain the sum of the first first order differences. In addition, based on the respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the second preset time window, the second means corresponding to the respiratory energy, heart rate energy, high frequency energy, bed power and pressure, and the sum of the second first order differences corresponding to the pressure information are calculated. The calculation process of the second sum of first order differences is similar to the calculation process of the first sum of first order differences, which will not be repeated here.

[0119] Further, based on the first mean and the second mean corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power, respectively, the mean ratios of the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power are calculated. For example, the first mean of the respiratory energy is divided by the second mean to obtain the mean ratio of the respiratory energy. Further, based on the mean ratios of the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power, respectively, the influencing factors corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power are calculated. Among them, the influencing factor of the respiratory energy The calculation method is as follows:

[0120]

[0121] Among them, max(·) is to find the maximum value, min(·) is to find the minimum value, is the average ratio of the respiratory energy in the second preset time window to the respiratory energy in the first preset time window. The calculation method of the influencing factor of heart rate energy and the influencing factor of high frequency energy is the same as that of the influencing factor of respiratory energy, which will not be repeated here.

[0122] The calculation method of the influencing factor of bed power is as follows:

[0123]

[0124] in, is the average ratio of the in-bed power in the second preset time window to the average in-bed power in the first preset time window, Represents the influencing factor of bed power.

[0125] In addition, a first difference between the first mean and the second mean corresponding to the pressure information is calculated, and a second difference between the sum of the first first-order differences and the sum of the second first-order differences is calculated. Further, based on the first difference and the second difference, the influence factor corresponding to the pressure information is determined, for example, if the first difference difference is greater than the preset pressure mean threshold, the influence factor of the pressure mean is set to 1, if the first difference difference is not greater than the preset pressure mean threshold, the influence factor of the pressure mean is set to 0; if the second difference is greater than the preset numerical threshold, the influence factor corresponding to the difference is set to 1, if the second difference is not greater than the preset numerical threshold, the influence factor corresponding to the difference is set to 0. Then, the influence factor of the pressure mean and the influence factor corresponding to the difference are added to obtain the influence factor corresponding to the pressure information.

[0126] Step S32, adding the influence factors corresponding to the breathing energy, the heart rate energy, the high frequency energy, and the in-bed power respectively to obtain the parameter influence factor at the target moment;

[0127] Specifically, the influencing factors corresponding to the respiratory energy, the heart rate energy, the high-frequency energy, and the in-bed power are directly added together to obtain the parameter influencing factor at the target moment.

[0128] Step S33, calculating the power impact factor of any target environmental variable according to the power impact factor at the previous moment and the parameter impact factor at the target moment;

[0129] Step S34, calculating the pressure influence factor of any of the target environmental variables according to the pressure influence factor of the pressure information at the previous moment and the influence factor of the pressure information at the target moment.

[0130] Specifically, determine the number of variables that have different environmental variables; determine the time interval between the target moment and the last update moment of the coefficient, in one embodiment, the unit of the time interval is set to day, wherein the update coefficient is used to characterize the power impact factor of the previous moment and the weight coefficient of the parameter impact factor of the target moment in the power impact factor update process; determine the power impact factor corresponding to each environmental variable at the previous moment; based on the time interval, the number of variables, and the power impact factor corresponding to each environmental variable at the previous moment, determine the update coefficient at the target moment. The calculation formula of the update coefficient is as follows:

[0131]

[0132] in, Represents the impact coefficient of environmental change quantity, where if the number of variables is greater than 1, Set to 1, otherwise is 0. Indicates a time interval; Indicates the environmental impact coefficient before the environmental change. During the process of updating the power impact factor, if the sum of the power impact factors of the environmental variables before the change is less than 1 and the sum of the pressure impact factors is less than 0.5, then is 1, otherwise is 0. Indicates the ratio of the power impact factor of the environment variable to be updated to the sum of the power impact factors of all target environment variables. The target environment variable refers to the environment variable in the first environment state that is different from the second environment state. For example, at the target moment, two environment variables that change, namely, vibration massage on and bed head lift, appear at the same time. When updating the power impact factor of vibration massage, It is the power influence factor of vibration massage divided by the sum of the power influence factor of vibration massage and the power influence factor of head of bed lifting.

[0133] Further, based on the update coefficient at the target moment, the power impact factor at the previous moment and the parameter impact factor at the target moment, the power impact factor of any target environmental variable at the target moment is calculated, and the calculation formula is as follows:

[0134]

[0135] in, represents the update coefficient, represents the power impact factor at the previous moment, Indicates the parameter influencing factor at the target moment.

[0136] It should be noted that the calculation process of the pressure influence factor is similar to the calculation process of the power influence factor, which will not be repeated here.

[0137] The embodiment of the present invention, through the above scheme, realizes updating the power influence factor and the pressure influence factor of any environmental variable that changes at the target time based on each first parameter feature within the first preset time window and each second parameter feature within the second preset time window, and then determining the target scene state of the smart mattress at the target time according to the power influence factor and the pressure influence factor, so as to make a judgment on leaving the bed by combining the scene state, effectively reducing the influence of environmental interference and improving the accuracy of leaving the bed detection.

[0138] like Figure 6 As shown, Figure 6 This is the fourth flow chart of the bed leaving detection method provided by the present invention; in one embodiment of the present invention, the self-learning update process of the parameter threshold is as follows:

[0139] Step S41, obtaining in-bed power and pressure information within a third preset time window;

[0140] Step S42, based on the in-bed power and pressure information within the third preset time window, determining whether the in-bed power and the pressure information are in a stable state;

[0141] Step S43: if the smart mattress is in a stable state and the power in bed at the target time is less than the first preset power threshold, then the parameter threshold corresponding to each parameter feature of the smart mattress in an unmanned state is updated;

[0142] Step S44: if the smart mattress is in a stable state and the power in bed at the target moment is greater than the second preset power threshold, the parameter thresholds corresponding to the various parameter characteristics of the smart mattress in a occupied state are updated.

[0143] Specifically, the in-bed power and pressure information within the third preset time window are obtained, and then it is determined whether the in-bed power and the pressure information are in a stable state. More specifically, based on the in-bed power within the third preset time window, the power standard deviation and the power mean are calculated, and based on the pressure information within the third preset time window, the pressure standard deviation is calculated; and then the ratio between the power standard deviation and the power mean is calculated; if the ratio is greater than a preset ratio threshold, or the pressure standard deviation is greater than a preset standard deviation threshold, it is determined to be in an unstable state, otherwise it is determined to be in a stable state.

[0144] Further, in the case of being in a stable state: determine whether the power in bed is less than the first preset power threshold, the first preset power threshold is determined in combination with the scene state at the target moment, optionally, the first preset power threshold is a specified multiple of the minimum power in bed under the scene state, for example, the multiple is 1.5 under the scene state of the third interference intensity, if the multiple is 2 under the scene state of the second interference intensity, the multiple is 2.5 under the strong interference scene state of the first interference intensity, in other embodiments, the first preset power threshold increases with a fixed coefficient over time, the threshold increases by 0.001 per second under the scene state of the third interference intensity, increases by 0.003 per second under the scene state of the second interference intensity, and increases by 0.009 per second under the strong interference scene state of the first interference intensity. Further, if the power in bed is less than the first preset power threshold, it is determined that the smart mattress is in an unmanned state, and then the parameter thresholds corresponding to the various parameter features of the smart mattress in the unmanned state are updated, that is, the parameter thresholds of the respiratory energy, heart rate energy, high frequency energy, and pressure information in the unmanned state are updated. The update rules of the above-mentioned parameter thresholds are the same. Taking the parameter threshold of pressure information as an example, the update rules are as follows:

[0145]

[0146] in, is the updated parameter threshold of the unmanned state, is the parameter threshold of the unmanned state before updating, P is the current pressure value, is the parameter update coefficient. If the current pressure value is less than , the parameter update coefficient is 0.2, if the current pressure value is not less than , the parameter update coefficient is 0.025. In other embodiments, the parameter threshold value will be automatically increased by a fixed value when it is determined that there is no one.

[0147] In addition, in the case of a stable state: if the in-bed power at the target moment is greater than the second preset power threshold, it is determined that the smart mattress is in an occupied state, and then the parameter thresholds corresponding to the various parameter characteristics of the smart mattress in the occupied state are updated, that is, the parameter thresholds of the breathing energy, heart rate energy, high-frequency energy, and pressure information in the occupied state are updated. Among them, the second preset power threshold is determined in combination with the scene state at the target moment, and the second preset power threshold is a specified multiple of the minimum in-bed power under the scene state. For example, the multiple is 6 in the non-interference scene state of the third interference intensity, if the multiple is 8 in the scene state of the second interference intensity, the multiple is 10 in the strong interference scene state of the first interference intensity. Among them, the update formula of each threshold in the occupied state is the same as the update of the threshold in the unoccupied state, that is:

[0148]

[0149] in, is the parameter threshold of the updated occupied state, is the parameter threshold of the occupied state before updating, P is the current pressure value, is the parameter update coefficient, the current pressure value is less than , the parameter update coefficient is 0.1, and the current pressure value is not less than , the parameter update coefficient is 0.02. In other embodiments, the parameter threshold will increase by a fixed value when it is determined that there is someone present.

[0150] Through the above scheme, the embodiment of the present invention realizes self-learning and updating of various parameter thresholds according to the bed power, pressure information and scene status, so that bed leaving detection can be performed according to the self-learning updated parameter thresholds, effectively improving the accuracy of bed leaving detection.

[0151] The bed exit detection device provided by the present invention is described below. The bed exit detection device described below and the bed exit detection method described above can be referenced to each other.

[0152] Figure 7 Schematic diagram of the structure of the bed leaving detection device provided by the present invention, such as Figure 7As shown, a bed leaving detection device according to an embodiment of the present invention comprises:

[0153] An acquisition module 51 is used to acquire a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extract various parameter features based on the cardiac impact signal and the pressure signal;

[0154] A determination module 52 is used to determine a first environmental state corresponding to the smart mattress at a target time, wherein the first environmental state refers to a working mode of the smart mattress at the target time and / or a working mode state of a smart device associated with the smart mattress;

[0155] A classification module 53, configured to classify the first environmental state based on the parameter characteristics, and obtain a target scene state of the smart mattress at a target time, wherein the target scene state is used to characterize scene states of different interference intensities;

[0156] A detection module 54 is used to perform bed leaving detection on each parameter feature using each parameter threshold updated in real time to obtain multiple detection results, wherein the parameter threshold is obtained by self-learning and updating based on bed power, pressure information and scene status;

[0157] The weighting module 55 is used to perform weighted processing on each of the detection results based on the weight associated with the target scene state to obtain the target bed leaving detection result.

[0158] It should be noted here that the above-mentioned device provided in the embodiment of the present invention can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects that are the same as the method embodiment in this embodiment will not be described in detail here.

[0159] Figure 8 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 8 As shown, the electronic device may include: a processor 61, a memory 62, a communication interface 63 and a communication bus 64, wherein the processor 61, the memory 62 and the communication interface 63 communicate with each other through the communication bus 64. The processor 61 may call the logic instructions in the memory 62 to execute the bed leaving detection method.

[0160] In addition, the logic instructions in the above-mentioned memory 62 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0161] On the other hand, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it is implemented to perform the bed leaving detection method provided by the above methods. The method includes: obtaining a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extracting each parameter feature based on the cardiac impact signal and the pressure signal; determining the first environmental state corresponding to the smart mattress at the target time, wherein the first environmental state refers to the working mode of the smart mattress at the target time and / or the working mode state of the smart device associated with the smart mattress; based on each of the parameter features, the first environmental state is scene-classified to obtain the target scene state of the smart mattress at the target time, wherein the target scene state is used to characterize the scene state of different interference intensities; using each parameter threshold updated in real time, each of the parameter features is respectively detected in bed leaving, and multiple detection results are obtained, wherein the parameter threshold is obtained by self-learning update based on the bed power, pressure information and scene state; based on the weight associated with the target scene state, each of the detection results is weighted to obtain the target bed leaving detection result.

[0162] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0163] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting when leaving bed, characterized in that: include: Acquire a cardiac impact signal and a pressure signal collected by a sensor within a preset time period, and extract various parameter features based on the cardiac impact signal and the pressure signal; Determine a first environmental state corresponding to the smart mattress at the target time, wherein the first environmental state refers to the working mode of the smart mattress at the target time and / or the working mode state of the smart device associated with the smart mattress; Based on the parameter characteristics, the first environmental state is subjected to scene classification to obtain a target scene state of the smart mattress at a target time, wherein the target scene state is used to characterize scene states of different interference intensities; Using the parameter thresholds updated in real time, each parameter feature is detected while leaving the bed to obtain multiple detection results, wherein the parameter threshold is obtained by self-learning and updating based on the power in bed, pressure information and scene status; Based on the weight associated with the target scene state, each of the detection results is weighted to obtain the target bed leaving detection result.

2. The method for detecting leaving bed according to claim 1, characterized in that: The first environmental state is subjected to scene classification based on the parameter characteristics to obtain the target scene state of the smart mattress at the target time, including: Determine whether each environmental variable in the first environmental state is the same as each environmental variable in the second environmental state at a previous moment, wherein the previous moment refers to a moment before the target moment; If there are different environmental variables in the first environmental state, determining each first parameter feature in a first preset time window before the target moment and each second parameter feature in a second preset time window after the target moment based on each parameter feature in a preset time period; Based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window, updating a power impact factor and a pressure impact factor of any target environmental variable at the target time, wherein the target environmental variable refers to an environmental variable in the first environmental state that is different from the second environmental state; Determining a target power influence coefficient and a target pressure influence coefficient at the target time based on the power influence factors and pressure influence factors of all target environmental variables; The target scene state is determined based on the target power influence coefficient and the target pressure influence coefficient.

3. The method for detecting leaving bed according to claim 2, characterized in that: The first parameter characteristics include respiratory energy, heart rate energy, high frequency energy, in-bed power and pressure information within a first preset time window; The second parameter characteristics include respiratory energy, heart rate energy, high frequency energy, in-bed power and pressure information within a second preset time window; Accordingly, updating the power impact factor and the pressure impact factor of any target environmental variable at the target time based on each first parameter feature in the first preset time window and each second parameter feature in the second preset time window includes: determining, based on each first parameter feature within the first preset time window and each second parameter feature within the second preset time window, influencing factors of the respiratory energy, the heart rate energy, the high frequency energy, the in-bed power and the pressure information; Adding the influencing factors corresponding to the breathing energy, the heart rate energy, the high-frequency energy, and the in-bed power, respectively, to obtain the parameter influencing factor at the target moment; Calculate the power impact factor of any target environmental variable according to the power impact factor at the previous moment and the parameter impact factor at the target moment; The pressure influence factor of any target environmental variable is calculated based on the pressure influence factor of the pressure information at the previous moment and the influence factor of the pressure information at the target moment.

4. The method for detecting leaving bed according to claim 3, characterized in that: The determining, based on each first parameter feature within the first preset time window and each second parameter feature within the second preset time window, the influencing factors of the respiratory energy, the heart rate energy, the high frequency energy, the in-bed power, and the pressure information comprises: Based on the respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the first preset time window, first means corresponding to the respiratory energy, the heart rate energy, the high frequency energy, the bed power and the pressure, respectively, and a sum of first first-order differences corresponding to the pressure information are calculated; Based on the respiratory energy, heart rate energy, high frequency energy, bed power and pressure information within the second preset time window, second means corresponding to the respiratory energy, the heart rate energy, the high frequency energy, the bed power and the pressure, respectively, and a sum of second first-order differences corresponding to the pressure information are calculated; Based on the first mean and the second mean respectively corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power, the mean ratios respectively corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power are calculated; Based on the mean ratios of the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power, respectively, the influence factors corresponding to the respiratory energy, the heart rate energy, the high frequency energy and the in-bed power are calculated; An impact factor corresponding to the pressure information is determined based on a first difference between a first mean and a second mean corresponding to the pressure information and a second difference between the sum of the first first-order differences and the sum of the second first-order differences.

5. The method for detecting leaving bed according to claim 3, characterized in that: The step of calculating the power impact factor of any target environmental variable according to the power impact factor at the previous moment and the parameter impact factor at the target moment includes: Determine the number of variables that have different environmental variables; Determine the time interval between the target moment and the last update moment of the coefficient, wherein the update coefficient is used to characterize the weight coefficient of the power impact factor at the previous moment and the parameter impact factor at the target moment in the power impact factor update process; Determine the power impact factor corresponding to each environmental variable at the previous moment; Determine the update coefficient at the target moment based on the time interval, the number of variables, and the power impact factor corresponding to each environmental variable at the previous moment; The power impact factor of any of the target environmental variables is calculated based on the update coefficient at the target moment, the power impact factor at the previous moment, and the parameter impact factor at the target moment.

6. The method for detecting leaving bed according to claim 2, characterized in that: The step of determining the target power influence coefficient and the target pressure influence coefficient at the target time based on the power influence factors and the pressure influence factors of all target environmental variables includes: Determine the power impact factor and the pressure impact factor of the remaining environmental variables in the activated state at the target time; Adding the power impact factors of all target environmental variables and the power impact factors of the remaining environmental variables to obtain the target power impact coefficient; The pressure influence factors of all target environmental variables and the pressure influence factors of the remaining environmental variables are added together to obtain the target pressure influence coefficient.

7. The method for detecting leaving bed according to claim 2, characterized in that: The determining the target scene state based on the target power influence coefficient and the target pressure influence coefficient includes: If the target power influence coefficient is greater than a first power coefficient threshold, or the target pressure influence coefficient is greater than a preset pressure coefficient threshold, determining that the target scene state is in a scene state of a first interference intensity; If the target power influence coefficient is greater than the second power coefficient threshold and less than or equal to the first power coefficient threshold, and the target pressure influence coefficient is less than or equal to the pressure coefficient threshold, determining that the target scene state is in a scene state of a second interference intensity; If the target power influence coefficient is less than or equal to the second power coefficient threshold, and the target pressure influence coefficient is less than or equal to the pressure coefficient threshold, determining that the target scene state is in a scene state of a third interference intensity; The first power coefficient threshold is greater than the second power coefficient threshold, the first interference intensity is greater than the second interference intensity, and the second interference intensity is greater than the third interference intensity.

8. The method for detecting leaving bed according to claim 1, characterized in that: The self-learning update process of the parameter threshold is as follows: Acquiring in-bed power and pressure information within a third preset time window; Based on the in-bed power and pressure information within a third preset time window, determining whether the in-bed power and the pressure information are in a stable state; If it is in a stable state, and the power in bed at the target time is less than the first preset power threshold, then the parameter thresholds corresponding to the various parameter characteristics of the smart mattress when no one is in the bed are updated; If it is in a stable state, and the power in bed at the target moment is greater than the second preset power threshold, then the parameter thresholds corresponding to the various parameter characteristics of the smart mattress when it is in a occupied state are updated; The first preset power threshold and the second preset power threshold are determined in combination with the scene state at the target moment.

9. The method for detecting leaving bed according to claim 8, characterized in that: The determining, based on the in-bed power and pressure information within the third preset time window, whether the in-bed power and the pressure information are in a stable state includes: Determine a power standard deviation and a power mean corresponding to the in-bed power within the third preset time window, and a pressure standard deviation between the pressure information within the third preset time window; determining a ratio between the power standard deviation and the power mean; If the ratio is greater than a preset ratio threshold, or the pressure standard deviation is greater than a preset standard deviation threshold, it is determined to be in an unstable state.

10. The method for detecting leaving bed according to claim 1, characterized in that: The weighted processing of each of the detection results based on the weight associated with the target scene state to obtain the target leaving bed detection result includes: Performing weighted sum processing on the weight associated with the target scene state and each of the detection results to obtain a weighted calculation result; If the weighted calculation result is greater than a preset threshold, it is determined that the target is in bed when leaving the bed. If the weighted calculation result is not greater than a preset threshold, it is determined that the target is in the out-of-bed state in the out-of-bed detection result.

Images