Sleep state determination device, sleep state determination method, program, and control system

Abstract

To accurately determine sleep state.SOLUTION: A sleep state determination device includes a sensor, a first moving average calculation unit that calculates a first moving average over a first period using a predetermined moving average method on the basis of the respiratory rate obtained from sensing data acquired by the sensor, a second moving average calculation unit that calculates a second moving average over a second period longer than the first period on the basis of the respiratory rate, and a sleep state determination unit that uses calculation results using the first and second moving averages to determine sleep states, including REM sleep and non-REM sleep.SELECTED DRAWING: Figure 5

Description

[Technical Field]

[0001] The present invention relates to a sleep state determination device, a sleep state determination method, a program, and a control system. [Background technology]

[0002] Techniques have been proposed for detecting a person's sleep state and controlling electronic devices in accordance with the detection results. For example, Patent Document 1 listed below describes a technique for detecting a person's sleep state and setting control parameters for an air conditioner in accordance with the detection results. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-202659 Summary of the Invention [Problem to be solved by the invention]

[0004] The technology described in Patent Document 1 calculates at least one of the respiratory cycle and the respiratory depth at a certain timing, and calculates a stability level according to the calculated respiratory cycle, etc. The calculated stability level is then used as an index of a person's sleep state. However, the technology described in Patent Document 1 calculates the stability level by looking at the fluctuation range of the respiratory cycle, etc., which occurs over time. Therefore, when the respiratory cycle, etc., temporarily fluctuates significantly, an accurate stability level cannot be calculated, which may result in an error in determining the sleep state.

[0005] An object of the present invention is to provide a sleep state determination device, a sleep state determination method, a program, and a control system that can solve the above problems and determine the sleep state with high accuracy. [Means for solving the problem]

[0006] The present invention provides The sensor and a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit that calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result using the first moving average and the second moving average; A sleep state determination device.

[0007] The present invention provides a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit calculates a second moving average for a second period longer than the first period based on the respiratory rate; the sleep state determination unit determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result using the first moving average and the second moving average; A sleep state determination method.

[0008] The present invention provides a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit calculates a second moving average for a second period longer than the first period based on the respiratory rate; the sleep state determination unit determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result using the first moving average and the second moving average; The present invention provides a program for causing a computer to execute a sleep state determination method.

[0009] The present invention provides A sleep state determination device and an electrical device are included, The sleep state determination device The sensor and a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit that calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result using the first moving average and the second moving average, Electrical equipment is an electrical appliance control unit that performs control in accordance with the determination result of the sleep state determination unit; It is a control system.

[0010] The present invention provides A sleep state determination device and an electrical device are included, The sleep state determination device The sensor and a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit that calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state, including a REM sleep state and a non-REM sleep state, using a calculation result obtained by using the first moving average and the second moving average; an electrical appliance control signal transmitter that transmits a control signal for controlling the electrical appliance in accordance with a determination result of the sleep state determiner; It is a control system. [Brief explanation of the drawings]

[0011] [Figure 1] 1A and 1B are diagrams for explaining an outline of the present invention. [Figure 2] 1 is a diagram illustrating a sleep state determination system according to an embodiment. [Figure 3] 1A and 1B are diagrams illustrating an example of the appearance of a sleep state determining device. [Figure 4]1A and 1B are diagrams illustrating an example of the internal configuration of a sleep state determining device. [Figure 5] 1 is a block diagram for explaining an example of the electrical configuration of a sleep state determining device according to a first embodiment. [Figure 6] 1A and 1B are diagrams to be referred to when describing the processing executed by the sleep state determining device according to the first embodiment. [Figure 7] 1A to 1C are diagrams to be referred to when describing the processing executed by the sleeping state determining device according to the first embodiment. [Figure 8] 1A shows the results of sleep state determination by the processing performed in the first embodiment, and FIG. 1B shows the results of sleep state determination using an electroencephalograph. [Figure 9] 4 is a flowchart showing a flow of processing performed by the sleep state determining device according to the first embodiment. [Figure 10] 1A and 1B are diagrams to be referred to when explaining the effects obtained in the first embodiment. [Figure 11] 10A and 10B are diagrams to which reference will be made when explaining issues to be considered in the second embodiment. [Figure 12] FIG. 10 is a block diagram illustrating an example of the electrical configuration of a sleep state determining device according to a second embodiment. [Figure 13] 10A to 10C are diagrams to be referred to when explaining an example of the operation of a respiratory rate variation calculation unit according to the second embodiment. [Figure 14] 1A shows a graph showing the binarized total amount of respiratory rate fluctuation, and FIG. 1B shows the results of sleep state determination using an electroencephalograph. [Figure 15] FIG. 10A is a diagram for explaining a determination method of a sleep state determination device according to a second embodiment, and FIG. 10B is a diagram showing a sleep state determination result using an electroencephalograph. [Figure 16] 10 is a flowchart showing a flow of processing performed by a sleep state determining device according to a second embodiment. [Figure 17]FIG. 10 is a block diagram illustrating an example of the electrical configuration of a sleep state determining device according to a third embodiment. [Figure 18] FIG. 10 is a diagram to be referred to when explaining a sleep detection unit according to a third embodiment. [Figure 19] 10A to 10D are diagrams to be referred to when explaining an example of the operation of the sleep detection unit according to the third embodiment. [Figure 20] FIG. 10 is a diagram illustrating an example of a control system according to an application example. [Figure 21] FIG. 10 is a diagram for explaining another example of a control system according to an application example. [Figure 22] FIG. 10 is a diagram for explaining a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Note that the embodiments described below are preferred specific examples of the present invention, and the content of the present invention is not limited to these embodiments. The description will be given in the following order. <Summary of the Invention> First Embodiment <Second embodiment> <Third embodiment> <Application example> <Modification>

[0013] In this specification, "larger than A (greater than A)" may be larger than A (not including A) or equal to or greater than A. Furthermore, "smaller than B (below B)" may be equal to or less than B, or may be less than B. These terms are to be interpreted in a manner that does not cause any technical contradiction.

[0014] <Summary of the Invention> An overview of the present invention will be described with reference to Figures 1A and 1B. Figure 1A is a graph showing an example of changes in the respiratory rate (unit: bpm (breaths per minute)) of a subject (person) according to the present invention. The subject's respiratory rate is measured, for example, using a radio wave sensor (details will be described later). The horizontal axis of the graph shown in Figure 1A represents time, and the vertical axis represents the respiratory rate. Furthermore, waveform WA in the graph of Figure 1A represents the respiratory rate at each timing, and waveform WB in the graph of Figure 1A represents a moving average of the respiratory rate over a predetermined time width.

[0015] The graph in FIG. 1B is intended as a comparison for verifying the accuracy of the sleep state determination method according to the present invention. It shows the measurement results (actual measurement results) of a person's sleep state measured using an electroencephalograph. The electroencephalograph is attached to, for example, a person whose sleep state is being measured. A known electroencephalograph with relatively high measurement accuracy, for example, an accuracy of about 85% or higher, is used as the electroencephalograph. The horizontal axis of the graph in FIG. 1B indicates the same time axis as the horizontal axis of the graph in FIG. 1A. That is, the time axes of FIGS. 1A and 1B correspond to each other. The vertical axis of the graph in FIG. 1B indicates sleep states classified based on the results of using the electroencephalograph. The sleep states in this example include a wakefulness state (not a sleep state), rapid eye movement (REM) sleep, and non-REM sleep. REM sleep is a state in which the brain is relatively active and memory is organized and consolidated. Non-REM sleep is a state in which the cerebrum is considered to be resting. Non-REM sleep is classified into three stages (Non-REM1 to 3 (3 indicates deeper sleep)) based on the depth of sleep. NoScored indicates a state that is not classified as any of the above states. The black areas on the time axis indicate the sleep state at that time (the sleep state shown on the left side of Figure 1B).

[0016] For example, the respiratory rate tends to be high during REM sleep, as seen in sections TA and TB in Figure 1A. In the present invention, the respiratory rate, which is correlated with the sleep state, is used to accurately determine the sleep state of a person. Specific details of the present invention will be described below with reference to embodiments.

[0017] First Embodiment [Sleep state determination system] 2 shows a sleep state determination system (sleep state determination system 100) according to a first embodiment. The sleep state determination system 100 is also applicable to other embodiments and modified examples described below. The sleep state determination system 100 includes, for example, a sleep state determination device 1 and a person whose sleep state is determined by the sleep state determination device 1. In the following description, the person whose sleep state is determined will also be referred to as a user U, as appropriate.

[0018] The sleep state determination device 1 determines the sleep state of a sleeping user U. The determination result is used, for example, as material for a medical diagnosis regarding the sleep of the user U or for controlling electrical appliances. As will be described in detail later, the sleep state determination device 1 determines the sleep state of the user U using a non-contact sensor that is not attached to the user U. The sleep state determination device 1 is installed, for example, about 20 to 30 cm away from the user U (for example, next to the user U's pillow). The sleep state determination device 1 may also be attached to a wall or ceiling of a bedroom.

[0019] [Sleep state determination device] (Example of appearance) 3A and 3B are diagrams showing an example of the appearance of the sleep state judging device 1. The sleep state judging device 1 has a housing 10. As shown in FIGS. 3A and 3B, the housing 10 has, for example, a substantially box-like (cubic) shape. However, the shape of the housing 10 may be a shape other than a box-like (for example, a tree-like or triangular prism-like) shape. The housing 10 includes, for example, a dish-shaped bottom 10A located on the lower side, and a case 10B attached to the bottom 10A. The bottom 10A and the case 10B are made of, for example, resin. Of these, the case 10B may be made of a transparent resin. The sleep state judging device 1 is used, for example, with the bottom 10A placed on a flat surface.

[0020] A power supply terminal 11 is provided on a predetermined side of the bottom 10A. The power supply terminal 11 is, for example, a type that receives power from an external source via a USB (Universal Serial Bus) connector, but is not limited to this. The sleep state determination device 1 according to the embodiment operates using power supplied from an external source such as a commercial power source, but may also operate using power supplied from a battery. The battery may be a primary battery such as a dry cell battery, or a rechargeable secondary battery.

[0021] (Internal configuration example) 4A and 4B are diagrams illustrating an example of the configuration inside the housing 10. A plate-shaped main board 12 is arranged inside the bottom 10A. The main board 12 is provided with ICs (Integrated Circuits) (not shown) such as a control unit and a communication module. Power supplied from the power supply terminal 11 is supplied to the main board 12, and after being appropriately converted into voltage by a regulator or the like (not shown) provided on the main board 12, is supplied to each part of the sleep state determination device 1.

[0022] A motor mechanism 13 is installed vertically near the center of the main board 12. The motor mechanism 13 is formed using a rotary motor so that the upper end of the motor mechanism 13 rotates. A circular LED (Light Emitting Diode) board 16 is attached to the end of the motor mechanism 13 via a cylindrical shaft. The LED board 16 rotates when the rotary motor described above operates. The LED board 16 is provided with, for example, three infrared LEDs 17. The infrared LEDs 17 can transmit, to electrical devices 30 described below, infrared signals that are the same as those transmitted from an infrared remote control compatible with each electrical device 30.

[0023] A sensor substrate 18 is provided on a side of the LED substrate 16. The sensor substrate 18 has, for example, a thin plate shape, and is fixed to the LED substrate 16 in a predetermined position with its main surface (the surface occupying a wide area) extending in the vertical direction and attached to the cylindrical shaft. A sensor unit 21 is provided on the outward-facing main surface of the sensor substrate 18 (the surface opposite to the surface facing the LED substrate 16). The orientation of the sensor unit 21 can be changed as desired by rotating the rotary motor. A rotary connection 14 is provided between the motor mechanism 13 and the LED substrate 16 to ensure electrical connection between the main substrate 12 and the sensor substrate 18 even when the cylindrical shaft rotates. The rotary connection 14 is a so-called slip ring.

[0024] (Example of electrical configuration) 5 is a block diagram showing an example of the electrical configuration of the sleep state judging device 1. In general, the sleep state judging device 1 has, in addition to the above-mentioned sensor unit 21, a control unit 20, a sleep state judging unit 22, and a motor 23. Of course, the sleep state judging device 1 may have other components besides these.

[0025] The control unit 20 is configured by, for example, a CPU (Central Processing Unit), and controls each unit of the sleep state judging device 1 in an integrated manner.

[0026] The sensor unit 21 includes, for example, a sensor 21A and a sensing data processing unit 21B.

[0027] Sensor 21A is, for example, a radio wave sensor that emits radio waves, receives the radio waves reflected by a measurement object (e.g., user U), and detects the movement of user U by the Doppler effect. Sensor 21A is also a non-contact sensor that does not need to be attached to user U. Radio wave sensors are well known, and examples thereof include microwave sensors and millimeter wave sensors. Radio wave sensors may also include Doppler sensors and FMCW (Frequency Modulated Continuous Wave) sensors. By configuring sensor 21A as a radio wave sensor, radio waves can pass through clothing even when user U is wearing clothing, so the movement of user U can be detected.

[0028] As described above, the orientation of the sensor unit 21, specifically the sensor 21A, can be changed by rotating the rotary motor. That is, the sleep state determination device 1 according to this embodiment is configured to change the radiation direction of the radio waves emitted from the sensor 21A by changing the orientation of the sensor 21A. In this way, the sleep state determination device 1 has a function that can change the radiation direction of the radio waves emitted from the sensor 21A, a so-called swivel function. The swivel function of the sleep state determination device 1 allows the radiation direction of the radio waves to be directed toward the sleeping position of the user U, thereby increasing the flexibility of the installation location of the sleep state determination device 1. Furthermore, even if there are multiple users U, the radiation direction of the radio waves can be directed toward the sleeping position of each user U as appropriate, so that the sleep states of multiple users U can be determined by a single sleep state determination device 1.

[0029] The sensing data processing unit 21B performs known signal processing on the sensing data acquired by the sensor 21A. Examples of known signal processing include AD (Analog to Digital) conversion, fast Fourier transform processing, and filter processing. The AD conversion processing is performed every predetermined period (for example, every 20 mec). The sensing data processing unit 21B performs signal processing on the sensing data acquired by the sensor 21A, thereby obtaining the respiration rate of the user U based on the sensing data. The obtained respiration rate is supplied from the sensor unit 21 to the sleep state determination unit 22. Respiration rate sensing is started, for example, when the sleep state determination device 1 is powered on. Respiration rate sensing may be started when a predetermined user operation is used as a trigger.

[0030] The sleeping state determination unit 22 includes, for example, a first moving average calculation unit 22A, a second moving average calculation unit 22B, and a sleeping state determination unit 22C. The sleeping state determination unit 22 can also be configured as an integrated unit with the control unit 20 by implementing functions corresponding to the first moving average calculation unit 22A, the second moving average calculation unit 22B, and the sleeping state determination unit 22C in the control unit 20.

[0031] The first moving average calculation unit 22A calculates a first moving average for a first period based on the respiratory rate obtained from the sensing data acquired by the sensor 21A. The respiratory rate obtained from the sensing data is the respiratory rate obtained by the sensor unit 21. The first period is not limited to a specific time. In this example, the first period is described as 10 minutes. The first moving average calculation unit 22A calculates the first moving average using a predetermined moving average method. Examples of moving average methods include the simple moving average method, the weighted moving average method, and the exponentially smoothed moving average method. Among these, the exponentially smoothed moving average method is preferred from the viewpoint of obtaining high accuracy. The exponentially smoothed moving average method can be generalized by the following equation 1.

[0032]

number

[0033] The second moving average calculation unit 22B calculates a second moving average for a second period based on the respiratory rate obtained from the sensing data acquired by the sensor 21A. The respiratory rate obtained from the sensing data is the respiratory rate obtained by the sensor unit 21. The second period is not limited to a specific time, but is set to a period longer than the first period described above. In this example, the second period is described as 30 minutes. The second moving average calculation unit 22B calculates the second moving average using a predetermined moving average method. Examples of moving average methods include a simple moving average method, a weighted moving average method, and an exponentially smoothed moving average method. It is preferable that the moving average used by the second moving average calculation unit 22B be the same as the predetermined moving average method used by the first moving average calculation unit 22A. Among these, the exponentially smoothed moving average method is preferable from the viewpoint of obtaining high accuracy.

[0034] The sleep state determination unit 22C determines the sleep state using a calculation result that uses the first moving average calculated by the first moving average calculation unit 22A and the second moving average calculated by the second moving average calculation unit 22B. Details of the process performed by the sleep state determination unit 22C will be described later.

[0035] The motor 23 is a general term for the rotary motor and the like that constitute the above-mentioned motor mechanism 13. The rotary operation of the motor 23 is controlled by the control unit 20.

[0036] [Example of operation] Next, an example of the operation of the sleep state judging device 1 will be described. In the following description, the first moving average corresponds to a short-term moving average, and therefore the first moving average is also referred to as a short-term moving average. Furthermore, the second moving average corresponds to a long-term moving average, and therefore the second moving average is also referred to as a long-term moving average.

[0037] The sleep state determination unit 22C of the sleep state determination device 1 performs a calculation using a short-term moving average, which is a short-term (10 minutes in this embodiment) moving average, and a long-term moving average, which is a long-term (30 minutes in this embodiment) moving average. The calculation is, for example, a calculation to obtain the difference between the short-term moving average and the long-term moving average (short-term moving average - long-term moving average). However, instead of the difference, the calculation may be, for example, a calculation to obtain the short-term moving average divided by the long-term moving average.

[0038] The sleep state determination unit 22C uses the results of the above calculations to determine the sleep state of the user U. The sleep state includes a REM sleep state and a non-REM sleep state. For example, as shown in FIG. 6A, when the short-term moving average (the dotted line in FIG. 6A) exceeds the long-term moving average (the solid line in FIG. 6A) (when the difference is equal to or greater than a certain value), the sleep state determination unit 22C determines that the breathing rate is on the rise and determines that the sleep state of the user U is a "REM sleep state."

[0039] On the other hand, as shown in Figure 6B, when the short-term moving average (the dotted line in Figure 6B) is lower than the long-term moving average (the solid line in Figure 6B), i.e., when the difference is equal to or greater than a certain value, the sleep state determination unit 22C determines that the breathing rate is showing a downward trend and determines that the sleep state of the user U is a "non-REM sleep state."

[0040] 7A to 7C, the processing performed by the sleep state determination unit 22C will be specifically described. The horizontal axis of the graph shown in FIG. 7A is the time axis, and the vertical axis is the value of the short-term moving average or the long-term moving average. In addition, line L1 in the graph shown in FIG. 7A indicates changes in the short-term moving average, and line L2 indicates changes in the long-term moving average. The horizontal axis of the graph shown in FIG. 7B is the time axis. The time axes shown in FIGS. 7A and 7B are common time axes. The vertical axis of FIG. 7B indicates values ​​(scale) corresponding to the calculation result obtained by subtracting the long-term moving average from the short-term moving average. In addition, line L3 in FIG. 7B indicates changes in the calculation result. For example, two threshold values ​​THA and THB are set for the calculation result. The threshold value THA is set to 0.5, for example. The threshold value THB is set to -0.5, for example. The values ​​of the threshold values ​​THA and THB are not limited to 0.5 or -0.5, but are appropriately set so that the tendency of the calculation result can be determined as accurately as possible. 7C shows the results of classifying the calculation results using the thresholds THA and THB. In the following description, the results of classifying the calculation results using the thresholds THA and THB as shown in FIG. 7C will also be referred to as trend prediction, as appropriate.

[0041] The sleep state determination unit 22C performs a calculation, for example in real time, to subtract the long-term moving average from the short-term moving average. In this example, the long-term moving average is a 30-minute moving average, so the calculation result is obtained 30 minutes after the sensor unit 21 starts measurement.

[0042] The sleep state determination unit 22C classifies the calculation results into three categories according to the thresholds THA and THB. For example, if the calculation result is greater than the threshold THA, it is determined that the respiratory rate is tending to increase. In this case, the sleep state determination unit 22C sets the logical value "1." For example, if the calculation result is greater than the threshold THB but less than the threshold THA, it is determined that the respiratory rate is tending to remain unchanged (the sleep state is not changing). In this case, the sleep state determination unit 22C sets the logical value "0." For example, if the calculation result is less than the threshold THB, it is determined that the respiratory rate is tending to decrease. In this case, the sleep state determination unit 22C sets the logical value "-1." The graph showing the changes in the logical values ​​set by the sleep state determination unit 22C is shown in FIG. 7C.

[0043] Then, if the respiratory rate trend is on the rise (becomes "1"), the sleep state determination unit 22C determines that the user U's sleep state is a "REM sleep state." If, after being determined to be a "REM sleep state," the respiratory rate trend is on the rise or remains unchanged (becomes "0"), the sleep state determination unit 22C determines that the "REM sleep state" is continuing. On the other hand, if the respiratory rate trend is on the decline (becomes "-1"), the sleep state determination unit 22C determines that the user U's sleep state is a "non-REM sleep state." If, after being determined to be a "non-REM sleep state," the respiratory rate trend is on the decline or remains unchanged (becomes "0"), the sleep state determination unit 22C determines that the "non-REM sleep state" is continuing.

[0044] Fig. 8A shows a trend forecast resulting from the sleep state determination unit 22C using the method described above. Fig. 8B shows temporal changes in the sleep state of the same user U determined based on measurement results using an electroencephalograph, and is intended as a comparison for verifying the accuracy of the sleep state determination according to the present invention. Note that in the examples of Figs. 8A and 8B, the sleep state upon transition from a wakefulness state to a sleep state is illustrated as a "non-REM sleep state."

[0045] For example, suppose that at timing ta (around 3:10), the calculation result is classified as "1." In this case, as described above, the sleep state determination unit 22C determines that the sleep state of the user U is a "REM sleep state." The measurement result using the electroencephalograph also determines that the sleep state of the user U is a "REM sleep state." At subsequent timing tb (around 3:40), the calculation result is classified as "-1." The sleep state determination unit 22C determines that the sleep state of the user U has transitioned from a "REM sleep state" to a "non-REM sleep state." After timing tb, since the calculation result is classified as "0," the sleep state determination unit 22C determines that the "non-REM sleep state" is continuing. The actual measurement result in FIG. 8B also shows that non-REM sleep is continuing. At subsequent timing tc (around 5:20), the calculation result is classified as "1." The sleep state determination unit 22C determines that the sleep state of the user U has transitioned from a "non-REM sleep state" to a "REM sleep state." The actual measurement results in FIG. 8B also show that the sleep state transitioned from non-REM sleep to REM sleep around 5:20. Thus, the results of sleep state determination using the method according to this embodiment and the results of sleep state determination using an actual electroencephalograph correspond substantially to each other. This means that the method according to this embodiment can determine the sleep state with high accuracy.

[0046] [Processing flow] The flow of processing performed by the sleeping state judging device 1 according to the first embodiment will be described with reference to the flowchart of FIG.

[0047] In step ST11, the first moving average calculation unit 22A calculates a moving average of the respiratory rate for the past 10 minutes from the current timing based on the respiratory rate output from the sensor unit 21. The calculated moving average corresponds to the short-term moving average. Then, the process proceeds to step ST12.

[0048] In step ST12, the second moving average calculation unit 22B calculates a moving average of the respiratory rate for the past 30 minutes from the current timing based on the respiratory rate output from the sensor unit 21. The calculated moving average corresponds to the long-term moving average. Then, the process proceeds to step ST13.

[0049] In step ST13, the sleep state determination section 22C calculates (short-term moving average - long-term moving average), and the process then proceeds to step ST14.

[0050] In step ST14, the sleep state determination unit 22C determines the sleep state according to the calculation result. Specifically, if the calculation result shows an upward trend, the sleep state determination unit 22C determines that the sleep state of the user U is a REM sleep state. If the calculation result shows a downward trend, the sleep state determination unit 22C determines that the sleep state is the non-REM sleep state. If the calculation result shows no change, the sleep state determination unit 22C determines that the most recently determined REM sleep state or non-REM sleep state is continuing. The sleep state determination unit 22C makes the above determination at each predetermined determination timing.

[0051] As described above, according to this embodiment, the sleep state of the user U is determined using trend prediction that predicts future trends from a moving average of the respiration rate, so it is possible to accurately determine the sleep state of the user U. Even if there is an error in part of the measurement result of the respiration rate, this is smoothed by the moving average, so it is possible to accurately determine the sleep state of the user U.

[0052] FIG. 10A shows an example of a sleep state of a user U determined using the method according to this embodiment. FIG. 10B shows an example of the sleep state of the same user U determined based on the electroencephalogram obtained by an electroencephalogram attached to the same user U, intended as a comparison for verifying the accuracy of the sleep state determination according to this embodiment. Comparing the two, the degree of agreement between the determination of "REM sleep state" and "non-REM sleep state" is approximately 92.2%. This indicates that the accuracy of the sleep state determination method according to this embodiment is high, at over 90%.

[0053] <Second embodiment> Next, a second embodiment will be described. In the description of the second embodiment, the same or similar components as those in the above description will be denoted by the same reference numerals, and duplicated descriptions will be omitted as appropriate. Furthermore, unless otherwise specified, the matters described in the first embodiment can be applied to the second embodiment.

[0054] [Issues to be considered in this embodiment] First, to facilitate understanding of this embodiment, issues to be considered in this embodiment will be described. Fig. 11A shows an example of a trend prediction resulting from the determination by the sleep state determination unit 22C using the method described in the first embodiment. Fig. 11B shows temporal changes in the sleep state of the same user U determined based on measurement results using an electroencephalograph, and is intended as a comparison for verifying the accuracy of the sleep state determination according to the present invention. Note that in the examples of Figs. 11A and 11B, the sleep state upon transition from a wakefulness state to a sleep state is illustrated as a "non-REM sleep state."

[0055] As explained above, the method described in the first embodiment can determine the sleep state of the user U with high accuracy. However, even with this method, there are cases where the sleep state of the user U is erroneously determined. For example, as shown in FIG. 11A, assume that the calculation result at timing td (around 4:20) is classified as an upward trend ("1"). In this case, the sleep state determination unit 22C determines the sleep state of the user U as a "REM sleep state." However, as shown in FIG. 11B, the determination result based on the electroencephalogram measured by the electroencephalograph is a "non-REM sleep state," resulting in a discrepancy (inconsistency). The present embodiment is an embodiment that minimizes such discrepancies and enables the sleep state to be determined with higher accuracy (preciseness).

[0056] [Configuration example of sleep state determination device] 12 is a block diagram illustrating an example of the electrical configuration of a sleeping state judging device (sleeping state judging device 1A) according to the second embodiment. Note that the following description will focus on components that are different from the sleeping state judging device 1 described in the first embodiment.

[0057] The sleep state determination device 1A includes a respiratory rate variation calculation unit 22D. The respiratory rate variation calculation unit 22D is included in, for example, the sleep state determination unit 22. The respiratory rate variation calculation unit 22D calculates the variation in the respiratory rate of the user U over a predetermined period of time in the past (for example, one minute). The respiratory rate variation calculation unit 22D then calculates a value based on the total respiratory rate variation, for example, the total respiratory rate variation, which is the sum of the respiratory rate variation over a predetermined period of time from the current timing. The respiratory rate of the user U is supplied from the sensor unit 21. The sleep state determination unit 22 according to this embodiment can also be configured as an integrated part of the control unit 20 by implementing functions corresponding to the first moving average calculation unit 22A, the second moving average calculation unit 22B, the sleep state determination unit 22C, and the respiratory rate variation calculation unit 22D in the control unit 20.

[0058] An example of the operation of respiratory rate variation calculator 22D will be described with reference to Fig. 13A to Fig. 13C. As shown in Fig. 13A, respiratory rate variation calculator 22D calculates the respiratory rate starting from the respiratory rate at a predetermined timing (5:30 in the example shown in Fig. 13A). As an example, respiratory rate variation calculator 22D calculates, as the respiratory rate variation, the difference (absolute value) between the respiratory rate at the timing when there is a change in the respiratory rate and the respiratory rate at the timing when there is the next change in the respiratory rate (however, the change in the respiratory rate is different from the previous change). The change in the respiratory rate includes when the respiratory rate no longer changes (becomes approximately constant).

[0059] For example, the "1" written on the leftmost side of the example shown in Fig. 13A indicates that the difference between the respiratory rate at the timing when the respiratory rate decreased and the respiratory rate at the timing when the next respiratory rate change, i.e., the respiratory rate fluctuation amount, is "1." For example, the "1" written second from the left in the example shown in Fig. 13A indicates that the difference between the respiratory rate at the timing when the respiratory rate increased and the respiratory rate at the timing when the next respiratory rate change, i.e., the respiratory rate fluctuation amount, is "1." For example, the "7" written third from the left in the example shown in Fig. 13A indicates that the difference between the respiratory rate at the timing when the respiratory rate began to decrease and the respiratory rate at the timing when the next respiratory rate change, i.e., the respiratory rate fluctuation amount, is "7." The same applies to the other respiratory rate fluctuation amounts.

[0060] The respiratory rate variation calculation unit 22D calculates the total respiratory rate variation by, for example, adding up the calculated respiratory rate variations over the past minute. For example, the total respiratory rate variation over one minute shown in FIG. 13A is 36. FIG. 13B is a graph showing temporal changes in the total respiratory rate variation. The horizontal axis of FIG. 13B indicates an example of time, and the vertical axis indicates the total respiratory rate variation (bpm). The respiratory rate variation calculation unit 22D binarizes the calculated total respiratory rate variation based on a threshold value THC set for the total respiratory rate variation. The threshold value THC is set to 200 (bpm), for example, but is not limited to this value. For example, the respiratory rate variation calculation unit 22D assigns a logical value of "1" when the total respiratory rate variation is greater than the threshold value THC, and assigns a logical value of "0" when the total respiratory rate variation is less than the threshold value THC.

[0061] 13C is a graph showing the binarized result of the total amount of respiratory rate fluctuation. For example, the binarized result is supplied to the sleep state determination section 22C.

[0062] FIG. 14A is a graph showing the binarization results of the total respiratory rate fluctuation. FIG. 14B, intended as a comparison for verifying the accuracy of the sleep state determination according to the present invention, shows temporal changes in the sleep state of the same user U, determined based on measurements using an electroencephalograph. Comparing FIG. 14A and FIG. 14B, at least when the binarization result is "0," the actual sleep state is a "non-REM sleep state," as shown in FIG. 14B. That is, when the binarization result is "0," the sleep state is likely to be a "non-REM sleep state." It is known that in a "non-REM sleep state," breathing generally becomes deeper and the respiratory rate stabilizes. The binarization result of "0" means that when the total respiratory rate fluctuation is smaller than the threshold value THC, i.e., the respiratory rate is in a state of being relatively stable. This is consistent with the estimation of a "non-REM sleep state" when the binarization result is "0." In this embodiment, this point is combined with the trend prediction described above to improve the accuracy of sleep state determination.

[0063] The sleep state determination unit 22C combines the classification result of the total amount of respiratory rate fluctuation (for example, the binarization result) with the classification result of the calculation result to determine the sleep state of the user U. Specifically, if the total amount of respiratory rate fluctuation is smaller than a predetermined threshold, the sleep state of the user U is determined to be a non-REM sleep state, and if the total amount of respiratory rate fluctuation is larger than the predetermined threshold, the sleep state of the user U is determined in the same manner as in the first embodiment, that is, based on the calculation result of the difference between the short-term moving average and the long-term moving average.

[0064] FIG. 15A shows three graphs. In each graph, the horizontal axis represents the time axis. The top graph is a binarized graph of the total respiratory rate fluctuation. The middle graph is the result of trend prediction. The bottom graph is a combination of the top and middle graphs according to the above rules. FIG. 15B shows the temporal change in the sleep state of the same user U, determined based on the measurement results using an electroencephalograph.

[0065] When the binarized result of the total respiratory rate fluctuation is "0", the sleep state of user U is determined to be a "non-REM sleep state". Looking at the actual measurement results shown in FIG. 15B, when the binarized result of the total respiratory rate fluctuation is "0", the sleep state of user U is determined to be a "non-REM sleep state". Also, for example, when the trend prediction becomes "1" as at timing td, the sleep state would normally be determined to be a "REM sleep state". However, because the binarized result of the total respiratory rate fluctuation is "0", this is given priority and the sleep state of user U is determined to be a "non-REM sleep state". This determination result is consistent with the actual measurement results of the sleep state shown in FIG. 15B.

[0066] If the binarized result of the total respiratory rate fluctuation is "1", the sleep state of user U is determined based on trend prediction. For example, at timing tb, the binarized result of the total respiratory rate fluctuation is "1", so the sleep state determination result based on trend prediction takes precedence. Since the trend prediction is "-1", the sleep state determination unit 22C determines that user U's sleep state has become a "non-REM sleep state" from this timing. This determination result coincides with the actual measurement results of the sleep state shown in FIG. 15B.

[0067] [Processing flow] The flow of processing performed by the sleeping state judging device 1A according to the second embodiment will be described with reference to the flowchart of FIG.

[0068] In step ST21, the first moving average calculation unit 22A calculates a moving average of the respiratory rate for the past 10 minutes from the current timing based on the respiratory rate output from the sensor unit 21. The calculated moving average corresponds to the short-term moving average. Then, the process proceeds to step ST22.

[0069] In step ST22, the second moving average calculation unit 22B calculates a moving average of the respiratory rate for the past 30 minutes from the current timing based on the respiratory rate output from the sensor unit 21. The calculated moving average corresponds to the long-term moving average. Then, the process proceeds to step ST23.

[0070] In step ST23, the respiratory rate variation calculation unit 22D calculates the respiratory rate variation and obtains the total respiratory rate variation, which is the sum of the respiratory rate variation over a predetermined period (e.g., one minute). The respiratory rate variation calculation unit 22D binarizes the total respiratory rate variation using the threshold value THC. The respiratory rate variation calculation unit 22D supplies the binarized result to the sleep state determination unit 22C. Then, the process proceeds to step ST24.

[0071] In step ST24, the sleep state determination section 22C calculates (short-term moving average - long-term moving average), and the process then proceeds to step ST25.

[0072] In step ST25, it is determined whether the total amount of respiratory rate fluctuation is smaller than the threshold value, in other words, whether the binarization result in step ST23 is "0." This determination is made, for example, by the sleep state determination unit 22C. If the determination result in step ST25 is Yes, the process proceeds to step ST26.

[0073] In step ST26, since the total amount of respiratory rate fluctuation is smaller than the threshold value, in other words, the binarization result in step ST23 is "0", the sleep state determination unit 22C determines that the sleep state of the user U is a "non-REM sleep state".

[0074] If the determination result in step ST25 is No, the process proceeds to step ST27. In step ST27, the sleep state determination unit 22C determines the sleep state of the user U based on the calculation result in step ST24. That is, the sleep state determination unit 22C determines the sleep state of the user U based on trend prediction, as described in the first embodiment. The above determination process is performed at each determination timing. Then, if the wake-up of the user U is detected by a known method, the process by the sleep state determination device 1A ends.

[0075] As described above, in this embodiment, when the total amount of respiratory rate fluctuation is smaller than the threshold, the sleep state of the user U is determined taking into consideration that the sleep state of the user U is substantially in a "non-REM sleep state." Therefore, the sleep state of the user U can be determined more accurately.

[0076] <Third embodiment> Next, a third embodiment will be described. In the description of the third embodiment, the same or similar components as those in the above description will be denoted by the same reference numerals, and duplicated descriptions will be omitted as appropriate. Furthermore, unless otherwise specified, the matters described in the first and second embodiments can be applied to the third embodiment.

[0077] 17 is a block diagram showing an example of the electrical configuration of a sleeping state judging device (sleeping state judging device 1B) according to the third embodiment. Note that the following description will focus on components that are different from the sleeping state judging device 1 described in the first embodiment.

[0078] The sleep state determination unit 22 of the sleep state determination device 1B includes a sleep onset detection unit 22E. As shown in Fig. 18, the sleep onset detection unit 22E detects the timing (timing te) at which the user U transitions from a wakeful state (Wake) to a sleep state (e.g., a "non-REM sleep state") as the timing at which the user U falls asleep. Note that the sleep state determination unit 22 according to this embodiment can also be configured as an integrated part of the control unit 20 by implementing functions corresponding to the first moving average calculation unit 22A, the second moving average calculation unit 22B, the sleep state determination unit 22C, and the sleep onset detection unit 22E in the control unit 20.

[0079] An example of the operation of the sleep detection unit 22E will be described with reference to Fig. 19A to Fig. 19D. Fig. 19A shows an example of changes in the respiration rate over two minutes when the user U is awake (wake state). As shown in Fig. 19A, in the wake state, the respiration rate fluctuates relatively greatly.

[0080] FIG. 19B shows an example of changes in the respiration rate when user U is in a "non-REM sleep state." When the sleep state is a "non-REM sleep state," there are times when the respiration rate remains substantially unchanged (stable) for a certain period of time, as shown by the area surrounded by a dotted line in FIG. 19B. For example, there are times when the respiration rate remains substantially unchanged for 20 seconds or more. "Substantially unchanged respiration rate" means, for example, that the difference between the maximum and minimum respiration rate over a certain period of time is 0 or a certain value or less (for example, 1 or less).

[0081] The sleep detection unit 22E monitors the respiratory rate supplied from the sensor unit 21 and monitors whether a time has come when the respiratory rate has stabilized for, for example, 20 seconds or more. When the sleep detection unit 22E detects that the respiratory rate has stabilized for 20 seconds or more, it sets a detection flag, for example, as shown in FIG. 19C. The detection flag is, for example, a flag that changes from a logical value of "0" to "1." In the example shown in FIG. 19C, the detection flag is set around 23:35. As shown in FIG. 19D, the sleep detection unit 22E determines that the user U has fallen asleep at the time the detection flag is set.

[0082] According to this embodiment, it is possible to detect that the user U has fallen asleep. For example, after the sleep detection unit 22E detects that the user U has fallen asleep, the first moving average calculation unit 22A, the second moving average calculation unit 22B, and the sleep state determination unit 22C may be configured to operate. This reduces power consumption. Note that the sleep state determination device 1A described in the second embodiment may be configured to include the sleep detection unit 22E.

[0083] <Application example> The present invention can be applied in a variety of ways. Fig. 20 shows an example of the configuration of a control system (control system 100A) that is one of the application examples. The control system 100A includes a sleeping state determination device 1 and an electric device 30. The sleeping state determination device 1 may be the sleeping state determination device 1A or the sleeping state determination device 1B.

[0084] The electrical appliance 30 has an electrical appliance communication unit 31 and an electrical appliance control unit 32. The electrical appliance communication unit 31 acquires the sleeping state of the user U determined by the sleep state determination device 1 through communication. The communication may be wired communication or wireless communication. The electrical appliance communication unit 31 supplies the sleeping state of the user U acquired through communication to the electrical appliance control unit 32. The electrical appliance control unit 32 performs control according to the sleep state determination result performed by the sleep state determination unit 22C.

[0085] Specific examples of the electrical appliance 30 include lighting devices, electric blinds, air conditioners, television devices, smart speakers, electric reclining beds, medical equipment, etc. When the electrical appliance 30 is an air conditioner, for example, when the sleep state of the user U transitions from a "REM sleep state" to a "non-REM sleep state," the electrical appliance control unit 32 performs control to, for example, weaken the strength of the air conditioning. Note that the control performed by the electrical appliance control unit 32 can be changed as appropriate depending on the type of electrical appliance 30.

[0086] FIG. 21 shows a configuration example of a control system (control system 100B) according to another application example. The control system 100B includes a sleep state determination device 1 and an electrical appliance 30. The sleep state determination device 1 may be the sleep state determination device 1A or the sleep state determination device 1B. As described above, the sleep state determination device 1 includes an infrared LED 17 serving as an electrical appliance control signal transmitter for transmitting, to the electrical appliance 30, an infrared signal identical to an infrared signal transmitted from an infrared remote control compatible with each electrical appliance 30. The control unit 20 stores information related to the infrared remote control signal transmitted from the infrared remote control compatible with each electrical appliance 30 in advance. Based on the sleep state determination result by the sleep state determination unit 22, the control unit 20 determines the necessary control content for the electrical appliance 30 and transmits an infrared remote control signal corresponding to the control content from the infrared LED 17 to the electrical appliance 30, thereby controlling the electrical appliance 30. The infrared remote control signal transmitted from the infrared LED 17 is received by the electrical appliance communication unit 31. The infrared remote control signal received by the electrical appliance communication unit 31 is supplied to the electrical appliance control unit 32. The electrical appliance control unit 32 performs control according to the content of the infrared remote control signal. If the electrical appliance 30 is an air conditioner, for example, when the sleep state of the user U transitions from a "REM sleep state" to a "non-REM sleep state," the electrical appliance control signal transmission unit transmits an infrared remote control signal to control the air conditioning to be weaker, for example.

[0087] <Modification> Although the embodiments of the present invention have been specifically described above, the present invention is not limited to the above-described embodiments, and various modifications based on the technical concept of the present invention are possible. Modifications will be described below.

[0088] Some of the functions of each of the above-described sleep state judging devices 1, 1A, and 1B may be performed by other devices. For example, as shown in FIG. 22, the sleep state judging device 1 may have a communication unit 25. The sleep state judging unit 22 may be configured to be included in a server 40. The server 40 further has a server communication unit 41. The sleep state judging device 1 and the server 40 can send and receive data via a communication line such as the Internet.

[0089] The respiration rate obtained by the sensor unit 21 is transmitted to the server 40 via the communication unit 25. Data indicating the transmitted respiration rate is supplied from the server communication unit 41 to the sleep state determination unit 22. The sleep state determination unit 22 uses the respiration rate supplied from the server communication unit 41 to perform the processing described in the embodiment.

[0090] The server 40 may be a device such as a smartphone that can be carried by the user U. Some of the functions of the sleep state determination device 1 and the like may be executed by the smartphone.

[0091] Other modified examples will be described. In the above-described embodiment, the sleep states of the user U are the "REM sleep state" and the "non-REM sleep state," but other sleep states may also be used. For example, the "non-REM sleep state" may be further classified according to the depth of the sleep state.

[0092] In the above-described embodiment, an example of determining the sleep state of one user U has been described, but the sleep states of multiple (e.g., two) users U may also be determined. For example, the direction of radio wave irradiation can be changed by operating the rotation unit 14, so that radio waves can be appropriately irradiated to each user U. Then, the sleep state of each user U may be determined by performing the processing described in the embodiment.

[0093] The user U may be an adult or a child. The present invention can also be applied to animals other than humans, such as pets.

[0094] The present invention can be realized not only as an apparatus, but also as a method, a program, etc. The present invention can also be realized as a learning model that executes a computer to input a respiratory rate and output a sleep state.

[0095] The configurations, methods, processes, shapes, materials, and numerical values ​​of the above-described embodiments can be combined or substituted without departing from the spirit of the present invention. Furthermore, one thing can be divided into two or more things, and two or more things can be combined into one. Furthermore, some parts can be omitted. [Explanation of symbols]

[0096] 1, 1A, 1B... Sleep state determination device 21 Sensor unit 21A···Sensor 22 Sleep state determination unit 22A...First moving average calculation section 22B...Second moving average calculation section 22C Sleep state determination unit 22D Respiratory rate variation calculation section 22E···Sleep detection unit 30 Electrical Equipment 31 Electrical Equipment and Communications Department 32 Electrical equipment control section 100 Sleep state determination system 100A, 100B... Control System

Claims

1. The sensor and a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit that calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result that uses the first moving average and the second moving average. Sleep state determination device.

2. the sleep state determination unit classifies the calculation result into one of an upward trend, a downward trend, and a no-change trend according to a predetermined threshold; When the calculation result shows the upward trend, the sleep state is determined to be the REM sleep state; when the calculation result shows the downward trend, the sleep state is determined to be the non-REM sleep state; and when the calculation result shows the no change trend, the sleep state is determined to be the REM sleep state or the non-REM sleep state determined immediately before is continuing. The sleep state determination device according to claim 1 .

3. a respiratory rate fluctuation calculation unit that calculates a total respiratory rate fluctuation, which is the sum of respiratory rate fluctuations for a predetermined period from the present timing, the sleep state determination unit determines the sleep state by combining a classification result obtained by classifying the total amount of respiratory rate fluctuation according to a predetermined threshold value and a classification result of the calculation result. The sleep state determination device according to claim 2 .

4. the sleep state determination unit determines the sleep state to be a non-REM sleep state when the total amount of respiratory rate fluctuation is smaller than a predetermined threshold, and determines the sleep state based on the calculation result when the total amount of respiratory rate fluctuation is larger than the predetermined threshold. The sleep state determination device according to claim 3 .

5. the calculation result is a result of subtracting the second moving average from the first moving average.

5. The sleep state determination device according to claim 1.

6. the first moving average calculation unit calculates the first moving average using an exponential smoothing moving average method as a predetermined moving average method; the second moving average calculation unit calculates the second moving average using an exponential smoothing moving average method; 5. The sleep state determination device according to claim 1.

7. A sleep detection unit that detects that the user has fallen asleep using the respiratory rate.

5. The sleep state determination device according to claim 1.

8. The sensor is a non-contact sensor that is not attached to the user.

5. The sleep state determination device according to claim 1.

9. The sensor is configured as a radio wave sensor, and the radiation direction of the radio waves emitted from the sensor is configured to be changeable.

5. The sleep state determination device according to claim 1.

10. a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result using the first moving average and the second moving average; Sleep state determination method.

11. a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result using the first moving average and the second moving average; A program that causes a computer to execute a sleep state determination method.

12. A sleep state determination device and an electrical device are included, The sleep state determination device includes: The sensor and a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit that calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result that uses the first moving average and the second moving average; and The electrical equipment includes: an electrical appliance control unit that performs control in accordance with the determination result of the sleep state determination unit; Control system.

13. A sleep state determination device and an electrical device are included, The sleep state determination device includes: The sensor and a first moving average calculation unit that calculates a first moving average for a first period using a predetermined moving average method based on the respiratory rate obtained from the sensing data acquired by the sensor; a second moving average calculation unit that calculates a second moving average for a second period longer than the first period based on the respiratory rate; a sleep state determination unit that determines a sleep state including a REM sleep state and a non-REM sleep state using a calculation result that uses the first moving average and the second moving average; an electrical appliance control signal transmission unit that transmits a control signal for controlling the electrical appliance in accordance with a determination result of the sleep state determination unit, Control system.

Images