TWM666728U - Alarm clock system incorporating artificial intelligence and alarm clock device carrying it - Google Patents

Abstract

The present invention provides an alarm system combined with artificial intelligence, which includes an artificial intelligence module configured to output a sleep efficiency judgment result; and an alarm effect module configured to activate an alarm effect according to the sleep efficiency judgment result; the present invention uses the artificial intelligence module to judge the sleep quality to instruct the alarm effect module to switch the sensory stimulation mode of the alarm device, and makes a comprehensive judgment on the sleep quality of the sleeper based on multiple factors to decide whether to use gradual or more intense means for sensory stimulation, thereby avoiding the sleeper from experiencing discomfort due to being forced to wake up, and improving the quality of life of the sleeper after waking up.

Description

Alarm system combined with artificial intelligence and alarm device equipped with it

This creation involves a sleep quality judgment system, which specifically combines artificial intelligence computing and is applied to alarm technology; the alarm technology involved in this creation also combines multiple sensory stimulations, including auditory, olfactory and visual sensory stimulations.

Usually, before falling asleep, people will set an alarm clock to wake the sleeper from deep sleep at a specific time. However, when the sleeper is in deep sleep, the sound stimulation may be insufficient because the auditory sensitivity has not yet recovered, or because the light in the sleeping place is dim, the sleeper is still in a semi-awake state at a specific time, and there is a situation of staying in bed.

Therefore, the alarm clock that relies on sound stimulation mode at present cannot fully play its role; in a deep sleep or semi-awake state, in addition to the continuous stimulation of hearing, the vision and smell can also be stimulated by sufficient intensity to promote the sleeper to gradually enter the awake state.

Based on this, this creation is based on combining artificial intelligence to judge sleep quality and combining multiple sensory stimulation modes. On the one hand, it can effectively promote the sleeper's awakening, and on the other hand, it can reduce the discomfort of the sleeper during the awakening process by switching modes, and take care of the sleeper's mood when waking up.

The invention provides an alarm system combined with artificial intelligence, which includes: an artificial intelligence module, which includes: a sleep quality information processing unit, which is configured to read a sleep physiological state information and calculate a sleep quality information based on the sleep physiological state information; and a sleep efficiency judgment unit, which is configured to read the sleep quality information and output a sleep efficiency judgment result based on the sleep quality information, wherein the sleep quality information includes a sleeping time information, a wake-up frequency information, a wake-up time information, a snoring time information, a wake-up frequency information or a combination thereof; and an alarm effect module, which is connected to the artificial intelligence module signal, and is configured to activate an alarm effect at a wake-up time or a pre-wake-up time according to the sleep efficiency judgment result, wherein: the pre-wake-up time is prior to the wake-up time.

As described above, the sleep efficiency determination unit further calculates a sleep quality score based on the sleep time information, the awakening times information, the awakening time information, the snoring time information, and the waking times information, and determines whether the sleep efficiency of the corresponding sleeper is good based on whether the sleep quality score is greater than a sleep efficiency score, wherein the sleep quality score is calculated based on the following formula (I):

Among them, a is an arbitrary value between 1 and 100, b is an arbitrary value between 0.01 and 0.1, c is an arbitrary value between 0.03 and 0.2 and greater than b, the time of falling asleep is the absolute value of the hour, the number of times of waking up is the absolute value of the number of times, the number of times of waking up is the absolute value of the number of times, the absolute value of the awakening time is the absolute value of the number of minutes, and the absolute value of the snoring time is the absolute value of the minute.

As in the alarm system described above, the sleep efficiency determination unit further calculates the sleep quality score based on a weighted deep sleep time, which can be calculated according to the following formula (II):

Among them, ΔN3 is the total weight value, which is the absolute value of the deep sleep time. The deep sleep time is calculated by the sleep quality information processing unit based on the sleep time information and the wake time information.

The alarm system as described above further includes a sleep sensing module, which is connected to the artificial intelligence module signal, and is configured to detect the state of the sleeper to output the sleep physiological state information, and the sleep physiological state information includes a breathing frequency, a heart rate, a snoring frequency or a combination thereof.

The alarm system as described above further includes at least one or more of the following: a speaker unit, connected to the alarm effect module signal, configured to respond to the sleep efficiency judgment result to activate a sound effect, wherein the speaker unit is connected to a speaker signal, and the speaker is configured to respond to the sound effect to emit an alarm sound with a first loudness; an aroma unit, connected to the alarm effect module signal, configured to respond to the sleep efficiency judgment result to emit an alarm sound with a first loudness; Activate a scent effect, wherein the scent unit is connected to a fragrance device signal, and the fragrance device is configured to respond to the scent effect to release a fragrance, which has a first release concentration; and a lighting unit is connected to the alarm effect module signal, and is configured to respond to the sleep efficiency judgment result to activate a lighting effect, wherein the lighting unit is connected to a light-emitting body signal, and the light-emitting body is configured to respond to the lighting effect to activate a flashing light, which has a first illumination.

As described above, the alarm system, wherein: when the sleep efficiency judgment result is good, the speaker, the fragrance device, and the light-emitting body synchronously respond to the alarm effect at the wake-up time to start; when the sleep efficiency judgment result is poor, the speaker, the fragrance device, and the light-emitting body synchronously respond to the alarm effect at the pre-wake-up time to start, wherein, between the pre-wake-up time and the wake-up time, the loudness of the alarm sound is gradually amplified to the first loudness, the release concentration of the fragrance is the first release concentration or increases from an initial release concentration gradient to the first release concentration, and the illumination of the flashing light is gradually enhanced to the first illumination.

As for the alarm system described above, when the first lying-in time is reached after the waking time, if the alarm effect module has not been turned off, the alarm effect module further activates a strong alarm effect, which includes a strong sound effect, a strong light effect, a strong smell effect or any combination thereof, wherein the loudness of the alarm sound is amplified from the first loudness to a second loudness, the release concentration of the fragrance is increased from the first release concentration to a second release concentration, and the illumination of the flashing light is increased from the first illumination to a second illumination.

As for the alarm system described above, after the waking time, if the alarm effect continues to be activated for more than the first over-bed time, the alarm effect module automatically turns off the alarm effect at a second over-bed time.

Another aspect of this invention is to provide an alarm device combined with artificial intelligence, which includes: a body, which defines a storage space inside; and an alarm system as described above, which is arranged in the storage space.

As described above, the alarm device, wherein the sleep sensing module is disposed on a wearable device or a mobile device.

This creation uses an artificial intelligence module to judge the sleep quality, and switches the sensory stimulation mode of the alarm device according to the sleep quality judgment result. The artificial intelligence makes a comprehensive judgment based on the combination of multiple factors, including the time of falling asleep, the number of awakenings, the awakening time, the snoring time, the number of waking up times, etc., which can more accurately test the sleep quality of the sleeper, and has the function of continuously recording the sleep quality score to dynamically update the sleep efficiency score; based on the sleep efficiency score to judge the quality of sleep, the system can judge whether to use gradual or more intense means of sensory stimulation, thereby reducing the discomfort experienced by the sleeper in the process of entering the awakening stage from the deep sleep stage, and improving the quality of life of the sleeper after waking up.

100: Alarm system

1: Artificial intelligence module

11: Sleep quality information processing unit

110: Sleep Analyzer

111:Snoring Analyzer

112: Wake-up analyzer

12: Sleep efficiency judgment unit

2: Alarm effect module

21: Speaker unit

210: Speaker

22: Odor unit

220: Fragrance device

23: Lighting unit

230: Luminous body

3: Sleep sensor module

31: Audio receiver

32: Vibration monitoring device

4: Wearable device

5: Mobile devices

200: Alarm device combined with artificial intelligence

20: Body

S: Storage space

201: Speaker hole

202: Odor release hole

203: light-transmitting part

FIG1A is a system block diagram for illustrating the configuration relationship of the alarm system components of the first embodiment; FIG1B is a waveform diagram for exemplifying the relationship between each stage of the sleep cycle and time; FIG1C is a block diagram for illustrating the configuration of the components of the sleep quality information processing unit; FIG2A to FIG2C are a series of system block diagrams for respectively illustrating the configuration relationship of the alarm system components of the second embodiment to the fourth embodiment; FIG3 is a system block diagram for illustrating the configuration relationship of the alarm system components of the fifth embodiment; FIG4A is a system block diagram for illustrating the configuration relationship of the alarm device components of the sixth embodiment; and FIG4B is an appearance schematic diagram for exemplifying the configuration relationship of the speaker hole, the odor release hole and the light-transmitting part on the main body of the alarm device.

Please refer to FIG. 1A . The first implementation of the present invention is an alarm system (100) combined with artificial intelligence, which includes an artificial intelligence module (1) configured to output a sleep efficiency judgment result; and an alarm effect module (2) configured to activate an alarm effect according to the sleep efficiency judgment result.

In a plurality of embodiments, the artificial intelligence module (1) comprises a sleep quality information processing unit (11) and a sleep efficiency determination unit (12), wherein the sleep quality information processing unit (11) is configured to read a sleep physiological state information and calculate a sleep quality information based on the sleep physiological state information, and the sleep efficiency determination unit (12) is configured to read the sleep quality information and output a sleep efficiency determination result based on the sleep quality information, wherein the sleep quality information comprises a sleep time point information, a wake-up number information, a wake-up time information, a snoring time information, a wake-up number information or a combination thereof.

In these embodiments, the alarm effect module (2) is signal-connected to the artificial intelligence module (1) and is configured to activate an alarm effect at a wake-up time or a pre-wake-up time according to the sleep efficiency judgment result, wherein: the pre-wake-up time is prior to the wake-up time, if the sleep efficiency judgment result is good, the alarm effect is activated at the wake-up time, if the sleep efficiency judgment result is poor, the alarm effect is activated at the pre-wake-up time.

It can be understood that the sleep efficiency judgment results output by the artificial intelligence module (1) can be specifically divided into multiple levels, each of which can correspond to the sleep stage of the sleeper, and the alarm effect module (2) makes decisions based on the corresponding sleep stage to output an appropriate alarm effect.

Specifically, a complete sleep cycle of humans consists of two cycles: non-rapid eye movement (NREM) and sleep and rapid eye movement (REM) sleep. A complete sleep cycle is an average of 90 minutes. Furthermore, the sleep cycle can be divided into four different stages: falling asleep (N1), light sleep (N2), deep sleep (N3), and rapid eye movement (R).

Please refer to Figure 1B, which specifically shows the progress of different stages of the sleep cycle; first, in the non-rapid eye movement period, from the falling asleep period (N1) gradually enters the light sleep period (N2), and then from the light to the deep sleep period (N3); then, generally speaking, it does not directly enter the rapid eye movement period from the deep sleep and deep sleep period (N3), but from the deep sleep and deep sleep period (N3) back to the light sleep period (N2), and after one or more cycles, it enters the rapid eye movement period; finally, with the end of the rapid eye movement period, it returns to the falling asleep period (N1) again.

Accordingly, in multiple embodiments, the artificial intelligence module (1) is further connected to a sleep sensing module (3) signal to capture the sleep physiological state information of the sleeper, including breathing frequency, heart rate, snoring frequency or any combination thereof.

In some embodiments, the sleep sensing module (3) can collect the breathing sound of the sleeper through an audio receiver (31) to generate a breathing signal or a muscle sound signal, or collect the pulse or heart rate of the sleeper through a vibration monitoring device (32) to generate a pulse signal or a heart rate signal, and the artificial intelligence module (1) records and judges it as a basis for calculating the subsequent sleep quality information; in one example, the sleep sensing module (3) records its breathing signal, pulse signal or heart rate signal on a first virtual time axis, and calculates its breathing rate, pulse rate or heart rate accordingly.

In these embodiments, the sleep quality information processing unit (11) further includes a memory device (not shown) for storing the breathing frequency, pulse frequency, heart rate or snoring frequency corresponding to different sleep stages, so as to serve as a basis for determining the sleep stage of the sleeper, and to calculate the time of falling asleep, number of awakening times, awakening time, snoring time and number of waking up times.

In these embodiments, please refer to FIG. 1C , the sleep quality information processing unit (11) further includes a sleep analyzer (110) for calculating the time of falling asleep, the number of awakenings and the awakening time; for example, when the sleeper enters the light sleep stage (N2) from the sleep stage (N1), the breathing rate and the pulse rate will gradually decrease. Here, the sleep analyzer (110) can preset a first sleep threshold value as a basis for judging whether the sleeper has entered the light sleep stage (N2). When the breathing rate or the pulse rate increases, the sleep analyzer (110) can calculate the time of falling asleep, the number of awakenings and the awakening time. When the frequency is lower than the first sleep threshold, the sleep analyzer (110) records a first sleep time on a second virtual time axis. A second sleep threshold can be preset as a basis for judging whether the sleeper has entered a deep sleep period (N3). When the breathing frequency or the pulse frequency is lower than the second sleep threshold, the sleep analyzer (110) records a second sleep time on the second virtual time axis. The first sleep time and/or the second sleep time are further output as sleep time information.

On the contrary, when the sleeping person's breathing frequency or pulse frequency is higher than the first sleep threshold on the second virtual time axis after the first sleep-onset time or the second sleep-onset time, the sleep analyzer (110) records a wake-up time on the virtual time axis, and simultaneously records the wake-up time and outputs it as wake-up time information. If the wake-up time repeatedly appears within the night sleep time range, the sleep quality information processing unit records and outputs the wake-up frequency information. And synchronously output the awake time information; it can be understood that the awake time information is based on the time between the awake time and the next sleep time, in this example, it is the time between the awake time and a third sleep time, wherein the third sleep time is when the breathing frequency or the pulse frequency is lower than the first sleep threshold, the sleeper enters the light sleep period (N2), and the sleep quality information processing unit records the sleep time on the second virtual time axis.

In contrast, in other embodiments, the sleep quality information processing unit (11) is equipped with a snoring analyzer (111), which performs audio analysis based on the frequency and waveform of the respiratory signal and muscle sound signal transmitted by the audio receiver (31) to determine whether snoring occurs and generate a snoring signal. The sleep quality information processing unit (11) starts to calculate the snoring time of the sleeper from reading the first snoring signal.

On the other hand, the sleep quality information processing unit (11) further includes an waking up analyzer (112) configured to determine whether the sleeper has made a waking up action based on a background sound signal and a background vibration signal to calculate the number of waking up times, wherein the background sound signal is obtained by the audio receiver (31) collecting and outputting the background sound in the sleeping environment, and the background vibration signal is obtained by the vibration monitoring device (32) collecting and outputting the background vibration in the sleeping environment. For example, the waking up analyzer (112) receives the background sound signal and the background vibration signal, and compares the sound characteristics to determine whether the sleeper has made a waking up action. For example, if the background sound signal meets the sound characteristics of the friction between the bed frame components and is accompanied by a background vibration signal with a larger amplitude than breathing or pulse, the waking up analyzer (112) records a late night waking up time on the second virtual time axis and outputs the waking up times information.

In these embodiments, the sleep efficiency determination unit (12) sets the wake-up time and the pre-wake-up time according to the alarm activation time set by the sleeper, and calculates and outputs the sleep efficiency determination result based on the sleep quality information at the pre-wake-up time; specifically, the sleep efficiency determination unit (12) calculates a sleep quality score according to the weight of each sleep quality information, and determines whether the sleep efficiency is good according to whether the sleep quality score is greater than a sleep efficiency score.

In some preferred embodiments, the sleep efficiency determination unit (12) calculates the sleep quality score according to the following formula (I):

It is understandable that the quality of sleep at night is significantly affected by the time of falling asleep. If the time of falling asleep is earlier, for example, falling asleep between 21:00 and 0:00 at night, it is better than falling asleep between 0:00 and 5:00 at night. The number of times of getting up involves whether the sleep cycle is interrupted. The more times, it can be inferred that the sleeper is less likely to return to the light sleep stage (N2) or deep sleep stage (N3). Therefore, in the above sleep quality calculation formula, the first weight score is calculated by the time of falling asleep and the number of times of getting up. The ratio is expressed as a ratio, and a weight value a is assigned, which can be any value between 1 and 100. Specifically, the absolute value of the hour is taken as the time of falling asleep, and the absolute value of the number of times of getting up is taken as the number of times. The following is an exemplary explanation. The sleep efficiency judgment unit (12) reads that the sleeper falls asleep at 21 o'clock at night, and the time of falling asleep is taken as 21, and reads that the sleeper gets up 2 times in total between the time of falling asleep and the time of getting up, and the number of times of getting up is taken as 2. Based on this, the calculation of the first weight score is as follows:

On the other hand, the number of awakenings and the awakening time are highly correlated and proportional. If the number of awakenings is higher and the awakening time is longer, it means that the sleeper is prone to sleep interruptions, resulting in poor sleep quality. Therefore, in the above sleep quality calculation formula, the second weight score is represented by the product of the number of awakenings and the awakening time, and is assigned a weight value b, which can be any value between 0.01 and 0.1; specifically, the number of awakenings is determined by whether the breathing rate or pulse rate is higher than the first sleep threshold to determine whether the person is awake, so it is not necessarily the same as the number of waking up, and the number of awakenings is the time from the first time of falling asleep to waking up. The number of times the sleeper enters the awake state between the first awake time and the third sleeping time is judged, and the specific value is the absolute value of the awake times; in addition, the awake time is the time between the first awake time and the third sleeping time, specifically calculated in minutes, and the absolute value of the number of minutes is taken; continuing with the above case, the sleeper falls asleep at 21:00 at night and sets 6:00 in the morning as the waking time. During this period, the sleep efficiency judgment unit (12) reads the number of awake times as 3 times, so the awake times are taken as 3, and the awake time each time is 15, 20 and 5 minutes, totaling 30 minutes, so the awake time is taken as 30; accordingly, the calculation of the second weighted score is as follows: b × number of awake times × awake time = b × 3 × 30

On the other hand, although the presence of snoring does not necessarily cause the sleeper to wake up directly or force the sleeper to get up, snoring is one of the important factors that disturb the sleeper from entering deep sleep or rapid eye movement period; however, compared with the number of awakenings and awake time, the impact of snoring on sleep quality is relatively low; therefore, in the above sleep quality calculation formula, the third weight score is expressed by snoring time and is assigned a weight value c, which can be any value greater than b between 0.03 and 0.2.

In a specific situation, since the frequency of snoring can be fixed or irregular, the snoring time is calculated as the sum of the time between the first time you fall asleep and the time you wake up, and the time between any snoring signal reading and the time you wake up closest to the snoring signal.

Continuing with the above case, the sleeper falls asleep at 21:00 at night and sets 6:00 in the morning as the time to wake up. During this sleep period, the sleep analyzer (110) determines that there are 3 awakening times, which are marked as the first awakening time, the second awakening time, and the third awakening time in chronological order. The snoring analyzer (111) outputs 3 snoring signals in total, among which the sleep quality information processing unit (11) records the first snoring signal to the first awakening time as the first snoring signal. During the period, 17 snoring signals are read, and the 18th snoring signal is recorded to the second awakening time as the second snoring time. During the period, 10 snoring signals are read, and the 29th snoring signal is recorded to the third awakening time as the third snoring time. Then, the sleep efficiency judgment module (12) reads the first snoring time, the second snoring time and the third snoring time and adds them up to obtain the snoring time, and the snoring time is specifically calculated in minutes, and the absolute value of the number of minutes is taken.

Taking the above example, the sleep efficiency determination module (12) reads the first snoring time as 12 minutes, the second snoring time as 17 minutes and the third snoring time as 3 minutes, and the total snoring time is 32 minutes. The snoring time is 32, and the third weighted score is calculated as follows: c × snoring time = c × 32

In the above example, the sleep quality score is calculated as follows:

In some specific embodiments, the sleep efficiency score can be calculated by total weighting of the deep sleep time. It is understandable that the longer the deep sleep time, the better the sleep quality of the sleeper. The deep sleep time is given a total weight value ΔN3 in the above formula (I), and the modified formula (I) is as follows:

The total weight value ΔN3 is calculated by adding up the time between the time of falling asleep and the time of waking up, any second time of falling asleep and the next first time of falling asleep or waking up, and taking the absolute value of the deep sleep time.

Continuing with the above case, the sleep analyzer (110) records a second sleep time before each wake-up time, and before the first wake-up time, records a cycle of the second sleep time-first sleep time and a cycle of the second sleep time-first wake-up time, which are 42 minutes and 45 minutes respectively. Between the first wake-up time and the second wake-up time, records a cycle of the second sleep time-first sleep time and The second sleep time and the second wake time of one cycle are 35 minutes and 43 minutes respectively, and between the second wake time and the third wake time, two cycles of the second sleep time and the first sleep time are recorded, which are 38 minutes and 33 minutes respectively. The total time of the six deep sleep stages is 236 minutes, which is converted to 3.93 hours. The total weight value ΔN3 is 3.93. In this example, the sleep quality score is calculated as follows:

In the above-mentioned embodiment, since the sleep quality varies from person to person, and different sleepers in different sleeping environments, the sleep information collected by the system (100) is inevitably different, so the sleep efficiency score is obtained by collecting the sleep quality scores of the sleeper on different days and performing an average or regression analysis calculation. There is no special restriction on the number of sleep quality scores collected, and at least 7 are collected for calculation; in some embodiments, the sleep efficiency score can be a normal sleep efficiency score established by collecting the sleep quality scores of different sleepers on different days and performing a regression analysis calculation, so as to provide the sleep efficiency judgment unit (12) with a comparison to judge whether the real-time sleep efficiency is good or bad.

The second embodiment of the invention is an alarm system (100) combined with artificial intelligence, and its components, connection relationships between components and actuation mechanisms are generally the same as those of the first embodiment, see FIG2A, but it further includes a speaker unit (21) connected to the alarm effect module (2) signal, configured to respond to the sleep efficiency judgment result to activate a sound effect; and a speaker (210) connected to the speaker unit (21) signal, configured to respond to the sound effect and emit an alarm sound, which has a first loudness.

In some embodiments, when the sleep efficiency judgment result is good, the speaker (210) is activated by ringing the sound effect at the time of waking up; when the sleep efficiency judgment result is poor, the speaker (210) is activated by ringing the sound effect at the pre-waking up time, wherein the loudness of the alarm sound is gradually amplified to the first loudness between the pre-waking up time and the waking up time, for example, 20 to 110 decibels, preferably 20 to 30 decibels, 30 to 40 decibels, 40 to 50 decibels, 50 to 60 decibels, 60 to 70 decibels, 70 to 80 decibels, 80 to 90 decibels, or 90 to 100 decibels.

The third embodiment of the invention is an alarm system (100) combined with artificial intelligence, and its components, connection relationships between components and actuation mechanisms are generally the same as those of the first embodiment, see FIG. 2B , but it further includes an aroma unit (22) connected to the alarm effect module (2) signal, configured to respond to the sleep efficiency judgment result to activate an aroma effect; and a fragrance device (220) connected to the aroma unit (22) signal, configured to respond to the aroma effect to release a fragrance having a first release concentration.

In some embodiments, when the sleep efficiency judgment result is good, the fragrance device (220) is activated in response to the scent effect at the time of waking up; when the sleep efficiency judgment result is poor, the fragrance device (220) is activated in response to the scent effect at the pre-waking up time, wherein the release concentration of the fragrance between the pre-waking up time and the waking up time is the first release concentration. or from an initial release concentration gradient to the first release concentration, wherein the ratio of the initial release concentration to the first release concentration is between 0.01 and 0.1; in some preferred embodiments, when the release concentration of the fragrance reaches the first release concentration, the odor unit (22) releases the odor effect so that the fragrance device (220) stops releasing the fragrance.

In many embodiments, it can be understood that in order to reduce the discomfort caused by the sleeper being forced to wake up due to the stimulation of the smell, the fragrance can be selected to have a pleasant smell, such as the smell of flowers or food; the flower aroma, such as peppermint, licorice root, rosemary, roselle, etc., can help soothe the sleeper's nerves; the food aroma, such as coffee, tea, milk, fruit, fried eggs, French fries, etc., can trigger The sleeper's hunger feeling is used to gradually move from the deep sleep stage (N3) to the light sleep stage (N2) to reduce the nervous tension caused by returning directly to the falling asleep stage (N1), so that the sleeper is naturally awakened; on the other hand, the so-called first release concentration refers to the concentration at which the presence of odor molecules can at least be detected by the human olfactory sense. For example, the olfactory threshold of coffee furanone is 0.005ppb, so only a very low concentration is required for the sleeper to detect it. The setting of the first release concentration will be different for different types of fragrances, and the initial release concentration will be adjusted accordingly, so there is no special restriction.

The fourth embodiment of the present invention is an alarm system (100) combined with artificial intelligence, and its components, connection relationships between components and actuation mechanisms are generally the same as those of the first embodiment, see FIG. 2C , but it further includes a lighting unit (23) connected to the alarm effect module (2) signal, configured to respond to the sleep efficiency judgment result to activate a lighting effect; and a light-emitting body (230), configured to respond to the lighting effect to activate a flashing light, which has a first illumination.

In this embodiment, the light-emitting body (230) can be any light source, such as a light-emitting diode (LED), a fluorescent lamp or an incandescent lamp, without any particular limitation.

In some embodiments, when the sleep efficiency judgment result is good, the light-emitting body (230) is activated by the lighting effect at the time of waking up; when the sleep efficiency judgment result is poor, the light-emitting body (230) is activated by the lighting effect at the pre-waking up time, wherein the illumination of the flashing light is gradually increased to the first illumination between the pre-waking up time and the waking up time, such as 100 to 300 lux or higher illumination, preferably 100 to 150 lux, 150 to 180 lux, 180 to 200 lux, 200 to 230 lux, 230 to 250 lux, 250 to 270 lux, or 270 to 300 lux.

In these embodiments, the flashing light is further responsive to the lighting effect by continuously flashing at a first flashing frequency, such as 1 to 10 times per second.

The fifth embodiment of the invention is an alarm system (100) combined with artificial intelligence. Its components, connection relationships between components and actuation mechanisms are generally the same as those of the first embodiment, see FIG3 , except that the alarm effect module (2) has a signal connection: a speaker unit (21) configured to respond to the sleep efficiency judgment result to activate a sound effect; an odor unit (22) configured to respond to the sleep efficiency judgment result to activate an odor effect; and a lighting unit (23) configured to respond to the sleep efficiency judgment result to activate a lighting effect.

In a plurality of embodiments, the speaker unit (21) is signal-connected to a speaker (210), which is configured to respond to the sound effect and emit an alarm sound having a first loudness; the scent unit (22) is signal-connected to a fragrance device (220), which is configured to respond to the scent effect to release a fragrance having a first release concentration; the lighting unit (23) is signal-connected to a light-emitting body (230), which is configured to respond to the lighting effect to activate a flashing light having a first illumination.

In this embodiment, the specific types, physical parameters and parameter selection of the alarm sound, the fragrance and the flashing light have been recorded in the second embodiment, the third embodiment and the fourth embodiment respectively, and will not be elaborated here.

In these embodiments, when the sleep efficiency judgment result is good, the speaker (210), the fragrance device (220), and the light-emitting body (230) synchronously respond to the alarm clock effect at the time of waking up to start; when the sleep efficiency judgment result is poor, the speaker (210), the fragrance device (220), and the light-emitting body (230) are activated at the time of waking up. The alarm effect is activated in synchronization with the waking up time, wherein, between the pre-waking up time and the waking up time, the loudness of the alarm sound is gradually amplified to the first loudness, the release concentration of the fragrance is the first release concentration or increases from an initial release concentration gradient to the first release concentration, and the illumination of the flashing light is gradually enhanced to the first illumination.

In some specific embodiments, after the first time of waking up, a first time of staying in bed, for example 15 to 30 minutes, is reached, if the alarm effect module (2) has not been turned off, the alarm effect module (2) further activates a strong alarm effect, which includes a strong sound effect, a strong light effect, a strong smell effect or any combination thereof, wherein the loudness of the alarm sound is increased from the first loudness to the second loudness. to a second loudness, such as 30 to 120 decibels and greater than the first loudness; the release concentration of the fragrance increases from the first release concentration to a second release concentration, wherein the second release concentration is at least 2 to 10 times the first release concentration; the illuminance of the flashing light is increased from the first illuminance to a second illuminance, such as 200 to 400 lux and greater than the first illuminance.

In a specific embodiment, when the alarm effect module (2) activates the strong alarm effect, the speaker unit (21) switches the alarm sound to a strong alarm sound, which has a louder volume, frequency or rhythm than the alarm sound, the scent unit (22) switches the fragrance to a stimulating atmosphere, which can be a smell or the smell of irritating food, such as mustard, chili, garlic, etc., and the lighting unit increases the flashing frequency of the flashing light to a second flashing frequency and continues to flash, wherein the second flashing frequency is 5 to 20 times per second and is higher than the first flashing frequency.

In the above embodiment, after the waking up time, if the alarm effect continues to be activated for more than the first bed-sleeping time, the alarm effect module (2) automatically turns off the alarm effect at a second bed-sleeping time, for example, 10 to 60 minutes after the first bed-sleeping time, preferably 10 to 20 minutes, 20 to 30 minutes, 30 to 40 minutes, 40 to 50 minutes or 50 to 60 minutes.

The sixth embodiment of the present invention provides an alarm device (200) combined with artificial intelligence, see Figures 4A and 4B, which includes: a body (20) defining a storage space (S) inside; and an alarm system (100) as provided in the fifth embodiment, wherein the artificial intelligence module (1), the alarm effect module (2), the speaker unit (21), the speaker (210), the scent unit (22), the fragrance device (220), the lighting unit (23) and the light-emitting body (230) are arranged in the storage space (S), and the sleep sensing module (3) is arranged in a wearable device (4) or a mobile device (5).

In many embodiments, the wearable device (4) includes but is not limited to a smart watch or a sports bracelet, and the mobile device (5) includes but is not limited to a smart phone, a tablet or a laptop.

In a plurality of embodiments, the body (20) further comprises: a speaker hole (201) disposed on the body (20) to connect the accommodating space (S) with the external space of the body (20), wherein the speaker (210) is disposed in the accommodating space (S) corresponding to the speaker hole (201); and an odor release hole (202) disposed on the body (20) adjacent to the speaker hole (201) to connect the accommodating space (S) with the external space of the body (20). The external space, wherein the fragrance device (220) and the odor release hole (202) are arranged in the accommodating space (S) in correspondence; a light-transmitting portion (203) is arranged on the main body (20) opposite to the odor release hole (202) and adjacent to the speaker hole (201) to connect the accommodating space (S) with the external space of the main body (20), wherein the light-emitting body (230) and the light-transmitting portion (203) are arranged in the accommodating space (S) in correspondence.

This invention provides an alarm system combined with artificial intelligence. The sleep sensing module reads the physical information such as sound and vibration of the sleeper and the environment. The sleep quality information processing unit records and judges the characteristics based on these physical information to generate sleep quality information. The sleep efficiency judgment unit judges the sleep quality. The system triggers the alarm effect module to produce a variety of sensory stimulation effects based on the sleep quality judgment results. Artificial intelligence collects a variety of factors that affect sleep and makes a comprehensive judgment. It can accurately judge the sleep quality of the sleeper and dynamically update the sleep efficiency score of the sleeper by continuous recording, providing customized sleep efficiency judgment results.

This creation also provides an alarm device equipped with the system, which can implement multiple sensory stimulations including alarms, fragrances, flashing lights, etc. to wake up the sleeper, and judge whether to use gradual or stronger stimulation methods based on the sleep efficiency score. In addition to creating an environment that guides the sleeper to gradually wake up, it can also reduce the discomfort experienced by the sleeper when entering the wakefulness period directly from the deep sleep period, and increase the comfort experience of the sleeper after waking up.

100: Alarm system

1: Artificial intelligence module

11: Sleep quality information processing unit

12: Sleep efficiency judgment unit

2: Alarm effect module

3: Sleep sensor module

31: Audio receiver

32: Vibration monitoring device

Claims (10)

An alarm system combined with artificial intelligence includes: an artificial intelligence module, which includes: a sleep quality information processing unit, which is configured to read a sleep physiological state information and calculate a sleep quality information based on the sleep physiological state information; and a sleep efficiency judgment unit, which is configured to read the sleep quality information and output a sleep efficiency judgment result based on the sleep quality information, wherein the sleep quality information includes a sleeping time information, a wake-up frequency information, a wake-up time information, a snoring time information, a wake-up frequency information or a combination thereof; and an alarm effect module, which is connected to the artificial intelligence module signal, and is configured to activate an alarm effect at a wake-up time or a pre-wake-up time according to the sleep efficiency judgment result, wherein: the pre-wake-up time is prior to the wake-up time. The alarm system as claimed in claim 1, wherein the sleep efficiency determination unit further calculates a sleep quality score based on the sleep time information, the awakening times information, the awakening time information, the snoring time information, and the waking times information, and determines whether the sleep efficiency of the corresponding sleeper is good based on whether the sleep quality score is greater than a sleep efficiency score, wherein the sleep quality score is calculated based on the following formula (I):

Among them, a is an arbitrary value between 1 and 100, b is an arbitrary value between 0.01 and 0.1, c is an arbitrary value between 0.03 and 0.2 and greater than b, the time of falling asleep is taken as the absolute value of the hour, the number of times of waking up is taken as the absolute value of the number of times, the number of times of waking up is taken as the absolute value of the number of times, the absolute value of the awake time is taken as the number of minutes, and the absolute value of the snoring time is taken as the minute. In the alarm system as claimed in claim 2, the sleep efficiency determination unit further calculates the sleep quality score based on a weighted deep sleep time, which can be calculated according to the following formula (II):

Among them, ΔN3 is the total weight value, which is the absolute value of the deep sleep time. The deep sleep time is calculated by the sleep quality information processing unit based on the sleeping time information and the waking time information. The alarm system as described in any one of claim 1 to 3 further includes a sleep sensing module connected to the artificial intelligence module signal, which is configured to detect the state of the sleeper to output the sleep physiological state information, and the sleep physiological state information includes a breathing frequency, a heart rate, a snoring frequency or a combination thereof. The alarm system as claimed in any one of claims 1 to 3 further comprises at least one or more of the following: a speaker unit connected to the alarm effect module signal, configured to respond to the sleep efficiency determination result to activate a sound effect, wherein the speaker unit is connected to a speaker signal, the speaker is configured to respond to the sound effect and emit an alarm sound having a first loudness; an aroma unit connected to the alarm effect module signal, configured to respond to the sleep efficiency determination result to activate a sound effect; The sleep efficiency judgment result is used to activate a scent effect, wherein the scent unit is connected to a fragrance device signal, and the fragrance device is configured to respond to the scent effect to release a fragrance with a first release concentration; and a lighting unit is connected to the alarm effect module signal, and is configured to respond to the sleep efficiency judgment result to activate a lighting effect, wherein the lighting unit is connected to a light-emitting body signal, and the light-emitting body is configured to respond to the lighting effect to activate a flashing light with a first illumination. The alarm system as described in claim 5, wherein: when the sleep efficiency judgment result is good, the speaker, the fragrance device, and the light-emitting body synchronously respond to the alarm effect at the wake-up time to start; When the sleep efficiency judgment result is poor, the speaker, the fragrance device, and the light-emitting body synchronously respond to the alarm effect at the pre-wake-up time to start, wherein, between the pre-wake-up time and the wake-up time, the loudness of the alarm sound is gradually amplified to the first loudness, the release concentration of the fragrance is the first release concentration or increases from an initial release concentration gradient to the first release concentration, and the illumination of the flashing light is gradually enhanced to the first illumination. The alarm system as described in claim 6, wherein, after the waking time reaches a first lying-in time, if the alarm effect module has not been turned off, the alarm effect module further activates a strong alarm effect, which includes a strong sound effect, a strong light effect, a strong smell effect or any combination thereof, wherein the loudness of the alarm sound is amplified from the first loudness to a second loudness, the release concentration of the fragrance is increased from the first release concentration to a second release concentration, and the illumination of the flashing light is increased from the first illumination to a second illumination. The alarm system as described in claim 7, wherein after the waking up time, if the alarm effect continues to be activated for more than the first bed-sleeping time, the alarm effect module automatically turns off the alarm effect at a second bed-sleeping time. An alarm device combined with artificial intelligence, comprising: a body, an interior of which defines a storage space; and an alarm system as described in any one of claims 1 to 3, disposed in the storage space. The alarm device as described in claim 9 further includes a sleep sensing module, which is signal-connected to the artificial intelligence module and is configured to detect the state of the sleeper to output the sleep physiological state information, wherein the sleep physiological state information includes a breathing frequency, a heart rate, a snoring frequency or a combination thereof, wherein the sleep sensing module is disposed on a wearable device or a mobile device.