WO2019037267A1 - Voice wake-up device and method - Google Patents

Abstract

A voice wake-up device, comprising a main processor unit (30), a voice control activating unit (10) and a microphone array unit (20), wherein the voice control activating unit (10) is used for collecting and processing a voice signal, generating a trigger signal when the voice signal meets a preset condition, and transmitting the trigger signal to the microphone array unit (20) so as to activate the microphone array unit (20); and the microphone array unit (20) is used for receiving and recognizing the voice signal, and outputting a wake-up signal when recognizing a key voice signal, the wake-up signal waking up the main processor unit (30). Only the voice control activating unit (10) works under a standby mode of the voice wake-up device, and the microphone array unit (20) only starts to work by means of the activation of the voice control activation unit (10), so that standby power consumption is greatly reduced, the standby current of the device, especially the standby current when the environment noise is not large, is greatly reduced compared with that of before, and a sufficient electric quantity may be provided even if a small battery is used. Therefore, the voice wake-up device may be applied to a mobile intelligent terminal, and the application range of the voice wake-up device is greatly increased.

Description

Voice wake-up device and method Technical field

The present invention relates to the technical field of a voice wake-up device, and in particular, to a voice wake-up device and a method thereof.

Background technique

At present, smart speakers, artificial intelligence devices, etc. all have voice wake-up and recognition functions. In order to achieve good far-field recognition effects, a multi-microphone array of two or more microphones is generally used, and a microphone or a digital signal processor is used to realize the sound source. Positioning, speech enhancement, and echo cancellation to improve recognition accuracy. However, if the system is always in the state of receiving signals in the wake-up of the keyword, the power consumption is generally large, and it is suitable for a device powered by a commercial power or a large-capacity battery, and is not suitable for a mobile smart device of a small battery, so that the application of the voice wake-up recognition is made. The scene is limited.

technical problem

The main object of the present invention is to provide a voice wake-up device and a method thereof, aiming at solving the technical problem that the standby power consumption of the voice wake-up device is too large.

Problem solution

Technical solution

The present invention provides a voice wake-up device, including a main processor unit, a voice activated unit, and a microphone array unit;

The voice activated unit is configured to collect and process a voice signal, and when the voice signal meets a preset condition, generate a trigger signal and transmit the signal to the microphone array unit to activate the microphone array unit to receive a voice signal;

The microphone array unit is configured to receive and recognize a voice signal, and when the key voice signal is recognized, output a wake-up signal, the wake-up signal is used to wake up the main processor unit; wherein, when the microphone array unit is activated The recognition time reaches a preset time, and the microphone array unit does not recognize the key voice signal, and the microphone array unit enters a non-operation state.

Further, the microphone array unit includes a single chip microcomputer and a microphone array;

The single chip is configured to receive a trigger signal of the voice activated unit, activate the microphone array, and enter a timing of a preset time, and after receiving the voice signal of the microphone array, determine whether the voice signal is a key voice signal And if yes, outputting a wake-up signal for waking up the main processor unit, and if not, outputting a sleep signal for sleeping the microphone array unit after a preset time;

The microphone array is configured to receive a voice signal and deliver the voice signal to the single chip microcomputer.

Further, the voice activated unit includes a first microphone, a first band pass filter amplifier, a first rectification filter, and a first comparator;

The first microphone is configured to pick up a sound and convert the voice into a voice signal;

The first band pass filter amplifier is configured to receive a voice signal output by the first microphone, and transmit a voice signal amplification process to the rectifier filter;

The first rectifying filter is configured to receive the amplified speech signal, and convert the speech signal processing into a DC signal, and send the signal to the first comparator;

The first comparator is configured to compare the DC signal, and when the DC signal exceeds a preset threshold value of the first comparator, the comparator outputs the trigger signal.

Further, a power switch is further included, and the power switch is respectively connected to the power unit, the voice activated unit, and the microphone array unit;

The voice activated unit outputs a trigger signal, the trigger signal controls the power switch to be turned on, and the power unit supplies power to the microphone array unit through the power switch to cause the microphone array unit to enter an identification working state;

The trigger signal controls the power switch to be turned off when the microphone array unit recognizes that the time reaches the preset time without recognizing a key voice signal.

Further, the power switch is further connected to the main processor unit, and the main processor unit controls the power switch to be turned off after receiving the sleep signal.

Further, the voice activated unit includes a second microphone, a second band pass filter amplifier, a second rectifier filter, a second comparator, and a monostable flip-flop;

The second microphone is configured to pick up a sound and convert the voice into a voice signal;

The second band pass filter amplifier is configured to receive a voice signal output by the second microphone, and transmit the voice signal amplification process to the second rectifier filter;

The second rectifying filter is configured to receive the amplified speech signal, and convert the speech signal processing into a DC signal, and send the signal to the second comparator;

The second comparator is configured to compare the DC signal, and when the DC signal exceeds a preset threshold value of the comparator, the second comparator outputs a first trigger signal;

The one-shot trigger is configured to receive the first trigger signal and convert the first trigger signal into a second trigger signal, where the second trigger signal controls the power switch.

The present invention also provides a voice wake-up method, which specifically includes the following steps:

Acquiring and processing the voice signal by the voice activated unit, and outputting the trigger signal by the voice activated unit when the voice signal meets the preset condition, the trigger signal activating the microphone array unit;

Receiving a voice signal through the microphone array unit, and outputting a wake-up signal through the microphone array unit when the voice signal is recognized as a key voice signal, the wake-up signal waking up the main processor unit; or calculating by the microphone array unit The length of the voice recognition time. If the length of the recognition time reaches a preset time, and the microphone array unit does not recognize the key voice signal, the microphone array unit is controlled to enter a non-working state.

Further, the specific steps of the voice activated unit to acquire and output the trigger signal include the following:

Receiving a sound through the first microphone and converting the voice into a voice signal, and transmitting the voice signal to the first band pass filter amplifier;

And the sound signal is amplified by the first band pass filter amplifier and then sent to the first rectifier filter;

Converting the amplified voice signal into a DC signal by the first rectifying filter and transmitting to the first comparator;

The DC signal is compared with a preset threshold by the first comparator, and the trigger signal is output by the first comparator when the DC signal exceeds a preset threshold.

Further, the specific steps of the microphone array unit to identify and output the wake-up signal include the following:

Receiving the trigger signal by the single chip microcomputer, and starting the microphone array and entering the timing of the preset time;

Receiving a voice signal through the microphone array, and transmitting the voice signal to the single chip microcomputer;

Receiving, by the single chip microcomputer, a voice signal output by the microphone array, and determining whether the voice signal is a key voice signal, and if the voice signal is a key voice signal, waking up through the single chip output And the wake-up signal wakes up the main processor unit, and if the voice signal is not a key voice signal, the sleep signal is outputted by the single-chip microcomputer after a preset time, and the sleep signal causes the microphone array unit to enter a sleep state.

Further, the method further includes: the voice activated unit outputs a trigger signal, the trigger signal controls the power switch to be turned on, and the power unit supplies power to the microphone array unit through the power switch to cause the microphone array unit to enter an identification working state; The microphone array unit recognizes that the time reaches the preset time without recognizing a key voice signal, and the trigger signal controls the power switch to be turned off.

Further, the method further includes: controlling, by the main processor unit, the power switch to be turned off after receiving the sleep signal.

Further, the voice wake-up method specifically includes the following steps:

Receiving a sound through the second microphone and converting the voice into a voice signal, and transmitting the voice signal to the second band pass filter amplifier;

And amplifying the voice signal by using the second band pass filter amplifier to be sent to the second rectifier filter;

Converting the amplified voice signal into a DC signal by the second rectifier filter and transmitting the signal to the second comparator;

Comparing the DC signal with a preset threshold by the second comparator, and outputting a first trigger signal by the second comparator when the DC signal exceeds a preset threshold;

Receiving the first trigger signal by a monostable trigger, and converting the first trigger signal into a second trigger signal, the second trigger signal controlling the power switch to be turned on, and the power supply unit supplies power to the microphone through the power switch An array unit; wherein the voice recognition time length is calculated by the microphone array unit, and if the recognition time length reaches a preset time, and the microphone array unit does not recognize the key voice signal, the second trigger signal controls the power source The unit switch is off.

Advantageous effects of the invention

Beneficial effect

The voice wake-up device provided in the present invention adopts a voice activated unit and a microphone array unit, wherein only the voice activated unit operates when the device is in the standby mode, and the microphone array unit must start working by activation of the voice activated unit, so that standby power consumption is achieved. Greatly reduced, the standby current of the device, especially when the ambient noise is not large, the standby current is greatly reduced, even if the small battery can provide enough power The amount of voice wake-up device can be applied to the mobile intelligent terminal, which greatly improves the application range of the voice wake-up device.

Brief description of the drawing

DRAWINGS

1 is a schematic structural diagram of a voice wake-up device according to an embodiment of the present invention;

2 is a schematic structural diagram of a voice wake-up device according to Embodiment 1 of the present invention;

3 is a schematic structural diagram of a voice activated unit according to Embodiment 1 of the present invention;

4 is a schematic structural diagram of a voice activated unit according to Embodiment 2 of the present invention;

FIG. 5 is a schematic diagram of a method for awakening a voice according to Embodiment 3 of the present invention; FIG.

FIG. 6 is a schematic diagram of a method of step S1 in Embodiment 3 of the present invention;

FIG. 7 is a schematic diagram of a method of step S2 in Embodiment 3 of the present invention;

FIG. 8 is a schematic diagram of a method for awakening a voice according to Embodiment 4 of the present invention.

The implementation, functional features, and advantages of the present invention will be further described in conjunction with the embodiments.

BEST MODE FOR CARRYING OUT THE INVENTION

BEST MODE FOR CARRYING OUT THE INVENTION

The specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

1-7, a voice wake-up device provided by the present invention includes a voice activated unit 10, a microphone array unit 20, a main processor unit 30, and a power supply unit 40.

The voice control activation unit 10 is configured to collect and process a voice signal, and is provided with a preset condition. When the processed voice signal satisfies a preset condition, the voice activation unit 10 generates a trigger signal and transmits the signal to the microphone array unit 20. To activate the microphone array unit 20 to receive a voice signal.

For example, if the preset voice signal is "hello", the preset voice is the pre-awake word "hello". When the pre-awake word is collected and processed by the voice activated unit 10, the voice signal of "hello" is obtained. The signal satisfies the preset condition, and the voice activated unit 10 outputs a trigger signal to activate the microphone array unit 20.

The above preset conditions can be set according to user requirements. Further, for higher voice sensitivity and anti-noise performance, the pre-awake words can use a two-syllable word or two monosyllabic words, and the fundamental frequency of the sound can be from 400HZ to 3KHZ. Between, in order to use the hardware bandpass filter to filter out-of-band noise.

The microphone array unit 20 is configured to receive a voice signal, and is pre-set with a program for identifying a key voice signal and a preset time for the microphone array unit 20 to recognize the wake-up voice signal. After the microphone array unit 20 is activated, when the received voice signal is Recognized as a key speech signal, the microphone array unit 20 outputs a wake-up signal, and the wake-up signal wakes up the main processor unit 30, while the microphone array unit 20 exits the program for recognizing the key speech signal and waits to receive and recognize subsequent speech signals, for example, using local The speech recognition engine or networked to the cloud server further identifies other speech signals; when the received speech signal is not a key speech signal, the speech signal cannot be recognized by the microphone array unit 20, and after a preset time, the microphone array unit 20 enters a non-operational state.

For example, the key wake-up word can be set to "alexa", then the key speech signal is "alexa" speech signal, and the preset time is 5s, when the microphone array unit 20 receives and passes beamforming, multi-grain noise reduction, speech enhancement, Echo cancellation and other methods to accurately identify the "alexa" speech signal, then output a wake-up signal, the wake-up signal wakes up the main processor, after 5s, the microphone array unit 20 still does not recognize the "alexa" speech signal, then the microphone array Unit 20 enters a non-working state.

When the main processor unit 30 is outputted with a wake-up signal and a corresponding operation is performed, until the sleep signal is received, the main processor unit 30 and the microphone array unit 20 are brought into an inoperative state.

The power supply unit 40 is powered by the voice activated unit 10, the microphone array unit 20, and the main processor unit 30, respectively, so that it can operate normally.

Thus, in the standby mode, only the voice activated unit 10 operates in the standby mode, the microphone array unit 20 has to start working by the activation of the voice activated unit 10, and the main processor unit 30 needs to wake up through the microphone array unit 20 to start working. The standby power consumption is greatly reduced, and the standby current is reduced, so that the voice wake-up device can be applied to the mobile smart terminal.

Embodiment 1:

As shown in FIG. 2 to FIG. 3, the voice activated unit 10 includes a first microphone 110, a first band pass filter amplifier 111, a first rectifying filter 112, and a first comparator 113, and the first microphone 110 and the first band pass. The wave amplifier, the first rectification filter 112, and the first comparator 113 are sequentially connected.

The first microphone 110 is configured to collect sound and convert the voice into a voice signal; the first band pass filter amplifier 111 is configured to receive the voice signal output by the microphone, and transmit the voice signal amplification process to the first rectification filter 112; A rectifying filter 112 is configured to receive the amplified speech signal and the speech signal The processing is converted into a DC signal and sent to the first comparator 113; the first comparator 113 is for comparing the DC signal, and when the DC signal exceeds the threshold value preset by the first comparator 113, the output ratio of the first comparator 113 is inverted. , that is, the trigger signal.

The preset condition is that the DC signal in this embodiment exceeds the threshold value preset by the first comparator 113.

The microphone array unit 20 includes a single chip 201 and a microphone array 202. The first comparator 113 is connected to an interrupt port of the single chip 201, and the single chip 201 is connected to the microphone array 202 and the main processor unit 30, respectively.

The single chip 201 is used to receive the trigger signal of the voice activated unit 10, and after receiving the trigger signal, the microphone array 202 is activated, that is, the level flip pulse edge outputted by the first comparator 113 triggers the interrupt of the single chip 201, and the microphone array 202 is controlled to enter the working state. And entering the preset time, the microphone array 202 is configured to receive the voice signal, and the voice signal is sent to the single chip 201; the single chip 201 receives the voice signal sent by the microphone array 202, and determines whether the voice signal satisfies the key voice signal, if Then, the single chip 201 outputs a wake-up signal for waking up the main processor unit 30. If not, the single chip 201 outputs a sleep signal for causing the microphone array unit 20 to sleep after the recognized preset time elapses.

Embodiment 2:

As shown in FIG. 4, the voice activated unit 10 includes a second microphone 120, a second band pass filter amplifier 121, a second rectifying filter 122, a second comparator 123, and a monostable flip-flop 124, a second microphone 120, The two-band pass wave amplifier, the second rectifying filter 122, the second comparator 123, and the one-shot flip-flop 124 are sequentially connected.

The second microphone 120 is configured to pick up the sound and convert the voice into a voice signal; the second band pass filter amplifier 121 is configured to receive the voice signal output by the microphone, and transmit the voice signal amplification process to the second rectifying filter 122; The second rectifying filter 122 is configured to receive the amplified speech signal, and convert the cover speech signal processing into a DC signal to the second comparator 123; the second comparator 123 is configured to compare the DC signal when the DC signal exceeds the second comparison. The preset threshold value of the first comparator 113 is outputted by the first comparator 113, that is, the first trigger signal; the one-shot trigger 124 is configured to receive the first trigger signal and convert the first trigger signal into the second trigger. The signal, i.e., the level flip pulse edge output by the second comparator 123, triggers the monostable flip-flop 124, which outputs a pulse with a time limit.

The preset condition is that the DC signal in the embodiment exceeds the threshold preset by the second comparator 123. value.

A power switch 125 is connected between the microphone array unit 20 and the power supply unit 40, and the monoflop 124 is connected to the power switch 125.

The time-limited pulsed control power switch 125 is turned on, and the power supply unit 40 can be powered by the power switch 125 to the microphone array unit 20, so that the microphone array unit 20 can enter a normal recognition working state.

The time-limited pulse may be a monostable pulse of 10 seconds. After the microphone array unit 20 has not recognized the key speech signal after 10 seconds, the pulse disappears, the power switch 125 is turned off, and the microphone array unit 20 is powered off. If the microphone array unit 20 recognizes the key speech signal within 10 seconds, the microphone array unit 20 outputs the wake-up signal and wakes up the main processor unit 30, while the microphone array unit 20 exits the process of recognizing the key speech signal and waits for reception and recognition. Subsequent voice signals.

The power switch 125 is also connected to the main processor unit 30. When the main processor unit 30 is woken up, the control signal is outputted to control the power switch 125 to remain open after the one-shot pulse ends, so that the power supply unit 40 can continue to be the microphone array. The unit 20 is powered; when the main processor unit 30 receives the sleep signal, the main processor unit 30 controls the power switch 125 to turn off, causing the microphone array unit 20 to be powered down.

As shown in FIG. 5, the present invention further provides a voice wake-up method corresponding to the voice wake-up device, which includes the following steps:

Step S1: collecting and processing the voice signal through the voice activated unit 10, when the voice signal meets the preset condition, the trigger signal is output through the voice activated unit 10, and the trigger signal activates the microphone array unit 20;

Step S2: receiving a voice signal through the microphone array unit 20, when the voice signal is recognized as a key voice signal, outputting a wake-up signal through the microphone array unit 20, the wake-up signal wakes up the main processor unit 30; or, is calculated by the microphone array unit 20 The length of the speech recognition time. If the recognition time length reaches the preset time, and the microphone wind array unit does not recognize the key speech signal, the microphone array unit 20 is controlled to enter the non-operation state.

Embodiment 3:

Further, as shown in FIG. 5 to FIG. 7 , in step S1 of the voice waking method, the voice activation unit 10 collects and outputs the trigger signal, and specifically includes the following steps:

Step S10: picking up the sound through the first microphone 110 and converting the voice into a voice signal, and inputting the voice signal Sent to the first band pass filter amplifier 111;

Step S11: the voice signal is amplified by the first band pass filter amplifier 111 and then sent to the first rectification filter 112;

Step S12: The amplified voice signal is converted into a DC signal by the first rectification filter 112, and sent to the first comparator 113;

Step S13: comparing the DC signal with the preset threshold by the first comparator 113, and outputting the trigger signal by the first comparator 113 when the DC signal exceeds the preset threshold.

In addition, in step 2, the specific steps of the microphone array unit 20 identifying and outputting the wake-up signal include the following:

Step S20: receiving the trigger signal by the single chip microcomputer 201, and starting the microphone array 202 and entering the timing of the preset time;

Step S21: receiving the voice signal outputted by the microphone array 202 through the single-chip microcomputer 201, and determining whether the voice signal is a key voice signal. If the voice signal is a key voice signal, the wake-up signal is output through the single-chip microcomputer 201, and the wake-up signal will be the main processor unit. 30 wake-up, if the voice signal is not a key voice signal, the sleep signal is outputted by the single-chip microcomputer 201 after a preset time, and the signal signal causes the microphone array unit 20 to go to sleep.

The preset condition is that the DC signal in the embodiment exceeds a preset threshold, and the trigger signal output by the first comparator 113 may be a level inversion of the output of the first comparator 113, and the pulse edge triggers the interrupt control of the single chip 201. The microphone array 202 starts to work and enters the timing of the preset time; wherein the key voice signal can be the voice signal of the preset "alexa" wake-up keyword, and when the single-chip microcomputer 201 receives the voice signal sent by the microphone array 202 as "alexa" The voice signal of the wake-up keyword, that is, the voice signal is judged as the key voice signal, and the single chip 201 outputs the wake-up signal; the microcontroller 201 does not receive the "alexa" wake-up keyword transmitted by the microphone array 202 within the preset time. The voice signal is outputted by the single chip microcomputer 201 to cause the microphone array unit 20 to sleep.

Embodiment 4:

As shown in FIG. 8, the step S1 of the above-mentioned voice waking method includes the following steps: the voice activation unit 10 collects and outputs the trigger signal.

Step S10': picking up the sound through the second microphone 120 and converting the voice into voice, and transmitting the voice signal To the second band pass filter amplifier 121;

Step S11': the speech signal is amplified by the second band pass filter amplifier 121 and then sent to the second rectification filter 122;

Step S12': the amplified voice signal is converted into a DC signal by the second rectification filter 122, and sent to the second comparator 123;

Step S13': comparing the DC signal of the second comparator 123 with a preset threshold, and when the DC signal exceeds a preset threshold, outputting the first trigger signal by the second comparator 123;

Step S14': receiving the first trigger signal through the monoflop 124 and converting the first trigger signal into a second trigger signal, the second trigger signal controlling the power switch 125 to be turned on, and the power unit 40 is powered by the power switch 125 to the microphone The array unit 20; wherein the voice recognition time length is calculated by the microphone array unit 20, and if the recognition time length reaches a preset time, and the microphone array unit 20 does not recognize the key voice signal, the second trigger signal controls the power switch 125 to be turned off. .

The preset condition is that the DC signal in the embodiment exceeds a preset threshold, and the first trigger signal output by the second comparator 123 may be a level inversion of the output of the second comparator 123, and the pulse edge triggering The steady trigger output pulse, when the pulse has a time limit, for example, the pulse is a 10 second pulse, the pulse control power switch 125 is turned on for 10 seconds, so that the power supply unit 40 is powered by the microphone array unit 20, and if the microphone is within 10 seconds The array unit 20 recognizes the key speech signal, the microphone array unit 20 outputs a wake-up signal waking up the main processor unit 30, and the power switch 125 is continuously turned on; otherwise, the power switch 125 is turned off after 10 seconds.

In addition, when the device is in operation or continuous ambient noise is relatively large, the microphone array unit 20 will cyclically switch between the inactive state and the wake-up keyword recognition control program, thereby avoiding missing the wake-up keyword.

In summary, the voice wake-up device provided in the present invention is combined with the voice activated unit 10 and the microphone array unit 20, wherein only the voice activated unit 10 operates when the device is in the standby mode, and the microphone array unit 20 has to pass the voice activated unit 10 After the activation starts, the main processor unit 30 needs to wake up through the microphone array unit 20 to start working, so that the standby power consumption is greatly reduced, and the standby current of the device is special, on the basis of ensuring the normal operation of the voice wake-up device. When the ambient noise is not large, the standby current is greatly reduced, even if the small battery can provide enough power to make the voice wake up. The device can be applied to the mobile intelligent terminal, which improves the application range of the voice wake-up device.

The above is only the preferred embodiment of the present invention, and is not intended to limit the scope of the invention, and the equivalent structure or equivalent process transformations made by the description of the invention and the drawings are directly or indirectly applied to other related The technical field is equally included in the scope of patent protection of the present invention.

Claims (12)

1. A voice wake-up device includes: a main processor unit, further comprising a voice activated unit and a microphone array unit;

   The voice activated unit is configured to collect and process a voice signal, and when the voice signal meets a preset condition, generate a trigger signal and transmit the signal to the microphone array unit to activate the microphone array unit to receive a voice signal;

   The microphone array unit is configured to receive and recognize a voice signal, and when the key voice signal is recognized, output a wake-up signal, the wake-up signal is used to wake up the main processor unit; wherein, when the microphone array unit is activated The subsequent recognition time reaches a preset time, and the microphone array unit does not recognize the key speech signal, and the microphone array unit enters a non-operation state.
2. The voice wake-up device according to claim 1, wherein said microphone array unit comprises a single chip microcomputer and a microphone array;

   The single chip is configured to receive a trigger signal of the voice activated unit, activate the microphone array, and enter a timing of a preset time, and after receiving the voice signal of the microphone array, determine whether the voice signal is a key voice signal And if yes, outputting a wake-up signal for waking up the main processor unit, and if not, outputting a sleep signal for sleeping the microphone array unit after a preset time;

   The microphone array is configured to receive a voice signal and deliver the voice signal to the single chip microcomputer.
3. The voice wake-up device according to claim 2, wherein the voice activated unit comprises a first microphone, a first band pass filter amplifier, a first rectifying filter, and a first comparator;

   The first microphone is configured to pick up a sound and convert the voice into a voice signal;

   The first band pass filter amplifier is configured to receive a voice signal output by the first microphone, and transmit a voice signal amplification process to the first rectifier filter;

   The first rectifying filter is configured to receive the amplified speech signal, and convert the speech signal processing into a DC signal, and send the signal to the first comparator;

   The first comparator is configured to compare the DC signal, and when the DC signal exceeds a preset threshold value of the first comparator, the first comparator outputs the trigger signal.
4. The voice waking device of claim 1 , further comprising a power switch, wherein the power switch is respectively connected to the power unit, the voice activated unit, and the microphone array unit;

   The voice activated unit outputs a trigger signal, the trigger signal controls the power switch to be turned on, and the power unit supplies power to the microphone array unit through the power switch to cause the microphone array unit to enter an identification working state;

   The trigger signal controls the power switch to be turned off when the microphone array unit recognizes that the time reaches the preset time without recognizing a key voice signal.
5. The voice wake-up device according to claim 4, wherein the power switch is further connected to the main processor unit, and the main processor unit controls the power switch to be turned off after receiving the sleep signal.
6. The voice wake-up device according to claim 5, wherein the voice activated unit comprises a second microphone, a second band pass filter amplifier, a second rectifying filter, a second comparator, and a monostable flip-flop;

   The second microphone is configured to pick up a sound and convert the voice into a voice signal;

   The second band pass filter amplifier is configured to receive a voice signal output by the second microphone, and transmit the voice signal amplification process to the second rectifier filter;

   The second rectifying filter is configured to receive the amplified speech signal, and convert the speech signal processing into a DC signal, and send the signal to the second comparator;

   The second comparator is configured to compare the DC signal, and when the DC signal exceeds a preset threshold value of the comparator, the second comparator outputs a first trigger signal;

   The one-shot trigger is configured to receive the first trigger signal and convert the first trigger signal into a second trigger signal, where the second trigger signal controls the power switch.
7. A voice wake-up method is characterized in that it specifically includes the following steps:

   Acquiring and processing the voice signal by the voice activated unit, and outputting the trigger signal by the voice activated unit when the voice signal meets the preset condition, the trigger signal activating the microphone array unit;

   Receiving a voice signal through the microphone array unit, and outputting a wake-up signal through the microphone array unit when the voice signal is recognized as a key voice signal, the wake-up signal waking up the main processor unit; or calculating by the microphone array unit The length of the voice recognition time. If the length of the recognition time reaches a preset time, and the microphone array unit does not recognize the key voice signal, the microphone array unit is controlled to enter a non-working state.
8. The voice wake-up method according to claim 7, wherein the specific steps of the voice activated unit to acquire and output the trigger signal include the following:

   Receiving a sound through the first microphone and converting the voice into a voice signal, and transmitting the voice signal to the first band pass filter amplifier;

   And the sound signal is amplified by the first band pass filter amplifier and then sent to the first rectifier filter;

   Converting the amplified voice signal into a DC signal by the first rectifying filter and transmitting to the first comparator;

   The DC signal is compared with a preset threshold by the first comparator, and the trigger signal is output by the first comparator when the DC signal exceeds a preset threshold.
9. The voice wake-up method according to claim 8, wherein the specific steps of the microphone array unit to recognize and output the wake-up signal include the following:

   Receiving the trigger signal by the single chip microcomputer, and starting the microphone array and entering the timing of the preset time;

   Receiving a voice signal through the microphone array, and transmitting the voice signal to the single chip microcomputer;

   Receiving, by the single chip microcomputer, a voice signal output by the microphone array, and determining whether the voice signal is a key voice signal, if the voice signal is a key voice signal Then, the wake-up signal is outputted by the single-chip microcomputer, and the wake-up signal wakes up the main processor unit. If the voice signal is not a key voice signal, the sleep signal is output after the preset time passes by the single-chip microcomputer, and the sleep signal makes the The microphone array unit goes to sleep.
10. The method for waking up a voice according to claim 7, further comprising:

    The voice activated unit outputs a trigger signal, the trigger signal controls the power switch to be turned on, and the power unit supplies power to the microphone array unit through the power switch to cause the microphone array unit to enter an identification working state;

    The trigger signal controls the power switch to be turned off when the microphone array unit recognizes that the time reaches the preset time without identifying a key voice signal.
11. [Correct according to Rule 26 12.03.2018]
    

    The method for waking up a voice according to claim 10, further comprising:

    The main processor unit controls the power switch to be turned off after receiving the sleep signal.
12. [Correct according to Rule 26 12.03.2018]
    

    The voice wake-up method according to claim 11, which specifically comprises the following steps:

    Receiving a sound through the second microphone and converting the voice into a voice signal, and transmitting the voice signal to the second band pass filter amplifier;

    And amplifying the voice signal by using the second band pass filter amplifier to be sent to the second rectifier filter;

    Converting the amplified voice signal into a DC signal by the second rectifier filter and transmitting the signal to the second comparator;

    Comparing the DC signal with a preset threshold by the second comparator, and outputting a first trigger signal by the second comparator when the DC signal exceeds a preset threshold;

    Receiving the first trigger signal by a monostable trigger, and converting the first trigger signal into a second trigger signal, the second trigger signal controlling the power switch to be turned on, and the power supply unit supplies power to the microphone through the power switch An array unit; wherein the voice recognition time length is calculated by the microphone array unit, and if the recognition time length reaches a preset time, and the microphone array unit does not recognize the key voice signal, the second trigger signal controls the power source The unit switch is off.

Images