Air conditioner, voice recognition method thereof, terminal server and readable storage medium

Abstract

The application discloses a voice recognition method of an air conditioner, the air conditioner and a computer readable storage medium, and comprises the following steps: a terminal server receives voice information uploaded by a master machine; the type of the voice information is recognized; when it is recognized that the type of the voice information is a wake-up voice, adjustment information is sent to a slave machine, so that the slave machine adjusts its operation parameters according to the adjustment information, or reduces the working noise of the slave machine after adjusting its working mode. The embodiment of the application improves the voice recognition effect of the slave machine of the air conditioner in a noisy environment and improves the use effect of the air conditioner.

Description

Technical Field

[0001] This invention relates to the field of speech recognition technology, and in particular to an air conditioner and its speech recognition method, terminal server and readable storage medium. Background Technology

[0002] Climate change and global warming have made air conditioners, as air conditioning devices, an indispensable household appliance. As people's demands increase, the simple structure of air conditioners can no longer meet their needs. Therefore, a type of air conditioner, the "main unit and sub-unit," has been introduced. This type of air conditioner includes a main unit and at least one sub-unit. The main unit is the primary unit and is installed in a fixed location. The sub-unit can be moved into the main unit for charging; alternatively, the sub-unit can be detached from the main unit and moved to different rooms to regulate the air in those rooms, achieving localized environmental control.

[0003] Currently, the following problems exist in the implementation of voice control for dual-zone air conditioners:

[0004] If the noise generated during the operation of the slave unit is too loud, the noise will interfere with the user's voice commands, causing the slave unit to be unable to pick up the user's voice commands or causing the slave unit's voice recognition to fail.

[0005] It should be noted that the above content is only used to help understand the technical problem solved by the present invention, and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main objective of this invention is to provide an air conditioner and its voice recognition method, terminal server, and readable storage medium. The aim is to enable the air conditioner's sub-unit to obtain the user's wake-up voice in a timely and accurate manner when operating in a noisy mode, so as to adjust the state in a timely manner, improve the success rate of picking up the sub-unit's control voice, enhance the voice recognition effect of the air conditioner, and improve the user experience.

[0007] Based on this, the present invention provides a voice recognition method for an air conditioner, applied to a terminal server, wherein the terminal server communicates with the air conditioner, the air conditioner comprising a main unit and a movable slave unit detachable from the main unit, and the method comprising:

[0008] The terminal server receives the voice information uploaded by the host machine;

[0009] Identify the type of the voice information;

[0010] When the voice information is identified as a wake-up voice, adjustment information is sent to the sub-machine, so that the sub-machine adjusts its own operating parameters or its own working mode according to the adjustment information, thereby reducing the operating noise of the sub-machine.

[0011] Optionally, the method further includes:

[0012] When the voice information is identified as control voice, the control object corresponding to the control voice is determined.

[0013] If the controlled object is the sub-machine, then a control command corresponding to the control voice is sent to the sub-machine, causing the sub-machine to execute the control operation corresponding to the control command.

[0014] Optionally, the step of sending adjustment information to the slave device when the type of the voice information is identified as a wake-up voice includes:

[0015] When the voice information is identified as a wake-up voice, the control object corresponding to the wake-up voice is determined. The wake-up voice of the slave unit is different from that of the master unit.

[0016] If the controlled object is a submachine, then adjustment information is sent to the submachine.

[0017] Optionally, before the step of the terminal server receiving the voice information uploaded by the host machine, the method further includes:

[0018] When the terminal server receives a working mode of the slave machine uploaded by the slave machine, it sends a voice information upload request to the master machine, so that the master machine collects the voice information and uploads the voice information to the terminal server. The first mode is a mode in which the working noise is greater than or equal to a preset threshold.

[0019] Alternatively, the step of sending adjustment information to the slave device when the type of the voice information is identified as a wake-up voice includes:

[0020] When the voice information is identified as a wake-up voice and the working mode uploaded by the slave device is the first mode, adjustment information is sent to the slave device.

[0021] Optionally, before the step of the terminal server receiving the voice information uploaded by the host machine, the method further includes:

[0022] When the distance between the working position of the slave unit and the user's position is greater than or equal to a preset distance, a voice information upload request is sent to the master unit, so that the master unit collects the voice information and uploads it to the terminal server.

[0023] Optionally, the adjustment information includes at least one of a wake-up voice, an operating parameter adjustment command, and a working mode switching command.

[0024] Optionally, applied to the sub-unit of an air conditioner, the method includes:

[0025] The terminal server receives adjustment information, which is voice information uploaded by the host machine and is sent when the voice information is wake-up voice.

[0026] Based on the adjustment information, adjust the operating parameters or the working mode to reduce operating noise.

[0027] Optionally, the steps for adjusting operating parameters include:

[0028] Reduce the fan speed of the sub-unit;

[0029] Alternatively, the steps to adjust the work mode include:

[0030] Switch the working mode of the slave unit to the second mode, which is a mode that makes the working noise less than a preset threshold.

[0031] Optionally, after the step of adjusting the operating parameters or adjusting the working mode according to the adjustment information to reduce operating noise, the method further includes:

[0032] When a control voice is detected, the control command corresponding to the control voice is executed.

[0033] To achieve the above objectives, the present invention also provides a terminal server, the terminal server including a memory, a processor, and a speech recognition program stored in the memory and executable on the processor, wherein the speech recognition program, when executed by the processor, implements the various steps of the speech recognition method for an air conditioner as described above.

[0034] The present invention also provides an air conditioner, the air conditioner including a memory, a processor, and a voice recognition program stored in the memory and executable on the processor, wherein when the voice recognition program is executed by the processor, it implements the steps of the voice recognition method of the air conditioner as described above.

[0035] In addition, a computer-readable storage medium stores a speech recognition program that, when executed by a processor, implements the various steps of the speech recognition method for an air conditioner as described above.

[0036] The air conditioner, its voice recognition method, terminal server, and readable storage medium provided by this invention allow the slave unit to operate in a noisy mode. By combining voice information collected by the master unit, the slave unit can be controlled, such as adjusting its operating parameters or switching its operating mode, thereby reducing noise and improving the environment for voice recognition, thus increasing the success rate of voice recognition. The voice information collected by the master unit is relayed to the slave unit, enabling it to receive voice commands that would otherwise be impossible to pick up in noisy conditions, such as wake-up commands. This not only prevents the slave unit from being unable to wake up in noisy operating modes but also allows it to improve its operating environment in a timely manner, thereby enhancing subsequent voice recognition performance, improving the air conditioner's voice recognition capabilities, and improving the user experience. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the hardware architecture of the air conditioner according to an embodiment of the present invention;

[0038] Figure 2 This is an interactive diagram of an embodiment of the voice recognition method for an air conditioner according to the present invention;

[0039] Figure 3 This is another interactive diagram of an embodiment of the voice recognition method for an air conditioner of the present invention;

[0040] Figure 4 This is an interactive diagram of another embodiment of the voice recognition method for an air conditioner according to the present invention;

[0041] Figure 5 This is an interactive diagram of another embodiment of the voice recognition method for an air conditioner according to the present invention;

[0042] Figure 6 This is a diagram of the working environment of the air conditioner involved in the voice recognition method of the air conditioner of the present invention.

[0043] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0044] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0045] Climate change and global warming have made air conditioners, as air conditioning devices, an indispensable household appliance. As people's demands increase, the simple structure of air conditioners can no longer meet their needs. Therefore, manufacturers have introduced a type of air conditioner called a "main unit and sub-unit." This type of air conditioner includes a main unit and at least one sub-unit. The main unit is the primary unit and is installed in a fixed location. The sub-unit can be moved into the main unit for charging; alternatively, the sub-unit can be detached from the main unit and moved to different rooms to regulate the air in those rooms, achieving localized environmental control.

[0046] As a modern intelligent appliance, voice recognition is an important function of dual-unit air conditioners. However, the following problems exist in the implementation of voice control in dual-unit air conditioners: If the noise generated during the operation of the slave unit is too loud, the noise will interfere with the user's voice commands, causing the slave unit to be unable to pick up the user's voice commands or causing the slave unit's voice recognition to fail.

[0047] Based on this, this invention provides a voice recognition method for air conditioners. The main unit is located in a fixed position and does not generate significant noise during operation. The main unit collects the user's voice information, and then forwards voice information for waking up or controlling the slave unit to it. The slave unit then responds accordingly based on the forwarded voice information. Thus, even when the slave unit operates in a noisy mode, it can still obtain the user's voice information from the main unit, avoiding situations where the slave unit cannot recognize voice commands or has a low voice information recognition rate in noisy operating modes. This improves the voice recognition effect of the air conditioner and enhances the user experience.

[0048] To better understand the above technical solutions, exemplary embodiments of this disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this disclosure to those skilled in the art.

[0049] As one implementation method, the hardware environment architecture involved in the voice recognition method for the air conditioner can be as follows: Figure 1 As shown.

[0050] Optionally, the hardware architecture involved in the voice recognition method for the air conditioner may include the air conditioner and / or a terminal server. The air conditioner, such as a mother-daughter air conditioner, includes a mother unit and a movable daughter unit that can be detached from the mother unit. Optionally, the daughter unit and the mother unit may be controlled by the same controller; alternatively, both the daughter unit and the mother unit may also have their own controllers, and each unit is independently controlled by its own controller, but they can communicate with each other. Optionally, the daughter unit and the mother unit communicate through the terminal server.

[0051] In one implementation, the air conditioner or terminal server includes: a processor 101, such as a CPU, a memory 102, and a communication bus 103. The communication bus 103 is used to establish communication between these components. The processor 102 is used to invoke an application program to perform voice recognition operations.

[0052] The memory 102 can be a high-speed RAM or a stable memory (non-volatile memory), such as a disk storage device.

[0053] It is understood that, in one embodiment, the voice recognition program for the air conditioner is stored in the memory 102 of the terminal server, or in a computer-readable storage medium. When the processor 101 retrieves the voice recognition program from the memory 102, it performs the following operations:

[0054] The terminal server receives the voice information uploaded by the host machine;

[0055] Identify the type of the voice information;

[0056] When the voice information is identified as a wake-up voice, adjustment information is sent to the sub-machine, so that the sub-machine adjusts its own operating parameters or its own working mode according to the adjustment information, thereby reducing the operating noise of the sub-machine.

[0057] It is understood that, in another embodiment, the voice recognition program for the air conditioner is stored in the air conditioner's memory 102 or in a computer-readable storage medium. When the processor 101 retrieves the voice recognition program from the memory 102, it performs the following operations:

[0058] The terminal server receives adjustment information, which is voice information uploaded by the host machine and is sent when the voice information is wake-up voice.

[0059] Based on the adjustment information, adjust the operating parameters or the working mode to reduce operating noise.

[0060] Based on the hardware architecture of the air conditioner and terminal server described above, the following embodiments of the present invention are proposed.

[0061] Please refer to Figure 2 This embodiment presents an example of a voice recognition method for an air conditioner. The air conditioner used in this embodiment differs from conventional air conditioners. (See reference...) Figure 6 The air conditioner includes a main unit and at least one sub-unit. The main unit is the host unit and is installed in a fixed position. The sub-unit can be moved into the main unit for charging. The sub-unit can also be detached from the main unit and moved to various rooms to regulate the air in each room.

[0062] Both the main unit and the slave unit of the air conditioner have voice recognition capabilities, enabling them to be controlled via voice recognition.

[0063] It's important to note that speech recognition includes online and offline speech recognition. Online speech recognition involves the slave or master unit collecting speech information, uploading it to a terminal server, which analyzes the speech information, converts it into control commands, and then returns them to the slave or master unit, which executes the corresponding control operation. Offline speech recognition involves the slave or master unit collecting speech information, analyzing it using a locally stored speech recognition model, converting it into control commands, and then responding to the commands by executing the corresponding control operation. Speech information includes wake-up voice and control voice. Wake-up voice refers to the voice used to wake up the device, such as "Xiao Mei Xiao Mei," "Da Mei Da Mei," "Xiao A Xiao A," "Da A Da A," etc. Control voice refers to the voice used to control the device, such as "Adjust the fan speed to low" or "Turn off the cooling." Because the recognition model for wake-up voice is relatively fixed and the recognition process is relatively simple, offline speech recognition is generally used to achieve rapid device response to users. Control voice commands are more numerous and complex, requiring a terminal server for recognition under stable network conditions; therefore, online speech recognition is generally used for control voice. Therefore, when a device captures a wake-up voice, it generally does not upload it to the terminal server, but instead directly wakes up the device to capture the control voice subsequently issued by the user.

[0064] However, in the application of a dual-unit air conditioner, the slave unit generates noise when operating certain modes. Excessive noise can affect the success rate of voice recognition on the slave unit, causing its voice recognition function to fail and the slave unit to not respond to user voice commands. It should be noted that the slave unit is configured to operate in a mode where the ambient noise is greater than or equal to a preset threshold, such as a high-power purification mode. When the slave unit operates in high-power purification mode, the fan speed is high, generating significant wind noise.

[0065] Therefore, in this embodiment of the invention, when the slave device operates in these modes, the master device assists in speech recognition to avoid situations where the slave device fails to respond to user voice commands. That is, the master device uploads the collected voice information to the terminal server, and the terminal server then forwards the voice control commands for the slave device to the slave device, thereby supplementing the voice data that the slave device cannot collect. Several implementation methods of this embodiment are listed below.

[0066] For one possible implementation, please refer to Figure 2 , Figure 2 The diagram shows an interaction of a voice recognition method for an air conditioner, which includes:

[0067] Step S21: Upload Working Mode. During operation, the slave device uploads its working mode to the terminal server for recording. Optionally, the slave device can upload the mode to the terminal server whenever it is running any mode, or it can upload the first mode only when running the first mode.

[0068] Step S22: Determine if the operating mode is the first mode. After receiving the operating mode of the slave unit, the terminal server determines whether the operating mode of the slave unit is the first mode, which is the operating mode that causes the operating noise to be greater than or equal to a preset threshold. If the operating mode is the first mode, then proceed to step S23. If the operating mode is not the first mode, then interactive control is performed according to the normal interaction with the slave unit and the master unit.

[0069] Step S23: Request the master unit to upload the collected voice information. The voice information includes at least one of wake-up voice and control voice. When the terminal server determines that the slave unit's operating mode is the first mode, it indicates that the slave unit's current operating mode generates a lot of noise, resulting in a low success rate for the slave unit to pick up user voice commands. Therefore, the terminal server requests the master unit to assist in collecting user voice commands and uploading them to the terminal server. This allows the terminal server to forward the voice information or the corresponding control command to the slave unit when it determines that the voice information pertains to slave unit control. In this way, user voice commands can also be transmitted to the slave unit. When the master unit receives the voice information upload request, it will upload all voice information it collects to the terminal server during subsequent operation.

[0070] Step S24: The master unit uploads the wake-up voice. During operation, the microphone of the master unit is turned on. If the microphone picks up the wake-up voice, the master unit uploads the picked-up wake-up voice to the terminal server.

[0071] Step S25: Send adjustment information to the submachine.

[0072] When the master unit uploads a wake-up voice message, this message could be a user waking up the master unit or a user waking up the slave unit (in scenarios where the wake-up words for the slave and master units are the same, such as when the user says "Xiao Mei Xiao Mei" to wake up the slave unit and / or the master unit). In this embodiment, regardless of whether the wake-up is to the slave unit or the master unit, adjustment information is sent to the slave unit. This allows the slave unit to adjust its operating parameters or its working mode according to the adjustment information, thereby reducing the operating noise of the slave unit. This facilitates listening to the user's subsequent control voice messages. If the control voice message is directed to the slave unit, it is responded to; if it is directed to the master unit, it is not responded to. The working mode refers to the slave unit's operating mode.

[0073] Furthermore, to further improve control accuracy, the terminal server can also send adjustment information to the slave device when it determines that the voice information is a wake-up voice and the slave device uploads the first working mode. In other words, the slave device only needs to forward the wake-up voice uploaded by the master device when it is in the first mode scenario.

[0074] The adjustment information may be a wake-up voice message, such as forwarding the wake-up voice message collected by the master unit to the slave unit to wake it up. Alternatively, the adjustment information may be at least one of an operating parameter adjustment command and a working mode switching command, such as instructing the slave unit to adjust its own operating parameters to reduce operating noise, or instructing the slave unit to switch its current working mode to a mode with less operating noise, such as a silent mode or a windless mode, etc.

[0075] Step S26: Adjust operating parameters or adjust the working mode. The slave unit receives the adjustment information and responds accordingly. For example, if the adjustment information is a wake-up voice command, the microphone is turned on to pick up the control voice, and the fan speed is reduced to decrease wind noise. If the adjustment information is an operating parameter adjustment command, the operating parameters are adjusted according to the command, such as reducing the fan speed. If the adjustment information is a working mode switching command, the unit switches to the indicated second mode, such as switching to silent mode. In this way, after the slave unit is adjusted, a good environment is provided for collecting the user's control voice, improving the success rate of voice recognition.

[0076] Step S27: Collect control voice and execute the control commands corresponding to the control voice. After the slave unit is woken up and the operating parameters are adjusted, or after switching the working mode, the microphone is turned on. At this time, the microphone of the slave unit collects voice information in the environment. In a relatively quiet environment, the slave unit can accurately collect the user's voice information and respond in a timely manner.

[0077] In this implementation, the master unit uploads the acquired wake-up voice. Upon receiving the wake-up voice, the operating parameters of the slave unit are adjusted to switch its operating mode, reducing operating noise and providing a quiet environment for the slave unit to acquire control voice, thus improving the success rate of voice recognition. Furthermore, after switching operating modes or adjusting operating parameters, the slave unit can directly acquire control voice, choosing between online or offline recognition. Even in environments with unstable networks, offline voice control is possible. Compared to relaying voice information through the master unit, this implementation offers greater timeliness and stability.

[0078] For another possible implementation, please refer to Figure 3 , Figure 3 This is an interactive diagram of an air conditioner voice recognition method, which includes the above-mentioned... Figure 2Steps S21, S22, and S23 are described above. Therefore, the specific descriptions of steps S21, S22, and S23 can be found in the above implementation method and will not be repeated here.

[0079] The difference lies in this implementation: both the wake-up and control voice commands of the slave device are relayed by the master device. In other words, the voice information uploaded by the master device includes both control and wake-up voice commands. One possible implementation involves the master device uploading the wake-up voice command; another involves uploading the control voice command. If the uploaded command is a wake-up voice command, then based on the above... Figure 2 Implementation method: If the uploaded data is control voice, then this implementation method will be followed.

[0080] For example, after step S23, the following steps are also included:

[0081] Step S28: The host machine uploads control voice. During operation, the host machine's microphone is turned on. If the microphone picks up control voice, the host machine uploads the picked-up control voice to the terminal server.

[0082] Step S29: Send the control command corresponding to the control voice to the slave unit.

[0083] In one possible implementation, the master unit only uploads control voices for the slave unit (the master unit's own control voices are recognized offline). When the terminal server receives the control voices uploaded by the master unit, it recognizes them as the slave unit's control voices. At this point, the terminal server sends the control command corresponding to the control voice to the slave unit, and the slave unit responds to the control command.

[0084] In another possible implementation, the master unit uploads all collected control voice messages to the terminal server for recognition. During the voice recognition process, the terminal server determines the control object corresponding to the control voice message. If the control object is the slave unit, the server sends a control command corresponding to the control voice message to the slave unit, which responds to the command and executes the corresponding control operation. If the control object is the master unit, the server returns the control command corresponding to the control voice message to the master unit, which responds to the command and executes the corresponding control operation (the process of the master unit recognizing voice messages online).

[0085] Step S30: The slave unit executes the control operation corresponding to the control command.

[0086] Compared to the above implementation, this method eliminates the need for the server to determine whether the slave machine is operating in the first mode. The server simply sends the control commands uploaded from the master machine to the slave machine, enabling it to execute the corresponding control operations. Even if the slave machine does not collect voice information in the first mode, it can still execute the control commands corresponding to that voice information. Furthermore, the slave machine does not need to adjust its operating parameters and can maintain operation in the first mode. However, this requires a stable network environment, and the forwarding of control voice commands from the master machine to the slave machine results in a delay in control command response and untimely reaction.

[0087] Understandably, in this implementation, if the server receives voice information uploaded by both the slave and master machines simultaneously, and the control object and control command corresponding to the voice information are the same, then it only needs to return a control command to either the slave or master machine.

[0088] In another possible implementation, in the scenario where the slave unit operates in the first mode, the voice recognition process of the air conditioner does not include the above. Figure 2 Steps S21, S22, and S23 shown above do not need to be performed. Figure 2 Steps S21, S22, and S23 are shown below. In this implementation, if the air conditioner detects that the slave unit is operating in the first mode (a working mode where the ambient noise is greater than or equal to a preset threshold), the air conditioner activates online voice recognition, which means offline voice recognition is turned off. Voice information collected by the master unit, or voice information collected by both the master and slave units, is uploaded to the terminal server for voice analysis. Specifically, after the master unit uploads the collected voice information to the terminal server, the terminal server performs voice recognition processing; the specific processing procedure is described above. Figure 2 Steps S24 to S27 shown, or refer to or Figure 3 Steps S28 to S29 are shown, so they will not be described again.

[0089] In this embodiment, when the slave unit operates in a noisy mode, it is controlled by combining the voice information collected by the master unit. This control includes adjusting the slave unit's operating parameters or switching its operating mode to reduce noise and improve the environment for voice recognition, thereby increasing the success rate of voice recognition. The voice information collected by the master unit is relayed to the slave unit, enabling it to receive voice commands that would otherwise be impossible to pick up in noisy conditions, such as wake-up commands. This not only prevents the slave unit from being unable to wake up in noisy operating modes but also allows it to improve its operating environment in a timely manner, thus enhancing subsequent voice recognition performance, improving the air conditioner's voice recognition capabilities, and improving the user experience.

[0090] Optionally, in order to make the voice recognition method for an air conditioner easier to implement and simplify the control flow, different wake-up words can be set for the slave unit and the master unit, so as to distinguish whether the user intends to wake up the slave unit or the master unit through different wake-up words. Further, when the slave unit is operating in an operating mode with relatively large working noise, if the master unit is used to forward the wake-up voice, it can also quickly determine whether the specific device to be woken up is the slave unit or the master unit based on different wake-up words. If the master unit is to be woken up, there is no need to adjust the operating mode or operating parameters of the slave unit. In this way, it can avoid adjusting the operating mode or operating parameters of the slave unit regardless of whether the user wakes up the slave unit or the master unit. That is, only when it is clear that the slave unit needs to be woken up, the operating mode or operating parameters of the slave unit are adjusted, which improves user experience and avoids frequent adjustments of the slave unit.

[0091] Based on this, an embodiment of the present invention proposes another embodiment of a voice recognition method for an air conditioner. In this embodiment of the present invention, as in the foregoing embodiment, the applied air conditioner includes a host and a master unit. The structural function and existing problems of the air conditioner are the same as those in the foregoing embodiment, and reference may be made to the foregoing embodiment for details, which will not be repeated herein.

[0092] Different from the foregoing embodiment, in the embodiment of the present invention, different wake-up words are configured for the slave unit and the master unit. For example, the wake-up word of the slave unit is "Little A Little A", and the wake-up word of the master unit is "Big A Big A", so as to distinguish whether the user intends to perform voice control on the slave unit or the master unit.

[0093] In the embodiment of the present invention, when the slave unit is operating in the first mode, the master unit is used to assist the slave unit in voice control, so as to avoid the situation that the slave unit does not respond to the user's voice instruction. That is, the master unit uploads the collected wake-up voice to the terminal server, and the terminal server then forwards the wake-up voice related to the slave unit to the slave unit, so as to achieve the purpose of supplementing the wake-up information that cannot be collected by the slave unit.

[0094] It should be noted that, in this embodiment, the master unit is configured to be able to collect the local wake-up voice and also collect the wake-up voice for the slave unit, and can recognize the wake-up voice to determine whether the wake-up voice is for waking up the local end or the slave unit. If the wake-up voice is for the slave unit, the wake-up voice is uploaded to the terminal server for processing by the terminal server. Alternatively, the master unit can collect the wake-up voice, upload all collected wake-up voices to the terminal server, and the terminal server recognizes whether the wake-up voice is for waking up the master unit or the slave unit.

[0095] Several implementation modes for implementing the embodiment of the present invention are listed below.

[0096] In a possible implementation mode, please refer to Figure 4 , Figure 4The diagram shows an interaction of a voice recognition method for an air conditioner, which includes:

[0097] Step S41: Upload Working Mode. During operation, the slave device uploads its working mode to the terminal server for recording. Optionally, the slave device can upload the mode to the terminal server whenever it is running any mode, or it can upload the first mode only when running the first mode.

[0098] Step S42: Determine if the operating mode is the first mode. After receiving the operating mode of the slave unit, the terminal server determines whether the operating mode of the slave unit is the first mode, which is the operating mode that causes the operating noise to be greater than or equal to a preset threshold. If the operating mode is the first mode, then proceed to step S43. If the operating mode is not the first mode, then interactive control is performed according to the normal interaction with the slave unit and the master unit.

[0099] Step S43: Request the master unit to upload the collected voice information. The voice information includes at least one of wake-up voice and control voice. When the terminal server determines that the slave unit's operating mode is the first mode, it indicates that the slave unit's current operating mode generates a lot of noise, resulting in a low success rate for the slave unit to pick up user voice commands. Therefore, the terminal server requests the master unit to assist in collecting user voice commands and uploading them to the terminal server. This allows the terminal server to forward the voice information or the corresponding control command to the slave unit when it determines that the voice information pertains to slave unit control. In this way, user voice commands can also be transmitted to the slave unit. When the master unit receives the voice information upload request, it will upload all voice information it collects to the terminal server during subsequent operation.

[0100] Optionally, in one possible implementation, when the slave unit is running in the first mode, the air conditioner's voice recognition process may not include steps S41, S42, and S43 mentioned above; that is, steps S41, S42, and S43 do not need to be executed. In other words, in this implementation, if the air conditioner detects that the slave unit is running in the first mode, it activates online voice recognition, thus disabling offline voice recognition. The voice information collected by the master unit, or the voice information collected by both the master and slave units, is uploaded to the terminal server for voice analysis. This means step S44 is executed directly.

[0101] Step S44: The master unit uploads the wake-up voice. During operation, the microphone of the master unit is turned on. If the microphone picks up the wake-up voice, the master unit uploads the picked-up wake-up voice to the terminal server.

[0102] Step S45: Whether the control object corresponding to the wake-up voice is a slave unit. A terminal server identifies the wake-up voice; if the wake-up voice is the wake-up voice of the slave unit, the control object is the slave unit, and if the wake-up voice is the wake-up voice of the master unit, the control object is the master unit. For example, the wake-up word of the slave unit is "Little A Little A", and the wake-up word of the master unit is "Big A Big A". If the voice uttered by the user is "Little A Little A", it means the slave unit is woken up; if the voice uttered by the user is "Big A Big A", it means the master unit is woken up.

[0103] Wherein, the master unit can upload only the wake-up words of the slave unit, or can upload all wake-up words to the terminal server. If the master unit only uploads the wake-up words of the slave unit, the server directly executes step S46 without executing step S45, that is, it is not necessary to determine the control object corresponding to the wake-up voice.

[0104] Step S46: sending adjustment information to the slave unit.

[0105] If the wake-up voice is a wake-up word for waking up the slave unit, adjustment information is sent to the slave unit.

[0106] Since the wake-up word of the slave unit is different from that of the master unit, after the wake-up voice is collected, it can be clarified whether the wake-up object is the slave unit or the master unit according to the control object corresponding to the wake-up voice. When the wake-up object is the slave unit, adjustment information is sent to the slave unit, so that the slave unit adjusts its own operating parameters or its own working mode according to the adjustment information, and then reduces the working noise of the slave unit, so as to monitor the user's subsequent control voice. If the control voice is a control for the slave unit, the control voice is responded to; if the control voice is a control for the master unit, the control voice is not responded to.

[0107] In addition, in order to further improve the control accuracy, the terminal server can also send adjustment information to the slave unit when it is determined that the control object corresponding to the wake-up voice is the slave unit and the working mode uploaded by the slave unit is the first mode. That is, only when the wake-up object is the slave unit and the slave unit does not collect the wake-up voice due to its own large working noise (the scenario where the slave unit is in the first mode), the wake-up voice uploaded by the master unit needs to be forwarded to the slave unit.

[0108] If the control object corresponding to the wake-up voice is the master unit, there is no need to forward the wake-up voice to the slave unit. The slave unit continues to operate according to the current first mode without adjusting operating parameters or switching working modes.

[0109] The adjustment information may be a wake-up voice message, such as forwarding the wake-up voice message collected by the master unit to the slave unit to wake it up. Alternatively, the adjustment information may be at least one of an operating parameter adjustment command and a working mode switching command, such as instructing the slave unit to adjust its own operating parameters to reduce operating noise, or instructing the slave unit to switch its current working mode to a mode with less operating noise, such as a silent mode or a windless mode, etc.

[0110] Step S47: Adjust operating parameters or adjust the working mode. The slave unit receives the adjustment information and responds accordingly. For example, if the adjustment information is a wake-up voice command, the microphone is turned on to pick up the control voice, and the fan speed is reduced to decrease wind noise. If the adjustment information is an operating parameter adjustment command, the operating parameters are adjusted according to the command, such as reducing the fan speed. If the adjustment information is a working mode switching command, the unit switches to the indicated second mode, such as switching to silent mode. In this way, after the slave unit is adjusted, a good environment is provided for collecting the user's control voice, improving the success rate of voice recognition.

[0111] Step S48: Collect control voice and execute the control commands corresponding to the control voice. The slave unit is woken up, and after adjusting the operating parameters or switching the working mode, it turns on the microphone. At this time, the slave unit's microphone collects voice information from the environment. In a relatively quiet environment, the slave unit can accurately collect the user's voice information and respond in a timely manner.

[0112] Optionally, as a modern intelligent appliance, voice recognition is an important function of a dual-unit air conditioner. However, the implementation of voice control in dual-unit air conditioners still has the following problems: the slave unit may be far from the user during operation, and when the user controls the slave unit by voice, the slave unit may not be able to pick up the voice command, or the picked-up voice volume may be too low to be recognized, resulting in voice recognition failure of the slave unit. Based on this, this embodiment of the invention proposes another embodiment of a voice recognition method, please refer to... Figure 5 , Figure 5 The diagram shows an interaction of a voice recognition method for an air conditioner, which includes:

[0113] Step S51: Upload working location. During operation, the slave device uploads its working location to the terminal server in real time or periodically, and the terminal server records its working location.

[0114] Step S52: Upload user location. The host machine is equipped with an infrared detector or radar detector to detect the user's location. The user's location is then uploaded to the terminal server, where it is recorded.

[0115] Step S53: Determine whether the distance between the working position of the submachine and the user position is greater than or equal to the preset distance.

[0116] During the operation of the slave unit, the terminal server determines in real-time or periodically whether the distance between the slave unit's working position and the user's position is greater than or equal to a preset distance. If so, it combines the data from the master unit for auxiliary speech recognition. Specifically, when the distance between the slave unit's working position and the user's position is greater than or equal to the preset distance, the success rate of the slave unit in picking up user speech information is considered lower than a preset value. Therefore, to improve the slave unit's speech recognition function in this scenario, speech information collected by the master unit is used to assist in speech recognition. The method of using speech information collected by the master unit to assist the slave unit in speech recognition is described in steps S54 to S57.

[0117] In other words, this embodiment is relative to the above. Figure 3 The difference in implementation lies in the fact that this embodiment determines whether to use the voice data collected by the master unit to assist the slave unit in voice recognition based on the distance between the user and the slave unit. The aforementioned... Figure 3 The embodiment shown determines whether to use the voice data collected by the master unit to assist the slave unit in speech recognition based on whether the slave unit's working mode is the first mode.

[0118] Step S54: Request the master unit to upload the collected voice information. The voice information includes at least one of wake-up voice and control voice. When the terminal server determines that the slave unit's operating mode is the first mode, it indicates that the noise generated by the slave unit's current operating mode is very high, and the success rate of the slave unit picking up user voice commands is low. Therefore, the terminal server requests the master unit to assist in collecting user voice commands and uploading them to the terminal server. This allows the terminal server to forward the voice information or the corresponding control command to the slave unit when it determines that the voice information is related to slave unit control. In this way, user voice commands can also be transmitted to the slave unit. When the master unit receives the voice information upload request, it will upload all voice information collected during subsequent operation to the terminal server.

[0119] Step S55: The master unit uploads control voice. During operation, the microphone of the master unit is turned on. If the microphone picks up control voice, the master unit uploads the picked-up wake-up voice to the terminal server.

[0120] Step S56: Send the control command corresponding to the control voice to the slave unit.

[0121] In one possible implementation, the master unit only uploads control voices for the slave unit (the master unit's own control voices are recognized offline). When the terminal server receives the control voices uploaded by the master unit, it recognizes them as the slave unit's control voices. At this point, the terminal server sends the control command corresponding to the control voice to the slave unit, and the slave unit responds to the control command.

[0122] In another possible implementation, the master unit uploads all collected control voice messages to the terminal server for recognition. During the voice recognition process, the terminal server determines the control object corresponding to the control voice message. If the control object is the slave unit, the server sends a control command corresponding to the control voice message to the slave unit, which responds to the command and executes the corresponding control operation. If the control object is the master unit, the server returns the control command corresponding to the control voice message to the master unit, which responds to the command and executes the corresponding control operation (the process of the master unit recognizing voice messages online).

[0123] In some embodiments, the terminal server can determine whether the distance between the working position of the slave device and the user's position is greater than or equal to a preset distance after the master device uploads the voice information. When the control object corresponding to the control voice is the slave device, and the distance between the working position of the slave device and the user's position is greater than or equal to the preset distance, adjustment information is sent to the slave device. In other words, it is only necessary to forward the control voice uploaded by the master device to the slave device in scenarios where the slave device is far from the user, so that the slave device can execute the user's voice control commands.

[0124] It is understandable that the relationship between the working position of the slave unit and the user's position is determined after the master unit uploads the voice information. The master unit can more accurately locate the user's specific position (the user's position is equivalent to the sound source position) through the voice information emitted by the user. In this way, the accuracy of the user's position uploaded by the master unit is higher, which further improves the voice recognition function.

[0125] Step S57: The slave unit executes the control operation corresponding to the control command.

[0126] Compared to the above implementation, this method eliminates the need for the server to determine whether the slave machine is operating in the first mode. The server simply sends the control commands uploaded from the master machine to the slave machine, enabling it to execute the corresponding control operations. Even if the slave machine does not collect voice information in the first mode, it can still execute the control commands corresponding to that voice information. Furthermore, the slave machine does not need to adjust its operating parameters and can maintain operation in the first mode. However, this requires a stable network environment, and the forwarding of control voice commands from the master machine to the slave machine results in a delay in control command response and untimely reaction.

[0127] Understandably, in this implementation, if the server receives voice information uploaded by both the slave and master machines simultaneously, and the control object and control command corresponding to the voice information are the same, then it only needs to return a control command to either the slave or master machine.

[0128] In one optional embodiment, after the terminal server uploads voice information from the master unit, if the voice information is a wake-up voice rather than a control voice, it can send adjustment information to the slave unit. This adjustment information includes adjusting the slave unit's working position, such as controlling the slave unit to move towards the user's location until the distance between the slave unit's working position and the user's location is less than a preset distance. Thus, when the slave unit moves to this position, subsequent control voice commands from the user can be picked up by the slave unit with a high recognition success rate.

[0129] In some embodiments, the terminal server can further determine whether the distance between the working position of the slave device and the user's position is greater than or equal to a preset distance after the master device uploads the voice information. If the voice information is a wake-up voice and the distance between the working position of the slave device and the user's position is greater than or equal to the preset distance, adjustment information is sent to the slave device. In other words, the wake-up voice uploaded by the master device is only forwarded to the slave device when the distance between the slave device and the user is relatively far. This also further improves the voice recognition function.

[0130] Optionally, in some possible implementations, the voice recognition method of the air conditioner can also employ a master unit to assist the slave unit in voice recognition when the slave unit is operating in a first mode and the distance between the slave unit's working position and the user's position is detected to be greater than or equal to a preset distance. That is, this embodiment combines the above embodiments to obtain an implementation with higher voice recognition accuracy.

[0131] Alternatively, other possible implementations in this embodiment can refer to the various implementations in the first embodiment.

[0132] In this embodiment, when the slave unit is operating at a distance from the user, it is controlled by combining the voice information collected by the master unit. That is, the voice collected by the master unit is transferred to the slave unit, and the slave unit executes the corresponding control operation. If the slave unit does not pick up the voice control command, the master unit forwards it. This allows the slave unit to respond to the user's voice control command in a timely manner even when no voice control command is collected, thereby improving the voice control effect of the air conditioner.

[0133] Optionally, this application embodiment also provides a computer program product, which includes speech recognition program code. When the speech recognition program code is executed by the processor of a computer or other device, it implements the above embodiments.

[0134] It should be noted that the above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A voice recognition method for an air conditioner, characterized in that, Applied to a terminal server that communicates with an air conditioner, the air conditioner comprising a main unit and a movable slave unit detachable from the main unit, the method includes: When the terminal server receives a working mode of the slave machine uploaded by the slave machine, it sends a voice information upload request to the master machine, so that the master machine collects the voice information and uploads the voice information to the terminal server. The first mode is a mode in which the working noise is greater than or equal to a preset threshold. The terminal server receives the voice information uploaded by the host machine; Identify the type of the voice information; When the voice information is identified as a wake-up voice, adjustment information is sent to the sub-machine, so that the sub-machine adjusts its own operating parameters or adjusts its own working mode according to the adjustment information, thereby reducing the operating noise of the sub-machine. The step of sending adjustment information to the slave device when the type of the voice information is identified as a wake-up voice includes: When the voice information is identified as a wake-up voice and the working mode uploaded by the slave device is the first mode, adjustment information is sent to the slave device.

2. The voice recognition method for an air conditioner as described in claim 1, characterized in that, The method further includes: When the voice information is identified as control voice, the control object corresponding to the control voice is determined. If the controlled object is the sub-machine, then a control command corresponding to the control voice is sent to the sub-machine, causing the sub-machine to execute the control operation corresponding to the control command.

3. The voice recognition method for an air conditioner as described in claim 1, characterized in that, The step of sending adjustment information to the slave device when the type of the voice information is identified as a wake-up voice includes: When the voice information is identified as a wake-up voice, the control object corresponding to the wake-up voice is determined. The wake-up voice of the slave unit is different from that of the master unit. If the controlled object is a submachine, then adjustment information is sent to the submachine.

4. The voice recognition method for an air conditioner as described in any one of claims 1-3, characterized in that, Before the step of the terminal server receiving the voice information uploaded by the host machine, the method further includes: When the distance between the working position of the slave unit and the user's position is greater than or equal to a preset distance, a voice information upload request is sent to the master unit, so that the master unit collects the voice information and uploads it to the terminal server.

5. The voice recognition method for an air conditioner as described in claim 1, characterized in that, The adjustment information includes at least one of the following: wake-up voice, operating parameter adjustment command, and working mode switching command.

6. A voice recognition method for an air conditioner, characterized in that, The method, applied to the sub-unit of an air conditioner, includes: The system receives adjustment information from a terminal server. This adjustment information is generated when the terminal server receives voice information uploaded by the host machine, and the voice information is of the wake-up voice type. Specifically, when the terminal server receives the voice information uploaded by the slave machine in the first working mode, it sends a voice information upload request to the host machine, causing the host machine to collect the voice information and upload it to the terminal server. The first mode is a mode where the operating noise is greater than or equal to a preset threshold. The terminal server receives the voice information uploaded by the host machine; identifies the type of the voice information; and when it identifies the voice information as a wake-up voice and receives the voice information uploaded by the slave machine in the first working mode, it sends adjustment information to the slave machine. Based on the adjustment information, adjust the operating parameters or the working mode to reduce operating noise.

7. The voice recognition method for an air conditioner as described in claim 6, characterized in that, The steps for adjusting operating parameters include: Reduce the fan speed of the sub-unit; Alternatively, the steps to adjust the work mode include: Switch the working mode of the slave unit to the second mode, which is a mode that makes the working noise less than a preset threshold.

8. The voice recognition method for an air conditioner as described in claim 6, characterized in that, Following the step of adjusting operating parameters or adjusting the working mode to reduce operating noise based on the aforementioned adjustment information, the method further includes: When a control voice is detected, the control command corresponding to the control voice is executed.

9. A terminal server, characterized in that, The terminal server includes a memory, a processor, and a voice recognition program stored in the memory and executable on the processor. When executed by the processor, the voice recognition program implements the steps of the voice recognition method for an air conditioner as described in any one of claims 1 to 5.

10. An air conditioner, characterized in that, The air conditioner includes a memory, a processor, and a voice recognition program stored in the memory and executable on the processor, wherein the voice recognition program, when executed by the processor, implements the steps of the voice recognition method for the air conditioner as described in any one of claims 6 to 8.

11. A computer-readable storage medium, characterized in that, The readable storage medium stores a speech recognition program, which, when executed by a processor, implements the steps of the speech recognition method for an air conditioner as described in any one of claims 1 to 5, or implements the steps of the speech recognition method for an air conditioner as described in any one of claims 6 to 8.

Images