Voice control method and system for bathroom equipment and electronic equipment

Abstract

The invention relates to the technical field of equipment control, and discloses a voice control method and system for bathroom equipment and electronic equipment, and the method comprises the steps: recognizing a current identity label of a user according to the biological characteristics of the user; acquiring voice information input by a user, and sending the voice information to the artificial intelligence large model; performing semantic reasoning based on the voice information and historical interaction data bound with the current identity label through an artificial intelligence large model to obtain a user intention, and generating a control instruction according to the user intention; according to the method, the dynamic intention is reasoned and the user instruction is recognized and analyzed by adopting the context semantics of the artificial intelligence large model, so that the response of the bathroom equipment better fits the use habit of a user, the limitation of a traditional fixed instruction is broken through, and the user experience is improved. More complex and more flexible user voice requirements can be understood, and the user experience is improved.

Description

Technical Field

[0001] This invention relates to the field of equipment control technology, specifically to a voice control method, system, and electronic device for bathroom equipment. Background Technology

[0002] With the advancement of digitalization and intelligentization, bathroom spaces have gradually upgraded from traditional functional areas to smart living scenarios, and smart bathroom equipment has become an important direction for home furnishing consumption upgrades. Among them, smart bathroom products, with smart toilets at their core, offer basic intelligent functions such as heated seats, warm water washing, and automatic flushing.

[0003] Currently, existing smart bathroom devices have initially integrated technologies such as voice control, sensor sensing, and remote operation via APP, aiming to enhance user convenience and comfort through technological empowerment. However, the intelligence of existing voice control methods needs improvement. In continuous interaction scenarios, existing systems lack the necessary contextual memory and associative capabilities, leading to frequent interruptions in the interaction process. For example, when a user first asks "What water temperature is suitable?" and then follows up with the instruction "Set according to the recommended value," the system cannot associate the two instructions and cannot accurately identify the core demand corresponding to the "recommended value." The user needs to completely rephrase their needs, greatly reducing interaction efficiency and user experience. Summary of the Invention

[0004] This invention provides a voice control method, system, and electronic device for bathroom equipment, in order to at least solve the problems of existing voice control systems being not intelligent enough and having a poor user experience.

[0005] In a first aspect, the present invention provides a voice control method for bathroom fixtures, the method comprising: Identify the user's current identity based on their biometric characteristics; It acquires the user's voice input and sends the voice information to the large artificial intelligence model; The system uses a large AI model to perform semantic reasoning based on voice information and historical interaction data bound to the current identity to obtain user intent and generate control commands based on the user intent. Control commands are sent to bathroom fixtures to cause them to perform corresponding actions.

[0006] In one alternative implementation, the historical interaction data includes pre-cached multi-turn historical dialogues; And / or, using large-scale artificial intelligence models to perform semantic reasoning based on voice information and historical interaction data bound to the current identity to obtain user intent, including: Convert speech information into text information; By using the state machine management mechanism of a large artificial intelligence model, user intent can be obtained by parsing text information and historical interaction data.

[0007] In one alternative implementation, after generating control instructions based on user intent, the process includes: The system detects whether a preset negative word exists in the text information. If a preset negative word exists, the execution chain in the control command is reversed.

[0008] In one optional implementation, after acquiring the user-input voice information, the method further includes: Extracting prosodic features from speech information; The system utilizes a large artificial intelligence model to analyze users' emotional categories based on prosodic features. Switch between command interaction modes based on emotion category; And / or, after converting voice information into text information, the method further includes: generating a response text based on the text information and the current instruction interaction mode; The response text is converted into a response voice signal by a speech synthesis engine and then played through a voice terminal.

[0009] In one alternative implementation, converting speech information into text information includes: The speech information is converted into text information through a multimodal dialect recognition engine; After generating control commands, the following is also included: Generate response text corresponding to control commands using a built-in dialect and slang database; The response text is converted into a response voice signal by a speech synthesis engine and then played through a voice terminal.

[0010] In one alternative implementation, after identifying the user's current identity based on the user's biometrics, the method further includes: Obtain the health baseline corresponding to the current identity identifier. The health baseline is generated by transfer learning using the user's historical physiological indicators. Collect users' real-time physiological indicators and preset environmental indicators, compare the real-time physiological indicators with the health baseline, and / or compare the preset environmental indicators with the pre-set environmental standards, and push health suggestions or adjust the operating parameters of bathroom equipment based on the comparison results.

[0011] In one alternative implementation, after identifying the user's current identity based on the user's biometrics, the method further includes: Obtain user settings preferences based on the current identity identifier; Obtain the operating parameters of bathroom equipment based on user settings preferences.

[0012] In one alternative implementation, before sending the voice information to the large artificial intelligence model, the following steps are included: The system detects the network status of the bathroom fixtures. If the network connection fails, it matches the voice information with the locally cached control commands and controls the bathroom fixtures to perform corresponding actions based on the matched control commands.

[0013] Secondly, the present invention provides a voice control system for a bathroom fixture, comprising: The identity recognition module is used to identify the user's current identity based on the user's biometric characteristics; The voice acquisition module is used to acquire the voice information input by the user and send the voice information to the large artificial intelligence model; The instruction generation module is used to obtain user intent by semantic reasoning based on voice information and historical interaction data bound to the current identity through a large artificial intelligence model, and to generate control instructions based on the user intent. The instruction sending module is used to send control instructions to the bathroom equipment so that the bathroom equipment can perform corresponding actions.

[0014] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the voice control method of the bathroom device described in the first aspect or any corresponding embodiment thereof.

[0015] The present invention has the following beneficial effects: The voice control method for bathroom equipment of the present invention identifies the user's current identity based on the user's biometrics, obtains the user's input voice information, and sends the voice information to an artificial intelligence (AI) big data model. The AI ​​big data model performs semantic reasoning based on the voice information and historical interaction data bound to the current identity to obtain the user's intent, and generates control commands based on the user's intent. This constructs a full-link system of voice acquisition, semantic parsing, big data model reasoning, and equipment control. By using the contextual semantic reasoning of the AI ​​big data model to dynamically infer intent, the system identifies and parses user commands, making the response of the bathroom equipment more in line with the user's usage habits. This breaks through the limitations of traditional fixed commands, can understand more complex and flexible user voice needs, and improves the user experience. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a voice control method for a bathroom device according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the architecture of a voice control method for a bathroom device according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating a voice control method for another bathroom device according to an embodiment of the present invention. Figure 4 This is a schematic diagram of the dialect speech processing according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the user personalization settings process according to an embodiment of the present invention; Figure 6 This is a structural block diagram of the voice control system of a bathroom device according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0020] Most current smart bathroom devices, such as smart toilets and water heaters, generally adopt an operating mode that combines a multi-button physical panel with a lengthy APP setup process. This results in complex operating logic and a high learning curve. For middle-aged and elderly people, the densely packed button icons and cumbersome APP setup steps are difficult to master, preventing them from fully utilizing the smart functions of the devices.

[0021] In existing voice control solutions, reliance on traditional rule engines or simple speech recognition modules results in a severe lack of flexibility and intelligence in voice interaction. In continuous interaction scenarios, existing systems lack the necessary contextual memory and associative capabilities, leading to frequent interruptions in the interaction process.

[0022] In addition, existing solutions only provide unified general parameter configurations and do not establish a user profile system, making it impossible to accurately adapt to differences in user identity.

[0023] In view of this, embodiments of the present invention propose a voice control method for bathroom fixtures. This voice control method can be executed via a terminal device or a server. Specifically, the terminal device can be a smartphone, tablet computer, laptop computer, PDA, or desktop computer, etc. The server can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services.

[0024] This invention provides a voice control method for bathroom equipment, constructing a full-link system of voice acquisition, semantic parsing, GPT inference, and equipment control. It adopts dynamic intent recognition and parsing of user commands using a large artificial intelligence model, and combines the state machine management mechanism of the large artificial intelligence model to realize multi-turn dialogue, thereby improving the intelligence of voice control and enhancing the user experience.

[0025] According to an embodiment of the present invention, a voice control method for a bathroom device is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0026] This embodiment provides a voice control method for bathroom equipment, such as Figure 1 As shown, the process includes the following steps: Step S100: Identify the user's current identity based on the user's biometric characteristics.

[0027] Specifically, near-infrared imaging modules, millimeter-wave radar, fingerprint recognition modules, and iris recognition modules can be integrated into bathroom fixtures to identify user biometrics and obtain the user's current identity. Each identity and its corresponding biometric feature are pre-stored in a database. The current user's biometric features are matched against those in the database to obtain the current identity. If the current user's biometric features are not in the database, a new identity is created and bound to the input biometric features in the database.

[0028] Each identity is linked to a corresponding user profile. For example, if the user type is elderly, child, or pregnant, the water pressure, toilet seat temperature, water temperature, and other parameters of the bathroom equipment can be automatically adjusted based on the user's current identity, without the need for manual settings.

[0029] In some embodiments, near-infrared imaging technology is used to penetrate the surface skin of the palm, capture the distribution pattern and blood flow signals of subcutaneous veins, and generate an infrared image of the user's palm.

[0030] By eliminating ambient light interference in near-infrared images through Gaussian filtering and combining it with adaptive threshold segmentation technology, the background of the palm skin and the subcutaneous vein area are separated, and irrelevant noise such as skin texture and wrinkles is removed to ensure clear vein contours.

[0031] Based on the branching points, endpoints, direction, and texture density of palm veins in infrared images, a vascular topology vector is constructed. Dynamic features of the palm veins are constructed based on blood flow signal fluctuations in several consecutive frames of infrared images. The live features of the palm veins include both static and dynamic features. Static features include the vascular topology vector constructed based on the branching points, endpoints, direction, and texture density, while dynamic features are obtained by analyzing minute fluctuations in blood flow signals across multiple image sequences. By verifying the blood flow characteristics of living blood vessels, attacks using forged palm vein images or models are prevented.

[0032] A multi-dimensional palm vein feature matrix is ​​formed by integrating vascular topology vectors and palm vein dynamic features. The multi-dimensional palm vein feature matrix of the current user is compared with the templates stored in the database. A combination of three methods is used: Manhattan distance calculation, feature point similarity matching, and blood flow signal consistency verification. The identity identifier bound to the user template with the highest similarity is used as the current identity identifier.

[0033] Only the extracted multi-dimensional palm vein feature matrix is ​​stored, without retaining the original palm vein image or complete biological data, which meets the requirements for protecting user data privacy. Employing palm vein infrared image recognition, and combining vascular topology vectors and dynamic blood flow features to construct a multi-dimensional feature matrix, it offers stronger resistance to forgery compared to single static biometric features, reducing false or missed identifications, and is suitable for identifying multiple family members.

[0034] Step S200: Obtain the voice information input by the user and send the voice information to the artificial intelligence big data model.

[0035] Specifically, the bathroom fixtures integrate an IP65-protected dual-microphone array to collect user-input voice information. Users input voice information through the microphones, such as "open the toilet lid" or "adjust the seat temperature to level 5." The microphones collect simulated voice signals from the environment, completing the signal input for the "starting point of human-computer interaction." The bathroom fixtures then send the voice information to an AI model. The AI ​​model uses a speech recognition algorithm to obtain the corresponding text information. Alternatively, the bathroom fixtures combine adaptive noise reduction algorithms and Automatic Speech Recognition (ASR) algorithms to convert the voice information into text information and send the text information to the AI ​​model. Subsequent reasoning processes of the AI ​​model are based on the text information.

[0036] The AI ​​big data model is a large language model (LLM), including but not limited to GPT, Wenxin Yiyan, and Tongyi Qianwen. The AI ​​big data model is deployed in the cloud, and bathroom appliances send text information to the AI ​​big data model through gateway devices.

[0037] Step S300: The user intent is obtained by semantic reasoning based on voice information and historical interaction data bound to the current identity through a large artificial intelligence model, and control instructions are generated according to the user intent.

[0038] Specifically, in combination Figure 2 and Figure 3 As shown, the AI ​​big data model uses the GPT model, which pre-stores historical interaction data bound to various identities from several days ago, such as historical interaction data within a 7-day or 14-day period. Upon receiving the currently input text information, the AI ​​big data model combines the historical interaction data to generate control commands that meet the user's needs. For example, if the user first asks "What water temperature is suitable?" in the historical interaction data, the AI ​​big data model recommends 36°C. If the current user inputs "Set the water temperature according to the recommended value," then the AI ​​big data model, based on the current user input text information and historical interaction data, combined with contextual reasoning or its own reasoning, derives the core intent to set the smart toilet's flushing water temperature to 36°C, and converts this core intent into machine-recognizable control commands, such as "Device ID: Smart Toilet_001 | Command: Seat Temperature | Parameter: Room Temperature," thus improving interaction efficiency and user experience.

[0039] It should be understood that after recognizing the user's control command, a status verification is required. Specifically, this involves querying the real-time status of bathroom devices on the IoT platform. For example, if the control command is to turn on the bidet mode, it is necessary to verify whether the bidet mode of the smart toilet is already turned on to avoid repeatedly executing invalid commands.

[0040] Furthermore, a dangerous command interception setting is added to the cloud, allowing users to set dangerous temperature, dangerous water pressure, and other parameters. When the parameters in a user's control command exceed the dangerous value, the corresponding control command will be intercepted, preventing the machine from executing normally after the user accidentally inputs a dangerous command, thus improving equipment safety.

[0041] Step S400: Send control commands to the bathroom fixtures so that the bathroom fixtures can perform corresponding actions.

[0042] Specifically, after the cloud-based AI model generates control commands, it locates the network address of the bathroom device, such as Wi-Fi IP or Bluetooth MAC address. Using a pre-set communication protocol, such as MQTT, LoRa, or Zigbee, it transmits the standardized control commands to the bathroom device, which then executes the commands and provides feedback. For example, upon receiving a control command to turn on the night light, the bathroom device drives a relay to close, energizing the smart toilet night light.

[0043] Furthermore, after the bathroom equipment completes its operation, it will broadcast the result, such as "the water temperature has been adjusted to the recommended value," to the user via a voice terminal, thus completing the closed loop.

[0044] It should be understood that after the bathroom equipment performs the corresponding action, it transmits the execution structure back to the cloud through the original communication protocol, and updates the equipment status in the cloud.

[0045] The voice control method for bathroom equipment in this invention identifies the user's current identity based on their biometrics, acquires the user's input voice information, and sends the voice information to an artificial intelligence (AI) big data model. The AI ​​big data model then performs contextual semantic reasoning based on the voice information and historical interaction data bound to the current identity to obtain the user's intent. Control commands are generated based on the user's intent, constructing a full-link system of voice acquisition, semantic parsing, big data model reasoning, and equipment control. By employing dynamic intent reasoning based on the contextual semantics of the AI ​​big data model to identify and parse user commands, the response of the bathroom equipment becomes more aligned with user habits, breaking through the limitations of traditional fixed commands. This allows the system to understand more complex and flexible user voice needs, thus improving the user experience.

[0046] In some embodiments, historical interaction data includes pre-cached multi-turn historical dialogues. Further, step S300 involves obtaining user intent through semantic reasoning using a large artificial intelligence model based on voice information and historical interaction data bound to the current identity, including: Step S301: Convert the voice information into text information.

[0047] Step S302: The user intent is obtained by parsing the text information and historical interaction data through the state machine management mechanism of the artificial intelligence big data model.

[0048] Specifically, bathroom fixtures or large-scale artificial intelligence models combine adaptive noise reduction algorithms and automatic speech recognition algorithms to convert speech information into text information.

[0049] For each user who has used the bathroom equipment, their identity is pre-stored and their historical interaction data with the bathroom equipment is bound. The historical interaction data includes, but is not limited to, multiple rounds of historical dialogues from 10 minutes ago, several hours ago, or several days ago. By caching the historical interaction data and using the state machine management mechanism of the artificial intelligence large model, the text information and multiple rounds of historical dialogues are parsed to obtain the user's intent, which can realize multi-round dialogue, improve the intelligence of voice control, and enhance the user experience.

[0050] Furthermore, after generating control instructions based on user intent, the process includes: The system detects whether a preset negative word exists in the text information. If a preset negative word exists, the execution chain in the control command is reversed.

[0051] Specifically, the AI ​​big data model introduces negative semantic capture. It scans text information through string matching algorithms (such as the KMP algorithm) and combines semantic analysis to determine the negative object. When it detects that the text information includes pre-set negative semantic words, such as "don't turn on the night light" or "it's too bright", the AI ​​big data model immediately reverses the execution chain, turns off the night light, and records it in the "user error correction library". In the next similar scenario, the control instructions recorded in the "user error correction library" will be adopted first.

[0052] By automatically recognizing negative intents and avoiding accidental execution, the execution chain reversal logic adapts to colloquial negative expressions, enhancing the natural interactive experience.

[0053] In some embodiments, after obtaining the user-inputted voice information in step S200, the method further includes: extracting prosodic features of the voice information; calling an artificial intelligence big data model to analyze the user's emotion category based on the prosodic features; and switching the instruction interaction mode according to the emotion category.

[0054] And / or, after converting the voice information into text information, it further includes: generating a response text based on the text information and the current instruction interaction mode; converting the response text into a response voice signal through a speech synthesis engine and then playing it through a voice terminal.

[0055] The prosodic features of speech information include intonation and speech rate. The AI ​​model uses these features to determine the user's emotional category, such as irritability or pleasure. For example, if both the intonation and speech rate exceed a threshold, the user's emotional category is determined to be irritability; otherwise, it is determined to be pleasure.

[0056] The command interaction mode includes a concise command mode and a normal command mode. When the user's emotion category is irritable, the command interaction mode is switched to concise command mode. If the user's emotion category is pleasant, the command interaction mode is normal command mode.

[0057] The AI ​​big data model determines the strategy for interacting with users based on the command interaction mode. In the normal command mode, the response strategy adopts a friendly or user-preset tone to interact with the user, while in the concise command mode, a concise response strategy is adopted.

[0058] The AI ​​model outputs response text based on the current command interaction mode. The response text differs depending on the command interaction mode. In particular, the response text output in the concise command mode is shorter than that output in the normal command mode, in order to avoid replying with too much content and causing user annoyance when the user is in a bad mood.

[0059] This invention analyzes user emotions by extracting speech prosody features, enabling the interaction mode of bathroom equipment to dynamically switch according to the user's emotions, thereby improving the humanization of the interaction.

[0060] In some embodiments, step S301, converting voice information into text information, includes: The speech information is converted into text information through a multimodal dialect recognition engine; After generating control commands, the following is also included: Generate response text corresponding to control commands using a built-in dialect and slang database; The response text is converted into a response voice signal by a speech synthesis engine and then played through a voice terminal.

[0061] Specifically, the bathroom fixtures integrate a dual-microphone array. Users input voice information through the microphones, which collect analog voice signals from the environment, thus completing the signal input for the "starting point of human-computer interaction." Afterwards, an A / D converter converts the collected analog voice signal into a digital voice signal for subsequent algorithm processing.

[0062] Audio noise reduction algorithms such as spectral subtraction and wavelet transform are used to filter out environmental noise like television sounds and footsteps, preserving clear user speech signals. Then, a multimodal dialect recognition engine converts the denoised digital speech signal into text.

[0063] The multimodal dialect recognition engine supports free mixing of Mandarin and multiple dialects, employing a pre-training and fine-tuning architecture. Pre-training includes unsupervised pre-training and supervised fine-tuning. Unsupervised pre-training utilizes 700,000 hours of unlabeled dialect data, learning pronunciation rules through a "teacher-student" network structure. Supervised fine-tuning mixes labeled data from multiple dialects, using a "character + dialect label" modeling unit to address the issue of different pronunciations for the same character.

[0064] Design a microphone array beamforming algorithm to optimize the signal-to-noise ratio for the reverberant environment in the bathroom. Combine the Mel cepstral coefficients and the unique tone features of dialects to construct a multi-dimensional acoustic feature space and achieve acoustic feature enhancement.

[0065] After the dialect speech is recognized as text by the multi-modal dialect recognition engine, it is mapped to the system instruction set through a dynamic intent classifier. For example, when the user says "Turn up the water temperature" in Sichuan dialect, it can be automatically parsed as "Adjust the flushing water temperature to 40°C. Combine the identity recognition result and generate a composite response through the personalized knowledge base.

[0066] The dialect slang library includes the correspondence between slang in dialects such as Cantonese and Wenzhou dialect and Mandarin. For example, the Cantonese phrase "得闲饮茶" corresponds to "有空来喝茶". Generate natural responses that conform to the regional culture through an artificial intelligence large model.

[0067] Such as Figure 4 As shown, the working process of the multi-modal dialect recognition engine is as follows: First, receive the user's dialect speech input, and then perform acoustic feature enhancement processing on it to optimize the speech quality; Subsequently, determine whether the speech is a dialect. If it is, perform dialect recognition and text conversion. If not, perform standard speech recognition processing. The recognition results of both paths will enter the dialect culture adaptation processing link. After optimizing the text semantics in combination with the regional cultural background, generate a response that conforms to the regional culture, and then maintain the coherence of the interaction through multi-round dialogue management. Finally, the process ends.

[0068] The embodiment of the present invention adopts a multi-modal dialect recognition engine, which supports the input of dialect speech in different regions, breaks through the limitation of Mandarin recognition, meets the usage needs of non-Mandarin users, and expands the audience range of the product. The built-in dialect slang library generates reply texts, which are fed back in dialect form after speech synthesis, making the interaction closer to the user's language habits and enhancing the sense of intimacy and convenience of dialect users.

[0069] In some embodiments, after identifying the current identity identifier of the user according to the user's biometric characteristics in step S100, it further includes: Step a1, obtain the health baseline corresponding to the current identity identifier, where the health baseline is generated by using the historical physiological indicators of the user for transfer learning.

[0070] Step a2, collect the user's real-time physiological indicators and preset environmental indicators, compare the real-time physiological indicators with the health baseline, and / or compare the preset environmental indicators with the pre-set environmental standards, and push health suggestions or adjust the operating parameters of the bathroom equipment based on the comparison results.

[0071] Specifically, historical physiological indicators of users are collected, including historical toilet time, heart rate, heart rate variability (HRV), blood pressure trends, leg temperature, etc. These indicators can be obtained through smart bracelets or bathroom-integrated sensors. Transfer learning algorithms, such as a ResNet-based transfer model, are used to train a health baseline, which is updated every month. The range of historical physiological indicators of users is obtained through the health baseline, for example, the health baseline includes normal body temperature values.

[0072] Preset environmental parameters include temperature and humidity. Body temperature is indirectly monitored through sensors integrated into bathroom fixtures, such as water temperature sensors, and blood pressure trends are correlated with seat pressure sensors. Alternatively, third-party health devices can be connected via Bluetooth to collect the user's real-time physiological indicators.

[0073] By comparing real-time physiological indicators with a health baseline, abnormal health fluctuations are identified. Combined with the results of health anomaly detection, personalized suggestions are generated through an artificial intelligence model, such as "Your uric acid level is high, it is recommended to reduce seafood intake." When abnormal toilet time is detected (>15 minutes), a health reminder is pushed through the artificial intelligence model, such as "You have been sitting for too long, it is recommended to get up and move around."

[0074] For example, the real-time physiological indicator is heart rate. The real-time heart rate is compared with the corresponding health baseline. If the difference between the real-time heart rate and the health baseline is greater than a set value, it is determined to be an abnormal heart rate, and health advice is actively pushed to remind the user to pay attention to abnormal heart rate when using the toilet and to seek medical attention in time.

[0075] Alternatively, real-time physiological indicators such as blood pressure can be used. If the difference between real-time blood pressure and a healthy baseline is greater than the set value, it indicates that the user needs a higher temperature. In this case, the device parameters can be adjusted to automatically lower the water temperature to below 38°C, providing the user with a suitable water temperature.

[0076] In addition, environmental sensors can be set to collect preset environmental indicators, including temperature and humidity sensors and air quality monitoring modules. For example, when the indoor humidity is detected to be lower than the preset humidity standard, it indicates that the indoor environment is relatively dry, and a suggestion is generated: "The moisturizing mode has been activated for you, and we also recommend drinking warm water," and the toilet seat heating power is automatically adjusted. When it is detected that the current time is night, the slow closing speed of the smart toilet seat is reduced, and the prompt sound is turned off.

[0077] In this embodiment of the invention, personalized health management is achieved by combining a health baseline with real-time physiological indicators. Simultaneously, bathroom equipment can act as a health assistant for users, proactively pushing health suggestions. By detecting preset environmental indicators in the bathroom, more comfortable equipment operating parameters can be provided to users, improving safety and comfort during use.

[0078] In some embodiments, after identifying the user's current identity based on the user's biometrics in step S100, the method further includes: Step b1: Obtain user settings preferences based on the current identity identifier.

[0079] Step b2: Obtain the operating parameters of the bathroom equipment according to the user's settings preferences.

[0080] Specifically, user preferences include personalized settings for bathroom fixtures, such as setting the bidet water temperature to 39℃-42℃, water pressure levels, seat heating levels, and automatic lid-opening settings. These preferences can be set by the user or obtained by recording their historical usage habits.

[0081] Upon detecting a human approaching, the bathroom fixtures open the toilet lid and control the operating parameters of the equipment according to the user's settings preferences based on the current identity, such as the water pressure setting and the seat heating setting. For example, when user A is detected, the seat temperature is automatically adjusted to 38°C and the flushing water pressure is set to level 5; when user B approaches, the temperature is switched to 35°C and the water pressure is set to level 3.

[0082] Combination Figure 5 As shown in the example, personalized decisions can be generated from multi-source data such as toilet time and heart rate to obtain the operating parameters of bathroom equipment, and a set of control instructions can be generated to control the operation of the equipment. Based on the execution results, it is determined whether feedback learning is needed. If feedback learning is needed, the user profile is updated, including updating the user's preferences and health records.

[0083] In the above solution, device operation is controlled based on user settings preferences, which meets the personalized needs of different users, fully realizes personalized parameter adaptation, and improves user comfort.

[0084] In some embodiments, before sending the voice information to the large artificial intelligence model, the following is included: The system detects the network status of the bathroom fixtures. If the network connection fails, it matches the voice information with the locally cached control commands and controls the bathroom fixtures to perform corresponding actions based on the matched control commands.

[0085] Specifically, an emergency semantic library is pre-configured in the edge gateway. The emergency semantic library contains a variety of commonly used control commands. The text information is matched with the locally cached control commands to obtain the corresponding control commands, ensuring that basic voice control is still supported when the network is down. After the network is connected, the new commands are automatically synchronized to the artificial intelligence big model.

[0086] The embodiments of the present invention can ensure that bathroom equipment can still respond to basic voice commands when the network is disconnected, thus avoiding the inability to use the equipment due to network problems.

[0087] This embodiment also provides a voice control system for a bathroom appliance. This system is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0088] This embodiment provides a voice control system for bathroom equipment, such as Figure 6 As shown, it includes: The identity recognition module 601 is used to identify the user's current identity based on the user's biometric characteristics; The voice acquisition module 602 is used to acquire the voice information input by the user and send the voice information to the artificial intelligence model. The instruction generation module 603 is used to obtain user intent by semantic reasoning based on voice information and historical interaction data bound to the current identity through a large artificial intelligence model, and to generate control instructions based on the user intent. The instruction sending module 604 is used to send control instructions to the bathroom equipment so that the bathroom equipment can perform corresponding actions.

[0089] The voice control system for bathroom equipment in this invention identifies the user's current identity based on their biometrics, acquires the user's input voice information, and sends the voice information to an artificial intelligence (AI) big data model. The AI ​​big data model then performs contextual semantic reasoning based on the voice information and historical interaction data bound to the current identity to obtain the user's intent. Control commands are generated based on the user's intent, constructing a full-link system of voice acquisition, semantic parsing, big data model reasoning, and equipment control. By employing dynamic intent reasoning based on the contextual semantics of the AI ​​big data model to identify and parse user commands, the response of the bathroom equipment becomes more aligned with user habits, breaking through the limitations of traditional fixed commands. It can understand more complex and flexible user voice needs, thus improving the user experience.

[0090] In some alternative implementations, the historical interaction data includes pre-cached multi-turn historical dialogues, and the instruction generation module 603 includes: The speech recognition module is used to convert speech information into text information; The intent parsing module is used to parse text information and multi-turn historical dialogues to obtain user intent through the state machine management mechanism of the large artificial intelligence model.

[0091] In some alternative implementations, the instruction generation module 603 further includes: The negation word recognition module is used to detect whether there is a preset negation word in the text information. If a preset negation word exists, the execution chain in the control instruction is reversed.

[0092] In some alternative implementations, the voice control system for bathroom fixtures also includes: The prosodic feature extraction module is used to extract prosodic features from speech information; The emotion analysis module is used to call upon a large artificial intelligence model to analyze the user's emotion category based on prosodic features; The mode switching module is used to switch the command interaction mode according to the emotion category; The response text acquisition module is used to output the response text based on the text information and the current command interaction mode; The response voice signal acquisition module is used to convert the response text into a response voice signal through a speech synthesis engine and then play it through a voice terminal.

[0093] In some alternative implementations, the speech recognition module is also used to convert speech information into text information through a multimodal dialect recognition engine; The instruction generation module also includes: The dialect response module is used to generate response text corresponding to control commands through a built-in dialect slang database. The response text is then converted into a speech signal by a speech synthesis engine and played through a voice terminal.

[0094] In some alternative implementations, the identity recognition module 601 includes: The health baseline acquisition module is used to acquire the health baseline corresponding to the current identity identifier. The health baseline is generated by using the user's historical physiological indicators for transfer learning. The baseline comparison module is used to collect users' real-time physiological indicators and preset environmental indicators, compare the real-time physiological indicators with the health baseline, and / or compare the preset environmental indicators with the pre-set environmental standards. Based on the comparison results, it pushes health suggestions or adjusts the operating parameters of bathroom equipment.

[0095] In some alternative implementations, the identity recognition module 601 includes: The personalized data acquisition module is used to obtain user settings preferences based on the current identity identifier; The parameter acquisition module is used to obtain the operating parameters of bathroom equipment according to user settings preferences.

[0096] In some alternative implementations, the voice control system for bathroom fixtures also includes: The offline operation module is used to detect the network status of the bathroom equipment. If the network cannot be connected, the voice information is matched with the locally cached control commands, and the bathroom equipment is controlled to perform corresponding actions based on the matched control commands.

[0097] The voice control system for bathroom equipment provided in this embodiment of the invention can execute the voice control method for bathroom equipment provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0098] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0099] The following is a detailed reference. Figure 7 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 701, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 702 or a program loaded from memory 708 into random access memory (RAM) 703. The RAM 703 also stores various programs and data required for the operation of the electronic device. The processor 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0100] Typically, the following devices can be connected to I / O interface 705: input devices 706 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 707 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 708 including, for example, magnetic tapes, hard disks, etc.; and communication devices 709. Communication device 709 allows electronic devices to exchange data via wireless or wired communication with other devices. Although Figure 7 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0101] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 709, or installed from a memory 708, or installed from a ROM 702. When the computer program is executed by the processor 701, it performs the functions defined in the voice control method for bathroom equipment according to embodiments of the present invention.

[0102] Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0103] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A voice control method for bathroom fixtures, characterized in that, include: Identify the user's current identity based on their biometric characteristics; Acquire user-inputted voice information and send the voice information to the large-scale artificial intelligence model; The user intent is obtained by semantic reasoning based on the voice information and historical interaction data bound to the current identity using a large artificial intelligence model, and control instructions are generated according to the user intent. The control command is sent to the bathroom equipment so that the bathroom equipment performs the corresponding action.

2. The method according to claim 1, characterized in that, The historical interaction data includes pre-cached multi-turn historical dialogues; And / or, using a large-scale artificial intelligence model to perform semantic reasoning based on the voice information and historical interaction data bound to the current identity to obtain user intent, including: Convert the voice information into text information; The user intent is obtained by parsing the text information and historical interaction data through the state machine management mechanism of the large artificial intelligence model.

3. The method according to claim 2, characterized in that, After generating control instructions based on the user's intent, the process includes: The system detects whether a preset negative word exists in the text information. If the preset negative word exists, the execution chain in the control instruction is reversed.

4. The method according to claim 2, characterized in that, After obtaining the user's voice input, it also includes: Extract the prosodic features of the speech information; The AI ​​model is invoked to analyze the user's emotion category based on the prosodic features; Switch the command interaction mode according to the emotion category; And / or, after converting the speech information into text information, it further includes: Generate a response text based on the text information and the current instruction interaction mode; The response text is converted into a response voice signal by a speech synthesis engine and then played through a voice terminal.

5. The method according to claim 2, characterized in that, Converting the voice information into text information includes: The speech information is converted into text information using a multimodal dialect recognition engine; After generating control commands, the following is also included: The system generates response text corresponding to the control commands by using a built-in dialect and slang database. The response text is converted into a response voice signal by a speech synthesis engine and then played through a voice terminal.

6. The method according to claim 1, characterized in that, After identifying the user's current identity based on their biometrics, the process also includes: Obtain the health baseline corresponding to the current identity identifier, wherein the health baseline is generated by transfer learning using the user's historical physiological indicators; Collect users' real-time physiological indicators and preset environmental indicators, compare the real-time physiological indicators with the health baseline, and / or compare the preset environmental indicators with pre-set environmental standards, and push health suggestions or adjust the operating parameters of the bathroom equipment based on the comparison results.

7. The method according to claim 1, characterized in that, After identifying the user's current identity based on their biometrics, the process also includes: Obtain user settings preferences based on the current identity identifier; The operating parameters of the bathroom equipment are obtained based on the user's settings preferences.

8. The method according to claim 1, characterized in that, Before sending the voice information to the large-scale artificial intelligence model, the following steps are included: The network status of the bathroom device is detected. If the network cannot be connected, the voice information is matched with the locally cached control commands, and the bathroom device is controlled to perform corresponding actions based on the matched control commands.

9. A voice control system for a bathroom fixture, characterized in that, include: The identity recognition module is used to identify the user's current identity based on the user's biometric characteristics; The voice acquisition module is used to acquire the voice information input by the user and send the voice information to the artificial intelligence big data model; The instruction generation module is used to obtain user intent by performing semantic reasoning based on the voice information and historical interaction data bound to the current identity through an artificial intelligence big data model, and to generate control instructions based on the user intent; The instruction sending module is used to send the control instructions to the bathroom equipment so that the bathroom equipment can perform corresponding actions.

10. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the voice control method for the bathroom device according to any one of claims 1 to 8.

Images