Household intelligent equipment integrated management and linkage system based on AI voice interaction

Abstract

The invention discloses a household intelligent equipment integrated management and linkage system based on AI voice interaction. The system comprises a voice interaction module which collects user voice input and generates a digital voice stream signal; the semantic analysis module receives the digital voice stream signal, and generates a standardized instruction signal containing an equipment object, an execution action and a control parameter through a natural language understanding model; the intelligent decision-making module receives the standardized instruction signal and combines the standardized instruction signal with a real-time environment state signal obtained from the environment sensing module; the environment sensing module collects multi-dimensional sensor data in a home environment in real time and generates a real-time environment state signal; and the equipment linkage execution module receives the multi-equipment cooperative control signal, and generates a control instruction signal corresponding to each intelligent equipment through the heterogeneous equipment protocol adapter. The AI voice interaction-based household intelligent equipment integrated management and linkage system can solve the problems of intelligent household equipment linkage fragmentation, passive response and non-intelligent interaction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart home, in particular to a home smart device integrated management and linkage system based on AI voice interaction. BACKGROUND

[0002] With the rapid development of Internet of Things and artificial intelligence technology, the concept of smart home is increasingly popular, and the types and quantities of smart devices on the market have surged, covering lighting, security, audio-visual, environmental regulation and many other categories. However, the current market situation is that there are many device brands and different communication protocols, such as wireless fidelity, Bluetooth mesh network, Zigbee protocol and other technical standards coexist, which makes it difficult for devices of different brands and protocols to effectively interconnect and interoperate, forming a "data island". Users often need to install multiple independent applications to control different brands of devices, which is cumbersome and fragmented, and seriously violates the original intention of smart home to pursue convenience and unity. Existing solutions mainly rely on smart speakers or single hub devices as control centers, or through application programming interfaces between cloud platforms for interfacing.

[0003] The former is limited by the processing power and coverage of the center node, and once the center node fails, the whole system will be paralyzed; the latter is heavily dependent on Internet connection, and network delay and stability problems directly affect the control experience, and there is a risk of data privacy leakage. In terms of interaction, existing voice assistants can only respond to simple and stereotyped single instructions, and have very limited understanding ability for fuzzy instructions, compound instructions and multi-round dialogues that need to combine environmental context (such as time, sensor state, user historical behavior). The linkage between devices is mostly a simple "if the condition is met, then execute the action" trigger rule set by the user in advance, and the system lacks the ability of autonomous learning and dynamic adjustment, and cannot provide truly intelligent predictive and proactive services according to the changing complex environment or user habits, with low level of intelligence. Therefore, the market urgently needs an integrated home device management and linkage system that can break down device barriers, deeply integrate environmental perception, have advanced artificial intelligence decision-making capabilities, and provide natural, intelligent and proactive services. SUMMARY

[0004] In view of the above prior art defects, the purpose of the present application is to provide an AI voice interaction-based integrated management and linkage system for smart home devices, which is used to solve the problems of smart home device linkage fragmentation, passive response and unintelligent interaction. The present application solves the problem by constructing a core system integrating semantic analysis, intelligent decision-making, environment perception and device linkage. The system first converts user voice into accurate instructions, then combines real-time environment data and user habits for intelligent decision-making, and finally controls different brands of devices to work cooperatively through a protocol adapter, thereby realizing the integrated, active and scene-based intelligent linkage and management of home devices across brands and protocols.

[0005] The present application provides an AI voice interaction-based integrated management and linkage system for smart home devices, comprising: A voice interaction module, which collects user voice input and generates a digitized voice stream signal, and transmits it to a semantic analysis module; A semantic analysis module, which receives the digitized voice stream signal and generates a standardized instruction signal containing device objects, execution actions and control parameters through a natural language understanding model, and transmits it to an intelligent decision-making module; An intelligent decision-making module, which receives the standardized instruction signal, combines real-time environment state signals obtained from an environment perception module, and generates a multi-device cooperative control signal through an adaptive scene calculation engine, and transmits it to a device linkage execution module; An environment perception module, which collects multi-dimensional sensor data in the home environment in real time and generates real-time environment state signals, and transmits them to the intelligent decision-making module; A device linkage execution module, which receives the multi-device cooperative control signal, generates control instruction signals corresponding to each smart device through a heterogeneous device protocol adapter, and transmits them to the target device group in the home network.

[0006] In an embodiment of the present application, the voice interaction module further comprises a distributed microphone array unit, which is used to synchronously collect voice input in different physical spaces in the home and generate enhanced voice stream signals with spatial orientation information. The semantic analysis module can identify the specific spatial position of the sound source by receiving the enhanced voice stream signals, and then integrate the spatial position information as an additional parameter into the standardized instruction signal, so that the subsequent multi-device cooperative control signal can accurately control the device group in a specific spatial position, thereby realizing the spatial perception interaction capability based on sound source positioning.

[0007] In an embodiment of the present application, a context understanding unit is integrated in the semantic parsing module, which maintains a dynamically updated dialogue context model by analyzing the logical relationship of instructions in the historical interaction sequence, and when receiving a new digitized voice stream signal, the unit can complete and resolve the referential words or omitted statements in the current instruction according to the dialogue context model, thereby generating a more complete and accurate standardized instruction signal, ensuring accurate understanding of the user's continuous and context-dependent complex voice instructions.

[0008] In an embodiment of the present application, a user habit learning unit is built in the adaptive scene computing engine in the intelligent decision-making module, which autonomously constructs and updates the user's personalized behavior pattern model by continuously analyzing the association between historical standardized instruction signals and historical real-time environment state signals, and when receiving a new standardized instruction signal, the engine can combine the current real-time environment state signal and call the user's personalized behavior pattern model for comprehensive reasoning, thereby generating a personalized multi-device collaborative control signal that is more in line with the user's long-term usage habits and preferences.

[0009] In an embodiment of the present application, the environment perception module includes a multi-source sensor data fusion unit that receives heterogeneous data streams from temperature sensors, humidity sensors, light sensors, human infrared sensors, and door and window opening and closing state sensors, and performs time synchronization and calibration processing on these heterogeneous data streams, and then generates a unified and comprehensive real-time environment state signal through a feature extraction algorithm, which fully describes the overall condition of the current home environment, providing more abundant and reliable decision-making basis for the intelligent decision-making module.

[0010] In an embodiment of the present application, the intelligent decision-making module also accesses an external service interface unit when generating the multi-device collaborative control signal, which is used to obtain external public service information such as future weather forecasts, current public transportation status, or schedule events in the user's calendar, and incorporates these external service information as a decision factor into the calculation process, so that the final generated multi-device collaborative control signal not only responds to the indoor environment but also forms a linkage with the changes in the external world, realizing intelligent decision-making based on the integration of internal and external contexts.

[0011] In an embodiment of the present application, the heterogeneous device protocol adapter in the device linkage execution module maintains an extended device protocol library, which contains the standard and private data packet format definitions of multiple Internet of Things communication protocols, and the adapter translates the received general multi-device collaborative control signal into the underlying control instruction signal that can be recognized by specific brand and specific model devices by querying the protocol library, thereby realizing unified management and seamless access to heterogeneous member devices with different communication protocols and technical standards.

[0012] In an embodiment of the present application, the device linkage execution module further comprises a local execution redundancy unit which is automatically activated in the case of system network connection interruption, takes over the output of the heterogeneous device protocol adapter and continues to execute the basic device control function in the local network according to the last known valid strategy, while recording all execution operations and performing data synchronization after network recovery, thereby ensuring the continuity and reliability of the core control function of the system in the case of network fluctuations.

[0013] In an embodiment of the present application, the system further comprises a virtual scene configuration module which receives complex scene rules defined by the user through the graphical interface, including multiple device states and their trigger conditions, and compiles the scene rules into scene description signals that can be recognized by the intelligent decision module. If the real-time environment state signal matches the preset trigger condition in the scene description signal during runtime, the intelligent decision module automatically generates and issues the corresponding multi-device coordinated control signal, thereby realizing automatic scene linkage without voice instructions.

[0014] In an embodiment of the present application, a user confirmation and feedback module is added between the semantic analysis module and the intelligent decision module. The module receives standardized instruction signals and converts them into visual or audible prompt signals in natural language to present to the user. After obtaining the user's explicit confirmation feedback signal, the instruction signal is further transmitted to the intelligent decision module. If the user's negative feedback is received, the operation is cancelled and the user is prompted to re-enter the instruction. This mechanism effectively avoids device misoperation caused by voice recognition or semantic understanding errors. The AI voice interaction-based home smart device integrated management and linkage system provided by the present application solves the problem by constructing a core system that integrates semantic analysis, intelligent decision, environment perception, and device linkage. The system first converts user voice into accurate instructions, then makes intelligent decisions in combination with real-time environment data and user habits, and finally controls different brands of devices to work cooperatively through protocol adapters, thereby realizing cross-brand, cross-protocol, integrated, proactive, and scene-based intelligent linkage and management of home devices. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0016] Figure 1 The system architecture diagram of the AI voice interaction-based home smart device integrated management and linkage system. Detailed Implementation

[0017] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0018] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0019] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0020] Please see Figure 1 The diagram illustrates the integrated management and linkage system for smart home devices based on AI voice interaction according to the present invention. It includes a voice interaction module that collects user voice input and generates a digital voice stream signal, which is then transmitted to a semantic parsing module. The semantic parsing module receives the digital voice stream signal and generates standardized instruction signals containing device objects, execution actions, and control parameters through a natural language understanding model, which are then transmitted to an intelligent decision-making module. The intelligent decision-making module receives the standardized instruction signals and, combined with real-time environmental status signals obtained from an environmental perception module, generates multi-device collaborative control signals through an adaptive scene calculation engine, which are then transmitted to a device linkage execution module. The environmental perception module collects multi-dimensional sensor data from the home environment in real time and generates real-time environmental status signals, which are then transmitted to the intelligent decision-making module. Finally, the device linkage execution module receives the multi-device collaborative control signals, generates control instruction signals corresponding to each smart device through a heterogeneous device protocol adapter, and sends them to the target device group in the home network.

[0021] Figure 1As shown, its core architecture consists of five interworking core modules, forming a complete intelligent processing chain from user instruction input to device group execution. The system begins with the voice interaction module, which serves as the physical interface for the system to directly interact with the user. Its primary function is to capture raw user voice input. These voice inputs are often generated in complex acoustic scenarios in a home environment, which may contain background noise, echo, and interference from multiple people speaking at the same time. The high-sensitivity microphone array built into this module is responsible for collecting these analog sound wave signals and immediately converting them into digitized voice stream signals through its internal analog-to-digital converter. This conversion process is the foundation of all subsequent advanced processing, ensuring accurate digital representation of voice information. Next, the digitized voice stream signal is immediately transmitted to the semantic analysis module in the system, thus opening a new stage of instruction understanding. The semantic analysis module is one of the "brains" of the entire system, responsible for understanding user intent. After receiving the digitized voice stream signal from the voice interaction module, its integrated automatic speech recognition unit first starts, converting continuous voice stream into corresponding text information. However, text conversion alone is far from intelligent requirements, and its core value lies in its integrated natural language understanding model. This model performs deep syntactic analysis and semantic mining on the generated text, including a series of complex calculations such as word segmentation, part-of-speech tagging, named entity recognition, and dependency syntax analysis. Its purpose is to accurately extract three core elements from seemingly casual daily speech: device object, execution action, and control parameter. For example, for the user's vague instruction "turn the living room lights up a little brighter," the model needs to accurately identify that the "device object" is "the living room lights," the "execution action" is "adjust," and the "control parameter" is "increase the brightness level." Finally, this module encapsulates these understanding results into a standardized instruction signal that is clear in structure and machine-readable. This signal has a pre-defined data format, ensuring unambiguous and efficient information transfer between subsequent modules. At this point, the user's intent has been successfully interpreted by the system, and the standardized instruction signal is immediately transmitted to the intelligent decision-making hub of the entire system - the intelligent decision-making module.

[0022] In particular, the intelligent decision module is the command center that implements the intelligent linkage of devices. It receives the standardized instruction signals from the semantic analysis module. However, its decision is not only dependent on the direct instructions of the user, but also introduces another dimension of key information - the environmental context. It simultaneously receives real-time environmental state signals from the environmental perception module, which provides an objective snapshot of the current home environment, such as the temperature, humidity, light intensity, whether there is human activity, etc. The adaptive scene calculation engine inside the module analyzes the fusion of the two signals, which involves complex logical judgment and strategy selection. The engine may judge that, in the case of sufficient current environmental light, the instruction to "turn on the light" should be rejected or the user should be prompted; or when the user issues the instruction "I am very hot", combined with the current temperature sensor data, the decision is made to turn on the air conditioner or open the fan to be more energy-efficient. After this comprehensive analysis, the module finally generates not a single device control command, but a multi-device collaborative control signal that may involve multiple devices and has a timing logic. This signal is a high-level action program that indicates the device group that needs to work together and the action sequence they need to perform. Subsequently, this macro action program is sent to the device linkage execution module, waiting to be translated into specific actions.

[0023] Further, the device linkage execution module is the "hands and feet" of the system connecting with the physical world, which is responsible for translating the abstract tasks issued by the intelligent decision module into bottom-level commands that can be understood and executed by specific devices in the home network. After receiving the multi-device collaborative control signal, the key component inside it - the heterogeneous device protocol adapter starts working. Smart devices in the home come from different manufacturers and use different communication protocols, such as wireless fidelity, Bluetooth mesh network, Zigbee protocol, etc. The adapter maintains a large protocol driver library that can identify the type and protocol of the target device. Then, it acts like a multilingual translator who can "translate" the general collaborative control signal into control instruction signals that can be recognized by specific devices, conforming to their private communication protocols and data packet formats. Finally, these specific instruction signals are sent to each target device group through the home local area network, driving lights, air conditioners, curtains, etc. to complete the specified linkage operation, thus achieving the overall scene effect expected by the user.

[0024] In an embodiment of the present application, the voice interaction module therein is enhanced. The claim indicates that the voice interaction module further comprises a distributed microphone array unit, which is an important extension at the basic hardware level. This unit is not a single microphone, but multiple microphone nodes are distributed in different functional spaces of the home, such as the living room, bedroom, kitchen, etc. These nodes can synchronously collect voice input from different spatial points and generate an enhanced voice stream signal with spatial orientation information through advanced signal processing algorithms such as beamforming and sound source positioning. This means that the signal not only contains the user's voice content itself, but also additional precise spatial position information of the sound source, such as "the sound comes from the east of the central part of the living room". When the semantic analysis module receives this enhanced signal, its processing capability is greatly improved. It can first analyze the spatial orientation information embedded therein and integrate this information as a crucial additional parameter into the standardized instruction signal generated. The technical effect of this improvement is revolutionary, as it enables the subsequent intelligent decision-making module to generate multi-device coordinated control signals based on spatial precision control. For example, when the user says "turn on the light here" in the dining room area, the system identifies that the user is located in the dining room through sound source positioning, and accurately controls the dining room lamps, not the lamps in the living room or kitchen. This greatly improves the interaction accuracy and user experience of the system in complex multi-room environments, realizes true spatial perception interaction, and effectively avoids device mis-triggering due to ambiguous location. The semantic analysis module therein is further defined. The claim indicates that the semantic analysis module integrates a context understanding unit, which is a key enhancement at the artificial intelligence understanding level. The core function of this unit is to maintain a dynamically updated dialogue context model, which is an internal state machine that can record and understand the logical relationship of the conversation history in a short period of time. It can understand the relevance between instructions through continuous analysis and learning of historical interaction sequences. When the module receives a new digitized voice stream signal, the unit does not process the current instruction in isolation, but actively calls the current dialogue context model as a reference background for interpreting the new instruction. This mechanism enables the system to have strong ability to handle referential words and omitted sentences. For example, the user first says "turn on the TV in the living room", and the system executes, then the user says "turn up the volume a little". For the second instruction, "volume" and "a little" are ambiguous references, lacking clear objects and parameters. At this time, the context understanding unit can accurately parse "volume" as "the volume of the TV in the living room" according to the dialogue history just established (the last instruction was for "the TV in the living room"), and generate a complete and accurate standardized instruction signal.Without this unit, the system will not be able to understand such coherent dialogues, and the user will have to say complete and accurate instructions every time, making the interactive experience very mechanical and cumbersome. Therefore, the introduction of this unit ensures the accurate understanding of the system for the complex dialogue process of human nature, continuity, and containing a large number of context-dependent relationships, and is a key technical feature to improve the intelligence and humanization of the system.

[0025] As shown in Figure 1 Further deepening and concretization of the intelligent decision module, the core of which is the introduction of a user habit learning unit with continuous learning ability, so that the system changes from a passive instruction executor to an active service provider. The focus of this claim is to describe a key subcomponent inside the adaptive scene computing engine and its working principle. This user habit learning unit is not a static, pre-defined rule base, but a dynamic, evolving machine learning model. Its operation relies on continuous analysis and mining of historical data, which mainly includes two sources: one is the historical standardized instruction signal sequence, which records all the instructions issued by the user in the past and their contents; the second is the historical real-time environment state signal sequence, which synchronously records the corresponding home environment state at the time of each instruction. The unit analyzes the association and pattern between the two types of historical signals through complex algorithms, for example, it may find that when the ambient light decreases to a certain threshold at night, the user has a high probability of issuing the instruction "turn on the living room light and dim the bedroom light"; or find that the user tends to set the air conditioner to a certain temperature in the morning of each working day. By mining these deep and recurring patterns, the unit can autonomously build and continuously update a highly personalized user behavior pattern model, which is essentially a digital abstraction of the user's living habits and preferences. After that, when the intelligent decision module receives a new standardized instruction signal again, its decision-making process will fundamentally change. Instead of making reactive decisions based on the current instruction and instantaneous environment state, the adaptive scene computing engine will actively invoke the user personalized behavior pattern model to participate in comprehensive reasoning. The engine will match the current real-time context with the historical patterns stored in the model and make predictions, generating more predictive multi-device collaborative control signals beyond the current simple instruction. For example, the user simply issues the instruction "I'm back", and the system, combined with the model, knows that the user's habit is to turn on the light first, then turn on the air conditioner and play the news after coming home, so it will automatically generate a collaborative signal containing these three operations; or when the user adjusts the temperature, the system will directly suggest and adjust to the temperature value that the user most commonly sets at that time. This decision-making mechanism based on long-term habit learning greatly improves the intelligence level and user experience of the system, making its service truly considerate and personalized.

[0026] As shown in Figure 1As shown, the environmental perception module is functionally expanded and refined, and its core is to construct a comprehensive, reliable and unified environmental perception system through multi-source sensor data fusion technology, to provide a solid data foundation for intelligent decision-making. The claim elaborates the working process of a special data processing unit in the environmental perception module, i.e. the multi-source sensor data fusion unit. The home environment is complex and changeable, and single sensor data can only reflect one side of the environment, while truly intelligent decision-making requires a comprehensive and three-dimensional environmental awareness. The unit receives raw heterogeneous data streams from various sensors deployed in the home environment, including temperature sensors for monitoring the hot and cold degree of the environment, humidity sensors for monitoring the dry or humid degree of the air, light sensors for monitoring the light and dark changes of the environment, human infrared sensors for sensing the presence or absence of human movement, and magnetic sensors for monitoring the opening and closing state of doors and windows. These data streams are "heterogeneous", meaning they may differ in data format, unit, sampling frequency and communication protocol. The primary task of the unit is to preprocess these disordered multi-source data, including time stamp synchronization to ensure that all data are snapshots under the same time reference, and data calibration to eliminate measurement errors and biases between different sensors. After completing these basic preparatory work, the unit will start a complex feature extraction algorithm. This algorithm does not simply stack data, but performs dimensionality reduction, denoising and correlation analysis on the data to extract high-level features that best represent the overall state of the current environment. Finally, the output of the unit is not a disordered data packet, but a unified, structured and comprehensive real-time environmental state signal. This signal is a highly refined information synthesis, which may use a vector or a state code to comprehensively describe the overall status of the current home environment, such as "no one, bright, warm, ventilated". Such a high-quality signal provides an incredibly rich, reliable and easy-to-process decision-making basis for the subsequent intelligent decision-making module, making decision-making no longer based on guesswork or single conditions, but on deep insight into the environment, which is a key prerequisite for achieving advanced intelligent linkage.

[0027] Further, the intelligent decision module is expanded in another dimension, which is to break the limitation of smart home system as a closed system, and to bring the dynamic information of the outside world into the decision-making process through the introduction of an external service interface unit, so as to realize the intelligent linkage between the home environment and the broader external environment. The claim points out that when the intelligent decision module performs its core decision-making function, its information input source is no longer limited to semantic instructions and sensor data within the system. It adds an external service interface unit, which acts as a bridge between the system and the external Internet service world. Through the calling of various application programming interfaces, this unit can obtain a variety of external public service information in real time. These information is of great value and includes a wide range of categories, such as future weather forecast information from the meteorological department (e.g. rainstorm in two hours), current public transportation status information from the traffic department (e.g. delay of the subway line frequently taken by the user), or schedule event information from the user's mobile device calendar (e.g. an important video conference in one hour). After obtaining these information, the unit converts it into a format that the system can understand and inputs it as a crucial decision factor into the calculation process of the adaptive scene calculation engine. This means that the decision logic of the intelligent decision engine becomes a multi-element function that integrates internal environment status, user instructions and external world events. As a result, the multi-device coordination control signals generated by the system will contain unprecedented foresight and contextuality. For example, the system obtains the information that "the future weather forecast is rainstorm", and may automatically close all windows before the user arrives home; it monitors the event that "there is a meeting in the user's calendar in one hour", and may automatically adjust the home environment to the meeting mode (e.g. adjust the light, turn off the TV, mute the sound) ten minutes before the meeting starts; or it adjusts the start time of the smart water heater according to the traffic congestion information. This deep integration of home internal intelligence and external digital ecology greatly expands the application scenarios and intelligent upper limit of the system, making it evolve from an automatic tool for a small environment to an intelligent life assistant that can understand and adapt to changes in the external world.

[0028] As Figure 1As shown, the device linkage execution module is the most critical technical refinement, and its core is to disclose a specific implementation scheme for the system to solve the industry core pain point of interconnection and intercommunication between intelligent devices of different brands and different protocols, that is, through a heterogeneous device protocol adapter with an embedded extended device protocol library. The claim goes deep into the internal working mechanism of the module. In the field of smart home, one of the biggest challenges is the fragmentation of devices and the diversity of protocols. There are countless brands of smart devices on the market, which may use wireless fidelity, Bluetooth mesh network, Zigbee protocol and other communication protocols, and even if the protocols are the same, different brands often use private and incompatible data packet formats to define their control instructions. The multi-device cooperative control signal received by the device linkage execution module from the intelligent decision module is a high-level, general and device-independent action command, such as "set the living room main light brightness to seventy percent". This general command cannot directly control a specific lamp of a specific brand. At this time, the heterogeneous device protocol adapter begins to play its role as a "general translator". The adapter maintains an extensible device protocol library inside, which is essentially a large database that stores the standard specifications of various mainstream and even non-mainstream Internet of Things communication protocols, as well as the private communication instruction sets and data packet structure definitions of many specific brands and specific models of devices. When the adapter receives a general control signal, it will first identify the identity of the target device (such as brand, model, protocol type), then query the protocol library to find the communication driving rules that exactly match the device. Then, it translates the general "set brightness" command into the underlying binary control instruction signal that the device can recognize and conforms to its private protocol specification. This process may be to encapsulate the command into a specific structure of wireless fidelity data packet, or to convert it into a specific format of Bluetooth broadcast information. Through this ingenious design, the invention successfully realizes the unified management, seamless access and control of various heterogeneous smart devices on the market that use different communication protocols and technical standards, ultimately breaking down the technical barriers and data silos between devices, enabling users to truly enjoy the convenience brought by integrated integration, which is the basic technical guarantee for the invention to realize its core value.

[0029] The AI voice interaction-based home smart device integrated management and linkage system of the present invention solves the problem by building a core system that integrates semantic analysis, intelligent decision, environment perception and device linkage. The system first converts user voice into accurate instructions, then makes intelligent decisions in combination with real-time environment data and user habits, and finally controls different brands of devices to work cooperatively through a protocol adapter, thereby realizing cross-brand, cross-protocol intelligent linkage and management of home devices in an integrated, proactive and scenario-based manner.

[0030] Therefore, through the AI voice interaction-based home smart device integrated management and linkage system, the problems of smart home device linkage fragmentation, passive response and non-intelligent interaction can be solved.

[0031] The above embodiments only exemplarily illustrate the principles and effects of the present application, and are not used to limit the present application. Any person skilled in the art can modify or change the above embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes completed by those skilled in the art without departing from the spirit and technical thought disclosed by the present application should be covered by the claims of the present application.

Claims

1. An integrated management and linkage system for smart home devices based on AI voice interaction, characterized in that: include: The voice interaction module collects user voice input and generates a digital voice stream signal, which is then transmitted to the semantic parsing module. The semantic parsing module receives digitized voice stream signals, generates standardized instruction signals containing device objects, execution actions, and control parameters through a natural language understanding model, and transmits them to the intelligent decision-making module. The intelligent decision-making module receives standardized command signals, combines them with real-time environmental status signals obtained from the environmental perception module, generates multi-device collaborative control signals through an adaptive scene computing engine, and transmits them to the device linkage execution module. An environmental sensing module collects multi-dimensional sensor data from the home environment in real time and generates real-time environmental status signals, which are then transmitted to the intelligent decision-making module. The device linkage execution module receives multi-device collaborative control signals, generates control command signals corresponding to each smart device through a heterogeneous device protocol adapter, and sends them to the target device group in the home network.

2. The integrated management and linkage system for home smart devices based on AI voice interaction as described in claim 1, characterized in that, The voice interaction module also includes a distributed microphone array unit, which is used to synchronously collect voice input in different physical spaces of the home and generate an enhanced voice stream signal with spatial orientation information. The semantic parsing module can identify the specific spatial location of the sound source by receiving the enhanced voice stream signal, and then integrate the spatial location information as an additional parameter into the standardized command signal, so that the subsequently generated multi-device collaborative control signal can accurately control the device group in a specific spatial location, thereby realizing the spatial perception interaction capability based on sound source localization.

3. The integrated management and linkage system for home smart devices based on AI voice interaction as described in claim 1, characterized in that, The semantic parsing module integrates a context understanding unit. This unit maintains a dynamically updated dialogue context model by analyzing the logical relationships of instructions in historical interaction sequences. When a new digital voice stream signal is received, the unit can complete and resolve referential words or ellipsis statements in the current instruction based on the dialogue context model, thereby generating a more complete and accurate standardized instruction signal. This ensures accurate understanding of complex voice instructions that are continuous and contain contextual dependencies.

4. The integrated management and linkage system for home smart devices based on AI voice interaction as described in claim 1, characterized in that, The adaptive scene computing engine in the intelligent decision-making module has a built-in user habit learning unit. This unit continuously analyzes the correlation between historical standardized command signals and historical real-time environmental state signals, and autonomously constructs and updates the user's personalized behavior pattern model. When a new standardized command signal is received, the engine can combine the current real-time environmental state signal and call the user's personalized behavior pattern model to perform comprehensive reasoning, thereby generating personalized multi-device collaborative control signals that are more in line with the user's long-term usage habits and preferences.

5. The integrated management and linkage system for home smart devices based on AI voice interaction as described in claim 1, characterized in that, The environmental perception module includes a multi-source sensor data fusion unit. This unit receives heterogeneous data streams from temperature sensors, humidity sensors, light sensors, human infrared sensors, and door and window opening / closing status sensors. It performs time synchronization and calibration processing on these heterogeneous data streams, and then generates a unified and comprehensive real-time environmental status signal through a feature extraction algorithm. This signal comprehensively describes the overall condition of the current home environment, providing richer and more reliable decision-making basis for the intelligent decision-making module.

6. The integrated management and linkage system for home smart devices based on AI voice interaction as described in claim 1, characterized in that, When generating multi-device collaborative control signals, the intelligent decision-making module also connects to an external service interface unit. This unit is used to obtain real-time information such as future weather forecasts, current public transportation status, or scheduled events in the user's calendar. It incorporates this external service information as a decision factor into the calculation process, so that the final multi-device collaborative control signal not only responds to the indoor environment but also interacts with changes in the external world, realizing intelligent decision-making based on the integration of internal and external contexts.

7. The integrated management and linkage system for home smart devices based on AI voice interaction according to claim 1, characterized in that, The heterogeneous device protocol adapter in the device linkage execution module maintains an extended device protocol library. This library contains standard and private data packet format definitions for various IoT communication protocols. The adapter translates the received general multi-device collaborative control signals into low-level control command signals that can be recognized by specific brand and model devices by querying this protocol library, thereby achieving unified management and seamless access for heterogeneous member devices with different communication protocols and technical standards.

8. The integrated management and linkage system for home smart devices based on AI voice interaction according to claim 1, characterized in that, The device linkage execution module also includes a local execution redundancy unit. This unit is automatically activated when the system's connection with the cloud network is interrupted. It takes over the output of the heterogeneous device protocol adapter and continues to execute basic device control functions in the local network according to the last known effective strategy. At the same time, it records all execution operations and performs data synchronization after the network is restored, thereby ensuring the continuity and reliability of the system's core control functions under network fluctuations.

9. The integrated management and linkage system for home smart devices based on AI voice interaction according to claim 1, characterized in that, The system also includes a virtual scene configuration module. This module receives complex scene rules defined by the user through a graphical interface, which include multiple device states and their triggering conditions. The module compiles these scene rules into scene description signals that the intelligent decision module can recognize. If the intelligent decision module finds that the real-time environmental state signal matches the preset triggering conditions in the scene description signal during operation, it automatically generates and sends out corresponding multi-device collaborative control signals, thereby achieving automated scene linkage without voice commands.

10. The integrated management and linkage system for home smart devices based on AI voice interaction according to claim 1, characterized in that, A user confirmation and feedback module is added between the semantic parsing module and the intelligent decision-making module. This module receives standardized command signals and converts them into visual or audible prompts in natural language and presents them to the user. Only after receiving a clear confirmation feedback signal from the user will the command signal be passed on to the intelligent decision-making module. If a negative feedback is received from the user, the operation is canceled and the user is prompted to re-enter the command. This mechanism effectively avoids device misoperation caused by errors in voice recognition or semantic understanding.

Images