Smart home control method and system, smart device and medium

Abstract

The invention relates to an intelligent home control method and system, intelligent equipment and a medium, and belongs to the technical field of artificial intelligence, and the method comprises the steps: generating a unique identifier based on own hardware behavior characteristics, carrying out the cross verification of environment data of peripheral intelligent equipment which is accessed to the network, autonomously establishing a trust relationship, and building a dynamic network without a central node; collecting user behavior data, environment characteristics and equipment capability data; processing the user behavior data, the environment characteristics and the equipment capability data to obtain scene data fragments; autonomous evolution of scene data is realized through distributed collaborative learning, and intelligent control logic adaptive to a current scene is formed; and responding to an instruction sent by a user, cooperatively verifying the intention of the user with other intelligent equipment, and matching the corresponding scene data fragment to execute a control operation. The stability of the smart home is improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence, and in particular to a smart home control method, system, smart device, and medium. Background Technology

[0002] With the development of IoT technology, smart home systems have become an important tool for improving quality of life. By networking and coordinating various devices in the home, such as lighting, appliances, security systems, and environmental monitoring devices, smart home systems enable diverse functions such as remote device control and real-time monitoring of environmental parameters, enhancing the convenience and comfort of living and driving the transformation of home life towards intelligence and energy efficiency.

[0003] Currently, traditional smart home systems generally adopt a centralized control architecture as their core solution. The core design idea of ​​this architecture is to connect all smart devices in the home to a central controller via wireless or wired means. The central controller is responsible for the unified management of device registration, command forwarding, data aggregation, and logical operations.

[0004] However, with the increasing number of smart home devices, their growing functional complexity, and users' rising demands for system continuity, the central controller, as the core hub of the entire system, becomes paralyzed when it stops working due to hardware failures, software crashes, network outages, or power interruptions. This results in all connected smart devices losing their unified scheduling and coordination capabilities. Therefore, a stable smart home control method is urgently needed. Summary of the Invention

[0005] To improve the stability of smart homes, this application provides a smart home control method, system, smart device, and medium.

[0006] Firstly, this application provides a smart home control method, which adopts the following technical solution: Generate a unique identifier based on its own hardware behavior characteristics, cross-verify with environmental data from surrounding networked smart devices, autonomously establish trust relationships, and build a dynamic network without a central node. Collect user behavior data, environmental characteristics, and device capability data; The user behavior data, environmental characteristics, and device capability data are processed to obtain scene data fragments; By using distributed collaborative learning, scene data can be autonomously evolved to form intelligent control logic that adapts to the current scene. In response to user commands, it collaborates with other smart devices to verify user intent and matches corresponding scene data fragments to execute control operations.

[0007] By adopting the above technical solutions, a decentralized dynamic network is constructed through unique hardware feature identification and cross-validation of environmental data, eliminating reliance on centralized nodes and improving network robustness and security. Multi-dimensional data collection and scenario-based processing combined with distributed collaborative learning enable scene data to autonomously evolve, allowing intelligent control logic to continuously adapt to user habits and environmental changes, significantly improving control accuracy. The multi-device collaborative verification and consensus mechanism in the user command response stage ensures accurate intent recognition and avoids misjudgment by a single device. Ultimately, this achieves autonomous networking, adaptive learning, and precise collaborative control of smart home devices, significantly optimizing user interaction experience, reducing manual intervention costs, and enhancing system intelligence and reliability.

[0008] Furthermore, the process of generating a unique identifier based on its own hardware behavior characteristics, cross-validating with environmental data from nearby networked smart devices, autonomously establishing trust relationships, and building a dynamic network without a central node includes: Obtain a trust verification request initiated by the smart device to be connected to the network, wherein the trust verification request includes the unique identifier; Multiple network-connected smart devices feed back scene data fragments to smart devices waiting to be connected to the network, enabling the smart devices waiting to be connected to the network to cross-verify the environmental data and automatically establish a trust relationship with the network-connected smart devices, thus forming a dynamic network without a central node.

[0009] By adopting the above technical solution, a unique identifier is generated based on hardware behavior characteristics. This identifier is then cross-verified using environmental data from multiple network-connected devices to establish a trust relationship, ensuring the uniqueness and legitimacy of the network-connected devices and eliminating the risk of unauthorized devices accessing the network.

[0010] Furthermore, when any smart device is moved or offline, the method further includes: Obtain the signal strength detection results and device trust values ​​of each network-connected smart device; A first sequence in descending order is generated based on the signal monitoring intensity; A second, descending sequence is generated based on the device trust value; The sequence numbers of the network-connected devices in the first sequence and the second sequence are calculated according to a preset ratio to obtain a comprehensive sequence. Temporary cooperative clusters are formed according to the order of the integrated sequence, and mobile or offline smart devices relay data transmission.

[0011] By adopting the above technical solution, when the device is moved or offline, a comprehensive sequence is generated by combining the signal strength and the device trust value, and a temporary cooperative cluster is formed to realize data transmission relay, avoid network interruption caused by single-point device failure, comprehensively evaluate the signal and trust value, and prioritize the participation of devices with good communication quality and high reliability in cooperation, thereby improving data transmission efficiency and security and enhancing the system's fault tolerance.

[0012] Furthermore, obtaining the device trust value of each network-connected smart device includes: Acquire historical behavior data and initial trust value for each smart device; In chronological order, each historical behavior data point is assigned a value to obtain a historical behavior curve; whereby, if the historical behavior data is accurate and stable, it is assigned a value of 1; if the current data is abnormal, it is assigned a value of 0. Analyze the historical behavior curve, identify stable data segments by identifying consecutive 1s, and obtain the number of stable data segments, the length of each stable data segment, and the length of the historical behavior curve; Based on a preset table, a stability level reference table is determined corresponding to the length of the historical behavior curve. The stability level reference table includes multiple stability levels, the number of data segments corresponding to each stability level, and the calculation ratio. The number of stable data segments is compared with the stability level lookup table to obtain the corresponding stability level, and the calculation ratio corresponding to the stability level is determined. The difference between the initial trust value and the maximum trust value is calculated, and the difference is multiplied by the ratio to obtain the preliminary additional trust value; The data change trend is determined based on the number of stable data segments and the length of each stable data segment. The influence coefficient of the change trend on the additional trust value is determined. The additional trust value is calculated by multiplying the influence coefficient by the additional trust value. The device trust value of the smart device is obtained by adding the initial trust value to the additional trust value.

[0013] By adopting the above technical solution, based on historical behavioral data and initial trust value, the device trust value is accurately calculated by quantifying the characteristics of stable data segments and combining the trend coefficient, which objectively reflects the reliability of device operation, provides a scientific basis for the selection of network collaborative devices, and reduces the impact of unreliable devices on the network.

[0014] Further, determining the data change trend based on the number of stable data segments and the length of each stable data segment, and determining the influence coefficient of the change trend on the additional trust value, includes: Determine whether the number of stable data segments is less than a preset value; if so, determine the influence coefficient as the first preset value. Otherwise, the stable data segments are arranged in chronological order to obtain the first sequence; The first stable data segment in the first sequence is determined as the current data segment; Execution of loop steps: Compare the next current data segment in the first sequence with the current data segment; If they are equal, then assign the first value; If it is greater than, then assign the second value; If it is less than, then assign the second value; The next stable data segment in the first sequence is determined as the new current data segment. The loop steps are repeated until the current data segment is the last stable data segment in the first sequence, thus obtaining a second sequence composed of various assignments. Calculate the ratio of the number of each type of assignment in the first sequence to the number of stable data segments; If the ratio corresponding to the first value is the largest, then the data is stable, and the influence coefficient is determined to be the second preset value; If the ratio corresponding to the second value is the largest, the data tends to stabilize, and the influence coefficient is determined to be the third preset value; If the ratio corresponding to the third data is the largest, then the data tends to fluctuate, and the influence coefficient is determined to be the fourth preset value; The first preset value > the second preset value > the third preset value > the fourth preset value.

[0015] By adopting the above technical solution, the influence coefficient is determined by quantifying the trend of changes in the number and length of stable data segments. This allows for precise differentiation between stable, trending-to-stable, and fluctuating states of equipment operation, enabling differentiated adjustments to the additional trust value. This makes the calculation of the equipment trust value more in line with the actual operating trend and avoids the one-sidedness of a single-dimensional assessment.

[0016] Furthermore, the process of achieving autonomous evolution of scene data through distributed collaborative learning to form intelligent control logic adapted to the current scene includes: The master intelligent device in the current scene decomposes the scene data segment into at least one functional unit, and broadcasts each functional unit to other intelligent devices in the dynamic network; The system receives optimization feedback from other smart devices based on their own device capability data and local environmental characteristics for the functional unit; wherein the optimization feedback includes suggestions for modifying the execution logic of the functional unit, adding new associated device capabilities, or adjusting the execution priority of the functional unit; Based on all the optimization feedback received, each optimization feedback is evaluated and filtered through a pre-set consensus algorithm to form an optimized set of functional units; The optimized set of functional units is reorganized into new scene data fragments and synchronized to all relevant intelligent devices in the dynamic network to update their stored scene data fragments.

[0017] By adopting the above technical solution, a distributed collaborative learning mechanism is used to decompose the main control device into functional units, provide optimization suggestions from multiple devices, and screen and integrate consensus algorithms. This enables the autonomous evolution of scene data, fully integrates the capabilities and environmental information of each device in the network, allows the intelligent control logic to continuously adapt to scene changes, improves the accuracy and adaptability of the control logic, and avoids the limitations of single device decision-making.

[0018] Furthermore, the step of responding to a user-sent command, collaborating with other smart devices to verify the user's intent, and matching the corresponding scene data fragment to execute control operations includes: Any smart device, acting as an instruction receiving device, parses the instruction based on an intent recognition model and extracts at least one first-level intent keyword. The instruction receiving device broadcasts the first-level intent keyword in the dynamic network, and other intelligent devices that receive the broadcast perform a first round of matching based on the scene data fragments stored locally, and provide feedback on the matching degree and related second-level environmental context information. The instruction receiving device summarizes the feedback information. If there is a scene data segment with a matching degree exceeding the first threshold and consistent with the second-level environmental context information, it directly triggers execution. If the first threshold is not reached or there is a conflict in the environmental context information, the instruction receiving device initiates a multi-round distributed voting consensus for the user's intent in the dynamic network based on the second-level environmental context information and historical collaboration records. According to the voting consensus results, it dynamically adjusts or merges multiple related scene data fragments, generates temporary control logic and executes it, and at the same time, it uses this execution record as a new scene data fragment for learning and evolution.

[0019] By adopting the above technical solutions, user intent is verified through multi-device collaboration. The accuracy of instruction execution is ensured by combining matching degree and environmental context dual verification. In case of conflict, distributed voting consensus is initiated to generate temporary control logic to avoid misjudgment by a single device. Execution records are synchronized with the evolving scenario data, allowing the system to continuously adapt to user needs.

[0020] Secondly, this application provides a smart home control system, which adopts the following technical solution: The dynamic network building module is used to generate a unique identifier based on its own hardware behavior characteristics, and to establish trust relationships autonomously through cross-verification of environmental data from surrounding smart devices already connected to the network, thereby building a dynamic network without a central node. The data acquisition module is used to collect user behavior data, environmental characteristics, and equipment capability data. The scene data fragment processing module is used to process the user behavior data, environmental features and device capability data to obtain scene data fragments; The control logic generation module is used to automate scene data through distributed collaborative learning, forming intelligent control logic that adapts to the current scene. The instruction response module is used to respond to instructions sent by the user, collaborate with other smart devices to verify the user's intent, and match the corresponding scene data fragments to execute control operations.

[0021] By adopting the above technical solutions, a decentralized dynamic network is constructed through unique hardware feature identification and cross-validation of environmental data, eliminating reliance on centralized nodes and improving network robustness and security. Multi-dimensional data collection and scenario-based processing combined with distributed collaborative learning enable scene data to autonomously evolve, allowing intelligent control logic to continuously adapt to user habits and environmental changes, significantly improving control accuracy. The multi-device collaborative verification and consensus mechanism in the user command response stage ensures accurate intent recognition and avoids misjudgment by a single device. Ultimately, this achieves autonomous networking, adaptive learning, and precise collaborative control of smart home devices, significantly optimizing user interaction experience, reducing manual intervention costs, and enhancing system intelligence and reliability.

[0022] Thirdly, this application provides a smart device, which adopts the following technical solution: A smart device, comprising: At least one processor; Memory; Wireless communication module; Hardware feature acquisition module; At least one computer program, wherein the at least one computer program is stored in the memory and configured to be executed by the at least one processor, the at least one computer program being configured to: perform the method as described in any one of the first aspects.

[0023] Fourthly, this application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium storing a computer program that can be loaded by a processor and execute the method as described in any one of the first aspects.

[0024] In summary, this application includes at least one of the following beneficial technical effects: 1. By using unique hardware identifiers and cross-validation with environmental data, a decentralized dynamic network is constructed, eliminating dependence on centralized nodes and improving network robustness and security; 2. Multi-dimensional data acquisition and scenario-based processing combined with distributed collaborative learning enable scenario data to autonomously evolve, allowing intelligent control logic to continuously adapt to user habits and environmental changes, significantly improving control accuracy. Attached Figure Description

[0025] Figure 1This is a flowchart illustrating the smart home control method in the embodiments of this application.

[0026] Figure 2 This is a structural block diagram of the smart home control system in the embodiments of this application.

[0027] Figure 3 This is a structural block diagram of the smart device in the embodiments of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0030] This application discloses a smart home control method. The method is implemented based on a network architecture without a central node. The smart device cluster includes, but is not limited to, smart lights, smart air conditioners, smart curtains, smart door locks, smart speakers, smart sensors, and other devices with data acquisition, wireless communication, and data processing capabilities. Each smart device has a built-in processor, memory, and wireless communication module, and can autonomously complete unique identifier generation, data acquisition, collaborative computing, and control operation execution.

[0031] Reference Figure 1 The method includes steps S101 to S105: Step S101: Generate a unique identifier based on its own hardware behavior characteristics, cross-verify with environmental data from surrounding smart devices already connected to the network, autonomously establish trust relationships, and build a dynamic network without a central node.

[0032] Specifically, when each smart device leaves the factory, the hardware feature acquisition module extracts its own hardware behavior features, including but not limited to the processor operating frequency fluctuation curve, the memory chip read / write response time series, the stable value of the communication module signal transmission power, and the initial calibration parameters of the sensors. All features are hashed to generate a unique, non-copyable device identifier. The method includes steps S1011 to S1012.

[0033] Step S1011: Obtain the trust verification request initiated by the smart device to be connected to the network. The trust verification request includes a unique identifier. Step S1012: Multiple network-connected smart devices feed back scene data fragments to the smart devices to be connected to the network, so that the smart devices to be connected to the network can cross-verify the environmental data fed back, automatically establish a trust relationship with the network-connected smart devices, and build a dynamic network without a central node.

[0034] Specifically, after the smart device to be connected to the network is powered on, it scans for other smart devices already connected to the network in the vicinity through the wireless communication module and sends a trust verification request to all the scanned devices. The trust verification request includes information such as the device's unique identifier, device type, and capability description.

[0035] Upon receiving the request, nearby smart devices already connected to the network extract a unique identifier and simultaneously call their own environmental data acquisition module to obtain environmental data from the last 10 seconds. They then associate this environmental data with their own device identifier and current timestamp and send it back to the smart device to be connected to the network.

[0036] The smart device to be connected to the network collects environmental data from at least three smart devices already connected to the network. The same type of environmental data is cross-validated. For example, the average value of all Danquil temperature data is calculated. If the temperature data reported by a single smart device already connected to the network deviates from the average value by no more than ±1℃, the environmental data of the smart device already connected to the network is determined to be valid. Otherwise, invalid feedback data is discarded.

[0037] If the valid environmental data collected by the smart device to be connected to the network comes from at least 3 smart devices already connected to the network, and the consistency of the same environmental parameter in the valid environmental data reaches more than 80%, then it is determined that the device to be connected to the network has established a trust relationship with these devices already connected to the network. The device to be added to the network stores the information of the devices already in the network that have established a trust relationship in the local network topology table. At the same time, it sends network access confirmation information to these devices already in the network. Upon receiving the confirmation information, the devices already in the network update their own network topology table and incorporate the device to be added to the network, thus completing the expansion of the dynamic network without a central node.

[0038] After generating the dynamic network, each smart device needs to understand the status and data of other devices. If any smart device moves or goes offline, the dynamic network needs to adjust for the impact of the device's offline status in a timely manner. To ensure the stable operation of the smart home, the method includes steps S11 to S15: Step S11: Obtain the signal strength detection results and device trust values ​​of each network-connected smart device.

[0039] Specifically, in a dynamic network, each connected smart device monitors the signal strength of other devices with which it has established a trust relationship in real time. When the signal strength of any smart device falls below a certain threshold multiple times consecutively, the device is determined to have moved or gone offline. At this point, any smart device immediately broadcasts a move / offline notification to the entire dynamic network. Upon receiving the notification, each connected smart device reports its own signal strength detection results and its own device trust value to the broadcast initiating device, which then aggregates all the feedback information.

[0040] The method for obtaining the device trust value of a network-connected smart device includes steps S111 to S118: Step S111: Obtain historical behavior data and initial trust value for each smart device.

[0041] Specifically, each smart device records its own historical behavior data in real time during operation, including data collection accuracy, data transmission stability, and command execution success rate, generating a historical record every hour and storing it in the local historical database.

[0042] The initial trust value of the device is preset by the manufacturer and is determined according to the device type and hardware configuration. The value range is 60-80 points (out of 100). The initial trust value of sensor devices is 70 points, the initial trust value of execution devices (such as air conditioners and curtains) is 75 points, and the initial trust value of gateway devices is 80 points.

[0043] Step S112: Assign values ​​to each historical behavior data point in chronological order to obtain the historical behavior curve; where, if the historical behavior data is accurate and stable, assign a value of 1; if the current data is abnormal, assign a value of 0.

[0044] Step S113: Analyze the historical behavior curve, identify the stable data segments by the data segments that are consecutively 1, and obtain the number of stable data segments, the length of each stable data segment, and the length of the historical behavior curve.

[0045] Step S114: Determine the stability level reference table corresponding to the length of the historical behavior curve based on the preset table. The stability level reference table includes multiple stability levels, the number of data segments corresponding to each stability level, and the calculation ratio.

[0046] Specifically, the length of the historical behavior curve is also one of the factors affecting the assessment of stability. Therefore, different lengths of historical behavior curves correspond to different stability level comparison tables.

[0047] The more data segments there are, the higher the stability level, and the larger the corresponding calculation ratio.

[0048] For example, if the historical behavior curve has 100 lines, the corresponding stability level table is as follows: Step S115: Compare the number of stable data segments with the stability level comparison table to obtain the corresponding stability level, and determine the calculation ratio corresponding to the stability level.

[0049] For example, there are 7 stable data segments, corresponding to the second-level stability level, with a calculation ratio of 0.6.

[0050] Step S116: Calculate the difference between the initial trust value and the maximum trust value, and multiply the difference by the ratio to obtain the preliminary additional trust value.

[0051] For example, if the difference is 25 points, the initial additional trust value = difference × calculation ratio = 25 × 0.6 = 15 points.

[0052] Step S117: Determine the data change trend based on the number of stable data segments and the length of each stable data segment, determine the influence coefficient of the change trend on the additional confidence value, and multiply the additional confidence value by the influence coefficient to calculate the additional confidence value. The method includes steps Sa to Sj: Step Sa: Determine if the number of stable data segments is less than the preset value; if so, proceed to step Sb: Determine the influence coefficient to be the first preset value.

[0053] Specifically, the preset number of stable data segments is 5. If the number of stable data segments is less than 5, the smart device has a stable foundation, the possibility of data fluctuation is small, and the trend of change has little impact on the trust value. The impact coefficient is the first preset value, such as 0.9.

[0054] Otherwise, proceed to steps Sc to S: Step Sc: Arrange the stable data segments in chronological order to obtain the first sequence.

[0055] Step Sd: Determine the first stable data segment in the first sequence as the current data segment.

[0056] Step Se: Execute the loop step: Compare the next current data segment in the first sequence with the current data segment; If they are equal, then assign the first value; If it is greater than, then assign the second value; If it is less than, then assign the second value; The next stable data segment in the first sequence is determined as the new current data segment; Step Sf: Repeat the loop step Se until the current data segment is the last stable data segment in the first sequence, and obtain the second sequence composed of various assignments.

[0057] Step Sg: Calculate the ratio of the number of each type of assignment in the first sequence to the number of stable data segments.

[0058] Step Sh: If the ratio corresponding to the first value is the largest, then the data is stable, and the influence coefficient is determined to be the second preset value.

[0059] Step Si: If the ratio corresponding to the second value is the largest, the data tends to be stable, and the influence coefficient is determined to be the third preset value.

[0060] Step Sj: If the ratio corresponding to the third data is the largest, then the data tends to fluctuate, and the influence coefficient is determined to be the fourth preset value.

[0061] First preset value > Second preset value > Third preset value > Fourth preset value.

[0062] If the influence coefficient is 0.8, then the final additional trust value = initial additional trust value × influence coefficient = 15 × 0.8 = 12.

[0063] Step S118: Add the initial trust value to the additional trust value to obtain the device trust value of the smart device.

[0064] Specifically, if the initial trust value is 75 points and the additional trust value is 12 points, then the device's trust value is 87 points. The initial trust value represents the factory quality, while the additional trust value reflects the current operating instructions; therefore, the device trust value can comprehensively reflect the reliability of the device.

[0065] Step S12: Generate a descending first sequence based on the signal monitoring intensity.

[0066] Step S13: Generate a descending second sequence based on the device trust value.

[0067] Step S14: Calculate the serial numbers of the network-connected devices in the first sequence and the second sequence according to a preset ratio to obtain a comprehensive sequence.

[0068] Step S15: Form temporary cooperative clusters according to the order of the integrated sequence, and relay the data transmission between mobile or offline smart devices.

[0069] Specifically, the first N devices are selected in sequence to form a temporary collaborative cluster. The device with the highest overall score in the temporary collaborative cluster becomes the temporary master device, taking over the data transmission and processing tasks of the offline devices until the offline devices rejoin the network or a replacement device is added to the network. Here, N is the number of associated devices of the offline device. For example, if the offline device is a temperature and humidity sensor and the associated devices are air conditioners and curtains, then N=2.

[0070] Step S102: Collect user behavior data, environmental characteristics, and device capability data.

[0071] Specifically, each smart device collects user behavior data through its own sensors or associated devices, including but not limited to the user's device usage time patterns, device parameter setting preferences, user movement trajectories, such as the time the air conditioner is turned on each day, the air conditioner's usual temperature, the time the lights are turned off, the usual brightness of the lights, and smart door lock records.

[0072] Sensor devices collect indoor environmental data, including temperature, humidity, light intensity, air quality, and noise levels.

[0073] Each smart device also performs regular self-checks, recording its own device capability parameters, including operating status, functional support, performance parameters, energy consumption data, etc.

[0074] Step S103: Process user behavior data, environmental characteristics and device capability data to obtain scene data fragments.

[0075] Specifically, each smart device transmits collected user behavior data, environmental characteristics, and device capability data to a local processor, where the raw data is preprocessed using data cleaning algorithms. The preprocessed data then undergoes feature extraction algorithms to extract key features. Based on these extracted key features, clustering and integration are performed according to scene dimensions to obtain scene data fragments. Each scene data fragment includes a scene identifier, triggering conditions, a list of associated devices, and device control parameters. For example, in a nighttime rest scene, the data fragment includes: triggering conditions (after 22:00, light intensity < 50 lux, user enters bedroom), associated devices (bedroom lights, bedroom air conditioner, bedroom curtains), and control parameters (light brightness 30%, air conditioner temperature 26℃, curtains closed).

[0076] Step S104: Implement autonomous evolution of scene data through distributed collaborative learning to form intelligent control logic adapted to the current scene. The method includes steps S1041 to S1044.

[0077] Step S1041: The master intelligent device in the current scene decomposes the scene data fragment into at least one functional unit and broadcasts each functional unit to other intelligent devices in the dynamic network.

[0078] Specifically, the master smart device is automatically elected in the dynamic network based on the scenario type. For home scenarios, the device with the highest usage frequency and the highest device trust value is prioritized as the master device, such as a smart speaker in the living room or an air conditioner in the master bedroom. If there are multiple candidate smart devices, they are determined through distributed voting, with the device receiving more than 50% of the votes being elected.

[0079] The main control intelligent device receives all scene data fragments in the current scenario and breaks down each scene data fragment into at least one functional unit. Each functional unit corresponds to a specific device operation or data processing task. For example, the data fragment in the nighttime rest scenario is broken down into: lighting adjustment functional unit, air conditioning control functional unit, curtain control functional unit, etc. Each functional unit includes the operation object, basic execution logic, parameter range, etc.

[0080] The master control intelligent device broadcasts the disassembled functional units to all other intelligent devices in the dynamic network through the wireless communication module. The broadcast information includes functional unit identifier, scene identifier, basic execution logic, and parameter range.

[0081] Step S1042: Receive optimization feedback from other smart devices on the functional unit based on their own device capability data and local environmental characteristics; wherein, the optimization feedback includes suggestions for modifying the execution logic of the functional unit, adding new associated device capabilities, or adjusting the execution priority of the functional unit.

[0082] Specifically, after receiving the broadcast from the functional unit, other smart devices analyze it in conjunction with their own device capability data and local environmental characteristics to generate optimization feedback: If the device is the object of operation in the functional unit, suggestions for modifying the execution logic should be made based on its own hardware performance. For example, if the device is a bedroom light corresponding to the lighting adjustment functional unit, then based on its own supported brightness adjustment range and color temperature adjustment capability, the suggestion for modifying the brightness adjustment is: "The brightness adjustment range can be expanded to 10%-100%, and it is recommended to add color temperature adjustment to 3000K."

[0083] If the device is associated with auxiliary devices, add associated device capabilities based on local environmental characteristics. For example, if the device is a temperature and humidity sensor, it is recommended to add a dehumidification mode option to the air conditioning control unit based on the current bedroom temperature.

[0084] If the device is the master control device for other scenarios, the priority of the functional units will be adjusted based on historical collaboration records. For example, in a nighttime rest scenario, the curtain control functional unit has a higher priority than the lighting adjustment functional unit.

[0085] Each smart device will send optimization feedback to the main control smart device.

[0086] Step S1043: Based on all the received optimization feedback, evaluate and filter each optimization feedback through a preset consensus algorithm to form an optimized set of functional units.

[0087] Specifically, the preset consensus algorithm is based on the Byzantine fault tolerance algorithm, and the evaluation dimensions include whether the feedback meets the device capabilities and scenario requirements, whether it can be implemented through existing hardware, and whether it can improve the user experience, with each dimension having a weight of 40%, 30%, and 30%, respectively.

[0088] The top 10 devices with the highest trust values ​​in the dynamic network are invited to form evaluation nodes. Each evaluation node scores each optimization feedback and calculates the average score for each feedback.

[0089] Given a score range of 0-10, an optimization feedback screening threshold of 7 points is set. Optimization feedback with an average score ≥ 7 points is retained. For multiple optimization feedbacks for the same functional unit, if there are conflicts, the feedback with the highest voting support rate at the evaluation node is selected.

[0090] Step S1044: Reorganize the optimized set of functional units into new scene data fragments and synchronize them to all relevant intelligent devices in the dynamic network to update their stored scene data fragments.

[0091] Specifically, the master control intelligent device integrates the filtered and optimized feedback into the corresponding functional units, forming an optimized set of functional units. Following the scenario logic, the optimized set of functional units is reorganized to generate new scenario data fragments. These new data fragments include optimized triggering conditions, associated devices, execution logic, and parameters. Each device receives the new scenario data fragments synchronized from the master control intelligent device and stores them locally, overwriting the old data.

[0092] Step S105: In response to the user's command, collaborate with other smart devices to verify the user's intent, match the corresponding scene data fragment, and execute the control operation. The method includes steps S1051 to S1054: Step S1051: Any smart device acts as an instruction receiving device, parses the instruction based on the intent recognition model, and extracts at least one first-level intent keyword.

[0093] Specifically, after any connected smart device receives a command sent by a user, it preprocesses the command, such as removing invalid characters, converting speech to text, and completing ambiguous commands, to obtain a standard command text.

[0094] The instruction receiving device invokes a pre-trained intent recognition model to perform semantic analysis on the instruction text and extract at least one first-level intent keyword. The intent recognition model is built on an LSTM neural network, and the training dataset contains multiple home control scenario instruction samples. The first-level intent keywords are core words representing the user's core needs. For example, if the user instruction is "Go to sleep at night, turn off the lights, and turn the air conditioner down," the extracted first-level intent keywords would be "sleep," "turn off the lights," and "cool down the air conditioner."

[0095] Step S1052: The instruction receiving device broadcasts the first-level intent keyword in the dynamic network. Other intelligent devices that receive the broadcast perform a first round of matching based on the scene data fragments stored locally, and provide feedback on the matching degree and related second-level environmental context information.

[0096] Specifically, the instruction receiving device encapsulates the first-level intent keyword, the instruction sending timestamp, and its own device identifier into a broadcast packet, and broadcasts it to all network-connected smart devices through the multicast communication channel of the dynamic network.

[0097] After receiving the broadcast packet, each smart device calculates the percentage overlap between the set of triggering keywords in the scene data fragment and the first-level intent keywords of the broadcast, thus obtaining the matching degree. It also collects its own corresponding second-level environmental context information. This second-level environmental context information consists of key data characterizing the current physical environment and device status, such as: Environmental parameters: current time, temperature of the area, light intensity, and human presence status; Device status parameters: its own operating status and the connection status of associated devices.

[0098] Each smart device generates a feedback packet by combining the matching degree, the second-level environmental context information, and its own device identifier, and sends it to the instruction receiving device.

[0099] Step S1053: The instruction receiving device summarizes the feedback information. If there is a scene data segment with a matching degree exceeding the first threshold and consistent with the second-level environmental context information, execution is triggered directly.

[0100] Specifically, the instruction receiving device presets a first threshold of 80%. When the matching degree of the scene data segment is ≥80%, it is considered that the segment is initially matched with the user's intent. Then, it determines whether the second-level environmental context is consistent. If it is a numerical parameter, it is determined to be consistent if the deviation between the mean of all feedback values ​​and a single feedback value is less than the preset error. If it is a state parameter, it is determined to be consistent if all feedback state information is completely consistent.

[0101] When at least one scene data segment meets the conditions, the instruction receiving device sends control instructions to all smart devices associated with that scene data segment, thereby triggering execution.

[0102] Step S1054: If the first threshold is not reached or there is a conflict in the environmental context information, then based on the second-level environmental context information and historical collaboration records, a multi-round distributed voting consensus for the user's intent is initiated in the dynamic network. According to the voting consensus results, multiple related scene data fragments are dynamically adjusted or merged to generate temporary control logic and execute it. At the same time, the execution record is used as a new scene data fragment for learning and evolution.

[0103] Specifically, the instruction receiving device summarizes all feedback second-level environmental context information, filters out conflicting parameters, and obtains a consensus basic environmental information table; at the same time, it extracts all scene data fragments with a matching degree >50% as relevant scene fragments and obtains a relevant scene fragment summary table.

[0104] Then, the instruction receiving device broadcasts the "Consensus Basic Environment Information Table," the "Related Scenario Fragment Summary Table," and the first-level intent keywords to all smart devices in the network with a trust value ≥80, inviting these devices to participate in multiple rounds of voting as consensus nodes. The first round filters scenario fragments with a score ≥6; the second round of voting determines the optimal parameters with a vote rate ≥60%; the final round selects a scheme with a vote rate ≥70%, and if no solution is found, the top two schemes are re-voted to determine the winner.

[0105] Based on the voting consensus results, the instruction execution device adjusts conflict parameters and merges multiple scenario fragments to generate temporary control logic. The instruction receiving device sends the logic to associated devices for execution and collects the results, records complete execution information, encapsulates it into a new scenario fragment with a timestamp, synchronizes it across the entire network, and incorporates it into the device database for subsequent distributed collaborative learning and evolution.

[0106] Furthermore, to optimize the communication mechanism between devices in the smart home system, reduce communication latency, and improve response speed, smart devices utilize machine learning algorithms to train predictive models based on device usage and environmental changes over a period of time. When the state of a smart device changes, the predictive model is immediately activated to assess whether other related smart devices need to adjust their behavior. If no adjustment is needed, they temporarily enter a low-power mode to reduce unnecessary communication activities. The predictive model is continuously updated to adapt to environmental changes and user habits, ensuring the accuracy and real-time nature of predictions.

[0107] On the other hand, to enhance the security of smart home systems and prevent unauthorized access and operation, users must authenticate themselves before using any device. Smart devices compare the user's biometric information with pre-stored data to ensure the authenticity and uniqueness of the user's identity. For sensitive operations, such as changing system settings or controlling important devices, users are required to undergo dual authentication, combining passwords and biometrics to increase the difficulty of unauthorized operations. Smart devices regularly perform security checks, monitor abnormal behavior on the network, and take immediate measures to isolate potential threats and protect system security once suspicious activity is detected.

[0108] To better implement the above methods, this application also provides a smart home control system, referring to... Figure 2 The smart home control system 200 includes: The dynamic network construction module 201 is used to generate a unique identifier based on its own hardware behavior characteristics, and to establish a trust relationship autonomously by cross-verifying the environmental data of surrounding smart devices already connected to the network, thereby constructing a dynamic network without a central node. Data acquisition module 202 is used to collect user behavior data, environmental characteristics and equipment capability data; The scene data fragment processing module 203 is used to process user behavior data, environmental characteristics and device capability data to obtain scene data fragments; The control logic generation module 204 is used to realize the autonomous evolution of scene data through distributed collaborative learning, forming intelligent control logic adapted to the current scene; The instruction response module 205 is used to respond to instructions sent by the user, collaborate with other smart devices to verify the user's intent, and match the corresponding scene data fragments to execute control operations.

[0109] Furthermore, the dynamic network construction module 201 is specifically used for: Obtain the trust verification request initiated by the smart device to be connected to the network. The trust verification request includes a unique identifier. Multiple network-connected smart devices feed back scene data fragments to smart devices waiting to be connected to the network, enabling the latter to cross-validate the environmental data and automatically establish a trust relationship with the network-connected smart devices, thus forming a dynamic network without a central node.

[0110] Furthermore, the smart home control system 200 also includes: The signal and trust value acquisition module is used to acquire the signal strength detection results and device trust values ​​of each network-connected smart device. The first sequence generation module is used to generate a descending first sequence based on the signal detection intensity; The second sequence generation module is used to generate a descending second sequence based on the device trust value; The integrated sequence generation module is used to calculate the sequence number of the network-connected devices in the first sequence and the second sequence according to a preset ratio to obtain the integrated sequence. The temporary collaboration cluster selection module is used to form temporary collaboration clusters in the order of the comprehensive sequence, and relay mobile or offline smart devices for data transmission.

[0111] Furthermore, obtain the device trust value of each network-connected smart device, including: Acquire historical behavior data and initial trust value for each smart device; In chronological order, each historical behavior data point is assigned a value to obtain a historical behavior curve; if the historical behavior data is accurate and stable, it is assigned a value of 1; if the current data is abnormal, it is assigned a value of 0. By analyzing historical behavior curves, the data segments with consecutive 1s are identified as stable data segments, and the number of stable data segments, the length of each stable data segment, and the length of the historical behavior curve are obtained. Based on a preset table, a stability level reference table is determined corresponding to the length of the historical behavior curve. The stability level reference table includes multiple stability levels, the number of data segments corresponding to each stability level, and the calculation ratio. Compare the number of stable data segments with the stability level comparison table to obtain the corresponding stability level, and determine the calculation ratio corresponding to the stability level. The difference between the initial trust value and the maximum trust value is calculated, and the difference is multiplied by the ratio to obtain the initial additional trust value; The data change trend is determined based on the number of stable data segments and the length of each stable data segment. The influence coefficient of the change trend on the additional trust value is determined. The additional trust value is calculated by multiplying the additional trust value by the influence coefficient. The device trust value of the smart device is obtained by adding the initial trust value to the additional trust value.

[0112] The data change trend is determined based on the number of stable data segments and the length of each stable data segment. The impact coefficient of this change trend on the additional confidence value is then determined, including: Determine if the number of stable data segments is less than a preset value; if so, set the influence coefficient to the first preset value. Otherwise, arrange the stable data segments in chronological order to obtain the first sequence; The first stable data segment in the first sequence is identified as the current data segment; Execution of loop steps: Compare the next current data segment in the first sequence with the current data segment; If they are equal, then assign the first value; If it is greater than, then assign the second value; If it is less than, then assign the second value; The next stable data segment in the first sequence is determined as the new current data segment. The loop steps are repeated until the current data segment is the last stable data segment in the first sequence, resulting in a second sequence composed of various assignments. Calculate the ratio of the number of each type of assignment in the first sequence to the number of stable data segments; If the ratio corresponding to the first value is the largest, then the data is stable, and the influence coefficient is determined to be the second preset value; If the ratio corresponding to the second value is the largest, the data tends to be stable, and the influence coefficient is determined to be the third preset value; If the ratio corresponding to the third data is the largest, then the data tends to fluctuate, and the influence coefficient is determined to be the fourth preset value; First preset value > Second preset value > Third preset value > Fourth preset value.

[0113] Furthermore, the control logic generation module 204 is specifically used for: The master intelligent device in the current scene decomposes the scene data fragment into at least one functional unit, and broadcasts each functional unit to other intelligent devices in the dynamic network; Receive optimization feedback from other smart devices based on their own device capability data and local environmental characteristics for the functional unit; among which, optimization feedback includes suggestions for modifying the execution logic of the functional unit, adding new associated device capabilities, or adjusting the execution priority of the functional unit; Based on all the optimization feedback received, each optimization feedback is evaluated and filtered through a pre-set consensus algorithm to form an optimized set of functional units; The optimized set of functional units is reorganized into new scene data fragments and synchronized to all relevant smart devices in the dynamic network to update their stored scene data fragments.

[0114] Furthermore, the instruction response module 205 is specifically used for: Any smart device acts as an instruction receiving device, parses the instruction based on the intent recognition model, and extracts at least one first-level intent keyword; The instruction receiving device broadcasts first-level intent keywords in the dynamic network. Other intelligent devices that receive the broadcast perform the first round of matching based on the scene data fragments stored locally, and provide feedback on the matching degree and related second-level environmental context information. The instruction receiving device summarizes the feedback information. If there is a scene data segment with a matching degree exceeding the first threshold and consistent with the second-level environmental context information, execution is triggered directly. If the first threshold is not reached or there is a conflict in the environmental context information, the instruction receiving device initiates a multi-round distributed voting consensus on the user's intent in the dynamic network based on the second-level environmental context information and historical collaboration records. According to the voting consensus results, multiple related scene data fragments are dynamically adjusted or merged to generate temporary control logic and execute it. At the same time, the execution record is used as a new scene data fragment for learning and evolution.

[0115] The various variations and specific examples of the methods in the foregoing embodiments are also applicable to the smart home control device of this embodiment. Through the foregoing detailed description of the smart home control method, those skilled in the art can clearly understand the implementation method of the smart home control device in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0116] To better implement the above methods, embodiments of this application provide a smart device, referring to... Figure 3 The intelligent device 300 includes a processor 301, a memory 303, and a wireless communication module 304. The memory 303 and the wireless communication module 304 are both connected to the processor 301 and are connected via a bus 302. The structure of this intelligent device 300 does not constitute a limitation on the embodiments of this application.

[0117] Processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 301 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0118] Bus 302 may include a pathway for transmitting information between the aforementioned components. Bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 302 may be divided into address bus, data bus, control bus, etc.

[0119] The memory 303 may be a ROM (Read Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0120] The memory 303 is used to store application code that executes the solution of this application, and its execution is controlled by the processor 301. The processor 301 is used to execute the application code stored in the memory 303 to implement the content shown in the foregoing method embodiments.

[0121] The wireless communication module 304 is a core component for data transmission between smart devices and between devices and the network, and it needs to adapt to the communication requirements of decentralized dynamic networks in smart homes. This includes, but is not limited to, Wi-Fi modules, Bluetooth modules, ZigBee modules, and LoRa modules.

[0122] Figure 3 The smart device 300 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0123] This application also provides a computer-readable storage medium storing a computer program. When executed by a processor, the program implements the smart home control method provided in the above embodiments. It constructs a decentralized dynamic network through unique hardware identifiers and cross-validation of environmental data, eliminating reliance on centralized nodes and improving network robustness and security. Multi-dimensional data acquisition and scenario-based processing combined with distributed collaborative learning enable autonomous evolution of scenario data, allowing the intelligent control logic to continuously adapt to user habits and environmental changes, significantly improving control accuracy. A multi-device collaborative verification and consensus mechanism in the user command response stage ensures accurate intent recognition and avoids misjudgment by a single device. Ultimately, it achieves autonomous networking, adaptive learning, and precise collaborative control of smart home devices, significantly optimizing user interaction experience, reducing manual intervention costs, and enhancing system intelligence and reliability.

[0124] In this embodiment, the computer-readable storage medium can be a tangible device that holds and stores instructions used by an instruction execution device. The computer-readable storage medium can be, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof. Specifically, the computer-readable storage medium can be a portable computer disk, a hard disk, a USB flash drive, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory stick, floppy disk, optical disk, magnetic disk, mechanical encoding device, or any combination thereof.

[0125] The computer program in this embodiment includes program code for performing all the aforementioned methods. The program code may include instructions corresponding to the method steps provided in the above embodiments. The computer program can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded to an external computer or external storage device via a network (e.g., the Internet, a local area network, a wide area network, and / or a wireless network). The computer program can be executed entirely on the user's computer as a standalone software package.

[0126] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

[0127] Additionally, it should be understood that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

Claims

1. A smart home control method, characterized in that, Executed by a cluster of smart devices, including: Generate a unique identifier based on its own hardware behavior characteristics, cross-verify with environmental data from surrounding networked smart devices, autonomously establish trust relationships, and build a dynamic network without a central node. Collect user behavior data, environmental characteristics, and device capability data; The user behavior data, environmental characteristics, and device capability data are processed to obtain scene data fragments; By using distributed collaborative learning, scene data can be autonomously evolved to form intelligent control logic that adapts to the current scene. Responding to user commands, it collaborates with other smart devices to verify user intent and matches corresponding scene data fragments to execute control operations.

2. The method according to claim 1, characterized in that, The process of generating a unique identifier based on its own hardware behavior characteristics, cross-validating with environmental data from surrounding networked smart devices, autonomously establishing trust relationships, and building a dynamic network without a central node includes: Obtain a trust verification request initiated by the smart device to be connected to the network, wherein the trust verification request includes the unique identifier; Multiple network-connected smart devices feed back scene data fragments to smart devices waiting to be connected to the network, enabling the smart devices waiting to be connected to the network to cross-verify the environmental data and automatically establish a trust relationship with the network-connected smart devices, thus forming a dynamic network without a central node.

3. The method according to claim 2, characterized in that, When any smart device is moved or offline, the method further includes: Obtain the signal strength detection results and device trust values ​​of each network-connected smart device; A first sequence in descending order is generated based on the signal monitoring intensity; A second, descending sequence is generated based on the device trust value; The sequence numbers of the network-connected devices in the first sequence and the second sequence are calculated according to a preset ratio to obtain a comprehensive sequence. Temporary cooperative clusters are formed according to the order of the integrated sequence, and mobile or offline smart devices relay data transmission.

4. The method according to claim 3, characterized in that, The process of obtaining the device trust value of each network-connected smart device includes: Acquire historical behavior data and initial trust value for each smart device; In chronological order, each historical behavior data point is assigned a value to obtain a historical behavior curve; whereby, if the historical behavior data is accurate and stable, it is assigned a value of 1; if the current data is abnormal, it is assigned a value of 0. Analyze the historical behavior curve, identify stable data segments by identifying consecutive 1s, and obtain the number of stable data segments, the length of each stable data segment, and the length of the historical behavior curve; Based on a preset table, a stability level reference table is determined corresponding to the length of the historical behavior curve. The stability level reference table includes multiple stability levels, the number of data segments corresponding to each stability level, and the calculation ratio. The number of stable data segments is compared with the stability level lookup table to obtain the corresponding stability level, and the calculation ratio corresponding to the stability level is determined. The difference between the initial trust value and the maximum trust value is calculated, and the difference is multiplied by the ratio to obtain the preliminary additional trust value; The data change trend is determined based on the number of stable data segments and the length of each stable data segment. The influence coefficient of the change trend on the additional trust value is determined. The additional trust value is calculated by multiplying the influence coefficient by the additional trust value. The device trust value of the smart device is obtained by adding the initial trust value to the additional trust value.

5. The method according to claim 4, characterized in that, The step of determining the data change trend based on the number of stable data segments and the length of each stable data segment, and determining the impact coefficient of the change trend on the additional trust value, includes: Determine whether the number of stable data segments is less than a preset value; if so, determine the influence coefficient as the first preset value. Otherwise, the stable data segments are arranged in chronological order to obtain the first sequence; The first stable data segment in the first sequence is determined as the current data segment; Execution of loop steps: Compare the next current data segment in the first sequence with the current data segment; If they are equal, then assign the first value; If it is greater than, then assign the second value; If it is less than, then assign the second value; The next stable data segment in the first sequence is determined as the new current data segment. The loop steps are repeated until the current data segment is the last stable data segment in the first sequence, thus obtaining a second sequence composed of various assignments. Calculate the ratio of the number of each type of assignment in the first sequence to the number of stable data segments; If the ratio corresponding to the first value is the largest, then the data is stable, and the influence coefficient is determined to be the second preset value; If the ratio corresponding to the second value is the largest, the data tends to be stable, and the influence coefficient is determined to be the third preset value; If the ratio corresponding to the third data is the largest, then the data tends to fluctuate, and the influence coefficient is determined to be the fourth preset value; The first preset value > the second preset value > the third preset value > the fourth preset value.

6. The method according to claim 5, characterized in that, The process of autonomously evolving scene data through distributed collaborative learning to form intelligent control logic adapted to the current scene includes: The master intelligent device in the current scene decomposes the scene data segment into at least one functional unit, and broadcasts each functional unit to other intelligent devices in the dynamic network; The system receives optimization feedback from other smart devices based on their own device capability data and local environmental characteristics for the functional unit; wherein the optimization feedback includes suggestions for modifying the execution logic of the functional unit, adding new associated device capabilities, or adjusting the execution priority of the functional unit; Based on all the optimization feedback received, each optimization feedback is evaluated and filtered through a pre-set consensus algorithm to form an optimized set of functional units; The optimized set of functional units is reorganized into new scene data fragments and synchronized to all relevant intelligent devices in the dynamic network to update their stored scene data fragments.

7. The method according to claim 1 or 6, characterized in that, The process of responding to user-sent commands, collaborating with other smart devices to verify user intent, and matching corresponding scene data fragments to execute control operations includes: Any smart device, acting as an instruction receiving device, parses the instruction based on an intent recognition model and extracts at least one first-level intent keyword. The instruction receiving device broadcasts the first-level intent keyword in the dynamic network, and other intelligent devices that receive the broadcast perform a first round of matching based on the scene data fragments stored locally, and provide feedback on the matching degree and related second-level environmental context information. The instruction receiving device summarizes the feedback information. If there is a scene data segment with a matching degree exceeding the first threshold and consistent with the second-level environmental context information, it directly triggers execution. If the first threshold is not reached or there is a conflict in the environmental context information, the instruction receiving device initiates a multi-round distributed voting consensus for the user's intent in the dynamic network based on the second-level environmental context information and historical collaboration records. According to the voting consensus results, it dynamically adjusts or merges multiple related scene data fragments, generates temporary control logic and executes it, and at the same time, it uses this execution record as a new scene data fragment for learning and evolution.

8. A smart home control system, characterized in that, include: The dynamic network building module is used to generate a unique identifier based on its own hardware behavior characteristics, and to establish trust relationships autonomously through cross-verification of environmental data from surrounding smart devices already connected to the network, thereby building a dynamic network without a central node. The data acquisition module is used to collect user behavior data, environmental characteristics, and equipment capability data. The scene data fragment processing module is used to process the user behavior data, environmental features and device capability data to obtain scene data fragments; The control logic generation module is used to automate scene data through distributed collaborative learning, forming intelligent control logic that adapts to the current scene. The instruction response module is used to respond to instructions sent by the user, collaborate with other smart devices to verify the user's intent, and match the corresponding scene data fragments to execute control operations.

9. A smart device, characterized in that, include: At least one processor; Memory; Wireless communication module; At least one computer program, wherein the at least one computer program is stored in the memory and configured to be executed by the at least one processor, the at least one computer program being configured to: perform a smart home control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored and can be loaded by a processor and executed as described in any one of claims 1 to 7.

Images