A local autonomous decision-making whole-house intelligent closed-loop control system and method based on a Matter 1.5 protocol

Abstract

The application discloses a local autonomous decision-making whole-house intelligent closed-loop control system and method based on a Matter 1.5 protocol. The three modules of a protocol unified modeling module, a local AI decision-making module and a local energy scheduling module are all solidified locally. In the case of network interruption or cloud server failure, the three modules operate in a local closed loop, and the device control instruction generation function and the energy consumption optimization instruction generation function remain 100% normal operation locally, completely solving the industry common pain point of "network interruption paralysis" of the existing system. Meanwhile, the protocol unified modeling module standardizes the bidirectional mapping of the private data format and control instructions of different communication protocol terminal devices, establishes a local unified communication link, and realizes local linkage control of cross-brand and cross-protocol devices without cloud API transfer, completely breaking the protocol heterogeneous island.

Description

Technical Field

[0001] This invention relates to the technical field of smart homes, specifically to a local autonomous decision-making whole-house intelligent closed-loop control system and method based on the Matter 1.5 protocol. Background Technology

[0002] The core logic of current mainstream smart home systems, such as scene decision-making, device linkage, and energy consumption scheduling, is deployed on cloud servers. Local gateways only handle data transmission and forwarding, lacking local computing power and independent decision-making units. When the network is interrupted or the cloud server fails, all intelligent linkage functions are completely paralyzed, leaving only manual device control capabilities. The system's stability and fault tolerance are extremely poor.

[0003] Furthermore, existing smart home devices are developed using proprietary Zigbee, Wi-Fi, and BLE protocols, resulting in incompatibility between device object models, control command sets, and communication frame formats across different brands. Traditional smart gateways only support one-way data transmission and simple forwarding for a single protocol, failing to achieve standardized parsing of cross-protocol data and unified command mapping. Cross-ecosystem device collaborative control can only be achieved through third-party cloud API relays, leading to issues such as high communication latency, limited control permissions, poor operational reliability, and easy leakage of user data. Summary of the Invention

[0004] This application provides a local autonomous decision-making whole-house intelligent closed-loop control system and method based on the Matter 1.5 protocol, which is used to solve the dual architectural defects of existing whole-house intelligent systems mentioned in the background technology, namely strong cloud dependence and protocol heterogeneity islands, which lead to system paralysis after network outage and inability of cross-brand devices to link locally.

[0005] To address the aforementioned technical problems, firstly, this application provides a local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol, comprising: The protocol unified modeling module is used to standardize the bidirectional mapping of private data formats and control commands of terminal devices with different communication protocols, and to establish a local unified communication link. The local AI decision-making module is used to receive and analyze human body status, environmental parameters and user operation data through the unified communication link, complete user behavior modeling and scene strategy autonomous iteration locally, and generate device control commands. The local energy dispatch module is used to receive the control instructions generated by the local AI decision module, and combine the whole house load power, photovoltaic power generation and grid peak and valley information to perform local energy-saving optimization dispatch and generate energy consumption optimization instructions. The instructions generated by the local AI decision-making module and the local energy dispatching module are all converted into control instructions that the target device can recognize through the protocol unified modeling module before being sent out for execution. In the event of network interruption or cloud failure, the protocol unified modeling module, local AI decision-making module, and local energy scheduling module all operate in a closed loop locally. The device control command generation function and the energy consumption optimization command generation function maintain normal local operation and do not rely on real-time cloud computing.

[0006] In one embodiment, the protocol unified modeling module has a built-in Matter standard device object dictionary and general Cluster mapping rules, supporting device access and bidirectional mapping for four communication protocols: Wi-Fi, Thread, BLE, and Zigbee.

[0007] In one embodiment, the system further includes a multi-mode adaptive traffic splitting module, which automatically allocates high-bandwidth devices to Wi-Fi links, low-power sensors to Thread links, and network configuration and debugging services to BLE links based on the bandwidth, latency, and power consumption requirements of the terminal devices.

[0008] In one embodiment, the local AI decision-making module includes a lightweight neural network inference model for receiving three-state data of human dynamic movement, static sitting, and no one leaving the site, collected at a preset frequency, and performing user behavior clustering and habit modeling locally.

[0009] In one embodiment, the local AI decision-making module further includes a multi-parameter coupling and control unit, which integrates four types of data: ambient light, indoor and outdoor temperature, seasonal time sequence, and user's historical operating habits, to dynamically adjust the color temperature and brightness of lights, the constant temperature range of air conditioning, the fresh air operation level, and the opening and closing angle of curtains.

[0010] In one embodiment, the local energy dispatching module includes a peak-valley electricity price adaptive dispatching unit, which is used to automatically start water heaters, charging piles and sweeping robot equipment during off-peak electricity price periods, and shut down unnecessary loads during peak electricity price periods.

[0011] In one embodiment, the local energy dispatch module includes a photovoltaic energy storage collaborative dispatch unit, which is used to prioritize the consumption of photovoltaic power generation to supply the entire house load, store surplus power in energy storage devices, and smoothly switch to grid power supply during periods of no sunlight.

[0012] In one embodiment, the system further includes a data acquisition module comprising one or more of a millimeter-wave human presence sensor, a temperature and humidity sensor, a light sensor, a water immersion sensor, a gas sensor, and a smoke sensor.

[0013] In one embodiment, the millimeter-wave human presence sensor is configured to collect three-state data of dynamic human movement, static sedentary state, and no-person leaving state at a preset sampling frequency, and output the data to the local AI decision-making module.

[0014] In a second aspect, the present application further provides a local autonomous decision-making whole-house intelligent closed-loop control method based on the Matter 1.5 protocol, which is configured to implement the method of the local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol, comprising the following steps: performing standardized bidirectional mapping on the private data formats and control instructions of terminal devices of different communication protocols through a protocol unified modeling module, to establish a local unified communication link; collecting human state data, environmental parameter data and user operation behavior data in real time through the unified communication link; performing user behavior modeling and autonomous iteration of scene strategies on the collected data locally by a local AI decision-making module, to generate device regulation and control instructions; receiving at least part of the device regulation and control instructions by a local energy scheduling module, and performing energy-saving optimization scheduling locally in combination with whole-house load power, photovoltaic power generation amount and power grid peak-valley information, to generate energy consumption optimization instructions; converting both the device regulation and control instructions generated by the local AI decision-making module and the energy consumption optimization instructions generated by the local energy scheduling module into control instructions recognizable by target devices through the protocol unified modeling module, and then issuing the converted instructions for execution.

[0015] The beneficial effects of the above-mentioned local autonomous decision-making whole-house intelligent closed-loop control system and method based on the Matter 1.5 protocol are as follows: In the present application, the protocol unified modeling module, the local AI decision-making module, and the local energy scheduling module are all solidified locally. Under the condition of network interruption or cloud server failure, all three modules operate in a local closed loop, and the device regulation and control instruction generation function and the energy consumption optimization instruction generation function maintain 100% normal local operation, which completely solves the common industry pain point of "paralysis immediately upon network disconnection" of existing systems. At the same time, the protocol unified modeling module performs standardized bidirectional mapping on the private data formats and control instructions of terminal devices of different communication protocols to establish a local unified communication link, which realizes local linkage control of cross-brand and cross-protocol devices without cloud API relay, completely breaking the protocol heterogeneous isolated island. Brief Description of the Drawings

[0016] Figure 1 is a schematic structural diagram of the local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol shown in the embodiment of the present application; Figure 2This is a flowchart illustrating a local autonomous decision-making whole-house intelligent closed-loop control method based on the Matter 1.5 protocol, as shown in an embodiment of this application. Detailed Implementation

[0017] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0018] It should be noted that when a component is considered to be "connected" to another component, it can be directly connected to the other component or there may be an intermediary component present. Conversely, when a component is said to be "directly" connected to another component, there is no intermediary component.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] like Figure 1 As shown, this application provides a local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol, including: Protocol Unified Modeling Module 1 is used to standardize the bidirectional mapping of private data formats and control commands of terminal devices with different communication protocols, and to establish a local unified communication link. The Protocol Unified Modeling Module 1 has a built-in Matter standard device object dictionary and general Cluster mapping rules, supporting device access and bidirectional mapping for four communication protocols: Wi-Fi, Thread, BLE, and Zigbee.

[0021] The unified protocol modeling module 1 is deployed on local edge computing nodes and is built based on the Matter 1.5 standard cluster model released by the CSA Alliance. The unified protocol modeling module 1 has a built-in Matter standard device object dictionary and general cluster mapping rules, supporting device access for four communication protocols: Wi-Fi, Thread, BLE, and Zigbee.

[0022] In practical operation, the unified protocol modeling module 1 performs standardized bidirectional mapping of the private data formats and control commands of terminal devices with different communication protocols: for uplink data, it converts the private data points of various sensors into the Matter standard format; for downlink commands, it converts the Matter standard commands into control commands recognizable by the target device's private protocol. Through this bidirectional mapping, the unified protocol modeling module 1 establishes a unified communication link within the whole-house local area network, enabling devices of different brands and protocols to achieve native local interoperability without the need for cloud API relay.

[0023] Local AI decision-making module 2 is used to receive and analyze human body status, environmental parameters and user operation data through the unified communication link, complete user behavior modeling and scene strategy autonomous iteration locally, and generate device control commands. The local AI decision-making module includes a lightweight neural network inference model, which is used to receive three-state data of human dynamic movement, static sitting, and no one leaving the site, collected at a preset frequency, and to complete user behavior clustering and habit modeling locally. The local AI decision-making module also includes a multi-parameter coupling and control unit, which integrates four types of data: ambient light, indoor and outdoor temperature, seasonal time sequence, and user's historical operating habits, to dynamically adjust the color temperature and brightness of lights, the constant temperature range of air conditioning, the fresh air operation level, and the opening and closing angle of curtains.

[0024] The local AI decision-making module 2 is also deployed on a local edge computing node and communicates with the unified protocol modeling module 1. Through the unified communication link established by the unified protocol modeling module 1, the local AI decision-making module 2 receives real-time human status data, environmental parameter data, and user operation behavior data. Human status data includes three states: dynamic movement, static sitting, and no one leaving the site; environmental parameter data includes temperature, humidity, and light intensity; user operation behavior data includes records of device on / off operations, parameter adjustments, etc. The local AI decision-making module 2 is equipped with a lightweight neural network inference model, such as a micro-inference model built using the TensorFlow Lite framework. This model continuously processes the received multi-dimensional data at a sampling frequency of 10Hz, completing data cleaning, feature extraction, user behavior clustering, and long-term usage habit modeling locally. Based on this, the system can autonomously iteratively optimize scene trigger thresholds, device control parameters, and delay execution strategies without requiring users to manually preset fixed linkage rules.

[0025] For example, through continuous learning, the system discovered that users typically set their living room lights to 80% brightness and 3000K color temperature, while also setting the air conditioner to 24℃, during the weekday period from 7:00 PM to 8:00 PM. After a week of learning, the local AI decision module 2 automatically modeled this pattern as a "homecoming scenario" and automatically generated corresponding device control commands for the same time period in the future, without requiring users to repeatedly set or manually trigger them.

[0026] In addition, the local AI decision-making module 2 also incorporates a multi-parameter coupling control unit. This unit integrates four types of data: ambient light, indoor and outdoor temperature, seasonal time sequence, and user's historical operating habits, to dynamically adjust the color temperature and brightness of lights, the constant temperature range of the air conditioner, the fresh air operation level, and the opening angle of curtains. For example, on a summer afternoon, if sufficient indoor light (≥500 lux), a high outdoor temperature (≥32℃), and the user's historical habit of partial blackout are detected, the system automatically adjusts the curtains to a 40% opening angle, reduces the light brightness to 20% or turns it off, and sets the air conditioner to a constant temperature range of 24°-26℃, achieving coordinated adaptive control of multiple devices.

[0027] The local energy dispatch module 3 is used to receive the control instructions generated by the local AI decision module, and combine the whole house load power, photovoltaic power generation and grid peak and valley information to perform local energy-saving optimization dispatch and generate energy consumption optimization instructions. The local energy dispatch module 3 communicates with the local AI decision-making module 2 and the unified protocol modeling module 1. The local energy dispatch module 3 is built based on the Matter 1.5 protocol energy management standard cluster model, and collects real-time data on the instantaneous power of the whole house load, cumulative energy consumption, real-time photovoltaic power generation, and grid peak and valley period signals locally.

[0028] The local energy dispatch module 3 receives at least some of the control instructions generated by the local AI decision module 2, which mainly involve energy-related equipment such as air conditioners, water heaters, and charging piles. It then combines the collected energy consumption and energy data to perform multi-objective global energy-saving optimization dispatch and generate energy consumption optimization instructions.

[0029] The specific scheduling logic includes: (1) Peak-valley electricity price adaptive scheduling: Automatically identify the peak and valley periods of the power grid. In the low-valley electricity price range, such as from 22:00 at night to 6:00 the next day, automatically start water heaters, charging piles, sweeping robots and other time-delayed high-power devices; in the peak electricity price range, such as from 10:00 to 15:00 during the day, automatically shut down unnecessary loads to reduce the peak electricity cost of the whole house.

[0030] (2) Photovoltaic energy storage coordinated dispatch: Prioritize the consumption of clean energy generated by photovoltaic power generation to supply the load of the whole house, and automatically store the surplus power into the energy storage device; at night or during periods of no light, smoothly switch to grid power supply to maximize the utilization rate of clean energy.

[0031] The instructions generated by the local AI decision-making module 2 and the local energy dispatching module 3 are all converted into control instructions that can be recognized by the target device through the protocol unified modeling module 1 and then sent out for execution. During the instruction issuance phase, the device control instructions generated by the local AI decision-making module 2, such as light brightness adjustment and color temperature adjustment, and the energy consumption optimization instructions generated by the local energy scheduling module 3, such as air conditioning temperature adjustment and device start / stop, are all converted into private control instructions that the target device can recognize through the protocol unified modeling module 1, and then sent to the target device for execution through the corresponding communication link.

[0032] For example, when the local AI decision-making module 2 decides to adjust the brightness of a smart bulb using a Zigbee protocol to 80%, the protocol unified modeling module 1 first generates a Matter standard control command, then maps the command into a control frame format recognizable by the Zigbee bulb's proprietary protocol, and finally sends it for execution via the Zigbee link. Similarly, when the local energy scheduling module 3 decides to start a water heater using a Wi-Fi protocol during off-peak hours, the protocol unified modeling module 1 converts the Matter standard On / Off command into a command format recognizable by the water heater's proprietary protocol and sends it via the Wi-Fi link.

[0033] In the event of network interruption or cloud failure, the protocol unified modeling module 1, the local AI decision-making module 2, and the local energy scheduling module 3 all operate in a local closed loop. The device control command generation function and the energy consumption optimization command generation function maintain normal local operation and do not rely on real-time cloud computing.

[0034] In one embodiment, the system further includes a multi-mode adaptive traffic splitting module 4, which is used to automatically allocate high-bandwidth devices to Wi-Fi links, low-power sensors to Thread links, and network distribution and debugging services to BLE links according to the bandwidth requirements, latency requirements, and power consumption requirements of terminal devices.

[0035] The multi-mode adaptive traffic splitting module 4 has a built-in link optimization algorithm that monitors the communication characteristic parameters of each terminal device in real time and automatically matches the optimal communication link according to the preset traffic splitting strategy. The communication characteristic parameters include data packet size, transmission frequency, latency sensitivity, and power consumption status. (1) High-bandwidth devices are allocated to Wi-Fi links For high-bandwidth demanding devices such as high-definition cameras (e.g., Matter 1.5 security cameras), central air conditioning systems, smart TVs, and smart speakers, the multi-mode adaptive traffic splitting module 4 automatically allocates them to the high-speed Wi-Fi link. The Wi-Fi link supports OFDMA multi-user parallel transmission and 1024-QAM high-order modulation, with a single-stream peak rate of up to 1.2Gbps, effectively carrying high-throughput services such as video and audio streams.

[0036] When the high-definition camera transmits 4K video streams via the Wi-Fi 6 link, the end-to-end control latency is stably controlled to ≤20ms, the video is smooth and stutter-free, and the control command response is jitter-free, which is significantly better than the traditional Wi-Fi 4 / 5 solution.

[0037] (2) Low-power sensors are assigned to Thread links For devices with low power consumption and low bandwidth requirements, such as temperature and humidity sensors, light sensors, human presence sensors, water immersion sensors, gas sensors, and smoke sensors, the multi-mode adaptive current distribution module 4 distributes them to the Thread network.

[0038] Thread Mesh uses the IEEE 802.15.4 physical layer, supports IPv6 low-power wireless self-organizing networks, and has the following technical advantages: Self-healing networking: When a node in the network fails or the signal deteriorates, the Thread network automatically reroutes and maintains the communication link through multiple relay nodes without manual intervention. Low power consumption operation: The sensor node adopts a sleep-wake mechanism, with standby power consumption as low as microwatts, and the device sleep standby time is ≥12 months; Low-latency transmission: Network latency ≤10ms, meeting the requirements for real-time acquisition of environmental parameters and control response.

[0039] For example, when the temperature and humidity sensor in a bedroom experiences signal attenuation due to its distance from the gateway, the Thread network automatically uses a nearby smart socket or curtain motor as a relay node to complete multi-hop data forwarding, ensuring reliable uploading of sensor data.

[0040] (3) Distribution network and commissioning services are allocated to BLE links For services with low bandwidth requirements and low power consumption, such as device initialization network configuration, key synchronization, and close-range debugging interaction, the multi-mode adaptive traffic splitting module 4 allocates them to BLE5.3 links for independent bearer.

[0041] The BLE 5.3 link uses an independent radio frequency channel and protocol stack, which does not occupy service communication bandwidth and ensures the stability of the daily control link. When a device first joins the network, the user completes the device discovery, Matter security key exchange, and network credential distribution network configuration process with the edge gateway through the BLE link via a mobile app. The network configuration process is end-to-end encrypted, ensuring high security.

[0042] (4) Zigbee bridging and adaptation for older equipment For existing Zigbee devices, such as smart bulbs and sockets that are already in use with the Zigbee protocol, the multi-mode adaptive splitter module 4 achieves bridging and adaptation through its built-in Zigbee protocol stack. Data from Zigbee devices is converted to the Matter standard format by the protocol unified modeling module 1 and then managed uniformly by the system.

[0043] Furthermore, the multi-mode adaptive splitter module 4 also has dynamic adjustment capabilities. When a link experiences congestion, signal degradation, or equipment relocation, the module automatically assesses the current link quality and triggers link reselection and seamless handover.

[0044] For example, a smart TV initially connects via a Wi-Fi link. When the user moves the TV to an area with a weak Wi-Fi signal, the multi-mode adaptive splitting module 4 detects that the signal strength is below a threshold and automatically assesses the available bandwidth of the Thread link. If the Thread link can meet the bandwidth requirements of the current service, it triggers a link switch to ensure that the service continuity is not affected.

[0045] In one embodiment, the local energy dispatch module 3 includes a peak-valley electricity price adaptive dispatch unit, which is used to automatically start water heaters, charging piles and sweeping robot equipment during off-peak electricity price periods, and shut down unnecessary loads during peak electricity price periods.

[0046] When the network is available, the system periodically synchronizes the latest peak-valley time information and real-time electricity price data from the power grid company's official API or cloud service platform and caches it in local non-volatile storage. Even if the network is subsequently interrupted, the system can still operate its scheduling logic normally based on the last synchronized peak-valley time data.

[0047] The peak-valley electricity price adaptive dispatching unit classifies and manages all load devices in the house. Based on the device's operating delay tolerance and usage rigidity, the load is divided into the following three priority levels, as shown in Table 1:

[0048] Table 1 The peak-valley electricity price adaptive dispatching unit dynamically adjusts the start-stop logic and operating power of each load according to the above priority classification.

[0049] During off-peak electricity pricing periods, such as from 22:00 to 6:00 the next day, the peak-valley electricity pricing adaptive dispatching unit performs the following dispatching operations: (1) Water heater scheduling: Detects the current water temperature of the storage water heater and the user's preset target water temperature. If the current water temperature is lower than the preset threshold, the water heater will automatically start heating. The system also has learning capabilities. Based on the user's historical water usage habits, such as the user usually uses water at 7:00 am and 8:00 pm, the system will preheat the water temperature to a reserve temperature slightly higher than the target water temperature during off-peak hours. For example, if the target water temperature is 55°C, the system will heat the water to 60°C, using off-peak low-priced electricity to complete heat storage and reduce the electricity consumption for supplementary heating during peak hours. (2) Charging Pile Scheduling: For electric vehicle charging piles connected to the system, the peak-valley electricity price adaptive scheduling unit detects the vehicle's connection status, current battery level, and user-set departure time. If the vehicle is connected and the current battery level is lower than the user-set charging threshold, and there is still sufficient time before the user-set departure time, such as ≥6 hours, the system will delay the charging task to the off-peak period to make full use of the low-priced off-peak electricity to charge the vehicle. If the user has an urgent need for vehicle use, they can trigger the "instant charging" command through the mobile app or vehicle terminal, and the system will start charging immediately, without being restricted by off-peak periods; (3) Robotic vacuum cleaner scheduling: For devices such as robotic vacuum cleaners that can be executed with a completely delayed start time, the peak-valley electricity price adaptive scheduling unit forces the start time of their tasks to be scheduled during off-peak hours. For example, if the user sets the daily cleaning time to 9:00 am, the system will automatically adjust the actual start time to 10:00 pm the previous night and complete the cleaning task before the user gets up the next day, which will not affect the user experience and will take advantage of the off-peak electricity price. (4) Other delayed equipment: For other delayed high-power equipment such as dishwashers, dryers, and electric heaters, the peak-valley electricity price adaptive scheduling unit adopts a similar scheduling strategy to concentrate the equipment start-up time in the off-peak period as much as possible, forming a "load filling valley" effect.

[0050] During peak electricity price periods, such as 10:00-15:00 during the day and 18:00-21:00 in the evening, the peak-valley electricity price adaptive dispatching unit performs the following dispatching operations: (1) Shutdown of unnecessary loads: For high-priority devices such as water heaters, charging piles, robot vacuum cleaners, dishwashers, and dryers, if they are not in operation or have completed their scheduled tasks during peak hours, the system will keep them shut down and prohibit them from starting new tasks during peak hours. If the system detects that the above devices are started during peak hours due to user manual operation or abnormal reasons, the system will push a prompt message to the user through the local interactive terminal or mobile app, suggesting that the user postpone their use to off-peak hours.

[0051] (2) Power limitation for comfort equipment: For medium-priority equipment such as air conditioners, fresh air systems, and underfloor heating, the system implements a power reduction operation strategy during peak hours: Air conditioning: When cooling in summer, raise the set temperature by 1°C-2°C, for example, from 24°C to 25°C-26°C; when heating in winter, lower the set temperature by 1°C-2°C, for example, from 22°C to 20°C-21°C, to reduce instantaneous power while ensuring basic comfort.

[0052] (3) Low-priority equipment protection: Low-priority equipment such as lighting, refrigerators, security equipment, and medical equipment are not affected by peak-hour scheduling strategies, maintain real-time response and normal operation, and ensure users' basic living needs and home safety.

[0053] In one embodiment, the local energy dispatch module 3 includes a photovoltaic energy storage collaborative dispatch unit, which is used to prioritize the consumption of photovoltaic power generation to supply the entire house load, store surplus power in energy storage devices, and smoothly switch to grid power supply during periods of no sunlight.

[0054] In one embodiment, the system further includes a data acquisition module 5, which comprises one or more of a millimeter-wave human presence sensor, a temperature and humidity sensor, a light sensor, a water immersion sensor, a gas sensor, and a smoke sensor. The millimeter-wave human presence sensor is used to collect three-state data of human dynamic movement, static sitting, and no one leaving the site at a preset sampling frequency, and outputs the data to the local AI decision-making module.

[0055] The data acquisition module 5 is communicatively connected to the protocol unified modeling module 1. The data acquisition module 5 contains various types of sensors for collecting indoor human status, environmental parameters, and security status data across the entire area. It encapsulates the collected raw data into standardized Matter format data frames and uploads them to the local AI decision-making module 2 and the local energy dispatching module 3 for processing through the unified communication link established by the protocol unified modeling module 1.

[0056] The data acquisition module 5 includes one or more of the following sensor types: millimeter-wave human presence sensor, temperature and humidity sensor, light sensor, water immersion sensor, gas sensor, and smoke sensor. These sensors can be flexibly selected according to the actual application scenario, for example: Millimeter-wave human presence sensors and temperature and humidity sensors are installed in main activity areas such as bedrooms and living rooms; light sensors are installed in locations such as balconies and near windows; water immersion sensors are installed in water-using areas such as kitchens and bathrooms; gas sensors are installed in kitchens and near gas pipelines; and smoke sensors are installed in locations such as living rooms and corridors.

[0057] Millimeter-wave human presence sensors utilize 60GHz or 77GHz millimeter-wave radar technology, enabling them to penetrate non-metallic materials such as plastic and thin wood panels, and are unaffected by changes in ambient light. These sensors continuously transmit frequency-modulated continuous wave signals at a preset sampling frequency (e.g., 10Hz, i.e., 10 samples per second), and extract the target's distance, speed, angle, and micro-motion characteristics by receiving the echo signals reflected from the human body. Based on these characteristics, the sensor can accurately identify three states of human presence: dynamic movement, static sitting, and absence of human presence. When a person moves actively in a space, such as walking, turning, or raising their hand, the echo signal shows a significant Doppler frequency shift and a prominent velocity component, indicating a dynamic movement state. When a person remains stationary in a sitting or lying position but exhibits micro-movements such as breathing and heartbeat, periodic micro-movement features are extracted from the echo signal, such as chest cavity fluctuations at a frequency of 0.2-0.5Hz, indicating a prolonged static sitting state. When no one is present in the space, the echo signal shows no human feature reflections and remains consistently above a preset threshold, indicating that no one has left the area. The millimeter-wave human presence sensor standardizes the above three-state data through the protocol unified modeling module 1 and outputs it in real time to the local AI decision-making module 2.

[0058] The temperature and humidity sensors employ digital temperature and humidity sensing chips, such as SHT30 and DHT22, to collect real-time temperature and relative humidity data for various areas within the room. The sensors collect data at a sampling frequency of 1Hz and upload it to the local AI decision-making module 2 and the local energy dispatch module 3 via the unified protocol modeling module 1.

[0059] The local AI decision-making module 2 dynamically adjusts the air conditioner's constant temperature range and fresh air operation level based on temperature and humidity data and the user's historical operating habits. The local energy scheduling module 3 participates in energy consumption optimization scheduling decisions based on temperature data.

[0060] The light sensor uses a photodiode or a digital illuminance chip, such as the BH1750, to collect ambient light intensity data in various indoor areas in real time. The light sensor collects data at a sampling frequency of 1Hz and uploads it to the local AI decision-making module 2 via the unified modeling module 1.

[0061] The local AI decision-making module 2 dynamically adjusts the color temperature and brightness of the lights and the opening angle of the curtains based on lighting data, seasonal timing, and user history. For example, when natural light is sufficient (≥500 lux), the system automatically reduces the brightness of the lights or turns them off; when natural light is insufficient (≤50 lux), the system automatically turns on the lights and adjusts them to a suitable brightness.

[0062] The water immersion sensor employs electrode-based or photoelectric water immersion detection principles and is deployed in areas prone to water accumulation, such as kitchens, bathrooms, near washing machines, and floor drains. The sensor monitors in real time for water overflow, and once a water immersion signal is detected, it immediately reports an alarm to the local security closed-loop unit via the unified protocol modeling module 1.

[0063] Gas sensors employ catalytic combustion or semiconductor gas-sensitive elements and are deployed near kitchen gas stoves and gas pipeline interfaces to monitor the concentration of combustible gases such as natural gas and liquefied petroleum gas in real time. When the detected gas concentration exceeds a preset safety threshold (e.g., natural gas concentration ≥ 1% LEL), the sensor automatically reports an alarm.

[0064] Each sensor collects raw physical quantities according to its own sampling frequency. The built-in microprocessor of the sensor or the centralized processing unit of the data acquisition module 5 filters, denoises, and scales the raw data to convert it into standard physical units. According to the acquisition module 5, the Matter standardized interface provided by the protocol unified modeling module 1 is called to encapsulate the preprocessed data according to the Matter 1.5 standard data model. The encapsulated Matter standard data frame is uploaded to the local AI decision module 2 and the local energy dispatch module 3 through a unified communication link, such as Wi-Fi, Thread or BLE.

[0065] like Figure 2 As shown, in a second aspect, this application also includes a method for local autonomous decision-making whole-house intelligent closed-loop control based on the Matter 1.5 protocol, used to implement a method for a local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol, comprising the following steps: S1. Through the protocol unified modeling module, standardize the private data formats and control commands of terminal devices with different communication protocols and establish a local unified communication link. S2. Real-time collection of human body status data, environmental parameter data, and user control behavior data through the unified communication link; S3. The local AI decision-making module performs user behavior modeling and scene strategy autonomous iteration on the collected data locally to generate device control commands; S4. The local energy dispatch module receives at least some of the equipment control instructions and, in conjunction with the whole house load power, photovoltaic power generation and grid peak and valley information, performs energy-saving optimization dispatch locally to generate energy consumption optimization instructions. S5. The device control instructions generated by the local AI decision-making module and the energy consumption optimization instructions generated by the local energy scheduling module are converted into control instructions that the target device can recognize through the protocol unified modeling module and then sent out for execution.

[0066] The options described in the above system embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application's embodiments can be found in the above system embodiments, and will not be repeated in this embodiment.

[0067] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0068] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol, characterized in that, include: The protocol unified modeling module is used to standardize the bidirectional mapping of private data formats and control commands of terminal devices with different communication protocols, and to establish a local unified communication link. The local AI decision-making module is used to receive and analyze human body status, environmental parameters and user operation data through the unified communication link, complete user behavior modeling and scene strategy autonomous iteration locally, and generate device control commands. The local energy dispatch module is used to receive the control instructions generated by the local AI decision module, and combine the whole house load power, photovoltaic power generation and grid peak and valley information to perform local energy-saving optimization dispatch and generate energy consumption optimization instructions. The instructions generated by the local AI decision-making module and the local energy dispatching module are all converted into control instructions that the target device can recognize through the protocol unified modeling module before being sent out for execution. In the event of network interruption or cloud failure, the protocol unified modeling module, local AI decision-making module, and local energy scheduling module all operate in a closed loop locally. The device control command generation function and the energy consumption optimization command generation function maintain normal local operation and do not rely on real-time cloud computing.

2. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 1, characterized in that, The protocol unified modeling module has a built-in Matter standard device object dictionary and general Cluster mapping rules, supporting device access and bidirectional mapping for four communication protocols: Wi-Fi, Thread, BLE, and Zigbee.

3. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 2, characterized in that, The system also includes a multi-mode adaptive traffic splitting module, which automatically allocates high-bandwidth devices to Wi-Fi links, low-power sensors to Thread links, and network distribution and debugging services to BLE links based on the bandwidth, latency, and power consumption requirements of the terminal devices.

4. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 3, characterized in that, The local AI decision-making module includes a lightweight neural network inference model, which is used to receive three-state data of human dynamic movement, static sitting and no one leaving the site, collected at a preset frequency, and to complete user behavior clustering and habit modeling locally.

5. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 4, characterized in that, The local AI decision-making module also includes a multi-parameter coupling and control unit, which integrates four types of data: ambient light, indoor and outdoor temperature, seasonal time sequence, and user's historical operating habits, to dynamically adjust the color temperature and brightness of lights, the constant temperature range of air conditioning, the fresh air operation level, and the opening and closing angle of curtains.

6. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 5, characterized in that, The local energy dispatch module includes a peak-valley electricity price adaptive dispatch unit, which is used to automatically start water heaters, charging piles and sweeping robot equipment during off-peak electricity price periods, and shut down unnecessary loads during peak electricity price periods.

7. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 6, characterized in that, The local energy dispatch module includes a photovoltaic energy storage collaborative dispatch unit, which is used to prioritize the consumption of photovoltaic power generation to supply the entire house load, store surplus electricity in energy storage devices, and smoothly switch to grid power supply during periods of no sunlight.

8. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 7, characterized in that, The system also includes a data acquisition module, which contains one or more of the following: a millimeter-wave human presence sensor, a temperature and humidity sensor, a light sensor, a water immersion sensor, a gas sensor, and a smoke sensor.

9. The locally autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol according to claim 8, characterized in that, The millimeter-wave human presence sensor is used to collect three-state data of human dynamic movement, static sitting, and no one leaving the site at a preset sampling frequency, and outputs them to the local AI decision-making module.

10. A method for local autonomous decision-making whole-house intelligent closed-loop control based on the Matter 1.5 protocol, used to implement the method of the local autonomous decision-making whole-house intelligent closed-loop control system based on the Matter 1.5 protocol as described in any one of claims 1-9, characterized in that, Includes the following steps: Through the protocol unified modeling module, the private data formats and control commands of terminal devices with different communication protocols are standardized and mapped bidirectionally to establish a local unified communication link. Through the unified communication link, human body status data, environmental parameter data, and user operation behavior data are collected in real time. The local AI decision-making module performs user behavior modeling and scene strategy autonomous iteration on the collected data locally, and generates device control commands. The local energy dispatch module receives at least some of the equipment control instructions and combines them with the whole house load power, photovoltaic power generation and grid peak and valley information to perform energy-saving optimization dispatch locally and generate energy consumption optimization instructions. The device control commands generated by the local AI decision-making module and the energy consumption optimization commands generated by the local energy dispatch module are both converted into control commands that the target device can recognize through the protocol unified modeling module and then sent out for execution.

Images