Ultralow-power-consumption pyroelectric infrared human body detector and networking control method thereof

Abstract

The invention discloses an ultra-low power consumption pyroelectric infrared human body detector and a networking control method thereof. The detector comprises a detection module, a control module, a power supply module, a communication module, an interaction module and a dynamic power consumption management module, human body infrared radiation is detected based on the pyroelectric effect, and ultra-low power consumption operation is achieved by dynamically switching working modes. The networking control method is based on a BLE Mesh protocol, a self-adaptive topology management algorithm is used for optimizing routing, and efficient networking with intelligent equipment is achieved. The system solves the problems of high power consumption, large networking delay, high false alarm rate and the like of a traditional detector, and is suitable for the fields of smart home, intelligent security and protection and the like.

Description

Technical Field

[0001] The present invention relates to the fields of sensor technology and Internet of Things communication, and particularly relates to an ultra-low-power pyroelectric infrared human body detector and its networking control method. Background Art

[0002] As the core front-end device of intelligent security, intelligent lighting and Internet of Things systems, the performance of human body detectors directly affects the level of scene intelligence. Although traditional pyroelectric infrared human body detectors have achieved the basic function of detecting human presence, they have exposed core defects such as high power consumption, insufficient networking ability, and limited detection accuracy in practical applications, making it difficult to meet the requirements of modern intelligent scenarios for long battery life, high scalability, and environmental adaptability of devices.

[0003] (I) Specific Defects of the Existing Technology Power Consumption Control Defect: Dependence on External Power Supply or Frequent Battery Replacement Traditional pyroelectric infrared human body detectors usually need to be externally powered or use large-capacity batteries. In scenarios such as outdoor monitoring and hidden corners of smart homes, external power supplies face difficulties in wiring and high construction costs; when using large-capacity batteries, the battery life of the device is generally only 3-6 months, and the batteries need to be replaced frequently, which not only increases the user's usage cost but also causes electronic waste pollution. For example, due to high power consumption, traditional indoor human body detectors cannot meet the long-term maintenance-free requirements under single-battery power supply, especially in scenarios where maintenance is difficult to perform frequently.

[0004] Networking Ability Defect: Protocol Secrecy and Scale Limitation Traditional detectors mostly use communication protocols such as Zigbee and Wi-Fi, and there are problems that it is difficult to balance low power consumption and large-scale networking: Although the Zigbee protocol supports networking of hundreds of nodes, the chip cost is high and the protocol stack is complex, which is not suitable for miniaturized and low-cost devices; The power consumption of the Wi-Fi protocol is significantly high and cannot meet the low-power requirements of the detector for "long-term dormancy and occasional communication".

[0005] In addition, traditional protocols have strong secrecy, and it is difficult for devices from different manufacturers to interoperate, forming "protocol islands". For example, in large office areas, the networking scale of traditional detectors is usually ≤50 nodes, and the relay delay is relatively high, making it impossible to achieve real-time linkage adjustment of intelligent lights in multiple areas, and the energy-saving advantages of intelligent lighting systems are difficult to fully utilize.

[0006] Detection Accuracy Defect: Insufficient Environmental Interference and Scene Adaptability The detection performance of traditional pyroelectric infrared sensors is significantly affected by environmental temperature, humidity, and electromagnetic interference: When the environmental temperature is close to the human body temperature, the sensitivity of the sensor decreases, and false negatives are likely to occur; Air conditioner airflow and electromagnetic radiation from electrical equipment may cause false triggers.

[0007] In addition, traditional detectors can only output binary signals of "occupied / unoccupied" and cannot extract features such as human movement speed, direction, and residence time, which limits refined scenario applications. For example, in a smart conference room, traditional detectors cannot distinguish between "participants sitting for a long time" and "passing by briefly", resulting in incorrect shutdown of the lighting system or energy waste.

[0008] (II) Limitations of Improvements in the Prior Art The improvement attempts in the prior art for the above problems have obvious shortcomings: Hardware low-power optimization: Only focuses on the selection of the MCU and does not systematically optimize the power consumption distribution of the entire process of "detection - communication - sleep". The standby power consumption of the communication module still accounts for a relatively high proportion. Networking protocol improvement: The hybrid networking solution increases the hardware complexity and cost, and the protocol switching logic is prone to cause communication delays and packet losses. Algorithm optimization: The detection algorithm based on machine learning relies on cloud processing, increases the communication power consumption and response delay of edge devices, and is difficult to meet real-time requirements.

[0009] The prior art has not solved the fundamental contradictions of traditional human detectors in terms of power consumption, networking, and detection accuracy. The present invention provides a front-end device solution with long battery life, high scalability, and high reliability for intelligent Internet of Things scenarios through multi-technology dimension collaborative innovation, and has significant technological progress and practical application value. Summary of the Invention

[0010] In view of this, the purpose of the present invention is to overcome the deficiencies of the prior art and provide an ultra-low-power pyroelectric infrared human detector and its networking control method, aiming to provide a pyroelectric infrared human detector and its networking control method that can achieve ultra-low power consumption, large-scale networking linkage, and accurate human detection through dynamic power management, BLE Bluetooth Mesh intelligent networking, and multi-dimensional detection algorithms, so as to solve the technical bottlenecks of traditional detectors in terms of power consumption, networking ability, and detection accuracy.

[0011] To achieve the above purpose, the first aspect of the present invention provides an ultra-low-power pyroelectric infrared human detector, including: A detection module, which uses a digital pyroelectric infrared sensor, integrates a signal processing circuit and a temperature compensation module, and is used to detect human infrared radiation signals and output characteristic electrical signals. The sensor dynamically adjusts the detection threshold through a voltage division circuit. A control module, including a Bluetooth MCU and a multi-task scheduling engine. The Bluetooth MCU is connected to the detection module through a GPIO interface or an SPI bus, and is used to receive and analyze the characteristic electrical signals, and dynamically switch the following working modes according to the detection status: Low-power standby mode, only maintaining basic detection functions; High-sensitivity detection mode, which is activated when a suspicious signal is detected; Full-function working mode, which activates the communication module after confirming human activities; The multi-task scheduling engine adopts a non-preemptive real-time operating system (RTOS), sets the priority of the detection task to the highest, and dynamically adjusts the priority of the communication task; Power supply module, using a single button battery and a voltage stabilization and filtering circuit to provide stable power for each module; Communication module, integrated in the Bluetooth MCU, supporting the BLEMesh protocol, with an adaptive topology management algorithm built-in, used to establish a Mesh network connection with smart lights and security devices; Interaction module, including an indicator light and an anti-shake button module. The indicator light is driven by the GPIO of the control module and is used to indicate the working mode through the flashing frequency; Dynamic power management module, working in cooperation with the control module, dynamically adjusting the detection threshold and communication frequency according to the ambient light intensity and battery voltage.

[0012] The second aspect of the present invention provides a networking control method based on the above detector, including the following steps: Initialization and Mesh networking: After the detector is powered on, initialize the Bluetooth Mesh protocol stack, send a networking request containing the device ID in broadcast mode, and the master node assigns the node role as a terminal node or a relay node according to the RSSI signal strength; Multi-level detection and mode switching: In the low-power standby mode, the pyroelectric sensor wakes up periodically for detection; When a suspicious signal is detected, switch to the high-sensitivity detection mode, increase the sampling rate and analyze the time-domain characteristics; After confirming human activities, activate the full-function working mode, and send a linkage control instruction to the smart device through Bluetooth Mesh; Dynamic power management: Enter the deep sleep mode when there is no signal continuously, and wake up only through the sensor or the button; Remote interaction and linkage: The user can remotely view the status and adjust the parameters through the mobile phone APP, and the detector is linked with the smart device to execute security or scene control.

[0013] Furthermore, the Mesh networking supports adaptive topology management: The master node dynamically updates the routing table according to the node RSSI value, selects the optimal relay node through the Dijkstra shortest path algorithm, and optimizes the instruction transmission delay.

[0014] Further, the linkage control instruction includes human activity characteristic parameters, including the moving direction and the residence time. After receiving the instruction, the intelligent device executes actions in combination with preset rules, including automatically adjusting the light to a preset brightness to save energy when it detects that a person has stayed for more than a preset duration.

[0015] Further, it also includes the following steps: Multi-modal feature extraction and classification: Extract the time-frequency features of the signal through the STFT transform, analyze the waveform complexity in combination with a machine learning model, construct a human activity feature vector and classify it; Dynamic Mesh network optimization: Use a reinforcement learning algorithm to adjust the relay node strategy, combine with adaptive channel hopping to avoid interference sources, and improve communication reliability; Context-aware linkage control: Construct a spatio-temporal association model to infer the movement path of people, and perform intelligent scene switching based on the residence time and activity type; Predictive power consumption management: Predict the probability of human activities based on historical data, dynamically adjust the detection period and implement a hierarchical sleep strategy.

[0016] Further, the multi-modal feature extraction and classification includes: Extract the time-frequency features of the pyroelectric signal through the short-time Fourier transform (STFT), and analyze the energy distribution in the frequency band of 0.1 - 10 Hz; Adopt a sliding window algorithm to extract the time-domain features of the signal; including the rising edge slope and the pulse width; Classify the signal feature vector based on a convolutional neural network (CNN) model to distinguish human activities from pet and curtain shaking interferences.

[0017] Further, the dynamic Mesh network optimization includes: The master node collects the RSSI value, remaining power, and load rate of each node every 2 minutes to construct a network state matrix; Calculate the relay node selection strategy through the Q-learning algorithm, and the objective function is to minimize the weighted sum of transmission delay and power consumption; Adopt the frequency hopping spread spectrum technology to switch the communication channel every 10 seconds to avoid interference sources in the 2.4 GHz frequency band.

[0018] Further, the context-aware linkage control includes: Based on the trigger time difference and signal strength difference of adjacent detectors, calculate the personnel position coordinates through the triangulation algorithm; Construct a time-location-activity type association rule library; According to the personnel residence time and activity type, automatically trigger scene modes; including turning on the range hood and adjusting the air conditioner temperature.

[0019] Furthermore, the predictive power consumption management includes: Using a long short-term memory network to learn the personnel activity data of historical days and predicting the activity probability curve for the next 24 hours; Dynamically adjusting the detection period according to the prediction result: High-probability period: Set the detection period to 1 second and the sampling rate to 5 Hz; Low-probability period: Extend the detection period to 5 seconds and the sampling rate to 0.5 Hz; When there is no prediction data, adopt the default period of 3 seconds and the uniform sampling strategy.

[0020] The present invention adopts the above technical solutions and has the following beneficial effects: Through the innovation of dynamic power consumption management, intelligent networking protocol and precise detection algorithm, the present invention achieves the following core technical effects: Ultra-low power consumption and long battery life: Powered by a single button battery, through three-level working mode switching (low-power standby / high-sensitivity detection / full-function operation) and dynamic power consumption management, the overall power consumption of the whole machine is significantly reduced, meeting the long-term maintenance-free requirement and solving the problem that traditional devices rely on external power supplies or frequent battery replacement.

[0021] Efficient Mesh networking and cross-device linkage: Based on the BLEMesh protocol, large-scale node networking is realized. Through adaptive topology management, the relay path is optimized, supporting low-latency linkage with intelligent lights and security devices; compatible with cross-protocol gateways, breaking the brand ecological barrier and enhancing the interoperability of Internet of Things devices.

[0022] Precise detection and anti-interference ability: Integrate a temperature compensation module and a dual-element array sensor to reduce the influence of environmental temperature, humidity and electromagnetic interference. Combine signal feature analysis algorithms (such as STFT transform, machine learning classification) to achieve multi-dimensional detection of human movement direction, residence time, etc., improving the scene adaptability and reliability.

[0023] Intelligent scene control and user experience: Support remote parameter configuration and local status indication. The linkage instruction includes human activity feature parameters, and intelligent devices can perform dynamic adjustment according to preset rules (such as light brightness, security response), taking into account energy conservation and user-friendly experience. Description of the Drawings

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0025] Figure 1Schematic diagram of the structure of the ultra-low power pyroelectric infrared human body detector of the present invention; Figure 2 Schematic diagram of the process steps of the present invention. Detailed implementation manners

[0026] In the following, only some exemplary embodiments are briefly described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present invention. Therefore, the drawings and descriptions are considered to be exemplary in nature rather than restrictive.

[0027] Embodiment 1 This embodiment provides an ultra-low power pyroelectric infrared human body detector, and its structure and connection relationship can be referred to in the attached Figure 1 (The pyroelectric human body detector is respectively connected to a pyroelectric module, an indicator light module, a CR2450 button battery, and a KEY button module through a Bluetooth MCU, and communicates and interacts with intelligent devices such as intelligent lamp 1, intelligent lamp 2, and intelligent lamp 3 through Mesh networking). The specific composition is as follows: Detection module: Adopt a digital pyroelectric infrared sensor, integrated with a signal processing circuit and a temperature compensation module, for detecting human infrared radiation signals and outputting characteristic electrical signals, and the sensor dynamically adjusts the detection threshold through a voltage dividing circuit.

[0028] Specifically, the digital pyroelectric infrared sensor is of the PR2231D model, works based on the pyroelectric effect, and is connected to the PA port of the Bluetooth MCU through the OUT pin. The sensor internally integrates a dual-element array structure, and the two sensing elements are orthogonally arranged, and environmental interference is eliminated through a differential signal processing algorithm, and the moving direction and speed of the human body can be detected simultaneously. The voltage dividing circuit is used to adjust the detection sensitivity threshold of the SET pin. For example, in a strong light environment, the threshold is automatically increased by 0.5V to reduce false triggering.

[0029] Control module: Includes a Bluetooth MCU and a multi-task scheduling engine. The Bluetooth MCU is connected to the detection module through a GPIO interface or an SPI bus, for receiving and analyzing characteristic electrical signals, and dynamically switching the following working modes according to the detection state: Low-power standby mode, only maintaining the basic detection function; High-sensitivity detection mode, which is started when a suspicious signal is detected; Full-function working mode, which activates the communication module after confirming human activities; The multi-task scheduling engine adopts a non-preemptive real-time operating system (RTOS), sets the detection task priority to the highest, and dynamically adjusts the communication task priority.

[0030] As a preferred embodiment, the Bluetooth MCU in this embodiment is the CH571F chip, which integrates a 32MHz crystal oscillator to provide a stable clock signal and is connected to the LED indicator through the chip's ports. The multitask scheduling engine is based on the FreeRTOS operating system. In the low-power standby mode, the CH571F controls the sensor to wake up and detect at a cycle of 500ms, and the overall power consumption of the whole machine is ≤15μA; when a suspicious signal is detected (signal strength > threshold and duration > 200ms), it switches to the high-sensitivity mode, the sampling rate is increased to 10Hz and the time-domain characteristics are analyzed (such as rising-edge slope, pulse width).

[0031] Power supply module: It uses a single button battery and a voltage stabilizing and filtering circuit to provide a stable power supply for each module.

[0032] Specifically, the button battery is of the CR2450 model. The power input is converted into a stable VCC voltage through an LDO voltage regulator (L8321), and the noise is filtered through a π-type filter network. The dynamic power consumption management module monitors the battery voltage in real time. When the voltage is lower than 2.8V, the communication frequency is extended to 1 time per 5 seconds; combined with a photosensitive resistor to detect the ambient light intensity, the detection threshold is automatically reduced by 0.3V in low-light environments to further optimize the energy consumption.

[0033] Communication module: Integrated within the Bluetooth MCU, it supports the BLEMesh protocol and has a built-in adaptive topology management algorithm for establishing a Mesh network connection with smart lights and security devices.

[0034] Furthermore, the communication module follows the BLE Bluetooth Mesh v1.3 protocol, establishes a multi-hop network with smart lamps through the on-board antenna (ANT). The master node dynamically updates the routing table every 5 minutes according to the RSSI value, and uses the Dijkstra shortest path algorithm to select the optimal relay node, and the instruction transmission delay is ≤50ms. It supports cross-protocol gateway access and is compatible with Zigbee3.0 devices to achieve seamless linkage with smart devices of different brands.

[0035] Interaction module: It includes an indicator light and an anti-shake button module. The indicator light is driven by the GPIO of the control module and is used to indicate the working mode through the blinking frequency.

[0036] Specifically, the indicator light module uses a two-color LED, which is driven by the GPIO of the CH571F through PWM modulation. It blinks once per second during normal operation and blinks quickly 3 times when a person is detected; the KEY button module includes a micro switch and an RC filter circuit. When pressed, it sends a manual instruction to the CH571F, and the delay for eliminating the interference of button jitter is ≤20ms, supporting the switching of the detection mode or the triggering of parameter reset.

[0037] New Modules and Functions: Debug Interface Module: A Header611 debugging interface is reserved and connected to external devices through RX and TX pins for program downloading, fault troubleshooting, and function debugging, facilitating developers to optimize the performance of the detector.

[0038] Structural Description Overall Architecture: The detector consists of a pyroelectric module, a CH571F Bluetooth MCU, a CR2450 button battery, an indicator light module, a KEY button module, and a Debug interface, and interacts with intelligent lamps through BLE Bluetooth Mesh networking. As shown in the attached figure, each module is connected by circuits to form a low-power hardware system.

[0039] Core Control Unit: CH571F Bluetooth MCU: As the core processor, it is connected to sensors, indicator lights, and button modules through pins, and has a built-in Bluetooth protocol stack to implement Mesh networking, and supports program burning and debugging through the Debug interface.

[0040] Crystal Oscillator): Provides a 32MHz clock signal for the CH571F to ensure the chip timing accuracy.

[0041] Power Supply Module: CR2450 Button Battery: Provides 3V raw power, which is converted to a stable 3.3V voltage (VCC) by an LDO voltage regulator (L8321).

[0042] Filter Circuit: Suppresses power supply ripple and ensures the stable operation of each module.

[0043] Detection and Interaction Module: PR2231D Sensor: Transmits detection signals to the CH571F through the OUT pin, and the voltage division circuit adjusts the sensitivity.

[0044] Indicator Light (LED) and Button (JS1): Are electrically connected to the CH571F respectively to achieve status indication and manual control.

[0045] Principle Description: Human Detection Principle: The PR2231D sensor is based on the pyroelectric effect. When the human body's infrared radiation irradiates the dual-element array sensing element, the temperature change of the pyroelectric material generates a charge difference, which is amplified and filtered by the internal circuit to form an electrical signal (in the 0.1 - 10Hz frequency band) and is transmitted to the CH571F through the pin. The CH571F determines whether a detection event is triggered by comparing the signal strength with a preset threshold (dynamically adjusted by the voltage division circuit).

[0046] Bluetooth Communication Principle: The CH571F is built-in with the BLEMesh protocol stack and completes device discovery and networking by transmitting broadcast packets through the antenna. The master node assigns node roles (terminal / relay) based on the RSSI value, optimizes the routing using the Dijkstra algorithm, and ensures that commands (such as "turn on the light") are transmitted with low latency (≤50ms) through multi-hop paths. Adaptive channel hopping is supported during communication to avoid interference sources in the 2.4GHz frequency band.

[0047] Power Management Principle Dynamic Power Consumption Grading: In the low-power standby mode, the CH571F enters the STOP2 sleep state, only retaining the RTC and wake-up circuit, with a power consumption ≤5μA; in the full-function mode, the Bluetooth module and the sensor are activated for high-frequency sampling, with a power consumption ≤5mA.

[0048] Hybrid Energy Strategy: A micro photovoltaic battery and a super capacitor (0.1F / 5.5V) can be optionally configured. When there is sufficient light, it automatically switches to solar power supply and stores energy, extending the battery life.

[0049] Manual Control Principle: When the user presses the KEY micro switch (JS1), after the RC filter circuit eliminates the jitter, the signal triggers an interrupt of the CH571F through the GPIO pin, and preset operations such as mode switching and parameter reset are executed.

[0050] Explanation of Action Relationships Startup and Initialization: After installing the CR2450 battery, the LDO voltage regulator supplies power to the CH571F. The chip initializes the Bluetooth protocol stack and the sensor interface, and the indicator light flashes 3 times to indicate the completion of startup. The master node scans for available networks, sends a networking request containing the device ID, and is assigned a unique node address after joining the Mesh network.

[0051] Human Detection and Linkage Process Low-Power Standby: The sensor wakes up at a 500ms interval to detect human signals (power consumption ≤15μA).

[0052] High-Sensitivity Detection: When the signal strength > the threshold and lasts >200ms, the CH571F wakes up the Bluetooth module, increases the sampling rate to 10Hz, and analyzes the time-domain characteristics of the signal (such as pulse width) to confirm human activities.

[0053] Full-Function Operation: Send a command (such as "adjust the brightness to 70%") to the smart light through the Mesh network, the indicator light flashes quickly 3 times, and the smart light executes the action and feedbacks the status (such as the change in light brightness).

[0054] Sleep and Wake-up Mechanism: When there is no signal for 30 consecutive minutes, CH571F enters the deep sleep mode (power consumption ≤ 10 μA), and only periodic wake-up by the sensor or trigger by the button can wake up the device. After waking up, CH571F resynchronizes the Mesh network status and resumes the normal detection process.

[0055] Remote Interaction and Maintenance: The user sends instructions through the mobile phone APP via the Mesh network, and CH571F adjusts parameters such as detection thresholds and communication frequencies after parsing (such as adjusting the sensitivity from "high" to "medium"); developers download logs through the Debug interface, analyze the device operation status, and optimize the program.

[0056] Embodiment 2 As Figure 2 shown, this embodiment provides a networking control method based on the above detector, including Initialization and Mesh Networking: After the detector is powered on, the Bluetooth Mesh protocol stack is initialized, and a networking request containing the device ID is sent in broadcast mode. The master node assigns the node role as an end node or a relay node according to the RSSI signal strength; Multi-level Detection and Mode Switching: In the low-power standby mode, the pyroelectric sensor wakes up periodically for detection; When a suspicious signal is detected, switch to the high-sensitivity detection mode, increase the sampling rate, and analyze the time-domain characteristics; After confirming human activities, activate the full-function working mode, and send linkage control instructions to the smart device via Bluetooth Mesh; Dynamic Power Consumption Management: Enter the deep sleep mode when there is no signal continuously, and wake up only through the sensor or the button; Remote Interaction and Linkage: The user remotely views the status and adjusts parameters through the mobile phone APP, and the detector and the smart device are linked to execute security or scene control.

[0057] As an implementation method, the Mesh networking in this embodiment supports adaptive topology management: The master node dynamically updates the routing table according to the node RSSI value, selects the optimal relay node through the Dijkstra shortest path algorithm, and optimizes the instruction transmission delay.

[0058] As an implementation method, the linkage control instruction in this embodiment includes human activity characteristic parameters, including the moving direction and the residence time. After receiving the instruction, the smart device executes actions in combination with the preset rules, including automatically adjusting the light to the preset brightness to save energy when it detects that a person has stayed for more than the preset duration.

[0059] As an implementation method, this embodiment further includes the following steps: Multi-modal Feature Extraction and Classification: Extract the time-frequency features of the signal through the STFT transform, analyze the waveform complexity in combination with a machine learning model, construct a human activity feature vector and classify it; Dynamic Mesh Network Optimization: Use a reinforcement learning algorithm to adjust the relay node strategy, combine with adaptive channel hopping to avoid interference sources, and improve communication reliability; Context-Aware Linkage Control: Construct a spatio-temporal association model to infer the movement path of personnel, and perform intelligent scene switching based on the residence time and activity type; Predictive Power Consumption Management: Predict the probability of personnel activities based on historical data, dynamically adjust the detection period and implement a hierarchical sleep strategy. Train historical data through an LSTM model to determine that a 60% activity probability is the critical value for whether personnel activities are frequent or not.

[0060] As an implementation, the multi-modal feature extraction and classification described in this implementation include: Extract the time-frequency features of the pyroelectric signal through the short-time Fourier transform (STFT), and analyze the energy distribution in the frequency band of 0.1 - 10 Hz; Adopt a sliding window algorithm to extract the time-domain features of the signal; including the rising edge slope and pulse width; Classify the signal feature vector based on a convolutional neural network (CNN) model to distinguish human activities from interference by pets and curtain shaking.

[0061] As an implementation, the dynamic Mesh network optimization described in this implementation includes: The master node collects the RSSI value, remaining power, and load rate of each node every 2 minutes to construct a network state matrix; Calculate the relay node selection strategy through the Q-learning algorithm, and the objective function is to minimize the weighted sum of transmission delay and power consumption; Adopt the frequency-hopping spread-spectrum technology to switch the communication channel every 10 seconds to avoid interference sources in the 2.4 GHz frequency band.

[0062] Furthermore, the context-aware linkage control includes: Based on the trigger time difference and signal strength difference of adjacent detectors, calculate the personnel position coordinates through the triangulation algorithm; Construct a time-location-activity type association rule library; Automatically trigger scene modes according to the personnel residence time and activity type; including turning on the range hood and adjusting the air conditioner temperature.

[0063] As an implementation, the predictive power consumption management described in this implementation includes: Use a long short-term memory network to learn the personnel activity data of historical days and predict the activity probability curve for the next 24 hours; Dynamically adjust the detection period according to the prediction result: High-probability period: When the activity probability is ≥60%, the detection period is set to 1 second and the sampling rate is 5Hz; Low-probability period: When the activity probability is <60%, the detection period is extended to 5 seconds and the sampling rate is 0.5Hz; When there is no prediction data, the default period of 3 seconds and the uniform sampling strategy are adopted.

[0064] The ultra-low-power pyroelectric infrared human body detector and its networking control method provided by the present invention achieve the following remarkable technical effects through the collaborative innovation of the hardware architecture, networking protocol and detection algorithm: I. Ultra-low-power operation and long battery life Dynamic power consumption hierarchical management By scheduling the priorities of detection and communication tasks through a non-preemptive RTOS and dynamically switching among three working modes (low-power standby, high-sensitivity detection, full-function operation), the average power consumption of the whole machine is significantly reduced. In the low-power standby mode, only the basic detection function is maintained. When in deep sleep, unnecessary circuits are turned off, and the detection threshold is dynamically adjusted in cooperation with the voltage-dividing circuit. A single button battery can meet the requirements of long-term stable operation, solving the problems of traditional detectors relying on external power supplies or frequent battery replacement.

[0065] The dynamic power management module optimizes the power consumption allocation in real time according to the ambient light intensity and battery voltage. For example, the communication interval is extended in the low-voltage state, and the detection threshold is increased in the strong light environment to reduce false triggers, further extending the battery life and being suitable for maintenance-free scenarios such as outdoor monitoring and hidden corners of smart homes.

[0066] Hybrid energy self-supply design An energy harvesting circuit (such as a micro photovoltaic battery and a super capacitor) can be optionally configured to automatically switch to the self-powered mode and store energy when the light is sufficient, forming a dual-power system of "light energy self-supply + battery backup", further enhancing the battery life of the device in extreme environments and reducing the battery replacement frequency and electronic waste pollution.

[0067] II. Efficient Mesh networking and cross-ecosystem linkage Large-scale networking and dynamic routing optimization The communication module is integrated into the Bluetooth MCU, supporting the BLEMesh protocol and the adaptive topology management algorithm. The optimal relay node is dynamically selected through the Dijkstra shortest path algorithm to achieve low-latency transmission of instructions in a multi-hop network (such as the response time of the smart light linkage instruction is significantly shortened). The master node regularly updates the routing table, optimizes the communication path according to the node RSSI value, and supports a large number of nodes to work together within a single network, breaking through the limitations of traditional protocols in terms of networking scale and latency.

[0068] Compatible with cross - protocol gateway technology, it can establish Mesh network connections with intelligent terminals of different protocols such as smart lights and security devices, breaking the "protocol island" and achieving seamless interoperability of cross - brand and cross - protocol devices, meeting the diverse linkage requirements of scenarios such as smart home and smart security.

[0069] Anti - interference communication and reliability improvement Adopt reinforcement learning algorithms (such as Q - learning) to dynamically adjust the relay node strategy, combined with adaptive channel hopping technology to avoid interference sources in the 2.4GHz frequency band (such as WiFi, microwave ovens), improve the communication success rate and anti - interference ability, ensure stable communication in complex electromagnetic environments, and solve the problems of packet loss and delay caused by interference in traditional networking.

[0070] III. Precise detection and scene - based intelligent control Environmental adaptive detection and anti - interference ability The detection module uses a digital pyroelectric infrared sensor, integrated with a signal processing circuit and a temperature compensation module. It real - time monitors the environmental temperature through a thermistor network and dynamically calibrates the sensor sensitivity, solving the problem of missed detection of traditional detectors due to the environmental temperature being close to the human body temperature. The dual - element array structure combined with the differential signal processing algorithm effectively distinguishes the moving direction and speed of the human body, eliminates false triggers from interference sources such as pets and curtain vibrations, and improves the reliability of detection results.

[0071] Extract the time - frequency characteristics of the signal through STFT transformation, analyze the waveform complexity combined with a machine learning model (such as a convolutional neural network), construct a human activity feature vector and classify it (such as walking, sitting still, waving, etc.), realizing the upgrade from "simple presence detection" to "behavior feature recognition", and meeting the requirements of refined detection in scenarios such as smart conference rooms and office areas.

[0072] Spatio - temporal correlation analysis and intelligent linkage The control module infers the personnel movement path through the triangulation algorithm based on the trigger time difference and signal strength difference of adjacent detectors, and constructs a spatio - temporal correlation rule library (such as "frequent micro - movements in the kitchen area trigger the range hood to turn on") combined with the dwell time and activity type. After receiving instructions containing feature parameters such as moving direction and dwell time, intelligent devices can perform scene - based linkage operations such as dynamic lighting adjustment and security alarm, improving the humanization and energy - saving effect of the intelligent system.

[0073] Predictive power consumption optimization and proactive service The predictive power consumption management module predicts the probability of human activities based on historical activity data (such as learning through an LSTM neural network), and dynamically adjusts the detection period and sampling rate: increasing the detection frequency during high-probability periods to ensure response speed, and extending the sleep period during low-probability periods to reduce power consumption. This "wake-on-demand" mechanism further optimizes the overall power consumption while ensuring detection accuracy, achieving a balance between device performance and energy efficiency.

[0074] IV. Hardware Integration and Interaction Experience Optimization Compact Low-Power Hardware Architecture The core control unit (such as the CH571F Bluetooth MCU) integrates rich peripheral interfaces, communicates in real-time with detection modules, interaction modules, etc. through GPIO or SPI buses, and cooperates with the crystal oscillator to provide a stable clock signal to ensure the timing accuracy of the coordinated operation of each module. The combined design of the voltage regulation and filtering circuit and the button battery realizes the miniaturization and low-power compatibility of the hardware system, suitable for installation scenarios with limited space.

[0075] Convenient Interaction and Remote Management The interaction module drives the indicator light through GPIO to indicate the working mode at different blinking frequencies, and supports local parameter settings in combination with the anti-shake button module; at the same time, it supports remote viewing of the device status, adjustment of detection sensitivity and linkage rules on the mobile phone APP, and cooperates with the OTA firmware upgrade and configuration rollback mechanism to improve the convenience of user operation and the security of system management.

[0076] V. Systematic Breakthrough of the Technical Solution Through the deep integration of "hardware low-power design + intelligent networking protocol + multi-dimensional detection algorithm", the present invention systematically solves the contradictions of traditional human detectors in terms of power consumption, networking ability, and detection accuracy: In terms of power consumption: It breaks through the dependence of traditional devices on large-capacity batteries or external power supplies, achieving long single-battery life and energy self-supply; In terms of networking: Based on BLEMesh, it constructs a large-scale, low-latency, and anti-interference Internet of Things network, supporting cross-protocol ecological integration; In terms of detection: Through environmental compensation, feature analysis, and positioning algorithms, it realizes an intelligent leap from "binary detection" to "scene understanding".

[0077] The collaborative realization of the above technical effects enables the present invention to have significant application value in the fields of intelligent security, intelligent lighting, smart home, etc., promoting the development of front-end devices towards low power consumption, high reliability, and self-evolution, and meeting the core needs of intelligent terminals in the Internet of Things era.

[0078] The ultra-low-power pyroelectric infrared human body detector of this embodiment realizes the functions of low-power and intelligent human body detection through the collaborative work of each module, based on their respective working principles, and can meet the requirements of various application scenarios such as smart home, providing users with a more convenient, safe and comfortable living experience.

Claims

1. An ultra-low power pyroelectric infrared human body detector, characterized in that, Comprising: A detection module, which uses a digital pyroelectric infrared sensor, integrates a signal processing circuit and a temperature compensation module, and is used to detect human infrared radiation signals and output characteristic electrical signals. The sensor dynamically adjusts the detection threshold through a voltage division circuit; A control module, which includes a Bluetooth MCU and a multi-task scheduling engine. The Bluetooth MCU is connected to the detection module through a GPIO interface or an SPI bus, and is used to receive and analyze the characteristic electrical signals, and dynamically switch the following working modes according to the detection status: A low-power standby mode, which only maintains the basic detection function; A high-sensitivity detection mode, which is started when a suspicious signal is detected; A full-function working mode, which activates the communication module after confirming human activities; The multi-task scheduling engine uses a non-preemptive real-time operating system (RTOS), sets the priority of the detection task to the highest, and dynamically adjusts the priority of the communication task; A power supply module, which uses a single button battery and a voltage stabilizing and filtering circuit to provide a stable power supply for each module; A communication module, which is integrated in the Bluetooth MCU, supports the BLEMesh protocol, and has an adaptive topology management algorithm built-in, and is used to establish a Mesh network connection with intelligent lights and security devices; An interaction module, which includes an indicator light and an anti-shake button module. The indicator light is driven by the GPIO of the control module and is used to indicate the working mode through the flashing frequency; A dynamic power consumption management module, which works in cooperation with the control module and dynamically adjusts the detection threshold and communication frequency according to the ambient light intensity and battery voltage.

2. A networking control method for the detector according to claim 1, characterized in that, Including the following steps: Initialization and Mesh networking: After the detector is powered on, initialize the Bluetooth Mesh protocol stack, and send a networking request containing the device ID in broadcast mode. The master node assigns the node role as a terminal node or a relay node according to the RSSI signal strength; Multi-level detection and mode switching: In the low-power standby mode, the pyroelectric sensor wakes up periodically for detection; When a suspicious signal is detected, switch to the high-sensitivity detection mode, increase the sampling rate and analyze the time-domain characteristics; After confirming human activities, activate the full-function working mode, and send a linkage control instruction to the intelligent device through Bluetooth Mesh; Dynamic power consumption management: Enter the deep sleep mode when there is no signal continuously, and wake up only through the sensor or the button; Remote interaction and linkage: The user remotely views the status and adjusts the parameters through the mobile phone APP, and the detector and the intelligent device are linked to execute security or scene control.

3. The method according to claim 2, characterized in that, The Mesh networking supports adaptive topology management: The master node dynamically updates the routing table according to the node RSSI value, and selects the optimal relay node through the Dijkstra shortest path algorithm to optimize the instruction transmission delay.

4. The method according to claim 2 or 3, characterized in that, The linkage control instruction contains human activity characteristic parameters, including the moving direction and the dwelling time. After receiving the instruction, the intelligent device executes actions in combination with the preset rules, including automatically adjusting the light to the preset brightness to save energy when it is detected that a person has stayed for more than the preset duration.

5. The method according to claim 2, wherein It also includes the following steps: Multi-modal feature extraction and classification: Extract the time-frequency features of the signal through STFT transformation, analyze the waveform complexity in combination with the machine learning model, and construct and classify the human activity feature vector; Dynamic Mesh Network Optimization: Use reinforcement learning algorithms to adjust the relay node strategy, combined with adaptive channel hopping to avoid interference sources, and improve communication reliability; Context-Aware Linkage Control: Construct a spatio-temporal association model to infer the movement path of personnel, and perform intelligent scene switching based on the dwell time and activity type; Predictive Power Consumption Management: Predict the probability of personnel activities based on historical data, dynamically adjust the detection period, and implement a hierarchical sleep strategy.

6. The method according to claim 5, characterized in that, The multi-modal feature extraction and classification include: Extract the time-frequency features of pyroelectric signals through Short-Time Fourier Transform (STFT), and analyze the energy distribution in the frequency band of 0.1 - 10 Hz; Use the sliding window algorithm to extract the time-domain features of the signal; including the rising edge slope and pulse width; Classify the signal feature vectors based on the Convolutional Neural Network (CNN) model to distinguish human activities from pet and curtain movement interferences.

7. The method according to claim 5, characterized in that The dynamic Mesh network optimization includes: The master node collects the RSSI values, remaining battery power, and load ratio of each node every 2 minutes to construct a network state matrix; Calculate the relay node selection strategy through the Q-learning algorithm, and the objective function is to minimize the weighted sum of transmission delay and power consumption; Adopt frequency hopping spread spectrum technology to switch the communication channel every 10 seconds to avoid interference sources in the 2.4 GHz frequency band.

8. The method according to claim 5, characterized in that, The context-aware linkage control includes: Calculate the personnel position coordinates through the triangulation algorithm based on the trigger time difference and signal strength difference of adjacent detectors; Construct a time-location-activity type association rule library; Automatically trigger scene modes according to the personnel dwell time and activity type; including turning on the range hood and adjusting the air conditioner temperature.

9. The method according to claim 5, wherein The predictive power consumption management includes: Use a long short-term memory network to learn the personnel activity data of historical days and predict the activity probability curve for the next 24 hours; Dynamically adjust the detection period according to the prediction results: High probability period: The detection period is set to 1 second, and the sampling rate is 5 Hz; Low probability period: The detection period is extended to 5 seconds, and the sampling rate is 0.5 Hz; When there is no prediction data, use the default period of 3 seconds and uniform sampling strategy.

Images