Intelligent smoke and gas dual-detection alarm device based on multi-mode sensing fusion

Abstract

The invention provides an intelligent smoke and gas dual-detection alarm device based on multi-mode sensing fusion, and the device comprises a housing which is internally provided with a cavity-divided physical isolation structure which comprises a smoke detection cavity and a gas detection cavity; a smoke detection module; a fuel gas detection module; the main control module is electrically connected with the smoke detection module and the fuel gas detection module and is used for receiving and processing data acquired by the smoke detection module and the fuel gas detection module; the intelligent algorithm module is arranged in the main control module, and the intelligent algorithm module is used for executing multi-modal sensing data fusion analysis and dynamic threshold calibration; the alarm response module is electrically connected with the main control module, and the alarm response module is used for executing graded alarm response according to the analysis result of the intelligent algorithm module; a communication module; and a power supply module. According to the invention, through a cavity-divided physical isolation structure, a multi-sensor fusion technology and an intelligent algorithm system, high-precision detection of smoke and fuel gas is realized, and the false alarm rate is reduced.

Description

Technical Field

[0001] The present invention relates to a detection alarm device, in particular to an intelligent smoke and gas dual detection alarm device based on multi-modal sensor fusion, belonging to the technical field of security monitoring. Background Art

[0002] With the acceleration of urbanization, the demand for home security equipment continues to grow. Smoke alarms and gas leak alarms have become standard fire safety features in modern buildings. However, existing technical solutions generally use a standalone device deployment model: smoke alarms are mostly based on ionization or photoelectric sensing principles and are deployed in ceiling areas for fire warning; gas alarms rely on semiconductor or catalytic combustion sensors and are installed near gas pipelines in kitchens and other places to monitor combustible gases such as methane and carbon monoxide.

[0003] This separate design has exposed significant defects in actual applications: First, the purchase of dual equipment leads to a 40%-60% increase in user costs, and the installation process needs to consider issues such as power wiring and wall drilling locations separately. Especially for the renovation of old residential buildings, the coexistence of multiple devices greatly reduces space utilization; second, independently operated sensor systems have a cumulative false alarm rate phenomenon. For example, cooking fumes can easily trigger false alarms in smoke alarms, and volatile organic compounds such as alcohol volatilization may interfere with gas sensor readings. According to statistics, the average false alarm rate of existing equipment is as high as 18%, which seriously undermines user trust; third, the split design hinders the cross-validation of multi-source data. When a compound disaster occurs, such as a gas leak accompanied by electrostatic spark ignition, a single sensor cannot realize risk correlation analysis, resulting in delayed emergency response.

[0004] In recent years, while manufacturers have attempted to improve device performance through multi-parameter sensor integration, these technologies still face inherent limitations. First, the shared detection cavity design causes smoke particles and gas molecules to adsorb to each other, accelerating sensor sensitivity degradation. Typical experimental data shows a drop in detection accuracy of over 30% after six months of operation. Second, the lack of environmental interference compensation mechanisms prevents effective differentiation between real dangers and everyday interference sources, leading to misjudgments, particularly in extreme environments with humidity exceeding 75% or temperatures below 0°C. Furthermore, existing integration solutions generally neglect device compatibility with smart home systems. Most products only support a single local audible and visual alarm, failing to meet the rigid demands of remote early warning and multi-device linkage in the IoT era.

[0005] Therefore, there is an urgent need for an intelligent smoke and gas dual detection alarm device that can solve the above problems. Summary of the Invention

[0006] Based on the above background, the purpose of the present invention is to provide an intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion. Through a cavity-type physical isolation structure, multi-sensor fusion technology and an intelligent algorithm system, high-precision detection of smoke and gas can be achieved, the false alarm rate can be reduced, the equipment life can be extended, and the linkage of smart home systems can be supported.

[0007] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0008] An intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion, comprising:

[0009] A housing having a divided-chamber physical isolation structure disposed therein, the divided-chamber physical isolation structure including a smoke detection chamber and a gas detection chamber;

[0010] a smoke detection module, disposed in the smoke detection chamber, comprising a laser scattering unit, a photoelectric detection unit, and a dust prevention unit;

[0011] A gas detection module is provided in the gas detection chamber, the gas detection module comprising a sensor array, a filter unit and an airflow drive unit;

[0012] a main control module, electrically connected to the smoke detection module and the gas detection module, and configured to receive and process data collected by the smoke detection module and the gas detection module;

[0013] An intelligent algorithm module is provided in the main control module, and is used to perform multimodal sensor data fusion analysis and dynamic threshold calibration;

[0014] An alarm response module, electrically connected to the main control module, and configured to execute a hierarchical alarm response according to the analysis result of the intelligent algorithm module;

[0015] a communication module, electrically connected to the main control module, the communication module being used to implement remote alarm and device linkage; and

[0016] The power supply module is electrically connected to the main control module and is used to provide power.

[0017] Preferably, in the divided-chamber physical isolation structure, the distance between the smoke detection chamber and the gas detection chamber is 5-10 mm, an inclined air inlet is provided on the top of the smoke detection chamber, and an air inlet and an air outlet are provided on the side wall of the gas detection chamber.

[0018] Preferably, the laser scattering unit includes a vertical cavity surface emitting laser, and the emission wavelength of the vertical cavity surface emitting laser is 830-870 nanometers; the photoelectric detection unit includes a photodiode array; the dustproof unit includes a maze-type dustproof grid and a light trap absorption device, and the light trap absorption device adopts a double-layer orthogonal grating structure.

[0019] Preferably, the sensor array includes a methane sensor, a carbon monoxide sensor and a hydrogen sensor, and the sensor array adopts an equilateral triangle layout; the filtration unit includes a molecular sieve filter layer, and the pore size of the molecular sieve filter layer is 2-4 angstroms; the airflow drive unit includes a micro-turbofan, and the micro-turbofan is used to achieve active suction of airflow.

[0020] Preferably, the main control module includes a microcontroller and a memory, and the microcontroller is used to execute a multi-tasking real-time operating system to coordinate the work of various functional modules; the intelligent algorithm module includes a bidirectional long short-term memory neural network and a dynamic threshold calibration system, and the bidirectional long short-term memory neural network has a multi-layer neuron structure, and the dynamic threshold calibration system automatically adjusts the alarm threshold according to environmental conditions.

[0021] Preferably, the alarm response module includes an audible and visual alarm unit, and executing the graded alarm response includes:

[0022] First-level response: triggering local alarms and mobile terminal notifications within 0-10 seconds after detecting danger signals;

[0023] The second level response triggers the linkage control of smart devices within 10-30 seconds after detecting the danger signal;

[0024] The third level response triggers positioning data transmission 30 seconds after detecting a danger signal.

[0025] Preferably, the communication module includes a near field communication unit and a long-distance communication unit, wherein the near field communication unit supports the smart home protocol; the long-distance communication unit supports mobile communication network connection, including a positioning system interface.

[0026] Preferably, the power supply module includes a battery unit and a power management unit, and the power management unit executes a low power management strategy, including an intermittent wake-up mode of the sensor and dynamic power consumption adjustment.

[0027] Preferably, the intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion also includes a self-diagnosis module, which is used to monitor sensor performance and system operating status, and execute a temperature drift compensation algorithm to eliminate the impact of temperature fluctuations on detection accuracy.

[0028] Preferably, the intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion also includes an edge computing module and a multi-device collaborative learning system. The multi-device collaborative learning system adopts a federated learning framework to support parameter sharing and model optimization among multiple devices.

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion of the present invention solves the problem of mutual interference between smoke particles and gas molecules through a cavity-type physical isolation structure; significantly reduces the false alarm rate through multi-sensor fusion and intelligent algorithm system; improves the efficiency of handling dangerous events through a three-level progressive emergency response mechanism; extends the service life of the equipment through low-power design and self-diagnosis function; and achieves seamless integration with smart home systems through the Internet of Things communication function. The present invention integrates smoke detection and gas detection functions, reduces user costs, and simplifies the installation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0032] Figure 1 This is a schematic diagram of an intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to the present invention;

[0033] Figure 2 This is a system architecture diagram of an intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion in the present invention;

[0034] Figure 3 This is a flow chart of the intelligent algorithm of an intelligent smoke and gas dual detection alarm device based on multi-modal sensor fusion of the present invention;

[0035] In the figure: 1. Shell; 2. Smoke detection module; 3. Gas detection module; 4. Main control module; 5. Intelligent algorithm module; 6. Alarm response module; 7. Communication module; 8. Power supply module; 9. Self-diagnosis module; 10. Multi-device collaborative learning system. DETAILED DESCRIPTION

[0036] The technical solution of the present invention will be further described in detail below through specific embodiments and in conjunction with the accompanying drawings. It should be understood that the implementation of the present invention is not limited to the following embodiments, and any form of modification and / or change made to the present invention will fall within the scope of protection of the present invention.

[0037] In the present invention, unless otherwise specified, all parts and percentages are by weight. The equipment and raw materials used are commercially available or commonly used in the art. The methods in the following embodiments, unless otherwise specified, are conventional methods in the art. The components or equipment in the following embodiments, unless otherwise specified, are all universal standard parts or components known to those skilled in the art. Their structures and principles are known to those skilled in the art through technical manuals or routine experimental methods.

[0038] The following detailed description of the embodiments of the present invention is made in conjunction with the accompanying drawings. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present invention. However, one or more embodiments may be implemented by those skilled in the art without these specific details.

[0039] like Figure 1 and Figure 2 As shown, an embodiment of the present invention discloses an intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion, including a shell 1, a smoke detection module 2, a gas detection module 3, a main control module 4, an intelligent algorithm module 5, an alarm response module 6, a communication module 7 and a power supply module 8.

[0040] The housing 1 is made of flame-retardant ABS material and features a separate, physically separate chamber structure that divides the interior into a smoke detection chamber and a gas detection chamber. The smoke detection chamber has a diameter of 25 mm and features a 1.2 mm diameter, fan-shaped air inlet at a 45° angle on the top. The gas detection chamber has a diameter of 18 mm and features an air inlet and outlet on the sidewalls. The spacing between the two detection chambers has been optimized to 8 mm using computational fluid dynamics (CFD) simulations. This design effectively prevents cross-talk between the two detection chambers while maintaining the compactness of the device.

[0041] The smoke detection module 2 is arranged in the smoke detection chamber and includes a laser scattering unit, a photoelectric detection unit and a dustproof unit. The laser scattering unit adopts a vertical cavity surface emitting laser (VCSEL) with a wavelength of 850 nanometers and a beam scattering angle of ±5°, which can accurately capture smoke particles in the air. The photoelectric detection unit adopts a Hamamatsu S1223 photodiode array with high sensitivity and can detect weak scattered light signals. The dustproof unit includes a labyrinth dustproof grid and a light trap absorption device. The labyrinth dustproof grid can filter 99.6% of PM10 particles to prevent dust from interfering with the detection results; the light trap absorption device adopts a double-layer orthogonal grating structure to eliminate the fluctuation of the laser scattering signal caused by airflow disturbances and improve the detection stability.

[0042] The gas detection module 3 is arranged in the gas detection chamber and includes a sensor array, a filter unit and an airflow drive unit. The sensor array includes a TGS5042 methane sensor, a TGS5141 carbon monoxide sensor and an H2S-AH hydrogen sensor. The three sensors are arranged in an equilateral triangle with a spacing of 8 mm between each other and fixed on a ceramic substrate. This layout can reduce thermal interference between sensors. The filter unit uses a molecular sieve filter layer with a pore size of 3 angstroms (3A zeolite molecular sieve). In an environment with a humidity greater than 75%, the adsorption rate of water molecules reaches 98.5%, which can effectively prevent the influence of humidity on gas detection. The airflow drive unit uses an NMB0615 micro-turbo fan with a maximum speed of 8000 rpm, which can achieve an active suction flow rate of 1.2 m / s, shortening the methane enrichment detection time to 2.8 seconds, which is 57% faster than the traditional passive diffusion mode.

[0043] Main Control Module 4 uses the STM32H743VIT6 chip with a main frequency of 480MHz, 2MB of built-in Flash and 1MB of RAM. It runs the FreeRTOS real-time operating system and coordinates the operations of various functional modules. Main Control Module 4 has multiple task threads: a sensor data acquisition thread reads environmental parameters every 1 millisecond; a data processing thread performs signal filtering and preprocessing; and an algorithm inference thread performs risk assessment every 5 seconds. The circuit layout of Main Control Module 4 complies with the IPC-2221A standard to ensure electromagnetic compatibility.

[0044] The intelligent algorithm module 5 includes a bidirectional long short-term memory (LSTM) neural network and a dynamic threshold calibration system. Figure 3 As shown, the bidirectional LSTM neural network consists of a hidden layer of 128 neurons. Its input dimensions are seven-dimensional time series data consisting of smoke concentration, methane / carbon monoxide concentration, temperature and humidity, air pressure, and VOC values. Feature fusion is performed using an attention mechanism weight distribution module. The dynamic threshold calibration system automatically adjusts the alarm threshold based on seasonal characteristics: raising the methane alarm threshold to 15% LEL (Lower Explosive Limit) in winter and returning it to 10% LEL in summer. The threshold calculation is based on temperature sensor data and historical alarm records. Validated in 100,000 sets of scenario data in the laboratory, this algorithm reduced the false alarm rate by 62% compared to the original algorithm when two devices were operating simultaneously.

[0045] Alarm response module 6 includes an audible and visual alarm unit and a three-level progressive emergency response system. The audible and visual alarm unit includes a high-decibel buzzer and LED indicator light, and can emit a 105-decibel pulse alarm. The three-level progressive emergency response system operates as follows: The first-level response is triggered within 0-10 seconds after the danger signal is detected, including a local 105-decibel pulse alarm and mobile app push notification; the second-level response is triggered within 10-30 seconds after the danger signal is detected, closing the smart gas valve and activating the variable-speed exhaust system via the ZigBee 3.0 protocol; the third-level response is triggered 30 seconds after the danger signal is detected, sending encrypted Beidou / GPS dual-mode positioning data (accuracy of ±3 meters) via the NB-IoT module. This progressive response mechanism can take appropriate measures based on the degree of danger, avoiding overreaction or underreaction.

[0046] Communication module 7 includes a near-field communication unit and a long-range communication unit. The near-field communication unit uses the Silicon Labs EFR32MG24 chip, supports ZigBee 3.0, Thread, and Bluetooth 5.0 protocols, and can act as a Thread border router to synchronously control 32 sub-devices. The long-range communication unit uses the Quectel BC95-GL900 module, supports NB-IoT networks, and has low-power wide area network (LPWAN) communication capabilities. Communication module 7 also integrates the Matter protocol stack, supporting interoperability with smart home devices from different brands.

[0047] The power supply module 8 includes a battery unit and a power management unit. The battery unit uses a 2600mAh removable 18650 lithium battery and supports Qi wireless charging (85% efficiency). The power management unit uses the TIBQ25619 chip to achieve dynamic power switching and implement a low-power management strategy: the gas sensor adopts an intermittent wake-up mode with a duty cycle of 0.33% (activated for 0.1 second every 30 seconds), and the smoke sensor is triggered to wake up by the light intensity mutation detection circuit. The overall standby power consumption is reduced to 0.8 milliwatts, and the battery life is extended to 3 years (traditional equipment has an average of 1.5 years). The power module also provides triple circuit protection to prevent overcharging, over-discharging and short circuit.

[0048] The device of this embodiment also includes a self-diagnostic module 9 for monitoring sensor performance and system operating status. The self-diagnostic module 9 monitors sensor sensitivity attenuation in real time and issues a 14-day advance warning of battery replacement requirements when the attenuation rate exceeds 5%. The self-diagnostic module 9 also includes a temperature drift compensation algorithm that eliminates the impact of temperature fluctuations between -20°C and 60°C on detection accuracy. The temperature drift compensation algorithm is based on the sensor's temperature characteristic curve and establishes a mathematical model to automatically adjust the sensor output value, ensuring high-precision detection in extreme temperature environments. Experiments have shown that this technology can control detection errors under extreme temperature conditions to within ±5% of the standard state.

[0049] The self-diagnostic module 9 also supports device health assessment, including battery health monitoring, sensor lifespan assessment, and communication quality analysis. Through regular self-tests and data analysis, the device predicts potential failures and provides maintenance recommendations, significantly improving system reliability. If sensor performance anomalies are detected, the system automatically adjusts operating parameters to maximize normal operation and sends maintenance notifications to the user via the app.

[0050] The apparatus of this embodiment also includes an edge computing module and a multi-device collaborative learning system 10. The edge computing module processes data locally, reducing cloud reliance and improving response speed and privacy protection. The multi-device collaborative learning system 10 uses a federated learning framework to support parameter sharing and model optimization across multiple devices, while also using differential privacy technology (gradient clipping parameter g = 0.5) to protect data security.

[0051] The federated learning framework allows multiple devices to collaboratively optimize detection models without sharing raw data. Each device trains the model using local data and then sends only the updated model parameters (not the raw data) to a central server. The central server aggregates the parameter updates from all devices to form a global model, which is then distributed to each device. This approach protects user privacy while leveraging the advantages of distributed data, allowing the model to adapt to the usage environments and habits of different users.

[0052] Through federated learning, the system significantly reduces its misjudgment rate in kitchen fume aerosol scenarios, significantly improving the device's practicality and user experience. The system also supports online incremental learning, and detection accuracy continues to improve over time.

[0053] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. It should be noted that those skilled in the art can make various improvements and modifications to the present invention without departing from the principles of the present invention, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. An intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion, characterized by: The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion includes: A housing (1), wherein a chamber-type physical isolation structure is provided in the housing (1), and the chamber-type physical isolation structure includes a smoke detection chamber and a gas detection chamber; A smoke detection module (2) is arranged in the smoke detection chamber, and the smoke detection module (2) comprises a laser scattering unit, a photoelectric detection unit, and a dust prevention unit; A gas detection module (3) is arranged in the gas detection cavity, and the gas detection module (3) comprises a sensor array, a filter unit and an airflow drive unit; A main control module (4) is electrically connected to the smoke detection module (2) and the gas detection module (3), and the main control module (4) is used to receive and process data collected by the smoke detection module (2) and the gas detection module (3); An intelligent algorithm module (5) is provided in the main control module (4), and the intelligent algorithm module (5) is used to perform multimodal sensor data fusion analysis and dynamic threshold calibration; An alarm response module (6) is electrically connected to the main control module (4), and the alarm response module (6) is used to perform a hierarchical alarm response according to the analysis result of the intelligent algorithm module (5); a communication module (7), electrically connected to the main control module (4), the communication module (7) being used to realize remote alarm and device linkage; and A power supply module (8) is electrically connected to the main control module (4) and is used to provide power.

2. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: In the divided-chamber physical isolation structure, the distance between the smoke detection chamber and the gas detection chamber is 5-10 mm, the top of the smoke detection chamber is provided with an inclined air inlet, and the side wall of the gas detection chamber is provided with an air inlet and an air outlet.

3. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The laser scattering unit includes a vertical cavity surface emitting laser, and the emission wavelength of the vertical cavity surface emitting laser is 830-870 nanometers; the photoelectric detection unit includes a photodiode array; the dustproof unit includes a labyrinth dustproof grid and a light trap absorption device, and the light trap absorption device adopts a double-layer orthogonal grating structure.

4. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The sensor array includes a methane sensor, a carbon monoxide sensor and a hydrogen sensor, and the sensor array adopts an equilateral triangle layout; the filtration unit includes a molecular sieve filter layer, and the pore size of the molecular sieve filter layer is 2-4 angstroms; the airflow drive unit includes a micro-turbofan, and the micro-turbofan is used to achieve active suction of airflow.

5. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The main control module (4) includes a microcontroller and a memory, and the microcontroller is used to execute a multi-tasking real-time operating system to coordinate the work of various functional modules; the intelligent algorithm module (5) includes a bidirectional long short-term memory neural network and a dynamic threshold calibration system, the bidirectional long short-term memory neural network has a multi-layer neuron structure, and the dynamic threshold calibration system automatically adjusts the alarm threshold according to environmental conditions.

6. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The alarm response module (6) includes an audible and visual alarm unit, and executing a graded alarm response includes: First-level response: triggering local alarms and mobile terminal notifications within 0-10 seconds after detecting danger signals; The second level response triggers the linkage control of smart devices within 10-30 seconds after detecting the danger signal; The third level response triggers positioning data transmission 30 seconds after detecting a danger signal.

7. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The communication module (7) comprises a near field communication unit and a long-range communication unit, wherein the near field communication unit supports the smart home protocol; the long-range communication unit supports mobile communication network connection and includes a positioning system interface.

8. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The power supply module (8) comprises a battery unit and a power management unit, wherein the power management unit executes a low power consumption management strategy, including an intermittent wake-up mode of the sensor and dynamic power consumption adjustment.

9. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion further comprises a self-diagnosis module (9), which is used to monitor sensor performance and system operating status and execute a temperature drift compensation algorithm to eliminate the influence of temperature fluctuations on detection accuracy.

10. The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion according to claim 1 is characterized by: The intelligent smoke and gas dual detection alarm device based on multimodal sensor fusion also includes an edge computing module and a multi-device collaborative learning system (10). The multi-device collaborative learning system (10) adopts a federated learning framework and supports parameter sharing and model optimization among multiple devices.

Images