A fire source recognition device based on STM32F4

Abstract

The utility model discloses a fire source recognition device based on STM32F4, include: main control circuit adopts STM32F4 series microcontroller, visible light camera unit includes image sensor module, and it is connected with the DCMI interface of main control circuit through parallel video interface, and is equipped with subdomain voltage stabilizing power supply circuit, infrared thermal imaging unit includes infrared thermal imaging sensor array, through I 2 C bus is connected with main control circuit, and is equipped with level conversion and pull -up circuit, smoke sensing unit includes gas sensor, and its output is connected to the ADC channel of main control circuit after conditioning, wireless communication unit includes wi the -Fi module, and communicates with main control circuit through UART, and is equipped with bootstrap circuit, power management unit includes lithium cell and voltage reducing module, and is equipped with the gate control power supply circuit for visible light camera unit. The device has the remarkable characteristics of low power consumption, low false alarm, fast response, easy mass production.

Description

Technical Field

[0001] This utility model relates to the field of fire source identification technology, and more specifically, to a fire source identification device based on STM32F4. Background Technology

[0002] Existing fire detection technologies mainly fall into three categories: point-type smoke / heat detectors (such as ionization and photoelectric detectors), infrared / ultraviolet flame sensors, and video image recognition. Point-type smoke detectors are sensitive to smoke particles, but their response is delayed or prone to false alarms in environments with strong ventilation, high dust levels, heavy oil fumes, or early smoldering (without obvious smoke plumes). Heat detectors rely on ambient temperature rise, and are not easily triggered by small early fires. Single infrared / ultraviolet flame sensors have limited ability to distinguish scenes with high reflectivity, welding arcs, and dappled sunlight. Video algorithms can identify flames through color, shape, motion, and flicker characteristics; however, complex lighting, low nighttime illumination, specular reflection, vehicle lights / LED screens, etc., can lead to high false alarm rates, and purely visual solutions are not sensitive to obstructions.

[0003] To improve robustness, the industry has proposed a multi-sensor fusion approach (visible light + thermal + gas): infrared thermal imaging can provide hotspot information independent of ambient light, while gas / smoke sensing can characterize combustion byproducts, thus complementing visual deficiencies. However, existing fusion systems often face challenges.

[0004] ① Synchronization and fusion difficulties: inconsistent frame rates / latencies / fields of view across different modalities, lack of stable temporal consistency judgment; missing or complex spatial calibration of thermal images and visible light, making it difficult to accurately project hot spots into visual ROIs.

[0005] ② Power consumption and size: Video solutions using high-performance SoC / DSP have high power consumption, cost and heat dissipation pressure, which is not conducive to battery power supply and small-size terminals.

[0006] ③ Insufficient suppression of false alarms: Heat sources (heaters, pipes, focused sunlight) and actual flames are both high-temperature areas in thermal images. Without a combined criterion of color / flicker / smoke, it is difficult to reduce false alarms.

[0007] ④ Field adaptability: Gas-sensitive devices (such as MQ-2) are affected by preheating, drift, humidity and temperature, and individual differences. They lack self-learning baseline and cross-scenario parameter tuning, resulting in insufficient long-term stability.

[0008] ⑤ Interface and Electromagnetic Compatibility: The parallel video bus (PCLK / VSYNC / HREF / data line) of the image sensor is sensitive to clock jitter and wiring integrity; thermal imaging and expansion peripherals often require I / O. 2 C-level conversion and isolation adaptation are used to improve bus stability; improper configuration of the power-on bootstrap pin of the wireless module can easily cause startup abnormalities.

[0009] ⑥ Power supply and timing: Transient currents from peak loads such as cameras and wireless devices can easily cause the 3.3V mains to drop, triggering main control reset or image frame loss; lack of gating power-on and power-on sequence management for camera units affects system reliability.

[0010] ⑦ Operation and Mass Production: The lack of standardized external interfaces and on-site / production calibration processes makes it difficult to achieve batch consistency and rapid maintenance; EMC / safety design has shortcomings, affecting certification pass rate and yield.

[0011] On the other hand, using a microcontroller with DCMI (Digital Camera Interface) (such as the STM32F4) allows for the connection of parallel cameras with low power consumption and low cost, enabling DMA double-buffered zero-copy acquisition. However, issues still need to be addressed: pixel clock (XCLK / 24MHz) quality, PCLK sampling edge, VSYNC / HREF timing matching, and buffer bandwidth allocation; simultaneously, it needs to be compatible with I... 2 The C-type thermal imaging array, ADC gas-sensitive signal, and UART wireless link work in parallel to avoid mutual blocking and resource contention. This places systematic design requirements on power timing, signal integrity, task scheduling, and fusion algorithms.

[0012] In summary, the market urgently needs a fire source identification device that is low-power, low-false-alarm, fast-response, and easy to mass-produce: at the hardware level, it should have stable power sequence, camera gating, and bus level / isolation adaptation; at the algorithm level, it should achieve geometric mapping of thermal images to visible light and ROI cascading, temporal consistency fusion with smoke signals, and adaptive thresholds; at the engineering level, it should provide standardized external access, debugging / indication, wireless connectivity, and maintenance capabilities to meet the early fire source identification and networked alarm needs of various scenarios such as warehouses, power distribution, computer rooms, kitchens, and public buildings. Utility Model Content

[0013] To address the problems existing in the prior art, the purpose of this utility model is to provide a fire source identification device based on STM32F4, which has significant features such as low power consumption, low false alarm, fast response, and easy mass production. It adopts general-purpose components and standard interfaces, has clear test points, strong mass production and EMC / safety compliance, and controllable cost and yield.

[0014] To solve the above problems, the present invention adopts the following technical solution.

[0015] A fire source identification device based on STM32F4, comprising:

[0016] The main control circuit uses an STM32F4 series microcontroller.

[0017] The visible light camera unit includes an image sensor module, which is connected to the DCMI interface of the main control circuit through a parallel video interface, and is equipped with a domain-specific voltage regulation power supply circuit.

[0018] The infrared thermal imaging unit includes an infrared thermal imaging sensor array, which is connected via I... 2 The C bus is connected to the main control circuit and is equipped with level conversion and pull-up circuits;

[0019] The smoke sensing unit includes a gas sensor, the output of which is conditioned and connected to the ADC channel of the main control circuit.

[0020] The wireless communication unit includes a Wi-Fi module, which communicates with the main control circuit via UART, and is equipped with a bootstrap circuit.

[0021] The power management unit includes a lithium battery and a step-down module, and provides a gated power supply circuit for the visible light camera unit.

[0022] Furthermore, the visible light camera unit employs an OV2640 image sensor module. In some preferred embodiments, the OV2640 image sensor module's data pins Y0-Y7 (or D0-D7), PCLK, VSYNC, and HREF are connected to the DCMI interface of the main control circuit via a parallel video interface, and XCLK is provided by the main control circuit; the camera unit is equipped with dedicated regulated power supply branches of 2.8V and 1.3V.

[0023] Furthermore, the infrared thermal imaging unit employs an MLX90640 thermal imaging sensor. In some preferred embodiments, the MLX90640 array thermal imaging sensor uses I... 2 The C bus is connected to the main control circuit, and bidirectional level conversion and 4.7kΩ pull-up are set on the SCL and SDA lines by AO3400 devices; the thermal imaging unit is powered by a 3.3V regulator.

[0024] Furthermore, the smoke sensing unit employs an MQ-2 gas sensor. In some preferred embodiments, the output of the MQ-2 sensor is connected to the ADC channel of the main control circuit after being divided or conditioned by a resistor; this unit works in conjunction with the system for comprehensive fire source determination.

[0025] Furthermore, the domain-specific voltage regulation power supply circuit of the visible light camera unit includes a 2.8V regulated branch and a 1.3V regulated branch; the 2.8V regulated branch is provided by an XC6206P282MR regulator, and the 1.3V regulated branch is provided by an XC6219B132MR regulator. Both provide power to the OV2640's AVDD / DOVDD / DVDD domains.

[0026] Furthermore, the visible light camera unit is also equipped with a 24.000MHz active crystal oscillator as a local clock source.

[0027] A 24.000MHz active crystal oscillator, working in conjunction with the XCLK allocation network, provides the pixel clock reference for the OV2640.

[0028] Furthermore, the level conversion and pull-up circuit of the infrared thermal imaging unit is composed of an AO3400 MOSFET device and a 4.7kΩ pull-up resistor. In some preferred embodiments, the I of the infrared thermal imaging unit... 2 The pull-up resistor C is 4.7kΩ; SCL_EXT and SDA_EXT are connected via AO3400 to achieve the level and isolation of the external / internal bus.

[0029] Furthermore, the wireless communication unit employs an ESP8266 Wi-Fi module, whose bootstrap circuit includes pull-up resistors with a resistance of 5.6kΩ connected to the GPIO0 and GPIO15 pins. In some preferred embodiments, the ESP8266 Wi-Fi module communicates with the main control circuit via UART and is configured with a bootstrap resistor and a 3.3V regulated power supply. In some preferred embodiments, the TXD / RXD pins of the wireless communication unit are connected to the USART1_TX / USART1_RX pins of the main control circuit, and the USART1 terminal is led out via the onboard header JP3. In some preferred embodiments, the bootstrap signals of the ESP8266 Wi-Fi module include GPIO0 and GPIO15, and a 5.6kΩ resistor is configured to ensure stability during power-on startup.

[0030] Furthermore, the power management unit's gated power supply circuit uses an N-channel MOSFET as the switching device and is equipped with an LED status indicator. In some preferred embodiments, the power management unit includes a 7.4V lithium battery and a step-down module to supply power to the main control circuit and each sensing unit, and a MOSFET gated and indicator light structure is provided on the power supply side of the camera unit.

[0031] Furthermore, the DCMI interface signals of the main control circuit are brought out through the PC6 to PC12, PB6, and PB7 pin groups of the microcontroller. In some preferred embodiments, the DCMI signal group of the main control circuit includes DCMI_D0 to DCMI_D7, DCMI_PCLK, DCMI_VSYNC, DCMI_HREF, and DCMI_XCLK, which are allocated to the PC6 to PC12, PB6, and PB7 pin groups of the microcontroller and brought out through onboard components or connectors.

[0032] Compared with existing technologies, the advantages of this utility model are:

[0033] 1) The device achieves low false alarms and high detection by using time-series consistency judgment of visible light, infrared thermal imaging and smoke three modes; it first quickly locates hot spots with thermal imaging, and then performs color / motion / flicker analysis within the ROI, which has low computational load and fast response; the camera is powered on on demand by MOSFET gate control, normally only thermal imaging + smoke inspection, and the frame rate is increased during events, which takes into account both low power consumption and long battery life.

[0034] 2) DCMI parallel acquisition combined with DMA double buffer and I 2 The C / ADC is non-blocking, keeping the end-to-end delay in the low second range; the AO3400 bidirectional level converter and 4.7kΩ pull-up form soft isolation between the internal and external buses, enhancing anti-interference and ESD / EMC margin; the OV2640's 2.8V / 1.3V domain regulation and 24MHz reference improve imaging quality and frame synchronization stability.

[0035] 3) 3.3V first, camera side delayed power-on and ESP8266GPIO0 / GPIO15 bootstrap resistors ensure reliable startup; factory geometric calibration + on-site calibration combined with self-learning smoke baseline and brightness compensation adapt to multiple scenarios; JP3 serial port and gated LED facilitate debugging and maintenance, and remote parameter / log / OTA are optional; the whole uses general-purpose components and standard interfaces, test points are clear, mass production and EMC / safety compliance are strong, and cost and yield are controllable. Attached Figure Description

[0036] Figure 1 This is a general hardware connection block diagram of the system of this utility model;

[0037] Figure 2 This is a schematic diagram of the Zhengdian Atom STM32F407 minimum system board of this utility model;

[0038] Figure 3 This is a schematic diagram of the ATK-OV2640 principle of this utility model;

[0039] Figure 4 This is a schematic diagram of the IR_CAMERA_MLX90640 principle of this utility model;

[0040] Figure 5 This is a schematic diagram of the principle of the Zhengdian Atom ATK-MW8266D of this utility model. Detailed Implementation

[0041] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present utility model. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present utility model without creative effort are within the protection scope of the present utility model.

[0042] The core protection of this utility model lies in the hardware physical structure of the device, the electrical connection relationship between each unit, and the resulting hardware-level data interaction and coordination mechanism. The functions achieved by this utility model rely entirely on the aforementioned specific hardware connection and structure, and do not depend on any specific computer program or algorithm. That is, the "intelligence" and "function" of this utility model are embodied in its hardware architecture itself. Any device using the same or equivalent hardware structure can achieve the fire source identification function described in this utility model without any creative effort.

[0043] The hardware technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention.

[0044] Example 1:

[0045] 1. Overall hardware connection relationship of the system

[0046] Figure 1 This is a general hardware connection block diagram of the system of this utility model. (See diagram below.) Figure 1 As shown, the device in this embodiment is essentially an electronic device composed of specific hardware modules connected by defined electrical relationships.

[0047] The device includes a main control circuit, a visible light camera unit, an infrared thermal imaging unit, a smoke detection unit, a wireless communication unit, and a power management unit. The hardware configuration and electrical connections of each unit are as follows:

[0048] Main control circuit: Utilizes STM32F407 series microcontroller chip, including DCMI and I2C. 2 Physical interfaces such as C, ADC, and USART are used to build dedicated data channels with each sensing unit.

[0049] Visible light imaging unit: Employs the OV2640 image sensor chip. The chip's physical pins (Y0-Y7, PCLK, VSYNC, HREF) are directly and electrically connected to the main control chip's DCMI physical pins (corresponding to PC6-PC12, PB6, PB7) via PCB traces or ribbon cables. The XCLK pin receives the physical clock signal from the main control MCO pin. Its domain power supply circuit (composed of XC6206P282MR and XC6219B132MR regulators) provides independent power to the chip's analog and digital sections.

[0050] Infrared thermal imaging unit: Employs the MLX90640 sensor chip. Its I 2 The C bus (SCL, SDA) is physically connected to the master controller's I / O pin via a level shifting circuit consisting of an AO3400 MOSFET and a 4.7kΩ resistor. 2Pin C is connected. This circuit is a purely passive physical adapter network.

[0051] Smoke sensing unit: Employs the MQ-2 gas-sensitive element. Its output is connected to a specific ADC input pin of the main control chip via a physical resistor voltage divider circuit.

[0052] Wireless communication unit: Employs the ESP8266 module. Its TXD / RXD pins are physically connected to the main controller's USART_TX / USART_RX pins via cross-connection. Its GPIO0 / GPIO15 pins are pulled up to VCC or GND via 5.6kΩ physical resistors; this hardware connection determines its fixed startup mode.

[0053] The power management unit (WMU) is essentially a physical switch circuit consisting of an SI2302 MOSFET, resistors, and LEDs. The gate of this MOSFET is directly and electrically controlled by a GPIO pin of the main controller, and its drain and source are connected in series in the power supply circuit of the visible light camera unit. This connection enables physical on / off control of the camera's power supply.

[0054] like Figure 1 As indicated by the arrows, the power management unit supplies power to each module, with one path controlled by a MOSFET gate circuit, which in turn is managed by the GPIO of the main control circuit. The visible light camera unit connects to the DCMI interface of the main control circuit via a parallel bus to transmit image data; the infrared thermal imaging unit connects via I... 2 The C-bus connects to the main controller to transmit thermal matrix data; the smoke sensing unit connects to the main controller's ADC via analog circuitry to transmit concentration signals; and the wireless communication unit connects to the main controller via UART to transmit serial data.

[0055] The functionality of this invention depends entirely on the specific hardware connection and structure described above, and does not depend on any specific computer program or algorithm.

[0056] 2. Hardware implementation details of each unit

[0057] 2.1 Main Control Circuit

[0058] Figure 2 This is a schematic diagram of the STM32F407 minimum system board used in this invention. The main control circuit of this invention is built on this hardware platform. The physical pins of the microcontroller are specifically configured as follows: the DCMI interface is assigned to pin groups PC6-PC12, PB6, and PB7; I 2 The C interface uses PB8 / PB9; the ADC channel uses PA0; and USART1 uses PA9 / PA10. This specific pin assignment and connection to external units constitute the hardware basis of this invention.

[0059] 2.2 Visible Light Imaging Unit

[0060] Figure 3 This is a schematic diagram of the ATK-OV2640 module used in this invention. The core hardware of this module is the OV2640 image sensor chip. Figure 3 As shown, its key hardware features are:

[0061] Separate power supply circuit: Regulator U1 (XC6206P282MR) provides 2.8V to the AVDD / DOVDD pins of the sensor, and regulator U2 (XC6219B132MR) provides 1.3V to the DVDD pin of the sensor. This physical power separation design is crucial to ensuring image quality.

[0062] Clock circuit: Crystal oscillator Y1 (24.000MHz) provides the pixel clock reference for the sensor.

[0063] Interface connector: Connects to the DCMI interface of the main control board via physical pins such as D0-D7, PCLK, VSYNC, and HREF on pin P1.

[0064] 2.3 Infrared Thermal Imaging Unit

[0065] Figure 4 This is a schematic diagram of the IR_CAMERA_MLX90640 principle of this utility model. The hardware core of this unit is the MLX90640 sensor chip. Figure 4 As shown, its key hardware features are:

[0066] Level shifting and isolation circuits: in I 2 On the C bus (SCL, SDA), a physical adapter circuit consisting of MOSFETs Q1 and Q2 (AO3400) and pull-up resistors R2 and R3 (4.7kΩ) is used. This pure hardware circuit realizes level conversion and isolation between the external bus and the internal bus, improving interface compatibility and anti-interference capability.

[0067] 2.4 Wireless Communication Unit

[0068] Figure 5 This is a schematic diagram of the ATK-MW8266D module used in this invention. The hardware core of this unit is the ESP8266 module. Figure 5 As shown, its key hardware features are:

[0069] Bootstrap configuration circuit: The GPIO0 and GPIO15 pins of the module are pulled up to 3.3V through physical resistors R2 and R3 (5.6kΩ). This fixed hardware connection determines the power-on startup mode of the module and ensures startup reliability.

[0070] Serial port connection: Physically connected to the host's USART interface via the TXD / RXD pins of HEADER4.

[0071] 3. Hardware Collaboration Workflow

[0072] The functions of this device are achieved through the physical interaction of the aforementioned hardware units:

[0073] Triggering mechanism: The infrared thermal imaging unit triggers via I 2 The C-type physical bus continuously transmits data to the master controller. When the data indicates the presence of a hot spot, a GPIO pin on the master controller outputs a high / low level signal. This electrical signal directly drives the gate of the MOSFET gate circuit, thereby physically connecting the power supply circuit of the visible light camera unit.

[0074] Data Flow: Visible light image data is transferred to the main controller's internal memory via a parallel DCMI physical bus. Smoke concentration is converted into a digital value via an ADC pin. These three data streams are handled by the main controller's internal hardware DMA controller; this process is completed automatically by hardware and does not consume CPU resources.

[0075] Output mechanism: The final fire source identification result is converted into a serial electrical signal by the main controller through its USART hardware peripheral and output to the ESP8266 module.

[0076] In summary, this utility model is a physical device that achieves its function through specific hardware connections and structures shown in the accompanying drawings. The specific circuits shown in the figures are preferred embodiments of this utility model. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the hardware technical solutions disclosed in this utility model, based on the hardware concept of this utility model, should be covered within the protection scope of this utility model.

[0077] The above description is only a preferred embodiment of the present utility model, but the protection scope of the present utility model is not limited thereto.

Claims

1. A fire source identification device based on STM32F4, characterized in that: include: The main control circuit uses an STM32F4 series microcontroller. The visible light camera unit includes an image sensor module, which is connected to the DCMI interface of the main control circuit through a parallel video interface, and is equipped with a domain-specific voltage regulation power supply circuit. The infrared thermal imaging unit includes an infrared thermal imaging sensor array, which is connected via I... 2 The C bus is connected to the main control circuit and is equipped with level conversion and pull-up circuits; The smoke sensing unit includes a gas sensor, the output of which is conditioned and connected to the ADC channel of the main control circuit. The wireless communication unit includes a Wi-Fi module, which communicates with the main control circuit via UART, and is equipped with a bootstrap circuit. The power management unit includes a lithium battery and a step-down module, and provides a gated power supply circuit for the visible light camera unit.

2. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The visible light camera unit uses an OV2640 image sensor module.

3. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The infrared thermal imaging unit uses the MLX90640 thermal imaging sensor.

4. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The smoke sensing unit uses an MQ-2 gas sensor.

5. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The domain-regulated power supply circuit of the visible light camera unit includes a 2.8V regulated branch and a 1.3V regulated branch; the 2.8V regulated branch is provided by an XC6206P282MR regulator, and the 1.3V regulated branch is provided by an XC6219B132MR regulator.

6. The fire source identification device based on STM32F4 according to claim 1 or 5, characterized in that: The visible light camera unit is also equipped with a 24.000MHz active crystal oscillator as a local clock source.

7. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The level conversion and pull-up circuit of the infrared thermal imaging unit is composed of an AO3400 MOSFET device and a 4.7kΩ pull-up resistor.

8. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The wireless communication unit uses an ESP8266 Wi-Fi module, whose bootstrap circuit includes pull-up resistors with a resistance of 5.6kΩ connected to the GPIO0 and GPIO15 pins.

9. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The gated power supply circuit of the power management unit uses an N-channel MOSFET as the switching device and is equipped with an LED status indicator.

10. The fire source identification device based on STM32F4 according to claim 1, characterized in that: The DCMI interface signal of the main control circuit is brought out through the PC6 to PC12, PB6, and PB7 pin groups of the microcontroller.

Images