Power battery fire extinguishing control method and device based on machine vision and electronic equipment

Abstract

The invention discloses a power battery fire extinguishing control method and device based on machine vision and electronic equipment, and the method comprises the steps: obtaining current and voltage data, temperature distribution data and smoke and fire induction data of a target power battery, and judging whether the target power battery is on fire or not according to the current and voltage data, the temperature distribution data and the smoke and fire induction data; when determining that the target power battery is on fire, acquiring a thermal imaging image of the target power battery, and inputting the thermal imaging image and the temperature distribution data into a pre-trained fire point identification model to obtain a target fire point position; the liquid nitrogen injection angle is determined according to the target fire point position, the liquid nitrogen injection amount is determined according to the temperature distribution data and the smoke and fire induction data, and then the liquid nitrogen injection device is controlled to inject liquid nitrogen to the target fire point position according to the liquid nitrogen injection angle and the liquid nitrogen injection amount. The fire extinguishing efficiency and the fire extinguishing effect of the power battery are improved, the driving safety of the vehicle is enhanced, and the method can be applied to the technical field of vehicle control.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, and in particular to a power battery fire suppression control method, device, and electronic device based on machine vision. Background Technology

[0002] Currently, extinguishing fires involving power batteries requires external fire-fighting equipment and manual intervention, which is inefficient and the accuracy of fire extinguishing agent spraying cannot be controlled. At the same time, the fire extinguishing agent's participation in the exothermic chemical reaction can easily amplify the leakage, potentially causing secondary accidents, directly or indirectly damaging electrical components on the vehicle, and the resulting emissions (such as wastewater) can pollute the environment again.

[0003] In summary, fire suppression solutions for power batteries have drawbacks such as low efficiency, poor fire suppression effect, and environmental pollution, which affect the driving safety of vehicles. Summary of the Invention

[0004] The purpose of this invention is to at least partially solve one of the technical problems existing in the prior art.

[0005] Therefore, one objective of this invention is to provide a power battery fire extinguishing control method based on machine vision. This method can monitor whether the power battery is on fire in real time and accurately identify the location of the fire point, thereby quickly controlling the liquid nitrogen injection device to spray liquid nitrogen towards the fire point, thereby improving the efficiency and effect of power battery fire extinguishing and enhancing vehicle driving safety.

[0006] Another objective of this invention is to provide a power battery fire extinguishing control device based on machine vision.

[0007] To achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of the present invention include: On one hand, embodiments of the present invention provide a machine vision-based fire suppression control method for power batteries, comprising the following steps: Acquire current and voltage data, temperature distribution data, and smoke and fire detection data of the target power battery, and determine whether the target power battery is on fire based on the current and voltage data, temperature distribution data, and smoke and fire detection data; When it is determined that the target power battery is on fire, a thermal imaging image of the target power battery is acquired, and the thermal imaging image and the temperature distribution data are input into a pre-trained fire point recognition model to obtain the location of the target fire point. The liquid nitrogen injection angle is determined based on the target ignition point location, the liquid nitrogen injection quantity is determined based on the temperature distribution data and the smoke and fire sensing data, and then the liquid nitrogen injection device is controlled to inject liquid nitrogen towards the target ignition point based on the liquid nitrogen injection angle and the liquid nitrogen injection quantity.

[0008] Furthermore, in one embodiment of the present invention, the liquid nitrogen injection device includes a liquid nitrogen storage tank, a multi-branch delivery pipeline, multiple adjustable nozzles, and corresponding multiple high-pressure pulse injection valves. The liquid nitrogen storage tank is disposed in the cavity of the left or right side skirt of the vehicle body or in the cavity of the outer foot pedal. Each of the adjustable nozzles is connected to the liquid nitrogen storage tank through the multi-branch delivery pipeline. The injection direction of each of the adjustable nozzles can be controlled and adjusted. The total injection range of all the adjustable nozzles covers the bottom of the battery pack of the target power battery. Each of the high-pressure pulse injection valves is respectively installed on the branch delivery pipeline where the corresponding adjustable nozzle is located.

[0009] Furthermore, in one embodiment of the present invention, the step of acquiring the current and voltage data, temperature distribution data, and smoke and fire detection data of the target power battery, and determining whether the target power battery is on fire based on the current and voltage data, the temperature distribution data, and the smoke and fire detection data, specifically includes: The current and voltage data of the target power battery are obtained through a current and voltage sensor, and the temperature distribution data of the target power battery are obtained through a temperature sensor. The voltage fluctuation and current mutation rate of the target power battery are determined based on the current and voltage data, and the real-time temperature and temperature change rate of each cell of the target power battery are determined based on the temperature sensor. When the voltage fluctuation exceeds a preset first threshold or the current mutation rate exceeds a preset second threshold, and the real-time temperature exceeds a preset third threshold or the temperature change rate exceeds a preset fourth threshold, a thermal runaway risk warning is generated, and the smoke and fire sensing data of the target power battery is obtained through smoke and gas sensors. The smoke particle concentration and carbon monoxide concentration in the space where the target power battery is located are determined based on the smoke sensing data. When the concentration of smoke particles is greater than a preset fifth threshold or the concentration of carbon monoxide is greater than a preset sixth threshold, the target power battery is determined to be on fire.

[0010] Furthermore, in one embodiment of the present invention, the ignition point identification model is trained through the following steps: Acquire thermal imaging samples and temperature distribution samples of sample power batteries under thermal runaway fire conditions; Spatially align the thermal imaging samples and the temperature distribution samples to obtain training samples, and determine the corresponding ignition point location labels through manual annotation. Then, construct a training dataset based on the training samples and the ignition point location labels. The training dataset is input into a pre-built multi-branch convolutional neural network for training to obtain the trained ignition point recognition model.

[0011] Furthermore, in one embodiment of the present invention, the multi-branch convolutional neural network includes a thermal imaging branch network, a temperature distribution branch network, a feature fusion layer, and an output layer. The step of inputting the training dataset into the pre-constructed multi-branch convolutional neural network for training to obtain the trained fire point recognition model specifically includes: The thermal imaging samples are convolved using the thermal imaging branch network to obtain the spatial features of the thermal imaging. The temperature distribution samples are convolved using the temperature distribution branch network to obtain the spatial features of the temperature distribution. The feature fusion layer performs feature fusion on the thermal imaging spatial features and the temperature distribution spatial features based on an attention mechanism to obtain fused spatial features. The output layer maps the fused spatial features to the predicted ignition point location. The loss value is determined based on the predicted ignition point location and the ignition point location label; The parameters of the multi-branch convolutional neural network are updated based on the loss value to obtain the trained fire point recognition model.

[0012] Furthermore, in one embodiment of the present invention, determining the liquid nitrogen injection angle based on the target ignition point location and determining the liquid nitrogen injection quantity based on the temperature distribution data and the smoke and fire sensing data specifically includes: Based on the location of the target ignition point, determine the nearest number of adjustable nozzles as target nozzles, and determine the corresponding liquid nitrogen injection angle based on the position of the target nozzles and the location of the target ignition point; The severity of the fire is determined based on the temperature distribution data and the smoke and fire sensing data. The total amount of liquid nitrogen to be injected is determined based on the severity of the fire. The amount of liquid nitrogen to be injected into each of the target nozzles is determined based on the total amount of liquid nitrogen to be injected.

[0013] Furthermore, in one embodiment of the present invention, the step of controlling the liquid nitrogen injection device to inject liquid nitrogen toward the target fire location according to the liquid nitrogen injection angle and the liquid nitrogen injection amount specifically includes: Adjust the injection angle of the target nozzle according to the liquid nitrogen injection angle; The target pulse width is determined based on the liquid nitrogen injection volume, and the target driving voltage is determined based on the target pulse width and the fire severity. Then, the equivalent valve opening and injection pressure of the high-pressure pulse injection valve corresponding to the target nozzle are controlled based on the target pulse width and the target driving voltage, thereby injecting liquid nitrogen towards the target fire location.

[0014] On the other hand, embodiments of the present invention provide a power battery fire extinguishing control device based on machine vision, comprising: The fire detection module is used to acquire the current and voltage data, temperature distribution data, and smoke and fire detection data of the target power battery, and to determine whether the target power battery is on fire based on the current and voltage data, temperature distribution data, and smoke and fire detection data. The ignition point identification module is used to acquire a thermal imaging image of the target power battery when it is determined that the target power battery is on fire, and input the thermal imaging image and the temperature distribution data into a pre-trained ignition point identification model to obtain the location of the target ignition point. The liquid nitrogen injection control module is used to determine the liquid nitrogen injection angle based on the target ignition point location, determine the liquid nitrogen injection quantity based on the temperature distribution data and the smoke and fire sensing data, and then control the liquid nitrogen injection device to inject liquid nitrogen towards the target ignition location based on the liquid nitrogen injection angle and the liquid nitrogen injection quantity.

[0015] On the other hand, embodiments of the present invention provide an electronic device, including: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the above-described machine vision-based power battery fire suppression control method.

[0016] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the aforementioned machine vision-based power battery fire extinguishing control method.

[0017] On the other hand, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the aforementioned machine vision-based power battery fire extinguishing control method.

[0018] The advantages and beneficial effects of the present invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention: This invention acquires current and voltage data, temperature distribution data, and smoke and fire detection data of a target power battery. Based on these data, it determines whether the target power battery is on fire. If fire is detected, it acquires a thermal imaging image of the target power battery. The thermal imaging image and temperature distribution data are then input into a pre-trained ignition point recognition model to obtain the location of the target ignition point. The liquid nitrogen injection angle is determined based on the location of the target ignition point, and the liquid nitrogen injection quantity is determined based on the temperature distribution data and smoke and fire detection data. Finally, the liquid nitrogen injection device is controlled to inject liquid nitrogen towards the target ignition location based on the liquid nitrogen injection angle and the liquid nitrogen injection quantity. The embodiments of this invention can monitor whether the power battery is on fire in real time and accurately identify the location of the fire point, and then quickly control the liquid nitrogen injection device to spray liquid nitrogen at the location of the fire point, thereby improving the efficiency and effect of extinguishing the power battery fire and enhancing the driving safety of the vehicle. At the same time, extinguishing the power battery fire by spraying liquid nitrogen can rapidly cool it down to block the thermal runaway chain reaction and prevent reignition, further improving the extinguishing effect. Moreover, after the liquid nitrogen evaporates, only nitrogen gas is produced, leaving no chemical residues and not corroding or damaging other electrical components, making it more environmentally friendly. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments of the present invention are described below. It should be understood that the drawings described below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the steps of a machine vision-based fire suppression control method for power batteries, provided in an embodiment of the present invention. Figure 2 A structural block diagram of a power battery fire extinguishing control device based on machine vision provided in an embodiment of the present invention; Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0023] The machine vision-based power battery fire suppression control method provided in this invention can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited thereto; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application implementing the machine vision-based power battery fire suppression control method, but is not limited to the above forms.

[0024] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0025] Reference Figure 1 This invention provides a machine vision-based fire suppression control method for power batteries, specifically including the following steps: S101. Obtain the current and voltage data, temperature distribution data, and smoke and fire detection data of the target power battery, and determine whether the target power battery is on fire based on the current and voltage data, temperature distribution data, and smoke and fire detection data. S102. When it is determined that the target power battery is on fire, acquire the thermal imaging image of the target power battery, and input the thermal imaging image and temperature distribution data into the pre-trained ignition point recognition model to obtain the location of the target ignition point. S103. Determine the liquid nitrogen injection angle based on the target ignition point location, determine the liquid nitrogen injection quantity based on temperature distribution data and smoke and fire sensing data, and then control the liquid nitrogen injection device to inject liquid nitrogen towards the target ignition point based on the liquid nitrogen injection angle and liquid nitrogen injection quantity.

[0026] The embodiments of this invention can monitor whether the power battery is on fire in real time and accurately identify the location of the fire point, and then quickly control the liquid nitrogen injection device to spray liquid nitrogen at the location of the fire point, thereby improving the efficiency and effect of extinguishing the power battery fire and enhancing the driving safety of the vehicle. At the same time, extinguishing the power battery fire by spraying liquid nitrogen can rapidly cool it down to block the thermal runaway chain reaction and prevent reignition, further improving the extinguishing effect. Moreover, after the liquid nitrogen evaporates, only nitrogen gas is produced, leaving no chemical residues and not corroding or damaging other electrical components, making it more environmentally friendly.

[0027] As a further optional implementation, the liquid nitrogen injection device includes a liquid nitrogen storage tank, a multi-branch delivery pipeline, multiple adjustable nozzles, and corresponding multiple high-pressure pulse injection valves. The liquid nitrogen storage tank is located in the cavity of the left or right side skirt of the vehicle body or in the cavity of the outer foot pedal. Each adjustable nozzle is connected to the liquid nitrogen storage tank through the multi-branch delivery pipeline. The injection direction of each adjustable nozzle can be controlled and adjusted. The total injection range of all adjustable nozzles covers the bottom of the target power battery pack. Each high-pressure pulse injection valve is installed on the branch delivery pipeline where the corresponding adjustable nozzle is located.

[0028] Specifically, the vehicle-mounted liquid nitrogen injection device of the present invention will be described first.

[0029] 1) Design of liquid nitrogen supply and injection system 1. Liquid nitrogen storage and transportation Liquid nitrogen storage tanks are located within the cavities of the left and right side skirts of the vehicle body or within the cavities of the external foot pedals. Miniature Dewar flasks are used as the liquid nitrogen storage containers, with a volume controllable between 20-50L to meet the needs of small spaces. Through a vacuum insulation layer design, the daily evaporation rate is controlled to within 1%, reducing liquid nitrogen loss.

[0030] The conveying pipeline uses stainless steel corrugated pipe, which combines flexibility and low-temperature resistance (-196℃) to prevent pipeline rupture caused by vehicle bumps. The pipeline diameter is optimized to 8-10mm to balance flow rate and space occupation.

[0031] 2. Spraying device design The valve utilizes a high-pressure pulse injection valve with a response time of ≤50ms, enabling instantaneous liquid nitrogen injection. The valve body is made of titanium alloy, reducing weight by 40% compared to traditional steel valves, while also exhibiting excellent low-temperature resistance.

[0032] The injection port uses multiple fan-shaped nozzles with a coverage angle of 120° to ensure that liquid nitrogen evenly covers the bottom of the battery pack. The number of nozzles is configured according to the size of the battery pack (e.g., 4-6), and the flow rate of a single nozzle is controlled at 0.5-1L / s.

[0033] 2) 12V power supply adapter solution 1. In the event of a power battery failure, the range extender engine is started to supply power to the small, undamaged battery (when the small battery is depleted or damaged, it can be stored by connecting a supercapacitor in series with the ENG and discharged in a timely manner through the inverter for voltage regulation).

[0034] 2. Power System Selection The main power supply uses a 12V lithium battery pack (capacity 200-300Ah), with an energy density ≥150Wh / kg, reducing weight by 30% compared to lead-acid batteries. It is equipped with a BMS system to monitor battery status in real time and prevent overcharging / over-discharging.

[0035] The auxiliary power supply is a supercapacitor (capacity 5-10F), used for instantaneous high-current discharge (such as injection valve start-up) to mitigate the high-current impact on the lithium battery.

[0036] 3. Power Distribution and Management Design a DC-DC converter to boost 12V voltage to 24V / 48V (depending on the requirements of the injection valve), with a conversion efficiency of ≥95%.

[0037] A distributed power architecture is adopted to independently power modules such as control units, sensors, and injection valves, reducing the risk of single point of failure.

[0038] 3) Achieving small size and lightweight design 1. Structural optimization design The liquid nitrogen tank and injection system are integrated into a modular chassis frame made of aluminum alloy, which is 50% lighter than a steel frame. The frame dimensions are customized according to the vehicle model (e.g., length × width × height = 1.2m × 0.6m × 0.3m).

[0039] Pipelines and lines are embedded in the chassis and laid along the longitudinal beams to reduce space occupation.

[0040] 2. Application of lightweight materials The outer shell of the liquid nitrogen tank is made of carbon fiber composite material, which is three times stronger than that of a steel tank and 60% lighter.

[0041] The injection valve drive mechanism uses piezoelectric ceramics instead of traditional solenoid valves, reducing weight by 70% and increasing response speed.

[0042] As a further optional implementation, current and voltage data, temperature distribution data, and smoke / fire detection data of the target power battery are acquired, and the determination of whether the target power battery is on fire is made based on the current and voltage data, temperature distribution data, and smoke / fire detection data. Specifically, this includes: S1011. Obtain the current and voltage data of the target power battery through a current and voltage sensor, and obtain the temperature distribution data of the target power battery through a temperature sensor; S1012. Determine the voltage fluctuation and current mutation rate of the target power battery based on the current and voltage data, and determine the real-time temperature and temperature change rate of each cell of the target power battery based on the temperature sensor. S1013. When the voltage fluctuation is greater than the preset first threshold or the current mutation rate is greater than the preset second threshold, and the real-time temperature is greater than the preset third threshold or the temperature change rate is greater than the preset fourth threshold, a thermal runaway risk warning is generated and the smoke and fire sensing data of the target power battery is obtained through the smoke sensor and the gas sensor. S1014. Determine the smoke particle concentration and carbon monoxide concentration in the space where the target power battery is located based on the smoke sensing data. S1015. When the smoke particle concentration is greater than the preset fifth threshold or the carbon monoxide concentration is greater than the preset sixth threshold, it is determined that the target power battery is on fire.

[0043] This invention utilizes current and voltage data, temperature distribution data, and smoke and fire sensing data to perform multi-level detection of thermal runaway and fire in power batteries, as detailed below: 1. Voltage and current sensor: Real-time monitoring of abnormal fluctuations in battery circuit voltage / current (such as sudden drop, overcharge, short circuit), serving as the first-level detection to capture early signs of electrochemical failure (such as surge in internal resistance, micro-short circuit).

[0044] Threshold setting: A primary alarm is triggered when voltage fluctuation exceeds the rated value by ±15% or current mutation rate exceeds 100A / s.

[0045] Advantages: Millisecond-level response speed, earlier than temperature rise and smoke generation.

[0046] 2. Distributed temperature sensors (NTC thermocouples): Deployed at key nodes of the battery module (between cells, busbars) to monitor local temperature rise, serving as a second-level detection to identify the initial stage of thermal runaway (temperature slope > 1℃ / s or single point > 60℃).

[0047] Deployment density: One sensor is placed for every 3-5 cells to cover the heat diffusion path.

[0048] When the voltage fluctuation exceeds the preset first threshold or the current mutation rate exceeds the preset second threshold, and the real-time temperature exceeds the preset third threshold or the temperature change rate exceeds the preset fourth threshold, a thermal runaway risk warning is generated and smoke and fire detection is initiated.

[0049] 3. Smoke and fire detection and monitoring instrument (smoke + gas sensor): detects CO, VOC gas and smoke particles released by electrolyte decomposition, and serves as the third level of verification to detect fire under the premise of thermal runaway risk warning (avoiding false alarms of electrical signals).

[0050] When the concentration of CO, VOC or PM2.5 particles is too high, the power battery is determined to be on fire, and then an infrared thermal imager is activated to identify the ignition point.

[0051] As an optional implementation, the ignition point identification model is trained through the following steps: S201. Obtain thermal imaging samples and temperature distribution samples of the sample power battery under thermal runaway fire conditions. S202. Spatial alignment of thermal imaging samples and temperature distribution samples is performed to obtain training samples. The corresponding ignition point location labels are determined by manual annotation. Then, a training dataset is constructed based on the training samples and ignition point location labels. S203. Input the training dataset into a pre-built multi-branch convolutional neural network for training to obtain a trained fire point recognition model.

[0052] Specifically, the training data for the ignition point identification model includes: Thermal imaging samples: Infrared thermal imagers capture the thermal radiation distribution of a sample's power battery under thermal runaway and fire conditions, creating a temperature visualization image (e.g., TIFF format). Its advantages include being unaffected by light and being able to penetrate smoke. Temperature distribution sample: The temperature values ​​of each temperature detection point of the sample power battery are stored in matrix form. It needs to be aligned with the thermal imaging sample space to ensure coordinate consistency.

[0053] Images and data need to undergo geometric correction and resolution unification (e.g., mapping temperature data to a thermal imaging pixel grid), and thermal imaging images need to be denoised (e.g., median filtering) and pseudo-color enhanced to improve contrast in high-temperature areas.

[0054] Dataset requirements: Must include synchronously acquired thermal imaging and temperature data, with the ignition point location marked (e.g., in COCO format).

[0055] As a further optional implementation, the multi-branch convolutional neural network includes a thermal imaging branch network, a temperature distribution branch network, a feature fusion layer, and an output layer. The training dataset is input into the pre-constructed multi-branch convolutional neural network for training to obtain a trained fire point recognition model, which specifically includes: S2031. Thermal imaging samples are convolved using a thermal imaging branch network to obtain thermal imaging spatial features. S2032. The temperature distribution samples are convolved using a temperature distribution branch network to obtain the spatial features of the temperature distribution. S2033. The feature fusion layer performs feature fusion on the spatial features of thermal imaging and spatial features of temperature distribution based on the attention mechanism to obtain fused spatial features; S2034. The fused spatial features are mapped to the predicted ignition point location through the output layer; S2035. Determine the loss value based on the predicted ignition point location and the ignition point location label; S2036. Update the parameters of the multi-branch convolutional neural network based on the loss value to obtain the trained fire point recognition model.

[0056] Specifically, the multi-branch convolutional neural network includes a thermal imaging branch network, a temperature distribution branch network, a feature fusion layer, and an output layer, wherein: Thermal imaging branch network: Use CNN (such as ResNet) to extract texture and edge features of thermal imaging samples to obtain thermal imaging spatial features; Temperature data branch: Use connected networks or 1D-CNN to extract the temperature gradient and peak distribution features of temperature distribution samples to obtain the spatial features of temperature distribution; Feature fusion layer: By concatenating features or applying attention weights (such as the SE module), dual-path features are integrated to enhance the response in high-temperature abnormal regions and obtain fused spatial features; Output layer: Maps the fused spatial features to the predicted ignition point location.

[0057] The loss value is determined based on the predicted ignition point location and the ignition point location label. The parameters of the multi-branch convolutional neural network are updated based on the loss value to complete one round of iterative training. When the number of iterations reaches a preset threshold or the loss value is lower than a preset threshold, the trained ignition point recognition model is obtained.

[0058] As a further optional implementation, the liquid nitrogen injection angle is determined based on the target ignition point location, and the liquid nitrogen injection volume is determined based on temperature distribution data and smoke detection data, specifically including: S1031. Determine the nearest adjustable nozzles as target nozzles based on the location of the target ignition point, and determine the corresponding liquid nitrogen injection angle based on the location of the target nozzles and the location of the target ignition point. S1032. Determine the severity of the fire based on temperature distribution data and smoke detection data, determine the total amount of liquid nitrogen to be injected based on the severity of the fire, and determine the amount of liquid nitrogen to be injected into each target nozzle based on the total amount of liquid nitrogen to be injected.

[0059] Specifically, after determining the location of the target ignition point, the nearest adjustable nozzles are identified as target nozzles. Then, the liquid nitrogen injection angle of each target nozzle is determined based on the location of the target nozzles and the location of the target ignition point. The severity of the fire is determined based on temperature distribution data and smoke and fire sensing data. The total amount of liquid nitrogen to be injected is determined based on the severity of the fire. Finally, the amount of liquid nitrogen to be injected by each target nozzle is determined based on the total amount of liquid nitrogen to be injected.

[0060] In some optional embodiments, the liquid nitrogen injection volume and injection frequency of each target nozzle can be dynamically adjusted based on the fire severity (temperature, gas concentration) using a fuzzy PID control algorithm. For example, a low-flow continuous injection is used for the initial stage of a fire, while a high-flow pulse injection is used for the intermediate stage of a fire.

[0061] As a further optional implementation, the liquid nitrogen injection device is controlled to inject liquid nitrogen toward the target fire location based on the liquid nitrogen injection angle and the liquid nitrogen injection volume, which specifically includes: S1033. Adjust the spray angle of the corresponding target nozzle according to the liquid nitrogen spray angle; S1034. Determine the target pulse width based on the liquid nitrogen injection volume, determine the target driving voltage based on the target pulse width and the severity of the fire, and then control the equivalent valve opening and injection pressure of the high-pressure pulse injection valve corresponding to the target nozzle based on the target pulse width and the target driving voltage, thereby injecting liquid nitrogen towards the target fire location.

[0062] Specifically, the embodiments of the present invention employ a multi-degree-of-freedom nozzle bracket, which uses a stepper motor or servo motor to drive a gear set to achieve adjustment of the horizontal and pitch angles of the target nozzle.

[0063] A mapping model of injection volume → pulse width is adopted, with the basic formula: Target pulse width (ms) = K × Liquid nitrogen demand (L / s), where K is the valve flow coefficient (requires actual measurement and calibration). Although the valve itself is an "on / off" action (fully open or fully closed), by changing the duration of each opening (i.e., pulse width), the amount of fluid flowing out per unit time can be precisely controlled, which is functionally equivalent to adjusting the "opening degree". This embodiment of the invention uses PWM (Pulse Width Modulation) technology, which can control the valve opening duration by sending electrical pulses of different widths through a microprocessor, thereby achieving control of the equivalent valve opening degree.

[0064] A control method linking driving voltage and injection pressure is employed. After determining the pulse width, its equivalent valve opening is calculated. Then, by consulting a driving voltage mapping table based on the fire severity, the corresponding target driving voltage is obtained. An example of the driving voltage mapping table is shown in Table 1 below.

[0065]

[0066] Table 1 By controlling the equivalent valve opening and injection pressure of the high-pressure pulse jet valve corresponding to the target nozzle according to the target pulse width and target driving voltage, liquid nitrogen can be sprayed onto the target fire location to achieve rapid and precise fire extinguishing.

[0067] The method steps of the embodiments of the present invention have been described above. It can be understood that the embodiments of the present invention can monitor in real time whether the power battery is on fire and accurately identify the location of the ignition point, thereby quickly controlling the liquid nitrogen injection device to spray liquid nitrogen towards the ignition point, thus improving the efficiency and effect of extinguishing the power battery fire and enhancing vehicle driving safety. Simultaneously, extinguishing the power battery fire by spraying liquid nitrogen can rapidly cool it to block the thermal runaway chain reaction and prevent reignition, further improving the extinguishing effect. Moreover, after the liquid nitrogen evaporates, only nitrogen gas is produced, leaving no chemical residues and not corroding or damaging other electrical components, making it more environmentally friendly.

[0068] Reference Figure 2 This invention provides a machine vision-based fire suppression control device for power batteries, comprising: The fire detection module is used to acquire the current and voltage data, temperature distribution data, and smoke and fire detection data of the target power battery, and to determine whether the target power battery is on fire based on the current and voltage data, temperature distribution data, and smoke and fire detection data. The ignition point identification module is used to acquire thermal imaging images of the target power battery when it is determined that the target power battery is on fire, and input the thermal imaging images and temperature distribution data into a pre-trained ignition point identification model to obtain the location of the target ignition point. The liquid nitrogen injection control module is used to determine the liquid nitrogen injection angle based on the target ignition point location, determine the liquid nitrogen injection quantity based on temperature distribution data and smoke and fire sensing data, and then control the liquid nitrogen injection device to inject liquid nitrogen towards the target ignition location based on the liquid nitrogen injection angle and liquid nitrogen injection quantity.

[0069] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0070] Reference Figure 3This invention provides an electronic device, comprising: At least one processor; At least one memory for storing at least one program; When the above-mentioned at least one program is executed by the above-mentioned at least one processor, the above-mentioned at least one processor implements the above-mentioned machine vision-based power battery fire extinguishing control method.

[0071] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0072] This invention also provides a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the aforementioned machine vision-based power battery fire suppression control method.

[0073] This invention provides a computer-readable storage medium that can execute a machine vision-based power battery fire extinguishing control method provided in the method embodiments of this invention. It can execute any combination of the implementation steps of the method embodiments and has the corresponding functions and beneficial effects of the method.

[0074] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned machine vision-based power battery fire suppression control method.

[0075] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0076] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0077] The embodiments described in this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0078] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0079] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the aforementioned blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0080] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the aforementioned functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0081] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0082] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0083] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the aforementioned program can be printed, because the aforementioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0084] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0085] In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments" indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0086] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0087] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A method for power battery fire extinguishing control based on machine vision, characterized in that, The method comprises the following steps: obtaining current-voltage data, temperature distribution data and fire induction data of a target power battery, and determining whether the target power battery is on fire according to the current-voltage data, the temperature distribution data and the fire induction data; when it is determined that the target power battery is on fire, obtaining a thermal imaging image of the target power battery, inputting the thermal imaging image and the temperature distribution data into a pre-trained fire point recognition model to obtain a target fire point position; determining a liquid nitrogen injection angle according to the target fire point position, determining a liquid nitrogen injection amount according to the temperature distribution data and the fire induction data, and then controlling a liquid nitrogen injection device to inject liquid nitrogen to the target fire position according to the liquid nitrogen injection angle and the liquid nitrogen injection amount. 2.The power battery fire extinguishing control method based on machine vision of claim 1, wherein: The liquid nitrogen injection device comprises a liquid nitrogen storage tank, a multi-branch conveying pipeline, a plurality of adjustable nozzles and a plurality of corresponding high-pressure pulse injection valves. The liquid nitrogen storage tank is arranged in a left or right side skirt cavity of a vehicle body or in a cavity of an outside footboard. Each adjustable nozzle is in communication with the liquid nitrogen storage tank through the multi-branch conveying pipeline, and the injection direction of each adjustable nozzle can be controlled and adjusted. The total injection range of all the adjustable nozzles covers the bottom of the battery pack of the target power battery. Each high-pressure pulse injection valve is installed on the branch conveying pipeline corresponding to the adjustable nozzle. 3.The power battery fire extinguishing control method based on machine vision of claim 1, wherein, The method for obtaining current-voltage data, temperature distribution data and fire induction data of a target power battery, and determining whether the target power battery is on fire according to the current-voltage data, the temperature distribution data and the fire induction data, specifically comprises: obtaining the current-voltage data of the target power battery through a current-voltage sensor, and obtaining the temperature distribution data of the target power battery through a temperature sensor; determining the voltage fluctuation and current mutation rate of the target power battery according to the current-voltage data, and determining the real-time temperature and temperature change rate of each cell of the target power battery according to the temperature sensor; when the voltage fluctuation is greater than a preset first threshold value or the current mutation rate is greater than a preset second threshold value, and the real-time temperature is greater than a preset third threshold value or the temperature change rate is greater than a preset fourth threshold value, generating a thermal runaway risk warning and obtaining the fire induction data of the target power battery through a smoke sensor and a gas sensor; determining the smoke particle concentration and carbon monoxide concentration of the space where the target power battery is located according to the fire induction data; when the smoke particle concentration is greater than a preset fifth threshold value or the carbon monoxide concentration is greater than a preset sixth threshold value, determining that the target power battery is on fire. 4.The method of claim 1, wherein, The fire point recognition model is obtained by the following steps: obtaining thermal imaging samples and temperature distribution samples of sample power batteries in a thermal runaway fire state; performing spatial alignment on the thermal imaging samples and the temperature distribution samples to obtain training samples, and determining corresponding fire point position labels through artificial labeling, and then constructing a training data set according to the training samples and the fire point position labels; The training data set is input into a pre-constructed multi-branch convolutional neural network for training to obtain the trained fire point recognition model.

5. The machine vision-based power battery fire extinguishing control method according to claim 4, characterized in that, The multi-branch convolutional neural network comprises a thermal imaging branch network, a temperature distribution branch network, a feature fusion layer, and an output layer. The training of the training data set into the pre-constructed multi-branch convolutional neural network to obtain the trained fire point recognition model specifically comprises: The thermal imaging sample is convoluted through the thermal imaging branch network to obtain thermal imaging spatial features; The temperature distribution sample is convoluted through the temperature distribution branch network to obtain temperature distribution spatial features; The thermal imaging spatial features and the temperature distribution spatial features are fused based on an attention mechanism through the feature fusion layer to obtain fused spatial features; The fused spatial features are mapped into a predicted fire point position through the output layer; A loss value is determined according to the predicted fire point position and the fire point position label; The multi-branch convolutional neural network is updated in parameters according to the loss value to obtain the trained fire point recognition model. 6.The machine vision-based power battery fire extinguishing control method according to claim 2, wherein The target fire point position is determined, the distance closest to the target fire point position is determined as a target nozzle, and the liquid nitrogen injection angle corresponding to the target nozzle is determined according to the position of the target nozzle and the target fire point position. The fire severity is determined according to the temperature distribution data and the smoke fire sensing data, the total amount of liquid nitrogen injection is determined according to the fire severity, and the liquid nitrogen injection amount of each target nozzle is determined according to the total amount of liquid nitrogen injection. The liquid nitrogen injection device is controlled to inject liquid nitrogen to the target fire position according to the liquid nitrogen injection angle and the liquid nitrogen injection amount.

7. The power battery fire extinguishing control method based on machine vision according to claim 6, characterized in that, The injection angle of the corresponding target nozzle is adjusted according to the liquid nitrogen injection angle. The target pulse width is determined according to the liquid nitrogen injection amount, the target driving voltage is determined according to the target pulse width and the fire severity, and then the equivalent valve opening degree and injection pressure of the high-pressure pulse injection valve corresponding to the target nozzle are controlled according to the target pulse width and the target driving voltage, so as to inject liquid nitrogen to the target fire position. It comprises:

8. A machine vision-based power battery fire extinguishing control device, characterized in that, The fire judgment module is used for acquiring the current voltage data, temperature distribution data, and smoke fire sensing data of the target power battery, and judging whether the target power battery is on fire according to the current voltage data, the temperature distribution data, and the smoke fire sensing data. The fire point recognition module is used for acquiring the thermal imaging image of the target power battery when it is determined that the target power battery is on fire, inputting the thermal imaging image and the temperature distribution data into the pre-trained fire point recognition model, and obtaining the target fire point position. ​ A liquid nitrogen spraying control module is configured to determine a liquid nitrogen spraying angle according to the target ignition point position, determine a liquid nitrogen spraying amount according to the temperature distribution data and the fire induction data, and control a liquid nitrogen spraying device to spray liquid nitrogen to the target ignition position according to the liquid nitrogen spraying angle and the liquid nitrogen spraying amount.

9. An electronic device, comprising: The method comprises the following steps: at least one processor; at least one memory for storing at least one program; when the at least one program is executed by the at least one processor, the at least one processor implements the machine vision-based power battery fire extinguishing control method according to any one of claims 1 to 7.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the machine vision-based power battery fire extinguishing control method according to any one of claims 1 to 7.

Images