Charging pile fire early warning method and system based on dynamic risk assessment

Abstract

The invention relates to the technical field of charging pile management, in particular to a charging pile fire early warning method and system based on dynamic risk assessment, and the method comprises the steps: deploying temperature sensors in a charging module internal region, a battery interface contact region, a heat dissipation air channel and a surrounding environment region of a charging pile; the method comprises the following steps: acquiring temperature data of different types of sensors in the same area to form a multi-area temperature sensing network, receiving original temperature data acquired by all sensors in the multi-area temperature sensing network, filtering the original temperature data by adopting a self-adaptive Kalman filtering algorithm, and performing weighted fusion on the temperature data of different types of sensors in the same area to obtain temperature data; based on the processed data, carrying out dynamic risk assessment from two dimensions of temperature deviation degree and temperature change rate, and carrying out multi-level fire risk level judgment; and performing grading response operation according to the judged fire risk grade. The false alarm rate and the missing report rate are reduced, and the problem of response lag is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of charging pile management, in particular to a charging pile fire warning method and system based on dynamic risk assessment. BACKGROUND

[0002] With the rapid development of the electric vehicle industry, the safety of charging piles as core infrastructure is crucial. However, the energy is concentrated during high-power charging, leading to charging pile fire accidents. The main causes include overload of charging module IGBT and other power devices, increased contact resistance due to poor contact of battery interface connectors, and internal cable aging and insulation damage. These faults can cause local high-temperature hot spots, which can easily cause serious fire accidents if not detected and disposed of in time.

[0003] Currently, the safety protection schemes commonly used by charging piles on the market mostly only arrange a single type of temperature sensor on the surface of the charging module radiator, resulting in a failure to monitor local overheating phenomena caused by poor contact and the like. The existing warning mechanism generally relies on fixed temperature thresholds for judgment and fails to consider the direct impact of charging power and environmental temperature on temperature rise. This method completely ignores the key indicator of the instantaneous rate of change of temperature and responds with serious lag to the sudden temperature rise caused by short circuits and other faults, often missing the best disposal opportunity when the alarm is sounded.

[0004] Therefore, there is an urgent need for a charging pile fire warning solution that can achieve comprehensive monitoring and accurate early warning to overcome the deficiencies of the existing technology. SUMMARY

[0005] To solve the above problems, the present application provides a charging pile fire warning method and system based on dynamic risk assessment.

[0006] In a first aspect, the present application provides a charging pile fire warning method based on dynamic risk assessment, comprising: S1, deploying temperature sensors in the internal area of the charging module, the contact area of the battery interface, and the heat dissipation air duct and the surrounding environment area of the charging pile to form a multi-area temperature sensor network, S2, receiving the original temperature data collected by all sensors in the multi-area temperature sensor network and performing the following processing: S21, filtering the original temperature data using an adaptive Kalman filter algorithm to filter out the instantaneous temperature jumps caused by electromagnetic interference during charging; S22, weighting and fusing the temperature data of different types of sensors in the same area to obtain temperature data; for areas with a single type of sensor, the filtered data is directly used; S3, based on the processed data, dynamic risk assessment is performed from two dimensions of temperature deviation and temperature change rate, including: S31, temperature deviation and temperature change rate of each region are calculated; S32, according to the temperature deviation and temperature change rate, and in combination with the duration of the temperature deviation or the temperature change rate exceeding the corresponding threshold and the risk space distribution characteristics of multiple regions, a multi-level fire risk level is determined; S4, according to the determined fire risk level, a hierarchical response operation is performed.

[0007] By deploying sensors in multiple regions, a three-dimensional monitoring network covering the key heating parts inside the charging pile and the surrounding environment is constructed, completely solving the problem of single monitoring point and existing early warning blind area in the prior art. Through adaptive filtering and weighted fusion processing, electromagnetic interference noise is effectively suppressed, and the advantages of heterogeneous sensors are comprehensively utilized, obtaining more reliable and more accurate temperature data than single sensor, laying a solid foundation for subsequent accurate early warning. By introducing dynamic threshold and temperature change rate two-dimensional parameters for risk assessment, the early warning mechanism can adapt to different working conditions and environments, and can respond sensitively and accurately to different fault modes such as slow overload and instantaneous short circuit, reducing the false alarm rate and the missing alarm rate, and solving the response lag problem.

[0008] Through the hierarchical response mechanism, the early warning signal is seamlessly connected with the active protection measures such as power-off, fire extinguishing and regional linkage, realizing a complete automatic safety closed loop from perception to decision to execution, which can effectively intervene in the fire budding stage, and solves the problem of disconnection between early warning and disposal.

[0009] As a preferred technical solution of the present application, S1 specifically includes: A contact temperature sensor is arranged in the internal region of the charging module for monitoring the temperature of the power device; A non-contact infrared temperature sensor and a contact temperature sensor are arranged in the battery interface contact region at the same time, the non-contact infrared temperature sensor is used for monitoring the surface temperature of the plug and socket contact surface, and the contact temperature sensor is installed on the socket body for monitoring the body temperature thereof; A distributed optical fiber temperature sensor is laid in the preset range of the heat dissipation channel and the surrounding region of the charging pile for monitoring the cable and environment temperature, and has a high temperature point positioning function.

[0010] By coordinating the deployment of non-contact infrared and contact temperature sensors in the battery interface area, the instantaneous hot spots on the contact surface and the average temperature of the body can be captured simultaneously, solving the industry problem of localized overheating caused by poor contact, which traditional sensors struggle to accurately monitor. The use of distributed fiber optic sensors enables wide-area, location-specific temperature monitoring of the cable and its surrounding environment, eliminating blind spots that could lead to the spread of external fire sources.

[0011] As a preferred embodiment of the technical solution of the present invention, in S21, the step of using an adaptive Kalman filter algorithm to filter the original temperature data to remove instantaneous temperature jumps caused by electromagnetic interference during charging includes: Initialize the state variables and covariance matrix of the Kalman filter based on the initial operating state of the charging pile and the initial measurement values ​​of the sensors; In each sampling period, based on the operating status of the charging pile and the measurement model of the sensor, the state variables and covariance matrix at the current moment are predicted; Receive raw temperature data collected by the sensor and calculate the measurement residual, which is the difference between the raw temperature data and the predicted temperature value; Update the Kalman gain based on the measurement residuals and covariance matrix; The original temperature data is filtered using the updated Kalman gain to obtain filtered temperature data. Update the state variables and covariance matrix based on the filtered temperature data.

[0012] Through a dual mechanism of adaptive Kalman filtering and outlier suppression, the system can intelligently identify and filter out instantaneous temperature jumps caused by strong electromagnetic interference, ensuring data stability and authenticity and avoiding malfunctions. By assigning higher weights to the infrared sensors for data fusion, the fusion results more accurately reflect the true temperature changes at the battery interface contact surface, highlighting the ability to monitor major fault types and further improving system reliability.

[0013] As a preferred embodiment of the technical solution of the present invention, the step of weighted fusion of temperature data from different types of sensors in the same area in S22 to obtain temperature data includes: For the battery interface contact area, the surface temperature data measured by the non-contact infrared temperature sensor after filtering is fused with the body temperature data measured by the contact temperature sensor based on confidence weights to obtain the fused temperature data for that area.

[0014] As a preferred embodiment of the technical solution of the present invention, the step of performing weighted fusion based on confidence weights includes: Assign a first weighting coefficient to the data from the non-contact infrared temperature sensor. Assign a second weighting coefficient to the contact temperature sensor data. Wherein, the first weighting coefficient With the second weighting coefficient The sum is 1, and > ; Based on the assigned weights, the fusion temperature is calculated using the following weighted average formula. :

[0015] in, This is the filtered surface temperature data of the infrared sensor. This is the filtered temperature data of the contact sensor body.

[0016] As a preferred embodiment of the technical solution of the present invention, in S31, the step of calculating the temperature deviation and temperature change rate of each region includes: S311, Based on real-time collected ambient temperature and charging power The dynamic threshold of each monitoring area at the current time is calculated using a dynamic threshold function. Among them, the region dynamic threshold With ambient temperature and charging power The dynamic threshold function formula is as follows:

[0017] in, For the first The baseline threshold for each region For real-time charging power, This refers to the rated maximum power of the charging station. For real-time collection of ambient temperature, For reference to ambient temperature, and These are the weighting coefficients; S312, Process the real-time temperature values ​​of each region. With the corresponding dynamic threshold Subtracting the two yields the temperature deviation of the region. The calculation formula is: = - ; S313. Obtain the temperature value at the current time k. Temperature value at the previous sampling time k-1 Calculate the temperature change per unit time to obtain the rate of temperature change. The calculation formula is: = ( - ) / in, This represents the sampling interval.

[0018] By dynamically adjusting the safety threshold based on ambient temperature and charging power, the system can understand the normal temperature rise level under the current operating conditions, fundamentally avoiding false alarms caused by high summer temperatures or high-power charging, while ensuring the sensitivity of early warning under harsh operating conditions.

[0019] As a preferred embodiment of the technical solution of the present invention, in S32, the step of determining the multi-level fire risk level includes: S321. For each region, determine whether its temperature deviation is greater than a first deviation threshold, and / or whether its temperature change rate is greater than a first rate threshold. If the conditions are met, the region is marked as entering a primary abnormal state, and the corresponding first timer is started. S322. Based on the abnormal status, duration, and spatial distribution of all areas, the global risk level is determined according to the following rules: When at least one region remains in the primary abnormal state and the duration of its corresponding first timer exceeds the first time threshold, a first-level warning is triggered. A Level 2 warning is triggered when one of the following conditions is met: (i) The temperature deviation in any region is greater than a second deviation threshold that is greater than the first deviation threshold; (ii) The rate of temperature change in any region is greater than a second rate threshold that is greater than the first rate threshold. (iii) In multiple adjacent or thermally conductive regions, two or more regions simultaneously satisfy the primary abnormal state condition; As a preferred embodiment of the technical solution of the present invention, S322 further includes: When the temperature deviation and temperature change rate of all regions are lower than their corresponding first deviation threshold and first rate threshold, clear the abnormal state flags of all regions and reset the timer.

[0020] By introducing duration and spatial distribution as criteria, the system can distinguish between transient interference and real faults, and can identify the spread trend of faults, making the judgment logic of risk level more intelligent and reasonable, and greatly improving the accuracy of early warning.

[0021] Spatial distribution correlation criteria can identify situations where multiple related areas are simultaneously abnormal, which is a strong characteristic of a fire occurring and spreading, giving the system a certain early fire diagnosis capability.

[0022] As a preferred embodiment of the technical solution of the present invention, step S4 includes: When a Level 1 warning is detected, a local audible and visual alarm is activated and a warning message is pushed to the remote management platform. When a Level II warning is issued, the power supply to the charging circuit breaker will be immediately cut off, and the fire extinguishing device will be activated to extinguish the fire. At the same time, the charging piles within the preset range will be linked to the wireless communication module to enter the protection state.

[0023] The specific response actions under different risk levels have been clarified, especially the power outage, fire extinguishing and regional coordination operations under the level 2 warning, forming an efficient and proactive emergency response process that minimizes fire losses and safety risks.

[0024] Secondly, the technical solution of the present invention also provides a charging pile fire early warning system based on dynamic risk assessment, including a data processing and control module, wherein the data processing and control module is connected to a multi-area sensing module, an early warning and execution module and an area linkage communication module. The multi-area sensing module includes: At least one contact temperature sensor is disposed in the internal area of ​​the charging module for monitoring the temperature of the power devices; A non-contact infrared temperature sensor and a contact temperature sensor are installed in the battery interface contact area to monitor the surface temperature of the plug and socket contact surface and the body temperature of the socket body, respectively. Distributed fiber optic temperature sensors, laid within a preset range around the heat dissipation duct and charging pile, are used to monitor cable and ambient temperature and have the function of locating high temperature points. The data processing and control module is configured as follows: Receive the raw temperature data collected by the multi-region sensing module; An adaptive Kalman filter algorithm is executed to filter out instantaneous temperature jumps caused by electromagnetic interference; Weighted fusion based on confidence weights is performed on heterogeneous sensor data in the battery interface contact area; Calculate the temperature deviation and rate of temperature change for each region; Based on temperature deviation, temperature change rate, duration of exceeding threshold, and spatial distribution characteristics of risk in multiple regions, a multi-level fire risk level is determined. The early warning and execution module includes: Audible and visual alarms are used to issue local audio-visual alarms in response to Level 1 warning signals; The wireless communication module is used to push early warning information to the remote management platform; The charging circuit breaker is used to cut off the charging power supply in response to a secondary warning signal; Fire extinguishing equipment is used to activate fire extinguishing in response to a level-two warning signal; The regional linkage communication module is used to send linkage warning signals to other charging piles within a preset range when a level-two warning is determined.

[0025] Through the regional linkage communication module, intelligent protection among charging piles is realized, transforming the safety risks of a single point into a coordinated response of the entire charging station, which greatly improves the level and scope of safety protection.

[0026] As can be seen from the above technical solutions, this application has the following advantages: setting up a multi-area temperature sensing network to solve the problem of single temperature monitoring points and neglecting key parts; conducting dynamic fire risk assessment based on processed temperature data to solve the problem of relying on fixed thresholds and being susceptible to environmental interference leading to false alarms; solving the response lag problem by introducing the temperature change rate; and performing graded response operations according to the determined fire risk level to solve the problem of the disconnect between early warning and protection measures, which makes it impossible to stop the spread of fire. Attached Figure Description

[0027] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart illustrating the method provided in an embodiment of the present invention.

[0029] Figure 2 A block diagram of a system provided in an embodiment of the present invention. Detailed Implementation

[0030] To make the purpose, features, and advantages of this application more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solution protected by this application. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0031] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this application and in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0032] like Figure 1 As shown, this embodiment of the invention provides a charging pile fire early warning method based on dynamic risk assessment, including: S1. Temperature sensors are deployed in the charging module internal area, battery interface contact area, heat dissipation duct and surrounding environment area of ​​the charging pile to form a multi-area temperature sensing network. In this embodiment of the invention, the specific components include: A contact-type temperature sensor is deployed inside the charging module to monitor the temperature of the power devices; In the battery interface contact area, a non-contact infrared temperature sensor and a contact temperature sensor are deployed simultaneously. The non-contact infrared temperature sensor is used to monitor the surface temperature of the contact surface between the plug and the socket, and the contact temperature sensor is installed on the socket body to monitor its body temperature. Distributed fiber optic temperature sensors are installed within a preset range around the heat dissipation channel and charging pile to monitor cable and ambient temperature and to locate high-temperature points.

[0033] S2. Receive the raw temperature data collected by all sensors in the multi-region temperature sensing network, and perform the following processing: S21. The original temperature data is filtered using an adaptive Kalman filter algorithm to remove instantaneous temperature jumps caused by electromagnetic interference during charging; the specific steps are as follows: S211: Initialize the state variables and covariance matrix of the Kalman filter, based on the initial operating state of the charging pile and the initial measurement values ​​of the sensors.

[0034] Initialize state variables Covariance Matrix .

[0035] Initialize state variables The initial temperature state of the charging pile can be represented by a vector containing the initial temperature values ​​of key components of the charging pile. (Covariance matrix) The uncertainty of the initial state is usually represented by a diagonal matrix, where the elements on the diagonal represent the initial variance of each state variable.

[0036] Assuming that the temperature change of the charging pile is mainly affected by the charging power, ambient temperature, and internal heat transfer, the following state equation can be established:

[0037] Where is the state vector, representing the temperature of each key component of the charging pile at time k, for example:

[0038] ,

[0039] State transition matrix A matrix representing the transition relationship of state variables is usually an identity matrix, which represents the linear transition of state variables.

[0040] It is a control input matrix, representing the effect of charging power and ambient temperature on temperature changes. Assuming the charging power... and ambient temperature The known control inputs are:

[0041] in, It is a proportionality coefficient, which indicates the degree of influence of charging power and ambient temperature on the temperature changes of various key components.

[0042] It is a control input vector, representing charging power and ambient temperature:

[0043] This is the process noise vector, representing the impact of unmodeled dynamics and external disturbances on temperature changes, assumed to be zero-mean Gaussian white noise:

[0044] Process noise covariance matrix The statistical characteristics of process noise are typically represented by a small diagonal matrix, which represents the variance of the process noise.

[0045] S212: In each sampling period, based on the operating status of the charging pile and the measurement model of the sensor, predict the state variables and covariance matrix at the current moment.

[0046]

[0047] It is the predicted state variable at time k, that is, the predicted value of the state variable based on all available information before time k.

[0048] S213: Receives raw temperature data collected by the sensor and calculates the measurement residual. .

[0049] formula:

[0050] Measurement residuals Its purpose is to provide information about the difference between the current observations and the predicted values. This difference can be used to adjust and update the state variables of the Kalman filter to make it closer to the true system state.

[0051] The observation equation describes the relationship between the sensor measurements and the actual temperature of the charging pile. Assuming the sensor measurements are affected by the actual temperature and measurement noise, the following observation equation can be established:

[0052] It is an observation vector, representing the temperature value measured by the sensor at time k, for example:

[0053] Observation matrix Mapping state variables to the measurement space represents the relationship between the actual temperature and the measured value. Assuming that the sensor measurement value directly reflects the actual temperature, it can be approximated as an identity matrix.

[0054] This is the measurement noise vector, representing the sensor measurement error, assumed to be zero-mean Gaussian white noise:

[0055] Measurement noise covariance matrix The statistical characteristics of the measurement noise are represented by a diagonal matrix, which represents the variance of the measurement noise.

[0056] S214: Update the Kalman gain based on the measurement residuals and covariance matrix. .

[0057]

[0058] S215: Filter the original temperature data using the updated Kalman gain to obtain filtered temperature data.

[0059]

[0060] S216: Based on the filtered temperature data, update the state variables and covariance matrix to prepare for the filtering process in the next sampling period.

[0061]

[0062] S22. Weighted fusion of temperature data from different types of sensors in the same area is performed to obtain temperature data; for areas with a single type of sensor, the filtered data is used directly. For the battery interface contact area, the surface temperature data measured by the non-contact infrared temperature sensor after filtering is fused with the body temperature data measured by the contact temperature sensor based on confidence weights to obtain the fused temperature data for that area.

[0063] S3. Based on the processed data, a dynamic risk assessment is conducted from two dimensions: temperature deviation and temperature change rate, including: S31. Calculate the temperature deviation and temperature change rate of each region; wherein, the temperature deviation is obtained by comparing the temperature value with a dynamic threshold, the dynamic threshold being dynamically adjusted according to the real-time ambient temperature and charging power; the temperature change rate is obtained by calculating the temperature rise slope of a continuous sampling period. S32. Based on the temperature deviation and temperature change rate, and combined with the duration of the temperature deviation or temperature change rate exceeding the corresponding threshold and the risk spatial distribution characteristics of multiple areas, a multi-level fire risk level determination is made. S4. Based on the determined fire risk level, implement tiered response procedures, specifically including: When a Level 1 warning is detected, a local audible and visual alarm is activated and a warning message is pushed to the remote management platform. When a Level II warning is issued, the power supply to the charging circuit breaker will be immediately cut off, and the fire extinguishing device will be activated to extinguish the fire. At the same time, the charging piles within the preset range will be linked to the wireless communication module to enter the protection state.

[0064] In some embodiments, the step of performing confidence-weighted fusion includes: Assign a first weighting coefficient to the data from the non-contact infrared temperature sensor. Assign a second weighting coefficient to the contact temperature sensor data. Wherein, the first weighting coefficient With the second weighting coefficient The sum is 1, and > ; Based on the assigned weights, the fusion temperature is calculated using the following weighted average formula. :

[0065] in, This is the filtered surface temperature data of the infrared sensor. This is the filtered temperature data of the contact sensor body.

[0066] The principle behind the confidence-weighted fusion method is as follows: Non-contact infrared sensors can detect tiny hot spots on the metal contact surface caused by increased contact resistance more directly and quickly, but their measurements are easily affected by environmental radiation, dust, and installation angle. Contact sensors, on the other hand, measure the average temperature of the socket body, offering good stability and strong anti-interference capabilities, but their response to instantaneous hot spots on the contact surface suffers from thermal conduction delay. By assigning higher confidence weights to the infrared data, the fusion results can more sensitively reflect the true temperature changes of the contact surface. Simultaneously, contact data is used to smooth and anchor the results, improving data reliability.

[0067] Its specific implementation may include the following steps: Static weight allocation: Weight values ​​are fixed based on prior knowledge. For example, because infrared sensors are more direct at detecting hotspots caused by poor contact, they are assigned a higher weight (e.g., ...). = 0.6), assigning lower weights to contact sensors (e.g., = 0.4).

[0068] In some embodiments, dynamic weight allocation is used: the weight coefficients can be dynamically adjusted according to the sensor's operating status or data quality.

[0069] For example, if the infrared sensor mirror is detected to be severely contaminated (which can be determined by periodically checking the signal attenuation), its weight will be automatically reduced. and correspondingly improve Until the pollution is removed.

[0070] For example, if the calculated variance of the infrared sensor data is consistently too high, it indicates that the readings are unstable, and its weight can be temporarily reduced.

[0071] In some embodiments, step S31, calculating the temperature deviation and temperature change rate of each region, includes: S311, Based on real-time collected ambient temperature and charging power Through dynamic threshold function Calculate the dynamic threshold of each monitoring area at the current time. Among them, the region dynamic threshold With ambient temperature and charging power Increased with the rise; S312, Process the real-time temperature values ​​of each region. With the corresponding dynamic threshold Subtracting the two yields the temperature deviation of the region. The calculation formula is: = - ; S313. Obtain the temperature value at the current time k. Temperature value at the previous sampling time k-1 Calculate the temperature change per unit time to obtain the rate of temperature change. The calculation formula is: = ( - ) / in, This represents the sampling interval.

[0072] Dynamic threshold function An exemplary implementation is a linear weighted formula:

[0073] in: For the first The baseline threshold for each region under standard conditions (e.g., ambient temperature 25℃, charging power 0kW). For real-time charging power, This refers to the rated maximum power of the charging station. For real-time collection of ambient temperature, For reference ambient temperature (usually taken as 25℃). and These are weighting coefficients, obtained through experimental calibration, used to adjust the influence of power and ambient temperature on the threshold. The formula means that the higher the charging power, the more heat is generated, and the higher the allowable temperature rise; the higher the ambient temperature, the more difficult heat dissipation, and the higher the allowable reference temperature. Therefore, the dynamic threshold... It then rises.

[0074] A value greater than 0 indicates that the current temperature has exceeded the safety threshold for this operating condition; the larger the positive value, the more severe the overheating. Sampling interval. It can be set to 1 second, 2 seconds or 5 seconds as needed. The smaller the value, the more sensitive it is to rapid heating, but it is also more susceptible to noise interference.

[0075] This directly reflects the rate of imbalance between heat production and heat dissipation. An extremely high... (e.g., 20℃ / s) is typically characteristic of short-circuit arcing; a moderate but persistently positive... (For example, a temperature of 2℃ / s for more than 5 seconds) is a typical sign of overload or poor contact.

[0076] In some embodiments, step S32, which involves determining the multi-level fire risk level, includes: S321. For each region, determine whether its temperature deviation is greater than a first deviation threshold, and / or whether its temperature change rate is greater than a first rate threshold. If the conditions are met, the region is marked as entering a primary abnormal state, and the corresponding first timer is started. S322. Based on the abnormal status, duration, and spatial distribution of all areas, the global risk level is determined according to the following rules: When at least one region remains in the primary abnormal state and the duration of its corresponding first timer exceeds the first time threshold, a first-level warning is triggered. A Level 2 warning is triggered when one of the following conditions is met: (i) The temperature deviation in any region is greater than a second deviation threshold that is greater than the first deviation threshold; (ii) The rate of temperature change in any region is greater than a second rate threshold that is greater than the first rate threshold. (iii) In multiple adjacent or thermally conductive regions, two or more regions simultaneously satisfy the primary abnormal state condition; S323. When the temperature deviation and temperature change rate of all regions are lower than their corresponding first deviation threshold and first rate threshold, clear the abnormal status markers of all regions and reset all timers.

[0077] A specific embodiment of the multi-level fire risk level determination rule can be configured as follows: Primary abnormal state judgment: The first deviation threshold is 5℃; The first rate threshold is 1.5℃ / s; If the temperature deviation in a certain area is >5℃ or rate of temperature change A value >1.5℃ / s is considered a primary anomaly.

[0078] Level 1 Warning: The first time threshold is 10 seconds; If the above-mentioned primary abnormal state lasts for more than 10 seconds, a Level 1 warning will be triggered (audio and visual alarm, and reporting to the platform).

[0079] Level 2 warning: The second deviation threshold is 15℃ (condition i: absolute danger); The second rate threshold is 5℃ / s (condition ii: sudden change, such as short circuit); Condition iii: Spatial distribution correlation. For example: Scenario 1: The charging module area and the nearest heat dissipation duct area report primary anomalies at the same time, indicating that overheating has spread from the components to the heat dissipation system, and the risk has escalated.

[0080] Scenario 2: Both the battery interface area and the cable area report primary anomalies simultaneously, indicating that poor contact has caused the cable to overheat, escalating the risk.

[0081] In this embodiment of the invention, when a Level 1 warning is detected, the following operations are performed: (a) Activate the local audible and visual alarm device, emitting a continuous buzzing and yellow flashing light signal; (b) The warning information, including the warning level, location of the abnormal area, temperature data, and timestamp, is pushed to the remote monitoring platform through the wireless communication module; (c) Maintain the current charging status of the charging pile; When a Level 2 warning is detected, perform the following actions: (d) Immediately send a trip command to the charging circuit breaker to cut off the charging power supply, with a response time less than the set threshold. (e) After the power is cut off, the built-in aerosol fire extinguishing device will be automatically activated to spray the fire source area in a directional manner. (f) Upgrade the warning level to Level II and push the emergency alarm information to the remote monitoring platform and the mobile terminal of the maintenance personnel through the wireless communication module; (g) Send a linkage warning signal to all charging piles within a preset range via a local communication module; the charging piles receiving the signal perform one or more operations, such as reducing charging power, suspending new charging requests, or activating local warnings, based on their distance and location from the risk source.

[0082] Sending a linkage warning signal to all charging piles within a preset range via the local communication module specifically involves sending a data frame, which includes at least: the unique identifier of the charging pile at risk, the risk level, the risk type code, and the geographical coordinates of the risk source. The charging station receiving the signal will perform one or more of the following actions based on its distance and location from the risk source: reducing charging power, suspending new charging requests, or activating a local alert. Specific steps include: (a) Distance and risk assessment steps: The charging pile that receives the linkage early warning signal parses the data frame and calculates its distance D from the risk source based on its stored geographical location information; (b) Intelligent response decision-making steps: Based on the distance D and the risk level, implement a tiered response strategy: If D ≤ D1 (the first distance threshold, e.g., 10 meters), it is determined as a high-risk area, and the strictest protection operations are immediately executed: send a stop charging request to the vehicle that is charging, stop charging after receiving the vehicle's response, and activate a red stroboscopic alarm; If D1 < D ≤ D2 (the second distance threshold, e.g., 30 meters), it is determined as a medium-risk area, and preventive operations are executed: suspend starting any new charging process, limit the maximum output power to 50% of the rated power, and activate a yellow rotating light warning; If D > D2, it is determined as a low-risk area, and only a blue slow-flashing warning light is activated to remind users to pay attention to abnormalities around, and the early warning information is uploaded to the management platform.

[0083] (c) Additional decision factor steps: In the intelligent response decision, if wind direction information is available, for the charging piles located downwind of the risk source, after automatically upgrading the risk level assessment result by one level (e.g., upgrading from a medium-risk area to a high-risk area), the corresponding response operations are then executed.

[0084] After receiving the alarm signal, each charging pile no longer simply forwards it, but conducts intelligent independent risk assessment based on its relative position to the risk source and executes the most appropriate local protection strategy.

[0085] Communication method: LAN communication technologies with long-distance and strong penetration capabilities such as LoRa or Wi-Fi Mesh can be adopted to ensure that the linkage network remains effective when the cellular network (4G / 5G) may be interrupted.

[0086] Risk type code: It can be used to distinguish whether it is "abnormal temperature", "short circuit" or "open fire", and the receiving party can adjust the degree of the response strategy accordingly.

[0087] Geographical location coordinates: They can be obtained through the built-in GPS / Beidou module or pre-entered during deployment.

[0088] High-risk area (adjacent): The primary task is to immediately stop energy transmission to prevent itself from becoming a new fire source or being spread.

[0089] Medium-risk area (proximate): Adopt reduced-power operation, which not only guarantees part of the needs of the charged users but also significantly reduces the overall system heat load and risk.

[0090] Low-risk area (long-distance): It mainly serves the functions of warning and information reporting.

[0091] Such as Figure 2As shown, this embodiment of the invention also provides a charging pile fire early warning system based on dynamic risk assessment, including a data processing and control module, wherein the data processing and control module is connected to a multi-area sensing module, an early warning and execution module and an area linkage communication module; The multi-area sensing module includes: At least one contact temperature sensor is disposed in the internal area of ​​the charging module for monitoring the temperature of the power devices; A non-contact infrared temperature sensor and a contact temperature sensor are installed in the battery interface contact area to monitor the surface temperature of the plug and socket contact surface and the body temperature of the socket body, respectively. Distributed fiber optic temperature sensors, laid within a preset range around the heat dissipation duct and charging pile, are used to monitor cable and ambient temperature and have the function of locating high temperature points. The data processing and control module is configured as follows: Receive the raw temperature data collected by the multi-region sensing module; An adaptive Kalman filter algorithm is executed to filter out instantaneous temperature jumps caused by electromagnetic interference; Weighted fusion based on confidence weights is performed on heterogeneous sensor data in the battery interface contact area; Calculate the temperature deviation and rate of temperature change for each region; Based on temperature deviation, temperature change rate, duration of exceeding threshold, and spatial distribution characteristics of risk in multiple regions, a multi-level fire risk level is determined. The early warning and execution module includes: Audible and visual alarms are used to issue local audio-visual alarms in response to Level 1 warning signals; The wireless communication module is used to push early warning information to the remote management platform; The charging circuit breaker is used to cut off the charging power supply in response to a secondary warning signal; Fire extinguishing equipment is used to activate fire extinguishing in response to a level-two warning signal; The regional linkage communication module is used to send linkage warning signals to other charging piles within a preset range when a level-two warning is determined.

[0092] In this embodiment of the invention, the data processing and control module is integrated into an edge computing gateway, which has a built-in AI inference chip for local real-time execution of data processing and risk assessment algorithms. The regional linkage communication module is a LoRa wireless communication module or a Zigbee wireless communication module. The system also includes an environmental sensor for collecting ambient temperature data and providing it to the data processing and control unit for calculating dynamic thresholds.

[0093] The fire extinguishing device is an aerosol fire extinguishing device.

[0094] In some embodiments, the step of the data processing and control module executing an adaptive Kalman filter algorithm to filter out instantaneous temperature jumps caused by electromagnetic interference includes: Based on the optimal temperature estimate at time k-1, the predicted temperature state at time k and its prediction error covariance are predicted using the state equation. Based on the latest temperature observation data sequence, the statistical characteristics of the observation noise are calculated in real time, and the observation noise covariance matrix at time k is dynamically adjusted; the adjustment range of the observation noise covariance matrix is ​​positively correlated with the intensity of the abrupt signal in the observation data. Based on the temperature observation value at time k, the predicted temperature state value, and the dynamically adjusted observation noise covariance matrix, the Kalman gain is calculated, and the optimal temperature estimate and its estimation error covariance at time k are updated. Compare the original temperature observation value at time k with the predicted temperature state value at time k; If the absolute value of the difference exceeds the safety threshold calculated based on covariance, the observation is determined to be an abnormal jump, and the update is abandoned. The predicted temperature state is directly used as the final filtered output data at time k. Otherwise, the optimal temperature estimate at time k is used as the final filtered output data.

[0095] In some embodiments, the step of the data processing and control module performing confidence-weighted fusion of heterogeneous sensor data in the battery interface contact area includes: For the battery interface contact area, the surface temperature data measured by the non-contact infrared temperature sensor after filtering is fused with the body temperature data measured by the contact temperature sensor based on confidence weights to obtain the fused temperature data for that area.

[0096] Assign a first weighting coefficient to the data from the non-contact infrared temperature sensor. Assign a second weighting coefficient to the contact temperature sensor data. Wherein, the first weighting coefficient With the second weighting coefficient The sum is 1, and > ; Based on the assigned weights, the fusion temperature is calculated using the following weighted average formula. :

[0097] in, This is the filtered surface temperature data of the infrared sensor. This is the filtered temperature data of the contact sensor body.

[0098] In some embodiments, the data processing and control module includes the following steps in calculating the temperature deviation and temperature change rate of each region: S311, Based on real-time collected ambient temperature and charging power The dynamic threshold of each monitoring area at the current time is calculated using a dynamic threshold function. Among them, the region dynamic threshold With ambient temperature and charging power The dynamic threshold function formula is as follows:

[0099] in, For the first The baseline threshold for each region For real-time charging power, This refers to the rated maximum power of the charging station. For real-time collection of ambient temperature, For reference to ambient temperature, and These are the weighting coefficients; S312, Process the real-time temperature values ​​of each region. With the corresponding dynamic threshold Subtracting the two yields the temperature deviation of the region. The calculation formula is: = - ; S313. Obtain the temperature value at the current time k. Temperature value at the previous sampling time k-1 Calculate the temperature change per unit time to obtain the rate of temperature change. The calculation formula is: = ( - ) / in, This represents the sampling interval.

[0100] In some embodiments, the data processing and control module performs multi-level fire risk level determination steps, including: S321. For each region, determine whether its temperature deviation is greater than a first deviation threshold, and / or whether its temperature change rate is greater than a first rate threshold. If the conditions are met, the region is marked as entering a primary abnormal state, and the corresponding first timer is started. S322. Based on the abnormal status, duration, and spatial distribution of all areas, the global risk level is determined according to the following rules: When at least one region remains in the primary abnormal state and the duration of its corresponding first timer exceeds the first time threshold, a first-level warning is triggered. A Level 2 warning is triggered when one of the following conditions is met: (i) The temperature deviation in any region is greater than a second deviation threshold that is greater than the first deviation threshold; (ii) The rate of temperature change in any region is greater than a second rate threshold that is greater than the first rate threshold. (iii) In multiple adjacent or thermally conductive regions, two or more regions simultaneously satisfy the primary abnormal state condition; S323. When the temperature deviation and temperature change rate of all regions are lower than their corresponding first deviation threshold and first rate threshold, clear the abnormal status markers of all regions and reset the timer.

[0101] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for early warning of fires in charging piles based on dynamic risk assessment, characterized in that, include: S1. Temperature sensors are deployed in the charging module internal area, battery interface contact area, heat dissipation duct and surrounding environment area of ​​the charging pile to form a multi-area temperature sensing network. S2. Receive the raw temperature data collected by all sensors in the multi-region temperature sensing network, and perform the following processing: S21. The original temperature data is filtered using an adaptive Kalman filter algorithm to filter out instantaneous temperature jumps caused by electromagnetic interference during the charging process. S22. Weighted fusion of temperature data from different types of sensors in the same area is performed to obtain temperature data; for areas with a single type of sensor, the filtered data is used directly. S3. Based on the processed data, a dynamic risk assessment is conducted from two dimensions: temperature deviation and temperature change rate, including: S31. Calculate the temperature deviation and temperature change rate of each region; S32. Based on the temperature deviation and temperature change rate, and combined with the duration of the temperature deviation or temperature change rate exceeding the corresponding threshold and the risk spatial distribution characteristics of multiple areas, a multi-level fire risk level determination is made. S4. Based on the determined fire risk level, implement graded response procedures.

2. The charging pile fire early warning method based on dynamic risk assessment according to claim 1, characterized in that, S1 specifically includes: A contact-type temperature sensor is deployed inside the charging module to monitor the temperature of the power devices; In the battery interface contact area, a non-contact infrared temperature sensor and a contact temperature sensor are deployed simultaneously. The non-contact infrared temperature sensor is used to monitor the surface temperature of the contact surface between the plug and the socket, and the contact temperature sensor is installed on the socket body to monitor its body temperature. Distributed fiber optic temperature sensors are installed within a preset range around the heat dissipation channel and charging pile to monitor cable and ambient temperature.

3. The charging pile fire early warning method based on dynamic risk assessment according to claim 2, characterized in that, In step S21, the step of using an adaptive Kalman filter algorithm to filter the raw temperature data to remove instantaneous temperature jumps caused by electromagnetic interference during charging includes: Initialize the state variables and covariance matrix of the Kalman filter based on the initial operating state of the charging pile and the initial measurement values ​​of the sensors; In each sampling period, based on the operating status of the charging pile and the measurement model of the sensor, the state variables and covariance matrix at the current moment are predicted; Receive raw temperature data collected by the sensor and calculate the measurement residual, which is the difference between the raw temperature data and the predicted temperature value; Update the Kalman gain based on the measurement residuals and covariance matrix; The original temperature data is filtered using the updated Kalman gain to obtain filtered temperature data. Update the state variables and covariance matrix based on the filtered temperature data.

4. The charging pile fire early warning method based on dynamic risk assessment according to claim 3, characterized in that, The step in S22 to weightedly fuse temperature data from different types of sensors in the same area to obtain temperature data includes: For the battery interface contact area, the surface temperature data measured by the non-contact infrared temperature sensor after filtering is fused with the body temperature data measured by the contact temperature sensor based on confidence weights to obtain the fused temperature data for that area.

5. The charging pile fire early warning method based on dynamic risk assessment according to claim 4, characterized in that, The steps for performing confidence-weighted fusion include: Assign a first weighting coefficient to the data from the non-contact infrared temperature sensor. Assign a second weighting coefficient to the contact temperature sensor data. Wherein, the first weighting coefficient With the second weighting coefficient The sum is 1, and > ; Based on the assigned weights, the fusion temperature is calculated using the following weighted average formula. : in, This is the filtered surface temperature data of the infrared sensor. This is the filtered temperature data of the contact sensor body.

6. The charging pile fire early warning method based on dynamic risk assessment according to claim 5, characterized in that, In S31, the steps for calculating the temperature deviation and temperature change rate of each region include: S311, Based on real-time collected ambient temperature and charging power The dynamic threshold of each monitoring area at the current time is calculated using a dynamic threshold function. Among them, the region dynamic threshold With ambient temperature and charging power The dynamic threshold function formula is as follows: in, For the first The baseline threshold for each region For real-time charging power, This refers to the rated maximum power of the charging station. For real-time collection of ambient temperature, For reference to ambient temperature, and These are the weighting coefficients; S312, Process the real-time temperature values ​​of each region. With the corresponding dynamic threshold Subtracting the two yields the temperature deviation of the region. The calculation formula is: = - ; S313. Obtain the temperature value at the current time k. Temperature value at the previous sampling time k-1 Calculate the temperature change per unit time to obtain the rate of temperature change. The calculation formula is: = ( - ) / in, This represents the sampling interval.

7. The charging pile fire early warning method based on dynamic risk assessment according to claim 6, characterized in that, In S32, the steps for determining multi-level fire risk levels include: S321. For each region, determine whether its temperature deviation is greater than a first deviation threshold, and / or whether its temperature change rate is greater than a first rate threshold. If the conditions are met, the region is marked as entering a primary abnormal state, and the corresponding first timer is started. S322. Based on the abnormal status, duration, and spatial distribution of all areas, the global risk level is determined according to the following rules: When at least one region remains in the primary abnormal state and the duration of its corresponding first timer exceeds the first time threshold, a first-level warning is triggered. A Level 2 warning is triggered when one of the following conditions is met: (i) The temperature deviation in any region is greater than a second deviation threshold that is greater than the first deviation threshold; (ii) The rate of temperature change in any region is greater than a second rate threshold that is greater than the first rate threshold. (iii) In multiple adjacent or thermally conductive regions, two or more regions simultaneously satisfy the primary abnormal state condition.

8. The charging pile fire early warning method based on dynamic risk assessment according to claim 7, characterized in that, Following S322 are: When the temperature deviation and temperature change rate of all regions are lower than their corresponding first deviation threshold and first rate threshold, clear the abnormal state flags of all regions and reset the timer.

9. The charging pile fire early warning method based on dynamic risk assessment according to claim 8, characterized in that, The steps in S4 include: When a Level 1 warning is detected, a local audible and visual alarm is activated and a warning message is pushed to the remote management platform. When a Level II warning is issued, the power supply to the charging circuit breaker will be immediately cut off, and the fire extinguishing device will be activated to extinguish the fire. At the same time, the charging piles within the preset range will be linked to the wireless communication module to enter the protection state.

10. A charging pile fire early warning system based on dynamic risk assessment, characterized in that, It includes a data processing and control module, which is connected to a multi-area sensing module, an early warning and execution module, and an area linkage communication module; Multi-area sensing module, including: At least one contact temperature sensor is disposed in the internal area of ​​the charging module for monitoring the temperature of the power devices; A non-contact infrared temperature sensor and a contact temperature sensor are installed in the battery interface contact area to monitor the surface temperature of the plug and socket contact surface and the body temperature of the socket body, respectively. Distributed fiber optic temperature sensors are installed within a preset range around the heat dissipation duct and charging pile to monitor cable and ambient temperature. The data processing and control module is configured as follows: Receive the raw temperature data collected by the multi-region sensing module; An adaptive Kalman filter algorithm is executed to filter out instantaneous temperature jumps caused by electromagnetic interference; Weighted fusion based on confidence weights is performed on heterogeneous sensor data in the battery interface contact area; Calculate the temperature deviation and rate of temperature change for each region; Based on temperature deviation, temperature change rate, duration of exceeding threshold, and spatial distribution characteristics of risk in multiple regions, a multi-level fire risk level is determined. The early warning and execution module includes: Audible and visual alarms are used to issue local audio-visual alarms in response to Level 1 warning signals; The wireless communication module is used to push early warning information to the remote management platform; The charging circuit breaker is used to cut off the charging power supply in response to a secondary warning signal; Fire extinguishing equipment is used to activate fire extinguishing in response to a level-two warning signal; The regional linkage communication module is used to send linkage warning signals to other charging piles within a preset range when a level-two warning is determined.

Images