Low-voltage power distribution cabinet fireproof system with electric arc detection function

Abstract

The application discloses a low-voltage power distribution cabinet fireproof system with electric arc detection function, and particularly relates to the technical field of power distribution cabinet electrical safety, which comprises a data acquisition module, which is used for collecting original parameter data in a low-voltage power distribution cabinet in real time; an analysis and processing module, which is used for preprocessing the original parameters and constructing a fireproof diagnosis data set; a fireproof diagnosis module, which is used for constructing a fireproof model based on the data set to divide risk levels; a fireproof execution module, which is used for executing responses to the risk levels; and a fireproof feedback module, which is used for feeding back electric arc fault types and execution results of the low-voltage power distribution cabinet to corresponding control terminals. The application constructs a low-voltage power distribution cabinet fireproof system with electric arc detection function through the data acquisition module, the analysis and processing module, the fireproof diagnosis module, the fireproof execution module and the fireproof feedback module, and solves the problems of misjudgment of electric arc fault identification, insufficient accuracy of risk assessment, excessive or insufficient protection and low coordination efficiency.

Description

Technical Field

[0001] This invention relates to the field of electrical safety technology for distribution cabinets, and more specifically, to a fire prevention system for low-voltage distribution cabinets with arc detection function. Background Technology

[0002] Low-voltage switchgear, as a core device in power systems for distributing electrical energy and controlling circuits, directly impacts the stability of power supply in industrial production and building applications. During actual operation, arcing faults can easily occur inside switchgear due to aging wiring, poor contact, overload, or short circuits. The high temperature generated by the arc can rapidly ignite surrounding insulation materials, leading to fires, equipment damage, and even personal injury. Therefore, arcing fault detection and fire prevention control in low-voltage switchgear have become critical technical requirements in the field of electrical safety.

[0003] Existing low-voltage switchgear fire protection systems with arc detection capabilities can generally meet usage requirements, but several shortcomings remain: First, traditional systems lack correlation analysis of arc characteristic parameters, electrical parameters, and environmental parameters, leading to misjudgments in arc fault identification. Second, existing systems lack standardized preprocessing of collected parameter data and fail to analyze parameter correlations, resulting in insufficient accuracy in fire assessment of low-voltage switchgear. Third, existing systems often employ a one-size-fits-all emergency protection approach, leading to either over- or under-protection of switchgear. Fourth, existing systems do not provide real-time feedback of detection results and actions to the monitoring center, hindering timely intervention and subsequent traceability by maintenance personnel, resulting in inefficient emergency coordination. Therefore, a low-voltage switchgear fire protection system capable of accurately detecting arc faults, quantitatively assessing fire risks, implementing tiered responses, and providing closed-loop feedback is urgently needed. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a fire protection system for low-voltage distribution cabinets with arc detection function, which solves the problems mentioned in the background art through the following solution.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a low-voltage distribution cabinet fire protection system with arc detection function, comprising:

[0006] Data acquisition module: The low-voltage distribution cabinet is preset as the target cabinet. The original parameters of the target cabinet are collected in real time by deploying multi-source sensors. The original parameters include electrical parameters, arc characteristic parameters, and environmental parameters.

[0007] Analysis and processing module: Receives raw parameters of the target electrical cabinet in real time, performs standardized preprocessing on the raw parameters, performs correlation analysis on the preprocessed raw parameters, and obtains a fire prevention diagnostic dataset;

[0008] Fire prevention diagnosis module: Based on the fire prevention diagnosis dataset, a fire prevention model for low-voltage distribution cabinets is constructed. According to the comprehensive risk coefficient of the fire prevention model, the risk level is divided and fire prevention control instructions are generated. The model includes an arc fault identification unit and a fire risk assessment unit. The risk levels include level three risk, level two risk and level one risk.

[0009] Fire prevention execution module: Receives fire prevention control commands in real time and matches fire prevention responses. When the command shows a level 3 risk, it activates the audible and visual alarm device. When the command shows a level 2 risk, it cuts off the power supply to the faulty branch circuit based on the level 3 risk. When the command shows a level 1 risk, it cuts off the main power supply to the electrical cabinet and activates the fire extinguishing device.

[0010] Fire feedback module: Feeds back fire control commands and fire response to the remote monitoring center and fire control center of the target electrical cabinet, and records the feedback process.

[0011] Preferably, the analysis and processing module performs correlation analysis on the preprocessed raw parameters to obtain a fire prevention diagnosis dataset: the correlation between raw parameters is calculated by Pearson correlation coefficient, and raw parameters related to electrical fire prevention diagnosis are selected based on the correlation results to form a fire prevention diagnosis dataset. The calculation of Pearson correlation coefficient includes: calculating the temporal correlation between electrical parameters and arc characteristic parameters, and calculating the spatial correlation between arc characteristic parameters and environmental parameters.

[0012] It is necessary to further explain the correlation between the original parameters used in the Pearson correlation coefficient calculation. Firstly, the parameters undergo spatiotemporal alignment, specifically including: unifying timestamps and aligning all preprocessed parameters according to their acquisition time to form a time-series matrix. Where n is the number of samples for all parameters, and m is 7 because seven parameter types were collected, including current parameters. Voltage parameters Arc duration Arc waveform ,temperature ,humidity carbon monoxide concentration Spatial grouping: All preprocessed parameters are divided according to the detection area of ​​the distribution cabinet. Each spatial group contains the synchronously collected values ​​of all parameters within the divided area. The area division includes the busbar area and the branch area.

[0013] Secondly, Pearson correlation coefficient linear calculation is performed on two different parameter variables to obtain the correlation degree between the two different parameters. Here, the voltage parameter is used as an example. With arc waveform Taking the correlation coefficient calculation as an example, let the collected voltage parameter set be... Where n represents the total number of voltage parameter samples and the collected arc waveform dataset. Where n represents the total number of arc waveform samples, and the specific formula for calculating the correlation coefficient is as follows:

[0014]

[0015] Where n represents the total number of voltage parameter samples and arc waveform samples. This represents the sample value of the specific voltage parameter in the sample. This represents the specific arc waveform value in the sample, and when When the absolute value of is close to 1, the two parameters show a significant correlation.

[0016] Preferably, the fire prevention diagnosis module constructs a low-voltage distribution cabinet fire prevention model based on the fire prevention diagnosis dataset: it receives the fire prevention diagnosis dataset output by the analysis and processing module, and constructs a low-voltage distribution cabinet fire prevention model including an arc fault identification unit and a fire risk assessment unit based on the dataset. The arc fault identification unit adopts a classification model using the random forest algorithm, takes the arc feature parameters and electrical parameters in the dataset as input, and outputs the arc fault type.

[0017] The fire risk assessment unit constructs a multi-factor assessment matrix based on the fault type and environmental parameters output by the arc fault identification unit, and calculates the comprehensive risk coefficient by substituting the original parameters into the assessment matrix. ;

[0018] It should be further explained that the fire risk assessment unit of the low-voltage distribution cabinet fire protection model constructs a multi-factor assessment matrix based on the arc fault type and environmental parameters. This assessment matrix includes: Arc fault types, temperature, Indicates humidity, Represents carbon monoxide concentration, weight Based on the results of Pearson correlation coefficient calculation, they are divided into the following categories: ,use The overall parameter set representing the arc fault type and environmental parameters The comprehensive risk coefficient is obtained by substituting all parameters into the evaluation matrix and performing a weighted calculation. The specific calculation formula is as follows:

[0019]

[0020] The comprehensive risk coefficient is calculated based on historical data of all parameters. Divided into the following four coefficient ranges: , , , .

[0021] The technical effects and advantages of this invention are as follows:

[0022] This invention collects arc characteristic parameters by deploying ultraviolet sensors and combines them with electrical parameters. It then uses a random forest algorithm to construct a fire prevention diagnostic model. This model can prevent fires by capturing arc characteristics before they occur, thus solving the problem of misjudging arc faults in traditional fire prevention systems.

[0023] This invention uses an analysis and processing module to standardize and preprocess electrical parameters, arc characteristic parameters, and environmental parameters, calculates Pearson correlation coefficients to screen out key correlation parameters, and forms a fire prevention diagnostic dataset, thereby improving the accuracy of fire risk assessment.

[0024] This invention classifies risks into three levels based on a comprehensive risk coefficient and matches them with differentiated execution strategies. This avoids the one-size-fits-all emergency protection methods in existing technologies, and ensures the normal operation of low-voltage distribution cabinets while ensuring safety. It also solves the problems of over-protection or under-protection in traditional systems.

[0025] The fire protection feedback module of this invention synchronizes control commands and execution results to the remote monitoring center and fire control center in real time via industrial Ethernet, realizing closed-loop feedback throughout the entire process and improving the monitoring and emergency coordination efficiency of the low-voltage distribution cabinet fire protection system. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the system logic structure of the present invention.

[0027] Figure 2 This is a schematic diagram of the system analysis and processing module structure of the present invention.

[0028] Figure 3 This is a schematic diagram of the overall system framework of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Please see Figures 1-3As shown, this embodiment of the invention provides a low-voltage distribution cabinet fire protection system with arc detection function. Through a data acquisition module, analysis and processing module, fire prevention diagnosis module, fire prevention execution module, and fire prevention feedback module, the low-voltage distribution cabinet fire protection system with arc detection function is constructed. This embodiment of the invention discloses a low-voltage distribution cabinet fire protection system with arc detection function, the system comprising:

[0031] Data acquisition module: The low-voltage distribution cabinet is preset as the target cabinet. The original parameters of the target cabinet are collected in real time by deploying multi-source sensors. The original parameters include electrical parameters, arc characteristic parameters, and environmental parameters.

[0032] Analysis and processing module: Receives raw parameters of the target electrical cabinet in real time, performs standardized preprocessing on the raw parameters, performs correlation analysis on the preprocessed raw parameters, and obtains a fire prevention diagnostic dataset;

[0033] Fire prevention diagnosis module: Based on the fire prevention diagnosis dataset, a fire prevention model for low-voltage distribution cabinets is constructed. According to the comprehensive risk coefficient of the fire prevention model, the risk level is divided and fire prevention control instructions are generated. The model includes an arc fault identification unit and a fire risk assessment unit. The risk levels include level three risk, level two risk and level one risk.

[0034] Fire prevention execution module: Receives fire prevention control commands in real time and matches fire prevention responses. When the command shows a level 3 risk, it activates the audible and visual alarm device. When the command shows a level 2 risk, it cuts off the power supply to the faulty branch circuit based on the level 3 risk. When the command shows a level 1 risk, it cuts off the main power supply to the electrical cabinet and activates the fire extinguishing device.

[0035] Fire feedback module: Feeds back fire control commands and fire response to the remote monitoring center and fire control center of the target electrical cabinet, and records the feedback process.

[0036] The data acquisition module collects the raw parameters of the target electrical cabinet in real time by deploying multi-source sensors, specifically including:

[0037] By deploying current and voltage sensors in the distribution cabinet, the current parameter in the electrical parameters can be collected in real time. With voltage parameters The arc duration, a key characteristic parameter of the electric arc, is collected in real time using an ultraviolet sensor. Arc waveform The system collects environmental parameters, including temperature, humidity, and smoke, in real time using temperature sensors, humidity sensors, and smoke sensors. ,humidity carbon monoxide concentration .

[0038] Further explanation is needed regarding the following: Current sensors are installed on the busbars at the incoming end of the distribution cabinet and on the cables at the outgoing ends of each branch; voltage sensors are installed on the voltage terminals at the incoming end of the distribution cabinet and on each branch; and the terminals are insulated. An ultraviolet (UV) sensor is installed at the center of the top inside the distribution cabinet, and another UV sensor is installed at each branch circuit breaker and cable joint. The sensor probes are pointed towards the arc-generating area, which includes the cable joints and contacts. Each monitoring point maintains a distance of 0.3. Up to 0.5 The distance; the temperature sensor and humidity sensor are installed inside the rear wall of the distribution cabinet away from the heat-generating element, and the smoke sensor is installed at the cabinet door to facilitate the detection of overflowing carbon monoxide gas.

[0039] The analysis and processing module receives the raw parameters of the target electrical cabinet in real time and performs standardized preprocessing on the raw parameters:

[0040] Firstly adopt The criteria involve removing outliers from the original parameters, filling in missing values ​​using linear interpolation, converting the analog signals of the original parameters into digital signals, and finally performing... Normalization maps the original parameters of different dimensions to... Interval.

[0041] It should be further explained that outlier removal is performed on the original parameters, and individual parameters within the original parameters are processed separately, with separate datasets set for each individual parameter. Where n is the total number of samples for a single parameter, the mean of the parameter samples is calculated. and standard deviation The specific formula for calculating the mean is as follows: The formula for calculating standard deviation is: in Indicates the first A sample parameter; when a sample parameter satisfy When, determine parameters Outlier parameters are identified and removed. After removing outliers, the mean of the remaining data is recalculated. and standard deviation Repeat the iteration 1 to 2 times until there are no outliers;

[0042] Linear imputation is performed to fill in missing parameters caused by sensor communication interruption. The imputation includes pre-setting the time series of parameters, collecting parameters every 1 second, and determining that there is a missing value at that time point if no parameter is collected after 1 second. ,in Indicates the time of missing value collection; missing values ​​are linearly imputed, if missing values... Valid parameters existing at previous and next time points and The specific formula for linear interpolation is as follows:

[0043]

[0044] in Indicates the moment before missing values ​​were collected. This indicates the moment immediately following the collection of missing values;

[0045] The analog signal of the original parameters is subjected to signal conditioning and quantization conversion. The signal conditioning involves filtering and amplifying the analog signal of the original parameters output by the sensor, and then quantizing the analog signal into a digital signal through an analog-to-digital converter (ADC). The formula is as follows:

[0046]

[0047] in Here are the conditioned analog input voltage parameters, and n is the number of bits in the analog-to-digital converter (ADC). Where is the reference voltage of the ADC (e.g., 0V and 5V), and D is the digital output value;

[0048] Perform the output digital quantity Normalization eliminates the dimensional differences between parameters of different dimensions and maps the parameters to... The interval facilitates subsequent model analysis, based on a dataset of individual parameters from the original parameters. Calculate the minimum value and maximum value Perform normalization processing, the normalization processing includes, for the first Sample parameters Normalization is obtained by transformation. The specific calculation formula is as follows: Among the arc characteristic parameters, the arc waveform Each one needs to be normalized separately.

[0049] The analysis and processing module performs correlation analysis on the preprocessed raw parameters to obtain a fire prevention diagnostic dataset:

[0050] The correlation between the original parameters is calculated by using the Pearson correlation coefficient. Based on the correlation results, the original parameters related to electrical fire prevention diagnosis are selected to form a fire prevention diagnosis dataset. The calculation of the Pearson correlation coefficient includes: calculating the temporal correlation between electrical parameters and arc characteristic parameters, and calculating the spatial correlation between arc characteristic parameters and environmental parameters.

[0051] It is necessary to further explain the correlation between the original parameters used in the Pearson correlation coefficient calculation. Firstly, the parameters undergo spatiotemporal alignment, specifically including: unifying timestamps and aligning all preprocessed parameters according to their acquisition time to form a time-series matrix. Where n is the number of samples for all parameters, and m is 7 because seven parameter types were collected, including current parameters. Voltage parameters Arc duration Arc waveform ,temperature ,humidity carbon monoxide concentration Spatial grouping: All preprocessed parameters are divided according to the detection area of ​​the distribution cabinet. Each spatial group contains the synchronously collected values ​​of all parameters within the divided area. The area division includes the busbar area and the branch area.

[0052] Secondly, Pearson correlation coefficient linear calculation is performed on two different parameter variables to obtain the correlation degree between the two different parameters. Here, the voltage parameter is used as an example. With arc waveform Taking the correlation coefficient calculation as an example, let the collected voltage parameter set be... Where n represents the total number of voltage parameter samples and the collected arc waveform dataset. Where n represents the total number of arc waveform samples, and the specific formula for calculating the correlation coefficient is as follows:

[0053]

[0054] Where n represents the total number of voltage parameter samples and arc waveform samples. This represents the sample value of the specific voltage parameter in the sample. This represents the specific arc waveform value in the sample, and when When the absolute value of the two parameters is close to 1, the two parameters show a significant correlation.

[0055] Calculate the temporal correlation between electrical parameters and arc characteristic parameters, and analyze the electrical parameters. With arc characteristic parameters The time-varying synchronous correlation includes: dividing electrical parameters and arc characteristic parameters into time windows; dividing the parameters by setting the window size to 50 acquisition points and the step size to 2; and performing Pearson correlation calculations on the parameters within each window. Specifically, this includes... , , , ;

[0056] Calculate the characteristic parameters of the electric arc With environmental parameters Spatial correlation analysis is conducted to examine the coexistence and correlation of arc characteristic parameters and environmental parameters within the same space, such as whether an increase in arc duration is accompanied by an increase in carbon monoxide concentration. Relevant arc characteristic parameters and environmental parameters are extracted from the spatially divided busbar and branch regions, and Pearson correlation calculations are performed on the parameters within each space. Specifically, this includes… , , , , , ;

[0057] Since electric arcs are a precursor to fires in low-voltage distribution cabinets, parameters that are significantly correlated with electric arc characteristics are selected and retained based on the Pearson correlation calculations between electrical parameters and electric arc characteristic parameters, and between electric arc characteristic parameters and environmental parameters, thus forming a fire prevention diagnostic dataset.

[0058] The fire prevention diagnostic module constructs a fire prevention model for low-voltage distribution cabinets based on the fire prevention diagnostic dataset.

[0059] The system receives the fire prevention diagnosis dataset output by the analysis and processing module, and builds a low-voltage distribution cabinet fire prevention model based on the dataset, which includes an arc fault identification unit and a fire risk assessment unit. The arc fault identification unit adopts a classification model using the random forest algorithm, takes the arc feature parameters and electrical parameters in the dataset as input, and outputs the arc fault type.

[0060] Further explanation is needed regarding the arc fault identification unit of the low-voltage distribution cabinet fire prevention model. Based on the correlation results of the fire prevention diagnosis dataset, arc feature parameters and electrical parameters are extracted from the dataset for arc fault identification. The arc feature parameters and electrical parameters are labeled according to the classification model of the random forest algorithm. The classification model of the random forest algorithm includes the number of decision trees, tree depth, number of split samples, and random seeds. The arc fault types are divided into four categories: first, normal arc when the low-voltage distribution cabinet is switched on; second, arc due to poor contact; third, arc due to high voltage breakdown of insulation; and fourth, arc due to high current short circuit.

[0061] The fire risk assessment unit constructs a multi-factor assessment matrix based on the fault type and environmental parameters output by the arc fault identification unit, and calculates the comprehensive risk coefficient by substituting the original parameters into the assessment matrix. .

[0062] It should be further explained that the fire risk assessment unit of the low-voltage distribution cabinet fire protection model constructs a multi-factor assessment matrix based on the arc fault type and environmental parameters. This assessment matrix includes: Arc fault types, temperature, Indicates humidity, Represents carbon monoxide concentration, weight Based on the results of Pearson correlation coefficient calculation, they are divided into the following categories: ,use The overall parameter set representing the arc fault type and environmental parameters The comprehensive risk coefficient is obtained by substituting all parameters into the evaluation matrix and performing a weighted calculation. The specific calculation formula is as follows:

[0063]

[0064] The comprehensive risk coefficient is calculated based on historical data of all parameters. Divided into the following four coefficient ranges: , , , .

[0065] The fire prevention diagnosis module classifies risk levels and generates fire prevention control instructions based on the comprehensive risk coefficient of the fire prevention model.

[0066] Risk levels are classified based on the comprehensive risk coefficient of the fire prevention model. When the comprehensive risk coefficient is within... The risk level is divided into three levels when the interval is within a certain range; when the comprehensive risk coefficient is within a certain range... When the risk level is within a certain range, it is classified as Level 2 risk; when the overall risk is within a certain range... The interval is divided into Level 1 risk. Based on the risk level, three levels of fire prevention control instructions are generated. The digital signals of the three risk levels are converted into electrical signals and transmitted to the fire prevention execution module.

[0067] The fire prevention execution module receives fire prevention control commands in real time and matches the fire prevention response accordingly.

[0068] Establish a communication connection between the fire prevention execution module and the fire prevention diagnosis module, receive fire prevention control commands in real time, parse the risk level in the command, extract the risk level to match the current fire prevention response level, and drive the target electrical cabinet to execute the matching response result;

[0069] The process of extracting risk levels to match the current fire response level includes: activating the audible and visual alarm device when the instruction indicates a level three risk; cutting off the power supply to the faulty branch when the instruction indicates a level two risk, based on the level three risk; and cutting off the main power supply to the electrical cabinet and activating the fire extinguishing device when the instruction indicates a level one risk.

[0070] Further explanation is needed regarding the risk level in the analysis instruction. The risk level is extracted to match the current fire response level. The main arc fault type in the level three risk analysis is a normal arc with poor contact. The temperature, humidity, and carbon monoxide concentration are within the safe range. At this time, the audible and visual alarm device will sound an alarm.

[0071] The main arc fault type in the secondary risk analysis is insulation damage and poor contact. The temperature of the distribution cabinet rises to 60℃ and the carbon monoxide concentration reaches 50ppm. At this time, the audible and visual alarm device continues to alarm and cuts off the branch circuit of the distribution cabinet within 0.5s to isolate the fault area, triggering the clean gas fire extinguishing device, which completes the spraying of the entire branch circuit of the cabinet within 10s.

[0072] The main type of arc fault in the Level 1 risk analysis is short-circuit arc. The temperature of the distribution cabinet is greater than 60℃ and the carbon monoxide concentration is greater than 50ppm. At this time, the main power supply of the low-voltage distribution cabinet is cut off and the clean gas fire extinguishing device is triggered, and the entire cabinet is sprayed within 10 seconds.

[0073] The fire protection feedback module feeds back fire protection control commands and fire protection responses to the remote monitoring center and fire control center of the target electrical cabinet:

[0074] The fire feedback module receives information from the fire execution module in real time via an Ethernet interface. It performs CRC format verification on the received information to verify the integrity of the data frames. The information includes fire control commands and fire responses. It establishes communication connections with the remote monitoring center and the fire control center through industrial Ethernet communication. At the same time, the fire feedback module generates a log file daily to record the information feedback process. The record includes transmission time, transmission content, transmission status, and forms a feedback log.

[0075] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0076] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A low voltage switchgear fire protection system with arc detection, characterized in that, include: Data acquisition module: The low-voltage distribution cabinet is preset as the target cabinet. The original parameters of the target cabinet are collected in real time by deploying multi-source sensors. The original parameters include electrical parameters, arc characteristic parameters, and environmental parameters. Analysis and processing module: Receives raw parameters of the target electrical cabinet in real time, performs standardized preprocessing on the raw parameters, and calculates the temporal correlation between electrical parameters and arc characteristic parameters, as well as the spatial correlation between arc characteristic parameters and environmental parameters, using Pearson correlation coefficient. Based on the correlation analysis results, it selects raw parameters related to fire prevention diagnosis to obtain a fire prevention diagnosis dataset. Fire prevention diagnostic module: A fire prevention model for low-voltage distribution cabinets is constructed based on a fire prevention diagnostic dataset. The model includes an arc fault identification unit and a fire risk assessment unit. The arc fault identification unit uses a random forest algorithm, taking arc feature parameters and electrical parameters as inputs, and outputs the arc fault type. The fire risk assessment unit constructs a multi-factor assessment matrix based on the fault type and environmental parameters and performs weighted calculations to obtain a comprehensive risk coefficient. Based on the comprehensive risk coefficient, risk levels are classified and fire prevention control instructions are generated. The risk levels include Level 3 risk, Level 2 risk, and Level 1 risk. Fire prevention execution module: Receives fire prevention control commands in real time and executes corresponding fire prevention responses. When the command shows a level 3 risk, it activates the audible and visual alarm device. When the command shows a level 2 risk, it cuts off the power supply to the faulty branch circuit based on the level 3 risk. When the command shows a level 1 risk, it cuts off the main power supply to the electrical cabinet and activates the fire extinguishing device. Fire protection feedback module: Feeds back fire protection control commands and fire protection responses to the remote monitoring center and fire control center of the target electrical cabinet, and records the feedback process to form a feedback log.

2. A low voltage switchgear fire protection system with arc detection according to claim 1, characterized in that, The method of collecting raw parameters of the target electrical cabinet in real time by deploying multi-source sensors specifically includes: By deploying current sensors and voltage sensors in the power distribution cabinet, real-time acquisition of current parameters in electrical parameters and voltage parameters ; through the ultraviolet sensor, real-time acquisition of arc light duration in arc characteristic parameters , arc light waveform ; through the temperature sensor, humidity sensor, smoke sensor, real-time acquisition of temperature , humidity , carbon monoxide concentration .

3. The low voltage switchgear fire protection system with arc detection according to claim 1, wherein, The original parameters are then standardized and preprocessed. Firstly adopt The criteria involve removing outliers from the original parameters, filling in missing values ​​using linear interpolation, converting the analog signals of the original parameters into digital signals, and finally performing... Normalization maps the original parameters of different dimensions to... Interval.

4. The low voltage switchgear fire protection system with arc detection according to claim 1, wherein, The analysis and processing module obtains a fire prevention diagnostic dataset including: The correlation between the original parameters is calculated by using the Pearson correlation coefficient. Based on the correlation results, the original parameters related to electrical fire prevention diagnosis are selected to form a fire prevention diagnosis dataset. The calculation of the Pearson correlation coefficient includes: calculating the temporal correlation between electrical parameters and arc characteristic parameters, and calculating the spatial correlation between arc characteristic parameters and environmental parameters. The time sequence correlation of the electrical parameters and the arc characteristic parameters is calculated, and the electrical parameters are analyzed and the arc characteristic parameters synchronous association changing over time, which includes time window partitioning of the electrical parameters and the arc characteristic parameters, partitioning the parameters by setting the window size to 50 collection points and the step length to 2, and performing Pearson correlation calculation on the parameters in the window 、 、 、 ; Computing arc characteristic parameters spatial correlation with environmental parameters , analyzing the coexistence correlation of arc characteristic parameters and environmental parameters in the same space, such as whether the increase of arc light duration is accompanied by the correlation of the increase of carbon monoxide concentration, extracting the relevant arc characteristic parameters and environmental parameters of the bus area and branch area of space division, and respectively performing Pearson correlation calculation on the parameters in the space, which specifically includes 、 、 、 、 、 ; Since electric arcs are a precursor to fires in low-voltage distribution cabinets, parameters that are significantly correlated with electric arc characteristics are selected and retained based on the Pearson correlation calculations between electrical parameters and electric arc characteristic parameters, and between electric arc characteristic parameters and environmental parameters, thus forming a fire prevention diagnostic dataset.

5. The low voltage switchgear fire protection system with arc detection of claim 1, wherein, The fire protection model for low-voltage distribution cabinets is constructed based on the fire prevention diagnostic dataset. The system receives the fire prevention diagnosis dataset output by the analysis and processing module, and builds a low-voltage distribution cabinet fire prevention model based on the dataset, which includes an arc fault identification unit and a fire risk assessment unit. The arc fault identification unit adopts a classification model using the random forest algorithm, takes the arc feature parameters and electrical parameters in the dataset as input, and outputs the arc fault type. The fire risk assessment unit constructs a multi-factor assessment matrix based on the fault type and the environmental parameters output by the arc fault identification unit, and calculates the comprehensive risk coefficient by substituting the original parameters into the assessment matrix.

6. The low voltage switchgear fire protection system with arc detection of claim 1, wherein, According to the comprehensive risk coefficient of the fire prevention model, the risk level is divided and the fire prevention control instruction is generated. Risk levels are classified based on the comprehensive risk coefficient of the fire prevention model. When the comprehensive risk coefficient is within... The risk level is divided into three levels when the interval is within a certain range; when the comprehensive risk coefficient is within a certain range... When the risk level is within a certain range, it is classified as Level 2 risk; when the overall risk is within a certain range... The time interval is divided into Level 1 risk, and three levels of fire control instructions are generated based on the risk level.

Images