Electrical fire early-warning method and system based on simulation
By using a simulation-based electrical fire risk feature database and a real-time monitoring system, the problems of lag and high cost of existing electrical fire early warning systems are solved, enabling early warning and risk classification of electrical fires. This system is applicable to high-rise buildings and other electrical fire safety situations.
Patent Information
- Application Number
- PCT/CN2025/143636
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-08
- Filing Date
- 2025-12-18
- Publication Date
- 2026-11-12
AI Technical Summary
Existing electrical fire early warning systems are slow to respond, have a high false alarm rate, and rely on experimental extraction of fire risk feature databases, which is costly and risky, and lacks advanced identification capabilities.
An electrical fire risk feature database is established based on simulation. By constructing an electrical fire simulation model, generating simulation result data, establishing a fire risk feature database, training a fire risk early warning model, using LSTM deep learning for prediction, and combining real-time monitoring equipment for graded early warning of electrical fire risks.
It enables early warning of electrical fires, reduces costs, improves the accuracy and reliability of warnings, is applicable to various occasions, has high portability, and guides fire safety and equipment maintenance.
Smart Images

Figure CN2025143636_12112026_PF_FP_ABST
Abstract
Description
Methods and systems for early warning of electrical fires based on simulation.
[0001] This disclosure claims priority to Chinese patent application No. 202510590108.9, filed on May 8, 2025, the entire contents of which are incorporated herein by reference. Technical Field
[0002] This disclosure relates to the field of early warning technology for electrical fires, and in particular to an early warning method and system for electrical fires based on simulation. Background Technology
[0003] With the acceleration of urbanization, the demand for electricity from residents and industries has increased significantly, and fire safety issues have become increasingly prominent. Electrical fires, due to their suddenness, rapid spread, and difficulty in being extinguished, pose a huge threat to people's lives and property.
[0004] Electrical fires are not only caused by traditional circuit faults, but can also be caused by modern household appliances, electric vehicles and other new electrical equipment, making fire prevention and control more complex. Summary of the Invention
[0005] In a first aspect, this disclosure provides a simulation-based early warning method for electrical fires, including:
[0006] Establish electrical fire simulation models for different fire types and environmental conditions, set basic parameters and simulation parameters of the simulation model, run the simulation model, and generate multiple sets of simulation result data.
[0007] A fire risk feature database is established by using multiple sets of simulation results data generated for different fire types and environmental conditions;
[0008] Using the established fire risk feature database, a fire risk early warning model is trained;
[0009] The trained fire risk prediction model is used to process real-time monitored temperature and gas concentration data to predict the trend of characteristic changes over a future period.
[0010] The trained fire risk early warning model is used to process the predicted trend of characteristic changes over a future period to determine the electrical fire risk classification and early warning results.
[0011] In some embodiments, electrical fire simulation models are established for different fire types and environmental conditions. Basic parameters and simulation parameters of the simulation model are set, the simulation model is run, and multiple sets of simulation result data are generated, including:
[0012] Establish a simulation model of the spatial area of a high-rise building, set the basic parameters of the spatial area, and define the spatial area as any independent space within the high-rise building;
[0013] Set the simulation parameters for the spatial area. The simulation parameters include: monitoring points and status observers, high-temperature points or heat sources, and ignition points or fire sources.
[0014] Using the established high-rise building spatial area simulation model, the set basic parameters of the spatial area and simulation parameters, the entire process of fire in the risky electrical equipment within the spatial area is simulated.
[0015] Once the simulation begins, the state observer records the state variables of the local area where the monitoring point is located in real time. These state variables include temperature, as well as the type and concentration of gases produced by the combustion or thermal decomposition of the substance.
[0016] After the simulation is completed, simulation result data is generated based on the change process of each state variable.
[0017] In some embodiments, each set of simulation result data includes temperature data and gas concentration at each monitoring point obtained after the simulation model is run; using multiple sets of simulation result data generated for different fire types and environmental conditions, a fire risk feature library is established, including:
[0018] The simulation results data are cleaned, including handling missing values, handling outliers, and standardizing the simulation results data.
[0019] Fire risk characteristics are extracted from the cleaned simulation data, including temperature change rate and peak gas concentration. These characteristics are then classified and labeled according to whether a fire risk exists and whether a fire has occurred.
[0020] After establishing the labeled fire risk characteristics, a fire risk characteristic database will be created for different fire types and environmental conditions. The fire risk characteristic database includes equipment fire risk characteristics and fire risk level information.
[0021] In some embodiments, the fire risk early warning model includes a feature extraction sub-model and a risk early warning sub-model; training the fire risk early warning model using an established fire risk feature library includes:
[0022] The feature extraction sub-model is used to preprocess and extract features from the raw data included in the fire risk feature database to generate processed data.
[0023] A risk warning sub-model is constructed based on decision trees or random forests. The processed data is then used to pre-train the constructed risk warning sub-model, thereby generating a trained fire risk warning model.
[0024] In some embodiments, the simulation result data for each group includes temperature data and gas concentration at each monitoring point obtained after the simulation model is run; the method further includes: training a fire risk prediction model, including:
[0025] The simulation results data are cleaned, including handling missing values, handling outliers, and standardizing the simulation results data.
[0026] Extract the time series of temperature change rate and gas concentration change rate from the cleaned simulation results data;
[0027] Using time series data on temperature change rate and gas concentration change rate, a fire risk prediction model based on Long Short Term Memory (LSTM) deep learning is trained. The trained fire risk prediction model is used to predict the trend of feature changes over a future period based on the input historical data. The trend of feature changes includes the trend of temperature change rate and the trend of gas concentration change rate.
[0028] Secondly, this disclosure provides an early warning real-time monitoring system for electrical fires, including: multiple real-time monitoring devices and a central monitoring device.
[0029] Multiple real-time monitoring devices are distributed in different areas within a preset range. These devices are used to monitor different areas within the preset range and output fire risk monitoring information.
[0030] The central monitoring equipment is used to receive fire risk monitoring information from multiple real-time monitoring devices in different areas within a preset range.
[0031] The central monitoring equipment includes a visual user interface, which displays the fire risk level of different areas within a preset range using audiovisual controls.
[0032] In some embodiments, the real-time monitoring device includes a temperature sensor, a gas sensor, and a data processing unit.
[0033] Real-time monitoring equipment is installed in any of the following locations: cable wells, control rooms or switch rooms leading to mezzanines, around cables passing through floors, walls, or between panels, or near hot bodies passing through high-temperature pipes.
[0034] The data processing unit uses the trained fire risk prediction model to process the temperature data monitored in real time by the temperature sensor and the gas concentration data monitored in real time by the gas sensor, and predicts the trend of characteristic quantity changes in the future.
[0035] The data processing unit is also used to process the predicted trend of characteristic changes over a future period using the trained fire risk early warning model, and to determine the electrical fire risk classification and early warning results.
[0036] In some embodiments, the electrical fire risk classification and early warning results include:
[0037] Low risk, but there may be a fire risk.
[0038] Medium risk, with a significant fire risk;
[0039] High risk, with a very high risk of fire.
[0040] Thirdly, this disclosure provides a terminal, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method in the first aspect above.
[0041] Fourthly, this disclosure provides a computer storage medium storing a computer program for performing the method described in the first aspect above. Attached Figure Description
[0042] The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this disclosure. Furthermore, the same reference numerals denote the same parts throughout the drawings.
[0043] Figure 1 is a schematic flowchart of an early warning method for electrical fires based on simulation according to an embodiment of the present disclosure.
[0044] Figure 2 is a schematic diagram of a simulation model of a high-rise building space area according to an embodiment of the present disclosure.
[0045] Figure 3 is a schematic diagram of a cross-section of a polyvinyl chloride (PVC) cable according to an embodiment of the present disclosure.
[0046] Figure 4 is a schematic diagram of the composition of an early warning real-time monitoring system for electrical fires according to an embodiment of the present disclosure.
[0047] Figure 5 is a schematic diagram of a visual user interface of a real-time monitoring system for early warning of electrical fires in high-rise buildings according to an embodiment of the present disclosure.
[0048] Figure 6 is a flowchart illustrating the development of a real-time monitoring system for early warning of electrical fires in high-rise buildings based on a simulation-based method for early warning of electrical fires in high-rise buildings, according to an embodiment of the present disclosure.
[0049] Figure 7 is a schematic diagram of the composition of a terminal according to an embodiment of the present disclosure.
[0050] Figure 8 is a schematic diagram of a program product according to an embodiment of the present disclosure. Detailed Implementation
[0051] Exemplary embodiments of this disclosure are described below with reference to the accompanying drawings. However, this disclosure can be implemented in many different embodiments and is not limited to the embodiments described herein. The terminology used in the exemplary embodiments represented in the drawings is not intended to limit this disclosure. In the drawings, the same units / elements are referred to by the same reference numerals.
[0052] Unless otherwise stated, the terms used herein (including technical terms) have their usual meanings as understood by one of ordinary skill in the art. Furthermore, it is understood that terms defined in commonly used dictionaries should be understood to have meanings consistent with the context of their relevant field, and not to be interpreted as idealized or overly formal.
[0053] According to the building fire protection design code, high-rise buildings are residential buildings with a building height greater than 27 meters (m) and non-single-story factories, warehouses and other civil buildings with a building height greater than 24 meters.
[0054] High-rise building fires often exhibit a "chimney effect." In high-rise building fires, air convection causes smoke and heat to rise, and the fire often spreads rapidly upwards in a short period of time through relatively enclosed spaces such as stairwells, pipe shafts, and gaps in glass curtain walls. This is one of the reasons why high-rise fires are difficult to control.
[0055] Electrical fires in high-rise buildings are characterized by rapid fire spread, difficulty in evacuating people, difficulty in fighting fires, complex electrical equipment and wiring, and a large number of flammable materials, requiring more stringent fire prevention strategies.
[0056] With the development of technology, modern fire early warning systems are beginning to adopt more advanced technologies, such as smoke detection technology, big data analysis, and artificial intelligence algorithms, to improve the accuracy of early warning and the efficiency of rescue. These advanced technologies enable real-time monitoring and intelligent analysis of electrical systems, such as monitoring changes in wire resistance to detect line aging, monitoring overload and short-circuit currents, detecting leakage current, and monitoring abnormal temperatures.
[0057] Traditional fire early warning systems, such as smoke detectors and automatic sprinkler systems, while playing a role in warning of and extinguishing fires to some extent, have limitations such as slow response and high false alarm rates. Considering the higher power supply reliability requirements of high-rise buildings, it is necessary to develop advanced early warning methods for electrical fires in high-rise buildings to achieve early identification of fire risks and avoid losses.
[0058] In recent years, significant progress has been made in the extraction of methods for predicting and warning electrical fire risks. For example, a method for predicting and warning electrical fire risks based on big data in fire protection has been proposed, providing a new approach for early detection of electrical fire risks through dynamic assessment.
[0059] In addition, multi-sensor information fusion technology has also been applied to electrical fire early warning extraction, which improves the accuracy and reliability of the early warning system by integrating data from different sensors.
[0060] Meanwhile, data integration technology has been incorporated into the fire early warning system, enabling real-time monitoring and display, intelligent data analysis algorithms to identify abnormal patterns, alarm and event management, and resource scheduling and command, providing intelligent support for fire early warning.
[0061] Fire warnings often occur in the early stages of a fire or after the fire has already started, resulting in a significant delay. Furthermore, establishing a fire risk feature database relies on specific experimental extraction methods, which are time-consuming and costly, and the experiments themselves carry inherent risks. Therefore, it is necessary to establish a fire risk feature database and achieve early warning of fire risks without conducting additional fire simulation experiments.
[0062] To address the above issues, this disclosure provides a simulation-based early warning method and system for electrical fires, which can be applied to fire safety and prevention of electrical fires in high-rise buildings.
[0063] As shown in Figure 1, the simulation-based early warning method for electrical fires according to this embodiment includes the following steps S100 to S500.
[0064] S100. Establish electrical fire simulation models for different fire types and environmental conditions, set basic parameters and simulation parameters of the simulation model, run the simulation model, and generate multiple sets of simulation result data.
[0065] S200: Utilize multiple sets of simulation results data generated for different fire types and environmental conditions to establish a fire risk feature database.
[0066] S300. Using the established fire risk feature database, train the fire risk early warning model.
[0067] S400 uses a trained fire risk prediction model to process real-time monitored temperature and gas concentration data, and predicts the trend of characteristic quantity changes over a future period.
[0068] S500: The trained fire risk early warning model is used to process the predicted trend of characteristic quantity changes over a future period to determine the electrical fire risk classification and early warning results.
[0069] This early warning method for electrical fires saves costs and avoids the risks of fire simulation experiments. Based on simulation, an electrical fire risk feature database is established, enabling differentiated early warnings. It is applicable to various scenarios and has high portability.
[0070] Thus, by utilizing simulation-based early warning methods for electrical fires, we can guide the deployment of fire safety measures and the replacement and maintenance of aging electrical equipment, ensure the safe and reliable operation of power distribution systems, and provide a reliable reference for maintenance work.
[0071] In some embodiments, as shown in Figures 2 and 3, an electrical fire simulation model is established for different fire types and environmental conditions. The basic parameters and simulation parameters of the simulation model are set, the simulation model is run, and multiple sets of simulation result data are generated, including the following:
[0072] Establish a simulation model of the spatial area of a high-rise building, set the basic parameters of the spatial area, and define the spatial area as any independent space within the high-rise building.
[0073] Set the simulation parameters for the spatial area. The simulation parameters include: monitoring points and status observers, high temperature points or heat sources, and ignition points or fire sources.
[0074] Using the established high-rise building spatial area simulation model, the set basic parameters of the spatial area, and the spatial area simulation parameters, the entire process of fire in the risky electrical equipment within the spatial area is simulated. After the simulation calculation begins, the state observer records the various state quantities of the local area where the monitoring point is located in real time. These state quantities include: temperature, and the type and concentration of gases produced by the combustion or thermal decomposition of substances.
[0075] After the simulation is completed, simulation result data is generated based on the change process of each state variable.
[0076] In some implementations, simulation software is used to simulate the fire-induced decomposition process or thermal decomposition process of electrical equipment. Based on this simulation, gas characteristic data of electrical fires are extracted, and an electrical fire risk characteristic database is established.
[0077] In this way, an electrical fire risk characteristic database can be established quickly, conveniently, and safely, providing a basis for judging electrical fire risk classification and early warning. Subsequently, electrical fire risk classification and early warning can be implemented based on the established electrical fire risk characteristic database.
[0078] In addition to fire safety scenarios in high-rise buildings, the early warning method is applicable to various applications with electrical fire safety early warning. The basic parameters or simulation parameters of the simulation model should be adjusted according to different application scenarios or monitoring objects.
[0079] In some implementations, the impact of floor height on the early characteristics of electrical fires can be ignored when the application scenario is not a high-rise building. In situations with high safety requirements or harsh environments (such as high-temperature or humid environments), the model parameters are corrected. This method has a degree of portability and can be applied to different objects or scenarios for early warning of electrical fire risks.
[0080] In some embodiments, each set of simulation result data includes temperature data and gas concentration at each monitoring point obtained after the simulation model is run. A fire risk feature library is established using multiple sets of simulation result data generated for different fire types and environmental conditions, including the following:
[0081] The simulation results data are cleaned, including handling missing values, handling outliers, and standardizing the simulation results data.
[0082] Fire risk characteristics are extracted from the cleaned simulation results data. These characteristics include: temperature change rate and peak gas concentration.
[0083] Fire risk characteristics are classified and labeled according to whether there is a fire risk and whether a fire has occurred.
[0084] After establishing the labeled fire risk characteristics, a fire risk characteristic database will be created for different fire types and environmental conditions. The fire risk characteristic database includes equipment fire risk characteristics and fire risk level information.
[0085] In some implementations, the process of ignition and decomposition of materials in a high-rise building is simulated using the visualization simulation software Pyrosim, including the following steps S110, S120 and S130.
[0086] Step S110: Establish a simulation model of the spatial area of the high-rise building as shown in Figure 2 and set the basic parameters of the spatial area. For example, the spatial area is an independent space within the high-rise building.
[0087] In some implementations, the basic parameters of the space area include: the length / width / height / area size and floor height of the floor, four walls and ceiling that constitute the space area; the distribution of ventilation openings in the space area; information on the distribution of flammable materials in the space area (such as the location of at least one flammable material (such as a cardboard box) in the room); and the gaseous environment in the space area.
[0088] In some implementations, the basic parameters of the spatial region also include the operating parameters of the hazardous electrical equipment within the spatial region; the operating parameters of the hazardous electrical equipment include: the type, material, and geometric parameters (length, width, height, perimeter, area, volume, or capacity) of the hazardous electrical equipment.
[0089] Furthermore, considering the special characteristics of fires in high-rise buildings, the height of the floor where the spatial area is located (the height of the spatial area) is taken into account and used as a basic parameter of the spatial area.
[0090] Furthermore, since fires in super high-rise buildings are affected by convection, chimney effect and high-speed wind, basic parameters such as air pressure and gas velocity related to height are set in the simulation model of the high-rise building space area.
[0091] Step S120: Set the simulation parameters for the spatial area. The simulation parameters include: monitoring points and status observers, high-temperature points or heat sources, and ignition points or fire sources.
[0092] In some implementations, high-temperature points or heat sources, as well as ignition points or fire sources, are typically placed on the surface or inside of the hazardous electrical equipment.
[0093] In some implementations, multiple monitoring points are set up within a spatial area; the monitoring points are generally set up on the surface of the hazardous electrical equipment, in the surrounding environment of the hazardous electrical equipment, and in the upper corner of the spatial area; a status observer is set up at the monitoring point, which is used to record various status quantities of the local area where the monitoring point is located during the entire process of the hazardous electrical equipment ignition, such as temperature, humidity, and the type and concentration of gas.
[0094] Step S130: Using the established simulation model of the spatial area of the high-rise building, the set basic parameters of the spatial area, and the simulation parameters of the spatial area, the entire process of fire of the risky electrical equipment in the spatial area is simulated.
[0095] In some implementations, after the simulation begins, the state observer records in real time the various state variables of the local area where the monitoring point is located. These state variables include temperature, humidity, and the types and concentrations of gases produced by combustion or thermal decomposition of substances. After the simulation ends, multiple time series are generated based on the changes in these state variables.
[0096] Returning to steps S110 and S120, a high-rise building electrical fire simulation model can be established for different fire types and environmental conditions. The basic parameters and / or simulation parameters of the set spatial area can be changed, such as changing the type and material of the electrical equipment, or changing the location of the high temperature point or ignition point. Then, step 130 is executed to run the simulation model, simulate the entire process of electrical equipment fire under various scenarios, record the state quantities of the local area where the monitoring point is located in real time, and generate simulation result data.
[0097] Record the basic parameters and simulation parameters of each simulation operation, as well as the corresponding simulation result data, such as the time series of each or multiple state variables mentioned above.
[0098] Therefore, by using digital simulation to study the early risk characteristics of electrical fires in high-rise buildings, and by adjusting the simulation parameters of the simulated space and the parameters of the at-risk electrical equipment to adjust the fire process simulation, not only can economic and time costs be reduced, but the fire hazard problem in the actual experimental process can also be avoided.
[0099] A risk feature database for electrical fires in high-rise buildings was established using simulation results from various scenarios. This database can be used for risk classification and early warning of electrical fires under various conditions, and is of great significance for early warning of electrical fire risks.
[0100] Figure 2 illustrates the spatial area of the high-rise building mentioned in this embodiment. For example, an enclosed space with length a, width b, and height h is constructed, and a ventilation opening with length w and width w is set at a position h1 from the ceiling on the front wall.
[0101] Polyvinyl chloride (PVC) cables are used as hazardous electrical equipment. Figure 3 illustrates the cross-section of the cable. As shown in Figure 3, the inner core material of the PVC cable is copper, with an inner radius r equal to p (r = p); the outer insulation material is PVC, with an outer radius R equal to q (R = q). As shown in Figure 2, the cable length is set to l, and three identical PVC cables d1, d2, and d3 are arranged parallel to each other at a vertical distance y from the ground, with adjacent cables spaced x apart along the width of the room.
[0102] As shown in Figure 2, a fire risk point (such as a heat source, fire source, high temperature point, or ignition point) is set at the center point Q of cable d1. Monitoring points are set at locations A, B, C, D, E, and F in the room. Point A coincides with the heat source Q (set on the surface of the risky electrical equipment), points B and C are at different distances directly above point A (set around the risky electrical equipment), and points D, E, and F are located in different corners inside the room, near the top of the room (set in the corners of the simulated space).
[0103] After the simulation starts, the temperature at each point and the type and concentration of characteristic gases (such as carbon monoxide, carbon dioxide, and hydrogen chloride) are monitored, and the electrical signals (current or voltage) of the three cables are also monitored.
[0104] Considering that high-rise building fires are affected by floor height and the surrounding environment, basic parameters are set for the upper and lower floors of the simulated space, as well as several other adjacent rooms on the same floor. These parameters include identifying hazardous electrical equipment, heat sources or fire sources, and multiple monitoring points. Simultaneously, the ambient air pressure and gas flow rate are set according to the floor height of the simulated space. At this point, a visual simulation model corresponding to the simulated space in Figure 2 can be created in Pyrosim software.
[0105] Set the heating source t1 at point Q to trigger the start of the simulation. Record the data acquisition status (temperature, type and concentration of characteristic gas) at each monitoring point. Monitor the electrical signals (current or voltage) of the three cables until obvious smoke spread is observed inside the simulation space, or the carbon particle concentration is set to reach a preset threshold in the simulation software. Then stop the simulation and store the simulation process data.
[0106] Furthermore, referring to the foregoing description, the working parameters or simulation parameters, such as the location of the fire risk point (e.g., moving point Q to cable d2), the initial temperature of the heating source t1, the location of the monitoring point, the floor height, etc., are modified to simulate the corresponding fire ignition process and record the information collection results.
[0107] Furthermore, the type of risky equipment can be changed, such as replacing PVC cables with meter boxes. Operating or simulation parameters, such as the location of fire risk points (e.g., moving point Q to cable d2), the initial temperature of heating source t1, the location of monitoring points, and floor height, can be modified to simulate the corresponding fire ignition process and record the information collection results, thereby changing the fire type and environmental conditions.
[0108] In some embodiments, the fire risk early warning model includes a feature extraction sub-model and a risk early warning sub-model; training the fire risk early warning model using an established fire risk feature library includes:
[0109] The feature extraction sub-model is used to preprocess and extract features from the raw data recorded in the fire risk feature database to generate processed data.
[0110] A risk warning sub-model is constructed based on decision trees or random forests. The processed data is then used to pre-train the constructed risk warning sub-model, thereby generating a trained fire risk warning model.
[0111] In some embodiments, each set of simulation result data includes temperature data and gas concentration at each monitoring point obtained after the simulation model is run.
[0112] The method also includes: training a fire risk prediction model. Training the fire risk prediction model includes:
[0113] The simulation results data are cleaned, including handling missing values, handling outliers, and standardizing the simulation results data.
[0114] Extract the time series of temperature change rate and gas concentration change rate from the cleaned simulation results data;
[0115] Using time series data on temperature change rate and gas concentration change rate, a fire risk prediction model based on Long Short Term Memory (LSTM) deep learning is trained. The trained fire risk prediction model is used to predict the trend of feature changes over a future period based on the input historical data; the trend of feature changes includes the trend of temperature change rate and the trend of gas concentration change rate.
[0116] As shown in Figure 4, an embodiment of this disclosure provides a simulation-based real-time monitoring system for early warning of electrical fires, which can be applied to the prevention of electrical fires in high-rise buildings. The system includes multiple real-time monitoring devices 10 and a central monitoring device 20.
[0117] Multiple real-time monitoring devices 10 are distributed in different areas within a preset range. These devices are used to monitor different areas within the preset range and output fire risk monitoring information.
[0118] The central monitoring device 20 is used to receive fire risk monitoring data output by multiple real-time monitoring devices in different areas within the preset range.
[0119] The central monitoring equipment includes a visual user interface, which includes audiovisual controls to display the fire risk levels of different areas within a preset range.
[0120] In some embodiments, the real-time monitoring device includes: a temperature sensor, a gas sensor, and a data processing unit;
[0121] Real-time monitoring equipment is installed in any of the following locations: cable wells, control rooms or switch rooms leading to mezzanines, around cables passing through floors, walls, or between panels, or near hot bodies passing through high-temperature pipes.
[0122] The data processing unit is used to process the temperature data monitored in real time by the temperature sensor and the gas concentration data monitored in real time by the gas sensor using the trained fire risk prediction model, and to predict the trend of characteristic quantity changes over a future period of time.
[0123] The data processing unit is also used to process the predicted trend of characteristic changes over a future period using the trained fire risk early warning model, and to determine the electrical fire risk classification and early warning results.
[0124] In some embodiments, the electrical fire risk classification and early warning results include:
[0125] Low risk, but there may be a fire risk.
[0126] Medium risk, with a significant fire risk;
[0127] High risk, with a very high risk of fire.
[0128] As shown in Figure 4, the real-time monitoring system for early warning of electrical fires in high-rise buildings includes real-time monitoring equipment 10 and central monitoring equipment 20. The real-time monitoring equipment 10 includes: a temperature sensor, a gas sensor, and a data processing unit.
[0129] Thus, the high-rise building electrical fire early risk warning real-time monitoring system of this embodiment can centrally monitor and predict the fire risk of electrical equipment in high-rise buildings within the area.
[0130] Figure 6 illustrates the process of developing a real-time monitoring system for early warning of electrical fires in high-rise buildings using a simulation-based method. As shown in Figure 6, the process for integrating the real-time monitoring system includes: deploying the real-time monitoring system in the real space.
[0131] Based on the aforementioned simulation model of the high-rise building space area and monitoring points, monitoring points are set at corresponding locations in the real rooms in the real space, and real-time monitoring equipment 10 is set at each monitoring point. The monitoring points include: cable wells, control rooms or switch rooms leading to mezzanine floors, around cables passing through floors, walls, and between panels, and near hot bodies passing through high-temperature pipes.
[0132] Each real-time monitoring device includes sensors, a storage unit, and a data processing unit. The device measures characteristic signals in the environment, stores the measurement data, and processes the data. When changes in temperature or gas concentration related to electrical fire risk occur in the environment, the real-time monitoring devices at each monitoring point automatically process the signals, identify the fire risk level, issue warnings, and record the data. Thus, the real-time monitoring devices at various monitoring points distributed in different areas within a preset range output the fire risk level respectively.
[0133] In some implementations, no warning signal is output under normal circumstances; when the system detects a fire risk, it outputs the severity level of the fire risk.
[0134] In some implementation methods, fire risk parameters are qualitatively classified based primarily on relevant standard documents (such as international / national safety standards, such as ISO / NFPA standards) and in conjunction with actual fire safety requirements. The severity levels of fire risk are divided into the following three levels: a. Level 1: Low risk, a fire risk may exist, and on-site personnel are advised to conduct an investigation or further monitoring; b. Level 2: Medium risk, a significant fire risk exists, and necessary fire-fighting measures should be taken, including fire warnings to personnel in the area; c. Level 3: High risk, a fire risk is confirmed, and immediate fire-fighting actions should be taken, such as cutting off power to the area to prevent the risk.
[0135] In this way, a fire risk classification and early warning work standard was established, which provides differentiated outputs for fire risks of different levels of danger or urgency and types of fire, and can efficiently guide staff to carry out fire safety activities.
[0136] The central monitoring equipment centrally monitors electrical fire risks in different areas within a preset range. For example, the central monitoring equipment includes a visual user interface. Real-time monitoring devices at various monitoring points distributed across different areas within the preset range uniformly send fire risk monitoring data for those areas to the central monitoring equipment, and the visual user interface displays the fire risk level using indicator lights or buzzers and other audiovisual controls. When a risk occurs, a warning signal can be issued quickly and intuitively. Furthermore, to enrich its functionality, other auxiliary functions have been designed, including displaying the connection status of monitoring devices, viewing historical trends of various indicators, entering a specified password to unlock higher operating privileges, providing users with personalized settings such as page backgrounds, receiving real-time messages from various devices, and prompts for handling pending tasks.
[0137] As shown in Figure 2, monitoring points are deployed in the actual room space, including sensor units, storage units, and data processing units. The sensor units include temperature sensors, smoke sensors, and gas sensors. The storage unit stores the established fire risk characteristic database and the pre-trained fire risk prediction model. The data processing unit integrates data preprocessing functions and prediction models. When data changes are collected, it can automatically start data processing and risk assessment.
[0138] Figure 5 shows a visual user interface that provides real-time signal output for staff. Figure 5 illustrates the normal user interface display. The icons at the bottom, from left to right, represent: 1. Network Status: Displays the connection status of monitoring devices; 2. Data Trend Table: Views historical trends of various indicators; 3. Permission Granting: Assigns and grants higher permissions; 4. Settings: Allows users to personalize page settings such as background; 5. Mailbox: Receives messages from various devices promptly; 6. To-Do Items: Red dots indicate whether there are currently any tasks to be processed, and the flashing speed indicates the urgency. When sensors detect abnormal data signals, the signals are first processed and classified within the integrated unit, while potential risks are predicted. The signals are then sent to the central control unit, and alarm signals are issued via indicator lights and buzzers. The user interface displays the on-site monitoring screen and thermal imaging image. Besides high-rise buildings, this real-time monitoring system can also be integrated into residential buildings, manufacturing plants, substations, and other locations to maintain electrical safety in different places, which is of great significance for ensuring electrical safety.
[0139] In some embodiments, as shown in FIG6, the process includes the following.
[0140] The obtained simulation results data are standardized, such as deleting abnormal data, identifying and removing interfering data, and standardizing and normalizing the data.
[0141] Based on the fire risk level corresponding to the entire fire process of each electrical equipment and the corresponding combination of simulation result data, a feature database is established to indicate the electrical fire risk level of high-rise buildings.
[0142] Furthermore, the simulation results data are processed.
[0143] First, the extracted result data is preprocessed, such as normalizing result data with different physical units or orders of magnitude to the range of 0-1, so as to improve the model training effect and reduce the impact of noise.
[0144] Secondly, the simulation results data under different fire types and environmental conditions are divided into normal stage and different stages of fire development (normal stage, low risk, medium risk and high risk) according to time sequence. These datasets contain various fire types and environmental conditions, as well as the corresponding fire risk levels.
[0145] Furthermore, a fire risk characteristic database should be established.
[0146] The data such as temperature and gas concentration obtained after running the simulation model are cleaned, including handling missing values, outliers, and standardizing the data.
[0147] Relevant features, such as temperature change rate and peak gas concentration, are extracted from the cleaned raw data. These features are then categorized according to whether a fire risk exists and whether a fire has occurred, creating a labeled dataset. Based on this, a fire risk feature library can be created, which includes the correspondence between equipment fire risk features (abnormal temperature, gas concentration, etc.) and fire risk levels (low, medium, high).
[0148] The fire risk early warning model is generally divided into two parts: feature extraction and risk warning.
[0149] The feature extraction part is not pre-trained and is designed to perform preprocessing and feature extraction on the raw data.
[0150] The early warning function is pre-trained. A suitable early warning model (such as decision tree, random forest, etc.) is selected, and the relationship between the feature combination obtained from A and the fire risk level is input into the model for training, so as to learn which feature combinations indicate fire risk.
[0151] The complete model has the following functions: after inputting temperature data and gas concentration data for a certain period of time, it can automatically preprocess the raw data (denoising, handling missing values, normalization, etc.) and extract features. The model can also analyze the extracted features through the early warning function to determine the current fire risk level.
[0152] The fire risk prediction model processes complete data such as temperature and gas concentration recorded in the simulation, extracts different feature sequences (such as temperature change rate, gas concentration change rate, etc.), and uses time series (such as deep learning models such as LSTM) to analyze the feature change patterns and trends.
[0153] The resulting model can predict the trend of feature changes over a future period based on the input historical data.
[0154] The characteristic changes predicted by the fire risk prediction model are input into the fire risk early warning model, which can then be used to assess future fire risks.
[0155] Thus, the fire risk feature database provides characteristic data about equipment risks and the relationship between these characteristics and fires, serving as the core data source for training early warning and prediction models. The fire risk early warning model is mainly responsible for short-term risk identification, detecting anomalies in real-time data and issuing warnings, enabling managers to respond and make decisions in real time. The fire risk prediction model identifies and judges fire risks over a future period, assisting in risk prevention and equipment maintenance.
[0156] The simulation-based early warning method and system for electrical fires disclosed herein can provide early warning of electrical fires in high-rise buildings, comprehensively reflect the electrical fire risk situation in monitored areas, and provide fire safety information accordingly. This disclosure provides methodological guidance on when to conduct fire safety inspections and replacements of electrical equipment in monitored areas, which can maximize the safety of electrical equipment in the area and help reduce the possibility of fire accidents.
[0157] This disclosure also provides a terminal for executing the methods described above. FIG7 shows a schematic diagram of a terminal provided in some embodiments of this disclosure. Referring to FIG7, the terminal 8 includes: a processor 800, a memory 801, a bus 802, and a communication interface 803. The processor 800, the communication interface 803, and the memory 801 are connected via the bus 802. The memory 801 stores a computer program that can run on the processor 800, and when the processor 800 runs the computer program, it executes the methods provided in any of the embodiments of this disclosure described above.
[0158] The memory 801 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between the device network element and at least one other network element is achieved through at least one communication interface 803 (which may be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.
[0159] Bus 802 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. The memory 801 is used to store programs. After receiving an execution instruction, the processor 800 executes the program. The methods described in any embodiment of this disclosure can be applied to the processor 800, or implemented by the processor 800.
[0160] The processor 800 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the method can be completed through integrated logic circuits in the hardware of the processor 800 or through software instructions. The processor 800 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components; it can implement or execute the methods, steps, and logic block diagrams in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method described in conjunction with the embodiments of this disclosure can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 801. Processor 800 reads the information in memory 801 and, in conjunction with its hardware, completes the steps of the above method.
[0161] The terminal provided in this disclosure embodiment is based on the same inventive concept as the method in this disclosure embodiment embodiment, and has the same beneficial effects as the method it adopts, operates or implements.
[0162] This disclosure also provides a computer-readable storage medium (e.g., a non-transitory computer-readable storage medium) corresponding to the methods provided in the above embodiments. The computer-readable storage medium is an optical disc, on which a computer program (i.e., the program product 900 shown in FIG8) is stored. When the computer program is run by a processor, it executes the methods provided in any of the above embodiments.
[0163] It should be noted that computer-readable storage media may also include, but are not limited to, phase-change random access memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other optical or magnetic storage media, which will not be elaborated here.
[0164] The computer-readable storage medium provided in the embodiments of this disclosure is based on the same inventive concept as the methods in the embodiments of this disclosure and has the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.
[0165] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its scope. Therefore, if such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, this disclosure is also intended to include such modifications and variations.
Claims
1. A simulation-based early warning method for electrical fires, comprising: A simulation model for electrical fires is established for different fire types and environmental conditions. The basic parameters and simulation parameters of the simulation model are set, the simulation model is run, and multiple sets of simulation result data are generated. A fire risk feature database is established using the multiple sets of simulation result data generated for the different fire types and environmental conditions. Using the established fire risk feature database, a fire risk early warning model is trained; The trained fire risk prediction model is used to process real-time monitored temperature and gas concentration data to predict the trend of characteristic changes over a future period. The trained fire risk early warning model is used to process the predicted trend of characteristic quantity changes over the future period to determine the electrical fire risk classification and early warning results.
2. The method according to claim 1, wherein, The simulation model for electrical fires is established for different fire types and environmental conditions. The basic parameters and simulation parameters of the simulation model are set, the simulation model is run, and multiple sets of simulation result data are generated, including: Establish a simulation model of the spatial area of a high-rise building, and set the basic parameters of the spatial area, wherein the spatial area is any independent space within the high-rise building; The simulation parameters for the space region are set, including: monitoring points and status observers, heat sources, and ignition points or fire sources; Using the established simulation model of the spatial area of the high-rise building, the set basic parameters of the spatial area, and the simulation parameters of the spatial area, the entire process of fire in the risky electrical equipment within the spatial area is simulated. Once the simulation begins, the state observer records in real time the various state quantities of the local area where the monitoring point is located. These state quantities include: temperature, and the type and concentration of gases produced by the combustion or thermal decomposition of the substance. After the simulation is completed, simulation result data is generated based on the change process of each state variable.
3. The method according to claim 1 or 2, wherein, Each set of simulation results data includes temperature data and gas concentration at each monitoring point obtained after the simulation model is run; The process of establishing a fire risk feature database using multiple sets of simulation result data generated for different fire types and environmental conditions includes: The simulation result data is cleaned, including handling missing values, handling outliers, and standardizing the simulation result data. Fire risk characteristics are extracted from the cleaned simulation results data, including: temperature change rate and peak gas concentration. The fire risk characteristics are classified and labeled according to whether there is a fire risk and whether a fire has occurred; The fire risk feature library is created by establishing the marked fire risk feature quantities for different fire types and environmental conditions. The fire risk feature library includes equipment fire risk feature quantities and fire risk level information.
4. The method according to any one of claims 1 to 3, wherein, The fire risk early warning model includes a feature extraction sub-model and a risk early warning sub-model; The step of training a fire risk early warning model using the established fire risk feature database includes: The feature extraction sub-model is used to preprocess and extract features from the raw data included in the fire risk feature database to generate processed data. The risk warning sub-model is constructed based on decision trees or random forests. The processed data generated is used to pre-train the constructed risk warning sub-model to generate the trained fire risk warning model.
5. The method according to any one of claims 1 to 4, wherein, Each set of simulation results data includes temperature data and gas concentration at each monitoring point obtained after the simulation model is run; The method further includes: training the fire risk prediction model, including: The simulation result data is cleaned, including handling missing values, handling outliers, and standardizing the simulation result data. Extract the time series of temperature change rate and gas concentration change rate from the cleaned simulation results data; Using the time series of temperature change rate and the time series of gas concentration change rate, a fire risk prediction model based on long short-term memory deep learning is trained. The trained fire risk prediction model is used to predict the trend of feature quantity changes over a future period based on the input historical data. The trend of feature quantity changes includes the trend of temperature change rate and the trend of gas concentration change rate.
6. A real-time monitoring system for early warning of electrical fires, comprising: Multiple real-time monitoring devices are distributed in different areas within a preset range. These multiple real-time monitoring devices are used to monitor the different areas within the preset range and output fire risk monitoring information. The central monitoring equipment is used to receive the fire risk monitoring information output by the multiple real-time monitoring devices in different areas within the preset range; The central monitoring equipment includes a visual user interface, which includes audiovisual controls for displaying the fire risk levels of different areas within the preset range.
7. The system according to claim 6, wherein, Each real-time monitoring device includes: a temperature sensor, a gas sensor, and a data processing unit; The real-time monitoring equipment is installed in any of the following locations: cable well, control room or switch room leading to the mezzanine, around cables passing through floors, walls, or between panels, or near hot bodies passing through high-temperature pipes; The data processing unit is used to process the temperature data monitored in real time by the temperature sensor and the gas concentration data monitored in real time by the gas sensor using the trained fire risk prediction model, and to predict the trend of characteristic quantity changes over a future period of time. The data processing unit is also used to process the predicted trend of feature quantity changes over the future period using the trained fire risk early warning model, and to determine the electrical fire risk classification and early warning results.
8. The system according to claim 7, wherein, The electrical fire risk classification and early warning results include: Low risk, but there may be a fire risk. Medium risk, with a significant fire risk; High risk, with a very high risk of fire.
9. A terminal, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to implement the method of any one of claims 1 to 5.
10. A computer storage medium storing a computer program for performing the method of any one of claims 1 to 5.
