Method and device for determining fire grade, electronic equipment and program product

Abstract

The invention discloses a fire hazard grade determination method and device, electronic equipment and a program product, and relates to the technical field of artificial intelligence, and the determination method comprises the steps: obtaining the time sequence data of a to-be-detected position in a target area in a current time period, determining a target meteorological feature vector based on a preset database and the time sequence data, and obtaining a target meteorological feature vector; the preset database comprises data of a plurality of positions in the target area; decomposing the time sequence data to obtain decomposed data; and inputting the time sequence data, the target meteorological feature vector and the decomposition data into a first preset model to obtain a fire danger probability value of the to-be-detected position in a first preset time period, the current time period being earlier than the first preset time period, and the fire danger probability value being used for determining a fire hazard level of the to-be-detected position. According to the invention, a technical problem that the fire hazard grade cannot be accurately determined in a complex mountain area environment in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and more specifically, to a method, apparatus, electronic device, and program product for determining fire severity. Background Technology

[0002] As power transmission networks continue to extend into mountainous and other complex terrains, mountain power transmission channels face the threat of wildfires. Wildfires can cause serious accidents such as power line tripping and tower damage, threatening the safe and stable operation of the power grid and the reliable supply of electricity.

[0003] Currently, wildfire monitoring mainly relies on manual patrols and observation towers. However, manual patrols are insufficient to cover large areas of mountainous power transmission corridors, and in the early stages of a wildfire, when the fire is small and there is little smoke, it is difficult to detect it in time. Observation towers are also limited by terrain and weather factors, making it impossible to achieve all-weather, all-round monitoring. With technological advancements, satellite remote sensing technology has been gradually applied to wildfire monitoring. However, the spatial resolution of satellite remote sensing data is relatively low, and it may not be able to accurately detect some small-scale fires or fires hidden in complex terrain such as valleys. Furthermore, satellite remote sensing has a high false negative rate for small-area wildfires, and the monitoring effect is often severely affected by rainy weather or thick cloud cover. In addition, while meteorological data is used to establish fire weather risk levels to assess the risk of wildfires, other factors that affect wildfire occurrence are ignored, resulting in insufficient accuracy of the prediction results.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This invention provides a method, apparatus, electronic device, and program product for determining fire level, to at least solve the technical problem in related technologies that it is impossible to accurately determine fire level in complex mountainous environments.

[0006] According to one aspect of the embodiments of this application, a method for determining the fire level is provided, comprising: acquiring time-series data of a location to be detected in a target area during the current time period, and determining a target meteorological feature vector based on a preset database and the time-series data, wherein the preset database includes data of multiple locations within the target area; decomposing the time-series data to obtain decomposed data; inputting the time-series data, the target meteorological feature vector, and the decomposed data into a first preset model to obtain a fire risk probability value of the location to be detected during a first preset time period, wherein the current time period is earlier than the first preset time period, and the first preset model includes at least an input layer, a hidden layer, and an output layer, wherein the input layer is used to receive input data, the hidden layer is used to extract features from the input data, and the output layer is used to output a fire risk probability value based on the features through a fully connected layer and an activation function, and the fire risk probability value is used to determine the fire level of the location to be detected.

[0007] Furthermore, before acquiring the time-series data of the locations to be detected in the target area during the current time period, the method further includes: acquiring data of multiple locations within the target area during a second preset time period, and for each location, determining the feature vector of the fire hazard factor based on the data, wherein the second preset time period is earlier than the current time period; inputting each feature vector into a second preset classification model to obtain the fire hazard classification result of the location, wherein the network model of the second preset classification model includes at least: a gating structure, which is used to extract information at each time step, and outputs the fire hazard classification result based on all information, wherein the time step is determined based on the feature vector, and the fire hazard classification result includes at least: the category and the probability value of the category; and constructing a preset database based on the data of all categories, all probability values, and all locations.

[0008] Furthermore, before acquiring data from multiple locations within the target area during the second preset time period, the method further includes: acquiring historical fire point time-series data, wherein the historical fire point time-series data includes at least: historical fire point data from multiple ignition points and a set of associated data corresponding to each historical fire point, the set of associated data including at least: meteorological data, geographic information data, and power transmission channel facility data; processing the historical fire point time-series data to obtain historical time-series processed data; constructing a historical feature vector of the fire hazard factor for each ignition point based on the historical time-series processed data; and training the initial classification model based on the historical feature vector to obtain the second preset classification model.

[0009] Furthermore, the step of constructing a historical feature vector of the fire hazard factor for each ignition point based on historical time-series processing data includes: identifying multiple fire hazard factors, wherein each fire hazard factor includes at least multiple fields; extracting features from the historical time-series processing data to obtain feature data; determining the field features of each field based on all feature data; and constructing a historical feature vector for each ignition point based on all field features.

[0010] Furthermore, the data includes at least: first meteorological sequence data. The step of constructing a preset database based on data of all categories, all probability values, and all locations also includes: for each category, determining whether the category indicates a fire at a location, and mapping the probability value of the category to obtain the fire risk level corresponding to the first meteorological sequence data if the category indicates a fire at a location; for each fire risk level, determining whether the fire risk level is within a preset level threshold range, and adding the first meteorological sequence data to the preset database if the fire risk level is within the preset level threshold range, thereby constructing the preset database.

[0011] Furthermore, the time-series data includes at least: second meteorological sequence data. The step of determining the target meteorological feature vector based on a preset database and the time-series data includes: calculating the similarity value between the second meteorological sequence data and each first meteorological sequence data in the preset database; comparing all similarity values ​​to obtain a comparison result; determining multiple target meteorological sequence data based on the comparison result; assigning weights to each target meteorological sequence data based on the similarity value corresponding to each target meteorological sequence data to obtain weighted target meteorological sequence data; and fusing all weighted target meteorological sequence data to obtain the target meteorological feature vector.

[0012] Further, the steps of decomposing the time series data to obtain the decomposed data include: determining the fire risk data corresponding to multiple fire risk factors from the time series data; using data decomposition technology to decompose all fire risk data to obtain trend terms, seasonal terms, and residual terms; and determining the decomposed data based on the trend terms, seasonal terms, and residual terms.

[0013] Furthermore, after inputting time-series data, target meteorological feature vectors, and decomposed data into the first preset model to obtain the fire risk probability value of the location to be detected in the first preset time period, the method further includes: mapping the fire risk probability value to obtain the target fire level of the fire risk probability value; and determining the fire handling strategy for the location to be detected based on the target fire level.

[0014] According to another aspect of the embodiments of this application, a fire level determination device is also provided, comprising: a determination unit, configured to acquire time-series data of a location to be detected in a target area during the current time period, and determine a target meteorological feature vector based on a preset database and the time-series data, wherein the preset database includes data of multiple locations within the target area; a decomposition unit, configured to decompose the time-series data to obtain decomposed data; and an input unit, configured to input the time-series data, the target meteorological feature vector, and the decomposed data into a first preset model to obtain a fire risk probability value of the location to be detected during a first preset time period, wherein the current time period is earlier than the first preset time period, and the first preset model includes at least an input layer, a hidden layer, and an output layer, wherein the input layer is configured to receive input data, the hidden layer is configured to extract features from the input data, and the output layer is configured to output a fire risk probability value based on the features through a fully connected layer and an activation function, and the fire risk probability value is used to determine the fire level of the location to be detected.

[0015] Furthermore, the fire level determination device also includes: a first determination module, used to acquire data of multiple locations in the target area during a second preset time period before acquiring time-series data of the locations to be detected in the target area during the current time period, and for each location, determine the feature vector of the fire hazard factor based on the data, wherein the second preset time period is earlier than the current time period; a first input module, used to input each feature vector into a second preset classification model to obtain the fire hazard classification result of the location, wherein the network model of the second preset classification model includes at least: a gating structure, the gating structure is used to extract information at each time step, and output the fire hazard classification result based on all information, the time step is determined based on the feature vector, and the fire hazard classification result includes at least: the category and the probability value of the category; and a first construction module, used to construct a preset database based on the data of all categories, all probability values, and all locations.

[0016] Furthermore, the fire level determination device also includes: a first acquisition module, used to acquire historical fire point time series data before acquiring data from multiple locations within the target area during a second preset time period, wherein the historical fire point time series data includes at least: historical fire point data from multiple ignition points and a set of associated data corresponding to each historical fire point, the set of associated data including at least: meteorological data, geographic information data, and power transmission channel facility data; a first processing module, used to process the historical fire point time series data to obtain historical time series processed data; a second construction module, used to construct a historical feature vector of the fire hazard factor for each ignition point based on the historical time series processed data; and a first training module, used to train an initial classification model based on the historical feature vector to obtain a second preset classification model.

[0017] Furthermore, the second construction module includes: a first determining submodule for determining multiple fire hazard factors, wherein each fire hazard factor includes at least multiple fields; a first extraction submodule for extracting features from historical time-series processed data to obtain feature data; a second determining submodule for determining the field features of each field based on all feature data; and a first construction submodule for constructing a historical feature vector for each ignition point based on all field features.

[0018] Furthermore, the data includes at least: first meteorological sequence data, and the first construction module includes: a first mapping submodule, used to determine whether a fire has occurred at a location indicated by the category for each category, and to map the probability value of the category to obtain the fire risk level corresponding to the first meteorological sequence data if a fire has occurred at a location indicated by the category; and a first adding submodule, used to determine whether a fire risk level is within a preset level threshold range for each fire risk level, and to add the first meteorological sequence data to a preset database if the fire risk level is within the preset level threshold range, thereby constructing the preset database.

[0019] Furthermore, the time-series data includes at least: second meteorological sequence data, and the determining unit includes: a first calculation module for calculating the similarity value between the second meteorological sequence data and each first meteorological sequence data in a preset database; a first comparison module for comparing all similarity values ​​to obtain a comparison result; a second determining module for determining multiple target meteorological sequence data based on the comparison result; a first allocation module for assigning weights to each target meteorological sequence data based on the similarity value corresponding to each target meteorological sequence data to obtain weighted target meteorological sequence data; and a first fusion module for fusing all weighted target meteorological sequence data to obtain a target meteorological feature vector.

[0020] Furthermore, the decomposition unit includes: a third determining module, used to determine the fire risk data corresponding to multiple fire risk factors from the time series data; a first decomposition module, used to decompose all fire risk data using data decomposition technology to obtain trend terms, seasonal terms and residual terms; and a fourth determining module, used to determine the decomposed data based on the trend terms, seasonal terms and residual terms.

[0021] Furthermore, the fire level determination device also includes: a first mapping module, used to map the fire probability value to obtain the target fire level of the fire probability value after inputting time series data, target meteorological feature vector and decomposed data into the first preset model to obtain the fire probability value of the location to be detected in the first preset time period; and a fifth determination module, used to determine the fire handling strategy for the location to be detected based on the target fire level.

[0022] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any of the above-described methods for determining the fire level.

[0023] According to another aspect of the embodiments of this application, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement any of the above-described methods for determining the fire level.

[0024] In this invention, time-series data of the location to be detected in the target area during the current time period is obtained. Based on a preset database and the time-series data, a target meteorological feature vector is determined. The time-series data is decomposed to obtain decomposed data. The time-series data, the target meteorological feature vector, and the decomposed data are input into a first preset model to obtain the fire risk probability value of the location to be detected during a first preset time period. The fire risk probability value is used to determine the fire level of the location to be detected, which solves the technical problem in related technologies that it is impossible to accurately determine the fire level in complex mountainous environments.

[0025] In this invention, time-series data of the target location in the target area during the current time period is first acquired. Based on a preset database including historical meteorological information and environmental data, and the time-series data, a target meteorological feature vector is determined. Then, the acquired time-series data is decomposed by removing noise and extracting long-term trends and seasonal patterns to obtain decomposed data. Subsequently, the time-series data, the target meteorological feature vector, and the decomposed data are input into a first preset model to obtain the fire risk probability value of the target location during a first preset time period. The first preset model includes an input layer, a hidden layer, and an output layer. The input layer is responsible for receiving all input data, the hidden layer mines and extracts complex features from the input data, and the output layer uses the extracted features, through a fully connected layer and an activation function, to calculate the fire risk probability value of the target location during the upcoming first preset time period. Continuous probability values ​​can be mapped to specific fire levels. By accurately determining the fire level, precise early warning information can be provided to power grid maintenance personnel, assisting them in deploying corresponding fire prevention measures in advance according to the fire level, ensuring the safe operation of transmission channels, and reducing damage caused by fire. Attached Figure Description

[0026] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0027] Figure 1A hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for determining fire severity is shown.

[0028] Figure 2 This is a flowchart of the method for determining the fire level according to Embodiment 1 of this application;

[0029] Figure 3 This is a schematic diagram of an optional fire level determination device according to an embodiment of this application;

[0030] Figure 4 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0031] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0033] It should be noted that all related information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) collected and involved in this invention are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with the relevant laws, regulations, and standards of the relevant regions, and necessary confidentiality measures have been taken. These measures do not violate public order and good morals, and corresponding operation entry points are provided for users to choose to authorize or refuse. For example, this system has an interface with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface. After receiving consent from the aforementioned user or organization, the relevant information is obtained. If the user chooses to refuse, the process proceeds to an expert decision-making process.

[0034] In this invention, by integrating multiple data sources such as satellite remote sensing, meteorological data, geographic information data, and power transmission channel facility data, and utilizing data analysis and machine learning techniques, a wildfire prediction model is constructed to achieve real-time and accurate prediction of peak wildfire periods for power transmission channels in mountainous areas. This improves the timeliness and accuracy of wildfire monitoring, while providing more targeted and refined wildfire early warning information for power transmission channel operation and maintenance personnel, optimizing the allocation of prevention and control resources, reducing the impact of wildfires on the safe operation of power transmission channels, and thus enhancing the overall effectiveness of wildfire prevention and control for power transmission channels in mountainous areas.

[0035] The present invention will now be described in detail with reference to various embodiments.

[0036] Example 1

[0037] According to an embodiment of this application, an embodiment of a method for determining the fire level is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0038] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing a method for determining fire severity is shown. Figure 1 As shown, computer terminal 10 (or mobile device) may include one or more ( Figure 1The processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions may also be included. In addition, it may include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera, wherein the network interface can be connected to wired and / or wireless networks. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0039] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0040] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the fire level determination method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned fire level determination method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0041] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0042] The display may be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0043] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for determining the fire rating is shown. Figure 2 This is a flowchart of the method for determining the fire rating according to Embodiment 1 of this application, as follows: Figure 2 As shown, the method includes the following steps:

[0044] Step S201: Obtain the time series data of the location to be detected in the target area in the current time period, and determine the target meteorological feature vector based on the preset database and the time series data. The preset database includes data of multiple locations in the target area.

[0045] Optionally, since wildfires may cause serious accidents such as power line tripping and tower damage, threatening the safe and stable operation of the power grid and the reliable supply of electricity, the target area can be the surrounding area of ​​the mountain power transmission channel or other mountainous areas not covered by the power transmission network.

[0046] In this embodiment of the invention, the time series data (e.g., fire point data, meteorological data, geographical data, etc.) of the locations to be detected in the target area (such as various monitoring points along the power transmission line or key points for fire risk prediction) are obtained in the current time period (i.e., the time window for data collection and analysis, such as the last 72 hours). Based on the preset database and the time series data, the target meteorological feature vector can be determined.

[0047] Optionally, the preset database is a database pre-generated based on historical fire data, meteorological data, geographic information data and other data collected at multiple time points. The database includes fire data from multiple locations within the target area, and the fire data is data from before the current time period (e.g., data from 72 hours ago).

[0048] Step S202: Decompose the time series data to obtain decomposed data.

[0049] In this embodiment of the invention, a method such as STL (Seasonal-Trend decomposition using Loess, a decomposition method for time series data) can be used to decompose time series data into a trend term, a seasonal term, and random noise. The trend term reflects the long-term trend of data over time, the seasonal term captures the periodic fluctuations in the data, and the residual term is the remaining, irregular, and random part of the data, which includes the impact of short-term fluctuations and sudden events, thereby enabling a clearer identification of dynamic factors affecting fire risk.

[0050] Step S203: Input the time series data, target meteorological feature vector, and decomposed data into the first preset model to obtain the fire risk probability value of the location to be detected in the first preset time period. The current time period is earlier than the first preset time period. The first preset model includes at least an input layer, a hidden layer, and an output layer. The input layer is used to receive input data, the hidden layer is used to extract features from the input data, and the output layer is used to output the fire risk probability value based on the features through a fully connected layer and an activation function. The fire risk probability value is used to determine the fire level of the location to be detected.

[0051] In this embodiment of the invention, the first preset model is a pre-trained model (such as LSTM). The first preset model includes at least an input layer, a hidden layer, and an output layer. The input layer receives input data. The hidden layer extracts the time dependency and key features in the input data through LSTM units. Its gating mechanism enables the model to selectively forget or retain information. The output layer outputs the fire risk probability value based on the features extracted by the hidden layer through a fully connected layer and an activation function. The fully connected layer may include 48 neurons, corresponding to the next 48 hours. The activation function may use Sigmoid. The output value range is [0, 1], representing the hourly fire risk probability value. The fire risk probability value is used to determine the fire level of the location to be detected.

[0052] In this embodiment of the invention, time-series data (including temperature, humidity, and wind speed), target meteorological feature vectors, and decomposed data (including trend terms and seasonal terms) can be input into a first preset model to obtain the fire risk probability value of the location to be detected in a first preset time period. Thus, the fire risk probability value can be mapped to the corresponding target fire level. For example, a fire risk probability value < 0.3 indicates a low-risk fire, 0.3 ≤ fire risk probability value < 0.7 indicates a medium-risk fire, and a fire risk probability value ≥ 0.7 indicates a high-risk fire.

[0053] In summary, by inputting meteorological time-series data, target meteorological feature vectors, and decomposed data into the first preset model, accurate prediction of future fire risk levels is achieved. The target meteorological feature vectors include historical similar pattern information, and the decomposed data includes trend and seasonal characteristics. By combining trend, seasonal characteristics, and similar pattern information, not only is the limited prediction accuracy of a single data source overcome, but also the adaptability and reliability of the model are improved through dynamic pattern matching and time series analysis. This enables real-time monitoring and prediction of peak wildfire periods in mountainous power transmission channels and determination of wildfire levels. Consequently, it solves the technical problem of accurately determining fire levels in complex mountainous environments, effectively ensuring the safe operation of the power grid, reducing the impact of wildfire disasters, and improving the stability and efficiency of power transmission.

[0054] To accurately construct a preset database, in the fire level determination method provided in Embodiment 1 of this application, data from multiple locations within a target area during a second preset time period are acquired. For each location, a feature vector of the fire hazard factor is determined based on the data, wherein the second preset time period is earlier than the current time period. Each feature vector is input into a second preset classification model to obtain the fire hazard classification result for the location. The network model of the second preset classification model includes at least a gating structure, which is used to extract information at each time step. Based on all information, the fire hazard classification result is output. The time step is determined based on the feature vector. The fire hazard classification result includes at least a category and a probability value for the category. A preset database is constructed based on data from all categories, all probability values, and all locations.

[0055] In this embodiment of the invention, the second preset time period is a time window earlier than the current time period, used to collect historical data on fire hazard factors (such as meteorological factors, geographical factors, time factors, and power transmission channel factors) at multiple locations within the target area. Data from multiple locations within the target area during the second preset time period is acquired, and for each location, a feature vector of the fire hazard factor is determined based on the data. This feature vector is then input into the second preset classification model to obtain the fire hazard classification result for that location.

[0056] Optionally, the network model of the second preset classification model includes at least a gating structure. The gating structure is used to extract information at each time step. Based on all the information, a fire risk classification result can be output. The time step is determined based on the feature vector; that is, the time step depends on the data properties in the feature vector and the typical development cycle of wildfire events. For example, every few hours can be selected as a time step, and this time step can be set manually. The fire risk classification result includes at least the category and the probability value of the category. Based on data from all categories, all probability values, and all locations, a preset database can be constructed.

[0057] To accurately obtain the second preset classification model, in the method for determining the fire level provided in Embodiment 1 of this application, historical fire point time series data is acquired. This historical fire point time series data includes at least: historical fire point data for multiple ignition points and a set of associated data corresponding to each historical fire point. The associated data set includes at least: meteorological data, geographic information data, and power transmission facility data. The historical fire point time series data is processed to obtain historical time series processed data. Based on the historical time series processed data, a historical feature vector of the fire hazard factor for each ignition point is constructed. The initial classification model is trained based on the historical feature vector to obtain the second preset classification model.

[0058] In this embodiment of the invention, sensors can be used to acquire data such as the geographical location, time, and intensity of historical fire points. Meteorological data, including temperature, humidity, wind speed, wind direction, and precipitation, is relevant to wildfire occurrences and can be obtained from weather stations. Geographic information data comes from collected topographic, vegetation, and soil data of mountain power transmission corridors. Topographic data includes elevation, slope, and aspect; vegetation data includes vegetation type, vegetation coverage, and vegetation index; and soil data includes soil moisture and soil type. Power transmission corridor facility data includes line data, which records the route, tower locations, and equipment types of the power transmission corridor.

[0059] Historical fire time series data can be processed, including but not limited to deduplication, missing value handling, outlier handling, standardization, and spatiotemporal matching, to obtain historical time series processed data. Based on the historical time series processed data, a historical feature vector of the fire risk factor for each fire point can be constructed. The historical feature vector can include the values ​​of multiple fields. Based on the historical feature vector, the initial classification model can be trained to obtain a second preset classification model.

[0060] Optionally, deduplication can be performed on historical fire point time series data to remove duplicate fire point records caused by sensor malfunctions or false alarms. For missing values ​​in meteorological and geographic information data, interpolation methods are used to fill in the missing values, and outliers in the data are detected and corrected. Box plots can be used to detect outliers, and they can be corrected or removed. Standardization processing can also be performed on meteorological and geographic information data (e.g., Z-score (standard score transformation, used to convert data to a standard normal distribution with a mean of 0 and a standard deviation of 1) standardization, and Min-Max (normalization, scaling data to a specified range) standardization).

[0061] Optionally, historical fire point data can be spatiotemporally matched with meteorological data and geographic information data. For each fire point record, meteorological data within a certain time window (e.g., 24 hours) before and after its collection time, as well as geographic information data within a certain spatial range (e.g., a 5km radius) around its geographical location, can be extracted.

[0062] In order to accurately construct historical feature vectors, the method for determining fire level provided in Embodiment 1 of this application determines multiple fire risk factors, wherein each fire risk factor includes at least: multiple fields; feature extraction is performed on historical time series processing data to obtain feature data; based on all feature data, the field features of each field are determined; and for each ignition point, a historical feature vector is constructed based on all field features.

[0063] Optionally, each fire risk factor includes at least several fields, such as meteorological factors including temperature, humidity, wind speed, wind direction, precipitation, etc.; geographical factors including latitude, longitude, altitude, slope, aspect, vegetation type, vegetation coverage, soil moisture, etc.; time factors including season, month, date, hour, etc.; and power transmission channel factors including distance from power transmission lines, distance from towers, equipment type, etc.

[0064] In this embodiment of the invention, feature extraction is performed on historical time-series processing data to obtain feature data. Based on all feature data, the field features of each field can be determined. For each ignition point, a historical feature vector can be constructed based on all field features.

[0065] Optionally, the temporal and spatial distribution of historical fire points can be analyzed. Temporal distribution analysis can statistically analyze the distribution patterns of historical fire points at different time scales (hourly, daily, monthly, seasonal), and by plotting time-series graphs of fire point occurrence frequency, peak periods for wildfires in summer (June-August) and the dry season (October-November) can be identified. Spatial distribution analysis can identify high-incidence areas of wildfires by plotting geographic heat maps of fire points, and generate fire point density maps by calculating kernel density estimates of fire points, visually displaying the spatial distribution hotspots of fire points. Intensity indicators of fire points, such as radiant energy, area, and duration, can be calculated. Furthermore, the relationship between fire point intensity and distance to power transmission corridor facilities and equipment type can be analyzed. By statistically analyzing the distribution of fire point intensity within different distance ranges (e.g., 1km, 2km, 5km), the potential threat of fire points to power transmission corridors can be assessed.

[0066] The data includes at least: first meteorological sequence data. In order to improve the accuracy of constructing the preset database, in the method for determining the fire level provided in Embodiment 1 of this application, for each category, it is determined whether the category indicates a fire location, and if the category indicates a fire location, the probability value of the category is mapped to obtain the fire risk level corresponding to the first meteorological sequence data; for each fire risk level, it is determined whether the fire risk level is within the preset level threshold range, and if the fire risk level is within the preset level threshold range, the first meteorological sequence data is added to the preset database, thereby constructing the preset database.

[0067] In this embodiment of the invention, the data in the second preset time period includes at least first meteorological sequence data, which includes at least information such as temperature, humidity, wind speed, wind direction, and precipitation. For each category, it is determined whether the category indicates a fire at its location, and if the category indicates a fire at its location, its probability value is mapped to obtain the fire risk level corresponding to the first meteorological sequence data. For example, the fire risk level can be divided into 1-5 levels according to the probability value, such as [0, 0.2) as level 1, [0.2, 0.4) as level 2, [0.4, 0.6) as level 3, [0.6, 0.8) as level 4, and [0.8, 1) as level 5.

[0068] It should be noted that the above examples are for illustrative purposes only and do not limit the present invention in any way.

[0069] For each fire risk level, it can be determined whether the fire risk level is within the preset level threshold range (e.g., level 4-level 5). If so, it can be identified as a historical high fire risk event. If the fire risk level is within the preset level threshold range, its first meteorological sequence data is added to the preset database to construct the preset database. Each first meteorological sequence data can be a multi-dimensional vector with a length of 72 hours, including indicators such as temperature, humidity, and wind speed.

[0070] The time-series data includes at least: second meteorological sequence data. In order to accurately obtain the target meteorological feature vector, in the method for determining the fire level provided in Embodiment 1 of this application, the similarity value between the second meteorological sequence data and each first meteorological sequence data in the preset database is calculated; all similarity values ​​are compared to obtain comparison results; based on the comparison results, multiple target meteorological sequence data are determined; based on the similarity value corresponding to each target meteorological sequence data, a weight is assigned to each target meteorological sequence data to obtain weighted target meteorological sequence data; all weighted target meteorological sequence data are fused to obtain the target meteorological feature vector.

[0071] Optionally, a dynamic time warping algorithm can be used to identify the similarity between current weather conditions and historical high fire risk events, that is, to calculate the similarity value between the second meteorological sequence data and each first meteorological sequence data in the preset database.

[0072] In this embodiment of the invention, all similarity values ​​can be compared to obtain the comparison results. The top 5 second meteorological sequence data with the highest similarity can be selected as the target meteorological sequence data. Based on the similarity value corresponding to each target meteorological sequence data (e.g., ...), Where Q represents the second meteorological sequence data, For the j-th first meteorological sequence data), assign weights to each target meteorological sequence data (e.g., ...). , where j represents the index variable in the preset database, used to traverse the first meteorological sequence data in the preset database (j=1,…,k), which can obtain the weighted target meteorological sequence data, and can fuse all the weighted target meteorological sequence data to obtain the target meteorological feature vector.

[0073] In order to accurately determine the decomposed data, in the method for determining the fire level provided in Embodiment 1 of this application, fire risk data corresponding to multiple fire risk factors are determined from time series data; data decomposition technology is used to decompose all fire risk data to obtain trend items, seasonal items and residual items; based on the trend items, seasonal items and residual items, the decomposed data is determined.

[0074] In this embodiment of the invention, fire risk data corresponding to each fire risk factor can be screened from time series data, and data decomposition technology (such as STL) can be used to decompose all fire risk data to obtain trend items, seasonal items and residual items, and the trend items, seasonal items and residual items can be used as decomposed data.

[0075] Optionally, when using STL for time series decomposition, different parameters need to be set first, such as seasonal period (e.g., 365), trend term smoothing window (e.g., 7×24), seasonal term smoothing window (e.g., 3×365), etc.

[0076] In order to determine the fire handling strategy for the location to be detected, in the method for determining the fire level provided in Embodiment 1 of this application, the fire hazard probability value is mapped to obtain the target fire level of the fire hazard probability value; based on the target fire level, the fire handling strategy for the location to be detected is determined.

[0077] In this embodiment of the invention, fire hazard probability values ​​can be mapped to obtain target fire levels, such as low fire risk [0-0.3], medium fire risk [0.3-0.7], and high fire risk [0.7-1]. If the fire hazard probability value is 0.75, the target fire level is a high-risk fire level. Based on the target fire level, fire handling strategies for the location to be detected can be determined. For example, for locations predicted to be high-risk fire levels, real-time monitoring of the area can be strengthened, fire-fighting equipment can be deployed in advance, and response time can be shortened.

[0078] The fire level determination method provided in this application can construct a wildfire high-incidence period prediction model by integrating multiple data sources such as historical fire point data, meteorological data, geographic information data, and power transmission channel facility data. Based on the fire risk probability value predicted by the model, a level mapping is performed, and refined fire handling strategies are generated for different fire risk levels. This enables real-time and dynamic prediction of wildfire high-incidence periods for mountain power transmission channels, solving the problems of insufficient wildfire prevention and control caused by monitoring delays, limited data, and incomplete models. It can not only detect wildfire hazards in a timely manner but also accurately predict the peak periods of wildfires, providing reliable early warning information for power transmission channel operation and maintenance personnel. This buys valuable time for early wildfire fighting and reduces the threat of wildfires to the safe operation of power transmission channels.

[0079] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0080] Example 2

[0081] This application also provides a fire level determination device. It should be noted that the fire level determination device of this application can be used to execute the fire level determination method provided in this application. The fire level determination device provided in this application is described below.

[0082] According to an embodiment of this application, an apparatus for implementing the above-described method for determining fire level is also provided. Figure 3 This is a schematic diagram of an optional fire level determination device according to an embodiment of this application, such as... Figure 3 As shown, the fire level determination device may include: a determination unit 30, a decomposition unit 31, and an input unit 32.

[0083] The determining unit 30 is used to acquire time series data of the location to be detected in the target area in the current time period, and determine the target meteorological feature vector based on the preset database and the time series data. The preset database includes data of multiple locations in the target area.

[0084] Decomposition unit 31 is used to decompose time-series data to obtain decomposed data;

[0085] Input unit 32 is used to input time series data, target meteorological feature vector and decomposed data into the first preset model to obtain the fire risk probability value of the location to be detected in the first preset time period, wherein the current time period is earlier than the first preset time period. The first preset model includes at least: an input layer, a hidden layer and an output layer. The input layer is used to receive input data, the hidden layer is used to extract features from the input data, and the output layer is used to output the fire risk probability value based on the features through a fully connected layer and an activation function. The fire risk probability value is used to determine the fire level of the location to be detected.

[0086] The fire level determination device provided in this application embodiment can obtain time series data of the location to be detected in the target area in the current time period through the determination unit 30, and determine the target meteorological feature vector based on the preset database and the time series data. The time series data can be decomposed by the decomposition unit 31 to obtain decomposed data. The time series data, the target meteorological feature vector and the decomposed data can be input into the first preset model through the input unit 32 to obtain the fire risk probability value of the location to be detected in the first preset time period.

[0087] Optionally, the fire level determination device further includes: a first determination module, used to acquire data of multiple locations in the target area during a second preset time period before acquiring time-series data of the locations to be detected in the target area during the current time period, and for each location, determine the feature vector of the fire hazard factor based on the data, wherein the second preset time period is earlier than the current time period; a first input module, used to input each feature vector into a second preset classification model to obtain the fire hazard classification result of the location, wherein the network model of the second preset classification model includes at least: a gating structure, the gating structure is used to extract information at each time step, and output the fire hazard classification result based on all information, the time step is determined based on the feature vector, and the fire hazard classification result includes at least: the category and the probability value of the category; and a first construction module, used to construct a preset database based on the data of all categories, all probability values, and all locations.

[0088] Optionally, the fire level determination device further includes: a first acquisition module, used to acquire historical fire point time series data before acquiring data from multiple locations within the target area during a second preset time period, wherein the historical fire point time series data includes at least: historical fire point data from multiple ignition points and a set of associated data corresponding to each historical fire point, the set of associated data including at least: meteorological data, geographic information data, and power transmission channel facility data; a first processing module, used to process the historical fire point time series data to obtain historical time series processed data; a second construction module, used to construct a historical feature vector of the fire hazard factor for each ignition point based on the historical time series processed data; and a first training module, used to train an initial classification model based on the historical feature vector to obtain a second preset classification model.

[0089] Optionally, the second construction module includes: a first determining submodule for determining multiple fire hazard factors, wherein each fire hazard factor includes at least multiple fields; a first extraction submodule for extracting features from historical time-series processed data to obtain feature data; a second determining submodule for determining the field features of each field based on all feature data; and a first construction submodule for constructing a historical feature vector for each ignition point based on all field features.

[0090] Optionally, the data includes at least: first meteorological sequence data, and the first construction module includes: a first mapping submodule, used to determine whether a fire has occurred at a location indicated by the category for each category, and to map the probability value of the category to obtain the fire risk level corresponding to the first meteorological sequence data if a fire has occurred at a location indicated by the category; and a first addition submodule, used to determine whether a fire risk level is within a preset level threshold range for each fire risk level, and to add the first meteorological sequence data to a preset database if the fire risk level is within the preset level threshold range, thereby constructing the preset database.

[0091] Optionally, the time-series data includes at least: second meteorological sequence data, and the determining unit 30 includes: a first calculation module for calculating the similarity value between the second meteorological sequence data and each first meteorological sequence data in a preset database; a first comparison module for comparing all similarity values ​​to obtain a comparison result; a second determining module for determining multiple target meteorological sequence data based on the comparison result; a first allocation module for assigning weights to each target meteorological sequence data based on the similarity value corresponding to each target meteorological sequence data to obtain weighted target meteorological sequence data; and a first fusion module for fusing all weighted target meteorological sequence data to obtain a target meteorological feature vector.

[0092] Optionally, the decomposition unit 31 includes: a third determining module, used to determine the fire risk data corresponding to multiple fire risk factors from the time series data; a first decomposition module, used to decompose all fire risk data using data decomposition technology to obtain trend terms, seasonal terms and residual terms; and a fourth determining module, used to determine the decomposed data based on the trend terms, seasonal terms and residual terms.

[0093] Optionally, the fire level determination device further includes: a first mapping module, used to map the fire probability value to obtain the target fire level of the fire probability value after inputting time series data, target meteorological feature vector and decomposed data into a first preset model to obtain the fire probability value of the location to be detected in a first preset time period; and a fifth determination module, used to determine the fire handling strategy for the location to be detected based on the target fire level.

[0094] The fire level determination device described above may also include a processor and a memory. The determination unit 30, the decomposition unit 31, the input unit 32, etc., are all stored in the memory as program units, and the processor executes the program units stored in the memory to realize the corresponding functions.

[0095] The aforementioned processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured. By adjusting the kernel parameters, time-series data, target meteorological feature vectors, and decomposed data are input into a first preset model to obtain the fire risk probability value of the location to be detected during a first preset time period.

[0096] The aforementioned memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0097] It should be noted that the aforementioned determining unit 30, decomposition unit 31, and input unit 32 correspond to steps S201 to S203 in Embodiment 1. The instances and application scenarios implemented by these units and their corresponding steps are the same, but they are not limited to the content disclosed in Embodiment 1. It should also be noted that these units can be hardware or software components stored in a memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). These units can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0098] Example 3

[0099] Embodiments of this application may provide a computer terminal, which may be any computer terminal device in a group of computer terminals. Optionally, in this embodiment, the aforementioned computer terminal may also be replaced with a mobile terminal or an electronic device, etc.

[0100] Optionally, in this embodiment, the computer terminal may be located in at least one of a plurality of network devices in a computer network.

[0101] In this embodiment, the computer terminal described above can execute the program code for the following steps in the method for determining the fire level: acquiring time-series data of the location to be detected in the target area during the current time period, and determining the target meteorological feature vector based on a preset database and the time-series data, wherein the preset database includes data of multiple locations within the target area; decomposing the time-series data to obtain decomposed data; inputting the time-series data, the target meteorological feature vector, and the decomposed data into a first preset model to obtain the fire risk probability value of the location to be detected during a first preset time period, wherein the current time period is earlier than the first preset time period, and the first preset model includes at least an input layer, a hidden layer, and an output layer, wherein the input layer is used to receive input data, the hidden layer is used to extract features from the input data, and the output layer is used to output the fire risk probability value based on the features through a fully connected layer and an activation function, and the fire risk probability value is used to determine the fire level of the location to be detected.

[0102] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the method for determining the fire level: acquiring data from multiple locations within the target area during a second preset time period, and for each location, determining a feature vector of the fire hazard factor based on the data, wherein the second preset time period is earlier than the current time period; inputting each feature vector into a second preset classification model to obtain the fire hazard classification result for the location, wherein the network model of the second preset classification model includes at least: a gating structure, the gating structure being used to extract information at each time step, outputting the fire hazard classification result based on all information, the time step being determined based on the feature vector, and the fire hazard classification result including at least: the category and the probability value of the category; constructing a preset database based on data from all categories, all probability values, and all locations.

[0103] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the fire level: acquiring historical fire point time-series data, wherein the historical fire point time-series data includes at least: historical fire point data of multiple ignition points and a set of associated data corresponding to each historical fire point data, the set of associated data including at least: meteorological data, geographic information data and power transmission channel facility data; processing the historical fire point time-series data to obtain historical time-series processed data; constructing a historical feature vector of the fire hazard factor for each ignition point based on the historical time-series processed data; and training the initial classification model based on the historical feature vector to obtain a second preset classification model.

[0104] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the fire level: determining multiple fire hazard factors, wherein each fire hazard factor includes at least: multiple fields; extracting features from historical time-series processed data to obtain feature data; determining the field features of each field based on all feature data; and constructing a historical feature vector for each ignition point based on all field features.

[0105] Optionally, the computer terminal described above can execute the program code for the following steps in the method for determining the fire level: for each category, determine whether the category indicates a fire at a location, and if the category indicates a fire at a location, map the probability value of the category to obtain the fire risk level corresponding to the first meteorological sequence data; for each fire risk level, determine whether the fire risk level is within a preset level threshold range, and if the fire risk level is within the preset level threshold range, add the first meteorological sequence data to a preset database to construct the preset database.

[0106] Optionally, the aforementioned computer terminal can execute the program code for the following steps in the method for determining the fire level: calculating the similarity value between the second meteorological sequence data and each first meteorological sequence data in the preset database; comparing all similarity values ​​to obtain a comparison result; determining multiple target meteorological sequence data based on the comparison result; assigning weights to each target meteorological sequence data based on the similarity value corresponding to each target meteorological sequence data to obtain weighted target meteorological sequence data; and fusing all weighted target meteorological sequence data to obtain a target meteorological feature vector.

[0107] Optionally, the aforementioned computer terminal can execute program code for the following steps in the method for determining the fire level: determining fire risk data corresponding to multiple fire risk factors from time-series data; using data decomposition technology to decompose all fire risk data to obtain trend items, seasonal items, and residual items; and determining the decomposed data based on the trend items, seasonal items, and residual items.

[0108] Optionally, Figure 4This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 4 As shown, the electronic device may include: one or more ( Figure 4 (Only one is shown) Processor 402, memory 404, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0109] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the fire level determination method and apparatus in this application embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the aforementioned fire level determination method. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0110] The processor can access the information and application programs stored in the memory via a transmission device to execute the steps described above in the method for determining the fire level.

[0111] The embodiments of this application provide a scheme for determining the fire level. By acquiring the time series data of the location to be detected in the current time period, and combining it with STL decomposition, a preset database and an LSTM network model, wildfire prediction for future time periods can be achieved. The predicted probability can be mapped to a level to obtain the fire level, thereby solving the technical problem in related technologies that it is impossible to accurately determine the fire level in complex mountainous environments.

[0112] Those skilled in the art will understand that Figure 4 The structure shown is for illustrative purposes only. Electronic devices can also be terminal devices such as smartphones, tablets, PDAs, and mobile internet devices (MIDs). Figure 4 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 4 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 4 The different configurations shown.

[0113] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0114] Example 4

[0115] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the fire rating determination method provided in Embodiment 1.

[0116] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0117] This application also provides a computer program product that, when executed on a data processing device, is adapted to perform the steps of a method for determining the fire rating.

[0118] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0119] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0120] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0121] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0122] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0123] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0124] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for determining the fire rating, characterized in that, include: The time-series data of the location to be detected in the target area in the current time period is obtained, and the target meteorological feature vector is determined based on the preset database and the time-series data. The preset database includes data of multiple locations in the target area. The time-series data is decomposed to obtain decomposed data; The time-series data, the target meteorological feature vector, and the decomposed data are input into a first preset model to obtain the fire risk probability value of the location to be detected in a first preset time period, wherein the current time period is earlier than the first preset time period. The first preset model includes at least an input layer, a hidden layer, and an output layer. The input layer is used to receive input data, the hidden layer is used to extract features from the input data, and the output layer is used to output the fire risk probability value based on the features through a fully connected layer and an activation function. The fire risk probability value is used to determine the fire level of the location to be detected.

2. The method for determining the fire level according to claim 1, characterized in that, Before acquiring the time-series data of the location to be detected in the target area within the current time period, the following steps are also included: Data of multiple locations within the target area during a second preset time period is obtained, and for each location, a feature vector of the fire risk factor is determined based on the data, wherein the second preset time period is earlier than the current time period; Each of the feature vectors is input into a second preset classification model to obtain the fire risk classification result of the location. The network model of the second preset classification model includes at least a gating structure, which is used to extract information at each time step and output the fire risk classification result based on all the information. The time step is determined based on the feature vector. The fire risk classification result includes at least a category and a probability value of the category. The preset database is constructed based on the data of all the categories, all the probability values, and all the locations.

3. The method for determining the fire level according to claim 2, characterized in that, Before acquiring data from multiple locations within the target area during a second preset time period, the method further includes: Acquire historical fire point time series data, wherein the historical fire point time series data includes at least: historical fire point data of multiple fire points and a set of associated data corresponding to each historical fire point data, wherein the set of associated data includes at least: meteorological data, geographic information data and power transmission channel facility data; The historical fire point time series data is processed to obtain historical time series processed data; Based on the historical time-series processing data, a historical feature vector of the fire risk factor for each of the fire ignition points is constructed. The initial classification model is trained based on the historical feature vectors to obtain the second preset classification model.

4. The method for determining the fire level according to claim 3, characterized in that, The step of constructing a historical feature vector of the fire hazard factor for each ignition point based on the historical time-series processing data includes: Multiple fire hazard factors are identified, wherein each fire hazard factor includes at least: multiple fields; Feature data is obtained by extracting features from the historical time-series processed data; Based on all the aforementioned feature data, determine the field features of each of the fields; For each of the aforementioned ignition points, a historical feature vector is constructed based on all the aforementioned field features.

5. The method for determining the fire level according to claim 2, characterized in that, The data includes at least: first meteorological sequence data. The step of constructing the preset database based on the data from all said categories, all said probability values, and all said locations further includes: For each category, determine whether the category indicates that a fire has occurred at the location, and if the category indicates that a fire has occurred at the location, map the probability value of the category to obtain the fire risk level corresponding to the first meteorological sequence data; For each fire risk level, it is determined whether the fire risk level is within a preset level threshold range. If the fire risk level is within the preset level threshold range, the first meteorological sequence data is added to the preset database to construct the preset database.

6. The method for determining the fire level according to claim 1, characterized in that, The time-series data includes at least: second meteorological sequence data. The step of determining a target meteorological feature vector based on a preset database and the time-series data includes: Calculate the similarity value between the second meteorological sequence data and each first meteorological sequence data in the preset database; All the aforementioned similarity values ​​are compared to obtain the comparison results; Based on the comparison results, multiple target meteorological sequence data were determined; Based on the similarity value corresponding to each target meteorological sequence data, a weight is assigned to each target meteorological sequence data to obtain weighted target meteorological sequence data; The target meteorological feature vector is obtained by fusing all the weighted target meteorological sequence data.

7. The method for determining the fire level according to claim 1, characterized in that, The step of decomposing the time-series data to obtain decomposed data includes: Fire risk data corresponding to multiple fire risk factors are determined from the time series data; Data decomposition technology is used to decompose all the fire risk data to obtain trend items, seasonal items, and residual items; The decomposed data is determined based on the trend term, the seasonal term, and the residual term.

8. The method for determining the fire level according to claim 1, characterized in that, After inputting the time-series data, the target meteorological feature vector, and the decomposed data into the first preset model to obtain the fire risk probability value of the location to be detected in the first preset time period, the method further includes: The fire probability value is mapped to obtain the target fire level of the fire probability value; Based on the target fire level, a fire handling strategy is determined for the location to be detected.

9. A device for determining the fire level, characterized in that, include: The determining unit is used to acquire time-series data of the location to be detected in the target area in the current time period, and determine the target meteorological feature vector based on a preset database and the time-series data, wherein the preset database includes data of multiple locations in the target area; The decomposition unit is used to decompose the time-series data to obtain decomposed data; An input unit is used to input the time-series data, the target meteorological feature vector, and the decomposed data into a first preset model to obtain the fire risk probability value of the location to be detected in a first preset time period, wherein the current time period is earlier than the first preset time period. The first preset model includes at least an input layer, a hidden layer, and an output layer. The input layer is used to receive input data, the hidden layer is used to extract features from the input data, and the output layer is used to output the fire risk probability value based on the features through a fully connected layer and an activation function. The fire risk probability value is used to determine the fire level of the location to be detected.

10. A computer program product, characterized in that, The method includes a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for determining the fire rating as described in any one of claims 1 to 8.

11. An electronic device, characterized in that, It includes one or more processors and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method for determining the fire level as described in any one of claims 1 to 8.

Images