A power transmission line mountain fire secondary geological disaster early warning method based on time attenuation effect
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ELECTRIC POWER RES INST OF GUANGXI POWER GRID CO LTD
- Filing Date
- 2026-04-09
- Publication Date
- 2026-08-07
AI Technical Summary
[0003]本发明提供了一种基于时间衰减效应的输电线路山火次生地质灾害预警方法，用于解决预警准确性和时效性不足的问题
Smart Images
Figure CN122531165A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of geological disaster early warning technology, and in particular to an early warning method for secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect. Background Technology
[0002] Transmission line corridors typically traverse areas with complex terrain, and geological disasters such as landslides and debris flows are among the main sources of risk threatening the safe operation of transmission lines. Existing geological disaster early warning methods for transmission lines employ statistical analysis models, which weight real-time rainfall data and static susceptibility indicators to obtain the probability or risk index of geological disasters occurring around each transmission line tower. When this probability or index exceeds a preset threshold, an early warning message is issued to maintenance personnel. This method ignores the long-term impact of historical wildfire events on the stability of geological bodies. When wildfires occur, they burn surface vegetation, destroying the soil's consolidation effect through the root system. At the same time, high temperatures cause soil organic matter to burn, forming a hydrophobic layer and reducing soil cohesion. These effects can persist in the stability of geological bodies for months or even years. However, if only the direct triggering factor of real-time rainfall is considered, and the stability of geological bodies is regarded as a static or linearly recovering parameter over time, it is impossible to quantify the residual impact of historical wildfires on current geological stability and its dynamic decay over time. When the rainfall intensity does not reach the conventional warning threshold, even if historical wildfires have severely damaged the stability of geological bodies, the existing model still judges it as a safe state, resulting in missed warnings of secondary geological disasters caused by wildfires, leading to problems with insufficient accuracy and timeliness of early warnings. Summary of the Invention
[0003] This invention provides an early warning method for secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect, which addresses the problems of insufficient accuracy and timeliness in early warning.
[0004] This invention provides a method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect, comprising: Acquire historical wildfire ignition parameters, vegetation type data, and terrain slope data within a preset geographical area of the power transmission line corridor; Based on the vegetation type data and terrain slope data, the time decay coefficient of each grid cell is determined; the time decay coefficient is used to characterize the rate at which the impact of wildfires on the stability of geological bodies decays over time. Based on the historical wildfire ignition parameters and the time decay coefficient, the geological stability impact base value of each grid cell at the current moment is calculated; the geological stability impact base value is used to characterize the degree of residual impact of historical wildfires on the stability of the current geological body. The rainfall forecast data of the transmission line corridor within a preset time period is obtained. The geological stability impact base value and the rainfall forecast data are input into a preset geological disaster probability prediction model, and the geological disaster occurrence probability of each grid cell is output. If the probability of a geological disaster exceeds a preset probability threshold, then the transmission line towers located in the corresponding grid unit are identified, and early warning information is generated for the transmission line towers.
[0005] Furthermore, determining the time decay coefficient of each grid cell based on the vegetation type data and terrain slope data includes: Obtain a preset vegetation restoration cycle mapping table, and retrieve the baseline restoration cycle of each grid unit from the vegetation restoration cycle mapping table. The vegetation restoration cycle mapping table is used to record the baseline restoration cycle of different vegetation types under standard slope conditions. Obtain a preset slope correction coefficient mapping table, and retrieve the slope correction coefficient of each grid unit from the slope correction coefficient mapping table; the slope correction coefficient mapping table is used to record the correction coefficients corresponding to different slope ranges; Multiply the baseline recovery period by the slope correction coefficient to obtain the comprehensive recovery period of each grid cell; The time decay coefficient of each grid cell is calculated based on the comprehensive recovery cycle.
[0006] Furthermore, the calculation of the time decay coefficient of each grid cell based on the comprehensive recovery period includes: A segmented calculation model is used to determine the time decay coefficient; wherein the segmented calculation model includes a first decay coefficient calculation model, a second decay coefficient calculation model, and a third decay coefficient calculation model, the first decay coefficient calculation model, the second decay coefficient calculation model, and the third decay coefficient calculation model correspond to different decay rates, and the decay rate of the first decay coefficient calculation model is greater than the decay rate of the second decay coefficient calculation model, and the decay rate of the second decay coefficient calculation model is greater than the decay rate of the third decay coefficient calculation model.
[0007] Furthermore, the determination of the time decay coefficient using a piecewise calculation model includes: Obtain the time difference between the current warning time and the occurrence time of each historical wildfire hotspot; If the time difference is less than or equal to the first preset ratio of the comprehensive recovery period, then the time attenuation coefficient is calculated using the first attenuation coefficient calculation model. If the time difference is greater than the first preset proportion of the comprehensive recovery period and less than or equal to the second preset proportion of the comprehensive recovery period, then the time attenuation coefficient is calculated using the second attenuation coefficient calculation model. If the time difference is greater than the second preset ratio of the comprehensive recovery period, then the time decay coefficient is calculated using the third decay coefficient calculation model.
[0008] Furthermore, the historical wildfire ignition parameters include the time of occurrence of each wildfire, the location of the ignition point, and intensity parameters used to characterize the intensity of the wildfire.
[0009] Furthermore, the calculation of the geological stability impact base value of each grid cell at the current moment based on the historical wildfire ignition parameters and the time decay coefficient includes: Based on the fire location, each historical wildfire fire point is mapped to the corresponding grid cell; Obtain the preset geological vulnerability factor for each grid cell, and calculate the geological vulnerability index for each grid cell based on the geological vulnerability factor; The geological stability impact base value of each grid cell at the current moment is determined based on the wildfire occurrence time, intensity parameters, and geological vulnerability index.
[0010] Furthermore, the determination of the geological stability impact base value of each grid cell at the current moment based on the wildfire occurrence time, intensity parameters, and the geological vulnerability index includes: The initial impact intensity value of each historical wildfire point is calculated based on the aforementioned intensity parameters; The attenuation weight of the fire point is determined based on the fire occurrence time and the time attenuation coefficient. Multiplying the initial influence intensity value by the attenuation weight yields the attenuated influence intensity of the fire point. Multiplying the attenuated influence intensity by the geological vulnerability index of the grid cell where the fire point is located yields the influence contribution value of the fire point to the grid cell. The impact contribution values of all historical wildfire hotspots within the same grid cell are summed to obtain the geological stability impact base value of the grid cell at the current moment.
[0011] Furthermore, the preset geological disaster probability prediction model is a fusion prediction model constructed based on machine learning algorithms; the input of the fusion prediction model includes the geological stability impact base value and the rainfall forecast data, and the output is the probability of geological disaster occurrence.
[0012] Furthermore, the step of inputting the geological stability impact baseline value and the rainfall forecast data into a preset geological hazard probability prediction model, and outputting the geological hazard occurrence probability for each grid cell, includes: The geological stability impact base value of each grid cell is used as the first input feature, and the rainfall forecast data is used as the second input feature, and both are input into the fusion prediction model. The fusion prediction model performs nonlinear mapping on the first and second input features to output the probability of geological disaster occurrence for each grid cell.
[0013] Furthermore, the early warning information includes at least one of pole identification information, risk level information, and recommended measures information.
[0014] As can be seen from the above technical solutions, the present invention has the following advantages: This invention acquires historical wildfire ignition parameters, vegetation type data, and terrain slope data within a preset geographical area of a power transmission line corridor. Based on the vegetation type data and terrain slope data, it determines the time decay coefficient of each grid unit to characterize the rate at which the impact of wildfires on geological stability decays over time. Based on the historical wildfire ignition parameters and time decay coefficients, it calculates the geological stability impact baseline value of each grid unit at the current moment, used to characterize the residual impact of historical wildfires on the current geological stability. It then acquires rainfall forecast data for the power transmission line corridor within a preset future time period, inputting the geological stability impact baseline value and rainfall forecast data into a preset geological disaster probability prediction model, outputting the probability of geological disaster occurrence for each grid unit. Finally, if the probability of geological disaster occurrence exceeds a preset probability threshold, it identifies the power transmission line towers located within the corresponding grid unit and generates early warning information for these towers. This invention dynamically quantifies the impact of historical wildfires through a time decay coefficient. Simultaneously, by fusing the geological stability baseline value after the decay of historical wildfires with future rainfall forecast data, it effectively improves the accuracy and timeliness of early warnings, providing reliable technical data support for the precise prevention and control of secondary geological disasters caused by wildfires along power transmission lines. Attached Figure Description
[0015] Figure 1 This is a schematic flowchart of an embodiment of a method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect in this invention. Figure 2 This is a schematic diagram of the process for determining the time decay coefficient in this invention; Figure 3 This is a schematic diagram of the calculation process for the basement value affected by geological stability in this invention. Detailed Implementation
[0016] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application 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 this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “corresponding to,” and any variations thereof, are intended to cover 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.
[0017] Example 1 The implementation method in this embodiment can be implemented in a system, on a server, or on a terminal; no specific limitation is made. The method in this application will be described below from the perspective of system implementation. Please refer to... Figure 1 The method provided in this application includes the following steps: S1. Obtain historical wildfire ignition point parameters, vegetation type data, and terrain slope data within a preset geographical range of the power transmission line corridor; In this embodiment, the preset geographical range is the area covered by a preset width extending to both sides of the transmission line's centerline. This preset width is determined based on the transmission line's voltage level and terrain complexity. For transmission lines with voltage levels of 500 kV and above, the preset width is set to 5 kilometers; for transmission lines with voltage levels of 220 kV and below, the preset width is set to 2 kilometers; and for areas with complex terrain or frequent wildfires, the preset width is increased to 10 kilometers. Historical wildfire ignition parameters are obtained through multi-source satellite remote sensing data, including multispectral satellite imagery and thermal infrared remote sensing data covering the transmission line corridor area within the past 5 to 10 years. Satellite data sources include medium-resolution imaging spectrometers, visible-infrared imaging radiometers, the Landsat series, and domestically produced environmental disaster reduction satellites. The acquired satellite imagery data is preprocessed to extract ignition parameters for each historical wildfire ignition point. These parameters include the time of occurrence of the wildfire, the location of the ignition point, and intensity parameters used to characterize the wildfire intensity. The timing of wildfire occurrence was determined by analyzing abnormal changes in the brightness temperature value in the thermal infrared band. When the brightness temperature value exceeded the preset flame detection threshold, it was identified as a fire point, and the time was recorded as the wildfire occurrence time. The flame detection threshold was set according to the satellite sensor type and seasonal variations. For medium resolution imaging spectrometer data, the flame detection threshold was set to 320 Kelvin to 330 Kelvin. The location of the fire point was determined by geometrically corrected geographic coordinates. Intensity parameters included radiative power and burned area. Radiative power was obtained by inverting the radiative transfer equations in the mid-infrared and thermal infrared bands, reflecting the burning intensity of the fire point. The burned area was extracted by analyzing the difference in the normalized vegetation index (NVI) over multiple time periods. Areas with a significant decrease in the NVI were identified as burned areas, and their areas were calculated. The threshold for the decrease in the NVI was 0.1 to 0.3.
[0018] Vegetation type data was acquired through high-resolution remote sensing image classification. A supervised classification method was used to divide the land cover within the transmission line corridor into different vegetation types, including forests, shrublands, grasslands, farmland, and bare land, creating a 1:10,000 scale vegetation type distribution map. Topographic slope data was extracted using a digital elevation model (DEM). A slope analysis algorithm was used to calculate the topographic slope value for each grid cell, ranging from 0 to 90 degrees. All historical wildfire ignition parameters, vegetation type data, and topographic slope data were spatially registered using uniform grid cells, with grid cell sizes set to 30 meters by 30 meters or 90 meters by 90 meters, ensuring spatial consistency across different data sources.
[0019] S2. Based on vegetation type data and terrain slope data, determine the time decay coefficient of each grid cell; the time decay coefficient is used to characterize the rate at which the impact of wildfires on the stability of geological bodies decays over time. Different vegetation types exhibit varying root-based soil-fixing capacity and surface cover recovery rates after wildfire damage. Furthermore, terrain slope influences soil erosion rates and vegetation recovery conditions. Therefore, it is necessary to comprehensively consider both vegetation type and terrain slope to determine the decay pattern of wildfire impact over time. This embodiment establishes a mapping relationship between vegetation recovery cycles and slope corrections, calculates the comprehensive recovery cycle of each grid unit, and then uses a piecewise model based on the comprehensive recovery cycle to calculate the time decay coefficient, thereby achieving differentiated quantification of the decay rate of wildfire impact under different geographical environments. Please refer to [link / reference]. Figure 2 The details are as follows: 101. Obtain the preset vegetation restoration cycle mapping table, and retrieve the baseline restoration cycle of each grid unit from the vegetation restoration cycle mapping table. The vegetation restoration cycle mapping table is used to record the baseline restoration cycle of different vegetation types under standard slope conditions. 102. Obtain the preset slope correction coefficient mapping table, and retrieve the slope correction coefficient of each grid cell from the slope correction coefficient mapping table; the slope correction coefficient mapping table is used to record the correction coefficients corresponding to different slope ranges; 103. Multiply the baseline recovery period by the slope correction factor to obtain the comprehensive recovery period of each grid cell; The pre-defined vegetation restoration cycle mapping table is constructed based on research findings on the restoration rates of different vegetation types in ecology. Specifically, for arbor forests, due to their deep root systems and high degree of lignification, the baseline restoration cycle is set at 5 to 7 years under standard slope conditions; for shrub forests with medium-depth root systems, the baseline restoration cycle is set at 3 to 5 years; for grasslands with shallow root systems but rapid growth, the baseline restoration cycle is set at 1 to 2 years; for farmland affected by human cultivation, the baseline restoration cycle is set at 0.5 to 1 year; and for bare land lacking a seed bank, the baseline restoration cycle is set at 7 to 10 years. The pre-defined slope correction coefficient mapping table is constructed based on the influence of terrain slope on vegetation restoration. A slope correction coefficient greater than 1 indicates a slower restoration rate, while a coefficient less than 1 indicates a faster restoration rate. Specifically, for gentle slopes less than 15 degrees, the correction factor is set to 0.8 to 0.9, indicating a relatively fast vegetation recovery rate; for moderate slopes between 15 and 30 degrees, the correction factor is set to 1.0, indicating a standard recovery rate; and for steep slopes greater than 30 degrees, the correction factor is set to 1.2 to 1.5, indicating a significantly slower recovery rate.
[0020] The aforementioned historical wildfire hotspot parameters, vegetation type data, and terrain slope data were all spatially registered using a unified grid unit. The grid unit is a basic calculation unit obtained by dividing the preset geographical area of the power transmission line corridor into 30m x 30m or 90m x 90m sections. The total number of grid units is determined based on the length and preset width of the power transmission line corridor. For example, a 100km-long power transmission line, divided into 5km-wide and 30m grid units, would have approximately 55,500 grid units. Each grid unit has a unique spatial index used to associate the vegetation type data, terrain slope data, and various parameters calculated subsequently. When calculating the comprehensive recovery period for each grid unit, firstly, based on the vegetation type of the grid unit, the corresponding baseline recovery period is retrieved from the vegetation recovery period mapping table. Then, based on the terrain slope value of the grid unit, the corresponding slope correction coefficient is retrieved from the slope correction coefficient mapping table. The baseline recovery period is multiplied by the slope correction coefficient to obtain the comprehensive recovery period for the grid unit. For grid units containing multiple vegetation types or slope variations, an area-weighted average method is used to determine the comprehensive value.
[0021] 104. Calculate the time decay coefficient of each grid cell based on the comprehensive recovery cycle.
[0022] In this embodiment, a segmented calculation model is used to determine the time decay coefficient. The segmented calculation model includes a first decay coefficient calculation model, a second decay coefficient calculation model, and a third decay coefficient calculation model. The first decay coefficient calculation model, the second decay coefficient calculation model, and the third decay coefficient calculation model correspond to different decay rates. The decay rate of the first decay coefficient calculation model is greater than that of the second decay coefficient calculation model, and the decay rate of the second decay coefficient calculation model is greater than that of the third decay coefficient calculation model.
[0023] 1. Obtain the time difference between the current warning time and the occurrence time of each historical wildfire hotspot; 2. If the time difference is less than or equal to the first preset ratio of the comprehensive recovery period, the time attenuation coefficient is calculated using the first attenuation coefficient calculation model; 3. If the time difference is greater than the first preset ratio of the comprehensive recovery period and less than or equal to the second preset ratio of the comprehensive recovery period, the time attenuation coefficient is calculated using the second attenuation coefficient calculation model. 4. If the time difference is greater than the second preset ratio of the comprehensive recovery cycle, the time decay coefficient is calculated using the third decay coefficient calculation model.
[0024] The first attenuation coefficient calculation model simulates the attenuation pattern during the initial rapid vegetation recovery phase after a wildfire. This model uses an exponential function with a high attenuation rate, ranging from 0.5 to 1.0, corresponding to the rapid recovery phase from 0 to 6 months after the wildfire. During this period, herbaceous plants grow rapidly, ground cover increases quickly, and the impact of the wildfire on the stability of the geological body decreases rapidly. The second attenuation coefficient calculation model simulates the attenuation pattern during the middle stage of slow vegetation recovery after a wildfire. This model uses an exponential function with a medium attenuation rate, ranging from 0.1 to 0.5, corresponding to the slow recovery phase from 6 to 24 months after the wildfire. During this period, shrubs and trees begin to grow, but their root systems have not yet fully recovered their soil-fixing capacity, and the attenuation rate of the wildfire's impact slows down. The third attenuation coefficient calculation model simulates the attenuation pattern during the later stage of basic vegetation recovery after a wildfire. This model uses an exponential function with a low attenuation rate, ranging from 0.01 to 0.1, corresponding to the long-term recovery phase after 24 months or more after the wildfire. At this stage, the vegetation structure has basically recovered, and the residual impact of the wildfire slowly approaches zero.
[0025] When determining the time decay coefficient using a segmented calculation model, the obtained time difference is calculated by subtracting the wildfire occurrence time from the current warning time, with the unit being months. Then, based on the ratio of this time difference to the comprehensive recovery period, the appropriate decay coefficient calculation model is determined. The first preset ratio is set to 0.3. When the time difference is less than or equal to 0.3 times the comprehensive recovery period, the fire is considered to be in a rapid recovery phase, and the first decay coefficient calculation model is used to calculate the time decay coefficient. The second preset ratio is set to 0.7. When the time difference is greater than 0.3 times but less than or equal to 0.7 times the comprehensive recovery period, the fire is considered to be in a slow recovery phase, and the second decay coefficient calculation model is used to calculate the time decay coefficient. When the time difference is greater than 0.7 times the comprehensive recovery period, the fire is considered to be in a long-term recovery phase, and the third decay coefficient calculation model is used to calculate the time decay coefficient. The values of the first and second preset ratios are obtained through statistical analysis of a large amount of vegetation recovery field monitoring data, effectively defining the time boundaries of different recovery stages. Using the above segmented calculation model, each historical wildfire point obtains a time decay coefficient that matches the actual recovery state of the fire point, based on the interval between its occurrence time and the current time, as well as the vegetation type and terrain slope conditions of the grid unit it belongs to.
[0026] S3. Based on historical wildfire ignition parameters and time decay coefficients, calculate the geological stability impact base value of each grid cell at the current moment; the geological stability impact base value is used to characterize the degree of residual impact of historical wildfires on the stability of the current geological body; The impact of historical wildfires on geological stability depends not only on the intensity of the fire itself, but also on the geological conditions at the fire site and the time elapsed since the fire. Wildfires of the same intensity occurring in areas with different geological conditions can cause significant differences in the degree of damage to geological stability. Furthermore, the impact of the same wildfire event on geological stability gradually weakens over time due to vegetation recovery and improved soil structure. This embodiment calculates the geological stability impact baseline value for each grid cell by integrating the spatial location, intensity, and time of each historical wildfire fire with the geological vulnerability conditions of its respective grid cell. This baseline value reflects the comprehensive residual impact of historical wildfires on the current geological stability, providing a foundational input for subsequent geological hazard probability prediction by integrating rainfall data. Please refer to [link / reference]. Figure 3 The details are as follows: 101. Based on the location of the fire points, map each historical wildfire fire point to the corresponding grid cell; 102. Obtain the preset geological vulnerability factor for each grid cell, and calculate the geological vulnerability index for each grid cell based on the geological vulnerability factor; Mapping each historical wildfire ignition point to its corresponding grid cell means determining the grid cell to which the ignition point belongs within a pre-defined grid cell system based on its geographic coordinates. Each grid cell has a unique spatial index and may contain zero, one, or more historical wildfire ignition points. For ignition points located near the grid cell boundary, the nearest neighbor rule is used to determine their affiliation.
[0027] The preset geological vulnerability factors include soil type, lithology, and slope factors. These factors are indicators that measure the inherent sensitivity of a geological body to geological hazards such as landslides and debris flows under external disturbances. Soil type factors are determined based on the physical and mechanical properties of the soil. Clay soils, due to their low shear strength, have factor values set between 0.8 and 1.0; sandy soils, due to their good permeability and low cohesion, have factor values set between 0.6 and 0.8; and gravelly soils, due to their larger particles and good drainage, have factor values set between 0.4 and 0.6. Lithology factors are determined based on the degree of weathering and joint development of the rocks. Weak rock formations such as mudstone and shale have factor values set between 0.8 and 1.0; medium-hard rock formations such as sandstone and limestone have factor values set between 0.5 and 0.8; and hard rock formations such as granite and basalt have factor values set between 0.2 and 0.5. The slope factor is determined based on the terrain slope value. For slopes less than 15 degrees, the factor value is set to 0.2 to 0.4; for slopes between 15 and 30 degrees, the factor value is set to 0.4 to 0.7; and for slopes greater than 30 degrees, the factor value is set to 0.7 to 1.0. The geological vulnerability index is obtained by weighted summation or product of the above factors. The weighting coefficients are determined according to the contribution of each factor to the occurrence of geological hazards. The soil type factor weight is set to 0.3 to 0.4, the lithology factor weight is set to 0.3 to 0.4, and the slope factor weight is set to 0.2 to 0.4. For each grid cell, a unique geological vulnerability index is calculated based on the values of each factor within that cell. The value ranges from 0 to 1; a higher value indicates a more vulnerable geological body in that grid cell, and a higher probability of geological hazards occurring under the same external disturbances.
[0028] 103. Determine the base value of the geological stability impact of each grid cell at the current moment based on the time of wildfire occurrence, intensity parameters, and geological vulnerability index.
[0029] 1. Calculate the initial impact intensity value of each historical wildfire hotspot based on intensity parameters; 2. Determine the attenuation weight of the fire point based on the fire occurrence time and time decay coefficient; 3. Multiply the initial impact intensity value by the attenuation weight to obtain the attenuated impact intensity of the fire point, and multiply the attenuated impact intensity by the geological vulnerability index of the grid cell where the fire point is located to obtain the impact contribution value of the fire point to the grid cell; 4. The impact contribution values of all historical wildfire hotspots within the same grid cell are summed to obtain the geological stability impact base value of the grid cell at the current moment.
[0030] The initial impact intensity value is calculated based on the intensity parameters of historical wildfire ignition points, including radiated power and burned area. Radiated power reflects the energy release intensity during combustion, ranging from several megawatts to several gigawatts; burned area reflects the spatial impact range of the fire, ranging from less than 1 hectare to hundreds of hectares. The initial impact intensity value is obtained by multiplying the radiated power by the burned area. This product comprehensively reflects the total energy release and impact scale of the wildfire event; a larger value indicates a stronger potential destructive capacity of the wildfire to the stability of the geological body. Each historical wildfire ignition point corresponds to an initial impact intensity value. For multiple ignition points within the same grid cell, each ignition point's initial impact intensity value is calculated independently.
[0031] The attenuation weight is determined based on the wildfire occurrence time and the time decay coefficient. It characterizes the remaining proportion of the fire's influence intensity after decay over time, from the moment of its occurrence to the current warning time. The attenuation weight ranges from 0 to 1. When a wildfire first occurs, the attenuation weight is close to 1; it gradually decreases over time. In this embodiment, the attenuation weight is calculated using an exponential decay function. The specific formula is: the attenuation weight equals a power function with a base of the natural constant and an exponent of a negative time decay coefficient multiplied by the time difference, where the time difference is in months. Through this calculation, each historical wildfire point obtains its corresponding attenuation weight based on the interval between its occurrence time and the current time.
[0032] The impact contribution value refers to the degree of influence of a single historical wildfire ignition point on the geological stability of its corresponding grid cell at the current moment. The initial impact intensity value is multiplied by the attenuation weight to obtain the residual impact intensity of the ignition point after time decay. Then, the residual impact intensity is multiplied by the geological vulnerability index of the grid cell containing the ignition point to obtain the impact contribution value of the ignition point to the grid cell. This calculation process reflects that the same wildfire event produces a larger impact contribution value in grid cells with higher geological vulnerability. The geological stability impact baseline value is obtained by summing the impact contribution values of all historical wildfire ignition points within the same grid cell. During the summation calculation, the impact contribution values of each ignition point are added together to obtain the comprehensive residual impact degree of the grid cell. For grid cells without historical wildfire ignition points, the geological stability impact baseline value is set to 0.
[0033] Through the above steps, each grid cell obtains a geological stability impact base value at the current warning time. This value comprehensively considers the intensity, occurrence time, attenuation pattern of all historical wildfire events within the grid cell, as well as the geological vulnerability conditions of the grid cell itself.
[0034] S4. Obtain rainfall forecast data for the transmission line corridor within a preset time period in the future, input the geological stability impact base value and rainfall forecast data into the preset geological disaster probability prediction model, and output the geological disaster occurrence probability of each grid cell; In this embodiment, the preset geological disaster probability prediction model is a fusion prediction model built based on machine learning algorithms. The input of the fusion prediction model includes the baseline value of geological stability impact and rainfall forecast data, and the output is the probability of geological disaster occurrence. The specific implementation is as follows: 1. The geological stability impact base value of each grid cell is used as the first input feature, and the rainfall forecast data is used as the second input feature, and both are input into the fusion prediction model; 2. By performing nonlinear mapping processing on the first and second input features through a fusion prediction model, the probability of geological disaster occurrence for each grid cell is output.
[0035] Specifically, the first input feature is the geological stability impact baseline value, calculated in step S3, ranging from 0 to positive infinity. A larger value indicates a higher degree of residual impact of historical wildfires on the stability of the current geological body. For grid cells without historical wildfire sites, this value is set to 0. The second input feature is rainfall forecast data, obtained through a high-resolution numerical weather prediction model. The specific duration of the preset future time period is determined based on early warning requirements. The rainfall forecast data includes indicators such as cumulative rainfall, maximum hourly rainfall intensity, and rainfall duration for each grid cell within the preset future time period. These rainfall indicators are normalized and then used as multiple components of the second input feature.
[0036] The fusion prediction model is a machine learning model pre-trained based on historical geological disaster event data. Training data includes the geological stability impact baseline values for each grid cell in historical periods, rainfall forecast data for the corresponding time periods, and labels indicating whether a geological disaster has occurred. The geological stability impact baseline values are calculated from historical wildfire data according to steps S1 to S3. Rainfall forecast data uses reanalysis rainfall data from historical periods, and geological disaster labels are obtained through spatial matching of historical geological disaster point data. During model training, the geological stability impact baseline values and rainfall forecast data are used as input features, and the geological disaster labels are used as supervision signals. Deep learning algorithms or gradient boosting tree algorithms are employed for training, enabling the model to learn a non-linear mapping relationship from input features to the probability of geological disaster occurrence.
[0037] In this embodiment, the fusion prediction model adopts a deep neural network structure, including an input layer, multiple hidden layers, and an output layer. The input layer receives a feature vector formed by combining the first and second input features, and the dimension of the feature vector is determined according to the number of input features. The hidden layers adopt a fully connected structure, with each hidden layer containing 64 to 256 neurons. The activation function is a linear rectified function, used to extract high-order interaction features and nonlinear relationships between input features. The output layer uses a single neuron, and the activation function is a logistic function, mapping the output value to the interval between 0 and 1, representing the probability of geological disaster occurrence. The probability of geological disaster occurrence ranges from 0 to 1. The closer the value is to 1, the higher the probability of geological disaster occurring in that grid cell under the current geological stability baseline and future rainfall conditions; the closer the value is to 0, the lower the probability of geological disaster occurrence. In the model application stage, for each grid cell, the geological stability influence baseline value calculated in step S3 is used as the first input feature, and the obtained rainfall forecast data is used as the second input feature. These are combined to form an input feature vector, which is then input into the trained fusion prediction model. The model performs forward propagation calculations, sequentially passing through nonlinear transformations in each hidden layer. The output layer outputs a probability value between 0 and 1, representing the probability of geological disaster occurrence for that grid cell. This process is repeated for all grid cells within the transmission line corridor to obtain a distribution map of the probability of geological disaster occurrence for each grid cell.
[0038] S5. If the probability of a geological disaster exceeds a preset probability threshold, identify the transmission line towers located in the corresponding grid unit and generate early warning information for the transmission line towers.
[0039] The early warning information here includes at least one of the following: tower identification information, risk level information, and recommended measures information. The preset probability thresholds are determined in advance based on the statistical analysis results of historical geological disaster events, including a first probability threshold and a second probability threshold. The first probability threshold is set to 0.3 to 0.5 to classify low-risk levels, and the second probability threshold is set to 0.6 to 0.8 to classify high-risk levels. When the probability of a geological disaster occurring in a grid unit exceeds the first probability threshold but does not exceed the second probability threshold, the grid unit is determined to be in a low-risk state; when the probability of a geological disaster occurring exceeds the second probability threshold, the grid unit is determined to be in a high-risk state. The specific values of the probability thresholds are adjusted according to the historical frequency of geological disasters in the area where the transmission line is located, the voltage level of the transmission line, and the safety operation requirements. For areas with frequent geological disasters or high-voltage transmission lines, the probability thresholds can be appropriately reduced to improve early warning sensitivity.
[0040] When identifying transmission line towers located within corresponding grid cells, the geographical coordinates of all towers along the entire transmission line are obtained, including the longitude, latitude, and tower number for each tower. Spatial overlay analysis is performed between the tower geographical coordinates and the spatial extent of the grid cells. For each grid cell, it is determined which towers are contained within its spatial extent. When the geographical coordinates of a tower fall within the boundary of a grid cell, that tower is identified as a transmission line tower within that grid cell. For towers located near the grid cell boundary, a buffer analysis method is used to determine the distance between the tower and adjacent grid cells, assigning it to the nearest grid cell. In cases where a single grid cell may contain multiple towers, all are identified as high-risk or low-risk towers, and prioritized according to their voltage level and importance.
[0041] When generating early warning information, differentiated warning content is adopted for different risk levels. For grid units in a low-risk state, a first-level early warning is generated, which includes pole identification information, risk level information, and recommended measures. Pole identification information includes the pole number, geographical coordinates, and the name of the line to which it belongs; the risk level information is marked as a yellow warning or a general warning; and recommended measures include advising maintenance personnel to increase patrol frequency, monitor subsequent rainfall changes, and prepare for emergencies. For grid units in a high-risk state, a second-level early warning is generated, which also includes pole identification information, risk level information, and recommended measures. The risk level information is marked as a red warning or an emergency warning; and recommended measures include advising immediate evacuation of personnel, activation of emergency rescue plans, standby of emergency repair teams, and notification of relevant departments to take control measures. Early warning information is transmitted to transmission line maintenance personnel, dispatch centers, and emergency management departments through mobile terminal push notifications, monitoring center pop-up displays, and SMS messages to ensure timely delivery of warning information and provide decision support for emergency response to secondary geological disasters caused by wildfires along transmission lines.
[0042] It is understood that those skilled in the art can combine various implementation methods in the above embodiments under the guidance of the above examples to obtain technical solutions with multiple implementation methods.
[0043] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect, characterized in that, include: Acquire historical wildfire ignition parameters, vegetation type data, and terrain slope data within a preset geographical area of the power transmission line corridor; Based on the vegetation type data and terrain slope data, the time decay coefficient of each grid cell is determined; the time decay coefficient is used to characterize the rate at which the impact of wildfires on the stability of geological bodies decays over time. Based on the historical wildfire ignition parameters and the time decay coefficient, the geological stability impact base value of each grid cell at the current moment is calculated; the geological stability impact base value is used to characterize the degree of residual impact of historical wildfires on the stability of the current geological body. The rainfall forecast data of the transmission line corridor within a preset time period is obtained, and the geological stability impact base value and the rainfall forecast data are input into a preset geological disaster probability prediction model to output the geological disaster occurrence probability of each grid cell. If the probability of a geological disaster exceeds a preset probability threshold, then the transmission line towers located in the corresponding grid unit are identified, and early warning information is generated for the transmission line towers.
2. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 1, characterized in that, The step of determining the time decay coefficient of each grid cell based on the vegetation type data and terrain slope data includes: Obtain a preset vegetation restoration cycle mapping table, and retrieve the baseline restoration cycle of each grid unit from the vegetation restoration cycle mapping table. The vegetation restoration cycle mapping table is used to record the baseline restoration cycle of different vegetation types under standard slope conditions. Obtain a preset slope correction coefficient mapping table, and retrieve the slope correction coefficient of each grid unit from the slope correction coefficient mapping table; the slope correction coefficient mapping table is used to record the correction coefficients corresponding to different slope ranges; Multiply the baseline recovery period by the slope correction coefficient to obtain the comprehensive recovery period of each grid cell; The time decay coefficient of each grid cell is calculated based on the comprehensive recovery cycle.
3. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 2, characterized in that, The calculation of the time decay coefficient of each grid cell based on the comprehensive recovery period includes: A segmented calculation model is used to determine the time decay coefficient; wherein the segmented calculation model includes a first decay coefficient calculation model, a second decay coefficient calculation model, and a third decay coefficient calculation model, the first decay coefficient calculation model, the second decay coefficient calculation model, and the third decay coefficient calculation model correspond to different decay rates, and the decay rate of the first decay coefficient calculation model is greater than the decay rate of the second decay coefficient calculation model, and the decay rate of the second decay coefficient calculation model is greater than the decay rate of the third decay coefficient calculation model.
4. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 3, characterized in that, The step of determining the time decay coefficient using a piecewise calculation model includes: Obtain the time difference between the current warning time and the occurrence time of each historical wildfire hotspot; If the time difference is less than or equal to the first preset ratio of the comprehensive recovery period, then the time attenuation coefficient is calculated using the first attenuation coefficient calculation model. If the time difference is greater than the first preset proportion of the comprehensive recovery period and less than or equal to the second preset proportion of the comprehensive recovery period, then the time attenuation coefficient is calculated using the second attenuation coefficient calculation model. If the time difference is greater than the second preset ratio of the comprehensive recovery period, then the time decay coefficient is calculated using the third decay coefficient calculation model.
5. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 1, characterized in that, The historical wildfire ignition parameters include the time of occurrence of each wildfire, the location of the ignition point, and intensity parameters used to characterize the intensity of the wildfire.
6. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 5, characterized in that, The calculation of the geological stability impact base value of each grid cell at the current moment, based on the historical wildfire ignition parameters and the time decay coefficient, includes: Based on the fire location, each historical wildfire fire point is mapped to the corresponding grid cell; Obtain the preset geological vulnerability factor for each grid cell, and calculate the geological vulnerability index for each grid cell based on the geological vulnerability factor; The geological stability impact base value of each grid cell at the current moment is determined based on the wildfire occurrence time, intensity parameters, and geological vulnerability index.
7. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 6, characterized in that, The determination of the geological stability impact base value of each grid cell at the current moment based on the wildfire occurrence time, intensity parameters, and geological vulnerability index includes: The initial impact intensity value of each historical wildfire point is calculated based on the aforementioned intensity parameters; The attenuation weight of the fire point is determined based on the fire occurrence time and the time attenuation coefficient. Multiplying the initial influence intensity value by the attenuation weight yields the attenuated influence intensity of the fire point. Multiplying the attenuated influence intensity by the geological vulnerability index of the grid cell where the fire point is located yields the influence contribution value of the fire point to the grid cell. The impact contribution values of all historical wildfire hotspots within the same grid cell are summed to obtain the geological stability impact base value of the grid cell at the current moment.
8. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 1, characterized in that, The preset geological disaster probability prediction model is a fusion prediction model built based on machine learning algorithms; the input of the fusion prediction model includes the geological stability impact base value and the rainfall forecast data, and the output is the probability of geological disaster occurrence.
9. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 8, characterized in that, The step of inputting the geological stability impact baseline value and the rainfall forecast data into a preset geological hazard probability prediction model, and outputting the geological hazard occurrence probability of each grid cell, includes: The geological stability impact base value of each grid cell is used as the first input feature, and the rainfall forecast data is used as the second input feature, and both are input into the fusion prediction model. The fusion prediction model performs nonlinear mapping on the first and second input features to output the probability of geological disaster occurrence for each grid cell.
10. The method for early warning of secondary geological disasters caused by wildfires along power transmission lines based on the time decay effect according to claim 1, characterized in that, The early warning information includes at least one of the following: tower identification information, risk level information, and recommended measures information.