128 results about "Debris flow" patented technology

The invention discloses a plateau mountain road disaster identification method and system based on multi-source remote sensing image restoration and super-resolution reconstruction. The method comprises the following steps: acquiring and preprocessing a multi-source remote sensing image of a plateau mountain region, and extracting landform measurement parameters based on a digital elevation model; a super-resolution reconstruction network fusing deformable convolution and Transform is constructed, and a low-resolution image is reconstructed by using constraint training of a composite loss function containing geomorphic measurement parameters; performing feature extraction and adaptive weighted fusion on the preprocessed image and the reconstructed high-resolution image; based on the fused image, utilizing a multi-task deep learning model to identify landslide, debris flow and roadbed subsidence disasters along the highway; and carrying out morphological optimization and boundary refinement under GIS constraint on an identification result, and outputting a disaster thematic map. According to the invention, the precision and reliability of road disaster identification in a complex terrain environment are effectively improved.

The invention relates to the technical field of earthquake monitoring, in particular to a geological disaster early warning device based on earthquake waveform recognition, which comprises an earthquake data receiving module, an earthquake wave type analysis module, a disaster mode recognition module, an early warning signal construction module, a risk judgment module and an early warning information issuing module. According to the invention, through identifying and analyzing the seismic waveform data, the characteristics of the earthquake and the type of the caused disaster, such as landslide, rock collapse, debris flow, ground fissure, surface collapse and the like, are rapidly judged, and through comprehensively evaluating the emergency and the influence range of the earthquake event, the targeted early warning information is generated and issued in time. The geological disaster response speed and efficiency can be remarkably improved, earthquake monitoring is not limited to recording of earthquake activities, disaster risks can be evaluated in real time, more accurate disaster prediction and early warning can be provided, for areas where geological disasters frequently occur, the early warning accuracy and timeliness can be remarkably improved, casualties and property losses can be reduced, and the economic benefit is improved. And the method has extremely high practical value and social significance.

The invention provides a physical simulation test device and method for development and evolution of debris flow disasters. The physical simulation test device comprises a material source supply device used for providing materials, a material source circulation device used for material flowing and a test monitoring device, and the material source supply device comprises a movable grid-shaped belt conveyor, a water source supply device and a material source supply box; the material source circulation device comprises a circulation device and a stacking device. According to the physical simulation test device and the simulation method for development and evolution of debris flow disasters, disaster formation mechanisms and disaster movement processes of debris flow disasters composed of different material sources and under different mountain terrains can be simulated.

The invention relates to the technical field of debris flow identification, and discloses a debris flow seismic oscillation signal identification method based on machine learning. The method comprises the following steps: acquiring a debris flow ground motion signal; constructing a debris flow seismic oscillation signal identification model based on a slice attention mechanism; the debris flow seismic oscillation signal identification model based on the slice attention mechanism is utilized to firstly perform instance normalization processing on debris flow seismic oscillation signals, then cut the debris flow seismic oscillation signals through overlapped sliding windows, and then perform Fourier transform processing on each slice. And then slice time sequence features are extracted from the slice sequence subjected to Fourier transform through an LSTM model, the attention weight is calculated by adopting a multi-head attention mechanism in combination with position codes of slices at different positions, and finally a debris flow seismic oscillation signal identification result is obtained through a full connection layer.

The present invention discloses a tiered liquid collection method for the in-situ mining process of ionic rare earth ores. For ionic rare earth mines with high mountains and steep slopes or large ore volumes, the method utilizes tiered divisions and liquid collection facility layouts based on geological conditions such as the area of ​​the injection mining area, topography, and rock formation occurrence. Multiple layers of leaching diversion holes, leaching collection trenches, and rainwater drainage trenches are arranged from top to bottom to collect leaching liquid and drain natural rainfall. An anti-seepage curtain is constructed along the foot of the mountain to prevent leaching liquid from overflowing and leaking. A leaching collection well with multiple layers of leaching diversion holes is installed in the liquid accumulation area inside the anti-seepage curtain to recover the leaching liquid and control the groundwater level in the accumulation area. This method can effectively control landslides, mudslides, and debris flows that are prone to occur during the injection process in mines with high mountains and steep slopes. It can also achieve tiered mining, effective leaching liquid collection, and reduced leaching agent loss in mines with large ore volumes, contributing to the establishment of a new model of ecological and green mining for sustainable development.

The invention provides equipment for landslide debris flow accumulation body deep rock-soil characteristic drilling imaging and a detection method.The equipment comprises a triangular support composed of three adjusting supporting legs, a supporting tray is fixed to the top of the triangular support, a lifting controller is fixed to the top of the supporting tray, and exploring tubes are coaxially installed on the upper portion and the lower portion of the lifting controller; a cable penetrates through the inside of the exploring tube, the cable is matched with the lifting controller and used for displaying the extension length of the cable, the end of the cable is connected with a probe used for conducting camera shooting detection on the exploring hole, and the lifting controller and the probe are connected with a main control computer. The equipment can stably work in a complex terrain of a landslide and debris flow accumulation body and obtain the characteristics of deep rock soil in real time, so that the requirement for obtaining deep, high-precision and real-time stable drilling imaging in the complex terrain is well met, and the problem that in the prior art, the deep rock soil detection capacity is insufficient is solved.

The present application discloses a method for measuring impact force of debris flow and application, and a design method for barrier energy dissipation pile array. The method for measuring vertical impact force of debris flow measures the tangential velocity of debris flow at the impact point to measure the vertical impact force distribution of debris flow on the impact point. Further, the design method for pile array simplifies the free section of energy dissipation pile as a cantilever beam, which generates a bending resistance when impacted by debris flow, and the bending resistance continuously works to dissipate the impact kinetic energy of debris flow. According to the motion characteristics of debris flow, the impact time and distance parameters are further derived to solve the problems of measuring and designing the stability parameters of single energy dissipation pile, the energy dissipation parameters of barrier energy dissipation pile array, and the number parameters of barrier energy dissipation pile array. The present application also provides the application of each measuring and designing method. The present application provides a more scientific and perfect solution for the energy dissipation design of barrier energy dissipation pile array to resist the impact damage of rockfall and debris flow caused by landslide disaster.

The invention discloses a geological disaster real-time monitoring system based on the Internet of Things, and relates to the technical field of geological disaster intelligent early warning. Original data such as earth surface displacement, rainfall and earth sound signals are acquired through a multi-source sensing acquisition module; the edge intelligent processing module extracts disturbance characteristics and calculates disturbance entropy and an energy index; constructing a regional topological graph by using a graph neural network, and predicting a propagation path and a risk probability of an induction factor; fusing meteorological prediction and topographic factors to construct a multi-dimensional mapping model, and identifying a disaster development trend; judging whether a critical state is entered or not based on a causal reasoning map and triggering early warning; the sampling frequency and the communication priority of the sensor are dynamically adjusted in combination with the historical anomaly similarity, and resource optimization configuration is achieved; the system has the advantages of high recognition precision, high response speed, low energy consumption and high deployment adaptability, and is suitable for intelligent monitoring and early warning of geological disasters such as landslide and debris flow.

The invention belongs to the technical field of geological disaster early-stage analysis, and particularly relates to a geological disaster early-stage analysis method and system based on multi-source data, and the method comprises the steps: identifying a risk source, determining an analysis range, and identifying a risk region and a region property in the analysis range; disaster formation mode analysis: analyzing geological disaster hidden dangers to carry out disaster pregnancy conditions and disaster formation processes, and establishing geological disaster geological identification marks and indexes; risk factor analysis: on the basis of a geological disaster forming mode, analyzing geological disaster movement and accumulation characteristics by adopting a mathematical model of rock-soil body movement and fluid-solid coupling movement, and estimating dynamic reserves of the accumulation characteristics; potential disaster-bearing body vulnerability analysis, including social vulnerability analysis, material vulnerability analysis and environmental vulnerability analysis; through cooperation of the units, the line can be prevented from being selected below a slope where debris flow easily occurs or near a channel according to an estimation result, and the risk that a project is damaged by the debris flow is reduced.

The invention discloses a valley-type ice rock collapse clastic flow disaster chain danger range prediction method, system and device and a storage medium, and belongs to the field of clastic flow disaster prevention and control, and the method comprises the steps: obtaining the average width of a valley; determining the volume and the ice content of the hidden danger body; the ratio H / L of the predicted vertical height difference H to the horizontal distance L in the whole movement process of the ice rock collapse clastic flow is obtained; the ratio Ht / Lt of the vertical height difference Ht and the horizontal distance Lt from the position of the hidden danger body to the position of each point of the gully is obtained, and when the first value smaller than H / L appears in the Ht / Lt, the position is determined to be the maximum danger range boundary of the ice rock collapse clastic flow disaster in the gully. The invention provides a set of efficient, quantitative and spatialized scheme, the static distribution information of the ice rock clastic flow hidden danger can be accurately quantified into visual and available geographic space information of the maximum danger range in combination with the dynamic influence of the valley terrain, and the risk assessment and emergency decision of ice rock collapse clastic flow disasters can be more scientifically supported.

The invention discloses a debris flow early warning method, system and device, a medium and a product, and relates to the field of debris flow early warning, and the method comprises the steps: obtaining a current infrasound signal of the earth surface; calculating a power spectrum entropy value of the current infrasound signal; comparing the power spectrum entropy value of the current infrasound signal with an early warning threshold value; when the power spectrum entropy value of the current infrasound signal is smaller than the early warning threshold value, the peak frequency of the current infrasound signal is calculated, and when the peak frequency is within a set frequency range, it is judged that the current infrasound signal is a debris flow precursor signal; and outputting an early warning signal according to the debris flow precursor signal. According to the invention, early warning can be provided and the accuracy of debris flow early warning can be enhanced.

The invention relates to the technical field of debris flow disaster monitoring, and discloses a debris flow type discrimination model training method and device based on RFECV multiple models, computer equipment, a computer readable storage medium and a computer program product, and the method comprises the steps: S101, data acquisition and preprocessing; s102, carrying out feature processing; s103, data division is carried out; s104, carrying out sample unbalance processing; s105, preliminarily training the model; s106, optimizing a feature subset; s107, carrying out hyper-parameter range convergence; s108, performing hyper-parameter assignment; s109, training a candidate type discrimination model; s110, performing loop optimization judgment; and S111, outputting a discrimination result. According to the automatic feature optimization method based on RFECV, redundant features are effectively eliminated, and the interpretability and stability of the model are improved; a hyper-parameter adaptive adjustment mechanism is introduced, dynamic model optimization is achieved, training efficiency and generalization ability are enhanced, the problems that traditional classification depends on subjective experience, feature selection is difficult, and the model is unstable are solved, and efficient and accurate classification of debris flow types in the cold and cold mountainous area is achieved.

The invention discloses an intelligent assessment method and system for regional geological disaster risk, and relates to the technical field of intelligent assessment. During operation of the system, various real-time data related to debris flow are acquired by integrating real-time sensors and satellite remote sensing data, original data are preprocessed and stored in a data storage system; based on the collected data, a physical model and a machine learning model are constructed, a debris flow disaster risk index MRI, a rainfall intensity coefficient RIC, a geological vulnerability coefficient GVC and a hydrodynamic pressure coefficient HPC are obtained after calculation, the probability of debris flow occurrence is predicted, and a result is output according to the model. The debris flow disaster risk index MRI is compared with a preset first safety threshold value M and a preset second safety threshold value N, the risk level of debris flow occurrence is evaluated, different areas are managed in a classified mode, debris flow early warning is given out according to real-time monitoring data and risk evaluation results, and automatic emergency response measures are supported.

The invention discloses a mountain torrent debris flow geometric parameter and surface flow field real-time calculation method and system, and belongs to the technical field of geological disaster monitoring, and the method comprises the following three core steps: S1, precisely calculating the geometric parameters of a riverway, determining the main axis of the riverway through principal component analysis, and calculating the width through a multi-thread parallel convex hull algorithm; efficient depth inversion is realized through a KDTree spatial index, and an Alpha shape is utilized to extract a central area form so as to estimate flow; s2, dynamically analyzing a surface flow field, constructing a grid model by using Delaunay triangulation, rendering and enhancing micro-topographic features based on gradient topographic shadow, and extracting a surface flow velocity field from continuous topographic shadow by using a PIV technology; and S3, data stream processing and transmission architecture: constructing a directed acyclic graph topological structure and a dual-channel data transmission mechanism. According to the invention, the problems of real-time performance and accuracy of debris flow monitoring in a complex mountain area environment are solved, the system delay is less than 300 milliseconds, and reliable technical support can be provided for early warning of mountain torrents and debris flow disasters.

To provide an electronic device for monitoring a moisture geological phenomenon, adapted to detect stress pf a control surface.SOLUTION: An electronic device (2) includes: a box-type casing (3) placed on a control surface (S); one or a plurality of detection elements (4a, 4b, 4c and 4d) configured to convert a mechanical signal (Fi) into an electric signal (Si_ril) having characteristics of a mechanical load acting on the control surface (S); a signal transceiver (5) configured to transmit and receive data; a memory unit (6) including a threshold value (SOG); a power supply (7) configured to supply power to various kinds of elements of the electronic device (2); and a processing unit (20) configured to process moisture geological risk monitoring data. The present invention also describes the electronic device (2) for monitoring a moisture geological phenomenon, adapted to detect stress of a debris flow. The present invention also describes a corresponding method for monitoring the moisture geological phenomenon.SELECTED DRAWING: Figure 2

An embodiment of the present application provides a high-position landslide debris flow test model device, which relates to the field of high-position landslide simulation devices. A high-position landslide debris flow test model device includes: an operation bracket, the operation bracket includes an upper flat plate and a lower support frame, a landslide simulation body is rotatably installed in a rectangular window of the upper flat plate of the operation bracket, a slope adjustment driving member is fixedly installed on the upper right part of the operation bracket, the slope adjustment driving member can drive the landslide simulation body to perform slope adjustment, a slope surface adjustment driving member is installed on the upper right part of the operation bracket, and a driving component is installed on a support rod under the operation bracket, and the driving component can drive the slope adjustment driving member and the slope surface adjustment driving member to work. Through the driving component, power can be provided to the slope adjustment driving member or the slope surface adjustment driving member as needed during the simulation test, so that one driving structure can drive and adjust two adjustment components in the simulation test, reducing the driving structure required for the simulation device.

The invention relates to the technical field of intelligent monitoring, in particular to a geological disaster intelligent monitoring method and system based on cloud-edge collaboration, and the system comprises a central agent and an edge agent which are in communication connection with each other. After a disaster monitoring task is set, the central intelligent agent obtains the disaster monitoring task, for example, flood is monitored, so that cloud-side collaborative real-time monitoring is carried out on geological disasters such as mountain torrents, debris flows and reservoir landslides; the disaster monitoring task is automatically decomposed to obtain a plurality of edge sub-tasks and a central sub-task, the edge sub-tasks are executed by the edge intelligent agents, the central sub-task is executed by the central intelligent agent, cloud-edge cooperation and division cooperation are achieved, and the edge intelligent agents can operate independently and cooperate through a network; a decentralized autonomous and centralized cooperative control mode is realized, and the robustness and the real-time performance are improved; and a plurality of edge intelligent agents are arranged, so that the problem of insufficient space coverage can be solved.

The invention relates to the technical field of mountain disaster risk identification, in particular to a debris flow identification method and device based on seismic array signal time-frequency characteristics, and the method comprises the steps: arranging a field debris flow vibration monitoring array at a preset position of a target debris flow gully, so as to collect continuous seismic oscillation waveform signals; performing short-time Fourier transform on the continuous seismic oscillation waveform signal to obtain local information of a time-frequency domain; solving the time-varying power spectral density of the continuous seismic oscillation waveform signal according to the local information of the time-frequency domain; and performing energy weighted average on the time-varying power spectral density to obtain a time-varying centroid frequency, and identifying whether a debris flow disaster event occurs according to the time-varying centroid frequency. Therefore, the problems that an existing classical long and short time window ratio identification method depends on amplitude characteristics of a high signal-to-noise ratio, the diagnosis distance is only within a range of tens of meters, and the detection effect fails under the condition that the observation distance is increased to hundreds of meters are solved.

According to the moraine lake outburst risk judgment method and the application, a moraine lake outburst susceptibility judgment factor and a conversion value thereof, a glacial lake outburst debris flow occurrence possibility evaluation factor and a conversion value thereof, a main river blockage possibility judgment factor and a conversion value thereof of debris flow, and a conversion value of moraine lake outburst in a main river flood peak flow are calculated; and calculating a comprehensive moraine lake outburst risk discriminant value according to the conversion value and judging a moraine lake outburst risk grade. Aiming at the characteristics of moraine lake outburst danger, according to multiple conditions such as outburst flood, debris flow, main river blockage and flood formed by re-outburst which influence the moraine lake outburst danger, one-by-one analysis and comprehensive consideration are performed, so that the accuracy of a moraine lake outburst danger judgment result is improved, and the accuracy of the moraine lake outburst danger judgment result is improved. And the flood peak flow factor which affects the dangerousness to the greatest extent is considered prominently, so that the dangerousness evaluation result is more reasonable and more reliable.

The invention discloses an intelligent early warning system for debris flow disasters in a mountainous area based on a remote sensing network, and particularly relates to the technical field of intelligent early warning of debris flow disasters. Calculating a brightness change coefficient and a color difference coefficient of the low-resolution remote sensing image at the same pixel point from the perspective of high time resolution, quantifying a brightness change fluctuation condition and a color difference condition of the low-resolution remote sensing image at the same pixel point, and calculating a first early warning coefficient; obtaining high-resolution remote sensing images at different moments, calculating boundary similarity, verifying debris flow disaster precursor signals captured in the low-resolution remote sensing images in the high-resolution remote sensing images according to the boundary similarity, and calculating a second early warning coefficient to construct a debris flow precursor comprehensive early warning model; and outputting a debris flow precursor comprehensive early warning index, and comprehensively evaluating the precursor risk of the debris flow disaster, thereby realizing efficient and accurate debris flow disaster precursor early warning.

The invention discloses debris flow monitoring equipment for a concrete bridge framework area. Accurate early warning is realized through multi-modal sensing and intelligent decision. The geological vibration sensing module captures debris flow characteristic vibration; the soil humidity module monitors sudden change of water content; the non-contact distance measuring module tracks debris flow movement in real time; and the structural deformation module evaluates bridge safety. The data processing module is used for performing EMD decomposition on the vibration signals to extract energy entropy, and performing Kalman filtering on distance measurement data to predict a track; and fusing multi-source features based on an LSTM / random forest model, and outputting a risk level and confidence. The response module is used for triggering a local sound-light alarm through primary response; the second-level response is used for pushing remote early warning; and the three-level response is linked with the traffic signal and the protective net. And multi-sensor dynamic weight adjustment, data consistency verification, hierarchical response and structural safety linkage are realized. The method is suitable for all-weather debris flow monitoring of mountainous bridges, and significantly improves the early warning timeliness and reliability.

The present invention relates to a method for classifying the genetic types of seepage barriers in deep-sea waterways, and belongs to the technical field of geological exploration for oil and gas field development. The present invention discloses a method for classifying the genetic types of seepage barriers in deep-sea waterways, comprising obtaining the sedimentation process, scale, lithology, distribution location, and gravity flow genetic type of the deep-sea waterway; classifying the genetic types of the seepage barriers in the deep-sea waterway into landslide-induced seepage barriers, debris flow-induced seepage barriers, low-density turbidite-induced seepage barriers, and semi-pelagic sediment-induced seepage barriers based on the gravity flow genetic type of the deep-sea waterway; and further finely classifying the genetic types of the seepage barriers in combination with the lithology, scale, distribution location, and sedimentation process of the seepage barriers. The present invention fills the gap in the classification scheme of the genetic types of seepage barriers within deep-sea waterways, thereby providing important geological basis and guidance for the evaluation of the continuity and connectivity within the reservoirs of deep-sea waterways.

The present invention discloses a method and application for determining the hazard of a moraine lake outburst. The method calculates a moraine lake outburst susceptibility factor and its conversion value, a moraine lake outburst debris flow probability assessment factor and its conversion value, a moraine lake outburst probability factor for debris flow blocking the main river and its conversion value, and a moraine lake outburst probability factor for the main river's peak flow. Based on the conversion values, a comprehensive moraine lake outburst hazard assessment value is calculated and the moraine lake outburst hazard level is determined. This method addresses the characteristics of moraine lake outburst hazard and comprehensively considers multiple factors that influence the risk of a moraine lake outburst, including outburst floods, debris flows, blocking the main river, and floods caused by re-outburst. This improves the accuracy of the moraine lake outburst hazard assessment results and prioritizes the peak flow factor, which has the greatest impact on the hazard, making the hazard assessment results more reasonable and reliable.

The invention discloses a method, a system, a storage medium and equipment for determining the occurrence time and frequency of debris flow in an unmanned area based on the size of lichen, and relates to the technical field of geological disaster monitoring and dating, and the method comprises the steps of S1, condition setting, S2, lichen data acquisition, S3, lichen growth rate fitting, and S4, time and frequency determination. The method is characterized in that the exposure time of lichens is calculated according to the average diameter of ground clothes and the actual growth rate of the lichens in combination with the initial diameter of the lichens, the occurrence time of debris flow is determined according to the exposure time of the ground clothes, and the occurrence frequency of the debris flow is obtained according to the occurrence time of the debris flow. The technical problem that the time and frequency of debris flow events in depopulated areas are difficult to determine is effectively solved.

The invention discloses a green recovery ecological restoration method for a mine goaf, and relates to the technical field of ecological restoration. Rock surface treatment is firstly carried out, the exposed rock surface of a mine is cleaned, and broken stones are prevented from falling off during construction; then, regreening planting is carried out, and seedling transplanting is carried out on the reconditioned rock surface; and then vegetation maintenance is carried out, the planted seedlings and the sown grass seeds are maintained, and the vegetation can root and germinate. According to the method, through vegetation reconstruction and ecological restoration, surface coverage is rapidly restored, water and soil loss is reduced, geological disaster risks such as landslide and debris flow are effectively restrained, and the life and property safety of the mining area and surrounding residents is prevented from being threatened; the hyperaccumulator can also be used for purifying soil heavy metal pollution, and is matched with vegetation to adsorb dust and regulate microclimate, so that the regional air quality and ecological environment are remarkably improved; meanwhile, the regreening mining area can be transformed into an ecological park, a carbon sink forest or an agricultural base, and the land value is increased.

The present invention discloses a rock-blocking dam device for intercepting debris flow, comprising: a fixed plate 1 for being fixed on the surface of a mountain; a bracket 3 fixedly connected to the fixed plate 1; a perforated plate 401, a horizontal rotating shaft being provided at the lower end, the horizontal rotating shaft being fixedly connected to the bracket 3; an inclined plate 602, located at the front side of the perforated plate 401, a vertical rod 601 being provided at one end, the vertical rod 601 being perpendicular to the bracket 3; when intercepting debris flow, the angle between the inclined plate 602 and the perforated plate 401 is reduced. The rock-blocking dam device of the present invention has strong adaptive adjustment capability and high repairability, can effectively block large-scale rocks and sediment, and can solve the problem in the prior art that the rock-blocking dam is easily impacted and eroded when facing large-scale debris flow, resulting in structural damage or failure.

The invention discloses a debris flow susceptibility evaluation method considering a whole starting process condition. The debris flow susceptibility evaluation method comprises the following steps: acquiring a remote sensing satellite image of a research area; the excess terrain volume, the InSAR erosion volume and the historical landslide source volume of the area are calculated firstly, and then a comprehensive source supply index is calculated in a fusion mode; calculating a river power index SPI and a daily maximum precipitation amount of the research area, and performing normalization processing on grids of the river power index and the daily maximum precipitation amount by a minimum value and maximum value synthesis method; performing weighted stacking on the two normalized factor layers to obtain a rainfall trigger index; calculating a transportation connectivity index; and carrying out three-dimensional joint probability modeling and susceptibility index calculation. According to the method, objective and accurate evaluation of the debris flow susceptibility is realized through quantitative description of the debris flow forming process and non-parametric probability modeling.

The application discloses a mountain torrent debris flow coverage range identification method based on visual geometry modeling and relates to the technical field of debris flow coverage range identification, and comprises the following steps: obtaining a mountain torrent debris flow area image to be identified and marking the image as a real-time target image; obtaining a real-time image gray value based on the real-time target image; obtaining historical target gray values under different light intensities; obtaining a first gray threshold value and a second gray threshold value based on the historical target gray values; obtaining a first threshold function based on the first gray threshold value; obtaining a second threshold function based on the second gray threshold value; obtaining a first real-time threshold value and a second real-time threshold value based on real-time light intensity, the first gray function and the second gray function; and obtaining a mountain torrent debris flow three-dimensional coverage area based on the real-time image gray value, the first real-time threshold value and the second real-time threshold value. The application is used to solve the problem that the existing technology adopts a fixed threshold value to segment a debris flow coverage area, thereby leading to inaccurate acquisition of the debris flow coverage area.

The invention discloses a debris flow gully mountain torrent early warning method. Historical data and monitoring data are acquired; performing abnormal value elimination, missing value filling, mi n-max normalization processing, data labeling and data segmentation on the historical data to obtain a training set, a verification set and a test set; performing abnormal value elimination, missing value filling, Z-score normalization processing and space-time matching on the monitoring data in sequence to obtain a time window matrix; training the LSTM neural network based on the training set to obtain a dynamic risk prediction model; inputting the time window matrix into a dynamic risk prediction model, obtaining a disaster probability and performing graded early warning; millimeter wave displacement monitoring, soil capacitance sensing and meteorological data fusion cross-domain integration are carried out; the geological parameter weights are alpha, beta and gamma, and the threshold value is dynamically adjusted through machine learning; and a special solution for geological disasters is formed.

The invention relates to the field of geological disaster recognition, in particular to a road slope landslide debris flow disaster recognition method, device and equipment and a medium. According to the invention, through the high-spatial-resolution optical remote sensing image of the highway slope in the alpine and valley region and DEM data in the same region, deep fusion of spectral features of the optical image and DEM elevation features is realized through the cross-modal cycle fusion module, and texture damage and earth surface deformation information of landslide debris flow can be captured at the same time. Meanwhile, when the optical image is seriously covered by cloud snow, the DEM data branch can be used as an independent data source for identification, and continuity and stability of identification work are guaranteed. According to the method, two kinds of remote sensing data with different modes and information complementation are organically combined, the accuracy of landslide debris flow geological disaster recognition is improved, the requirement for rapid general survey of disaster points along highways and major engineering areas in alpine and valley areas can be met, the manual workload can be reduced, and the working efficiency is improved.