323 results about "Risk classification" patented technology

The embodiment of the invention relates to the field of artificial intelligence, and provides a model context protocol injection attack protection method and device based on dynamic semantic analysis, computer equipment and a storage medium, and the method comprises the steps: obtaining request data corresponding to a call request initiated by a user through a model context protocol, carrying out the preprocessing of the request data, obtaining the preprocessed request data; performing semantic vectorization on request text and context historical information in the request data through a lightweight bidirectional encoder representation model to output an initial risk score; correcting the initial risk score according to context historical information carried in a model context protocol to obtain a corrected final risk score; and performing hierarchical defense decision according to the final risk score, and determining risk grading information corresponding to each piece of request data so as to execute a protection action corresponding to each piece of risk grading information. By adopting the method, the accuracy of identifying the protocol injection attack can be improved.

The invention discloses a cold region tunnel freeze injury diagnosis, risk grading and control system and method, and relates to the technical field of tunnel engineering. The system comprises a basic data layer for storing disease, index and measure databases; the core analysis layer is used for diagnosing a frost heaving mechanism, identifying single-factor, double-factor and three-factor frost heaving types, calculating freezing depth and frost heaving force and analyzing sensitivity; the risk level evaluation layer is used for dividing a disease zone, a frost heaving level, a freezing injury risk and a freezing injury risk response level; and the intelligent decision-making layer constructs an intelligent comprehensive decision-making analysis platform to realize risk-measure accurate matching and closed-loop management and control. The method comprises the four steps of data acquisition and storage, frost heaving mechanism diagnosis and quantification, risk grade evaluation and intelligent decision measure matching, through multi-source data fusion, multi-factor coupling analysis and zoning and grading prevention and control, the whole-process accurate management and control of the cold region tunnel frost damage is realized, and the frost damage treatment efficiency and the tunnel operation safety are improved.

The invention discloses a driver controller detection system, which comprises an acquisition module, an acquisition module, a processing module, an identification module, a judgment module, an analysis module and a generation module, the acquisition module captures contact information and electromagnetic interference data of contacts in real time by means of a high-dynamic voltage sensor and a broadband EMC probe; the acquisition module is triggered by a locomotive signal and records dynamic working condition data; the processing module aligns the data, performs denoising and extracts fault parameters; the identification module scans the voltage waveform and judges the transient contact failure; the judgment module monitors EMC data and compares an instruction position, and judges misoperation; the analysis module analyzes fault association, judges a composite type and triggers an alarm; the generation module generates a detection report and a log. According to the system, a multi-dimensional monitoring network is constructed under a dynamic working condition, data reliability is guaranteed, faults are accurately identified, a compound fault source is positioned, and the safety and stability of locomotive operation are improved through risk grading early warning and automatic reporting.

The invention provides a puree production line microbial pollution risk real-time monitoring and early warning method, which comprises the following steps: deploying multiple types of sensors, an industrial camera and an operation log interface, collecting environmental temperature and humidity, pH value, dissolved oxygen, image and operation behavior multi-modal data in a fermentation tank in real time, and carrying out edge calculation and standardized preprocessing to obtain the microbial pollution risk of a puree production line. Data missing, abnormity and time sequence difference are eliminated; multi-modal features and a dynamic knowledge graph constructed based on a production process and a pollution event are fused, and a deep learning model and a graph neural network are utilized to realize microbial pollution risk probability intelligent prediction; risk grading and a dynamic weighting algorithm are introduced, grading early warning signals and disposal suggestions are automatically generated, and a production control system is linked to execute response measures; according to the subsequent production state, continuous feedback is carried out, the model and the early warning threshold are adjusted in a self-adaptive mode, and the accuracy, response efficiency and production safety of pollution early warning are remarkably improved.

The invention relates to the technical field of image recognition, in particular to a burn scar hyperplasia risk assessment method combining semantic segmentation and texture feature analysis, which comprises the following steps: acquiring an image gray gradient, color jump and texture variance, calculating pixel mutation, extracting a mutation boundary, constructing a periodic direction field, and extracting a continuous offset region. According to the method, by extracting the pixel gray gradient, the color jump and the texture variance, calculating the boundary sudden change intensity and generating the layer, the structure change characteristics can be refined, the image anomaly perception precision can be enhanced, the semantic boundary can be identified based on the sudden change sequence, and the physical continuity is prevented from interfering the segmentation accuracy. A periodic evolution record is constructed through direction gradient, the dynamic trend of the structure is disclosed, an expansion area is locked in combination with direction continuous offset and change stability, a classification label is constructed through point location density, direction consistency and a gradient module value, and the interpretability and accuracy of risk identification are improved.

A soft rock roadway reinforced roof anchoring thickness collaborative design method comprises the steps that the proportion of grout filling cracks in a grouting reinforcement area is counted, a grouting dispersion uniformity index is calculated, and grouting effect evaluation is conducted; measuring the actual boundary depth of the grouting reinforcement area by adopting a method of combining drilling radar scanning and rock core sampling; mechanical parameters of the grouted rock mass are obtained, and a grouting enhancement coefficient is calculated; a stress transfer efficiency evaluation model is established, and the critical cooperative bearing surface depth is determined by analyzing the stress attenuation law at the boundary of the grouting reinforcement area; the critical anchoring layer thickness is calculated under the constraint condition that it is ensured that an anchoring system can effectively penetrate through a grouting strengthening area and go deep into a stable rock stratum; the cooperative work performance of a grouting reinforcement area and an anchoring system is quantitatively evaluated by calculating a system cooperation degree index, and a cooperative bearing efficiency evaluation system based on dynamic risk grading and control is established. According to the method, collaborative optimization of grouting and anchoring parameters can be achieved, and formation of a bolting-grouting integrated bearing structure can be ensured.

The invention provides an intelligent monitoring and risk early warning method for urban road settlement, and belongs to the technical field of urban management based on machine learning. The method comprises the following steps: firstly, collecting four types of multi-source space-time monitoring data, including urban road settlement data, underground environment data, pavement structure data and dynamic load and environment data; secondly, constructing a multi-source data space-time completion model, performing unsupervised completion on sparse monitoring area data, and generating space-time continuous settlement field data; thirdly, constructing a road settlement health index prediction model fusing multiple factors, inputting complementation data and original features, and outputting a health index of a 0-1 continuous interval; and finally, in combination with the health index and the road function level, four-level risk classification and dynamic early warning are realized. According to the method, the problems of space-time faults and data islands of a traditional method are solved, urban road global real-time monitoring is achieved, and early warning upgrading from qualitative judgment to quantitative grading is achieved.

The invention discloses an urban low-altitude landing risk assessment method based on three-dimensional space grids, and belongs to the technical field of low-altitude flight risk assessment. The objective of the invention is to solve the problem of accurate assessment of the low-altitude falling risk in a complex city scene. The method comprises the steps of performing three-dimensional rasterization processing on a low-altitude airspace of an evaluation airspace to obtain voxels of the evaluation airspace, and obtaining a voxel set of the evaluation airspace; building shelter formed by population density grids or mobile signaling inversion population, sensitive facility vectors, tree canopy shielding and three-dimensional building models is quantified, values are assigned to voxels of an evaluation airspace, and a voxelization environment parameter database covering the whole domain is obtained; constructing an aircraft failure rate model; carrying out an aircraft falling trajectory uncertainty simulation test to obtain landing coverage probability density distribution and tail end speed of the aircraft; and carrying out voxel risk synthesis by considering the probability of collision with people, and carrying out risk grading on the obtained voxel risk, thereby completing urban low-altitude landing risk assessment based on the three-dimensional space grid.

The invention provides an underground space multi-disaster coupling effect quantitative evaluation and quantitative control system and method. The system is characterized in that a multi-source sensing module, a data fusion and preprocessing module, a coupling risk evaluation module and a quantitative control decision module are connected in sequence; the method comprises the following steps: collecting multi-source heterogeneous monitoring data for a long time based on the multi-source sensing module; performing space-time registration, denoising and missing value interpolation on the multi-source heterogeneous monitoring data, and calculating an original disaster intensity index; performing normalization processing on different disaster intensity indexes, constructing a time-varying coupling factor between disasters, and calculating a comprehensive risk index by using a coupling risk assessment module; the quantitative control decision module performs dynamic risk grading based on the comprehensive risk index; and the quantitative control decision module calls and executes a quantitative control strategy from a preset strategy library based on the risk grading result and the dominant disaster mode combination. According to the invention, accurate monitoring and early warning of mine underground engineering disasters can be realized, and corresponding control decisions can be automatically generated.

The invention discloses a charging pile thermal runaway intelligent protection method and system based on edge calculation, and relates to the technical field of charging pile thermal runaway protection. Comprising the following steps: S1, collecting thermal runaway multi-mode sensing data in real time, and carrying out data preprocessing; a charging pile thermal runaway multi-mode abnormal state is judged, and an abnormal feature data packet is generated; s2, multi-modal abnormal feature vectors are constructed, the thermal runaway risk probability is evaluated, and thermal runaway risk grading protection of the charging pile is carried out; s3, the response effect of thermal runaway risk grading protection is quantified, and thermal runaway risk grading protection is adjusted; and S4, monitoring and feeding back the thermal runaway risk, and optimizing algorithm parameters and a thermal runaway risk grading protection strategy. The problems that an existing charging pile thermal runaway protection system is difficult to adapt to complex working conditions, multiple in false alarm and missing alarm, lagging in response and insufficient in protocol compatibility and data stability, and consequently the thermal runaway risk is difficult to recognize and protect timely and accurately are solved.

The invention relates to the technical field of automobile auxiliary driving, in particular to an intelligent driving auxiliary system and method based on multi-modal emotion recognition, and the method comprises the steps: building a driver baseline model; synchronously acquiring a face image, a biological signal and vehicle operation data of a target driver in real time, and performing preprocessing; performing feature extraction on the preprocessed facial image and biological signal of the driver, and constructing a fused emotion feature based on a driver baseline model; performing emotional state recognition and confidence calculation according to the fused emotional features; calculating an emotional risk index according to the emotional state, the confidence coefficient and the vehicle operation data, and performing risk grading; and executing a corresponding dynamic response strategy according to a risk grading result. According to the invention, accurate perception and graded active intervention on the emotion and risk of the driver are realized, and the driving safety is obviously improved.

The invention provides a postoperative bleeding risk monitoring method and system for a hepatobiliary patient. The method comprises the following steps: collecting real-time multi-modal data; constructing an LSTM-CNN hybrid model by using the time-frequency decomposition features, and obtaining a local tissue hypoxia index and a vasomotor function anomaly probability; the low-frequency impedance change rate and the albumin level are fused through a random forest algorithm, and the ascites occurrence probability and the effusion amount predicted value are obtained; and generating a bleeding point positioning coordinate and a thermodynamic diagram risk grade. And calculating a comprehensive bleeding risk probability and positioning a bleeding area. And generating graded early warning signals and recommending personalized treatment schemes or nursing suggestions. According to the invention, the LSTM-CNN hybrid model, the random forest algorithm and the three-dimensional convolutional neural network are adopted to deeply extract different data features, so that the limitation of single index evaluation is avoided. A causal relationship model is established through the Bayesian network, and the comprehensive risk probability calculation preciseness is improved; the early warning threshold is dynamically adjusted by combining the individual characteristics of the patient, and the traditional problem of easy misjudgment is solved.

The invention relates to the technical field of safety monitoring of power transmission equipment, in particular to an intelligent early warning nut and a monitoring method thereof.The nut comprises a quantitative anti-interference sensing and dynamic filtering module, a verification parameter is dynamically calculated based on an interference-verification quantitative model by synchronously collecting loosening distance, vibration and electromagnetic interference data, and the verification parameter is obtained; a real loosening event is accurately identified; the environment coupling trend quantitative prediction module is used for fusing multi-environment factor dynamic weighting such as humidity, wind speed and temperature difference after real loosening is judged, and predicting a future loosening distance through an environment-loosening quantitative correlation model; and the scenarized risk grading response module is used for calculating a scenarized risk quantized value R and triggering grading early warning in combination with the prediction result and the structural importance weight of the bolt mounting position, and generating a strategy instruction containing a maintenance priority. According to the invention, the problems of high false alarm rate, incapability of predictive maintenance and short endurance in the prior art are solved, and accurate sensing, intelligent prediction and graded early warning of the bolt loosening state are realized.

The embodiment of the invention provides a vehicle control method which comprises the steps that first road condition data and second road condition data of a preset road are obtained, the first road condition data are from a platform data source, and the second road condition data are from a vehicle-mounted data source; performing risk classification processing based on the first road condition data to obtain a target risk type of the preset road; determining a target risk level corresponding to the target risk type based on the second road condition data; based on the target risk type and the corresponding target risk level, determining a target driving strategy when driving on the preset road; and performing driving control on the vehicle based on the target driving strategy. Due to the fact that the first road condition data of the platform data source and the second road condition data of the vehicle-mounted data source are integrated, the road condition of an emergency can be obtained in advance, the corresponding driving strategy is determined, prospective vehicle control is achieved, and the response speed during vehicle control can be increased.

The invention relates to the technical field of data processing, in particular to a data auditing method, system and device based on AI model assistance and a medium. The method comprises the following steps: firstly, performing text extraction on to-be-audited data through an OCR (Optical Character Recognition) module, then performing semantic analysis by utilizing an NLP (Network Length Polymorphism) module, and converting an unstructured text into a structured feature; processing the structured features based on a preset knowledge base to generate auditing feature vectors containing business semantics; then inputting the feature vector into a machine learning model composed of an anomaly detection model and a risk classification model, and generating a preliminary auditing conclusion from three dimensions of a basic rule, a business rule and a risk rule; and finally, carrying out confidence evaluation on the conclusion of the three dimensions, and triggering a corresponding processing flow according to an evaluation result. Through cooperation of a plurality of AI models, conversion from surface text recognition to deep semantic understanding is realized, a comprehensive evaluation mechanism based on multiple dimensions is established, and the accuracy and efficiency of auditing are improved.

The invention discloses a water supply network leakage monitoring method, which is characterized in that a multi-source monitoring system of inlet flow and partition pressure is constructed based on a water supply network metering partition, and a partition reference operation model is established in a stage of judging stable operation. Real-time operation data are preprocessed, multi-dimensional features such as inlet flow trend, pressure change, flow-pressure correlation deviation and night low-load stability are extracted, and the features are fused to form a comprehensive deviation index. Constructing a self-adaptive threshold value in combination with historical same-type operation data, and realizing leakage abnormity identification and risk grading output by adopting a persistence discrimination rule; after early warning is triggered, a suspected leakage area is output by combining a pipe network topological relation based on the change amplitude difference and the change sequence relation of the pressure monitoring points in the subareas, and an auxiliary decision is provided for on-site refined leakage detection.

The invention discloses a paralytic nursing intelligent early warning system and method based on the Internet of Things technology. The system comprises a multi-modal physiological parameter acquisition module, an edge calculation preprocessing unit, an intelligent data transmission module, a multi-scale time sequence feature extraction module, a space-time diagram convolutional network module, a cross-modal attention fusion module, a multi-task risk prediction module and a model training and optimization module. According to the system, multi-mode data such as electrocardio, blood pressure, blood oxygen, eye movement tracks and voice are collected, preprocessed at an edge end and then transmitted to a cloud end; a multi-scale convolutional network is adopted to extract time sequence features, parameter association is modeled through space-time diagram convolution, cross-modal data fusion is realized by using an attention mechanism, and finally risk classification, anomaly detection and trend prediction are completed through a multi-task network. The early-stage, accurate and explainable early warning of the stroke risk is realized, and the early warning accuracy and clinical practicability are remarkably improved.

The invention relates to the technical field of emergency management, and discloses an emergency response early warning system for sudden fire blast events, which comprises a multi-source data access module used for collecting target equipment data and environment reference equipment data of energy storage equipment or a centralized charging place in real time, including temperature, current, voltage, fan state and smoke / smell sensor data, preprocessing the collected target equipment data and the environment reference equipment data; the abnormal index presetting module is used for presetting a composite abnormal index and a risk grading standard for representing the early-stage signs of thermal runaway; and the same-field benchmarking benchmark module is based on the acquired target equipment data and environment benchmark equipment data. A same-field benchmarking reference module is adopted, and based on environment reference data and abnormal accelerated change trend calculation, the technical effect of identifying a composite abnormal index in an extremely early stage is achieved, pre-warning is achieved, and the defect that a safety window period reserved for an operator to adopt an initial treatment action is short is overcome.

The invention discloses a geological disaster multi-scale environment control factor identification system based on machine learning, and relates to the technical field of geological disaster identification, and the system comprises a multi-scale environment factor collection module which obtains historical geological disaster data of a target area, extracts geological disaster multi-scale environment features, and constructs a geological disaster-environment factor correlation feature set; the quantitative evaluation module is used for establishing a geological disaster risk probability prediction model, quantitatively evaluating the marginal contribution degree of each environmental factor on different spatial scales, identifying the environmental factors dominating geological disaster development in multiple scales of a region, a drainage basin and a slope, and establishing a geological disaster risk probability curved surface of a target region and an environmental factor contribution spatial distribution diagram; and the prevention and control module introduces a human activity intensity index and an ecological system state index in a target area, establishes an environmental factor dynamic coupling risk propagation model and a real-time multi-scale risk transmission path, and generates a dynamic risk grading early warning map. According to the invention, the accuracy of geological disaster risk assessment is improved.

The invention provides a rockburst risk grading prediction method based on an improved LSTM neural network. The rockburst risk grading prediction method comprises the following steps: S1, collecting rockburst influence factor data; s2, preprocessing the rock burst influence factor data; s3, constructing an improved LSTM, residual connection and SVR collaborative fusion neural network model; s4, performing model training and hyper-parameter optimization; s5, performing double-label coding on the rockburst risk level; s6, performing model performance evaluation and optimization; and S7, performing rockburst risk grading prediction application. According to the method, accurate, efficient and dynamic grading prediction of the rockburst risk can be achieved, the method is suitable for rockburst risk grading prediction of deep underground engineering such as tunnels, mine pits and underground workshops, accurate technical support can be provided for construction safety decision making and protective measure making, and casualties and property losses caused by rockburst disasters are effectively reduced.

The invention discloses a weather-resistant steel corrosion trend pre-judgment method and system based on multi-modal data fusion and a dynamic graph neural network, and belongs to the technical field of structural health monitoring and industrial artificial intelligence. The method comprises the steps of collecting a time sequence corrosion image and environmental data, performing multi-scale visual feature extraction and environmental feature enhancement, and constructing graph structure data; a dynamic space-time diagram neural network model is constructed and trained, the model adaptively captures spatial dependence through a dynamic diagram structure learning device, space-time laws are jointly mined through a hierarchical space-time feature fusion module, and physical rationality of prediction is ensured by using a physical knowledge constraint module; and performing uncertainty quantification and risk grading based on model output. The system comprises a data acquisition module, a preprocessing module, a graph construction module, a dynamic space-time graph neural network model module, a trend prediction and analysis module, a visualization module, an alarm module and the like. According to the method, through deep fusion of the dynamic graph structure and multiple modals, more accurate and reliable space-time pre-judgment with physical significance on the corrosion trend of the weathering resistant steel is realized, and key technical support is provided for preventive maintenance of the structure.

The embodiment of the invention provides a container transportation state monitoring method based on multi-parameter perception, which is applied to the technical field of logistics monitoring and comprises the following steps: by taking a Hall switch trigger signal as a time reference and a logic starting point, acquiring acceleration, attitude angle, position and temperature and humidity signals in a time window before and after triggering; performing time sequence correlation analysis on the acceleration, the attitude angle and the position signal to generate initial event judgment; when the initial event is determined to be physical door opening or hoisting start, extracting an impact decay time constant and an attitude angle deviation retention amount, and constructing a two-dimensional feature space to perform mechanical impact classification; asynchronously calling a temperature and humidity signal according to a judgment result for root verification; and generating a joint event code and executing differentiated resource control. According to the invention, triggering events with different properties can be accurately distinguished, risk grading of mechanical impact and source tracing of environmental abnormity are realized, terminal endurance time is obviously prolonged, and monitoring reliability is ensured.

Detection of a malicious application or website by a transaction processing application includes inputting a transactional request having a transactional data record and a screenshot; sending the input data record and input screenshot to a backend controller; requesting to a prompt selector, a string comprising a feature-extraction prompt; sending the input screenshot and the received prompt string to a Large Vision Model, LVM; receiving a string having risk classification features from said LVM; verifying the received string by a format parser; if the received string fails the verification, requesting by the backend controller, a string having a feature-extraction prompt which explicitly mentions format parsing compatibility, and repeating the preceding steps; sending the received string to a risk classification model for providing a risk classification; sending the risk classification to the backend controller; determining if the application or website is determined as malicious, and accepting or rejecting the transactional request accordingly.

The invention relates to the technical field of geological disaster prevention and control, and discloses a mountain and valley area landslide hazard source identification and dynamic risk assessment method, which is characterized by comprising the following steps: S1, regional disaster-pregnant mode analysis and identification target region delineating: constructing a comprehensive susceptibility evaluation model, and delineating investigation target regions with different priorities in combination with InSAR deformation monitoring; s2, fine identification of a potential sliding source: accurately identifying a landslide boundary and a deep structure by adopting a'satellite-aircraft-ground 'cooperative technology combining multi-stage LiDAR topographic change detection, unmanned aerial vehicle multispectral and thermal infrared detection and geophysical exploration; and S3, dynamic risk assessment: establishing a dynamic and static load coupling stability model considering rainstorm and earthquake working conditions, calculating an instability probability by adopting Monte Carlo simulation, and realizing quantitative risk classification in combination with vulnerability analysis of a disaster-bearing body. According to the invention, early identification, accurate positioning and dynamic risk quantification of landslide hazard sources are realized, and the prevention and control capability of landslide disasters in alpine and valley areas is improved.

The invention discloses an identification analysis data management and association method and system based on an AI large model, and relates to the technical field of identification analysis, and the method comprises the steps: carrying out the semantic recognition and structural analysis of a multi-source heterogeneous financial document through the AI large model, and extracting key business identification elements; constructing a dynamic knowledge graph oriented to the order checking task based on the key business identification elements; in combination with a preset business rule set and a preset anomaly detection model, performing cooperative verification on the entities in the dynamic knowledge graph and the association relationship between the entities to generate a verification result; establishing business causal association and flow circulation association between the key business identification elements according to the verification result; and a risk grading decision is generated for abnormal data which are not qualified through verification. According to the method, the processing efficiency of the multi-source heterogeneous data in a financial service scene is improved, and the deep correlation analysis capability between the data is enhanced.

The application discloses a kind of urban lake algal bloom measurement method based on algal vertical distribution structure analysis, belong to environmental science field, the application constructs with column plane concentration level, column plane algal community composition, column plane aggregation characteristics and column plane profile form as key dimension, including 14 single index algal vertical distribution structure comprehensive evaluation index system, realizes the quantitative analysis of algal vertical distribution structure.The application establishes a set of fast response and risk classification as core algal bloom measurement process, including algal in-situ stereoscopic observation, algal vertical distribution structure comprehensive evaluation, algal vertical distribution structure classification and algal bloom and risk identification etc.Step.The method breaks through the limitation of traditional method in monitoring dimension, observation data is not fully mined and algal bloom determination standard is fuzzy etc.Limit, lays a theoretical foundation for algal bloom monitoring and evaluation system research, application in urban lake can provide technical means and scientific basis for algal bloom prediction and early warning and algal bloom precision prevention and control.

This invention discloses an artificial intelligence-based method for monitoring soil erosion risk. To address the problems of untimely identification of soil erosion risk, difficulty in quantifying disturbance information, and insufficient early warning accuracy within the responsibility area of ​​production and construction projects, this invention acquires high-resolution remote sensing images and divides the responsibility areas into prevention and control units. It utilizes a disturbance identification model to identify pixel-level engineering disturbances and extract disturbance features such as disturbance area ratio, disturbance morphology, and spatial relationship between disturbances and drainage channels. Combining rainfall, topography, soil, vegetation, and soil and water conservation measures data, it calculates key factors such as rainfall erosivity, vegetation cover management, slope length and gradient, soil erodibility, and measure factors. These factors are then substituted into a modified general soil loss equation to obtain baseline soil erosion. The baseline soil erosion, disturbance features, key factors, and rainfall forecast are input into a deep residual network model to obtain target soil erosion risk indicators. Finally, based on risk classification thresholds, the risk level of each prevention and control responsibility unit is output, achieving the technical effect of unit-scale soil erosion risk monitoring and early warning.

The invention belongs to the technical field of artificial intelligence, and relates to a service alarm method based on multi-modal data and a related device thereof, and the method comprises the steps: obtaining the multi-modal data when a target event occurs; carrying out cross-modal feature fusion on the extracted multi-modal features; generating a first fusion feature, a second fusion feature and a third fusion feature; respectively inputting the first fusion feature, the second fusion feature and the third fusion feature to the business risk classification prediction model after contrast learning enhancement, and obtaining business risk prediction values respectively output by the model for the first fusion feature, the second fusion feature and the third fusion feature; and based on the risk prediction value, determining whether to carry out business risk alarm on a target event. According to the invention, the method achieves the business alarm judgment of the target event from the perspective of multi-modal data, achieves the business risk prediction from different feature fusion dimensions, achieves the comprehensive alarm judgment of the target event, and improves the alarm research and judgment intensity.

The invention discloses a traffic alarm automatic processing method and system based on a multi-modal large model. The method comprises the following steps: step 1, multi-source heterogeneous data acquisition and real-time access; 2, carrying out multi-modal space-time alignment and fusion preprocessing; 3, event recognition and semantic reasoning based on a multi-modal large model; 4, generating a disposal strategy and arranging an instruction; 5, safe and controllable instruction execution and linkage control are carried out; step 6, performing treatment effect feedback and closed-loop optimization; and step 7, full-link auditing and visual redisk are carried out. According to the invention, the event detection accuracy in a complex environment is improved, and automation of alarm generation, risk grading, strategy generation and facility linkage is realized; the adaptability and expansibility of the system in different traffic environments are improved, the real-time response requirements of typical application scenes such as expressways and tunnels are met, and a reliable basis is provided for subsequent disposal strategies.

The invention discloses a debris flow silt arrester disaster prevention benefit evaluation method, and the method comprises the steps: constructing a hierarchical evaluation system which faces a future operation state and can integrate landform shape constraints, reservoir capacity operation conditions and a safety risk state on the premise of not remarkably increasing the data acquisition cost; according to the invention, an integrated technology of AHP weighting, two-level fuzzy synthesis and maximum membership level judgment is provided; wherein a calculation framework based on threshold determination and piecewise linear membership functions, qualitative index feature vector assignment, penalty mapping of interval optimal indexes and two-stage fuzzy comprehensive evaluation jointly form core features of the method; the weight is determined based on the analytic hierarchy process, the consistency check is passed, the two-stage fuzzy comprehensive evaluation and the maximum membership principle are further adopted to output the levels of'excellent, good, medium and poor ', and a quantitative basis is provided for inspection and maintenance of the silt arrester, desilting modification and risk grading management and control.