487 results about "Markov chain" patented technology

The invention relates to the field of welding process control, and discloses a pipeline all-position automatic TIG (Tungsten Inert Gas) welding method which comprises the following steps: collecting pipeline geometric parameters, welding position angles and material attribute data in real time through multi-source data fusion; constructing a Gaussian process regression dynamic response model, coupling a nonlinear mapping relationship among the process parameters, the molten pool morphology and the corrosion tendency index, and dynamically adjusting the weight of the model based on the welding position angle; a hierarchical strategy of Bayesian optimization and model prediction control is adopted to generate a parameter solution set meeting the fusion depth constraint and the corrosion threshold value, and the molten pool oscillation frequency is used as feedback to correct the current in real time; the heat accumulation evolution trend is predicted through a hidden Markov chain, and parameter closed-loop migration and trajectory compensation are achieved in combination with molten pool flow field coupling correction. The problems of uneven forming quality and corrosion risk caused by space displacement, dissimilar metal interface effect and dynamic disturbance in all-position welding are solved.

The invention discloses an unmanned aerial vehicle flight path optimization system and method based on artificial intelligence and Internet of Things, and relates to the technical field of unmanned aerial vehicle control, and the method comprises the steps: collecting communication delay, environment data and an unmanned aerial vehicle state in real time; analyzing historical communication delay data by using a Markov chain, and predicting a basic delay value in a short time in the future; correcting the basic delay value through a machine learning model; outputting a final delay prediction result; calculating a delay risk index, and judging a delay risk level; based on the delay risk level, when the delay risk level is low risk, sending a control instruction in advance according to the final delay prediction result; and when the delay risk level is medium or high risk, calling a pre-trained reinforcement learning strategy library, and generating a self-adaptive control instruction. According to the method, the multi-dimensional risk index is constructed, multiple strategies are provided to adjust the flight path according to the real-time data and the risk index, and the flight stability of the unmanned aerial vehicle is improved.

The invention relates to the technical field of airport safety protection, and discloses an airport bird strike prevention method and system based on multi-sensor fusion. The method comprises the steps of collecting airport peripheral airspace radar detection, radio detection, laser monitoring, weather and other multi-mode sensor data, and performing time synchronization and space registration to generate a fusion data stream; features are extracted based on a dynamic weight distribution algorithm, target types are identified by using a multi-modal classification model, and birds and unmanned aerial vehicles are distinguished; constructing a dynamic risk assessment model based on a Markov chain to generate an intrusion risk level; and when the risk level exceeds a preset threshold value, triggering a multi-level alarm strategy and linking the expelling equipment to execute a self-adaptive expelling instruction. According to the method and system, through multi-sensor fusion, the monitoring precision is improved, the target is accurately recognized, the risk is evaluated in real time, efficient alarming and expelling are achieved, the airport bird strike prevention capacity is effectively improved, and aviation safety is guaranteed.

The invention relates to a method for obtaining a hint injection attack test case for a large language model, which combines a conditional variation auto-encoder cVAE and a Markov chain, and obtains a large language model test case through data generation and context modeling, gradual exposure of malicious instructions and simulation of multiple rounds of dialogue attacks in reality. Realizing multiple rounds of dialogue attacks on the large language model, and challenging the defense capability of the large language model; the design scheme introduces a concealment technology, role play attack, state transition and other technologies, improves the complexity and concealment of attacks, has the core advantages of automation, higher concealment, wide coverage, batch testing and the like, can evaluate the security defense capability of the large language model more truly and comprehensively, and improves the security defense capability of the large language model. The defects of an existing defense mechanism are found, and research on multi-round prompt injection attacks and improvement of a security defense mechanism are promoted.

The invention discloses a water surface target tracking method and system based on multi-physical parameter fusion, and the method comprises the steps: obtaining a physical layer observation parameter of a target through a water surface monitoring radar, and constructing a multi-dimensional nonlinear observation vector; establishing a multi-mode motion model set including constant speed, variable speed and turning, and realizing dynamic switching among motion modes through a Markov chain; adjusting a process noise covariance and an observation noise covariance based on a residual covariance estimation result in the sliding window by adopting an adaptive extended Kalman filtering algorithm; distributing weights according to the measurement variance of each physical parameter, optimizing the Kalman gain through a weighted least square method, and completing the updating and estimation of a target state; dynamic switching of motion modes is achieved through a Markov chain, and when state estimation residual errors of continuous preset times exceed a preset threshold value, model mismatch is judged, and a Markov chain model switching mechanism is triggered; according to the method, the sea condition adaptability, the calculation efficiency and the engineering expandability can be improved.

The invention relates to the technical field of corrosion detection, in particular to a grounding grid corrosion intelligent monitoring system which comprises an electrochemical data capture module, a corrosion behavior analysis module, a corrosion probability evaluation module and a monitoring strategy optimization module. According to the method, data errors and abnormal values are filtered and corrected, the overall quality of data is improved, the accuracy of analysis results is ensured, and in corrosion behavior analysis, the future trend of corrosion is effectively predicted through dynamic change detection of electrochemical properties and calculation of the corrosion rate; the time sequence analysis of the electrochemical impedance data not only enhances the prediction accuracy, but also improves the understanding depth of corrosion behaviors, the corrosion probability evaluation module adds consideration of random changes to the whole prediction model through the application of the Markov chain, so that the model is closer to the uncertainty of the real world, and through the optimization of the processing flow, the prediction accuracy is improved. The real-time performance and the early warning capability of the monitoring system are obviously improved, and the reliability and the operation accuracy of the system are enhanced.

The invention relates to an encrypted traffic detection method based on multi-dimensional feature parallel fusion, and belongs to the technical field of network security. According to the method, when a multi-dimensional feature parallel fusion framework is constructed, the limitation of traditional statistical features on dynamic evolution characterization of encryption behaviors and the dependency of graph neural network topology modeling on computing resources are fully considered; through collaborative optimization of a Markov chain dynamic quantization protocol interaction state transition rule and a lightweight graph attention hierarchical compression mechanism, a detection model gives consideration to deep feature perception capability and efficient reasoning capability at the same time; based on the classification decision realized by the fusion mechanism, the recognition robustness and real-time defense efficiency of the encrypted malicious traffic are remarkably improved, and the active security protection level of the network is effectively enhanced.

The invention discloses a back corona rapping electric dust remover fault early warning method based on all-condition monitoring, and relates to the technical field of fault early warning, and the method comprises the steps: obtaining a rapping signal, and extracting back corona features based on wavelet transform and spectral analysis; constructing a dynamic coupling model of rapping strength-dust layer thickness, and combining back corona characteristics to obtain working condition characteristic vectors; obtaining a working condition data set, and obtaining a health degree scoring sequence under each time window in combination with the working condition feature vectors; decomposing the health degree scoring sequence into a plurality of intrinsic mode components, identifying abnormal fluctuation features based on a Markov chain, and performing risk level judgment; through all-condition monitoring and multi-modal data fusion, a back corona feature extraction, dust layer thickness prediction and health degree scoring model is constructed, dynamic identification and risk early warning of the operation state of the electric dust remover are achieved, and the problems that a traditional method is single in monitoring dimension and lags in response are solved.

A dynamic vision SLAM (Simultaneous Localization and Mapping) method based on an extended Bayesian model belongs to the technical field of autonomous robot navigation and computer vision crossing, and mainly comprises the following steps: performing prior dynamic object recognition on a current frame image by using an improved PWt-YOLO network, and outputting a binary mask and semantic probability distribution; oRB feature extraction is carried out by using a hierarchical quadtree algorithm and combining an adaptive threshold value; establishing an extended Bayesian probability model, fusing semantic prior, optical flow residual and epipolar geometric constraints of two-dimensional Gaussian distribution constructed based on dynamic object boundaries, realizing time sequence transmission of dynamic feature probabilities through a Markov chain, and executing feature point filtering according to joint dynamic probabilities; and robust pose estimation is realized through RANSAC-PnP, a sliding window and the like. According to the method, the problem of SLAM system positioning drift in a dynamic environment can be effectively solved by constructing a multi-modal fusion dynamic feature discrimination system.

The invention relates to the technical field of computers, and discloses a network threat detection method and device, a storage medium and computer equipment, the method can be applied to a high-risk scene involving a user operation process in insurance services, and the method comprises the following steps: preprocessing user behavior data in network operation data to obtain structured data; key behavior features are extracted based on multiple dimensions, an abnormal state transition path is determined by using a Markov chain algorithm, flow deviation degree features are generated, and a user behavior feature set is constructed through feature fusion; performing threat detection based on a security detection rule or by using a machine learning model to obtain a threat detection result; when threats exist in the user behavior feature set, abnormal user behavior data are determined and responded and handled, and a threat detection and processing report is generated. According to the method, the capability of detecting novel, disguise and internal threats is improved, the response time is shortened, the adaptability is enhanced, and a more efficient and reliable solution is provided for network security protection in the insurance industry.

The invention relates to the field of digital core modeling, in particular to a stomatal basalt three-dimensional digital core reconstruction method, which comprises the following steps: extracting target characteristic parameters from CT (Computed Tomography) scanning gray volume data of a stomatal basalt sample, the target characteristic parameters comprising porosity, cluster number, cluster size statistics, spatial uniformity index and simplified compactness; generating an initial three-dimensional digital core model according to the target characteristic parameters; carrying out iterative optimization on the initial three-dimensional digital core model by utilizing a self-adaptive Markov chain-Monte Carlo algorithm so as to enable cluster features of the optimized three-dimensional digital core model to approach target feature parameters; and performing curvature smoothing post-processing on the optimized three-dimensional digital core model to obtain the pore basalt three-dimensional digital core model. The model not only is highly matched with real pore basalt in the aspect of macroscopic statistical characteristics, but also shows natural and smooth curved surface characteristics in the aspect of microscopic pore boundary morphology, and provides a reliable digital basis for subsequent rock physical property analysis.

The invention discloses a dynamic tracking method for an air pollution source, which relates to the technical field of pollution source tracking and comprises the following steps of: acquiring a pollution concentration raster data set, a UAV sniffing set and a meteorological field data set, reconstructing a wind field and predicting pollution transportation, generating a candidate source hypothesis, matching fingerprints and calculating weights, adaptively scheduling sensors, and performing Bayesian inversion fusion and result output. Unified pollution concentration and meteorological data are generated through multi-source data collection and three-dimensional flow field reconstruction, second-level wind field calculation is achieved through a dimensionality reduction CFD model, candidate source hypotheses are generated in combination with Lagrange backtracking and Euler transport, and a weight matrix is formed through UAV mass spectrum fingerprint bidirectional matching. Secondary cruise of the unmanned aerial vehicle and densification sampling of ground nodes are driven through confidence ranking, and sampling density is adaptively optimized. Bayesian joint likelihood and Markov chain are adopted to fuse multi-source data, dynamic source coordinates and emission rate are output, and a closed-loop iterative air pollution source tracking system is realized.

The invention discloses an infrared small target detection method based on a conditional diffusion model. Specifically, the conditional diffusion model comprises a conditional perception coding module, a conditional guidance feature fusion coding module and a feature decoding module. According to the method, multi-scale feature extraction and fusion are carried out on an infrared image through a condition-guided condition perception coding module to generate condition features, feature fusion is carried out on the condition features and a noise mask of a time step t, and a model is guided to directionally optimize a target mask. In the training process of the conditional diffusion model, a mean square error loss function based on a time step t is adopted to supervise the accuracy of mask generation, and model parameters are dynamically adjusted through an Adam optimizer to minimize noise prediction errors. In the target detection stage, denoising is carried out step by step from the initial state of Gaussian noise through Markov chain iteration in the reverse diffusion process, and finally a high-precision target mask is generated.

A predictive maintenance method based on an elevator operation and maintenance time sequence knowledge graph comprises the steps that firstly, the elevator operation and maintenance time sequence knowledge graph is constructed, predictive maintenance of electromechanical equipment parts is achieved based on elevator part maintenance period optimization, a theoretical distribution model based on Weibull distribution is constructed, and parameters of Weibull distribution are fitted through a least square method. Therefore, the model can accurately reflect the failure rule of the elevator parts; calculating an inference result of the elevator operation and maintenance time sequence knowledge graph through a graph convolution model and a Horkes process model, and carrying out recursive relation calculation of component fault rate functions in N maintenance cycles; an elevator component maintenance comprehensive cost simulation model is constructed, and the optimal maintenance cycle of elevator components is calculated through a continuous time Markov chain method; the fault prediction accuracy is improved.

The invention provides a method and system for predicting corrosion evolution of a main cable of a suspension bridge, and the method comprises the steps: dividing the main cable into a plurality of cable segments, collecting the non-real-time physical information of the cable segments, and constructing a cable segment-level data structure model; a multi-dimensional input vector is generated by utilizing the characteristics of structural stress, environmental exposure, bridge age, maintenance records and the like, and corrosion risk scoring is realized in combination with the trained XGBoost model; further constructing a multi-state corrosion evolution model based on a Markov chain, dividing the corrosion state into pitting, expansion, precursor and critical stages, and establishing a time evolution path; risk distribution is visually displayed in a corrosion thermodynamic diagram mode, and residual life prediction and early warning information is output in combination with a threshold strategy; according to the method, the limitation of dependence on a real-time sensor is broken through, active prediction can be completed based on historical and structural parameters under a passive condition, and the corrosion identification accuracy and early warning timeliness are effectively improved.

The invention relates to the technical field of virtual display video rendering, in particular to a multi-scene adaptive virtual display video rendering method, which comprises the steps of scene feature extraction and dynamic modeling, scene switching prediction and resource pre-allocation, dynamic rendering parameter optimization, multi-thread task scheduling and load balancing, real-time feedback and adaptive adjustment and the like. According to the method, scene switching can be predicted through the Markov chain model, rendering parameters and resource allocation are optimized, the rendering efficiency and the adaptation capability in a complex scene are improved, dynamic adjustment is achieved based on real-time performance monitoring, system resource consumption is reduced, and the multi-scene high-quality rendering requirement is met.

The invention relates to the technical field of complex equipment damage detection, in particular to a damage trend prediction method based on a hierarchical graph structure and Markov residual correction, which is superior to a traditional time sequence regression model in the aspects of prediction precision and generalization ability. According to the method, a damage trend prediction model based on a hierarchical graph structure and Markov residual correction is established, the damage trend prediction model based on the hierarchical graph structure and the Markov residual correction comprises a prediction module and a correction module, and the prediction module converts damage size data obtained through engine borescope inspection into a graph structure; a damage prediction network of a hierarchical graph structure is constructed to process the graph structure, a damage trend prediction problem is converted into prediction of edge weights among nodes in the graph, and internal relations among damage data are revealed; and the correction module designs a residual error correction method based on multiple types of Markov chains so as to construct a Markov chain model to correct a prediction result.

The invention discloses a bad weather degraded image restoration method and device based on dynamic degradation generation, and relates to the field of image processing, and the method comprises the steps: employing an E step of an expectation maximization algorithm to optimize a hidden variable in a training process of a noise predictor and a dynamic degradation generator, and obtaining an optimized hidden variable; in M steps of an expectation maximization algorithm, inputting the optimized hidden variables and state variables into a dynamic degradation generator, generating a degradation layer to perform semi-supervised training on a noise predictor and the dynamic degradation generator, and obtaining a trained noise predictor; in an image restoration stage, an image degraded in severe weather and random Gaussian noise are input into a trained noise predictor, an iterative denoising step obeying a Markov chain is executed, and a clean restored image is obtained. The problems that an existing severe weather image restoration model is limited in generalization ability and unstable in restoration effect are solved.

The invention discloses a blurred image segmentation method based on a mixed diffusion model, and relates to the technical field of artificial intelligence image processing and computer vision. The invention provides a novel fast, efficient and accurate hybrid denoising segmentation model HAD-Net. The HAD-Net is an end-to-end architecture composed of a diffusion model and a segmentation model; wherein underlying distribution of data is learned by using a diffusion model, and noise distribution is fully captured through a Markov chain, so that an image is denoised in a reasoning stage; a clear denoised image is used as prior input to a segmentation model, a feature encoder containing wavelet convolution is designed, and high and low frequency information is fully extracted through wavelet convolution to enable a prior image to realize guide segmentation; reasonable construction of the overall architecture greatly reduces training cost, and reasonable network depth and internal module design solve the problem of low segmentation precision caused by too large difference between morphological distribution and size of segmentation targets.

The invention discloses a new energy joint output scene generation system based on data driving, and the system comprises a data preprocessing module which collects and preprocesses wind power and photovoltaic historical output data; the joint probability distribution model construction module obtains wind power and photovoltaic edge distribution through probability density function fitting, and performs correlation analysis on wind power and photovoltaic through a Kendall rank correlation coefficient and a tail dependency coefficient to obtain a matching error of each connection function; the Euclidean distance between each connection function and the preprocessed wind power and photovoltaic historical output data is calculated, and an optimal connection function is obtained by combining the matching error, so that a wind power and photovoltaic joint probability distribution model is established; the scene generation module uses a Markov chain to form a joint probability distribution model with time correlation, and generates a random scene set through Markov chain Monte Carlo algorithm sampling. The stability of the power system is improved, and the operation cost is reduced.

The invention relates to the cross technical field of AI multi-modal analysis and dialect processing, in particular to a dialect content crawling and auditing system and method based on AI analysis, a simulation terminal generates a real person behavior entropy interval event stream through a Markov chain and reinforcement learning hybrid model, a dialect exclusive operation library and a differentiated interest strategy are integrated, and a real person behavior entropy interval event stream is generated through a real person behavior entropy interval event stream; anti-crawling is avoided in combination with a proxy IP pool and a Bezier curve trajectory, a search unit realizes directional crawling of dialect keywords through a hot updateable script, a blind patrol mode locates high-risk content depending on a triple knowledge graph and risk prediction, whole-course block chain evidence storage is performed, and a collaborative analysis unit analyzes dialect audio and video features in a layered architecture. The multi-modal large model is combined with the professional small model, the audio spectrum, the text translation and the video picture are fused to realize cross-modal violation detection, the application service unit alarms violation content in real time, a supervision report containing a violation distribution thermodynamic diagram is generated, and the accuracy and traceability of dialect violation content crawling and auditing are improved.

The invention relates to the technical field of pipeline safety monitoring and water seepage prevention early warning, in particular to a water seepage prevention early warning system driven by complex temperature and humidity environment edge calculation, which comprises the following modules: an environment and water immersion data acquisition module used for acquiring temperature, relative humidity, water immersion signals and auxiliary evidence in real time at edge nodes and performing preprocessing; and the temperature and humidity coupling dew point judgment module is used for calculating a dew point at the edge node based on the temperature and humidity data and generating condensation negative evidence. According to the method, temperature and humidity and water immersion data are collected at edge nodes and preprocessed, condensation negative evidence is generated based on the temperature and humidity data and fused with water immersion positive evidence to form a Markov chain, meanwhile, an environment base line is established, and a threshold value is adjusted in a self-adaptive mode, so that continuous monitoring and closed-loop judgment of the water seepage state are achieved; therefore, the problems that most traditional pipeline water seepage monitoring systems depend on a remote server, and judgment lags and misinformation is caused by the complex environment and the static threshold value are solved.

The invention discloses an old-age tumor patient early-stage weakness risk prediction system based on an improved population algorithm, and the system comprises a data collection module which constructs an old-age tumor patient data set; the data preprocessing module is used for generating a fused elderly tumor patient feature matrix; the feature processing module forms an initial feature set; the feature screening module inputs the initial feature set into a discrete empire butterfly optimization algorithm to generate a key feature subset; the multi-scale Markov chain modeling module is used for constructing a multi-scale Markov chain model based on the key feature subset to form a risk state sequence; and the dynamic risk prediction module is used for judging whether the early-stage weakening risk of the old tumor patient reaches an early-warning standard or not and generating corresponding risk early-warning information. A layered prediction mechanism is established, so that the model can perform comprehensive analysis from short-term health fluctuation and medium-term change trend to long-term risk evolution.

The invention discloses a Teensy virus identification and protection method for a mobile hard disk, and relates to the technical field of virus identification and protection.The method effectively identifies high-simulation attack equipment through composite equipment fingerprint construction and combination of hardware ripple features and dynamic protocol response, and breaks through the limitation of traditional single feature detection; meanwhile, the protocol state transition probability model adopts hidden Markov chain real-time analysis, microsecond-level protocol switching abnormity is accurately captured, and the problem of missing detection caused by rough time granularity of an existing scheme is solved; in addition, a window context instruction association mechanism binds a system focus state with an HID operation, and blocks a hidden attack chain formed by legal instruction combination; the hierarchical fusing strategy fuses protocol endpoint control and physical layer isolation, so that the availability of a storage function is ensured, meanwhile, attack blocking is realized, and the situation that normal use is influenced due to full-port forbidding in a traditional scheme is avoided.

The invention relates to the technical field of market trend prediction, and discloses a market trend prediction system driven by macroeconomic indicators. An index acquisition module of the system dynamically acquires core economic indexes such as GDP growth rate, CPI, PMI and currency supply; the data preprocessing module is used for carrying out layered noise reduction processing on the multi-source heterogeneous data; the feature engineering module constructs a market sensitive feature set through spatio-temporal feature fusion; the prediction model building module is used for building a multi-layer nonlinear prediction model based on a deep belief network; the dynamic adjustment module adopts reinforcement learning to optimize a decision threshold value and combines a Markov chain to carry out state transition planning; and the feedback iteration module analyzes and predicts deviation through Bayesian filtering and realizes strategy updating. According to the method, deep learning and reinforcement learning technologies are creatively fused, the prediction precision is remarkably improved through a dynamic calibration mechanism, and the method can be widely applied to the macroeconomic analysis fields of financial investment, industrial planning and the like.

The invention relates to the technical field of power system inertia evaluation, in particular to a random mode switching driven power distribution network equivalent inertia online evaluation method and system, and the method comprises the steps: carrying out the priori definition of a group of hidden operation modes based on the equivalent inertia of a power distribution network, and constructing an unobservable Markov chain; a double-layer hidden Markov jump system model is constructed, the bottom layer is an unobservable Markov chain, and the upper layer is a continuous dynamic behavior model for describing the power distribution network; on the basis of PMU high-resolution time sequence disturbance data, deducing the probability of a system maximum probability dominant operation mode and the probability of each hidden operation mode; and carrying out multi-mode probability adaptive weighting on each equivalent inertia estimation value based on the probability of the dominant operation mode and each hidden operation mode so as to obtain the equivalent inertia estimation value of the power distribution network on line. Through the method, the problem that the equivalent inertia of the power distribution network is difficult to quickly, accurately, dynamically and adaptively assess online under the conditions of high permeability of new energy and frequent switching of working conditions is effectively solved.

The invention relates to a seismic liquefaction assessment method based on conditional random field simulation, which comprises the following steps: firstly, obtaining a logarithmic normal distribution random field of a target area under a corresponding SPT-N value, then resampling through a Bootstrap method, constructing a weighted prior probability density function of the target area in combination with a likelihood function, and finally calculating the seismic liquefaction of the target area according to a Bayesian theory. A Markov chain Monte Carlo sampling method is combined, through posterior probability density distribution, an optimal horizontal direction correlation distance is determined, a covariance matrix is constructed to generate a conditional random field, and then through multiple times of simulation, the conditional random field is converged; and finally, aiming at the target area, through calculation of a cyclic stress ratio and a cyclic resistance ratio, constructing a liquefaction probability distribution diagram corresponding to the target area. According to the method, a conditional random field simulation method is inferred and improved by combining Bootstrap and Bayesian theories, the precision and reliability of geological parameter simulation are remarkably improved, and reliable data support is provided for seismic liquefaction assessment of deep and uneven site engineering.

The invention relates to the technical field of clean energy market optimization, and particularly provides an optimal scheduling method and device for a power system, and the method comprises the steps: obtaining a policy factor adjustment vector of the power system through a pre-trained Markov chain prediction model based on a policy factor vector of a current power system; adjusting the policy factor vector of the current power system by using the policy factor adjustment vector of the power system; wherein the policy factor comprises at least one of the following: carbon quota demand, carbon emission reduction and green evidence obligation consumption. According to the technical scheme, the market mechanism can be optimized, development of renewable energy sources can be promoted, the electric power structure can be optimized, and the aim of carbon emission reduction can be achieved.

The invention relates to the technical field of intelligent tourism management, and discloses a tourist behavior dynamic modeling method based on space-time big data, which comprises the steps of collecting and preprocessing multi-source positioning data, inputting the multi-source positioning data into a multi-source positioning fusion engine, and combining statistical optimization and deep learning model fusion to obtain a high-precision positioning model. Track reconstruction and time sequence aggregation are carried out on continuous position points of tourists based on a high-precision positioning model, a structured activity data graph is generated, activity data of the tourists at all positions are obtained, activity preference characteristics are extracted through an association rule mining and clustering algorithm, and a Markov chain and a spatio-temporal evolution model are combined to predict a tourist flow trend. And forming a flow prediction result, and generating a visual dynamic decision support by using the prediction result. According to the invention, visual and dynamic decision support is provided, the crowd congestion is relieved, and the operation safety of the scenic area is guaranteed.

The invention discloses a method for exploring a heterogeneous catalytic reaction network based on a deep potential energy surface model, and the method comprises the steps: carrying out the accurate prediction of the free energy change of all possible element reactions on a catalyst through the combination of transfer learning and a density functional theory, and constructing a reaction network diagram based on a graph theory method; an improved Dijkstra method is used for searching a reaction network diagram to determine a reaction mechanism, and the selectivity of the catalyst is explored based on a Markov chain method. According to the technical scheme provided by the invention, when the CO2 reduction reaction mechanism on the monatomic alloy is explored, the prediction error MAE of the electron energy of each intermediate is only 0.16 eV, the calculated amount is reduced by 90%, the dominant path of the CO2 reduction reaction is visually displayed, and a new normal form is provided for complex mechanism exploration and catalyst screening.