318 results about "Dependability analysis" patented technology

The invention discloses a power grid maintenance plan reliability post-evaluation method based on Monte Carlo simulation and data driving, and belongs to the technical field of power system operation and reliability analysis. The method comprises the following steps: firstly, collecting multi-source data such as historical load, renewable energy output, equipment operation state and maintenance record of a power grid, constructing a time sequence database through cleaning, time alignment and feature extraction, and establishing a load and renewable energy probability model; constructing a maintenance plan model containing a state variable, a constraint condition and a peak clipping weight mechanism, and establishing a continuous time Markov chain state model for the key equipment to generate an availability sequence; generating a large-scale random operation scene set through a Monte Carlo method based on multiple models, and carrying out supply-demand balance and power flow analysis on each scene; multi-dimensional indexes of reliability, economy and safety are calculated and subjected to weighted fusion, a comprehensive post-evaluation report is generated after results are counted, and finally the maintenance plan is optimized according to the report. According to the method, the uncertainty of the power system can be comprehensively considered, multi-dimensional quantitative evaluation and closed-loop optimization of the maintenance plan are realized, intelligent support is provided for power grid maintenance decision making, and the method is suitable for a power transmission network, a power distribution network and a micro-grid.

The invention provides a non-probabilistic time-varying reliability analysis method based on ellipsoid correlation modeling. The method comprises the following steps: carrying out interval modeling on uncertainty input parameters; constructing a multi-dimensional ellipsoid uncertain domain based on correlation information among parameters; a system time-varying performance function is established, and an interval process of output response changing along with time is obtained through uncertainty propagation solution; identifying the maximum value point of the upper boundary and the minimum value point of the lower boundary in the interval process as reliability key points; respectively calculating non-probability reliability indexes based on the ellipsoid model at the moments of the two key points; and taking the smaller value of the two indexes as the non-probability time-varying reliability index of the system in the target time interval. According to the method, the problem that the evaluation result is conservative due to the fact that variable correlation is ignored in a traditional non-probability method is effectively solved, time-varying evolution of system performance can be accurately quantified, and an accurate and practical dynamic safety evaluation tool is provided for a complex engineering system with scarce data.

A Predictive Diagnostic Information Capability-Technology (PreDICT™) system (100) enables users including expert and nonexpert users to provide information regarding a condition of a subject and receive timely and accurate information regarding risk stratification, treatment options and other medical evaluation information. The illustrated system (100) generally includes a user device (102) for use by a user assisting a subject (104), a processing platform (108), and a network (106) for connecting the user device (102) to the processing platform (108). The system (100) may also involve an emergency response network (130) that includes public-safety answering points (PSAPs) (132). The processing platform (108) processes the sensor information and other information from the user device (102), determines risk stratification information as well as medical diagnosis and treatment option information based on machine learning technology, and provides output information to the user device to assist the user in treating the subject (104).

The invention belongs to the technical field of electric energy parameter monitoring, and particularly relates to an electric energy meter reliability analysis method and system, a computer device and a storage medium. The method comprises the steps of applying disturbance of a certain amplitude under a preset disturbance condition to original operation data of an electric energy meter to generate disturbance data under the corresponding disturbance condition, inputting the original operation data and the disturbance data under each disturbance condition into an inversion model to carry out parameter backstepping, and obtaining an inversion model under the corresponding disturbance condition. An inversion model under a corresponding disturbance condition is utilized to determine a predicted output value of the electric energy meter under the disturbance condition, the reliability of the electric energy meter is evaluated based on a residual error between a theoretical output value of the electric energy meter under a standard working condition and the predicted output value of the electric energy meter under each disturbance condition, and the inversion model is endowed with sensitive response capability to external fluctuation. The behavior change of the electric energy meter under a non-ideal condition can be observed, so that the performance change of the electric energy meter caused by disturbance can be identified.

The invention provides an all-component multi-source solid waste asphalt mixture design method and system, and relates to the technical field of data processing, and the method comprises the steps: carrying out the preprocessing of obtained test proportion data of a solid waste asphalt mixture, and generating a mixture data set; screening the mixture data set by using a preset constraint condition to obtain a plurality of candidate matching points meeting the performance index characteristics; the performance index characteristics comprise gradation structure characteristics related to particle size, bonding system characteristics related to high-temperature fluidity and water sensitivity, molding volume and construction characteristics related to porosity and saturation, and context embedding characteristics related to a service scene; splicing the performance index features and inputting the spliced performance index features into a trained gating expert network to obtain a corresponding performance prediction mean value and a performance fluctuation mean value; and performing reliability analysis on the candidate matching points based on the performance prediction mean value and the performance fluctuation mean value, and determining the candidate matching points of which the reliability is greater than a preset threshold value as target matching points.

The invention discloses a fault tree Boolean function equivalent mapping method based on an untrained neural network, and relates to the field of fault tree analysis. In order to solve the problems that in the prior art, a Boolean function mapping structure is not beneficial to parallel expansion, the calculation efficiency is limited, and the Boolean function mapping structure is difficult to efficiently realize on high-parallel platforms such as a GPU, the invention provides a method for generating topological structure data by analyzing a fault tree model; the basic events, the intermediate events and the top events are mapped into neurons of an input layer, a hidden layer and an output layer respectively, a feedforward network with fixed weight and bias is constructed, and a logic activation function is defined in nodes to realize Boolean logic propagation. The input layer receives a basic event state vector, outputs a top event result through forward propagation, and can realize large-scale Boolean function mapping on a parallel platform through batch input matrixes. The method is suitable for reliability analysis, minimum cut set simplification, top event probability calculation, parallelization fault tree solving and the like of a large-scale complex system.

The invention discloses a time-varying reliability analysis method based on a double-layer BP neural network, and relates to the field of time-varying structure reliability analysis, and the method comprises the steps: generating a sample pool based on a joint probability density function which affects a random design variable of a to-be-analyzed time-varying structure; calculating weights of sample points, screening part of the sample points and establishing a first training data set; an inner-layer BP neural network model is constructed and trained, the minimum value of the time-varying performance function corresponding to each sample point in the first training data set about time is calculated, and a second training set is formed; an outer-layer BP neural network model is constructed, and a second training set is adopted for training; the reliability of a time-varying structure is calculated by adopting a trained outer-layer BP neural network model, the minimum value of a time-varying performance function about time is solved through an inner-layer model, and the outer-layer model is constructed based on parameters of sample points and the minimum value solved by the inner-layer model, is used for calculating the failure probability and has good applicability. The method can be used for time-varying reliability analysis of complex structures.

The invention discloses a multi-mode fusion power semiconductor module full-automatic detection system and method, relates to the technical field of semiconductor testing, constructs a multi-topology loading module, and realizes mechanical load testing under a frequency converter working condition, electrical load simulation under a grid-connected working condition and static electrical parameter testing. And covering dynamic and static performance detection of the module in a multi-application environment. A multi-source temperature monitoring and weighted fusion technology is introduced, and precise regulation and control of local junction temperature, surface temperature and radiator temperature of the module are realized. Dynamic switch characteristics are extracted from various collected data through wavelet transformation, and module failure probability prediction is completed in combination with a Weibull life model. Automation of whole-process data analysis and detection report generation is achieved, and the test efficiency and the evaluation intelligent level are improved. The method is suitable for multi-scene performance evaluation and reliability analysis of the power semiconductor module.

The invention provides a composite beam reliability optimization method based on a failure probability function. The method comprises the following steps: 1) defining a random input variable, a design variable, a target function and a performance function; 2) calculating a failure sample and an extended failure probability based on an autonomous learning Kriging agent model and a Monte Carlo method according to the Bayesian theory; the method comprises the steps of (1) calculating a design parameter failure sample, (2) calculating a joint probability density implicit function on the basis of an adaptive K-nearest neighbor method and autonomous learning Kriging according to the calculated design parameter failure sample, and (3) calculating a failure probability implicit function on the basis of the Bayesian theorem.By means of the method, the problem that in the reliability optimization process, the calculation cost is difficult to bear due to reliability analysis is solved.

The invention discloses an MEMS inertial sensor reliability analysis method based on long and short term memory deep learning, and belongs to the technical field of microelectronic reliability evaluation, and the method comprises the steps: collecting an original output signal and environmental parameter data of an MEMS inertial sensor in a target scene, and recording a collection timestamp; then preprocessing is carried out to obtain a time sequence processing data set; performing feature extraction on the time sequence processing data set to form a multi-dimensional feature vector; carrying out correlation analysis on the multi-dimensional feature vectors, and screening out a key feature set; constructing an LSTM model based on the key feature set, and performing training and parameter optimization on the model to enable the model to capture a time sequence dependency relationship of sensor performance degradation; inputting the real-time key feature set subjected to preprocessing and feature extraction into the trained LSTM model, and outputting probability distribution of the current health state of the sensor; and then reliability evaluation is carried out and an early warning signal is output through the system terminal to prompt maintenance personnel to carry out timely intervention.

The invention belongs to the technical field of transistor detection, and particularly relates to a field effect transistor quality detection system and equipment thereof.The system comprises a transistor scanning module, a structural reliability analysis module, a signal excitation generation module, an intelligent feature analysis module, a transistor quality evaluation decision module and an interaction alarm end; according to the invention, the structural reliability analysis module analyzes based on the transistor image to assess the structural reliability of the field effect transistor, and generates various excitation signals suitable for field effect transistor detection when generating the structural reliability signal; the intelligent feature analysis module collects various electrical parameters and non-electrical parameters of the field effect transistor under signal excitation and performs deep analysis, and performs comprehensive and objective quality evaluation on the field effect transistor according to an analysis report in combination with a preset quality standard and an evaluation model, so that comprehensive, efficient and accurate detection on the field effect transistor is realized, and the detection accuracy is improved. And the quality detection level and the production efficiency can be improved.

The invention discloses an air-space-ground Internet of Things performance analysis method and system based on SWIPT and NOMA, and belongs to the technical field of air-space-ground integrated Internet of Things. Comprising the following steps: constructing an air-space-ground Internet of Things system model comprising an LEO satellite, an unmanned aerial vehicle relay and a ground user, and considering heterogeneous channel characteristics and node dynamics; deducing a closed-form solution of the outage probability of the system through joint transformation of a moment generation function and a characteristic function of a random variable; a high-order Taylor expansion and moment matching method is adopted to carry out approximation on the closed-form solution, a first-order approximate solution and a second-order approximate solution are obtained, and an upper bound of an approximate error is deduced. According to the method, the problem of accurate modeling of the outage probability under heterogeneous link coupling in the air-space-ground Internet of Things is solved, theoretical preciseness and engineering practicability are both achieved, and an effective tool is provided for network reliability analysis and optimization.

The invention relates to the technical field of unmanned aerial vehicles. Particularly relates to an unmanned aerial vehicle cluster multi-stage task reliability analysis method and system, a storage medium and a program. According to the method, a task stage process, a subsystem dependency relationship and reliability parameters are embedded into a SysML architecture model, a fault tree of each stage is automatically generated and converted into a binary decision diagram (BDD), and single-machine multi-stage task reliability calculation is achieved. On this basis, the reliability of the cluster in each stage and in the whole task is evaluated in combination with a k-out-of-n system model and a Monte Carlo simulation method, and the minimum delivery quantity is automatically optimized based on a reliability target. Finally, simulation and optimization results are automatically written back to the MBSE model through a parameter diagram, and design-simulation-decision closed-loop linkage is achieved. The method has the advantages of high modeling consistency, high calculation efficiency, optimal resource configuration and the like, and is suitable for reliability evaluation and optimal configuration of unmanned aerial vehicle cluster tasks in a complex environment.

The invention discloses a decoupling design method and system for reliability optimization of an angular contact ball bearing. The method comprises the following steps: determining initial design parameters of the angular contact ball bearing; constructing a reliability optimization design model which takes maximization of the rated dynamic load as a target and comprises probability reliability constraint processed by a penalty function; constructing a design variable and failure probability mapping agent model, specifically, identifying and quantifying uncertainty parameters, constructing a quasi-statics model and a limit state function, calculating the failure probability of a sample point through a probability reliability method, and training the agent model taking the design variable as input and the failure probability as output; and based on the proxy model, solving the reliability optimization design model by using an optimization algorithm, and outputting optimal bearing design parameters. According to the method, the reliability analysis process and the optimization iteration process are decoupled, so that the optimization efficiency is greatly improved while the design robustness is ensured.

The invention discloses a vehicle structural strength reliability analysis method and device, electronic equipment and a medium. The method comprises the following steps: acquiring a vehicle component structure to be subjected to reliability analysis, determining a corresponding component function according to the vehicle component structure, and determining a fault mode generated due to the influence of structural strength based on the component function; establishing a structural strength finite element simulation calculation model corresponding to the vehicle part structure based on the vehicle part structure, extracting node stress data based on the structural strength finite element simulation calculation model, and determining strength data based on the vehicle part structure; and determining a strength reliability probability, a strength failure probability and a structural strength reliability index according to the node stress data and the strength data, and generating a vehicle structural strength reliability analysis result according to the component function, the failure mode, the strength reliability probability, the strength failure probability and the structural strength reliability index. And carrying out reliability evaluation on the vehicle part structure based on the vehicle structure strength reliability analysis result.

The invention belongs to the technical field of spacecraft product quality and reliability analysis modeling, and particularly relates to a fault tree generation method and device based on a knowledge graph and a large model fusion technology. The method comprises the following steps: constructing a spacecraft product fault domain ontology model, establishing a tag system comprising entity tags and relation tags, and extracting related knowledge from a technical document, a SysML model and an FMEA table by using a large language model to form a triple; according to entity types and relation types of different source knowledge, fusion is carried out, and a fault knowledge graph is constructed; obtaining a function dependence sub-graph, a fault behavior sub-graph and a redundant sub-graph of a function or a composition structure from the fault knowledge graph, and retrieving a specific sub-graph from the fault knowledge graph according to a user question; analyzing fault propagation relation and redundant structure information based on function dependence in the specific sub-graph to construct an OR gate and an AND gate, and mapping nodes in the specific sub-graph to a basic event, an intermediate event, a top event and a gate to generate a fault tree.

The application discloses a kind of abrasive grain flow microchannel polishing quality reliability analysis systems, including abrasive grain flow polishing system, quality detection system and reliability analysis system, the abrasive grain flow polishing system is used in the polishing processing of microchannel in microfluidic chip;The quality detection system is used to carry out quality detection to the microchannel after processing, and obtains detection value R1;The reliability analysis system carries out reliability analysis to processing quality, and the information basis of analysis is processing parameter P and the microchannel processing quality detection value R1 obtained by detection;The reliability analysis system transmits adjusted processing parameter to the abrasive grain flow polishing system, and is used to control microchannel polishing.The application also discloses a kind of abrasive grain flow microchannel polishing quality reliability analysis method.The application utilizes new learning function, and the correlation coefficient introduced by new learning function, improves the accuracy of model in threshold value vicinity, and can avoid the accuracy in low correlation area, ensures the accuracy of reliability analysis.

The invention discloses a space station on-orbit maintenance SA network analysis method based on DSA and a Bayesian network, and the method comprises the steps: firstly building a topological structure of an SA reliability network, then obtaining the structure parameters of the SA reliability network, carrying out the SA reliability analysis, building an applicable proposition network according to a given manned spaceflight extravehicular activity research object, and carrying out the analysis of the SA reliability. And finally, carrying out systematic and qualitative DSA analysis on the researched manned spaceflight extravehicular activity tasks based on the propositional network. The method can effectively provide support for personnel decision making of the task system, provide support for training design of task personnel, evaluate rationality of information interaction design of the current task system, identify key nodes of information interaction and provide support for information interaction design of the task system. The method is suitable for evaluation and optimization of EVA task system design of space stations in China and astronaut ground training system design.

The invention provides a reliability evaluation system and method for a two-degree-of-freedom rotational vibration energy collector, and belongs to the field of nonlinear rotational vibration energy collection, and the method comprises the steps: S1, constructing a kinetic model, and converting the kinetic model into a Hamiltonian form; s2, reducing dimensions by adopting a random average method; s3, calculating a drift coefficient and a diffusion coefficient; s4, constructing and solving a boundary value problem to verify the calculation precision; and S5, parameter sensitivity analysis and design optimization. According to the reliability evaluation system and method for the two-degree-of-freedom rotary vibration energy collector, the problems that in reliability analysis of an existing two-degree-of-freedom rotary vibration energy collector, high-dimensional coupling characteristics are difficult to capture, an analysis frame is incomplete, calculation efficiency is low, and parameter optimization lacks theoretical guidance are solved.

The application provides a reliability analysis method for positioning accuracy of a shield tool changing robot, realizes sensitivity analysis of error parameters by constructing a kinematic model and an error proxy model, and accurately evaluates the failure probability of positioning accuracy. The assembly and machining errors are comprehensively considered, the error parameters are treated as random variables, the accuracy and efficiency of the analysis are improved. The mean and standard deviation of the random variable are used for sensitivity calculation, and a failure calculation method of positioning accuracy sensitivity is obtained, which provides an important basis for optimization design. The application not only helps to reduce the tool changing time, improve the construction efficiency and reduce the safety risk, but also provides an accurate and reliable tool changing robot manufacturing scheme for designers. The method has universality and flexibility, and can be widely applied to performance improvement and fault prevention of the shield tool changing robot. The implementation of the application will promote the development of shield construction technology and provide strong support for industrial automation and efficient production.

This invention provides a high-end scientific instrument reliability analysis and design method based on SysML, belonging to the field of instrument reliability analysis technology. The method includes: constructing a system model of the target instrument using SysML, the system model including a requirements diagram, internal block diagram, and attribute library; extracting all failure modes from all failure mode libraries and assigning them global numbers to establish a failure mode mapping matrix; performing failure mode and impact analysis (FMEA) for the target instrument based on the system model to obtain a failure mode and impact analysis table; establishing a parameterized reliability model for each functional path; calculating the reliability constraints of each functional path based on the system model and the impact analysis table; and performing reliability analysis based on the measured attributes and reliability constraints of sub-components to generate an optimization scheme. The advantages of this invention are improved accuracy, consistency, and automation of reliability analysis.

The application discloses a civil aircraft minimum risk bomb position structure reliability analysis method, which comprises the following steps: S10, establishing a finite element model of an LRBL structure, and determining dangerous positions of the civil aircraft minimum risk bomb position structure; S20, determining uncertainty input variables of the LRBL structure, and performing explosion simulation and solving in combination with the uncertainty input variables; S30, determining failure criteria of the dangerous positions of the LRBL structure under explosion action, and obtaining a limit state function; S40, determining a failure probability of the dangerous positions of the LRBL structure according to the limit state function; and S50, obtaining reliability of the LRBL structure by using a fault tree analysis. Explosion simulation is used to avoid the work that can be completed only by a large number of real explosion tests, efficient estimation of the failure probability of the civil aircraft minimum risk bomb position structure is realized, and the efficiency of reliability analysis and calculation is improved.

A method, computing device, and a non-transitory computer-readable medium are provided. The computing device determines an initial node-local burst buffer content at a start of a time period T. A current node-local burst buffer content is received by the computing device during the time period. For each checkpoint / restart time interval, the computing device: estimates stochastic transition rates λ and μ; estimates input flow data rates of data entering the node-local burst buffer from a compute node; and estimates drain data rates of the data leaving the node-local burst buffer to a parallel file system. The computing device models an average statistical reliability function of the node-local burst buffer within the time period T with respect to not exceeding a predetermined threshold value. When the average statistical reliability function has a value that is less than a predetermined threshold, the computing device performs an action.

The invention discloses a method for analyzing the strength and rigidity related degradation failure reliability of an automobile crane boom. The method comprises the following steps: firstly, establishing a load-stress equation of the boom; constructing a single strength degradation model and a rigidity degradation model based on the equation; then quantizing uncertain factors influencing the reliability to obtain a failure probability function corresponding to single degradation; and finally, establishing a strength and rigidity related degradation failure model based on the functions, and analyzing the reliability. According to the method, an automobile crane boom system is used as a research object, uncertainty factors and strength and rigidity degradation laws in actual service are deeply explored, and the system reliability is evaluated through a reliability analysis method; aiming at the limitation that a traditional method only considers single degradation failure, a related failure analysis theory is introduced, and a related degradation failure model is established on the basis of a traditional single degradation model in combination with the working condition characteristics of the boom, so that a reliability analysis method considering strength and rigidity related degradation failure is formed.

The invention discloses a nuclear chemical engineering facility safety analysis method and system and electronic equipment, and belongs to the technical field of nuclear chemical plant safety analysis. The analysis method comprises the steps of obtaining an originating event list of the nuclear chemical engineering facility; according to the originating event list and the event sequence logic, respectively constructing an event tree model corresponding to each originating event; a fault tree model of a key fault branch in each event tree model is constructed, and the fault tree model comprises basic events connected with logic gates; calculating a basic event probability based on correlation analysis, human factor reliability analysis and an equipment failure database; and according to each event tree model, the fault tree model and the basic event probability, calculating a safety analysis result corresponding to each originating event. According to the method, the problems that a security analysis method based on a deterministic theory in the prior art is too conservative and low in accuracy can be solved.

The invention discloses a Bayesian reasoning method for high-voltage power supply reliability under common cause failure influence, and belongs to the technical field of high-voltage power supply reliability analysis. Constructing a dynamic fault tree of the high-voltage power supply; acquiring the proportion of common cause failure in each pair of bottom event failures, and collecting evaluation suggestions for the proportion of the bottom event to construct a fuzzy matrix; determining common cause failure factors of common cause failure of each pair of bottom events; the dynamic fault tree of common cause failure is converted into a discrete time Bayesian network, and a Bayesian network graph, a conditional probability table and a marginal probability table are obtained respectively; and performing bidirectional reasoning on the dynamic reliability of the high-voltage power supply, and calculating reliability and maintainability indexes of the high-voltage power supply in each time period. According to the method, the common cause failure factor is determined through the fuzzy theory, and the dynamic fault tree considering common cause failure is mapped into the Bayesian network for reliability analysis, so that the problems that the common cause failure factor is difficult to determine and the reliability analysis calculation amount is too large are solved, and the high-voltage power supply reliability analysis precision and calculation efficiency are effectively improved.

The invention belongs to the technical field of automatic driving automobiles, and particularly relates to an automatic driving automobile reliability analysis method based on section-by-section decision Markov. Comprising the steps that a mixed state space model containing a discrete state space and a continuous state space is constructed, discrete states comprise an automatic driving mode, a manual driving mode and a takeover process mode, and the continuous state comprises a situation awareness level; defining a deterministic evolution mechanism of continuous state variables, and setting different situation awareness level descent parameters for different driving modes to simulate the influence of passive fatigue and active fatigue; a random transfer mechanism between discrete states is defined, and a transfer rate is set as a function of a situation awareness level. The method can accurately describe the influence of various factors such as driver fatigue, the take-over process, hand release reminding and non-driving related tasks on the reliability of the self-driving automobile.

The invention relates to the technical field of computer software, in particular to a human factor reliability analysis method and device based on a cognitive mechanism, and the method comprises the steps: refining a target nuclear power plant operation instruction text into executable operation behavior units; modeling the executable operation behavior unit according to the cognitive psychology model to obtain an operation cognitive model; repeatedly running different preset operation tasks for multiple times by using the operation cognition model so as to quantify and estimate human factor event occurrence time parameters; and inputting the human factor event occurrence time parameter and a preset cognitive failure probability into a pre-constructed human factor reliability analysis model to calculate a human factor error probability. Therefore, the problems that in an existing IDHEAS-ECA method, human factor event occurrence time mainly depends on subjective expert judgment, cognitive mechanism support is lacked, result precision is limited and the like are solved, and the method is suitable for operator behavior modeling and human factor error probability estimation in high-risk industries such as nuclear power plants and the like.

The invention discloses a turbine disc fatigue reliability life influence factor importance analysis method, which comprises the following steps of: converting fatigue reliability life sensitivity into unconditional fatigue reliability life and conditional fatigue reliability life calculation under characteristic points through Gauss integral, and establishing a self-adaptive updating criterion of a finite element Kriging agent model; the finite element Kriging agent model is iteratively updated, so that the analysis time of single-time fatigue life reliability in the fatigue reliability life analysis process is shortened, and unconditional fatigue reliability life analysis and conditional fatigue reliability life analysis are based on a structural finite element model of the same structure; the idea of sequence updating structure finite element is provided and is applied to the fatigue reliability analysis life process of the semi-proxy algorithm, so that the sequence sharing of the training sample set in the fatigue life reliability analysis method under different fatigue life thresholds is realized; and the solving efficiency of the unconditional fatigue reliability life and the conditional fatigue reliability life is improved.

This invention relates to the field of image segmentation technology and discloses a construction site risk behavior recognition system based on big data. The system constructs a multi-view spatiotemporal representation of construction by uniformly aligning multi-view video streams, personnel positioning information, equipment operating status, and construction space area data at the construction site. Simultaneously, it constructs a segmentation model including an image encoder and a Bayesian mask decoder, and introduces a dynamic evidence adaptation mechanism based on Set Transformer and Hypernetwork to achieve dynamic segmentation of construction targets and hazardous areas in complex construction scenarios. Furthermore, by combining spatial posterior distribution, semantic posterior distribution, and a visual language model, it achieves reliability analysis and semantic discrimination of risk behaviors, thereby improving the reliability and adaptability of risk behavior recognition in complex construction scenarios.