25 results about "Risk avoidance" patented technology

The application provides a tower crane construction safety monitoring method and system, and belongs to the technical field of engineering construction safety monitoring. The method comprises the following steps: acquiring multi-source data of a target tower crane, wherein the multi-source data comprises sensor data, video monitoring data and environmental data; determining a plurality of operating characteristics of the target tower crane based on the multi-source data; the plurality of operating characteristics comprise operating state characteristics, importance characteristics and environmental risk characteristics; determining a monitoring strategy of the target tower crane based on the plurality of operating characteristics; monitoring safety parameters of the target tower crane based on the monitoring strategy to obtain safety monitoring results; performing risk assessment based on the safety monitoring results and generating a risk avoidance strategy; wherein the target tower crane is one of a plurality of tower cranes in a construction site. The application improves the accuracy of tower crane construction safety monitoring, discovers risks in a timely manner and generates a risk avoidance strategy, thereby ensuring the safety of tower crane construction.

The invention relates to the technical field of engineering management, in particular to an engineering scene-oriented salary performance risk identification and control method and system. According to the technical scheme, the method comprises the steps of obtaining a risk tag type, a risk tag standard value and risk tag data; and judging whether the risk tag data corresponding to the risk tag type is empty or not. And constructing salary performance weights in one-to-one correspondence with the risk tag types. And calculating the coincidence coefficient of the contractor. And calculating the project capability goodness of fit, the project risk adaptation value and the risk avoidance degree of each contractor. And carrying out sorting recommendation on the contractor according to the risk avoidance degree from large to small. The higher the project capability goodness of fit and the higher the project risk adaptation value are, the higher the contractor is matched with the project, so that manual matching of related workers is avoided, and the management burden is reduced. Calculation is carried out in two aspects of project capability and project risk adaptation, calculation is more comprehensive, and the obtained recommended contractor better meets the requirement.

This application relates to a vehicle risk avoidance method and electronic device, belonging to the field of vehicle technology. The vehicle risk avoidance method includes: when the vehicle is parked and an animal recognition event is detected, acquiring animal status information and personnel location information, wherein the animal recognition event indicates the presence of a dangerous animal target within a preset area around the vehicle, and the animal status information is determined based on the state of the dangerous animal target; the personnel location information is used to represent the relative position between the personnel associated with the vehicle and the vehicle; determining a target risk avoidance strategy based on the animal status information and personnel location information; and controlling the vehicle to execute the target risk avoidance strategy. This application establishes a complete risk avoidance process from risk perception to targeted risk avoidance strategy formulation, and then to automatic execution of the risk avoidance strategy, enabling vehicles to provide users with proactive warnings and risk avoidance assistance in outdoor scenarios such as camping, thereby reducing the safety risks posed by dangerous animals in outdoor environments.

This invention discloses a construction supervision decision-making method and system based on game optimization and multi-agent reinforcement learning, relating to the field of construction supervision decision-making. This invention constructs a digital twin environment based on a Stackelberg master-slave game architecture, combines graph attention networks to extract features from the dynamic spatiotemporal interaction graph of the construction site, applies continuously differentiable hard safety constraints to the game process using a safety risk potential energy field based on physical field theory, and employs a multi-agent reinforcement learning algorithm based on Lagrange duality optimization to solve the Nash equilibrium strategy. This enables the supervision agent and the construction / resource agent to achieve an automatic balance between risk avoidance and schedule optimization in complex dynamic games, realizing intelligent collaborative decision-making that balances safety compliance and construction efficiency.

The invention discloses an intelligent community management system based on artificial intelligence, and relates to the field of intelligent community management, and the intelligent community management system comprises a scene perception pre-judgment module, a resource dynamic scheduling module, a man-machine cooperation interaction module and a full-period data closed-loop module, the scene perception pre-judgment module is a core data input end, and the full-period data closed-loop module is a core data output end. The information acquisition module is responsible for acquiring community multi-dimensional information and pre-judging a potential scene and providing a decision basis for subsequent modules; and the resource dynamic scheduling module is used for undertaking a result of the scene perception pre-judgment module and realizing adaptive allocation of human and material resources of the community. According to the intelligent community management system based on artificial intelligence, active risk avoidance and demand response can be realized, through the multi-modal fusion acquisition and behavior intention pre-judgment sub-module of the scene perception pre-judgment module, the existing passive alarm limitation is broken through, the problem can be recognized in advance, sufficient time is reserved for disposal, and the risk avoidance and demand response can be realized. The community safety risk and the resident life inconvenience are reduced.

The method for power collaborative distribution and risk avoidance of fracturing vehicle fleet relates to the technical field of fracturing vehicles, and is executed by a central collaborative controller which is in communication connection with local controllers corresponding to the fracturing vehicles; the method comprises the following steps: collecting running state parameters of each fracturing vehicle and receiving a total target of operation; constructing a dynamic collaborative state model of the vehicle fleet based on the running state parameters; establishing a multi-target collaborative optimization model; performing real-time solution on the multi-target collaborative optimization model to calculate target speed setting values; judging whether any target speed setting value is close to a preset safe working limit; judging whether there is a potential risk; selecting to adjust the target speed setting value according to the judgment result; and distributing final control instructions to the local controllers of the fracturing vehicles. The application can realize a global optimization control method with high robustness for power dynamic collaborative distribution of fracturing vehicle fleet, avoidance of running risks and the like.

This technology provides the ability to further improve the safety of those being monitored. [Solution] A danger avoidance support system, server device, information processing method, program, and a monitored terminal are provided, comprising a monitored terminal carried by the person being monitored and a server. The server acquires the current location information of the monitored terminal, acquires danger information associated with area and time, and transmits predetermined warning information to the monitored terminal when the current location meets notification conditions based on the risk level and reliability of the acquired danger information. The reliability level is based on information source attributes, etc., and the risk level can be corrected by environmental factors. The server updates through relearning, and the terminal outputs vibrations, etc., according to the danger level.

This invention discloses an end-to-end autonomous driving method based on a physical information fusion model, comprising: S1, collecting environmental observations, vehicle control actions, and calculated ground truth values ​​of physical features, and storing them in an experience replay pool; S2, training a world model that fuses physical information, which jointly learns latent state evolution and physical indicator prediction through an encoder, a state transition network, and a physical prediction head, and optimizes by minimizing the joint loss incorporating physical prediction errors; S3, based on the trained world model, performing policy learning in the latent space that considers physical indicator constraints to achieve preventative risk avoidance and smooth control; S4, deploying the trained policy to the vehicle controller for online control. This invention significantly improves the safety, stability, and physical consistency of autonomous driving systems in complex scenarios by explicitly modeling and predicting physical indicators in the world model and using them as forward-looking constraints in policy learning.

The application provides a financial abnormal behavior pattern recognition method and system, the method comprising: converting behavior anchor information into a high-dimensional feature vector, calculating the projection value of each high-dimensional feature vector on different behavior intention axes, generating a preliminary intention vector, optimizing the intention vector accuracy, and generating a final behavior intention vector; embedding the behavior intention vector into a preset behavior configuration manifold, identifying the atypical pattern of the behavior intention vector, and labeling the high-risk nodes in the atypical pattern; judging the spatial distance of each behavior intention vector to the labeled high-risk nodes in the behavior configuration manifold, introducing a high-risk weight based on the spatial distance and the basic disturbance density, obtaining a weighted disturbance density, dynamically adjusting the future deduction depth, and forming a prediction result with probability distribution; performing abnormal scoring on the prediction result and triggering a warning action, the application is converted from post-recognition to pre-prediction, when the intention structure deviates, it can still be identified in the configuration manifold, and the risk of deliberately avoiding identification of financial abnormal behavior is reduced.

The application provides an emergency rescue path planning method and system with adaptive risk avoidance, comprising: establishing an emergency rescue road network structure; before an actual natural disaster occurs, performing a risk entropy driven adaptive Monte Carlo simulation analysis, performing path information statistical analysis on the Monte Carlo simulation result, and establishing a pre-planned path library; when the actual natural disaster occurs, establishing a prior risk distribution of each planned path in the pre-planned path library; using a sequential Bayesian inference to dynamically correct the prior risk distribution of each planned path in the emergency rescue process; based on the path posterior risk mean of each planned path after the sequential Bayesian inference, performing dynamic planning of the emergency rescue path, selecting an optimal rescue path, and performing an emergency rescue task until the emergency rescue activity ends. Through the path dynamic adjustment strategy, the dispersed transportation strategy, the path pre-planning strategy, and the network communication free strategy, the application ensures that the rescue resources efficiently and safely reach the target location under the natural disaster scene of flood, typhoon, and the like.

This invention discloses a disaster avoidance planning method based on dynamic spatiotemporal evolution, relating to the field of building safety technology. The method includes: real-time collection of environmental status data and personnel location data through a multi-source sensor network and positioning system deployed within the building; based on a pre-set disaster dynamics model, using the environmental status data and personnel location data as initial conditions, extrapolating the disaster spatiotemporal evolution set within a specific future time window; abstracting the building space into a graph structure composed of multiple disaster avoidance planning units based on a BIM model, using the disaster spatiotemporal evolution set as input to the evacuation route planning units, and providing a set of parallel planning routes for evacuation groups in different zones within the building; continuously monitoring the personnel density on each route and the real-time updated disaster spatiotemporal evolution set, and automatically triggering a route redistribution mechanism based on the monitoring results. This invention can significantly improve the safety and efficiency of personnel evacuation during building emergencies and reduce the risk of secondary disasters.

The application discloses a chemical industrial park safety risk intelligent management and control system and method, belongs to the technical field of risk management and control, and solves the problems of incomplete tank area risk perception, data transmission delay and insufficient dynamic path risk avoidance. The system comprises a sensing device, a 5G communication module and an industrial Ethernet, a data transmission network is formed by the 5G communication module and the industrial Ethernet, and data is transmitted to a cloud platform; a time sequence database is used by the cloud platform to establish a device code index, a rule engine executes leakage judgment, and when methane is continuously 3-5 cycles > 25% LEL and the pressure drop rate > 0.5 MPa / min, a warning is triggered; a risk analysis module calls three-dimensional geographic information data to generate a 500-meter-2000-meter radius evacuation path through a Dijkstra algorithm, avoids the upwind area of the real-time wind direction and preferentially connects a fire station; and an application terminal receives a warning instruction containing a leakage device code and a dynamic path. The system is used for rapid warning and evacuation planning of chemical safety accidents.

The invention relates to a food detection scheduling method, system and device based on reinforcement learning and a medium. The method comprises the following steps: generating a system state containing a real-time risk assessment vector based on dynamic technical feature vectors of all samples to be detected and state codes of detection equipment; inputting the system state into a pre-constructed strategy network to obtain an output scheduling decision; on the basis of a dynamic weight generation network, calculating according to the real-time risk assessment vector to obtain an instant reward for executing a scheduling decision, monitoring whether a preset high-risk event occurs in the detection equipment system, and if the high-risk event occurs, generating an event-driven reward; and when an event-driven reward is generated, backtracking a historical decision information sequence corresponding to the high-risk event according to the historical decision tracking list, and calculating to obtain a tracing reward of the historical decision information sequence based on the event-driven reward and the historical decision information sequence. By adopting the method, scheduling risk avoidance and dynamic balance and collaborative optimization of the overall operation efficiency can be realized.

The invention discloses a power grid emergency material dispatching double-layer optimization method and system under an extreme weather. The method comprises the following steps: firstly, acquiring real-time meteorological monitoring data and vulnerability characteristics of power grid equipment, calculating a dynamic time sensitivity coefficient, and generating a material demand priority parameter in combination with a disaster point load level; secondly, constructing a double-layer material scheduling model, and introducing constraints to realize degraded supply under resource limitation by taking the maximization of a key load recovery index as a target at the upper layer; the lower layer achieves dynamic risk avoiding and time-space synchronization of people and goods through constraints with the purpose of minimizing the comprehensive scheduling cost; and finally, a double-layer iteration solving strategy based on interlayer interaction is adopted, an upper-layer topology construction mask is utilized to execute dimension reduction locking and dynamic truncation on lower-layer flow, and real-time weather and road condition data are accessed in the solving process to execute online rolling correction. According to the method, the problem that traditional scheduling lacks environment perception and supply and demand flexibility is solved, and the real-time adaptability, decision robustness and cooperation efficiency of power grid emergency repair under the extreme weather are improved.

The invention provides a financial abnormal behavior pattern recognition method and system. The method comprises the steps that behavior anchoring information is converted into high-dimensional feature vectors, the projection value of each high-dimensional feature vector on different behavior intention axes is calculated, a preliminary intention vector is generated, intention vector precision is optimized, and a final behavior intention vector is generated. Embedding the behavior intention vector into a preset behavior configuration manifold, identifying an atypical mode of the behavior intention vector, and labeling a high-risk node in the atypical mode; judging the spatial distance of each behavior intention vector marked with a high-risk node in the behavior configuration manifold, introducing a high-risk weight based on the spatial distance and the basic disturbance density, obtaining the weighted disturbance density, dynamically adjusting the future deduction depth, and forming a prediction result with probability distribution; according to the financial abnormal behavior identification method and device, post-event identification is converted into pre-event prediction, when the intention structure deviates, the intention structure can still be identified in the configuration manifold, and the deliberate avoidance risk of financial abnormal behavior identification is reduced.

The invention discloses an intelligent safety management method and device for engineering construction and a medium. The method comprises the following steps: acquiring a monitoring video stream; identifying a plurality of to-be-detected targets contained in the monitoring video stream according to predefined detection target features in the intelligent camera, and extracting a motion trail sequence of the plurality of to-be-detected targets in real time; performing cooperative relationship identification on the to-be-detected target to determine a construction cooperative group included in the monitoring video stream; identifying an abnormal behavior node corresponding to each constructor in the construction cooperation group to obtain an estimated risk event corresponding to the abnormal behavior node; simulating a risk conduction path of the pre-estimated risk event to obtain constructors affected by the abnormal behavior nodes; and calculating a dangerous area corresponding to the risk conduction path to determine a risk avoiding instruction of the constructor. Through the method, the technical problem that in the prior art, a long-distance engineering construction safety management method cannot predict safety risks caused by individual abnormity in a construction group is solved.

This disclosure relates to an industrial equipment control method and apparatus for early warning anomaly detection and risk avoidance. The method includes: acquiring the controllable state and statistical state of preset parameters of the equipment to be controlled at the current moment; inputting the controllable state and statistical state into a control action decision model, and determining the selection probability of each preset control action by using the objective function and constraints of the control action decision model with the goal of maximizing benefits; determining the target control action based on the preset action selection strategy and the selection probability of each preset control action; and controlling the equipment to be controlled according to the target control action. Through the above technical solution, the controllable state and statistical state of preset parameters at the current moment can be input into the control action decision model to generate control actions in real time and control the equipment. This allows for timely implementation of corresponding control actions to handle anomalies when abnormal characteristics are detected, reducing losses caused by equipment anomalies.

The invention discloses a visual management method for on-site construction of a construction project perceived by the Internet of Things, and belongs to the technical field of construction engineering management. Comprising the following steps: constructing a digital twin base model; deploying a multi-mode Internet of Things sensing network and performing data semantic mapping; creating and driving a dynamic digital twinborn body in real time; constructing and solving a construction element coupling risk potential field model; a transmissive visual interaction is provided. Dynamic elements of a construction site are abstracted into'potential points' carrying instantaneous risk potential, and interaction force among the potential points and composite risk potential of any point in space are calculated in real time through a mathematical formula defining risk gravitation and safety repulsive force. By solving the model, the method can quantify and visualize a dynamic composite risk generated by superposition of a plurality of risk sources, realizes predictive early warning from an event result to a risk cause, and can generate a navigation type optimal risk avoiding path instruction.

A fracturing truck fleet power collaborative distribution and risk avoidance method relates to the technical field of fracturing trucks, and is executed by a central collaborative controller, and the central collaborative controller is in communication connection with local controllers corresponding to a plurality of fracturing trucks; the method comprises the following steps: collecting operation state parameters of each fracturing truck, and receiving a total operation target; based on the operation state parameters, constructing a dynamic collaborative state model of the motorcade; establishing a multi-objective collaborative optimization model; solving the multi-target collaborative optimization model in real time, and calculating a target rotating speed set value; judging whether any target rotating speed set value is close to a preset safety working limit or not; judging whether a potential risk exists or not; according to the judgment result, the target rotating speed set value is selected to be adjusted; and distributing the final control instruction to a local controller of each fracturing truck. According to the method, dynamic collaborative distribution of fracturing fleet power can be realized, operation risks are avoided, and the method is a global optimization control method with high robustness.

The invention relates to an automatic traffic management method and device based on Beidou and digital twinning, equipment and a medium, and the method comprises the steps: obtaining the positioning data of each cargo transportation vehicle if a current Beidou signal is not lost; updating a pre-generated twin road grid database to obtain a real-time risk grid model; obtaining the abnormal state coordinates in the respective carriages; based on a real-time risk grid model, the abnormal state coordinates are bound to a road grid where the cargo transport vehicle is located currently, and risk superposition information is generated; generating a plurality of risk avoidance paths based on the risk superposition information and the real-time traffic flow density; based on the road gradient and refrigeration equipment power consumption data of the current vehicle, selecting a path with the lowest comprehensive energy consumption as a target bypass path; and simulating the influence of the target bypassing path on the regional traffic flow density. According to the method, the randomness of manual selection is avoided, the path energy consumption calculation error is reduced, and the congestion avoidance success rate is improved.

A traffic accident evidence chain generation system, based on a multimodal large language model, includes an input management module, a multimodal reasoning module, a fact extraction and consistency judgment module, a reverse correction module, a structured information extraction module, a causal analysis module, and an evidence chain construction module. This invention does not handle keyframe screening itself, but instead uses the set of key collision frames and the set of narrative frames output by external modules as predetermined inputs, establishing a strict sequential processing chain around text generation—fact alignment—summary reconstruction—frame-level correction—structured extraction—causal analysis—evidence chain generation—risk avoidance suggestions. This invention uses the facts of key collision frames as constraints to perform natural language reasoning and summary reconstruction on the narrative summary, suppressing model illusion and narrative drift at the source.

The invention discloses a risk prediction management system and method for a cross-border supply chain, and belongs to the technical field of risk management, and the system comprises a plan decomposition module which is used for obtaining a sequential net demand; the supplier retrieval module is used for obtaining qualified suppliers; the task distribution module is used for obtaining a sequential production task sequence; the production compensation module is used for obtaining compensation time sequence production redundancy; the transportation compensation module is used for obtaining a compensation time sequence transportation task; the compensation fusion module is used for generating a first cross-border supply chain toughness strategy; and the risk avoidance module is used for executing active risk avoidance so as to solve the technical problem that full-link active risk avoidance is difficult to realize due to insufficient supply chain toughness caused by multi-source uncertainty of production and transportation links in a cross-border supply chain, and the purposes of compensating production redundancy and transportation congestion through dynamic prediction and improving the safety of the cross-border supply chain are achieved. The active risk avoidance capability and the resource optimization efficiency of the cross-border supply chain are improved, and the delivery delay and the operation cost are reduced.

The invention relates to the technical field of data processing, and discloses a data auditing method, computer equipment and a computer program product. The method comprises the steps of obtaining to-be-audited data of multiple application architecture hierarchies, binding the to-be-audited data of the multiple application architecture hierarchies based on a preset association strategy, determining the risk degree of the bound to-be-audited data based on association features of the bound to-be-audited data, and executing a risk avoidance strategy corresponding to the risk degree according to the risk degree. Wherein the plurality of application architecture hierarchies comprise a user layer, a request layer and a database layer, and the preset association strategy represents a mapping relationship of the to-be-audited data among the plurality of application architecture hierarchies. The efficiency and accuracy of the audit evaluation process are improved.

The invention discloses a high-risk vehicle scene recognition method and system, electronic equipment and a storage medium, and relates to the field of vehicle control, and the method comprises the steps: S1, obtaining the historical risk assessment information of a scene where a vehicle is located; s2, generating a risk avoiding strategy model according to the historical risk assessment information of the scene where the vehicle is located; s3, acquiring current sensing information of a scene where the vehicle is located; s4, according to the current sensing information of the scene where the vehicle is located and the risk avoiding strategy model, a vehicle system control instruction is generated and executed; and S5, updating historical risk assessment information of a scene where the vehicle is located according to execution of a vehicle system control instruction, and iterating the risk avoiding strategy model.