External field test safety early warning method and system based on simulation extrapolation
By using a simulation-based extrapolation method, combined with discrete event simulation and dynamic principles, and utilizing multi-source detection data for model correction and risk assessment, the subjectivity and lag issues of traditional field test safety early warning methods are resolved, enabling real-time safety assessment and early warning for diverse test scenarios and equipment types.
Patent Information
- Application Number
- CN202511257331.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-04
- Publication Date
- 2025-12-23
AI Technical Summary
Traditional field test safety early warning methods rely on manual observation and experience judgment, which are highly subjective, have a slow response, and are difficult to detect and warn of potential safety hazards in a timely manner. Moreover, existing simulation models are difficult to meet the real-time safety assessment needs of diverse test scenarios and equipment types.
The simulation extrapolation method is used to create an initial simulation model by obtaining the model and application scenario parameters of the device under test. The state transition matrix is determined by combining discrete event simulation and dynamic principles. Multi-source detection data is fused to generate a predicted motion trajectory. Risk assessment and early warning are then performed based on a Bayesian network.
It improves the safety of field tests and the generalization ability of prediction models, enhances adaptability to changing environments and prediction accuracy, and realizes dynamic assessment and real-time safety early warning.
Smart Images
Figure CN121189141A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of virtual simulation technology, and in particular to a method and system for safety early warning of field tests based on simulation extrapolation. Background Technology
[0002] Field testing refers to the testing and evaluation of equipment or systems in real-world environments. It involves complex scenarios and various equipment types, and carries inherent safety risks. For example, during UAV field test flights, especially for new models, extensive hardware-in-the-loop (HIL) simulations are required to ensure UAV reliability and reduce flight risks. Traditional field test safety warning methods rely primarily on manual observation and experience, which are subjective, slow to react, and difficult to detect and warn of potential safety hazards in a timely manner.
[0003] With the development of computer simulation technology, safety early warning technology based on simulation models has gradually gained attention. Existing technologies have established some simulation models for specific scenarios or equipment types, which can predict and assess safety risks during testing. However, due to the diversity of test scenarios and equipment types, a single simulation model is insufficient to meet comprehensive safety early warning needs. Furthermore, existing technologies lack the ability to monitor and fuse real-time data, failing to acquire dynamic information from the test site in a timely manner, thus affecting the real-time performance and accuracy of early warnings. Summary of the Invention
[0004] In view of this, the present invention proposes a field test safety early warning method and system based on simulation extrapolation.
[0005] The technical solution of this invention is implemented as follows: The first aspect of this invention provides a field test safety early warning method based on simulation extrapolation, comprising: Obtain the model parameters and application scenario parameters of the device under test, and create an initial simulation model; Based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, and combined with discrete event simulation and dynamic principles, the state transition matrix of the motion state of the device under test over time is determined, and the initial simulation model is corrected using the state transition matrix to obtain the target prediction model; the motion performance parameters include velocity parameters, acceleration parameters, temperature parameters, pressure parameters, and displacement parameters; The multi-source detection data acquired in real time are fused to obtain multi-source fused data, which is then input into the target prediction model to obtain the predicted motion trajectory of the device under test.
[0006] Based on the above technical solutions, preferably, the device under test includes aerial operating equipment, ground operating equipment, and water operating equipment; the step of determining the state transition matrix of the motion state of the device under test over time based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, includes: Based on the principles of fluid mechanics, Newton's second law and aerodynamics, the flight state of the aerial operating equipment is simulated to determine the first state transition matrix of the motion state of the aerial operating equipment as a function of time. Based on the vehicle dynamics model combined with motion control theory and machine learning algorithm, the driving state of the ground-based equipment is simulated, and the second state transition matrix of the motion state of the ground-based equipment changing with time is determined. Based on fluid mechanics combined with the momentum theorem and the law of conservation of energy, the navigation state of the waterborne equipment is simulated, and the third state transition matrix of the motion state of the waterborne equipment as a function of time is determined.
[0007] Based on the above technical solutions, preferably, the step of determining the state transition matrix of the motion state of the device under test over time, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, includes: The application scenario parameters are characterized to establish a three-dimensional occupancy matrix; the three-dimensional occupancy matrix includes obstacle areas and free space; Based on the Euclidean distance from the device under test to the nearest obstacle area, the event behavior of the device under test is modeled to generate safe behavior event rules and determine the initial path planning strategy. The optimal path strategy is extracted from the initial path planning strategy using the gradient descent method.
[0008] Based on the above technical solutions, preferably, the airborne operating equipment includes multiple flight altitude layers, and the ground-based operating equipment and the waterborne operating equipment include multiple parallel travel layers. The motion path of the device under test is layered by a viscosity correction function corresponding to each altitude layer or travel layer; the viscosity correction function is related to the aggression factor.
[0009] Based on the above technical solutions, preferably, the step of determining the state transition matrix of the motion state of the device under test over time, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, includes: Based on the motion performance parameters and layered driving strategy of the device under test, and combined with discrete event simulation and dynamics principles, the number of transitions of the device under test from the first state to the second state is counted; the first state is the state corresponding to the second state at the previous moment. The state transition probability of the device under test is determined based on the proportion of the number of transitions in the total number. The state transition matrix of the device under test is determined based on the state transition probabilities.
[0010] Based on the above technical solutions, preferably, the step of fusing the real-time acquired multi-source detection data to obtain multi-source fused data includes: For airborne equipment, the extended Kalman filter algorithm is used to fuse the multi-source detection data in combination with the characteristics of the sensors and the noise model; For ground-based devices, deep learning algorithms are used to extract and match features from the multi-source detection data and complete data fusion. For equipment operating on water, a Bayesian network method is used to fuse the multi-source detection data.
[0011] Based on the above technical solutions, preferably, after inputting the multi-source fusion data into the target prediction model to obtain the predicted motion trajectory of the device under test, the method further includes: Risk warnings are issued based on the predicted motion trajectory and preset safety distance thresholds. A Bayesian network is used to determine the risk probability, and the warning level is determined according to the threshold range in which the risk probability falls.
[0012] Furthermore, a second aspect of the present invention provides a field test safety early warning system based on simulation extrapolation, comprising: a parameter acquisition module, a model correction module, and a trajectory prediction module; wherein, The parameter acquisition module is configured to acquire the model parameters and application scenario parameters of the device under test and create an initial simulation model. The model correction module is configured to determine the state transition matrix of the motion state of the device under test over time based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, and to correct the initial simulation model using the state transition matrix to obtain the target prediction model; the motion performance parameters include velocity parameters, acceleration parameters, temperature parameters, pressure parameters, and displacement parameters; The trajectory prediction module is configured to fuse real-time multi-source detection data to obtain multi-source fused data, and input the multi-source fused data into the target prediction model to obtain the predicted motion trajectory of the device under test.
[0013] More preferably, a third aspect of the present invention provides an electronic device, including a processor and a memory; the memory stores a computer program, wherein the computer program, when executed by the processor, implements the field test safety early warning method based on simulation extrapolation described in the first aspect.
[0014] More preferably, in a fourth aspect of the present invention, a computer storage medium is provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the field test safety early warning method based on simulation extrapolation described in the first aspect.
[0015] The field test safety early warning method and system based on simulation extrapolation of the present invention has the following advantages over the prior art: 1. By analyzing the motion performance parameters and hierarchical driving strategies of the device under test (DUT) in various application scenarios, and combining discrete event simulation and dynamic principles, the state transition matrix of the DUT's motion state over time is determined. This state transition matrix is then used to correct the initial simulation model, reflecting the device's true behavior patterns in different scenarios and avoiding the subjectivity of human assumptions in traditional models. Dividing the driving strategy into hierarchical levels allows the model to dynamically adjust state transition rules for different scenarios, improving adaptability to changing environments, avoiding overfitting to a single scenario, and enhancing the generalization ability of the prediction model.
[0016] 2. The equipment under test is divided into air-operated equipment, ground-operated equipment, and water-operated equipment, and different state transition equations are determined for each type to adapt to the dynamic characteristics of different physical environments. This enhances the environmental adaptability of the prediction model and improves its accuracy.
[0017] 3. Based on the simulation extrapolation algorithm, the safety early warning strategy was optimized, realizing the dynamic assessment and early warning of safety risks in field tests. This provided test personnel with intuitive safety early warning information, effectively solving the problem of the lack of dynamic assessment and early warning capabilities in existing technologies, and improving the safety of field tests. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 A flowchart illustrating a field test safety early warning method based on simulation extrapolation provided in an embodiment of the present invention; Figure 2A schematic diagram of a field test safety early warning system based on simulation extrapolation provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation
[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.
[0021] In some embodiments, such as Figure 1 As shown, Figure 1 This invention provides a flowchart illustrating a field test safety early warning method based on simulation extrapolation, as an embodiment of the present invention. The field test safety early warning method based on simulation extrapolation provided by the present invention includes: S110: Obtain the model parameters and application scenario parameters of the device under test, and create an initial simulation model.
[0022] In this embodiment, various relevant parameter information is collected for different types of application scenarios and equipment characteristics. For example, professional modeling tools and algorithms can be used to accurately model different scenarios and equipment based on the collected data. During the modeling process, methods such as discrete event simulation and system dynamics are used to model complex test scenarios. The equipment is modeled based on the physical model and operating principle of the equipment, combined with mathematical models and computer simulation technology, to obtain an initial simulation model.
[0023] In one example, establishing a high-precision simulation model library for airborne operational equipment involves collecting detailed parameters of numerous airborne operational equipment, such as the mass, wingspan, engine power, maximum flight speed, and range of different models. It also involves analyzing the flight performance data of different airborne operational equipment under various weather conditions (such as wind direction, wind speed, temperature, and humidity). Here, the airborne operational equipment can be a UAV (Unmanned Aerial Vehicle).
[0024] In another example, the establishment of a high-precision simulation model library for ground-based vehicles involves collecting a large amount of data on the power systems of these vehicles, such as motor torque, battery range, chassis and suspension system parameters, sensor accuracy and field of view, etc.; and analyzing the impact of different road conditions (such as rugged mountain roads, highways, and urban roads) on the operation of the ground-based vehicles. Here, the ground-based vehicle can be a UGV.
[0025] In another example, the establishment of a high-precision simulation model library for waterborne equipment involves collecting data on the equipment's dimensions, draft, displacement, and the impact of different hydrological conditions (such as water flow velocity, water temperature, and water quality) and marine meteorological conditions (such as wave size and direction) on the navigation of the waterborne equipment. Here, the waterborne equipment can be a USV (Unmanned Aerial Vehicle).
[0026] S120, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, determine the state transition matrix of the motion state of the device under test as a function of time, and use the state transition matrix to correct the initial simulation model to obtain the target prediction model; the motion performance parameters include velocity parameters, acceleration parameters, temperature parameters, pressure parameters and displacement parameters.
[0027] In some embodiments, the device under test includes an airborne device, a ground-based device, and a waterborne device; S120, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, and combining discrete event simulation and dynamic principles, the state transition matrix of the motion state of the device under test over time is determined, including: Based on the principles of fluid mechanics, Newton's second law and aerodynamics, the flight state of airborne equipment is simulated, and the first state transition matrix of the motion state of airborne equipment as a function of time is determined. Based on vehicle dynamics models combined with motion control theory and machine learning algorithms, the driving state of ground-based equipment is simulated, and the second state transition matrix of the motion state of ground-based equipment as it changes over time is determined. Based on fluid mechanics combined with the momentum theorem and the law of conservation of energy, the navigation state of waterborne equipment is simulated, and the third state transition matrix of the motion state of waterborne equipment as a function of time is determined.
[0028] In this embodiment, aerodynamic principles and fluid dynamics equations are used to simulate the dynamic response of rotors / wings in complex airflows such as turbulence and gusts. This provides physical constraints for the operation of ground-based equipment, avoids the extrapolation failure of pure data models, and addresses the energy conversion relationships between propulsion, wave interference, and fluid resistance involved in water-based equipment. All three types of state transition matrices use the same mathematical form, facilitating the construction of a cross-scenario joint simulation platform. By sharing state variables, such as position and velocity, spatiotemporal alignment is achieved.
[0029] In one optional embodiment, a computational fluid dynamics-based modeling method is employed, combining classic Newton's second law and aerodynamic formulas to accurately simulate the flight state of airborne equipment. Simultaneously, methods such as discrete event simulation are used to simulate events during coordinated flight and mission execution. When calibrating and validating the constructed model, actual flight test data is used to compare the model's predicted flight trajectory and attitude changes, continuously adjusting model parameters to ensure accuracy. The calibrated models are categorized and stored in a high-precision simulation model library, such as according to different mission purposes (reconnaissance, transportation, etc.) and different flight performance characteristics (long endurance, high speed, etc.). A dynamic model is used, combined with motion control theory and machine learning algorithms, to construct a driving model for ground-based equipment. For example, PID control algorithms are used to adjust the speed and steering angle of the ground-based equipment. Similarly, actual road tests are conducted to collect driving data of the ground-based equipment under different road conditions for model calibration and validation. Finally, the models are categorized and stored in the model library according to different application scenarios (logistics delivery, intelligent driving testing, etc.) and drive types (electric, fuel, etc.). By utilizing the principles of fluid mechanics and combining the momentum theorem and the law of conservation of energy, a navigation model for waterborne equipment is constructed. After calibrating and verifying the model using actual waterborne test data, the model is classified and stored according to the type of operation (marine monitoring, water rescue, etc.) and the structure of the equipment (single body, multi-body, etc.).
[0030] In some embodiments, S120, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, and combining discrete event simulation and dynamic principles, the state transition matrix of the motion state of the device under test over time is determined, including: The application scenario parameters are characterized to establish a three-dimensional occupancy matrix; the three-dimensional occupancy matrix includes obstacle areas and free space; Based on the Euclidean distance from the device under test to the nearest obstacle area, the event behavior of the device under test is modeled to generate safe behavior event rules and determine the initial path planning strategy. The optimal path strategy is extracted from the initial path planning strategy using the gradient descent method.
[0031] In this embodiment, the environment is characterized by establishing a 3D occupancy matrix with a 150×150×50 grid (3m resolution) and defining obstacle regions. With free space , This indicates that the entire space consists of obstacle areas and free space, and simultaneously defines... , This means that within the obstacle area, the propagation speed is 0, i.e., passage is impossible. Here, p represents a point in space. It is a propagation velocity function related to point p; A two-stage computation method is used to model the distributed node event behavior of the device under test and generate safe behavioral event rules. The first stage calculates the obstacle distance field. ,in, represents the Euclidean distance from point p to the nearest obstacle; this formula measures how close point p is to the obstacle. q is the obstacle region. The point in the middle, This is the Euclidean distance from point p to point q. By taking the minimum value, we can obtain the distance from point p to the nearest obstacle; then, we normalize it into a viscosity matrix. , This represents the viscosity value at point p, reflecting the "viscosity" of that point in event behavior, such as path planning. The larger the value, the farther away from the obstacle, the faster the propagation speed, and the easier it is to pass through during path planning. yes The maximum value in the entire space is used for normalization. The values are mapped to the [0,1] interval; the second stage performs path planning based on the viscosity matrix and solves the equation. ,in, This represents the time required for an object to propagate from its starting point to point p. yes The gradient, which represents the rate of change of the time field at point p, has a magnitude of . Related to the reciprocal of the viscosity matrix, by solving this equation, we can obtain the propagation time of the object in space, and then extract the optimal path strategy using the gradient descent method.
[0032] In some embodiments, the air-based operating equipment includes multiple flight altitude layers, and the ground-based and water-based operating equipment includes multiple parallel travel layers. The motion path of the device under test is layered by a viscosity correction function corresponding to each altitude layer or travel layer; the viscosity correction function is related to the aggression factor.
[0033] In this example, through the viscosity correction function Forced path layering; here, The coordinates of point p on the target layer are... These are the coordinates of the target layer. It is an attack factor used to control the constraint strength on the target layer. The larger the value, the stronger the constraint on the UAV to maintain at the target altitude layer. 0.005 and 0.01 are used to adjust the parameters of the viscosity correction function so that the calculation results are more in line with actual needs.
[0034] In some embodiments, S120, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, and combining discrete event simulation and dynamic principles, the state transition matrix of the motion state of the device under test over time is determined, including: Based on the motion performance parameters and layered driving strategy of the device under test, combined with discrete event simulation and dynamic principles, the number of transitions from the first state to the second state of the device under test is counted; the first state is the state corresponding to the second state at the previous moment. The state transition probability of the device under test is determined based on the proportion of the number of transitions in the total number. The state transition matrix of the device under test is determined based on the state transition probability.
[0035] In an optional embodiment, a state vector is defined for the motion performance parameters: , The displacement at the current moment, in meters, is collected by a displacement sensor. Velocity at the current moment, in units Data is collected by a speed sensor; Acceleration at the current moment, unit It is obtained through velocity data differential calculation or acceleration sensor; Temperature at the current moment, in units The temperature data is collected by a temperature sensor. Due to the pressure at present, the unit The data is collected by a pressure sensor; the state transition equation of the device under test satisfies... , where the state transition matrix Describe the recursive relationship of states over time: ; The time interval (s) between the current moment and the previous moment is determined by the sensor sampling period. The first three rows of the matrix correspond to the recursion of the motion state (displacement, velocity, acceleration), and the last two rows correspond to the assumption of constant and slowly changing temperature and pressure.
[0036] S130 fuses the real-time acquired multi-source detection data to obtain multi-source fused data, and inputs the multi-source fused data into the target prediction model to obtain the predicted motion trajectory of the device under test.
[0037] In some embodiments, multi-source detection data acquired in real time are fused to obtain multi-source fused data, including: For airborne equipment, the extended Kalman filter algorithm is used to fuse the multi-source detection data in combination with the characteristics of the sensors and the noise model; For ground-based devices, deep learning algorithms are used to extract and match features from multi-source detection data and complete data fusion. For equipment operating on water, a Bayesian network method is used to fuse multi-source detection data.
[0038] In this embodiment, the focus of data fusion differs depending on the device under test. For data monitoring of airborne equipment, an integrated airborne sensor suite is used, including an inertial measurement unit (IMU), a global positioning system (GPS) module, an optical flow sensor, and a barometric pressure sensor. These sensors collect real-time data on the UAV's acceleration, angular velocity, position, flight speed, altitude, etc. Wireless communication modules, such as wireless links based on the 802.11p protocol or satellite communication equipment, are used to transmit the data quickly and stably to the ground data processing center. During the data fusion stage, the Extended Kalman Filter (EKF) algorithm is used to perform fusion estimation of multiple sensor data based on sensor characteristics and noise models, thereby improving the accuracy and reliability of the data. For data monitoring of ground-based equipment, multiple sensors, such as lidar, millimeter-wave radar, and cameras, are installed on the ground-based equipment to acquire real-time information on obstacles, speed, road boundaries, etc., around the equipment. Vehicle-to-everything (V2X) technology or Dedicated Short Range Communication (DSRC) is used to transmit the data to the cloud data processing center. Data fusion employs multi-sensor fusion algorithms, combining feature-level fusion and decision-level fusion to enhance the perception of environmental information. For example, deep learning algorithms are used to extract and match features from camera images and LiDAR point cloud data to reduce erroneous judgments. For data monitoring of waterborne equipment, various hydrological sensors (such as current meters, thermometers, and water quality analyzers), Automatic Identification System (AIS) transponders, and radar are used to collect hydrological data, information from surrounding equipment, and its own navigation data in real time. Data is transmitted to the shore-based control center via satellite communication links or VHF / UHF communication equipment. Bayesian network methods are used for data fusion, combining prior knowledge and real-time sensor data to comprehensively assess the navigation status and environmental conditions of the waterborne equipment.
[0039] In some embodiments, after inputting multi-source fused data into a target prediction model to obtain the predicted motion trajectory of the device under test, the method further includes: Risk warnings are issued based on predicted motion trajectories and preset safe distance thresholds. A Bayesian network is used to determine the risk probability, and the warning level is determined based on the threshold range in which the risk probability falls.
[0040] In this embodiment, a security early warning software platform is constructed. Based on a simulation extrapolation algorithm, security risks are dynamically assessed. When the risk value exceeds a threshold, a security early warning is issued. Specifically, this includes: Based on the selected development framework and programming language, the basic architecture of the software platform is constructed using Java and Python combined with a relevant database management system to support the storage, management, and efficient processing of massive amounts of data. A validated simulation extrapolation algorithm is embedded into the software platform. This algorithm dynamically assesses the safety status of the current field test based on the predicted motion trajectory obtained in the above embodiments and the data transmitted by the real-time data monitoring and fusion system. The assessment process can be carried out through a series of mathematical models and rule bases, setting different risk indicator weights, comparing the collected data with the risk indicators, and developing corresponding early warning modules based on pre-set risk thresholds. When the calculated risk value exceeds the set threshold, the early warning module is triggered in a timely manner, sending clear and intuitive safety warning information to the test personnel in various forms (such as pop-ups, voice alarms, SMS messages, etc.) to remind them to take appropriate measures.
[0041] For the application module of airborne equipment, the software platform uses deep learning algorithms to build trajectory prediction and fault diagnosis models. By learning from a large amount of historical flight data, it predicts the future flight trajectory of the airborne equipment and promptly identifies potential faults. Combining the predicted trajectory, the platform adjusts the flight plan of the airborne equipment in advance when encountering complex weather conditions or mission changes to ensure flight safety. For example, if it predicts that the airborne equipment may be unable to complete its mission due to insufficient battery power, it issues a timely warning and generates a return-to-base instruction. For the application module of ground-based equipment, reinforcement learning algorithms from machine learning are introduced to dynamically evaluate the driving performance of ground-based equipment under different road conditions. By continuously interacting with the environment, it learns the optimal driving strategy. Combining the predicted trajectory and preset safety distance thresholds, it estimates safety risks in different scenarios, such as predicting potential traffic congestion or sudden traffic accidents at intersections ahead, adjusting the driving route of the ground-based equipment in advance, and triggering a timely warning mechanism. For example, when an obstacle is detected on the road ahead, the safety warning software platform issues a warning and controls the ground-based equipment to slow down or avoid it. For the application modules of waterborne equipment, big data analysis and intelligent algorithms are used to monitor and provide early warnings of marine environmental conditions and the navigation status of waterborne equipment in real time. For example, when severe sea conditions are predicted by analyzing historical marine meteorological data and real-time marine environmental data, the navigation status of waterborne equipment under severe sea conditions is simulated by combining a high-precision simulation model library, and instructions are sent to unmanned vessels in advance to guide them to a safe area. Intelligent algorithms are used to evaluate the status of the mechanical and electronic systems of waterborne equipment, promptly detect potential equipment failures, issue early warnings, and provide relevant maintenance suggestions.
[0042] In some embodiments, the process of constructing simulation extrapolation and dynamic early warning includes first establishing a risk prediction model, defining a safety threshold for potential collisions between distributed nodes of the device under test during the task. and The risk prediction model is defined as follows: , ,in, It is a safety distance threshold defined in the horizontal direction. This is a safety distance threshold defined in the vertical direction, required for equipment operating in the air, but not for equipment operating on land or water. t represents time. It is the start time. It is the predicted time window. and These are the position coordinates of drones i and j at time t. It is the distance in the horizontal direction. The distance in the vertical direction is considered to be the potential risk of conflict when the distance between airborne equipment and the horizontal and vertical directions is less than the corresponding safety threshold within the prediction time window. The distance between land-based or water-based equipment and the horizontal direction is considered to be less than the corresponding safety threshold. Secondly, the implementation of the extrapolation algorithm is established, and a dynamic Bayesian network is built. Each node contains a large number of variables, including environmental conditions, equipment parameters, and historical accident data, defining the probability of risk. ; It represents the probability of a risk occurring given observed evidence E, where E is the set of evidence including information such as environmental conditions and equipment parameters, and N is the number of samples taken. It is an indicator function, indicating when a risk occurs in the i-th sample. hour, Otherwise, the value is 0. The risk probability is estimated by statistically analyzing the results of multiple samplings.
[0043] Different warning levels and response measures are set according to the probability of risk: Level I: Immediately cease the test and initiate emergency evacuation; Level II: Limit equipment operating parameters; manual confirmation required; Level III: Sound and light warnings are issued, and alternative routes are pushed out.
[0044] In some embodiments, please refer to Figure 2 , Figure 2This is a schematic diagram of a field test safety early warning system based on simulation extrapolation provided in an embodiment of the present invention. The present invention provides a field test safety early warning system 200 based on simulation extrapolation, comprising: a parameter acquisition module 210, a model correction module 220, and a trajectory prediction module 230; wherein, The parameter acquisition module 210 is configured to acquire the model parameters and application scenario parameters of the device under test and create an initial simulation model. The model correction module 220 is configured to determine the state transition matrix of the motion state of the device under test as a function of time, based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, and to correct the initial simulation model using the state transition matrix to obtain the target prediction model; the motion performance parameters include velocity parameters, acceleration parameters, temperature parameters, pressure parameters and displacement parameters; The trajectory prediction module 230 is configured to fuse real-time multi-source detection data to obtain multi-source fused data, and input the multi-source fused data into the target prediction model to obtain the predicted motion trajectory of the device under test.
[0045] In some embodiments, the device under test includes an airborne device, a ground-based device, and a waterborne device; the model correction module 220 is specifically configured as follows: Based on the principles of fluid mechanics, Newton's second law and aerodynamics, the flight state of airborne equipment is simulated, and the first state transition matrix of the motion state of airborne equipment as a function of time is determined. Based on vehicle dynamics models combined with motion control theory and machine learning algorithms, the driving state of ground-based equipment is simulated, and the second state transition matrix of the motion state of ground-based equipment as it changes over time is determined. Based on fluid mechanics combined with the momentum theorem and the law of conservation of energy, the navigation state of waterborne equipment is simulated, and the third state transition matrix of the motion state of waterborne equipment as a function of time is determined.
[0046] In some embodiments, the model correction module 220 is specifically configured as follows: The application scenario parameters are characterized to establish a three-dimensional occupancy matrix; the three-dimensional occupancy matrix includes obstacle areas and free space; Based on the Euclidean distance from the device under test to the nearest obstacle area, the event behavior of the device under test is modeled to generate safe behavior event rules and determine the initial path planning strategy. The optimal path strategy is extracted from the initial path planning strategy using the gradient descent method.
[0047] In some embodiments, the air-based operating equipment includes multiple flight altitude layers, and the ground-based and water-based operating equipment includes multiple parallel travel layers. The motion path of the device under test is layered by a viscosity correction function corresponding to each altitude layer or travel layer; the viscosity correction function is related to the aggression factor.
[0048] In some embodiments, the model correction module 220 is specifically configured as follows: Based on the motion performance parameters and layered driving strategy of the device under test, combined with discrete event simulation and dynamic principles, the number of transitions from the first state to the second state of the device under test is counted; the first state is the state corresponding to the second state at the previous moment. The state transition probability of the device under test is determined based on the proportion of the number of transitions in the total number. The state transition matrix of the device under test is determined based on the state transition probability.
[0049] In some embodiments, the trajectory prediction module 230 is specifically configured as follows: For airborne equipment, the extended Kalman filter algorithm is used to fuse the multi-source detection data in combination with the characteristics of the sensors and the noise model; For ground-based devices, deep learning algorithms are used to extract and match features from multi-source detection data and complete data fusion. For equipment operating on water, a Bayesian network method is used to fuse multi-source detection data.
[0050] In some embodiments, a risk warning module is also included, which is specifically configured as follows: Risk warnings are issued based on predicted motion trajectories and preset safe distance thresholds. A Bayesian network is used to determine the risk probability, and the warning level is determined based on the threshold range in which the risk probability falls.
[0051] It should be noted that the field test safety early warning system based on simulation extrapolation provided in this application embodiment and the field test safety early warning method based on simulation extrapolation provided in this application embodiment are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned field test safety early warning method based on simulation extrapolation, and the repeated parts will not be described again.
[0052] In some embodiments, please refer to Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device 300 provided in this embodiment includes a processor 310 and a memory 320; the memory 320 stores a computer program, wherein the computer program, when executed by the processor, implements the aforementioned field test safety early warning method based on simulation extrapolation.
[0053] Specifically, processor 310 may include, for example, a general-purpose microprocessor, an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. Processor 310 may also include onboard memory for caching purposes. Processor 310 may be a single processing unit or multiple processing units for performing different actions of the method flow according to embodiments of this application.
[0054] The memory 320 may be any medium capable of containing, storing, transmitting, propagating, or transmitting instructions. For example, the memory 320 may include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, apparatuses, or propagation media. Specific examples of the memory 320 include: magnetic storage devices such as magnetic tape or hard disk drives (HDDs); optical storage devices such as optical discs (CD-ROMs); and may also be random access memory (RAM) or flash memory; and / or wired / wireless communication links.
[0055] This application also provides a computer-readable medium storing a computer program thereon, which, when executed by a processor, implements the above-described field test safety early warning method based on simulation extrapolation. This computer-readable medium may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The aforementioned computer-readable medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.
[0056] According to embodiments of this application, a computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wired, optical fiber, radio frequency signals, etc., or any suitable combination thereof.
[0057] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments and / or claims of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application. Therefore, the scope of this application should not be limited to the above embodiments, but should be defined not only by the appended claims, but also by their equivalents. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the protection scope of this invention.
Claims
1. A field test safety early warning method based on simulation extrapolation, characterized in that, include: Obtain the model parameters and application scenario parameters of the device under test, and create an initial simulation model; Based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, and combined with discrete event simulation and dynamic principles, the state transition matrix of the motion state of the device under test over time is determined, and the initial simulation model is corrected using the state transition matrix to obtain the target prediction model; the motion performance parameters include velocity parameters, acceleration parameters, temperature parameters, pressure parameters, and displacement parameters; The multi-source detection data acquired in real time are fused to obtain multi-source fused data, which is then input into the target prediction model to obtain the predicted motion trajectory of the device under test.
2. The field test safety early warning method based on simulation extrapolation as described in claim 1, characterized in that, The device under test includes aerial, ground-based, and water-based devices; the determination of the state transition matrix of the device under test's motion state over time, based on its motion performance parameters and layered driving strategies in various application scenarios, combined with discrete event simulation and dynamic principles, includes: Based on the principles of fluid mechanics, Newton's second law and aerodynamics, the flight state of the aerial operating equipment is simulated to determine the first state transition matrix of the motion state of the aerial operating equipment as a function of time. Based on the vehicle dynamics model combined with motion control theory and machine learning algorithm, the driving state of the ground-based equipment is simulated, and the second state transition matrix of the motion state of the ground-based equipment changing with time is determined. Based on fluid mechanics combined with the momentum theorem and the law of conservation of energy, the navigation state of the waterborne equipment is simulated, and the third state transition matrix of the motion state of the waterborne equipment as a function of time is determined.
3. The field test safety early warning method based on simulation extrapolation as described in claim 1, characterized in that, The process of determining the state transition matrix of the device under test (DUT) over time based on its motion performance parameters and layered driving strategies in various application scenarios, combined with discrete event simulation and dynamics principles, includes: The application scenario parameters are characterized to establish a three-dimensional occupancy matrix; the three-dimensional occupancy matrix includes obstacle areas and free space; Based on the Euclidean distance from the device under test to the nearest obstacle area, the event behavior of the device under test is modeled to generate safe behavior event rules and determine the initial path planning strategy. The optimal path strategy is extracted from the initial path planning strategy using the gradient descent method.
4. The field test safety early warning method based on simulation extrapolation as described in claim 2, characterized in that, The process of determining the state transition matrix of the device under test (DUT) over time based on its motion performance parameters and layered driving strategies in various application scenarios, combined with discrete event simulation and dynamics principles, includes: The aerial operating equipment includes multiple flight altitude layers, and the ground operating equipment and the water operating equipment include multiple parallel travel layers. The motion path of the device under test is layered by a viscosity correction function corresponding to each altitude layer or travel layer; the viscosity correction function is related to the aggression factor.
5. The field test safety early warning method based on simulation extrapolation as described in claim 1, characterized in that, The process of determining the state transition matrix of the device under test (DUT) over time based on its motion performance parameters and layered driving strategies in various application scenarios, combined with discrete event simulation and dynamics principles, includes: Based on the motion performance parameters and layered driving strategy of the device under test, and combined with discrete event simulation and dynamics principles, the number of transitions of the device under test from the first state to the second state is counted; the first state is the state corresponding to the second state at the previous moment. The state transition probability of the device under test is determined based on the proportion of the number of transitions in the total number. The state transition matrix of the device under test is determined based on the state transition probabilities.
6. The field test safety early warning method based on simulation extrapolation as described in claim 1, characterized in that, The process of fusing real-time acquired multi-source detection data to obtain multi-source fused data includes: For airborne equipment, the extended Kalman filter algorithm is used to fuse the multi-source detection data in combination with the characteristics of the sensors and the noise model; For ground-based devices, deep learning algorithms are used to extract and match features from the multi-source detection data and complete data fusion. For equipment operating on water, a Bayesian network method is used to fuse the multi-source detection data.
7. The field test safety early warning method based on simulation extrapolation as described in claim 1, characterized in that, After inputting the multi-source fusion data into the target prediction model to obtain the predicted motion trajectory of the device under test, the method further includes: Risk warnings are issued based on the predicted motion trajectory and preset safety distance thresholds. A Bayesian network is used to determine the risk probability, and the warning level is determined according to the threshold range in which the risk probability falls.
8. A field test safety early warning system based on simulation extrapolation, characterized in that, include: The module consists of a parameter acquisition module, a model correction module, and a trajectory prediction module; among which, The parameter acquisition module is configured to acquire the model parameters and application scenario parameters of the device under test and create an initial simulation model. The model correction module is configured to determine the state transition matrix of the motion state of the device under test over time based on the motion performance parameters and layered driving strategies of the device under test in various application scenarios, combined with discrete event simulation and dynamic principles, and to correct the initial simulation model using the state transition matrix to obtain the target prediction model; the motion performance parameters include velocity parameters, acceleration parameters, temperature parameters, pressure parameters, and displacement parameters; The trajectory prediction module is configured to fuse real-time multi-source detection data to obtain multi-source fused data, and input the multi-source fused data into the target prediction model to obtain the predicted motion trajectory of the device under test.
9. An electronic device comprising a processor and a memory; said memory having a storage for a computer program, wherein, When the computer program is executed by the processor, it implements the field test safety early warning method based on simulation extrapolation as described in any one of claims 1 to 7.
10. A computer storage medium, characterized in that, It stores a computer program, wherein when the computer program is executed by a processor, it implements the field test safety early warning method based on simulation extrapolation as described in any one of claims 1 to 7.