Unmanned aerial vehicle emergency control management method and system based on digital twinning

Abstract

The invention discloses an unmanned aerial vehicle emergency control management method and system based on digital twinning, and belongs to the technical field of unmanned aerial vehicle control. The system comprises an unmanned aerial vehicle data acquisition module which acquires image data in an emergency area and parameter records of an unmanned aerial vehicle; the virtual scene construction module is used for constructing a virtual scene by using a digital twinning technology, and mapping and dividing a three-dimensional grid in real time; the emergency route analysis module is used for planning a feasible route for each unmanned aerial vehicle in the virtual scene and screening out key routes; the execution scheme planning module is used for analyzing each scheme according to the key route planning scheme so as to set a communication scheme; and the operation visualization module is used for supervising each unmanned aerial vehicle to fly according to the communication scheme and displaying a virtual scene in real time through a visualization page of the command center. According to the invention, through digital twin-driven multi-source sensing fusion, dynamic grid adaptation and nonlinear anti-interference path planning, the environment modeling precision and communication link reliability of the unmanned aerial vehicle group in a disaster environment are improved.

Description

Technical Field

[0001] The present invention relates to the field of drone control technology, and specifically to a drone emergency control management method and system based on digital twins. Background Art

[0002] In emergencies like industrial accidents, drones, due to their flexibility and rapid response capabilities, have become a critical tool for emergency rescue. However, disaster environments are often accompanied by complex physical disturbances such as smoke, high temperatures, and significant communication signal attenuation. Furthermore, the dynamic evolution of disaster situations further exacerbates regional risk uncertainty. This requires drone swarms to possess dynamic environmental perception, interference-resistant path planning, and efficient coordinated control capabilities to ensure the reliable execution of critical missions.

[0003] While existing technologies have attempted to incorporate digital twins and multi-machine collaboration, they still face technical bottlenecks in addressing nonlinear interference and dynamic coupling scenarios. Traditional solutions rely on single sensors for local environmental detection, resulting in significant deviations in long-distance smoke concentration and temperature predictions. Furthermore, digital twin modeling employs static meshing, which cannot adapt to the nonlinear spatial attenuation characteristics of communication signals. In path planning, the impact of environmental parameters on signal strength is often simplified into linear models, ignoring the scattering enhancement effect of high-concentration smoke or the deterioration caused by high-temperature thermal noise, resulting in insufficient route reliability. Furthermore, inter-UAV collaboration relies on fixed relay nodes, lacking dynamic self-organization capabilities and susceptible to communication interruptions in weak signal areas. The system suffers from poor real-time performance and struggles to respond promptly to sudden disaster events, leading to delayed visualization and decision-making, severely hampering emergency response effectiveness. Therefore, current research urgently needs to develop a digital twin framework that integrates dynamic environmental inversion, adaptive communication optimization, and self-organizing collaboration to achieve precise control of UAV swarms and reliable data transmission across the entire link in high-risk disaster areas. Summary of the Invention

[0004] The purpose of the present invention is to provide a UAV emergency control management method and system based on digital twins to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides a UAV emergency control management method based on digital twins, comprising:

[0006] S100: Real-time collection of image data within the emergency area and parameter records of each UAV.

[0007] S200, using digital twin technology to build virtual scenes based on image data, and to map and divide the three-dimensional grid in real time.

[0008] S300: Planning feasible routes in the virtual scene based on the drone parameter information, and selecting key routes.

[0009] S400: Analyze each plan based on the key route planning plan to set a communication plan.

[0010] S500 and each drone fly according to the communication plan, and display the virtual scene in real time through the visualization page.

[0011] The emergency area refers to the specific geographical area where the disaster occurs. Image data refers to video images captured by various drones within the emergency area.

[0012] Parameter records include environmental parameters, operational parameters, and command parameters. Environmental parameters refer to the smoke concentration and temperature of the surrounding environment collected by the drone at different times. Operational parameters refer to the drone's spatial position, control signal strength, and cooperative signal strength at different times. Command parameters include the flight objectives issued by the control center.

[0013] Environmental parameters can be collected in two ways: one is to collect smoke concentration and temperature at close range around the drone using smoke and temperature sensors. The other is to predict smoke concentration and temperature at long distances around the drone through real-scene image analysis and thermal imaging analysis.

[0014] Short distance refers to the maximum effective distance that the sensor can detect, and long distance refers to a distance that exceeds the maximum effective distance that the sensor can detect and is less than or equal to the maximum effective distance of real-scene image analysis or thermal imaging analysis.

[0015] The prediction of smoke concentration usually requires a correlation analysis based on the smoke transmittance in the corresponding area of ​​the real-life image and the actual smoke concentration collected by sensors in the corresponding area by other drones at the same time, to obtain a relationship formula between the two, and thus predict and analyze the smoke concentration based on the real-life image.

[0016] Spatial position refers to the real-time three-dimensional coordinates of the drone in the air. Control signal strength refers to the signal strength between the drone and the command center, and collaborative signal strength refers to the signal strength between the drone and other drones.

[0017] By integrating sensor detection and image analysis technologies, comprehensive environmental parameter collection is achieved. UAV spatiotemporal state parameters and flight instructions are simultaneously integrated to build a multidimensional data pool. This provides high-precision, multi-dimensional, real-time data support for subsequent virtual scene construction and path planning, enhancing robust perception of complex and catastrophic environments.

[0018] S200 includes:

[0019] S201: Obtain image data and position parameters collected by each drone in the emergency area, extract common feature points in different images for matching, and mark overlapping areas.

[0020] S202. Use a multi-perspective 3D reconstruction algorithm to combine images from multiple perspectives to build a virtual scene, map entities in the image to the virtual scene through digital twin technology, and update the mapped entity objects in real time.

[0021] S203: Obtain parameter records of each UAV, analyze changes in control signal strength at different times, and draw a line graph for each UAV with time as the horizontal axis and control signal strength as the vertical axis.

[0022] S204 , setting a duration q, dividing the sampling intervals on the line graph according to the duration q, and marking the sampling intervals where the extreme difference in control signal strength is greater than a threshold w.

[0023] S205 , analyzing the spatial positions within the time period corresponding to each marked sampling interval, and using the Euclidean distance algorithm to select the two spatial positions with the farthest distance for calculation, to obtain the spatial distance of the corresponding marked sampling interval.

[0024] S206 , selecting the minimum value among the spatial distances of all marked sampling intervals as the sensing distance of the corresponding UAV, and calculating the average sensing distance of all UAVs as the grid scale L.

[0025] S207. Divide the virtual scene into three-dimensional grid areas, each of which is a cube with an edge length of L.

[0026] A digital twin scene is dynamically constructed based on a multi-view 3D reconstruction algorithm, and adaptive grid scales are generated based on the attenuation characteristics of the drone's control signal strength. This ensures real-time synchronization between the virtual scene and the physical world, and the grid division matches the actual communication attenuation characteristics of the drone, providing a spatial reference unit for traffic zone screening and route planning.

[0027] S300 includes:

[0028] S301: A three-dimensional grid area in a virtual scene where no physical objects exist is used as a passage area, feasible routes are planned according to the current spatial position of the UAV and the flight destination, and the passage areas passed by each feasible route are marked.

[0029] S302: Analyze the parameter records of all drones, filter out the control signal strength, smoke concentration and temperature at different times when the spatial position is within the marked passage area, and fit the influence relationship expression of each marked passage area. Specifically, it includes:

[0030] S302-1. Establish a training set for each marked passage area and obtain the control signal strength, smoke concentration and temperature of each UAV when its spatial position is in the marked passage area.

[0031] S302-2. The control signal strength of the same UAV at the same time is used as the dependent variable, and the smoke concentration and temperature are used as independent variables and packaged into samples.

[0032] The control signal strength, smoke concentration, and temperature in each sample belong to the environmental and operating parameters collected simultaneously by a specific drone at a specific moment. Data in different samples may belong to different drones or be collected at different times.

[0033] S302-3. All samples are put into the training set of the marked passage area to which they belong, and a polynomial model is established. Each training set is input into the polynomial model for training, and the influence relationship expression of each marked passage area is obtained by fitting:

[0034] ;

[0035] Where, To control the signal strength, is the smoke concentration, is the temperature, is the baseline signal strength, is the linear attenuation coefficient of smoke concentration, is the temperature linear attenuation coefficient, is the secondary attenuation coefficient of smoke concentration, is the temperature quadratic attenuation coefficient, is the error coefficient.

[0036] Baseline signal strength It refers to the theoretical signal strength under normal conditions, which serves as a benchmark reference for signal strength attenuation.

[0037] Smoke concentration linear attenuation coefficient It refers to the exponential attenuation contribution of the increase in smoke concentration to the signal intensity, and is used to control the power-law attenuation rate of smoke concentration.

[0038] Temperature linear attenuation coefficient It refers to the exponential decay contribution of the temperature increase to the signal intensity, which is used to capture the linear effect of temperature on the dielectric constant.

[0039] Secondary attenuation coefficient of smoke concentration It refers to the nonlinear correction of the exponential decay by the square term of the smoke concentration, such as the particle aggregation effect or complex scattering mechanism, and is used to describe the accelerated decay at high smoke concentrations.

[0040] Temperature quadratic attenuation coefficient It refers to the direct attenuation of signal strength by the square of the temperature, such as the nonlinear degradation of device thermal noise, and is used to quantify the additional deterioration of device performance at high temperatures.

[0041] Error coefficient refers to random noise or measurement error not explained by the model and is used for the uncertainty of the statistical fit.

[0042] S303: Obtain the latest smoke concentration and temperature of each marked passage area, substitute them into the corresponding influence relationship expression, and calculate the predicted control signal strength.

[0043] To obtain the latest smoke concentration and temperature in a marked area, the drone does not need to be in the marked area to collect data. As long as the environmental parameter collection requirements are met, whether through sensors or image analysis methods, as long as the time meets the definition of the latest data, it can be used as the latest smoke concentration and temperature. The definition of the latest data is pre-defined by the management.

[0044] S304: Calculate the pass coefficient of the corresponding feasible route based on the predicted control signal strength of each marked pass area, and select the route with a pass coefficient greater than the coefficient threshold for each drone. The feasible routes are taken as key routes. Specifically including:

[0045] S304-1. Obtaining a feasible route All marked passage areas are arranged in the order of flight and the difference in the predicted control signal strength between adjacent marked passage areas is calculated. The standard deviation of all differences is calculated as the fluctuation coefficient. .

[0046] S304-2. Statistics of feasible routes Number of all marked passage areas , calculate the average value of all predicted control signal strengths as the reference signal strength , substitute into the formula to calculate the feasible route The traffic coefficient :

[0047] ;

[0048] Where, 、 and is a constant, is the average reference signal strength of all feasible routes, is the average fluctuation coefficient of all feasible routes, It is the average number of marked passage areas of all feasible routes.

[0049] S304-3, and so on, respectively calculate the traffic coefficient of each feasible route.

[0050] A polynomial model is used to predict the control signal strength in each traffic zone. Combined with a traffic coefficient calculation formula based on volatility, signal strength, and path complexity, this method eliminates high-interference routes. Dynamically quantifying the impact of environmental physical parameters on communication quality helps avoid high-risk routes, such as those in dense fog and high-temperature interference zones, and improves route feasibility.

[0051] S400 includes:

[0052] S401: Count the number of all drones, a, and multiply the number of key routes for all drones to obtain b. Create b plans, and place a key route belonging to a different drone in each plan.

[0053] All drones are paired together, and the collaborative signal strength of each pair of associated drones at different spatial locations is analyzed based on parameter records, and the collaborative relationship expression of each pair of associated drones is fitted. Specifically, it includes:

[0054] S401-1. Establish a training set for each pair of associated drones, calculate the spatial distance of each pair of associated drones at different spatial positions, and obtain the cooperative signal strength at different spatial distances.

[0055] S401-2. The collaborative signal strength at the same time is used as the dependent variable, and the spatial distance is used as the independent variable and packaged into samples.

[0056] S401-3. All samples are placed in the training set of the corresponding related drones, and a linear regression model is established. Each training set is input into the model for training, and the collaborative relationship expression of each related drone is obtained by fitting. The specific training content includes:

[0057] The independent variable of each sample in the training set is used as the input value of the linear regression model, and the difference between the output result of each sample and the dependent variable is used as the gap value. Set the error threshold and adjust the intercept and the regression coefficient Make the gap values ​​of all samples in the training set smaller than the error threshold, and obtain the expression after training:

[0058] ;

[0059] Where, is the dependent variable, is the independent variable.

[0060] S402, obtain the drone in solution i Key routes , set the standard flight speed, calculate the drone Reach key routes Next to the first marked passage Time , analysis time The marked passage areas where other drones in scheme i fly along their respective key routes .

[0061] S403, according to time Get off the drone Calculate the spatial distance with other drones in the marked passage area, and substitute the corresponding collaborative relationship expression to calculate the predicted collaborative signal strength . Analyzing drones In marked traffic areas The predicted control signal strength , and other drones in their respective marked passage areas The predicted control signal strength .

[0062] S404, Marking and Are greater than drone, select and The minimum value is used as the forwarding signal strength of the marking drone. The marking drone with the largest forwarding signal strength is selected as the drone In marked traffic areas The auxiliary object, the forwarding signal strength of the auxiliary object is used as the UAV In marked traffic areas New predictions control signal strength.

[0063] S405. Analyze key routes one by one Mark each traffic area and set a new prediction control signal strength. After the analysis is completed, recalculate the key routes Analyze the marked traffic areas of other key routes under plan i in turn and recalculate the traffic coefficients.

[0064] S406: After all key routes are analyzed, the traffic coefficients of all key routes are summed up to form the traffic index of solution i. The traffic index of each solution is calculated separately, and the solution with the largest traffic index is selected as the communication solution.

[0065] The collaborative relationship expression is used to predict the effectiveness of multi-drone collaboration, and a dynamic forwarding mechanism is used to iteratively optimize the traffic index of each path combination. This enables self-organizing collaboration and redundant communication link planning between drones, ensuring reliable data transmission in weak signal areas and improving the overall robustness of mission execution.

[0066] In S500, all drones fly along the key routes in the communication plan. When they arrive at the marked passage area with auxiliary objects, they transmit data to the auxiliary objects, which then forward it to the command center.

[0067] During flight, the traffic coefficients for each marked traffic zone are updated in real time, routes and auxiliary objects are dynamically adjusted, and environmental and operating parameters are recorded in real time and stored in the parameter log. The command center displays the virtual scene in real time through a visual interface.

[0068] The digital twin interface maps drone trajectories, environmental parameter heat maps, and signal strength changes in real time, combined with a dynamic path optimization mechanism. This empowers the command center with global situational awareness and rapid intervention capabilities, and improves the real-time and operability of emergency response through visual closed-loop feedback.

[0069] The UAV emergency control and management system based on digital twins includes a UAV data acquisition module, a virtual scene construction module, an emergency route analysis module, an execution plan planning module, and an operation visualization module.

[0070] The UAV data acquisition module is used to collect image data within the emergency area and record the parameters of the UAV.

[0071] The virtual scene construction module uses digital twin technology to build virtual scenes, map and divide three-dimensional grids in real time.

[0072] The emergency route analysis module is used to plan feasible routes for each drone in a virtual scene and screen out key routes.

[0073] The execution plan planning module is used to plan plans based on key routes, analyze each plan and set communication plans.

[0074] The operation visualization module is used to supervise each drone to fly according to the communication plan, and display the virtual scene in real time through the visualization page of the command center.

[0075] The drone data acquisition module uses drones equipped with smoke and temperature sensors and cameras to collect real-time video images of the emergency area and three types of parameters. This environmental parameter collection combines short-range sensor detection with long-range image prediction technology, and records operational data such as 3D coordinates and dual signal strength.

[0076] It provides multi-level environmental perception capabilities to compensate for the blind spots of single sensors. It also improves data robustness through redundant acquisition, laying a data foundation for virtual scene construction and path analysis.

[0077] The virtual scene construction module uses a multi-view 3D reconstruction algorithm to fuse images from multiple drone perspectives, updating the digital twin scene in real time. By analyzing the fluctuation range of control signal strength, the grid scale is dynamically calculated and divided into standard cube grids.

[0078] Achieve high-precision dynamic modeling of disaster environments, grid division to match the drone signal attenuation characteristics, and provide spatial computing units for traffic zone classification and path planning.

[0079] The emergency route analysis module combines traffic zone screening with a polynomial model to predict control signal strength. A traffic coefficient formula is designed to filter high-risk routes.

[0080] Quantify the comprehensive impact of environmental interference factors on communication quality, screen out flight paths with stable signals and reliable equipment, and reduce the risk of communication interruption during mission execution.

[0081] The execution plan planning module constructs the UAV collaborative relationship expression and dynamic assistance mechanism, and selects the communication plan that maximizes the collaborative efficiency by traversing b route combinations and iteratively updating the traffic index.

[0082] Optimize the collaborative communication strategy of drone swarms, compensate for the quality of weak signal areas through dynamic relay forwarding, and improve the overall success rate of multi-machine missions in complex environments.

[0083] The operation visualization module integrates functions such as parameter record update, path dynamic adjustment, and auxiliary object switching, and annotates drone trajectories, signal strength heat maps, and environmental parameter change curves in real time in the digital twin scene.

[0084] Providing a decision-making support visualization interface enables the command center to understand the overall situation and quickly intervene in abnormal conditions, enhancing the timeliness of emergency response.

[0085] Compared with the prior art, the present invention has the following beneficial effects:

[0086] Multi-source environmental perception and data fusion: Breaking through the limitations of single-sensor perception, through the dual complementarity of close-range measurement by smoke sensors and long-range prediction models based on thermal imaging / transmittance analysis, we build the ability to dynamically invert global smoke concentration and temperature, addressing the large deviation in long-range parameter estimation in existing technologies and enhancing the environmental mapping accuracy of digital twin models.

[0087] Communication perception-driven dynamic meshing: Abandoning the static meshing method, the communication attenuation characteristics are reversely calculated based on the fluctuations in the drone control signal strength, and the three-dimensional grid scale is automatically generated. This allows the virtual scene grid to match the actual signal propagation characteristics in real time, improving the spatial adaptability of path planning to electromagnetic interference.

[0088] Path prediction model resistant to nonlinear interference: To address the defect of existing technologies that simplify signal attenuation into a linear model, nonlinear factors of the quadratic term of smoke concentration and the square term of temperature are introduced to quantify the superimposed effect of high-concentration smoke scattering and high-temperature equipment degradation, significantly improving the prediction reliability of the path passability coefficient and avoiding the problem of misselection of high-risk routes.

[0089] Dynamic self-organizing collaborative relay mechanism: Abandoning the preset static relay strategy, a dynamic correlation model of spatial distance and collaborative signal strength between drones is constructed, and auxiliary nodes are selected in real time for signal relay forwarding, ensuring the self-repair capability of communication links in weak signal areas, and avoiding the problem of drone loss of connection caused by node failure in traditional solutions.

[0090] Closed-loop visualization and real-time correction capabilities: Integrating dynamic parameter updates, path replanning, and collaborative strategy tuning functions, the system provides real-time feedback on communication quality heat maps and environmental evolution trends through a virtual-real linkage visualization interface, providing closed-loop control support for emergency decision-making and breaking through the bottleneck of delayed response to sudden disaster situations caused by existing technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0092] Figure 1 It is a flow chart of the UAV emergency control management method based on digital twins of the present invention;

[0093] Figure 2 It is a structural diagram of the UAV emergency control and management system based on digital twins of the present invention. DETAILED DESCRIPTION

[0094] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0095] See also Figure 1 The present invention provides a UAV emergency control management method based on digital twins, comprising:

[0096] S100: Real-time collection of image data within the emergency area and parameter records of each UAV.

[0097] S200, using digital twin technology to build virtual scenes based on image data, and to map and divide the three-dimensional grid in real time.

[0098] S300: Planning feasible routes in the virtual scene based on the drone parameter information, and selecting key routes.

[0099] S400: Analyze each plan based on the key route planning plan to set a communication plan.

[0100] S500 and each drone fly according to the communication plan, and display the virtual scene in real time through the visualization page.

[0101] The emergency area refers to the specific geographical area where the disaster occurs. Image data refers to video images captured by various drones within the emergency area.

[0102] Parameter records include environmental parameters, operational parameters, and command parameters. Environmental parameters refer to the smoke concentration and temperature of the surrounding environment collected by the drone at different times. Operational parameters refer to the drone's spatial position, control signal strength, and cooperative signal strength at different times. Command parameters include the flight objectives issued by the control center.

[0103] Environmental parameters can be collected in two ways: one is to collect smoke concentration and temperature at close range around the drone using smoke and temperature sensors. The other is to predict smoke concentration and temperature at long distances around the drone through real-scene image analysis and thermal imaging analysis.

[0104] Short distance refers to the maximum effective distance that the sensor can detect, and long distance refers to a distance that exceeds the maximum effective distance that the sensor can detect and is less than or equal to the maximum effective distance of real-scene image analysis or thermal imaging analysis.

[0105] The prediction of smoke concentration usually requires a correlation analysis based on the smoke transmittance in the corresponding area of ​​the real-life image and the actual smoke concentration collected by sensors in the corresponding area by other drones at the same time, to obtain a relationship formula between the two, and thus predict and analyze the smoke concentration based on the real-life image.

[0106] Spatial position refers to the real-time three-dimensional coordinates of the drone in the air. Control signal strength refers to the signal strength between the drone and the command center, and collaborative signal strength refers to the signal strength between the drone and other drones.

[0107] By integrating sensor detection with image analysis technologies (such as smoke transparency correlation models), comprehensive collection of near- and long-range environmental parameters is achieved. Drone spatiotemporal state parameters (spatial position, signal strength) and flight instructions are simultaneously integrated to build a multidimensional data pool. This provides high-precision, multi-dimensional, real-time data support for subsequent virtual scene construction and path planning, enhancing robust perception of complex and catastrophic environments.

[0108] S200 includes:

[0109] S201: Obtain image data and position parameters collected by each drone in the emergency area, extract common feature points in different images for matching, and mark overlapping areas.

[0110] S202. Use a multi-perspective 3D reconstruction algorithm to combine images from multiple perspectives to build a virtual scene, map entities in the image to the virtual scene through digital twin technology, and update the mapped entity objects in real time.

[0111] S203: Obtain parameter records of each UAV, analyze changes in control signal strength at different times, and draw a line graph for each UAV with time as the horizontal axis and control signal strength as the vertical axis.

[0112] S204 , setting a duration q, dividing the sampling intervals on the line graph according to the duration q, and marking the sampling intervals where the extreme difference in control signal strength is greater than a threshold w.

[0113] S205 , analyzing the spatial positions within the time period corresponding to each marked sampling interval, and using the Euclidean distance algorithm to select the two spatial positions with the farthest distance for calculation, to obtain the spatial distance of the corresponding marked sampling interval.

[0114] S206 , selecting the minimum value among the spatial distances of all marked sampling intervals as the sensing distance of the corresponding UAV, and calculating the average sensing distance of all UAVs as the grid scale L.

[0115] S207. Divide the virtual scene into three-dimensional grid areas, each of which is a cube with an edge length of L.

[0116] A digital twin scene is dynamically constructed using a multi-view 3D reconstruction algorithm. This algorithm, combined with the attenuation characteristics of the drone's control signal strength (calculating the minimum sensing distance through range marker intervals), generates an adaptive grid scale (edge ​​length L). This ensures real-time synchronization between the virtual scene and the physical world, and that the grid division matches the drone's actual communication attenuation characteristics, providing a spatial reference unit for traffic zone screening and route planning.

[0117] S300 includes:

[0118] S301: A three-dimensional grid area in a virtual scene where no physical objects exist is used as a passage area, feasible routes are planned according to the current spatial position of the UAV and the flight destination, and the passage areas passed by each feasible route are marked.

[0119] S302: Analyze the parameter records of all drones, filter out the control signal strength, smoke concentration and temperature at different times when the spatial position is within the marked passage area, and fit the influence relationship expression of each marked passage area. Specifically, it includes:

[0120] S302-1. Establish a training set for each marked passage area and obtain the control signal strength, smoke concentration and temperature of each UAV when its spatial position is in the marked passage area.

[0121] S302-2. The control signal strength of the same UAV at the same time is used as the dependent variable, and the smoke concentration and temperature are used as independent variables and packaged into samples.

[0122] The control signal strength, smoke concentration, and temperature in each sample belong to the environmental and operating parameters collected simultaneously by a specific drone at a specific moment. Data in different samples may belong to different drones or be collected at different times.

[0123] S302-3. All samples are put into the training set of the marked passage area to which they belong, and a polynomial model is established. Each training set is input into the polynomial model for training, and the influence relationship expression of each marked passage area is obtained by fitting:

[0124] ;

[0125] Where, To control the signal strength, is the smoke concentration, is the temperature, is the baseline signal strength, is the linear attenuation coefficient of smoke concentration, is the temperature linear attenuation coefficient, is the secondary attenuation coefficient of smoke concentration, is the temperature quadratic attenuation coefficient, is the error coefficient.

[0126] Baseline signal strength It refers to the theoretical signal strength under normal conditions, which serves as a benchmark reference for signal strength attenuation.

[0127] Smoke concentration linear attenuation coefficient It refers to the exponential attenuation contribution of the increase in smoke concentration to the signal intensity, and is used to control the power-law attenuation rate of smoke concentration.

[0128] Temperature linear attenuation coefficient It refers to the exponential decay contribution of the temperature increase to the signal intensity, which is used to capture the linear effect of temperature on the dielectric constant.

[0129] Secondary attenuation coefficient of smoke concentration It refers to the nonlinear correction of the exponential decay by the square term of the smoke concentration, such as the particle aggregation effect or complex scattering mechanism, and is used to describe the accelerated decay at high smoke concentrations.

[0130] Temperature quadratic attenuation coefficient It refers to the direct attenuation of signal strength by the square of the temperature, such as the nonlinear degradation of device thermal noise, and is used to quantify the additional deterioration of device performance at high temperatures.

[0131] Error coefficient refers to random noise or measurement error not explained by the model and is used for the uncertainty of the statistical fit.

[0132] S303: Obtain the latest smoke concentration and temperature of each marked passage area, substitute them into the corresponding influence relationship expression, and calculate the predicted control signal strength.

[0133] To obtain the latest smoke concentration and temperature in a marked area, the drone does not need to be in the marked area to collect data. As long as the environmental parameter collection requirements are met, whether through sensors or image analysis methods, as long as the time meets the definition of the latest data, it can be used as the latest smoke concentration and temperature. The definition of the latest data is pre-defined by the management.

[0134] S304: Calculate the pass coefficient of the corresponding feasible route based on the predicted control signal strength of each marked pass area, and select the route with a pass coefficient greater than the coefficient threshold for each drone. The feasible routes are taken as key routes. Specifically including:

[0135] S304-1. Obtaining a feasible route All marked passage areas are arranged in the order of flight and the difference in the predicted control signal strength between adjacent marked passage areas is calculated. The standard deviation of all differences is calculated as the fluctuation coefficient. .

[0136] S304-2. Statistics of feasible routes Number of all marked passage areas , calculate the average value of all predicted control signal strengths as the reference signal strength , substitute into the formula to calculate the feasible route The traffic coefficient :

[0137] ;

[0138] Where, 、 and is a constant, is the average reference signal strength of all feasible routes, is the average fluctuation coefficient of all feasible routes, It is the average number of marked passage areas of all feasible routes.

[0139] S304-3, and so on, respectively calculate the traffic coefficient of each feasible route.

[0140] A polynomial model (incorporating the linear and nonlinear attenuation patterns of smoke concentration and temperature on signal strength) is used to predict control signal strength in each traffic zone. A traffic coefficient calculation formula based on volatility, signal strength, and path complexity is used to eliminate high-interference routes. Dynamically quantifying the impact of environmental physical parameters on communication quality helps avoid high-risk routes, such as those in dense smoke and high-temperature interference areas, and improve route feasibility.

[0141] S400 includes:

[0142] S401: Count the number of all drones, a, and multiply the number of key routes for all drones to obtain b. Create b plans, and place a key route belonging to a different drone in each plan.

[0143] All drones are paired together, and the collaborative signal strength of each pair of associated drones at different spatial locations is analyzed based on parameter records, and the collaborative relationship expression of each pair of associated drones is fitted. Specifically, it includes:

[0144] S401-1. Establish a training set for each pair of associated drones, calculate the spatial distance of each pair of associated drones at different spatial positions, and obtain the cooperative signal strength at different spatial distances.

[0145] S401-2. The collaborative signal strength at the same time is used as the dependent variable, and the spatial distance is used as the independent variable and packaged into samples.

[0146] S401-3. All samples are placed in the training set of the corresponding related drones, and a linear regression model is established. Each training set is input into the model for training, and the collaborative relationship expression of each related drone is obtained by fitting. The specific training content includes:

[0147] The independent variable of each sample in the training set is used as the input value of the linear regression model, and the difference between the output result of each sample and the dependent variable is used as the gap value. Set the error threshold and adjust the intercept and the regression coefficient Make the gap values ​​of all samples in the training set smaller than the error threshold, and obtain the expression after training:

[0148] ;

[0149] Where, is the dependent variable, is the independent variable.

[0150] S402, obtain the drone in solution i Key routes , set the standard flight speed, calculate the drone Reach key routes Next to the first marked passage Time , analysis time The marked passage areas where other drones in scheme i fly along their respective key routes .

[0151] S403, according to time Get off the drone Calculate the spatial distance with other drones in the marked passage area, and substitute the corresponding collaborative relationship expression to calculate the predicted collaborative signal strength . Analyzing drones In marked traffic areas The predicted control signal strength , and other drones in their respective marked passage areas The predicted control signal strength .

[0152] S404, Marking and Are greater than drone, select and The minimum value is used as the forwarding signal strength of the marking drone. The marking drone with the largest forwarding signal strength is selected as the drone In marked traffic areas The auxiliary object, the forwarding signal strength of the auxiliary object is used as the UAV In marked traffic areas New predictions control signal strength.

[0153] S405. Analyze key routes one by one Mark each traffic area and set a new prediction control signal strength. After the analysis is completed, recalculate the key routes Analyze the marked traffic areas of other key routes under plan i in turn and recalculate the traffic coefficients.

[0154] S406: After all key routes are analyzed, the traffic coefficients of all key routes are summed up to form the traffic index of solution i. The traffic index of each solution is calculated separately, and the solution with the largest traffic index is selected as the communication solution.

[0155] The collaborative efficiency of multiple drones is predicted using a collaborative relationship expression (a linear model of spatial distance and collaborative signal strength). Combined with a dynamic forwarding mechanism (preferably using auxiliary drones for relay communication), the traffic index of each path combination is iteratively optimized. This enables self-organizing collaboration among drones and redundant communication link planning, ensuring reliable data transmission in weak signal areas and improving the overall robustness of mission execution.

[0156] In S500, all drones fly along the key routes in the communication plan. When they arrive at the marked passage area with auxiliary objects, they transmit data to the auxiliary objects, which then forward it to the command center.

[0157] During flight, the traffic coefficients for each marked traffic zone are updated in real time, routes and auxiliary objects are dynamically adjusted, and environmental and operating parameters are recorded in real time and stored in the parameter log. The command center displays the virtual scene in real time through a visual interface.

[0158] The digital twin interface maps drone trajectories, environmental parameter heatmaps, and signal strength changes in real time. Combined with a dynamic path optimization mechanism (updating traffic coefficients and auxiliary objects based on real-time data), this empowers the command center with global situational awareness and rapid intervention capabilities, improving the real-time and operability of emergency response through visual closed-loop feedback.

[0159] See also Figure 2 The present invention provides a UAV emergency control and management system based on digital twins, including a UAV data acquisition module, a virtual scene construction module, an emergency route analysis module, an execution plan planning module and an operation visualization module.

[0160] The UAV data acquisition module is used to collect image data within the emergency area and record the parameters of the UAV.

[0161] The virtual scene construction module uses digital twin technology to build virtual scenes, map and divide three-dimensional grids in real time.

[0162] The emergency route analysis module is used to plan feasible routes for each drone in a virtual scene and screen out key routes.

[0163] The execution plan planning module is used to plan plans based on key routes, analyze each plan and set communication plans.

[0164] The operation visualization module is used to supervise each drone to fly according to the communication plan, and display the virtual scene in real time through the visualization page of the command center.

[0165] The drone data acquisition module uses drones equipped with smoke / temperature sensors and cameras to collect real-time video images of the emergency area and three types of parameters (environmental parameters, operational parameters, and command parameters). This environmental parameter collection combines short-range sensor detection with long-range image prediction technology (thermal imaging and smoke transmission correlation analysis), and records operational data such as 3D coordinates and dual signal strength.

[0166] It provides multi-level environmental perception capabilities to compensate for the blind spots of single sensors. It also improves data robustness through redundant acquisition, laying a data foundation for virtual scene construction and path analysis.

[0167] The virtual scene construction module uses a multi-view 3D reconstruction algorithm to fuse images from multiple drone perspectives, updating the digital twin scene in real time. By analyzing the fluctuation range of control signal strength, the grid scale is dynamically calculated (based on the mean minimum sensing distance) and divided into standard cube grids.

[0168] Achieve high-precision dynamic modeling of disaster environments, grid division to match the drone signal attenuation characteristics, and provide spatial computing units for traffic zone classification and path planning.

[0169] The emergency route analysis module combines traffic zone screening with a polynomial model (including bilinear and quadratic attenuation terms for smoke and temperature) to predict control signal strength. A traffic coefficient formula (taking into account volatility, reference strength, and path complexity) is designed to filter high-risk routes.

[0170] Quantify the comprehensive impact of environmental interference factors on communication quality, screen out flight paths with stable signals and reliable equipment, and reduce the risk of communication interruption during mission execution.

[0171] The execution plan planning module constructs the UAV collaborative relationship expression (spatial distance-signal strength linear model) and dynamic assistance mechanism (forwarding relay strategy). By traversing b route combinations and iteratively updating the pass index, it selects the communication plan that maximizes the collaborative efficiency.

[0172] Optimize the collaborative communication strategy of drone swarms, compensate for the quality of weak signal areas through dynamic relay forwarding, and improve the overall success rate of multi-machine missions in complex environments.

[0173] The operation visualization module integrates functions such as parameter record update, path dynamic adjustment, and auxiliary object switching, and annotates drone trajectories, signal strength heat maps, and environmental parameter change curves in real time in the digital twin scene.

[0174] Providing a decision-making support visualization interface enables the command center to understand the overall situation and quickly intervene in abnormal conditions, enhancing the timeliness of emergency response.

[0175] Example 1: Assuming a feasible route The fluctuation coefficient of all feasible routes is 2.4, the number of all marked pass zones is 12, and the reference signal strength is -40dBm; the average fluctuation coefficient of all feasible routes is 3, the average number of marked pass zones of all feasible routes is 10, and the average reference signal strength of all feasible routes is -60dBm;

[0176] When the constant 、 and When the values ​​are 1, 0.6, and 0.4 respectively, substitute them into the formula to calculate the feasible route. The traffic coefficient:

[0177] ;

[0178] A feasible route The traffic coefficient is 1.16.

[0179] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0180] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. The UAV emergency control and management method based on digital twin is characterized by: The method includes: S100, real-time collection of image data within the emergency area and parameter records of each drone; S200, using digital twin technology to build virtual scenes based on image data, and to map and divide the 3D grid in real time; S300, planning feasible routes in the virtual scene based on the drone parameter information, and screening out key routes; S400, analyzing each plan based on the key route planning plan to set a communication plan; S500 and each drone fly according to the communication plan, and display the virtual scene in real time through the visualization page.

2. The UAV emergency control management method based on digital twin according to claim 1 is characterized in that: In S100, the emergency area refers to a specific geographical area where a disaster occurs; the image data refers to video images captured by each drone within the emergency area; Parameter records include environmental parameters, operating parameters, and command parameters. Environmental parameters refer to the smoke concentration and temperature of the surrounding environment collected by the drone at different times. Operating parameters refer to the spatial position, control signal strength, and collaborative signal strength of the drone at different times. Command parameters include the flight objectives issued by the control center. Spatial position refers to the real-time three-dimensional coordinates of the drone in the air; control signal strength refers to the signal strength between the drone and the command center, and collaborative signal strength refers to the signal strength between the drone and other drones.

3. The UAV emergency control management method based on digital twin according to claim 2 is characterized in that: S200 includes: S201: Obtain image data and location parameters collected by each drone in the emergency area, extract common feature points from different images for matching, and mark overlapping areas; S202. Using a multi-view 3D reconstruction algorithm to combine images from multiple viewpoints to build a virtual scene, mapping entities in the images to the virtual scene using digital twin technology, and updating the mapped entity objects in real time; S203, obtaining parameter records for each UAV, analyzing changes in control signal strength at different times, and drawing a line graph for each UAV with time as the horizontal axis and control signal strength as the vertical axis; S204: Set a duration q, divide the sampling intervals on the line graph according to the duration q, and mark the sampling intervals where the extreme difference in control signal strength is greater than a threshold w; S205, analyzing the spatial positions within the time period corresponding to each marked sampling interval, and using the Euclidean distance algorithm to select the two spatial positions with the greatest distance to calculate, to obtain the spatial distance of the corresponding marked sampling interval; S206: Select the minimum value among the spatial distances of all marked sampling intervals as the sensing distance of the corresponding drone, and calculate the average sensing distance of all drones as the grid scale L; S207. Divide the virtual scene into three-dimensional grid areas, each of which is a cube with an edge length of L.

4. The UAV emergency control management method based on digital twin according to claim 3 is characterized in that: S300 includes: S301: A three-dimensional grid area in the virtual scene where no physical objects exist is used as a passage area, and a feasible route is planned based on the current spatial position of the UAV and the flight destination, and the passage area passed by each feasible route is marked; S302: Analyze the parameter records of all drones, filter out the control signal strength, smoke concentration, and temperature at different times when the spatial position is within the marked passage area, and fit the influence relationship expression of each marked passage area; S303, obtaining the latest smoke concentration and temperature of each marked passage area, substituting them into the corresponding influence relationship expression to calculate the predicted control signal strength; S304: Calculate the pass coefficient of the corresponding feasible route based on the predicted control signal strength of each marked pass area, and select the route with a pass coefficient greater than the coefficient threshold for each drone. The feasible routes are taken as the key routes.

5. The UAV emergency control management method based on digital twin according to claim 4 is characterized in that: The establishment of the influence relationship expression in S302 includes: S302-1. Establish a training set for each marked passage area and obtain the control signal strength, smoke concentration, and temperature of each UAV when its spatial position is in the marked passage area; S302-2: The control signal strength of the same UAV at the same time is used as the dependent variable, and the smoke concentration and temperature are used as independent variables and packaged into samples; S302-3. All samples are put into the training set of the marked passage area to which they belong, and a polynomial model is established. Each training set is input into the polynomial model for training, and the influence relationship expression of each marked passage area is obtained by fitting: ； Where, To control the signal strength, is the smoke concentration, is the temperature, is the baseline signal strength, is the linear attenuation coefficient of smoke concentration, is the temperature linear attenuation coefficient, is the secondary attenuation coefficient of smoke concentration, is the temperature quadratic attenuation coefficient, is the error coefficient.

6. The UAV emergency control management method based on digital twin according to claim 4 is characterized in that: The calculation of the traffic coefficient in S304 includes: S304-1. Obtaining a feasible route All marked passage areas are arranged in the order of flight and the difference in the predicted control signal strength between adjacent marked passage areas is calculated. The standard deviation of all differences is calculated as the fluctuation coefficient. ; S304-2. Statistics of feasible routes Number of all marked passage areas , calculate the average value of all predicted control signal strengths as the reference signal strength , substitute into the formula to calculate the feasible route The traffic coefficient : ； Where, 、 and is a constant, is the average reference signal strength of all feasible routes, is the average fluctuation coefficient of all feasible routes, is the average number of marked passage areas for all feasible routes; S304-3, and so on, respectively calculate the traffic coefficient of each feasible route.

7. The UAV emergency control management method based on digital twin according to claim 4 is characterized in that: S400 includes: S401. Count the number of all drones (a), and multiply the number of key routes for all drones by the sum to obtain b; establish b plans, and include a key route belonging to a different drone in each plan; associate all drones with each other, analyze the cooperative signal strength of each pair of associated drones at different spatial locations based on the parameter records, and fit the cooperative relationship expression for each pair of associated drones; S402, obtain the drone in solution i Key routes , set the standard flight speed, calculate the drone Reach key routes Next to the first marked passage Time , analysis time The marked passage areas where other drones in scheme i fly along their respective key routes ; S403, according to time Get off the drone Calculate the spatial distance with other drones in the marked passage area, and substitute the corresponding collaborative relationship expression to calculate the predicted collaborative signal strength ;Analysis of drones In marked traffic areas The predicted control signal strength , and other drones in their respective marked passage areas The predicted control signal strength ; S404, Marking and Are greater than drone, select and The minimum value is used as the forwarding signal strength of the marking drone; the marking drone with the largest forwarding signal strength is selected as the drone In marked traffic areas The auxiliary object, the forwarding signal strength of the auxiliary object is used as the UAV In marked traffic areas New predictions control signal strength; S405. Analyze key routes one by one Mark each traffic area and set a new prediction control signal strength. After the analysis is completed, recalculate the key routes The traffic coefficient of ; analyze the marked traffic areas of other key routes under scheme i in turn and recalculate the traffic coefficient; S406. After all key routes are analyzed, the traffic coefficients of all key routes are summed up to obtain the traffic index of solution i; the traffic index of each solution is calculated separately, and the solution with the largest traffic index is selected as the communication solution.

8. The UAV emergency control management method based on digital twin according to claim 7 is characterized in that: The establishment of the collaborative relationship expression in S401 includes: S401-1. Establish a training set for each pair of associated drones, calculate the spatial distance between each pair of associated drones at different spatial positions, and obtain the collaborative signal strength at different spatial distances; S401-2. The collaborative signal strength at the same time is used as the dependent variable and the spatial distance is used as the independent variable and packaged into samples; S401-3. All samples are placed in the training set of the corresponding related UAVs, and a linear regression model is established. Each training set is input into the model for training, and the collaborative relationship expression of each related UAV is obtained by fitting: ； Where, is the dependent variable, is the independent variable.

9. The UAV emergency control management method based on digital twin according to claim 7 is characterized in that: In S500, all drones fly along the key routes in the communication plan. When they reach the marked passage area with auxiliary objects, they transmit data to the auxiliary objects, which then forward it to the command center. During the flight, the traffic coefficient of each marked traffic area is updated in real time, the route and auxiliary objects are adjusted dynamically, and the environmental parameters and operating parameters are recorded in real time and stored in the parameter record; the command center displays the virtual scene in real time through a visual interface.

10. The UAV emergency control and management system based on digital twin is characterized by: The system includes a drone data acquisition module, a virtual scene construction module, an emergency route analysis module, an execution plan planning module, and an operation visualization module; The UAV data acquisition module is used to collect image data within the emergency area and record the parameters of the UAV; The virtual scene construction module uses digital twin technology to build virtual scenes, map them in real time, and divide them into three-dimensional grids; The emergency route analysis module is used to plan feasible routes for each drone in the virtual scene and screen out key routes; The execution plan module is used to plan plans based on key routes, analyze each plan and set communication plans; The operation visualization module is used to supervise each drone to fly according to the communication plan, and display the virtual scene in real time through the visualization page of the command center.

Images