Geographic data acquisition method and device, equipment and storage medium

Abstract

The invention relates to a geographic data acquisition method and device, equipment and a storage medium. According to the main technical scheme, task data of a to-be-executed task, equipment data of alternative unmanned aerial vehicles, real-time weather data and real-time environment data are obtained, at least one target unmanned aerial vehicle executing the to-be-executed task is matched from the alternative unmanned aerial vehicles according to the task data and the equipment data, and the target unmanned aerial vehicle is matched with the to-be-executed task according to the task data and the equipment data. The method comprises the steps of determining an optimal flight path of at least one target unmanned aerial vehicle, collecting initial geographic data based on the optimal flight path of the at least one target unmanned aerial vehicle, automatically adjusting the optimal flight path and a data collection strategy according to real-time weather data, real-time environment data and task data, performing data analysis and fusion on the initial geographic data, and obtaining the target unmanned aerial vehicle. The flight strategy and the data acquisition strategy of the unmanned aerial vehicle can be dynamically adjusted in real time, and the data acquisition efficiency and precision are improved.

Description

Technical Field

[0001] The present application relates to the field of electronic information technology, and in particular to a geographic data collection method, device, equipment and storage medium. Background Art

[0002] At present, using drones for geographic data collection has become an efficient and economical way. With its advantages of rapid response, flexible mobility, and wide coverage, drones have been widely used in land surveying, resource surveys, environmental monitoring, natural disaster assessment and other fields. However, existing drone geographic data collection systems mostly rely on preset flight paths for data collection. Although this method ensures the continuity and stability of data collection to a certain extent, it lacks the ability to adjust data collection strategies in real time and dynamically, resulting in limited data collection efficiency and accuracy. Summary of the invention

[0003] Based on this, the present application provides a geographic data collection method, device, equipment and storage medium to solve the problem that drones lack the ability to adjust data collection strategies in real time and dynamically when collecting data, resulting in limited data collection efficiency and accuracy.

[0004] In a first aspect, a method for collecting geographic data is provided, the method comprising:

[0005] Obtain mission data for missions to be performed, equipment data for candidate drones, real-time weather data, and real-time environmental data;

[0006] According to the mission data and the equipment data, at least one target drone for executing the mission to be executed is matched from the candidate drones;

[0007] Determine an optimal flight path for at least one target UAV based on the mission data and the equipment data;

[0008] Based on the optimal flight path of at least one target UAV, initial geographic data is collected, and the optimal flight path and data collection strategy are automatically adjusted according to real-time weather data, real-time environmental data and mission data;

[0009] The initial geographic data is analyzed and integrated to obtain geographic data.

[0010] According to an achievable method in an embodiment of the present application, at least one target drone that performs the task to be performed is matched from candidate drones according to the task data and the device data, including:

[0011] Determine the task type of the task to be executed according to the task data of the task to be executed;

[0012] Evaluate the demand information of the tasks to be performed according to the task types to be performed;

[0013] Determine the priority of the tasks to be executed based on the demand information of the tasks to be executed;

[0014] Evaluate the overall capability score of the drone based on the drone’s equipment data;

[0015] At least one target drone for executing the task to be executed is matched from the candidate drones based on the priority of the task to be executed and the overall capability score of the drone.

[0016] According to an achievable method in an embodiment of the present application, determining an optimal flight path of at least one target UAV according to mission data and device data includes:

[0017] Define the objective function, decision variables and constraints based on the mission data and equipment data, where the decision variables are the multiple waypoints that the target UAV needs to pass through;

[0018] Generating at least one initial population according to the plurality of waypoints, wherein the initial population includes initial individuals composed of the plurality of waypoints in different arrangement orders;

[0019] Based on the objective function, evaluating the fitness of at least one initial individual;

[0020] According to the fitness of at least one initial individual, a new generation of individuals of at least one initial individual is generated until a preset number of iterations is reached, and the fitness fluctuation value of at least one initial individual is less than a preset threshold, and the optimal flight path of at least one target UAV is output.

[0021] According to an achievable method in an embodiment of the present application, automatically adjusting the optimal flight path according to real-time weather data includes:

[0022] According to the real-time weather data, the flight risk information of at least one target UAV is predicted based on the pre-trained weather impact model;

[0023] According to the flight risk information of at least one target UAV, the order or number of waypoints in the optimal flight path or the flight altitude and flight speed of at least one target UAV are automatically adjusted.

[0024] According to an achievable method in an embodiment of the present application, the data collection strategy is automatically adjusted according to the real-time environment data and task data, including:

[0025] Determine whether the initial geographic data meets the data collection standards based on real-time environmental data and mission data;

[0026] If the initial geographic data does not meet the data collection criteria, adjust the data collection area and increase the data collection frequency based on the real-time environmental data and the current flight position of at least one target UAV;

[0027] A re-collection instruction is generated according to the adjusted data collection area and the increased data collection frequency, and the re-collection instruction is sent to at least one target UAV.

[0028] According to an achievable method in an embodiment of the present application, the initial geographic data includes data collected by various sensors; the initial geographic data is analyzed and fused to obtain geographic data, including:

[0029] Extracting target features of the collected data of each sensor and geometric features of position and posture changes, wherein the target features include position change rate, acceleration change and ambient light intensity features;

[0030] The principal component analysis algorithm is used to reduce the dimension of target features and geometric features to obtain reduced dimension data;

[0031] Data analysis and fusion are performed on the dimension-reduced data to obtain geographic data.

[0032] According to an achievable method in the embodiment of the present application, data analysis and fusion are performed on the dimension-reduced data to obtain geographic data, including:

[0033] Analyze the reduced-dimensional data through a pre-trained learning model to obtain target geographic data;

[0034] If the target geographic data is data collected by sensors of different types, a consistency check is performed on the target geographic data to obtain corrected geographic data;

[0035] If the measurement data in the target geographic data contains noise, the target geographic data is filtered using a Kalman filter algorithm to obtain filtered geographic data;

[0036] If the target geographic data is data collected by multiple sensors of the same type, the target geographic data is fused by a preset weighted formula to obtain fused geographic data;

[0037] The corrected geographic data, filtered geographic data and fused geographic data corresponding to different sensors are combined to obtain geographic data.

[0038] In a second aspect, a geographic data collection device is provided, the device comprising:

[0039] The acquisition module is used to obtain the task data of the task to be executed, the equipment data of the candidate drones, the real-time weather data and the real-time environmental data;

[0040] A matching module, used for matching at least one target drone for executing the task to be executed from the candidate drones according to the task data and the equipment data;

[0041] A determination module, for determining an optimal flight path of at least one target UAV based on the mission data and the equipment data;

[0042] An adjustment module, for collecting initial geographic data based on an optimal flight path of at least one target UAV, and automatically adjusting the optimal flight path and data collection strategy according to real-time weather data, real-time environmental data, and mission data;

[0043] The fusion module is used to perform data analysis and fusion on the initial geographic data to obtain geographic data.

[0044] In a third aspect, a computer device is provided, comprising:

[0045] at least one processor; and

[0046] a memory communicatively connected to the at least one processor; wherein,

[0047] The memory stores computer instructions that can be executed by the at least one processor, and the computer instructions are executed by the at least one processor to enable the at least one processor to perform the method involved in the first aspect above.

[0048] In a fourth aspect, a computer-readable storage medium is provided, on which computer instructions are stored, characterized in that the computer instructions are used to enable a computer to execute the method involved in the first aspect above.

[0049] According to the technical content provided in the embodiments of the present application, by obtaining the task data of the task to be performed, the equipment data of the alternative drones, the real-time weather data and the real-time environmental data, at least one target drone that performs the task to be performed is matched from the alternative drones according to the task data and the equipment data, and the optimal flight path of the at least one target drone is determined according to the task data and the equipment data. Based on the optimal flight path of the at least one target drone, initial geographic data is collected, and the optimal flight path and data collection strategy are automatically adjusted according to the real-time weather data, the real-time environmental data and the task data. The initial geographic data is analyzed and fused to obtain geographic data, which can dynamically adjust the flight strategy and data collection strategy of the drone in real time to improve data collection efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is a schematic structural diagram of a geographic data collection system in one embodiment;

[0051] Figure 2 A schematic diagram of a flow chart of a geographic data collection method in one embodiment;

[0052] Figure 3 is a structural block diagram of a geographic data collection device in one embodiment;

[0053] Figure 4 FIG. 4 is a schematic structural diagram of a computer device in one embodiment. DETAILED DESCRIPTION

[0054] The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0055] For ease of understanding, the system to which this application is applicable is first described. The geographic data collection method provided in this application can be applied to Figure 1 The geographic data collection system architecture shown in FIG. 100 includes at least one drone 110 and a ground control center 120, wherein the drones 110 use a wireless network for real-time communication, and at least one drone 110 uses a wireless network for real-time communication with the ground control center 120. The wireless network can use long-range radio (Long Range Radio, LoRa), the fifth generation mobile communication technology (5th Generation Mobile Communication Technology, 5G), etc. to ensure the stability and reliability of data transmission. The drone 110 is equipped with a high-precision Global Navigation Satellite System (Global Navigation Satellite System, GNSS) receiver and sensors that adapt to different collection requirements, wherein the GNSS receiver ensures positioning accuracy, and the sensors include cameras, laser radars, temperature sensors, humidity sensors, anemometers, and barometers. The drone 110 has a long endurance and load capacity, and can perform long-term data collection tasks. The ground control center 120 is responsible for mission planning, monitoring drone status, and preliminary processing and storage of data. The UAV 110 collects real-time weather data and real-time environmental data and reports them to the ground control center 120. The ground control center 120 obtains the mission data of the mission to be performed, the equipment data of the candidate UAVs, the real-time weather data and the real-time environmental data. According to the mission data and the equipment data, the ground control center 120 matches at least one target UAV that performs the mission to be performed from the candidate UAVs. According to the mission data and the equipment data, the optimal flight path of the at least one target UAV is determined. Based on the optimal flight path of the at least one target UAV, the UAV 110 collects initial geographic data. The ground control center 120 automatically adjusts the optimal flight path and data collection strategy according to the real-time weather data, the real-time environmental data and the mission data, and performs data analysis and fusion on the initial geographic data to obtain geographic data.

[0056] Figure 2 A flowchart of a geographic data collection method provided in an embodiment of the present application, the method can be performed as follows Figure 1 The geographic data collection system 100 shown in FIG. Figure 2 As shown, the method may include the following steps:

[0057] S210, obtaining task data of the task to be executed, equipment data of the candidate drones, real-time weather data and real-time environmental data.

[0058] Among them, the task data is the specific content of the task to be executed, which may include the task execution location, task execution time, task execution scope, data collection type, etc.

[0059] The alternative drone is a drone that has the ability to perform the task to be performed. The equipment data may include the drone's endurance data, load data, sensor capabilities, battery power, flight altitude and other data.

[0060] Real-time weather data refers to parameters and information related to current weather conditions, which may include the current environment's temperature, humidity, air pressure, wind direction, wind speed, weather phenomena (such as sunny, cloudy, overcast, rain, snow, fog, etc.), precipitation, UV index, etc.

[0061] Real-time environmental data is information about environmental conditions collected and transmitted in real time by various environmental monitoring devices, which may include the concentration of pollutants in the air (such as PM2.5, PM10, sulfur dioxide, nitrogen dioxide, etc.), oxygen content, carbon dioxide concentration, etc.

[0062] Real-time weather data can use real-time data from weather stations, or use real-time data from international weather services, such as the National Oceanic and Atmospheric Administration and the Global Weather Data Service. It can also be based on real-time data collected by meteorological sensors installed on drones, such as temperature sensors, humidity sensors, anemometers and barometers. The sensors have cellular network capabilities so that data can be uploaded to the control center in real time.

[0063] The control center can obtain the task data input by the user through the input device in real time, the equipment data of the alternative drones pre-stored in the control center, and the real-time environmental data can be obtained in real time by the drone's sensors and uploaded to the ground control center or cloud server through wireless communication.

[0064] S220, matching at least one target drone that executes the task to be executed from the candidate drones according to the task data and the equipment data.

[0065] First, according to the task data of the tasks to be performed, the tasks to be performed are classified to obtain the task types of the tasks to be performed. The task types may include topographic mapping, vegetation monitoring, and environmental monitoring. Then, according to the task types of the tasks to be performed, the demand information of the tasks to be performed is evaluated. The demand information may include the area scope of the tasks to be performed, the required data type, the acquisition accuracy, the time limit, etc. Next, according to the demand information of the tasks to be performed, the importance, urgency, resource requirements and expected benefit factors of the tasks to be performed are understood, and then the priority of the tasks to be performed is determined. According to the equipment data of the drone, all available candidate drones are evaluated, including flight capabilities, sensor capabilities, and current status, and the overall capability score of the drone is calculated. The tasks to be performed with high priority are matched with drones with higher overall capability scores, and at least one target drone to perform the tasks to be performed is selected from the candidate drones. The target drone is a candidate drone used to perform a certain task to be performed.

[0066] S230, determining an optimal flight path of at least one target UAV based on the mission data and the equipment data.

[0067] According to the mission data and equipment data, a genetic algorithm is used to determine the optimal flight path of at least one target UAV. First, the problem and parameters are defined according to the mission data and equipment data, which may include defining the objective function, decision variables and constraints, wherein the decision variables are multiple waypoints that the target UAV needs to pass through. Using a coding method suitable for the genetic algorithm, the problem is converted into a form suitable for calculation, and multiple waypoints are genetically encoded, with each gene representing a waypoint. An initial population is randomly generated, and the initial population includes multiple initial individuals. The fitness of each initial individual is evaluated, and a new generation of individuals of at least one initial individual is generated according to the fitness of at least one initial individual. The new generation of individuals is mutated to increase the diversity of the population, and the survivor selection strategy in the genetic algorithm is used to replace the old generation of population to ensure that excellent individuals are retained in the new population. When the new generation of individuals generating the initial individuals reaches the termination condition, the optimal flight path of at least one target UAV is output.

[0068] S240, based on the optimal flight path of at least one target UAV, collects initial geographic data, and automatically adjusts the optimal flight path and data collection strategy according to real-time weather data, real-time environmental data, and mission data.

[0069] The initial geographic data is the data collected by the sensors of each target drone. Based on the optimal flight path, the target drone collects initial geographic data, and the ground control center determines whether the current weather will affect the drone flight based on real-time weather data, and sets key weather thresholds, such as wind speed exceeding a certain value and precipitation reaching a certain level, as the basis for dynamic adjustment.

[0070] Combine historical flight data and historical weather data to pre-establish a weather impact model for weather and flight safety. According to real-time weather data, the flight risk index of at least one target drone is predicted based on the pre-trained weather impact model. According to the flight risk index of at least one target drone, the order or number of waypoints in the optimal flight path, as well as the flight altitude and flight speed of at least one target drone are automatically adjusted. By obtaining data collection requirements and environmental changes based on real-time environmental data and mission data, the data collection frequency and path are automatically adjusted, and re-collection is performed when data quality decreases.

[0071] S250, performing data analysis and fusion on the initial geographic data to obtain geographic data.

[0072] The ground control center uses machine learning algorithms to analyze and fuse the initial geographic data to generate high-quality geographic data. First, the target features of the data collected by each sensor and the geometric features of the position and attitude changes are extracted, where the target features include the position change rate, acceleration change, and ambient light intensity features. Then, the principal component analysis algorithm is used to reduce the dimension of the target features and geometric features to obtain reduced-dimensional data. Finally, the reduced-dimensional data is fused using weighted averaging, Kalman filtering, consistency checking and other methods to obtain geographic data.

[0073] It can be seen that the embodiment of the present application obtains the task data of the task to be performed, the equipment data of the alternative drones, the real-time weather data and the real-time environmental data, and matches at least one target drone that performs the task to be performed from the alternative drones according to the task data and the equipment data, determines the optimal flight path of at least one target drone according to the task data and the equipment data, collects initial geographic data based on the optimal flight path of at least one target drone, and automatically adjusts the optimal flight path and data collection strategy according to the real-time weather data, the real-time environmental data and the task data, performs data analysis and fusion on the initial geographic data to obtain geographic data, and can dynamically adjust the flight strategy and data collection strategy of the drone in real time to improve data collection efficiency and accuracy.

[0074] As an achievable manner, at least one target drone for executing the task to be executed is matched from candidate drones according to the task data and the equipment data, including:

[0075] Determine the task type of the task to be executed according to the task data of the task to be executed;

[0076] Evaluate the demand information of the tasks to be performed according to the task types to be performed;

[0077] Determine the priority of the tasks to be executed based on the demand information of the tasks to be executed;

[0078] Evaluate the overall capability score of the drone based on the drone’s equipment data;

[0079] At least one target drone for executing the task to be executed is matched from the candidate drones based on the priority of the task to be executed and the overall capability score of the drone.

[0080] According to the task execution location, task execution time, task execution scope, data collection type, etc. in the task data of the task to be executed, the task type of the task to be executed is determined to be one of topographic mapping, vegetation monitoring, and environmental monitoring. According to the task type of the task to be executed, the demand information of the task to be executed is evaluated, and the demand information may include multiple aspects such as regional scope, required data type, collection accuracy, time limit, etc.

[0081] If the task type is topographic mapping, the area scope: the boundaries and area of ​​the mapping area must be clearly defined, including large and complex areas such as cities and mountains; data type: high-precision terrain data is required, including ground height, landform features, and building outlines; acquisition accuracy: positioning accuracy is required to reach centimeters to ensure the accuracy of the topographic map. There are also clear requirements for the accuracy of optical images and lidar point clouds; time limit: it must be completed within a reasonable time frame, suitable for the weather window and the urgency of the project.

[0082] If the task type is vegetation monitoring, its regional scope: the monitoring scope can be forest areas, farmlands, and urban green spaces, and the specific area needs to be clarified; data type: a variety of remote sensing data is required, such as NDVI (normalized vegetation index), vegetation density, and species distribution, and corresponding sensors are adapted, such as multispectral cameras; acquisition accuracy: it is required to be able to quantify the accuracy of vegetation health status and coverage indicators, and a spatial resolution of meters is required; time limit: set the time according to the growing season and observation period to ensure the timeliness of the monitoring results.

[0083] If the task type is environmental monitoring, the area scope: the environmental area that needs to be monitored, including the surrounding areas of pollution sources and sensitive areas of rivers and lakes; data type: mainly environmental parameters that need to be detected, such as air quality (PM2.5, CO2 concentration), water quality (turbidity, pH value), temperature and humidity, and corresponding sensors such as gas sensors, temperature and humidity sensors; acquisition accuracy: real-time monitoring accuracy and accuracy are required, and the ability to respond quickly to changes in environmental parameters should be able to be responded to; time limit: flexible arrangements are required according to the monitoring purpose, and continuous monitoring and regular sampling are required.

[0084] Based on the demand information of the tasks to be executed, the importance, urgency, and resource requirements of the tasks to be executed are evaluated, and weight factors are set for different tasks to be executed in advance based on the importance, urgency, resource requirements of the tasks to be executed and the expected benefit factors. The priority of each task to be executed is calculated using the weighted scoring mechanism. The calculation formula is as follows:

[0085] Priority(Ti)=w1*R(Ti)+w2*E(Ti)+w3*I(Ti) (1)

[0086] Among them, Priority(Ti) represents the priority of the task to be executed, is the execution window of the task, R(Ti) is the resource requirement of the task to be executed, E(Ti) is the execution window of the task to be executed, I(Ti) is the importance index of the task to be executed, and w1, w2, and w3 represent the corresponding weight factors.

[0087] All candidate drones are evaluated, including flight capability (endurance, load), sensor capability (different types of sensor support), and current status (battery power, faults), etc. Based on the drone's equipment data, the overall capability score of the drone is evaluated, and the calculation formula is as follows:

[0088] G=k1*[T-(C+W)]+k2*[(P÷Pmax)*100%]+k3*[(N÷Ntotal)*100%]+k4*[(B÷Bmax)*100%]-k5*F; (2)

[0089] Among them, G represents the overall capability score, k1, k2, k3, k4, and k5 are weight coefficients, T is the flight time of the UAV without a payload, C is the energy consumption of the UAV during navigation, W is the energy consumption of the sensor and payload, P is the actual payload carried by the UAV, Pmax is the maximum design payload of the UAV, N is the number of sensor types supported by the UAV, Ntotal is the total number of sensor types required to complete the task to be performed, B is the current power of the UAV, Bmax is the maximum power of the UAV, and F is the fault state. When there is no fault, F = 0, and when there is a fault, F = 100.

[0090] Through the optimal matching algorithm, the priority algorithm is applied to assign at least one drone with a high overall capability score to the high-priority tasks to be executed. The specific score can be determined according to the task content of the tasks to be executed, whether the tasks to be executed require multiple drones to assist in data collection, how many drones are needed, and at least one target drone to execute the tasks to be executed is matched from the alternative drones.

[0091] As an achievable approach, an optimal flight path of at least one target UAV is determined based on the mission data and the equipment data, including:

[0092] Define the objective function, decision variables and constraints based on the mission data and equipment data, where the decision variables are the multiple waypoints that the target UAV needs to pass through;

[0093] Generating at least one initial population according to the plurality of waypoints, wherein the initial population includes initial individuals composed of the plurality of waypoints in different arrangement orders;

[0094] Based on the objective function, evaluating the fitness of at least one initial individual;

[0095] According to the fitness of at least one initial individual, a new generation of individuals of at least one initial individual is generated until a preset number of iterations is reached, and the fitness fluctuation value of at least one initial individual is less than a preset threshold, and the optimal flight path of at least one target UAV is output.

[0096] Among them, the objective function is to minimize the flight path length while considering power consumption. Let the objective function S = min[L÷(F+b)], L is the sum of the path lengths, F is the power consumption, and b is a constant. The decision variables are the multiple waypoints that the target drone needs to pass through. Let the target drone need to pass through multiple waypoints A, B, C, and D. The constraints are the flight altitude limit of the target drone and the order of visiting each waypoint.

[0097] Use a coding method suitable for genetic algorithms to convert the problem into a form suitable for calculation. Use 505 gene coding, each gene represents a waypoint. Suppose the waypoints are A, B, C, and D, and the coding method is ABCD, which means flying in order according to the four waypoints A, B, C, and D.

[0098] Randomly generate the initial population, set the population size to 100, and generate the initial individuals: generate the initial individuals by randomly arranging the order of waypoints. The initial population example can be expressed as:

[0099] [A, B, C, D]

[0100] [C, B, A, D]

[0101] [D, A, C, B]

[0102] …

[0103] [A, B, C, D], [C, B, A, D] and [D, A, C, B] are the initial individuals.

[0104] Evaluate the fitness of each initial individual, calculate the path length and other indicators according to the objective function, including power consumption and time consumption, and use the Euclidean distance and the actual flight route to calculate the path length. The fitness function can be expressed as: fitness = 1 ÷ (path length + epsilon), where epsilon is a constant set to avoid division by zero. From the fitness function, it can be seen that the smaller the path length, the higher the fitness of the initial individual.

[0105] Use the roulette wheel selection method to select initial individuals with higher fitness from the current initial population for crossover. Use the roulette wheel selection method to extract initial individuals from the fitness for mating. Perform a crossover operation on the selected initial individuals to generate a new generation of individuals. Partial matching crossover (PMX) or sequential crossover (OX) can be used to retain some parental genes. Crossover example:

[0106] Parent 1: [A, B, C, D]

[0107] Parent 2: [C, A, D, B]

[0108] After crossover, offspring are generated:

[0109] Sub 1: [A, C, D, B]

[0110] Sub 2: [C, B, A, D]

[0111] Among them, father 1 and father 2 are the initial individuals, and son 1 and son 2 are the new generation individuals.

[0112] Perform mutation operations on the new generation of individuals to increase the diversity of the population. Exchange mutation can be used. Mutation example:

[0113] Original gene: [A, C, D, B]

[0114] After mutation: [A, B, D, C] (the positions of C and B are swapped)

[0115] Use the survivor selection strategy in the genetic algorithm to replace the old generation of population to ensure that excellent individuals are retained in the new population. You can use complete replacement, that is, all the newly generated individuals replace the old individuals, or you can use elite retention, that is, select some individuals with high fitness to enter the next generation.

[0116] The preset iteration number is the specified number of generations, which can be set to 100 generations. The fitness converges, and the fitness fluctuation is less than the set threshold within several generations. The optimal flight path and its corresponding fitness value are output. The fitness value includes the path length and power consumption. The set threshold can be set according to actual needs.

[0117] Output example:

[0118] Optimal path: [A, B, D, C]

[0119] Minimum trail length: 12.5 km

[0120] Total power consumption: 80%

[0121] Through the task priority algorithm and path optimization algorithm, drone resources can be efficiently allocated, flight paths can be optimized, duplicate coverage and invalid flights can be reduced, thereby reducing operating costs and time consumption.

[0122] As a feasible approach, the optimal flight path can be automatically adjusted based on real-time weather data, including:

[0123] According to the real-time weather data, the flight risk information of at least one target UAV is predicted based on the pre-trained weather impact model;

[0124] According to the flight risk information of at least one target UAV, the order or number of waypoints in the optimal flight path or the flight altitude and flight speed of at least one target UAV are automatically adjusted.

[0125] Among them, a weather impact model of weather and flight safety is established by combining historical flight data and weather data. The initial training model is trained using a machine learning algorithm (support vector machine) to predict flight risk information under different meteorological conditions, and the initial training model parameters (including kernel function type, C value and γ value) are preliminarily selected. The processed data is divided into a training set and a test set, using 70% of the data as a training set and 30% of the data as a test set. Support vector machines are used for training on the training set to find the best hyperplane by minimizing the loss function to achieve the prediction of flight risk information.

[0126] Flight risk information is information about the impact of weather conditions on flight safety, including information about weather conditions that have a minor impact on flight safety and information about weather conditions that have a major impact on flight safety. Real-time weather data may include ideal weather conditions such as sunny, cloudy, and windless, as well as adverse weather conditions such as thunderstorms, low visibility, strong winds, and extreme turbulence, all of which may affect flight safety. For example, thunderstorms may cause the risk of UAVs being struck by lightning or hail. The pre-trained weather impact model is input with real-time weather data to output the forecast of the weather conditions on the flight risk of the target UAV. For example, the risk of sunny to cloudy, the risk of strong winds, and the risk of light rain are all risks that have a minor impact on flight safety due to weather conditions, while the risk of lightning strikes, the risk of hail attacks, and the risk of hurricanes are all risks that have a major impact on flight safety due to weather conditions.

[0127] According to the flight risk information of at least one target UAV, the order or number of waypoints in the optimal flight path, as well as the flight altitude and flight speed of at least one target UAV are automatically adjusted.

[0128] If the flight risk information of at least one target UAV is that the weather conditions have little impact on flight safety, the basic flight strategy is executed to automatically maintain the waypoint sequence and number of waypoints for executing the optimal flight path, as well as the flight altitude and flight speed of at least one target UAV.

[0129] If the flight risk information of at least one target UAV is information that the weather conditions have a great impact on flight safety, a dynamic adjustment strategy is executed to automatically adjust the order or number of waypoints for executing the optimal flight path or the flight altitude and flight speed of at least one target UAV.

[0130] In the event of sudden weather, the waypoint sequence can be dynamically modified to avoid severe weather areas while maintaining the original waypoints. For example, a nearby waypoint can be reached first to bypass storms and rainstorms.

[0131] In extreme weather conditions, add new temporary waypoints. By introducing new waypoints, optimize the optimal flight path to ensure the safety of drone flight. For example, if you encounter a strong cyclone, add a waypoint in a safe area to avoid the directly affected airspace. When encountering strong winds, choose to increase the flight altitude or reduce the flight speed to ensure the safety of the drone flight.

[0132] The ground control center can automatically adjust flight strategies according to weather changes without human intervention. It can automatically identify weather conditions and autonomously select flight paths.

[0133] As a feasible approach, data collection strategies can be automatically adjusted based on real-time environment data and task data, including:

[0134] Determine whether the initial geographic data meets the data collection standards based on real-time environmental data and mission data;

[0135] If the initial geographic data does not meet the data collection criteria, adjust the data collection area and increase the data collection frequency based on the real-time environmental data and the current flight position of at least one target UAV;

[0136] A re-collection instruction is generated according to the adjusted data collection area and the increased data collection frequency, and the re-collection instruction is sent to at least one target UAV.

[0137] According to the mission data, the data collection standards are obtained, including clarity, screen size, etc. The impact of the flight environment of the target UAV on data collection is determined through real-time environmental data. During the data collection process, real-time monitoring is implemented to ensure that the data quality meets the data collection standards. If the initial geographic data does not meet the data collection standards, that is, the data quality decreases, noise increases, the picture is blurred, etc., the ground control center will automatically identify and record it. At the same time, the flight path will be re-planned, and the flight path will be readjusted to a good quality area based on the current flight position and environmental conditions of the target UAV, and the data collection frequency will be increased to obtain supplementary data. The adjusted data collection area and the increased data collection frequency will generate a re-collection instruction and send it to the target UAV corresponding to the task to be executed.

[0138] As a feasible method, the initial geographic data includes the data collected by each sensor; the initial geographic data is analyzed and fused to obtain geographic data, including:

[0139] Extracting target features of the collected data of each sensor and geometric features of position and posture changes, wherein the target features include position change rate, acceleration change and ambient light intensity features;

[0140] The principal component analysis algorithm is used to reduce the dimension of target features and geometric features to obtain reduced dimension data;

[0141] Data analysis and fusion are performed on the dimension-reduced data to obtain geographic data.

[0142] Extract relevant features, namely target features, from the data of each sensor, including position change rate, acceleration change, and ambient light intensity features. Combined with the feature point information in the visual data, extract the geometric features of position and posture changes, including aspect ratio, square perimeter to area ratio, feature angle, center of mass position, etc.

[0143] The principal component analysis algorithm is used to reduce the dimension of the target features and geometric features, retaining the main information for subsequent analysis. It mainly includes: 1. Standardized data: In order to eliminate the influence of dimension, all features are standardized so that their mean is 0 and variance is 1. 2. Calculate the covariance matrix: Calculate the covariance matrix from the standardized data to understand the relationship between each feature. The covariance matrix reflects the common changes of the variables. 3. Calculate eigenvalues ​​and eigenvectors: Perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​and corresponding eigenvectors. The eigenvalue represents the amount of information carried by each principal component. 4. Select the principal component: Select the first k principal components based on the size of the eigenvalue. Usually, the principal component with a cumulative contribution rate of 85% to 95% is selected to ensure that sufficient information is retained. 5. Convert the original data: Project the standardized original data onto the selected principal component to form a reduced-dimensional data set, that is, reduced-dimensional data. After obtaining the reduced-dimensional data, perform data analysis and fusion on the reduced-dimensional data to obtain geographic data.

[0144] As a feasible approach, data analysis and fusion are performed on the dimension-reduced data to obtain geographic data, including:

[0145] Analyze the reduced-dimensional data through a pre-trained learning model to obtain target geographic data;

[0146] If the target geographic data is data collected by sensors of different types, a consistency check is performed on the target geographic data to obtain corrected geographic data;

[0147] If the measurement data in the target geographic data contains noise, the target geographic data is filtered using a Kalman filter algorithm to obtain filtered geographic data;

[0148] If the target geographic data is data collected by multiple sensors of the same type, the target geographic data is fused by a preset weighted formula to obtain fused geographic data;

[0149] The corrected geographic data, filtered geographic data and fused geographic data corresponding to different sensors are combined to obtain geographic data.

[0150] The training process of the learning model is as follows: select a decision tree as the machine learning model based on the data characteristics and fusion objectives. For time series data, a long short-term memory network (LSTM) model can be used. Obtain sample training data, divide the data into training sets and test sets, use the training set to train the model, and adjust the hyperparameters to improve model performance. Use the test set to evaluate the performance of the model and calculate evaluation indicators such as accuracy, precision, and recall. Use cross-validation and grid search techniques to optimize the model.

[0151] The reduced-dimensional data is analyzed through a pre-trained learning model to deeply understand and extract the reduced-dimensional data, obtain the target geographic data, and provide support and optimization for subsequent data fusion and decision-making.

[0152] If the target geographic data is collected by sensors of different types, the consistency check is performed on the target geographic data and a consistency threshold is set. This threshold can be set based on specific application scenarios, sensor accuracy requirements, historical data and other factors. The mean and standard deviation of the output values ​​of the same type of sensors are calculated, and the difference between the output value of each sensor and the mean is calculated. If the difference exceeds the consistency threshold, it is considered that the data has a large deviation. The weight coefficient of the problematic sensor is reduced, or its data is temporarily removed to obtain the corrected geographic data to avoid negative impact on the final fusion result.

[0153] If the measurement data in the target geographic data contains noise, the Kalman filtering algorithm is used to filter the target geographic data to obtain filtered geographic data.

[0154] Let the state transfer matrix be F, and the previous state be X k-1 , then the current state prediction is X k =F*X k-1 ; After receiving new sensor data, calculate the difference between the current measured value and the predicted value, that is, the innovation N k =Z k -H*X k , where H is the observation matrix, Z k is the actual measured value. According to the covariance of the error covariance matrix and the measurement noise, the Kalman gain K is calculated. k =P k-1 *H T *(H*P k-1 *H T +R) -1 , where R is the covariance matrix of the measurement noise, H T is the transposed matrix of H to determine how to adjust the measurement information into the state prediction. Combined with the innovative weighted update of the current state estimate: X k =X k +K k *N k , N k is the measurement noise. Update the error covariance matrix P k =(IK k *H)*P k-1 , where I is the identity matrix, P k is the current error covariance matrix, P k-1is the error covariance matrix of the previous state. Through the above steps, the Kalman filter can effectively process the noisy sensor data, obtain the filtered geographic data, and improve the accuracy of the final fusion result.

[0155] If the target geographic data is data collected by multiple sensors of the same type, the target geographic data is fused by a preset weighting formula to obtain fused geographic data; wherein the preset weighting formula can be expressed as:

[0156]

[0157] Among them, Z represents the fused geographic data, u j represents the target geographic data, v j It represents the weight coefficient of the sensor, and n represents the number of sensors of the same type.

[0158] Corrected geographic data, filtered geographic data and fused geographic data are high-quality geographic data. Corrected geographic data, filtered geographic data and fused geographic data corresponding to different sensors are merged to obtain the final geographic data, and the position and posture are corrected, thereby improving the accuracy and efficiency of data processing.

[0159] It should be understood that although Figure 2 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in the application, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0160] Figure 3 A schematic diagram of a geographic data collection device provided in an embodiment of the present application, for performing the following steps: Figure 2 The method flow shown in Figure 3 As shown, the geographic data collection device 300 may include: an acquisition module 310, a matching module 320, an adjustment module 340 and a fusion module 350. The main functions of each component module are as follows:

[0161] An acquisition module 310 is used to acquire task data of the task to be executed, equipment data of the candidate drones, real-time weather data, and real-time environmental data;

[0162] A matching module 320 is used to match at least one target drone that performs the task to be performed from the candidate drones according to the task data and the equipment data;

[0163] A determination module 330, for determining an optimal flight path of at least one target UAV based on the mission data and the equipment data;

[0164] An adjustment module 340, for collecting initial geographic data based on the optimal flight path of at least one target UAV, and automatically adjusting the optimal flight path and data collection strategy according to real-time weather data, real-time environmental data and mission data;

[0165] The fusion module 350 is used to perform data analysis and fusion on the initial geographic data to obtain geographic data.

[0166] As an implementable manner, the matching module 320 is specifically used to: determine the task type of the task to be performed based on the task data of the task to be performed; evaluate the requirement information of the task to be performed based on the task type of the task to be performed; determine the priority of the task to be performed based on the requirement information of the task to be performed; evaluate the overall capability score of the UAV based on the equipment data of the UAV; and match at least one target UAV that performs the task to be performed from the alternative UAVs based on the priority of the task to be performed and the overall capability score of the UAV.

[0167] As an implementable manner, the determination module 330 is specifically used to: define an objective function, decision variables and constraints based on mission data and equipment data, wherein the decision variables are a plurality of waypoints that the target UAV needs to pass through; generate at least one initial population based on the plurality of waypoints, wherein the initial population includes initial individuals composed of a plurality of waypoints in different arrangement orders; based on the objective function, evaluate the fitness of at least one initial individual; based on the fitness of at least one initial individual, generate a new generation of individuals of at least one initial individual, until a preset number of iterations is reached, and the fitness fluctuation value of at least one initial individual is less than a preset threshold, and output the optimal flight path of at least one target UAV.

[0168] As an achievable method, the adjustment module 340 is specifically used to: predict the flight risk information of at least one target UAV based on real-time weather data and a pre-trained weather impact model; and automatically adjust the order of waypoints or the number of waypoints in the optimal flight path or the flight altitude and flight speed of at least one target UAV based on the flight risk information of at least one target UAV.

[0169] As an achievable method, the adjustment module 340 is specifically used to: determine whether the initial geographic data meets the data collection standard based on the real-time environmental data and the mission data; if the initial geographic data does not meet the data collection standard, adjust the data collection area and increase the data collection frequency based on the real-time environmental data and the current flight position of at least one target UAV; generate a re-collection instruction based on the adjusted data collection area and the increased data collection frequency, and send the re-collection instruction to at least one target UAV.

[0170] As an implementable manner, the initial geographic data includes data collected by various sensors; the fusion module 350 is specifically used to: extract target features of the data collected by various sensors and geometric features of position and posture changes, wherein the target features include position change rate, acceleration change and ambient light intensity features;

[0171] The principal component analysis algorithm is used to reduce the dimension of target features and geometric features to obtain reduced-dimensional data; the reduced-dimensional data is analyzed and fused to obtain geographic data.

[0172] As a feasible method, the fusion module 350 is specifically used to: analyze the reduced dimension data through a pre-trained learning model to obtain target geographic data; if the target geographic data is data collected by sensors of different types, perform a consistency check on the target geographic data to obtain corrected geographic data; if the measurement data in the target geographic data is noisy, use a Kalman filter algorithm to filter the target geographic data to obtain filtered geographic data; if the target geographic data is data collected by multiple sensors of the same type, fuse the target geographic data through a preset weighted formula to obtain fused geographic data; merge the corrected geographic data, filtered geographic data and fused geographic data corresponding to different sensors to obtain geographic data.

[0173] The same or similar parts between the above embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0174] According to an embodiment of the present application, the present application also provides a computer device and a computer-readable storage medium.

[0175] like Figure 4 , is a block diagram of a computer device according to an embodiment of the present application. The computer device is intended to represent various forms of digital computers or mobile devices. The digital computer may include a desktop computer, a portable computer, a workbench, a personal digital assistant, a server, a mainframe computer, and other suitable computers. The mobile device may include a tablet computer, a smart phone, a wearable device, etc.

[0176] like Figure 4 As shown, computer device 400 includes computing unit 401, ROM 402, RAM 403, bus 404, and input / output (I / O) interface 405, and computing unit 401, ROM 402, and RAM 403 are connected to each other through bus 404. Input / output (I / O) interface 405 is also connected to bus 404.

[0177] The computing unit 401 can perform various processes in the method embodiment of the present application according to the computer instructions stored in the read-only memory (ROM) 402 or the computer instructions loaded from the storage unit 408 to the random access memory (RAM) 403. The computing unit 401 can be various general and / or special processing components with processing and computing capabilities. The computing unit 401 may include, but is not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. In some embodiments, the method provided in the embodiment of the present application can be implemented as a computer software program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 408.

[0178] RAM 403 may also store various programs and data required for the operation of computer device 400. Part or all of the computer program may be loaded and / or installed on computer device 400 via ROM 402 and / or communication unit 409.

[0179] The input unit 406, output unit 407, storage unit 408 and communication unit 409 in the computer device 400 can be connected to the I / O interface 405. The input unit 406 can be, for example, a keyboard, a mouse, a touch screen, a microphone, etc. The output unit 407 can be, for example, a display, a speaker, an indicator light, etc. The computer device 400 can exchange information, data, etc. with other devices through the communication unit 409.

[0180] It should be noted that the device may also include other components necessary for normal operation, or may only include components necessary for implementing the solution of the present application, rather than all the components shown in the figure.

[0181] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof.

[0182] The computer instructions for implementing the method of the present application can be written in any combination of one or more programming languages. These computer instructions can be provided to the computing unit 401, so that when the computer instructions are executed by the computing unit 401 such as a processor, the steps involved in the method embodiment of the present application are executed.

[0183] The computer-readable storage medium provided in the present application may be a tangible medium that may contain or store computer instructions for executing the steps involved in the method embodiments of the present application. The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, and other forms of storage media.

[0184] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A geographic data collection method, characterized in that: The method comprises: Obtain mission data for missions to be performed, equipment data for candidate drones, real-time weather data, and real-time environmental data; According to the task data and the equipment data, matching at least one target drone that performs the task to be performed from the candidate drones; Determining an optimal flight path for the at least one target UAV based on the mission data and the device data; Based on the optimal flight path of the at least one target UAV, initial geographic data is collected, and the optimal flight path and data collection strategy are automatically adjusted according to the real-time weather data, the real-time environmental data and the mission data; The initial geographic data is analyzed and integrated to obtain geographic data.

2. The method according to claim 1, characterized in that The step of matching at least one target drone for executing the task to be executed from the candidate drones according to the task data and the equipment data includes: Determining the task type of the task to be executed according to the task data of the task to be executed; Evaluate the requirement information of the task to be performed according to the task type of the task to be performed; Determining the priority of the task to be executed according to the demand information of the task to be executed; Assessing an overall capability score of the drone based on the equipment data of the drone; Based on the priority of the task to be performed and the overall capability score of the drone, at least one target drone for performing the task to be performed is matched from the candidate drones.

3. The method according to claim 1, characterized in that Determining the optimal flight path of the at least one target UAV according to the mission data and the equipment data includes: Defining an objective function, decision variables, and constraints according to the mission data and the device data, wherein the decision variables are a plurality of waypoints that the target UAV needs to pass through; Generating at least one initial population according to the plurality of waypoints, wherein the initial population includes initial individuals composed of the plurality of waypoints in different arrangement orders; Based on the objective function, evaluating the fitness of the at least one initial individual; According to the fitness of the at least one initial individual, a new generation of individuals of the at least one initial individual is generated until a preset number of iterations is reached, and the fitness fluctuation value of the at least one initial individual is less than a preset threshold, and the optimal flight path of the at least one target UAV is output.

4. The method according to claim 1, characterized in that The automatically adjusting the optimal flight path according to the real-time weather data comprises: According to the real-time weather data, predicting the flight risk information of the at least one target UAV based on a pre-trained weather impact model; According to the flight risk information of the at least one target UAV, the order or number of waypoints in the optimal flight path or the flight altitude and flight speed of the at least one target UAV are automatically adjusted.

5. The method according to claim 1, characterized in that The automatically adjusting the data collection strategy according to the real-time environment data and the task data includes: Determining whether the initial geographic data meets data collection standards based on the real-time environment data and the task data; If the initial geographic data does not meet the data collection criteria, adjusting the data collection area and increasing the data collection frequency according to the real-time environmental data and the current flight position of the at least one target UAV; A re-collection instruction is generated according to the adjusted data collection area and the increased data collection frequency, and the re-collection instruction is sent to the at least one target UAV.

6. The method according to claim 1, characterized in that The initial geographic data includes data collected by various sensors; the data analysis and fusion of the initial geographic data to obtain geographic data includes: Extracting target features of the collected data of each sensor and geometric features of position and posture changes, wherein the target features include position change rate, acceleration change and ambient light intensity features; Using a principal component analysis algorithm to reduce the dimension of the target feature and the geometric feature to obtain reduced dimension data; The dimension-reduced data are subjected to data analysis and fusion to obtain geographic data.

7. The method according to claim 6, characterized in that The step of performing data analysis and fusion on the dimension reduction data to obtain geographic data includes: Analyzing the dimension-reduced data through a pre-trained learning model to obtain target geographic data; If the target geographic data is data collected by sensors of different types, a consistency check is performed on the target geographic data to obtain corrected geographic data; If the measurement data in the target geographic data contains noise, the target geographic data is filtered using a Kalman filter algorithm to obtain filtered geographic data; If the target geographic data is data collected by multiple sensors of the same type, the target geographic data are fused by a preset weighting formula to obtain fused geographic data; The corrected geographic data, filtered geographic data and fused geographic data corresponding to different sensors are combined to obtain geographic data.

8. A geographic data collection device, characterized in that: The device comprises: The acquisition module is used to obtain the task data of the task to be executed, the equipment data of the candidate drones, the real-time weather data and the real-time environmental data; A matching module, configured to match at least one target drone for executing the task to be executed from the candidate drones according to the task data and the device data; A determination module, configured to determine an optimal flight path of the at least one target UAV according to the mission data and the device data; an adjustment module, configured to collect initial geographic data based on the optimal flight path of the at least one target UAV, and automatically adjust the optimal flight path and data collection strategy according to the real-time weather data, the real-time environmental data and the mission data; The fusion module is used to perform data analysis and fusion on the initial geographic data to obtain geographic data.

9. A computer device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores computer instructions that can be executed by the at least one processor, and the computer instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: The computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 7.

Images