UAV conflict prediction method based on safety interval progressive decision tree

Abstract

The present invention discloses a method for predicting UAV conflicts based on a safety interval progressive decision tree, including: establishing a UAV safety interval parameter library; establishing a coordinate conversion model parameter library for converting WGS-84 coordinate system and GLONASS-PZ90 coordinate system to CGCS2000 coordinate system, and encapsulating the constructed Bursa model as a coordinate conversion function; parsing flight dynamic data and performing standardized conversion; predicting airspace UAV conflicts, and dividing the prediction results into a data set to be observed, a data set of key attention, and a data set of warnings according to the conflict risk level; calculating the time when the UAV pairs in the data set of key attention may reach the warning level; calculating the time when the conflicting UAV pairs in the warning data set are expected to collide. The present invention can maximize the utilization of airspace space resources, thereby effectively improving the flight flow in the airspace.

Description

Technical Field

[0001] The invention belongs to the technical field of low-altitude aircraft flight management, and in particular relates to a UAV conflict prediction method based on a safety interval progressive decision tree. Background Art

[0002] At present, the basic classification method of airspace divides airspace into seven categories based on many factors such as aircraft flight rules and performance requirements, airspace environment, and air traffic control service content; among them, Class W airspace (true altitude below 120 meters) is determined to be suitable for micro, light, and small unmanned aircraft. In this context, various cities and municipalities have actively responded and formulated implementation measures for drone flight management, clearly stipulating that there is no need to apply for a flight plan separately except for prohibited flight airspace below 120 meters.

[0003] At present, in most low-altitude drone-suitable airspaces, there are no public routes, which makes drones lack clear route planning and guidance when flying in the airspace. In particular, ordinary consumers use drones for non-commercial purposes such as personal entertainment, photography, and video shooting, which often tend to be disorderly. This situation may not only increase the risk of collision between drones and other aircraft, but also bring certain challenges to airspace management. As the number of drone flights continues to increase, the above situation will occur more and more frequently. Therefore, there is an urgent need for a method that can effectively identify drone flight risks in airspace based on the safe interval of drone flight. Through this method, potential conflicts can be identified in advance, and the conflict risks can be submitted in advance to the relevant management departments, drone operators or individual drone users, so that relevant personnel can take measures to control and deal with risks as soon as possible, thereby effectively ensuring the safety of drone operations in low-altitude airspace. Summary of the invention

[0004] In view of the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a UAV conflict prediction method based on a safety interval progressive decision tree to solve the problem of conflict identification and early warning when UAVs fly disorderly in airspace in the prior art. The method of the present invention establishes a UAV safety interval model and a conflict risk level according to the requirement that UAVs must maintain an appropriate safety distance during flight. Through the unified processing of the UAV flight dynamic data, the position, speed, altitude and other information of the UAV are determined to determine whether the distance between the UAVs reaches the safety interval threshold of different conflict risk levels. The present invention adopts a progressive decision tree method to classify the UAV pairs with risks into three types of data sets: to be observed, focused, and alarm prompts, and predict the collision time of the UAV pairs in the alarm prompt data set, so that users can solve key conflicts in an orderly and timely manner according to different risk levels, and effectively avoid the occurrence of UAV collisions.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] The present invention provides a UAV conflict prediction method based on a safety interval progressive decision tree, and the steps are as follows:

[0007] 1) Calculate the safe interval distance of drones based on drone performance parameters and establish a drone safety interval parameter library;

[0008] 2) Establish a coordinate conversion model parameter library for converting the WGS-84 coordinate system and GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system, encapsulate the constructed Bursa model into a coordinate conversion function, and define the function calling method;

[0009] 3) Receive the flight dynamics data sent by the drone, parse the data and perform standardized conversion, and store the conversion results in the drone flight dynamics database;

[0010] 4) The classification method of progressive decision tree is used to predict airspace UAV conflicts, and the prediction results are divided into observation data set, focus data set, and warning data set according to the conflict risk level;

[0011] 5) Calculate the time when the UAV pairs in the focus dataset may reach the warning level to prompt users to plan conflict resolution in advance; calculate the expected collision time of the conflicting UAV pairs in the warning dataset to prompt users to resolve the conflict.

[0012] Furthermore, the step 1) specifically includes:

[0013] 11) Establish a UAV performance parameter table, including the following parameters: UAV model, total thrust, vertical forward tilt angle, UAV mass, maximum horizontal speed, and maximum vertical speed;

[0014] 12) Calculate the safe separation distance of the drone, including the following steps:

[0015] 121) Obtain the total thrust of the drone motor / rotor from the drone performance parameter table , the angle of the UAV motor / rotor tilting forward in the vertical direction is Φ, calculate the maximum thrust of the UAV in the horizontal direction and maximum vertical thrust , the calculation formula is as follows;

[0016] ;

[0017]

[0018] 122) Get the drone quality from the drone performance parameter table , calculate the horizontal acceleration of the drone and the vertical acceleration , the calculation formula is as follows:

[0019] ;

[0020] ;

[0021] in, g is the acceleration due to gravity;

[0022] 123) Get the maximum horizontal speed of the drone from the drone performance parameter table and the maximum vertical speed , calculate the safe distance of the drone in the horizontal direction And vertical safety distance , the calculation formula is as follows:

[0023] ;

[0024] ;

[0025] in, and are the redundancy time in the horizontal and vertical directions respectively; and Configure parameters for safety intervals in horizontal and vertical directions respectively;

[0026] 13) Establish a drone safety interval parameter library, the parameter items include: drone model, drone mass, horizontal safety interval and vertical safety interval; save the calculation results in step 12) in the drone safety interval parameter library.

[0027] Furthermore, the step 2) specifically includes:

[0028] 21) Establish a coordinate transformation model parameter library, the parameter items include: original coordinate system type, new coordinate system type, geographical scope of application, translation parameters, rotation parameters and scale parameters;

[0029] 22) Establish coordinate conversion function and define function calling method;

[0030] 221) Construct the Bursa model as follows:

[0031] ;

[0032] ;

[0033] ;

[0034] in, is the translation parameter, is the scale parameter, is the rotation parameter; The original coordinates of the drone’s position in the WGS-84 or GLONASS-PZ90 coordinate system; The new coordinates of the UAV position in the CGCS2000 coordinate system;

[0035] 222) Encapsulate the Bursa model in step 221) into a coordinate conversion function, and define the function call method for use in coordinate conversion calculation; the function input parameters are: the original coordinate system type, the new coordinate system type, and the specific position of the drone in the original coordinate system ; The function's output parameter is: the specific position of the drone in the new coordinate system ;

[0036] 223) Use the coordinate conversion function to convert the WGS-84 coordinate system or GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system. The specific steps are as follows:

[0037] The original coordinate system type, new coordinate system type, and specific position of the drone in the original coordinate system in the function input parameters are By comparing the three parameter items of original coordinate system type, new coordinate system type and regional applicable scope in the coordinate transformation model parameter library, the translation parameter, rotation parameter and scale parameter are obtained; when the data record obtained from the coordinate transformation model parameter library is not unique, the coordinate transformation calculation is terminated and the user is prompted with an error; otherwise, the coordinate transformation calculation is continued;

[0038] The specific position of the drone in the original coordinate system , translation parameters, rotation parameters, and scale parameters are substituted into the coordinate conversion function to obtain the specific position of the drone in the new coordinate system , and the specific position of the drone in the new coordinate system The result is returned.

[0039] Furthermore, the step 3) specifically includes:

[0040] 31) Establish a UAV flight dynamics database, the data items include: flight record number, UAV serial number, take-off time, flight status, flight dynamics data reception time, UAV model, new coordinate system type, UAV position on the X axis, UAV position on the Y axis, UAV position on the Z axis, flight speed on the X axis, flight speed on the Y axis, flight speed on the Z axis, track angle, horizontal safety interval and vertical safety interval;

[0041] 32) Receive the flight dynamic data sent by the drone, parse the received flight dynamic data of the drone, and obtain the flight record number, flight status, current time, drone model, coordinate system type, position longitude, position latitude, flight altitude, vertical flight speed, horizontal flight speed and track angle;

[0042] 33) Decompose the flight record number according to the order and number of digits of the string to obtain the drone serial number, take-off time and random code respectively; the flight record number is composed of the drone serial number, take-off time and random code, and the specific format is: 20-digit drone serial number + 8-digit take-off time + 8-digit random code;

[0043] 34) Convert the WGS-84 coordinate system or GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system; specifically:

[0044] 341) The WGS-84 coordinate system, CGCS2000 coordinate system, and GLONASS-PZ90 coordinate system are expressed in the Earth-centered Earth-fixed rectangular coordinate system;

[0045] 3411) Define the position of the drone expressed in latitude, longitude and altitude coordinates as ,as follows:

[0046] ;

[0047] In the formula, B is the location dimension, L is the location longitude, H is the flight altitude;

[0048] 3412) Calculate the radius of curvature of the y-axis , the calculation formula is as follows:

[0049] ;

[0050] in, is the semi-major axis of the reference ellipsoid, e is the first eccentricity;

[0051] 3413) Define the position of the UAV expressed in Earth-centered rectangular coordinates as ;

[0052] The position of the drone expressed in latitude, longitude and altitude coordinates The position of the drone converted into Earth-centered rectangular coordinates , the calculation formula is as follows:

[0053] ;

[0054] ;

[0055] ;

[0056] 342) Determine the coordinate system type. If it is WGS-84 or GLONASS-PZ90, perform coordinate conversion calculation. The specific steps are: convert the obtained UAV position represented by the Earth-centered Earth-fixed rectangular coordinates , coordinate system type, and new coordinate system type are substituted into the coordinate conversion function to obtain the specific position of the UAV in the CGCS2000 coordinate system. , and set the drone position in the CGCS2000 coordinate system ; If it is CGCS2000 coordinate system, ;

[0057] 35) Calculate the speed of the drone on the X-axis, Y-axis, and Z-axis respectively ;

[0058] ;

[0059] ;

[0060] ;

[0061] in, is the vertical flight speed, is the horizontal flight speed, is the track angle;

[0062] Track angle It is the angle between the projection of the actual motion trajectory of the drone on the horizontal plane and the geographic north direction, measured in a clockwise direction, ranging from 0 to 360 degrees;

[0063] 36) Use the drone model obtained in step 32) to match the drone model parameter item in the drone safety interval parameter library to obtain the horizontal safety interval and the vertical safety interval; when the data record obtained from the drone safety interval parameter library is not unique, an error is prompted and the method ends; otherwise, execute step 37);

[0064] 37) Save the standardized converted data into the UAV flight dynamic database. The specific correspondence of the saved data is as follows:

[0065] The flight record number of the UAV flight dynamic database = the flight record number obtained in step 32);

[0066] The drone serial number in the drone flight dynamic database = the drone serial number obtained in step 32);

[0067] The take-off time of the UAV flight dynamic database = the take-off time obtained in step 33);

[0068] The flight status of the UAV flight dynamic database = the flight status obtained in step 32);

[0069] The flight dynamics data receiving time of the UAV flight dynamics database = the current time obtained in step 32);

[0070] The drone model of the drone flight dynamic database = the drone model obtained in step 32);

[0071] New coordinate system type for UAV flight dynamic database = CGCS2000;

[0072] The X-axis position of the drone in the drone flight dynamic database = obtained in step 342 ;

[0073] The Y-axis position of the drone in the drone flight dynamic database = obtained in step 342 ;

[0074] The Z-axis position of the drone in the drone flight dynamic database = obtained in step 342 ;

[0075] The flight speed on the X axis of the UAV flight dynamic database = obtained in step 35) ;

[0076] The flight speed on the Y axis of the UAV flight dynamic database = obtained in step 35) ;

[0077] The flight speed on the Z axis of the UAV flight dynamic database = obtained in step 35) ;

[0078] The track angle of the UAV flight dynamic database = the track angle obtained in step 32);

[0079] The horizontal safety interval of the UAV flight dynamic database = the horizontal safety interval obtained in step 36;

[0080] The vertical safety interval of the UAV flight dynamic database = step 36) obtains the vertical safety interval.

[0081] Furthermore, the step 4) specifically includes:

[0082] 41) Define the UAV conflict risk level based on the safety interval, and divide the conflict risk level into: to be observed, key attention, and warning prompt according to the judgment criteria configured by the user;

[0083] To be observed: The risk level is low, indicating that some key indicators show the possibility of potential risk escalation, but it is uncertain whether it will further develop into a higher risk state, and its changing trend will be continuously observed;

[0084] Focus: The risk level is medium, indicating that most key indicators show the possibility of potential risk escalation, and users need to pay close attention to its dynamic changes;

[0085] Warning: A high risk level indicates that the risk is obvious and has a great impact on the target, requiring users to take immediate measures to control and handle the risk;

[0086] 42) Receive the time range provided by the user, match the flight dynamic data receiving time data item in the UAV flight dynamics database according to the time range, and form a UAV test data set ;

[0087] ;

[0088] ;

[0089] in, and middle is the total number of matched data records; It is a UAV flight dynamic data column. Each data item in the data column comes from a specific data item in the UAV flight dynamic database, as follows:

[0090] The flight record number, is the position of the drone on the X axis, is the position of the drone on the Y axis, is the position of the drone on the Z axis, is the flight speed on the X axis, is the flight speed on the Y axis, is the flight speed on the Z axis, is the horizontal safety interval, is the horizontal safety interval, is the vertical safety interval, is the track angle;

[0091] 43) UAV test dataset As the sample set of the decision tree, the decision tree set is established with the feature that the distance between the positions of two drones on the X, Y, and Z axes in the sample set is less than or equal to the sum of the safety intervals between the two drones on the X, Y, and Z axes. , the specific steps are as follows:

[0092] For drone test dataset Any two points and , , the distances on the X, Y, and Z axes are , , ; The sum of the safety intervals on the X, Y, and Z axes are , , ;

[0093] Calculate the decision tree set ,as follows:

[0094] ;

[0095] 44) Ensemble decision trees As a new sample set, the drone The distance between the two drones on the X, Y, and Z axes is less than or equal to the safety interval between the two drones on the X, Y, and Z axes. The decision tree set is calculated progressively. , , , the specific steps are as follows:

[0096] ;

[0097] ;

[0098] ;

[0099] 45) According to the conflict risk level, establish the dataset to be observed, the dataset of key concerns, and the dataset of warning prompts;

[0100] 451) Establish an alarm prompt data set. When the drone meets the decision tree set , , When all three items are combined, the drone belongs to the warning prompt data set , the decision tree set , , As a sample data set, calculate the alarm prompt data set ,as follows:

[0101] ;

[0102] 452) Establish a focused data set, when the drone only needs to satisfy the decision tree set , , When any two-item set is , the decision tree set , , As a sample data set, calculate the focus data set ,as follows:

[0103] ;

[0104] 453) Establish a data set to be observed. When the drone pair only needs to satisfy the decision tree set , , When any item of the set is , the decision tree set , , As a sample data set, calculate the data set to be observed ,as follows:

[0105] .

[0106] Furthermore, the step 5) specifically includes:

[0107] 51) Computational focus on data sets The specific steps are as follows:

[0108] We will focus on the dataset As a sample dataset;

[0109] When focusing on datasets UAV ,satisfy , flight record number and Time when the warning level may be reached ;

[0110] When focusing on datasets UAV ,satisfy , flight record number and Time when the warning level may be reached ;

[0111] When focusing on datasets UAV ,satisfy , flight record number and Time when the warning level may be reached ;

[0112] Focus on the dataset The estimated time for all drones to reach the warning level ,as follows:

[0113] ;

[0114] 52) Calculate the alarm data set The specific steps for determining the estimated collision time of the conflicting drones are as follows:

[0115] Alarm prompt data set As a sample data set, calculate the UAV pair Estimated time of collision ,as follows:

[0116] ;

[0117] Alarm prompt dataset The estimated collision time of all drone pairs in ,as follows:

[0118] .

[0119] Beneficial effects of the present invention:

[0120] 1. The present invention fully considers the significant differences in performance and characteristics of different drones; on the one hand, it can ensure that there is sufficient time and space to implement emergency measures in an emergency situation, thereby minimizing potential risks; on the other hand, the established safety interval can maximize the utilization of airspace space resources, thereby effectively increasing the flight flow in the airspace.

[0121] 2. The present invention divides the conflict risk levels from low to high into three categories: to be observed, focused, and alarm, and sets different safety interval thresholds according to the risk levels; on the one hand, the risk can be identified early in its infancy, allowing users to intervene as early as possible; on the other hand, users can handle it in a timely and orderly manner according to the determined risk level in order from high to low.

[0122] 3. The present invention uses a progressive decision tree method, which has a relatively simple construction process and can process various types of data. On the one hand, when faced with large-scale data sets, the decision tree can be quickly trained and predicted; on the other hand, through a progressive approach, the calculation of each stage can be relatively independent, thereby improving the parallel efficiency of the operation and the flexibility of the function implementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0123] Figure 1 is a flow chart of the method of the present invention;

[0124] Figure 2Schematic diagram of a drone flying at a safe interval at low altitude in a city. DETAILED DESCRIPTION

[0125] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.

[0126] Reference Figure 1 , Figure 2 As shown, a UAV conflict prediction method based on a safety interval progressive decision tree of the present invention comprises the following steps:

[0127] 1) Calculate the safe interval distance of drones based on drone performance parameters and establish a drone safety interval parameter library;

[0128] Safety interval refers to the certain distance and time interval that needs to be maintained between two drones in order to ensure the safety of drone flight; in the air, safety interval includes longitudinal interval (front and back distance), lateral interval (left and right distance) and vertical interval (up and down height difference).

[0129] 11) Establish a UAV performance parameter table, including UAV model, total thrust, vertical forward tilt angle, UAV mass, maximum horizontal speed, and maximum vertical speed;

[0130] 12) Calculate the safe separation distance of the drone, including the following steps:

[0131] 121) Obtain the total thrust of the drone motor / rotor from the drone performance parameter table , the angle of the UAV motor / rotor tilting forward in the vertical direction is Φ, calculate the maximum thrust of the UAV in the horizontal direction and maximum vertical thrust , the calculation formula is as follows;

[0132] ;

[0133]

[0134] 122) Get the drone quality from the drone performance parameter table , calculate the horizontal acceleration of the drone and the vertical acceleration , the calculation formula is as follows:

[0135] ;

[0136] ;

[0137] in, gis the acceleration due to gravity;

[0138] 123) Get the maximum horizontal speed of the drone from the drone performance parameter table and the maximum vertical speed , calculate the safe distance of the drone in the horizontal direction And vertical safety distance , the calculation formula is as follows:

[0139] ;

[0140] ;

[0141] in, and They are the redundancy time in the horizontal and vertical directions (including reaction time, etc.); and They are the safety interval configuration parameters in the horizontal and vertical directions, which can be configured by the user according to their actual situation;

[0142] 13) Establish a drone safety interval parameter library, the parameter items include drone model, drone mass, horizontal safety interval, and vertical safety interval; save the calculation results in step 12) in the drone safety interval parameter library.

[0143] 2) Establish a coordinate conversion model parameter library for converting the WGS-84 coordinate system and GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system, encapsulate the constructed Bursa model into a coordinate conversion function, and define the function calling method for use in coordinate conversion calculation;

[0144] 21) Establish a coordinate transformation model parameter library, the parameter items include the original coordinate system type, the new coordinate system type, the geographical scope of application, the translation parameter, the rotation parameter, and the scale parameter;

[0145] 22) Establish a coordinate conversion function and define the function calling method for use in coordinate conversion calculation;

[0146] 221) Construct the Bursa model as follows:

[0147] ;

[0148] ;

[0149] ;

[0150] in, is the translation parameter, is the scale parameter, is the rotation parameter; The original coordinates of the drone’s position in the WGS-84 or GLONASS-PZ90 coordinate system; The new coordinates of the UAV position in the CGCS2000 coordinate system;

[0151] Among them, WGS-84 is used in GPS navigation system;

[0152] CGCS2000 is used in BeiDou satellite navigation system;

[0153] GLONASS-PZ90 is used in Russia's GLONASS satellite navigation system;

[0154] 222) Encapsulate the Bursa model in step 221) into a coordinate conversion function, and define the function call method for use in coordinate conversion calculation; the function input parameters are: the original coordinate system type, the new coordinate system type, and the specific position of the drone in the original coordinate system ; The function's output parameter is: the specific position of the drone in the new coordinate system ;

[0155] 223) Use the coordinate conversion function to convert the WGS-84 coordinate system or GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system. The specific steps are as follows:

[0156] The original coordinate system type, new coordinate system type, and specific position of the drone in the original coordinate system in the function input parameters are By comparing the three parameter items of original coordinate system type, new coordinate system type and regional applicable scope in the coordinate transformation model parameter library, the translation parameter, rotation parameter and scale parameter are obtained; when the data record obtained from the coordinate transformation model parameter library is not unique, the coordinate transformation calculation is terminated and the user is prompted with an error; otherwise, the coordinate transformation calculation is continued;

[0157] The specific position of the drone in the original coordinate system , translation parameters, rotation parameters, and scale parameters are substituted into the coordinate conversion function to obtain the specific position of the drone in the new coordinate system , and the specific position of the drone in the new coordinate system The result is returned.

[0158] 3) Receive the flight dynamics data sent by the drone, parse the data and perform standardized conversion, and store the conversion results in the drone flight dynamics database;

[0159] 31) Establish a UAV flight dynamics database, the data items include flight record number, UAV serial number, take-off time, flight status, flight dynamics data reception time, UAV model, new coordinate system type, UAV position on the X axis, UAV position on the Y axis, UAV position on the Z axis, flight speed on the X axis, flight speed on the Y axis, flight speed on the Z axis, track angle, horizontal safety interval and vertical safety interval;

[0160] 32) Receive the flight dynamic data sent by the drone, parse the received flight dynamic data of the drone, and obtain the flight record number, flight status, current time, drone model, coordinate system type, position longitude, position latitude, flight altitude, vertical flight speed, horizontal flight speed and track angle;

[0161] 33) Decompose the flight record number according to the order and number of digits of the string to obtain the drone serial number, take-off time and random code respectively; the flight record number is composed of the drone serial number, take-off time and random code, and the specific format is: 20-digit drone serial number + 8-digit take-off time + 8-digit random code;

[0162] 34) Convert the WGS-84 coordinate system or GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system;

[0163] 341) The WGS-84 coordinate system, CGCS2000 coordinate system and GLONASS-PZ90 coordinate system are expressed in the Earth-centered Earth-fixed rectangular coordinate system (ECEF);

[0164] 3411) Define the position of the drone expressed in latitude, longitude and altitude coordinates as ,as follows:

[0165] ;

[0166] In the formula, B is the location dimension, L is the location longitude, H is the flight altitude;

[0167] 3412) Calculate the radius of curvature of the y-axis , the calculation formula is as follows:

[0168] ;

[0169] in, is the semi-major axis of the reference ellipsoid, e is the first eccentricity;

[0170] 3413) Define the position of the UAV expressed in Earth-centered rectangular coordinates as ;

[0171] The position of the drone expressed in latitude, longitude and altitude coordinates The position of the drone converted into Earth-centered rectangular coordinates , the calculation formula is as follows:

[0172] ;

[0173] ;

[0174] ;

[0175] 342) Determine the coordinate system type. If it is WGS-84 or GLONASS-PZ90, perform coordinate conversion calculation. The specific steps are: convert the obtained UAV position represented by the Earth-centered Earth-fixed rectangular coordinates , coordinate system type, and new coordinate system type are substituted into the coordinate conversion function to obtain the specific position of the UAV in the CGCS2000 coordinate system. , and set the drone position in the CGCS2000 coordinate system ; If it is CGCS2000 coordinate system, ;

[0176] 35) Calculate the speed of the drone on the X-axis, Y-axis, and Z-axis respectively ;

[0177] ;

[0178] ;

[0179] ;

[0180] in, is the vertical flight speed, is the horizontal flight speed, is the track angle;

[0181] Track angle It is the angle between the projection of the actual motion trajectory of the drone on the horizontal plane and the geographic north direction, measured in a clockwise direction, ranging from 0 to 360 degrees;

[0182] 36) Use the drone model obtained in step 32) to match the drone model parameter item in the drone safety interval parameter library to obtain the horizontal safety interval and the vertical safety interval; when the data record obtained from the drone safety interval parameter library is not unique, an error is prompted and the method ends; otherwise, continue to step 37);

[0183] 37) Save the standardized converted data into the UAV flight dynamic database. The specific correspondence of the saved data is as follows:

[0184] The flight record number of the UAV flight dynamic database = the flight record number obtained in step 32);

[0185] The drone serial number in the drone flight dynamic database = the drone serial number obtained in step 32);

[0186] The take-off time of the UAV flight dynamic database = the take-off time obtained in step 33);

[0187] The flight status of the UAV flight dynamic database = the flight status obtained in step 32);

[0188] The flight dynamics data receiving time of the UAV flight dynamics database = the current time obtained in step 32);

[0189] The drone model of the drone flight dynamic database = the drone model obtained in step 32);

[0190] New coordinate system type for UAV flight dynamic database = CGCS2000;

[0191] The X-axis position of the drone in the drone flight dynamic database = obtained in step 342 ;

[0192] The Y-axis position of the drone in the drone flight dynamic database = obtained in step 342 ;

[0193] The Z-axis position of the drone in the drone flight dynamic database = obtained in step 342 ;

[0194] The flight speed on the X axis of the UAV flight dynamic database = obtained in step 35) ;

[0195] The flight speed on the Y axis of the UAV flight dynamic database = obtained in step 35) ;

[0196] The flight speed on the Z axis of the UAV flight dynamic database = obtained in step 35) ;

[0197] The track angle of the UAV flight dynamic database = the track angle obtained in step 32);

[0198] The horizontal safety interval of the UAV flight dynamic database = the horizontal safety interval obtained in step 36;

[0199] The vertical safety interval of the UAV flight dynamic database = step 36) obtains the vertical safety interval.

[0200] 4) The classification method of progressive decision tree is used to predict airspace UAV conflicts, and the prediction results are divided into three types of conflict risk level data sets: to be observed, focused, and warning according to the conflict risk level;

[0201] 41) Define the UAV conflict risk level based on the safety interval, and divide the conflict risk level into: to be observed, key attention, and warning prompt according to the judgment criteria configured by the user;

[0202] To be observed: The risk level is low, indicating that some key indicators show the possibility of potential risk escalation, but it is uncertain whether it will further develop into a higher risk state, and its changing trend will be continuously observed;

[0203] Focus: The risk level is medium, indicating that most key indicators show the possibility of potential risk escalation, and users need to pay close attention to its dynamic changes so as to take timely response measures;

[0204] Warning: The risk level is high, indicating that the risk is obvious and has a great impact on the target. Users need to take immediate measures to control and handle the risk.

[0205] 42) Receive the time range provided by the user (when calculating the UAV conflict in real time, the time is the current time), match the flight dynamic data receiving time data item in the UAV flight dynamics database according to the time range, and form a UAV test data set ;

[0206] ;

[0207] ;

[0208] in, and middle is the total number of matched data records; It is a UAV flight dynamic data column. Each data item in the data column comes from a specific data item in the UAV flight dynamic database, as follows:

[0209] The flight record number, is the position of the drone on the X axis, is the position of the drone on the Y axis, is the position of the drone on the Z axis, is the flight speed on the X axis, is the flight speed on the Y axis, is the flight speed on the Z axis, is the horizontal safety interval, is the horizontal safety interval, is the vertical safety interval, is the track angle;

[0210] 43) UAV test dataset As the sample set of the decision tree, the decision tree set is established with the feature that the distance between the positions of two drones on the X, Y, and Z axes in the sample set is less than or equal to the sum of the safety intervals between the two drones on the X, Y, and Z axes. , the specific steps are as follows:

[0211] For drone test dataset Any two points and , , the distances on the X, Y, and Z axes are , , ; The sum of the safety intervals on the X, Y, and Z axes are , , ;

[0212] Calculate the decision tree set ,as follows:

[0213] ;

[0214] 44) Ensemble decision trees As a new sample set, the drone The distance between the two drones on the X, Y, and Z axes is less than or equal to the safety interval between the two drones on the X, Y, and Z axes (the value is the larger safety interval between the two drones), and the decision tree set is progressively calculated. , , , the specific steps are as follows:

[0215] ;

[0216] ;

[0217] ;

[0218] 45) According to the conflict risk level, establish the dataset to be observed, the dataset of key concerns, and the dataset of warning prompts;

[0219] 451) Establish an alarm prompt data set. When the drone meets the decision tree set , , When all three items are combined, the drone belongs to the warning prompt data set , the decision tree set , , As a sample data set, calculate the alarm prompt data set ,as follows:

[0220] ;

[0221] 452) Establish a focused data set, when the drone only needs to satisfy the decision tree set , , When any two-item set is , the decision tree set , , As a sample data set, calculate the focus data set ,as follows:

[0222] ;

[0223] 453) Establish a data set to be observed. When the drone pair only needs to satisfy the decision tree set , , When any item of the set is , the decision tree set , , As a sample data set, calculate the data set to be observed ,as follows:

[0224] .

[0225] 5) Computational focus on data sets The time when the drone pair in the system may reach the warning prompt level is used to prompt the user to plan the conflict resolution plan in advance; calculate the warning prompt data set The estimated time of collision between conflicting drones can be used to prompt the user to resolve the conflict;

[0226] 51) Computational focus on data sets51) Computational focus on data sets The specific steps are as follows:

[0227] We will focus on the dataset As a sample dataset,

[0228] When focusing on datasets UAV ,satisfy , flight record number and Time when the warning level may be reached ;

[0229] When focusing on datasets UAV ,satisfy , flight record number and Time when the warning level may be reached ;

[0230] When focusing on datasets UAV ,satisfy , flight record number and Time when the warning level may be reached ;

[0231] Focus on the dataset The estimated time for all drones to reach the warning level ,as follows:

[0232] ;

[0233] 52) Calculate the alarm data set The specific steps for determining the estimated collision time of the conflicting drones are as follows:

[0234] Alarm prompt data set As a sample data set, calculate the UAV pair Estimated time of collision ,as follows:

[0235] ;

[0236] Alarm prompt dataset The estimated collision time of all drone pairs in ,as follows:

[0237] .

[0238] The present invention has many specific application paths. The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements can be made without departing from the principle of the present invention. These improvements should also be regarded as the protection scope of the present invention.

Claims

1. A UAV conflict prediction method based on a safety interval progressive decision tree, characterized in that: Here are the steps: 1) Calculate the safe interval distance of the drone based on the drone performance parameters and establish a drone safety interval parameter library; 2) Establish a coordinate conversion model parameter library for converting the WGS-84 coordinate system and the GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system, encapsulate the constructed Bursa model into a coordinate conversion function, and define the function calling method; 3) Receive the flight dynamics data sent by the drone, parse the data and perform standardized conversion, and store the conversion results in the drone flight dynamics database; 4) The classification method of progressive decision tree is used to predict airspace UAV conflicts, and the prediction results are divided into observation data set, focus data set, and warning data set according to the conflict risk level; 5) Calculate the time when the drone pairs in the focus data set may reach the warning level, so as to prompt users to plan conflict resolution solutions in advance; Calculate the estimated collision time of conflicting drone pairs in the warning prompt dataset to prompt users to resolve the conflict; The step 4) specifically includes: 41) Define the UAV conflict risk level based on the safety interval, and divide the conflict risk level into: waiting to be observed, key attention, and warning prompt according to the judgment criteria configured by the user; To be observed: The risk level is low, indicating that some key indicators show the possibility of potential risk escalation, but it is uncertain whether it will further develop into a higher risk state, and its changing trend will be continuously observed; Focus: The risk level is medium, indicating that most key indicators show the possibility of potential risk escalation, and users need to pay close attention to its dynamic changes; Warning: A high risk level indicates that the risk is obvious and has a great impact on the target, requiring users to take immediate measures to control and handle the risk; 42) Receive the time range provided by the user, match the flight dynamic data receiving time data item in the UAV flight dynamic database according to the time range, and form the UAV test data set D c ; <h2 style=";text-align:left;direction:ltr">D<h2 style=";text-align:left;direction:ltr"> c <h2 style=";text-align:left;direction:ltr"> (W1, W2, W3, W4, W5, W6…W)<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr"> ,|n=1,2,3…N} W n (XH n ,SX n ,SY n ,SZ n ,VX n ,VY n ,VZ n IX n ,IY n ,IZ n ,AN n |m=1,2,3...N) Among them, D c and W n Where N is the total number of matched data records; W n It is a UAV flight dynamic data column. Each data item in the data column comes from a specific data item in the UAV flight dynamic database, as follows: XH n The flight record number is SX n is the position of the drone on the X axis, SY n is the position of the drone on the Y axis, SZ n is the position of the drone on the Z axis, VX n is the flight speed on the X axis, VY n is the flight speed on the Y axis, VZ n is the flight speed on the Z axis, IX n is the safety interval of the horizontal X axis, IY n is the safety interval of the horizontal Y axis, IZ n is the vertical safety interval, AN n is the track angle; 43) The drone test dataset D c As the sample set of the decision tree, the decision tree set D is established with the feature that the distance between the positions of two drones on the X, Y, and Z axes in the sample set is less than or equal to the sum of the safety intervals between the two drones on the X, Y, and Z axes. y , the specific steps are as follows: For the drone test dataset D c Any two points W i and W j , i≠j, i∈n, j∈n, the distances on the X, Y, and Z axes are DX ij =|SX i -SX j |, DY ij =|SY i -SY j |, DZ ij =|SZ i -SZ j |; The sum of the safety intervals on the X, Y, and Z axes is IX respectively ij =IX i +IX j , IY ij =IY i +IY j , IZ ij =IZ i +IZ j ; Calculate the decision tree set D v ,as follows: D y ={(W i ,W j )|(DX ij ≤IX ij )and(DY ij ≤,Y ij )and(DZ ij ≤IZ ij )}; 44) Set the decision tree set D y As a new sample set, the drone pairs (W i , W j ) is characterized by the mutual distance between the two drones on the X, Y, and Z axes being less than or equal to the safety interval between the two drones on the X, Y, and Z axes, and the decision tree set is progressively calculated. The specific steps are as follows: 45) According to the conflict risk level, establish the data set to be observed, the data set of key attention, and the data set of warning prompts; The step 45) specifically includes: 451) Establish an alarm prompt data set. When the drone meets the decision tree set When all three items are combined, the drone belongs to the warning prompt dataset D gj , the decision tree set As a sample data set, calculate the alarm prompt data set D gj ,as follows: 452) Establish a focused data set, when the drone only needs to meet the decision tree set When any two-item set is used, the drone pair belongs to the focus data set Dgz, and the decision tree set As a sample data set, the focus data set Dgz is calculated as follows: 453) Establish a data set to be observed. When the drone only needs to satisfy the decision tree set When any item of the set is As a sample data set, the observed data set Dgc is calculated as follows:

2. The UAV conflict prediction method based on safety interval progressive decision tree according to claim 1 is characterized in that: The step 1) specifically includes: 11) Establish a UAV performance parameter table, the parameters include: UAV model, total thrust, vertical forward tilt angle, UAV mass, maximum horizontal speed and maximum vertical speed; 12) Calculate the safe separation distance of drones; 13) Establish a drone safety interval parameter library, the parameter items include: drone model, drone mass, horizontal safety interval and vertical safety interval; save the calculation results in step 12) in the drone safety interval parameter library.

3. The UAV conflict prediction method based on safety interval progressive decision tree according to claim 2 is characterized in that: The step 12) specifically includes: 121) Obtain the total thrust F of the drone motor / rotor from the drone performance parameter table, the angle of the drone motor / rotor tilted forward in the vertical direction is Φ, and calculate the maximum thrust F of the drone in the horizontal direction h and the maximum vertical thrust F v , the calculation formula is as follows; F v =F cosΦ F h =F sinΦ; 122) Obtain the drone mass M from the drone performance parameter table and calculate the drone's horizontal acceleration α h and the vertical acceleration α v , the calculation formula is as follows: a h =F h / M a v =MAX((F v -Mg) / M,g) Where g is the acceleration due to gravity; 123) Get the maximum horizontal speed of the drone from the drone performance parameter table and the maximum vertical speed Calculate the safe distance DI of the drone in the horizontal direction h And the vertical safety interval DI v , the calculation formula is as follows: Among them, T eh and T ev are the redundancy time in the horizontal and vertical directions respectively; N h and N v Configure parameters for horizontal and vertical safety intervals separately.

4. The UAV conflict prediction method based on safety interval progressive decision tree according to claim 3 is characterized in that: The step 2) specifically includes: 21) Establish a coordinate transformation model parameter library, the parameter items include: original coordinate system type, new coordinate system type, geographical application scope, translation parameters, rotation parameters and scale parameters; 22) Establish a coordinate transformation function and define the function calling method.

5. The UAV conflict prediction method based on safety interval progressive decision tree according to claim 4 is characterized in that: The step 22) specifically includes: 221) Construct the Bursa model as follows: Among them, ΔX, ΔY, ΔZ are translation parameters, m is the scale parameter, are rotation parameters; X1, Y1, Z1 are the original coordinates of the drone position in the WGS-84 coordinate system or the GLONASS-PZ90 coordinate system; X2, Y2, Z2 are the new coordinates of the drone position in the CGCS2000 coordinate system; 222) Encapsulate the Bursa model in step 221) as a coordinate conversion function, and define the function call method for use in coordinate conversion calculation; the function input parameters are: the original coordinate system type, the new coordinate system type, and the specific position of the drone in the original coordinate system X1, Y1, Z1; the function output parameters are: the specific position of the drone in the new coordinate system X2, Y2, Z2; 223) Use the coordinate conversion function to convert the WGS-84 coordinate system or GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system.

6. The method for predicting UAV conflicts based on a safety interval progressive decision tree according to claim 5, characterized in that: The step 223) is specifically as follows: Compare the original coordinate system type, new coordinate system type, and specific position X1, Y1, and Z1 of the drone in the original coordinate system in the function input parameters to the three parameter items of original coordinate system type, new coordinate system type, and regional applicable scope in the coordinate conversion model parameter library to obtain translation parameters, rotation parameters, and scale parameters; when the data record obtained from the coordinate conversion model parameter library is not unique, terminate the coordinate conversion calculation and prompt the user that an error has occurred; otherwise, continue to execute the coordinate conversion calculation; Substitute the specific position X1, Y1, Z1 of the drone in the original coordinate system, the translation parameters, rotation parameters, and scale parameters into the coordinate conversion function to obtain the specific position X2, Y2, Z2 of the drone in the new coordinate system, and return the specific position X2, Y2, Z2 of the drone in the new coordinate system.

7. The UAV conflict prediction method based on safety interval progressive decision tree according to claim 6 is characterized in that: The step 3) specifically includes: 31) Establish a UAV flight dynamics database, the data items include: flight record number, UAV serial number, take-off time, flight status, flight dynamics data reception time, UAV model, new coordinate system type, UAV position on the X axis, UAV position on the Y axis, UAV position on the Z axis, flight speed on the X axis, flight speed on the Y axis, flight speed on the Z axis, track angle, horizontal safety interval and vertical safety interval; 32) Receive the flight dynamic data sent by the drone, parse the received flight dynamic data of the drone, and obtain the flight record number, flight status, current time, drone model, coordinate system type, position longitude, position latitude, flight altitude, vertical flight speed, horizontal flight speed and track angle; 33) Decompose the flight record number according to the order and number of digits of the string to obtain the drone serial number, take-off time and random code respectively; the flight record number is composed of the drone serial number, take-off time and random code, and the specific format is: 20-digit drone serial number + 8-digit take-off time + 8-digit random code; 34) Convert the WGS-84 coordinate system or GLONASS-PZ90 coordinate system to the CGCS2000 coordinate system; specifically: 341) Express the WGS-84 coordinate system, CGCS2000 coordinate system, and GLONASS-PZ90 coordinate system in the Earth-centered Earth-fixed rectangular coordinate system; 3411) Define the position of the drone represented by latitude, longitude and altitude coordinates as P L ,as follows: P L =(B,L,H) Where B is the location latitude, L is the location longitude, and H is the flight altitude; 3412) Calculate the radius of curvature N of the yoke circle using the following formula: Where a is the semi-major axis of the reference ellipsoid, and e is the first eccentricity; 3413) Define the position of the UAV expressed in earth-centered rectangular coordinates as P Z =(X1, Y1, Z1); The drone position P represented by latitude, longitude and altitude coordinates L The UAV position P converted into Earth-centered rectangular coordinates Z , the calculation formula is as follows: X1=(N+H)cos(B)cos(L) Y1=(N+H)cos(B)sin(L) Z1=(N(1-e 2 )+H)sin(B); 342) Determine the coordinate system type. If it is the WGS-84 coordinate system or the GLONASS-PZ90 coordinate system, perform coordinate conversion calculation. The specific steps are: convert the obtained UAV position P represented by the Earth-centered Earth-fixed rectangular coordinates Z =(X1, Y1, Z1), coordinate system type, and new coordinate system type are substituted into the coordinate conversion function to obtain the specific position X2, Y2, Z2 of the drone in the CGCS2000 coordinate system, and set the drone position P in the CGCS2000 coordinate system C =(X2, Y2, Z2); if it is CGCS2000 coordinate system, P C =P Z =(X1, Y1, Z1); 35) Calculate the speed v of the drone on the X-axis, Y-axis, and Z-axis respectively x , v y , v z ; v x =v t v y =v l sinθ v z =v l cosθ Among them, v t is the vertical flight speed, v l is the horizontal flight speed, θ is the track angle; 36) Using the drone model obtained in step 32) to match the drone model parameter item in the drone safety interval parameter library, obtain the horizontal safety interval and the vertical safety interval; when the data record obtained from the drone safety interval parameter library is not unique, an error is prompted and the method ends; otherwise, step 37) is executed; 37) The standardized converted data is saved in the UAV flight dynamic database. The specific correspondence of the saved data is as follows: The flight record number of the UAV flight dynamic database = the flight record number obtained in step 32); The drone serial number in the drone flight dynamic database = the drone serial number obtained in step 32); The take-off time of the UAV flight dynamic database = the take-off time obtained in step 33); The flight status of the UAV flight dynamic database = the flight status obtained in step 32); The flight dynamics data receiving time of the UAV flight dynamics database = the current time obtained in step 32); The drone model of the drone flight dynamic database = the drone model obtained in step 32); New coordinate system type for UAV flight dynamic database = CGCS2000; The drone position on the X axis of the drone flight dynamic database = X2 obtained in step 342); The drone position on the Y axis of the drone flight dynamic database = Y2 obtained in step 342); The drone position on the Z axis of the drone flight dynamic database = Z2 obtained in step 342); The flight speed on the X axis of the UAV flight dynamic database = v obtained in step 35) x ; The flight speed on the Y axis of the UAV flight dynamic database = v obtained in step 35) y ; The flight speed on the Z axis of the UAV flight dynamic database = v obtained in step 35) z ; Track angle of the UAV flight dynamic database = track angle obtained in step 32); The horizontal safety interval of the UAV flight dynamic database = the horizontal safety interval obtained in step 36; The vertical safety interval of the UAV flight dynamic database = the vertical safety interval obtained in step 36).

8. The UAV conflict prediction method based on safety interval progressive decision tree according to claim 1 is characterized in that: The step 5) specifically includes: 51) Calculate the time when the drone pairs in the focus data set Dgz may reach the warning level. The specific steps are as follows: We will focus on dataset D gz As a sample dataset; When focusing on dataset D gz UAV (W i , W j ), satisfying |SX i -SX j |>MAX(IX i , IX j ), flight record number XH i and XH j The time T that may reach the warning level gz1 =|SX i -SX j | / |VX i -VX j |; When focusing on dataset D gz UAV (W i , W j ), satisfying |SY i -SY j |>MAX(IY i , IY j ), flight record number XH i and XH j The time T that may reach the warning level gz2 =|SY i -SY j | / |VY i -VY j |; When focusing on dataset D gz UAV (W i , W j ), satisfying |SZ i -SZ j |>MAX(IZ i , IZ j ), flight record number XH i and XH j The time T that may reach the warning level gz3 =|SZ i -SZ j | / |VZ i -VZ j |; Focus on dataset D gz The time set T of all drones expected to reach the warning level gz ,as follows: T gz ={XH i ,XH j ,(T gz1 ,T gz2 ,T gz3 )|(W i ,W j )∈D gz }; 52) Calculate the alarm prompt data set D gj The specific steps for determining the estimated collision time of the conflicting drones are as follows: The warning dataset D gj As a sample data set, calculate the UAV pair (W i , W j ) is expected to collide at a time T gj ",as follows: Alarm prompt dataset D gj The estimated collision time set T of all drone pairs in gj ,as follows: T gj ={XH i ,XH j ,T gj ″|(W i ,W j )∈D gj }。

Images