A multi-UAV conflict risk assessment method considering macro network characteristics

Abstract

The present invention discloses a multi-UAV conflict risk assessment method considering macro network characteristics, the steps are as follows: determine the conflict relationship between any two UAVs, calculate the shortest expected conflict time of each pair of conflicts in the airspace; based on the conflict relationship, establish a UAV conflict network, calculate the macro network characteristic indicators, including degree, average weighted degree, average strength and clustering coefficient; normalize the macro network characteristic indicators, and calculate the multi-UAV conflict risk assessment indicators. The method of the present invention can perform risk assessment on multi-UAV conflict scenarios and quantify the conflict risk level, thereby providing more accurate alarm information.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) traffic management, and in particular relates to a multi-UAV conflict risk assessment method considering macro network characteristics. Background Art

[0002] With the rapid development of low-altitude drones, the demand for flight is increasing day by day, air traffic activities are becoming more frequent, and the risk of flight conflicts between drones and buildings, and between drones, is also increasing day by day. Therefore, conducting conflict risk assessments for drone operations is an important means to promote the safe and standardized development of drones.

[0003] The operational risks of drones can be divided into mid-air collision risks and ground collision risks, among which mid-air collision risks include collisions with other aircraft and collisions with static obstacles. Current research on drone conflict risk assessment focuses on conflicts between drones and static obstacles or conflicts between two drones. With the increase in the density of drone operations in the airspace, the conflict relationships between drones are complex and changeable, which puts higher requirements on the drone operation risk assessment method. Therefore, when conducting conflict risk assessment on drones, in addition to considering the conflict characteristics between two drones, it is necessary to further extract macro features to realize the assessment of the overall operational conflict risks of multiple drones in the airspace at future moments. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a multi-UAV conflict risk assessment method that takes into account macro network characteristics, so as to solve the problems in the prior art that the multi-UAV conflict risk is difficult to assess and the assessment angle is one-sided; the method of the present invention can perform risk assessment on multi-UAV conflict scenarios and quantify the conflict risk level, thereby providing more accurate alarm information.

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

[0006] The present invention provides a multi-UAV conflict risk assessment method considering macro network characteristics, and the steps are as follows:

[0007] 1) Determine the conflict relationship between any two UAVs and calculate the shortest expected conflict time for each pair of conflicts in the airspace;

[0008] 2) Based on the conflict relations in step 1), a UAV conflict network is established and macro network characteristic indicators are calculated, including degree, average weighted degree, average strength and clustering coefficient;

[0009] 3) Normalize the macro network characteristic indicators in step 2) and calculate the multi-UAV conflict risk assessment index.

[0010] Furthermore, the step 1) specifically includes:

[0011] 11) Define conflict protection zones for drones;

[0012] The location of the drone is taken as the midpoint of the protection zone in the horizontal and vertical directions, and the shape of the conflict protection zone is set to be a cylinder. The size calculation formula is as follows:

[0013]

[0014]

[0015] Where h is the height of the cylinder in the conflict protection zone; r is the radius of the cylinder in the conflict protection zone; h u is the height of the drone’s overall dimensions; r u G is half of the maximum horizontal size of the drone; y With G h They are the vertical and horizontal hovering accuracy of the drone in GPS mode respectively;

[0016] 12) Determine the conflict relationship between any two drones in the airspace;

[0017] For any two drones in the airspace: drone A and drone B, drone A and drone B are regarded as mass points, and the predicted three-dimensional position coordinates of drone B at the tth moment are judged whether they are located in the conflict protection zone of drone A in the horizontal and vertical directions. The specific judgment formula is as follows:

[0018]

[0019] |P(z) t-A -P(z) t-B |≤max{h A ,h B} (4)

[0020] Where P(x) t-A 、P(y) t-A 、P(z) t-A are the horizontal x-coordinate, horizontal y-coordinate and vertical z-coordinate of drone A at time t; P(x) t-B 、P(y) t-B 、P(z) t-B are the x-coordinate, y-coordinate and z-coordinate of drone B at the tth moment respectively; A ,r B are the radii of the cylinders of the conflict protection zones of UAV A and UAV B, h A ,h Bare the heights of the cylinders of the conflict protection zones of UAV A and UAV B respectively; if formula (3) or formula (4) is not true, UAV A and UAV B will not collide at time t, and the conflict relationship l AB (t) = 0; if formula (3) and formula (4) are both true, UAV A and UAV B collide at time t, and the conflict relationship is l AB (t) = 1;

[0021] 13) Calculate the shortest expected conflict time;

[0022] According to the conflict relationship between drones obtained in step 12), when drone A and drone B collide, drone A does not perform any obstacle avoidance measures, and drone B flies directly to the direction of drone A at the maximum flight speed. The calculation of the expected shortest conflict time is as follows:

[0023]

[0024]

[0025] Where, T AB (t) is the expected shortest conflict time at time t, dis AB (t) is the distance between UAV A and UAV B at the tth moment, is the maximum flight speed of UAV B.

[0026] Furthermore, the step 2) specifically includes:

[0027] 21) Establishing a drone deconfliction network;

[0028] During the movement, each UAV is regarded as a particle with a direction and a node of the conflict network; based on the conflict relationship between any two UAVs obtained in step 12), when l AB = 0, there is no edge between node A and node B; when l AB =1, there is an edge between node A and node B;

[0029] 22) Calculate the degree of the UAV conflict network;

[0030] At time t, the number of edges connected to node A is the degree of node A, and the number of all connected edges in the drone conflict network is the degree of the network. The calculation formula is as follows:

[0031]

[0032]

[0033] In the formula, k A(t) is the degree of node A at time t, k(t) is the degree of the UAV conflict network at time t, and the probability of a UAV with a higher degree colliding with other UAVs is higher; n is the total number of UAVs in the network;

[0034] 23) Calculate the average weighted degree of the UAV conflict network;

[0035] The reciprocal of the shortest expected conflict time between node A and node B at time t is used as the weight of the edge between node A and node B. The calculation formula is as follows:

[0036]

[0037] In the formula, w AB (t) is the weight of the edge between node A and node B at time t;

[0038] At time t, the sum of the edge weights connected to node A is the weighted degree of node A, and the sum of the weighted degrees of all nodes in the drone conflict network is the weighted degree of the network. The calculation formula is as follows:

[0039]

[0040]

[0041]

[0042] In the formula, is the weighted degree of node A at time t, k w (t) is the weighted degree of the UAV conflict network at time t, is the average weighted degree of the UAV conflict network at time t. The higher the average weighted degree, the greater the security risk of the UAV conflict network;

[0043] 24) Calculate the average strength of the drone conflict network;

[0044] The average strength of the conflict network represents the degree of association between nodes in the network. The specific calculation formula is as follows:

[0045]

[0046] In the formula, is the average strength of the conflict network at time t;

[0047] 25) Calculate the clustering coefficient of the UAV conflict network;

[0048] The clustering coefficient of node A represents the ratio of the number of edges actually connected between nodes among all nodes connected to node A to the maximum number of possible connected edges of node A's neighboring nodes. The clustering coefficient is used to reflect the clustering situation between conflicting network nodes. The specific calculation formula is as follows:

[0049]

[0050] In the formula, c w-A (t) is the clustering coefficient of node A at the tth moment, l BC (t), l CA (t) are the conflict relationships between nodes B and C, and between nodes C and A at time t, respectively. AC (t) is the weight of the edge between node A and node C at time t; the clustering coefficient of the entire network is:

[0051]

[0052] In the formula, is the clustering coefficient of the network at the tth moment, The larger the size, the higher the conflict risk of the drone conflict network; when , it means that there is an edge between any two nodes in the network, and the risk of conflict is high.

[0053] Furthermore, the step 3) specifically includes:

[0054] 31) Index normalization processing;

[0055] The min-max method is used to normalize the four macro network characteristic indicators of the conflict network, namely, degree, average weighted degree, average strength and clustering coefficient. The formula is as follows:

[0056]

[0057] In the formula, x * is the normalized index, x is the index to be normalized, x min and x max are the minimum and maximum values ​​of each indicator respectively;

[0058] 32) Calculate UAV conflict risk assessment indicators;

[0059] Based on the normalized macro network characteristic index at time t obtained in step 31), the conflict risk assessment index Γ(t) of the UAV in the airspace at time t is constructed. The specific formula is as follows:

[0060]

[0061] In the formula, and are the normalized average weighted degree, average strength and clustering coefficient of the UAV conflict network at the tth moment, η s and η c The indicators are and the relative importance of

[0062] The calculation formula of the drone conflict risk assessment index is as follows:

[0063] R(t)=||Γ(t)||2 (18)

[0064] Where R(t) is the UAV conflict risk assessment index, R(t)∈[0,1]. The larger R(t) is, the greater the risk of conflict between UAVs will be as the risk probability and proximity increase. The higher the conflict risk between multiple UAVs operating in the airspace at time t is.

[0065] Beneficial effects of the present invention:

[0066] Aiming at the scenario where conflicts arise from the operation of multiple drones in an airspace, the present invention establishes a drone conflict network based on the conflict relationship between two drones, extracts macro network characteristic indicators, and then obtains drone conflict risk assessment indicators, thereby achieving effective assessment of the conflict risk of multiple drones operating in low-altitude airspace.

[0067] The conflict risk assessment result obtained by the method of the present invention can better reflect the overall operating status of drones in the airspace, and can provide decision-making support for security situation awareness and conflict resolution in multi-drone conflict scenarios, which is in line with the development needs of high-density operation of drones in low-altitude airspace in the future. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Flow chart of the method of the present invention.

[0069] Figure 2 Schematic diagram of conflicting connections in a UAV conflict network at different prediction moments according to an embodiment of the present invention. DETAILED DESCRIPTION

[0070] 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.

[0071] Reference Figure 1 As shown, a multi-UAV conflict risk assessment method considering macro network characteristics of the present invention comprises the following steps:

[0072] 1) Determine the conflict relationship between any two UAVs and calculate the shortest expected conflict time for each pair of conflicts in the airspace; specifically:

[0073] 11) Define conflict protection zones for drones;

[0074] The location of the drone is taken as the midpoint of the protection zone in the horizontal and vertical directions, and the shape of the conflict protection zone is set to be a cylinder. The size calculation formula is as follows:

[0075]

[0076]

[0077] Where h is the height of the cylinder in the conflict protection zone; r is the radius of the cylinder in the conflict protection zone; h u is the height of the drone’s overall dimensions; r u G is half of the maximum horizontal size of the drone; y With G h They are the vertical and horizontal hovering accuracy of the drone in GPS mode respectively;

[0078] 12) Determine the conflict relationship between any two drones in the airspace;

[0079] For any two drones in the airspace: drone A and drone B, drone A and drone B are regarded as mass points, and the predicted three-dimensional position coordinates of drone B at the tth moment are judged whether they are located in the conflict protection zone of drone A in the horizontal and vertical directions. The specific judgment formula is as follows:

[0080]

[0081] |P(z) t-A -P(z) t-B |≤max{h A ,h B} (4)

[0082] Where P(x) t-A 、P(y) t-A 、P(z) t-A are the horizontal x-coordinate, horizontal y-coordinate and vertical z-coordinate of drone A at time t; P(x) t-B 、P(y) t-B 、P(z) t-B are the x-coordinate, y-coordinate and z-coordinate of drone B at the tth moment respectively; A ,r B are the radii of the cylinders of the conflict protection zones of UAV A and UAV B, h A ,h B are the heights of the cylinders of the conflict protection zones of UAV A and UAV B respectively; if formula (3) or formula (4) is not true, UAV A and UAV B will not collide at time t, and the conflict relationship l AB(t) = 0; if formula (3) and formula (4) are both true, UAV A and UAV B collide at time t, and the conflict relationship is l AB (t) = 1;

[0083] 13) Calculate the shortest expected conflict time;

[0084] According to the conflict relationship between drones obtained in step 12), when drone A and drone B collide, drone A does not perform any obstacle avoidance measures, and drone B flies directly to the direction of drone A at the maximum flight speed. The calculation of the expected shortest conflict time is as follows:

[0085]

[0086]

[0087] Where, T AB (t) is the expected shortest conflict time at time t, dis AB (t) is the distance between UAV A and UAV B at the tth moment, is the maximum flight speed of UAV B.

[0088] 2) Based on the conflict relations in step 1), a UAV conflict network is established and macro network characteristic indicators are calculated, including degree, average weighted degree, average strength and clustering coefficient;

[0089] 21) Establishing a drone deconfliction network;

[0090] During the movement, each UAV is regarded as a particle with a direction and a node of the conflict network; based on the conflict relationship between any two UAVs obtained in step 12), when l AB = 0, there is no edge between node A and node B; when l AB =1, there is an edge between node A and node B;

[0091] In this example, the network connection status of the drone conflict at the predicted time t of 10 seconds and 20 seconds is as follows Figure 2 shown.

[0092] 22) Calculate the degree of the UAV conflict network;

[0093] At time t, the number of edges connected to node A is the degree of node A, and the number of all connected edges in the drone conflict network is the degree of the network. The calculation formula is as follows:

[0094]

[0095]

[0096] In the formula, k A (t) is the degree of node A at time t, k(t) is the degree of the UAV conflict network at time t, and the probability of a UAV with a higher degree colliding with other UAVs is higher; n is the total number of UAVs in the network;

[0097] 23) Calculate the average weighted degree of the UAV conflict network;

[0098] The reciprocal of the shortest expected conflict time between node A and node B at time t is used as the weight of the edge between node A and node B. The calculation formula is as follows:

[0099]

[0100] In the formula, w AB (t) is the weight of the edge between node A and node B at time t;

[0101] At time t, the sum of the edge weights connected to node A is the weighted degree of node A, and the sum of the weighted degrees of all nodes in the drone conflict network is the weighted degree of the network. The calculation formula is as follows:

[0102]

[0103]

[0104]

[0105] In the formula, is the weighted degree of node A at time t, k w (t) is the weighted degree of the UAV conflict network at time t, is the average weighted degree of the UAV conflict network at the tth moment. The higher the average weighted degree, the greater the security risk of the UAV conflict network.

[0106] 24) Calculate the average strength of the drone conflict network;

[0107] The average strength of the conflict network represents the degree of association between nodes in the network. The specific calculation formula is as follows:

[0108]

[0109] In the formula, is the average strength of the conflict network at time t;

[0110] 25) Calculate the clustering coefficient of the UAV conflict network;

[0111] The clustering coefficient of node A represents the ratio of the number of edges actually connected between nodes among all nodes connected to node A to the maximum number of possible connected edges of node A's neighboring nodes. The clustering coefficient is used to reflect the clustering situation between conflicting network nodes. The specific calculation formula is as follows:

[0112]

[0113] In the formula, c w-A (t) is the clustering coefficient of node A at the tth moment, l BC (t), l CA (t) are the conflict relationships between nodes B and C, and between nodes C and A at time t, respectively. AC (t) is the weight of the edge between node A and node C at time t; the clustering coefficient of the entire network is:

[0114]

[0115] In the formula, is the clustering coefficient of the network at the tth moment, The larger the size, the higher the conflict risk of the drone conflict network; when , it means that there is an edge between any two nodes in the network, and the risk of conflict is high.

[0116] In this example, the results of the macro network characteristic indicators when the prediction time t is 5 seconds, 10 seconds, 15 seconds, and 20 seconds are shown in Table 1:

[0117] Table 1

[0118]

[0119] 3) Normalizing the macro network characteristic indicators in step 2) and calculating the multi-UAV conflict risk assessment index;

[0120] 31) Index normalization processing;

[0121] The min-max method is used to normalize the four macro network characteristic indicators of the conflict network, namely, degree, average weighted degree, average strength and clustering coefficient. The formula is as follows:

[0122]

[0123] In the formula, x * is the normalized index, x is the index to be normalized, x min and x max are the minimum and maximum values ​​of each indicator respectively;

[0124] 32) Calculate UAV conflict risk assessment indicators;

[0125] Based on the normalized macro network characteristic index at time t obtained in step 31), the conflict risk assessment index Γ(t) of the UAV in the airspace at time t is constructed. The specific formula is as follows:

[0126]

[0127] In the formula, and are the normalized average weighted degree, average strength and clustering coefficient of the UAV conflict network at the tth moment, η s and η c The indicators are and The relative importance of η s and η c Take 1 / 3;

[0128] The calculation formula of the drone conflict risk assessment index is as follows:

[0129] R(t)=||Γ(t)||2 (18)

[0130] Where R(t) is the UAV conflict risk assessment index, R(t)∈[0,1]. The larger R(t) is, the greater the risk of conflict between UAVs will be as the risk probability and proximity increase. The higher the conflict risk between multiple UAVs operating in the airspace at time t is.

[0131] 33) Calculate the prediction accuracy of conflict risk assessment indicators;

[0132] In this example, R(t) is divided into four risk intervals with an interval of 0.25 to evaluate the conflict risk at the future time t. When the predicted value and the actual value of the conflict risk assessment index are in the same interval, the true label is yes, otherwise it is no. Through multiple Monte Carlo simulation experiments, the ratio of the number of correct predictions of the conflict risk assessment index to the number of simulation experiments is obtained as the risk prediction accuracy. The risk prediction accuracy when the prediction time t is 5 seconds, 10 seconds, 15 seconds, and 20 seconds is shown in Table 2 below. For the prediction analysis of the multi-UAV conflict risk assessment index in the airspace, as the time step increases, the uncertainty of the risk assessment increases, and the prediction accuracy decreases accordingly; Table 2 is as follows:

[0133] Table 2

[0134]

[0135] 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 multi-UAV conflict risk assessment method considering macro network characteristics, characterized in that: Here are the steps: 1) Determine the conflict relationship between any two UAVs and calculate the shortest expected conflict time for each pair of conflicts in the airspace; 2) Based on the conflict relationship in step 1), a UAV conflict network is established and macro network characteristic indicators are calculated, including degree, average weighted degree, average strength and clustering coefficient; 3) Normalizing the macro network characteristic indicators in step 2) and calculating the multi-UAV conflict risk assessment index; The step 1) specifically includes: 11) Define conflict protection zones for drones; The location of the drone is taken as the midpoint of the protection zone in the horizontal and vertical directions, and the shape of the conflict protection zone is set to be a cylinder. The size calculation formula is as follows: Where h is the height of the cylinder in the conflict protection zone; r is the radius of the cylinder in the conflict protection zone; h u is the height of the drone’s overall dimensions; r u G is half of the maximum horizontal size of the drone; y With G h They are the vertical and horizontal hovering accuracy of the drone in GPS mode respectively; 12) Determine the conflict relationship between any two drones in the airspace; For any two drones in the airspace: drone A and drone B, drone A and drone B are regarded as mass points, and the predicted three-dimensional position coordinates of drone B at the tth moment are judged whether they are located in the conflict protection zone of drone A in the horizontal and vertical directions. The specific judgment formula is as follows: |P(z) t-A -P(z) t-B |≤max{h A ,h B } (4) Where P(x) t-A 、P(y) t-A 、P(z) t-A are the horizontal x-coordinate, horizontal y-coordinate and vertical z-coordinate of drone A at time t; P(x) t-B 、P(y) t-B 、P(z) t-B are the x-coordinate, y-coordinate and z-coordinate of drone B at the tth moment respectively; A ,r B are the radii of the cylinders of the conflict protection zones of UAV A and UAV B, h A ,h B are the heights of the cylinders of the conflict protection zones of UAV A and UAV B respectively; if formula (3) or formula (4) is not true, UAV A and UAV B will not collide at time t, and the conflict relationship l AB (t) = 0; if formula (3) and formula (4) are both true, UAV A and UAV B collide at time t, and the conflict relationship is l AB (t) = 1; 13) Calculate the shortest expected conflict time; According to the conflict relationship between drones obtained in step 12), when drone A and drone B collide, drone A does not perform any obstacle avoidance measures, and drone B flies directly to the direction of drone A at the maximum flight speed. The calculation of the expected shortest conflict time is as follows: Where, T AB (t) is the expected shortest conflict time at time t, dis AB (t) is the distance between UAV A and UAV B at the tth moment, is the maximum flight speed of UAV B; The step 2) specifically includes: 21) Establishing a drone deconfliction network; During the movement, each UAV is regarded as a particle with a direction and a node of the conflict network; based on the conflict relationship between any two UAVs obtained in step 12), when l AB = 0, there is no edge between node A and node B; when l AB =1, there is an edge between node A and node B; 22) Calculate the degree of the UAV conflict network; At time t, the number of edges connected to node A is the degree of node A, and the number of all connected edges in the drone conflict network is the degree of the network. The calculation formula is as follows: In the formula, k A (t) is the degree of node A at time t, k(t) is the degree of the UAV conflict network at time t, and the probability of a UAV with a higher degree colliding with other UAVs is higher; n is the total number of UAVs in the network; 23) Calculate the average weighted degree of the UAV conflict network; The reciprocal of the shortest expected conflict time between node A and node B at time t is used as the weight of the edge between node A and node B. The calculation formula is as follows: In the formula, w AB (t) is the weight of the edge between node A and node B at time t; At time t, the sum of the edge weights connected to node A is the weighted degree of node A, and the sum of the weighted degrees of all nodes in the drone conflict network is the weighted degree of the network. The calculation formula is as follows: In the formula, is the weighted degree of node A at time t, k w (t) is the weighted degree of the UAV conflict network at time t, is the average weighted degree of the UAV conflict network at time t; 24) Calculate the average strength of the drone conflict network; The average strength of the conflict network represents the degree of association between nodes in the network. The specific calculation formula is as follows: In the formula, is the average strength of the conflict network at time t; 25) Calculate the clustering coefficient of the UAV conflict network; The clustering coefficient of node A represents the ratio of the number of edges actually connected between nodes among all nodes connected to node A to the maximum number of possible connected edges of node A's neighboring nodes. The clustering coefficient is used to reflect the clustering situation between conflicting network nodes. The specific calculation formula is as follows: In the formula, c w-A (t) is the clustering coefficient of node A at the tth moment, l BC (t), l CA (t) are the conflict relationships between nodes B and C, and between nodes C and A at time t, respectively. AC (t) is the weight of the edge between node A and node C at time t; the clustering coefficient of the entire network is: In the formula, is the clustering coefficient of the network at the tth moment, The larger the size, the higher the conflict risk of the drone conflict network; when , it means that there is an edge between any two nodes in the network, and the risk of conflict is high.

2. The multi-UAV conflict risk assessment method considering macro network characteristics according to claim 1 is characterized in that: The step 3) specifically includes: 31) Index normalization processing; The min-max method is used to normalize the four macro network characteristic indicators of the conflict network, namely, degree, average weighted degree, average strength and clustering coefficient. The formula is as follows: In the formula, x * is the normalized index, x is the index to be normalized, x min and x max are the minimum and maximum values ​​of each indicator respectively; 32) Calculate UAV conflict risk assessment indicators; Based on the normalized macro network characteristic index at time t obtained in step 31), the conflict risk assessment index Γ(t) of the UAV in the airspace at time t is constructed. The specific formula is as follows: In the formula, and are the normalized average weighted degree, average strength and clustering coefficient of the UAV conflict network at time t, η kw , η s and η c The indicators are and the relative importance of The calculation formula of the drone conflict risk assessment index is as follows: R(t)=||Γ(t)||2 (18) Where R(t) is the UAV conflict risk assessment index, R(t)∈[0,1]. The larger R(t) is, the greater the risk of conflict between UAVs will be as the risk probability and proximity increase. The higher the conflict risk between multiple UAVs operating in the airspace at time t is.

Images