Improved speed barrier method for cluster division

Through the improved speed obstacle method for cluster division, the elliptical protected area and clustering model is used to solve the problem of conflict identification and obstacle avoidance between aircraft in high-density aviation environment, the aircraft's obstacle avoidance efficiency and path planning effect are improved, and the aircraft's obstacle avoidance efficiency and path planning effect are reduced. Waste of resources and costs.

CN118095056BActive Publication Date: 2024-11-15CIVIL AVIATION FLIGHT UNIV OF CHINA
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410071065.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-01-17
Publication Date
2024-11-15
Estimated Expiration
2044-01-17

AI Technical Summary

Technical Problem

The prior art is difficult to effectively deal with real-time conflict identification and obstacle avoidance between aircraft in high-density, multi-objective and diversified aviation environments. The calculation complexity is high and the efficiency is low, and it cannot adapt to high-density flight environments.

Method used

Using the improved speed obstacle method for cluster division, by constructing a variable precise elliptical protection zone, clustering aircraft with similar velocity, heading and close distances based on a specific rule, and using the cluster control algorithm to micro-adjust the speed vector. , combined with plane geometric analysis, the speed and heading liberation strategies of aircraft clusters are achieved.

Benefits of technology

It effectively improves the aircraft's obstacle avoidance capabilities and path planning effects, reduces the amount of changes in the speed of escape and heading, improves operational efficiency, reduces the frequency of conflict identification and resolution, and reduces resource waste and costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118095056B_ABST
    Figure CN118095056B_ABST
Patent Text Reader

Abstract

The invention discloses an improved speed obstacle method for cluster division, comprising: constructing a variable accurate elliptical protection zone; constructing a clustering model based on specific rules to cluster aircraft to form clusters; characterizing the mass points, velocity vectors and protection zone size of the cluster, wherein the protection zone size is determined by establishing an optimization model with the minimum cluster protection zone area as the goal and with the constraints of no conflict between aircraft and the aircraft being located inside the cluster; using a cluster control algorithm to make a slight adjustment to the velocity vectors of the aircraft in the cluster to ensure that the aircraft in the cluster will not leave the cluster and conflict during obstacle avoidance; improving and optimizing the speed obstacle method for the concept of cluster, combining the constraint condition of the maximum change per unit time, and using plane geometry analysis to solve the obstacle avoidance strategy of the aircraft cluster including speed release and heading release; the invention improves the obstacle avoidance capability and path planning effect of the aircraft.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of aircraft path planning, in particular to an improved speed barrier method oriented to cluster division. Background Art

[0002] Driven by aviation technology and the trend of globalization, the widespread popularity of civil aviation passenger aircraft has not only broadened the route network, but also provided passengers with faster and more convenient travel options, thereby strengthening the connection of the global economy and promoting cultural exchanges. However, with the transportation of a large number of passengers, ensuring flight safety has become an issue that needs urgent attention. Especially when performing flight missions, dynamic route path planning becomes crucial. Generally, path planning can be divided into global path planning and local path planning. The former focuses on the optimal path, that is, avoiding known static obstacle aircraft based on prior knowledge to find the best path, while the latter focuses on real-time perception and analysis to avoid collisions. During flight, aircraft face unknown and dynamic obstacle aircraft, so a local path planning algorithm is needed to avoid obstacles in real time to ensure the safe navigation of the aircraft.

[0003] At present, the commonly used local path planning algorithms include artificial potential field method, guidance method, obstacle avoidance planning algorithm based on machine learning, speed obstacle method and dynamic window method. The artificial potential field method guides the aircraft to avoid obstacles by calculating gravity and repulsion, but it has limitations in accurately reaching the target point, path stability and flyability; the guidance method draws on the principle of missile guidance, but it still faces challenges in multi-obstacle aircraft invasion and considering subsequent flight actions; obstacle avoidance planning algorithms and intelligent algorithms based on machine learning have high requirements for computing resources and time, and are more suitable for simplified scenarios such as global path planning; the dynamic window rule is based on speed and acceleration constraints, combined with local perception information, to find the best path in a real-time environment; the speed obstacle method judges the conflict between aircraft and obtains the range of release by constructing a speed obstacle pile. The advantage is that it can directly consider speed information and can better adapt to dynamic environments. This method is not only used for conflict identification of two moving objects, but also for multiple moving objects. Therefore, it is widely used in obstacle avoidance and trajectory planning of aircraft, drones and aircraft.

[0004] Existing studies have all considered adding environmental factors and improving the accuracy of path planning. Although significant progress has been made in collision avoidance, path planning and risk assessment of aircraft and UAVs, they also have some common limitations or shortcomings, such as increased computational complexity and high dependence on training data, which leads to long calculation time, low efficiency, and limited model generalization ability and application scope. They are unable to handle high-density, multi-target and diversified aviation environment problems, which brings greater work pressure to controllers. Therefore, existing technologies cannot effectively deal with the problem of real-time conflict identification and obstacle avoidance between aircraft under high density. Summary of the invention

[0005] In order to solve the problems of aircraft conflict identification and dynamic path planning in high-density, multi-target and diversified aviation environments and improve the obstacle avoidance capability and path planning effect of aircraft, the present invention proposes an improved speed obstacle method for cluster division.

[0006] To achieve the above object, the technical solution adopted by the present invention is: an improved speed barrier method for cluster division, comprising the following steps:

[0007] Step 1: construct a variable precise elliptical protection zone according to the aircraft speed difference;

[0008] Step 2: Build a clustering model based on specific rules to cluster aircraft with similar speed, heading, and close distance to form clusters;

[0009] Step 3: Characterize the mass points, velocity vectors, and protected area size of the cluster, where the protected area size is determined by establishing an optimization model with the goal of minimizing the protected area of ​​the cluster and the constraints of no conflict between aircraft and the aircraft being located inside the cluster;

[0010] Step 4: Use the cluster control algorithm to make slight adjustments to the velocity vectors of the aircraft in the cluster to ensure that the aircraft in the cluster will not leave the cluster and collide during the obstacle avoidance period;

[0011] Step 5: Improve and optimize the speed obstacle method based on the concept of clustering, combine the constraint of the maximum change per unit time, and use plane geometry analysis to solve the obstacle avoidance strategy of the aircraft cluster including speed release and heading release.

[0012] As a further improvement of the present invention, the following steps are also included:

[0013] Step 6: Use Python to simulate the conflicts between clusters and verify the reliability of the speed obstacle avoidance strategy for cluster division.

[0014] As a further improvement of the present invention, in step 1, an ellipse is used to depict the safety range of aircraft of different types, and the center of the circle is defined as P i =(x i ,y i ), the velocity vector is The velocity is v i , the velocity vector Located on the long axis and with the positive x-axis at an angle of θ i ,θ i ∈(0,2π), the aircraft safety range is determined by the precise safety interval of different aircraft types and the current cruising speed, where the short semi-axis b iTake the precise safety interval corresponding to different models, the major axis a i It is related to the cruising speed, λ is the correlation coefficient, and the above variables are calculated by the following formula:

[0015]

[0016] As a further improvement of the present invention, in step 2, a clustering condition is defined: the distance, heading difference, and speed difference between aircraft are all not greater than a certain threshold. The clustering model construction method based on specific rules is as follows:

[0017] (1) Traverse the points on the ellipse and calculate the shortest distance between two points belonging to different ellipses to obtain the aircraft interval;

[0018] (2) Check whether the three clustering conditions are met simultaneously;

[0019] (3) The data that meet the clustering conditions are a cluster, and each cluster is distinguished.

[0020] As a further improvement of the present invention, the step 3 is specifically as follows:

[0021] Each divided cluster must satisfy the requirement that there is no internal conflict. If the aircraft within the cluster, i.e., the ellipses, overlap, it is considered that there is a conflict between aircraft. The conflict area S of the aircraft is calculated by the following formula:

[0022]

[0023] Where: (x i ,y i ) is the coordinate of the center of the target aircraft ellipse; (x, y) is the coordinate point on the protection zone ellipse formed by the potential conflicting aircraft; a i The major semi-axis of the protection zone ellipse for the target aircraft; b i The major semi-axis of the protection zone ellipse for the target aircraft;

[0024] The cluster is characterized by an ellipse, which is called a cluster ellipse. According to the characteristics of the aircraft position, velocity vector, and protected area size in the cluster, a mass point, velocity vector, and protected area size representing the cluster ellipse are found. The angle between the cluster ellipse and the x-axis adopts the combined velocity direction angle. The optimal centroid is the midpoint of the rectangle formed by all the ellipse boundary points in this direction. The mean value of each aircraft coordinate is used as the cluster centroid. n is the number of aircraft. Let the cluster centroid be P. Ni =(x Ni ,y Ni ), which is calculated as follows:

[0025]

[0026] Velocity vector of the cluster Including velocity direction and velocity magnitude, where the angle θ between the velocity direction and the positive half axis of the x-axis Ni is the angle between the aircraft's combined velocity vector and the positive half axis of the x-axis. The velocity magnitude v Ni is the average speed of each aircraft and is calculated as follows:

[0027]

[0028] The protection zone of the cluster ellipse is solved by constructing a cluster optimization model.

[0029] As a further improvement of the present invention, the method for constructing the optimization model of the cluster ellipse is specifically as follows:

[0030] (1) Taking the minimum area of ​​the cluster ellipse as the objective function;

[0031] (2) Traverse the points on each ellipse and make sure that each point substituted into the cluster ellipse formula satisfies a constraint of less than or equal to 1, so that each ellipse, i.e., the aircraft, is inside the cluster ellipse;

[0032] (3) Taking the centroid as the cluster center, θ Ni is the cluster velocity vector The angle with the x-axis is used to generate a cluster ellipse with the smallest area that contains all ellipses, and the major semi-axis a at this time is obtained. Ni and the minor axis b Ni The value of .

[0033] As a further improvement of the present invention, in step 4, the cluster control algorithm satisfies the alignment rule, that is, aircraft i calculates the combined velocity direction of all aircraft and the average velocity of the aircraft to obtain the cluster velocity vector, and then tries to align it. The heading adjustment method is as follows:

[0034]

[0035] As a further improvement of the present invention, the step 5 is specifically as follows:

[0036] In the distributed multi-aircraft motion coordination, each aircraft regards other aircraft as motion obstacle aircraft. Assume that the square radiation distance of the aircraft is a i , the side radiation distance is b i , the square radiation distance of a dynamic obstacle aircraft is a j , the side radiation distance is b j At time t1, the positions of the aircraft and the obstacle aircraft are P i , P j , the velocity vectors are The speeds are v i and vj , the yaw angles are θ i and θ j , the aircraft's position information is The position information of the obstacle aircraft is The above variables are calculated by the following formula:

[0037]

[0038] P j is the center of the circle, a ij is the major axis, b ij As the minor axis, draw an ellipse with P i Cone and ellipse P for the vertex j Tangent, tangent to rays l1 and l2, where is the relative velocity vector of the two, l m is a vector The above variables are calculated as follows:

[0039]

[0040] If the vector If the end is within the speed barrier cone, the dynamic obstacle aircraft poses a threat to the aircraft; define the collision area set CC in the speed space as follows:

[0041]

[0042] where ⊙P j The safety protection zone that represents the expansion of the obstructing aircraft is the elliptical area;

[0043] P i and ⊙P j Along vector Direction translation distance Get Q i =(x′ i ,y′ i ), Q j =(x′ j ,y′ j ), the unit time interval is Δt, and the above variables are calculated by the following formula:

[0044]

[0045] Move CC horizontally The resulting area is called the new collision area NCC, as shown below:

[0046]

[0047] in represents the Minkowski vector and operation;

[0048] By improving the speed barrier method after dividing the clusters, a speed barrier pile is established on the basis of the clusters to realize the identification and resolution of conflicts between clusters. At time t1, assuming that the square radiation distance of the aircraft cluster is a Ni , the side radiation distance is b Ni , the square radiation distance of a dynamic threat obstacle aircraft is a Nj , the side radiation distance is b Nj ; The positions of the cluster aircraft and the cluster obstacle aircraft are P Ni , P Nj , the velocity vectors are The speeds are v Ni and v Nj , the yaw angles are θ Ni and θ Nj , and then by P Ni and ⊙P Nj Along vector Direction translation distance Get Q Ni =(x′ Ni ,y′ Ni ), Q Nj =(x′ Nj ,y′ Nj ) Nj is the center of the circle, a Nij is the major axis, b Nij As the minor axis, draw an ellipse with Q Ni Cone and ellipse Q for the vertex Nj Tangent, tangent to rays l3 and l4, the above variables are calculated by the following formula,

[0049]

[0050] According to the improved speed obstacle model, the speed of the aircraft Now it falls into the NCC area and should be adjusted Make it fall outside the NCC area; because the aircraft is subject to its own performance constraints, the adjustment range can only be within the constraints; define the velocity vector The angle with the horizontal plane is θ Ni , aircraft release speed and the release heading θ′ Ni The allowable adjustment range within unit time Δt is as follows:

[0051]

[0052] The key parameters of the NCC area boundary are solved through geometric relationships, and the resulting release strategy is the optimal release strategy that satisfies the minimum speed and heading change.

[0053] First, let the tangent point of l3 and the optimized ellipse be P τ =(x τ ,y τ ); secondly, the optimized ellipse standard equation is differentiated with respect to x; thirdly, the tangent point P τ The derivative obtained at is equivalent to the slope of l3, and the derivative with respect to x is obtained. τ and τ Fourth, substitute the tangent point into the equation of the ellipse and combine the equations obtained in the previous step to solve for x τ ; Finally, we get y τ , The function of the line where the boundary l3 lies is The above variables are calculated as follows:

[0054]

[0055] Solve for the release velocity v′ Ni When the heading θ is not changed Ni , The straight line and the tangent equation The intersection point M = (x m ,y m ), through point M relative to point P Ni The horizontal and vertical components are obtained as and Solve for v′ Ni , calculated as follows:

[0056]

[0057] Similarly, for the NCC region boundary l4, a similar conclusion is obtained, and the release heading v″ Ni ;

[0058] Solve for the release heading θ′ Ni When the velocity v does not change Ni ,vector With tangent Intersecting at point G, Relative to P Ni (x Ni ,y Ni ) is the horizontal and vertical components of and point Bring in The equation is solved by combining the relationship equation between the two components to obtain θ′ Ni , calculated as follows:

[0059]

[0060] Similarly, for the NCC region boundary l4, a similar conclusion is obtained, and the release heading θ″ Ni .

[0061] The beneficial effects of the present invention are:

[0062] Conflict detection and release for aircraft clusters is a core issue to ensure flight safety under high traffic. The traditional speed barrier method not only includes complex calculation methods and low real-time performance, but also uses traditional aircraft intervals, which causes serious waste of resources, including airspace, cost, environment, etc. In response to the above problems, the present invention improves the speed barrier method based on the dynamic elliptical protection zone model combined with the concept of clusters. Finally, python simulation is used to verify the reliability of the model. Simulation experiments show that compared with traditional aircraft intervals, the present invention obtains precise protection zones according to different aircraft performances, which can reduce the changes in release speed and release heading, and effectively improve the operating efficiency of aircraft; compared with the traditional speed barrier method, the present invention is oriented to the concept of clusters, which effectively reduces the number of comparisons of the speed barrier method. When faced with a large number of conflicts, it can obtain release strategies more quickly and reduce the frequency of conflict resolution. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A model diagram of a safety protection zone in an embodiment of the present invention;

[0064] Figure 2 This is a diagram of aircraft cluster division in an embodiment of the present invention;

[0065] Figure 3 It is a schematic diagram of a two-dimensional plane traditional speed obstacle model in an embodiment of the present invention;

[0066] Figure 4 In the embodiment of the present invention, Schematic diagram of the speed barrier model after translation;

[0067] Figure 5 Schematic diagram of a speed barrier model between an aircraft and a cluster in an embodiment of the present invention;

[0068] Figure 6 Schematic diagram of collision avoidance of an aircraft freeing from an obstacle in an embodiment of the present invention; wherein (a) is a schematic diagram of speed adjustment of an obstacle aircraft for collision avoidance at time t1, and (b) is a schematic diagram of heading adjustment of an obstacle aircraft for collision avoidance at time t1;

[0069] Figure 7 It is a flow chart of aircraft conflict resolution based on the improved speed barrier method in an embodiment of the present invention;

[0070] Figure 8 This is a diagram of an aircraft clustering model in an embodiment of the present invention;

[0071] Fig. 9 This is a cluster ellipse information diagram obtained by cluster S1 in an embodiment of the present invention;

[0072] Fig.10 This is a cluster ellipse information diagram obtained by cluster S2 in an embodiment of the present invention;

[0073] Fig.11 This is a traditional speed barrier model diagram;

[0074] Fig.12 This is a diagram of an improved speed barrier model in an embodiment of the present invention. DETAILED DESCRIPTION

[0075] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0076] Example 1

[0077] An improved speed barrier method for cluster division includes the following steps:

[0078] A. Construct a variable precise elliptical protection zone according to the speed difference of aircraft;

[0079] Construct a clustering model based on specific rules to cluster aircraft with similar speeds, headings, and close distances to form so-called "clusters";

[0080] B. Characterize the mass points, velocity vectors, and protected area size of the cluster. The protected area size is determined by establishing an optimization model with the goal of minimizing the protected area of ​​the cluster and the constraints of no conflict between aircraft and the aircraft being located inside the cluster;

[0081] C. Use the cluster control algorithm to make slight adjustments to the velocity vectors of the aircraft in the cluster to ensure that the aircraft in the cluster will not leave the cluster and collide during obstacle avoidance;

[0082] D. Improve the concept of clusters and optimize the speed obstacle method. Combined with the constraint of the maximum change per unit time, use plane geometry analysis to solve the obstacle avoidance strategy of aircraft clusters (including speed release and heading release);

[0083] E. Conduct simulation experiments on inter-cluster conflicts through Python to verify the reliability of the speed obstacle avoidance strategy for cluster division.

[0084] In order to facilitate subsequent research, the following explanations and assumptions are made for the model:

[0085] (1) There are horizontal and vertical flight conflicts between aircraft. This paper mainly studies the horizontal conflict, so the model is simplified in a two-dimensional plane. It is not strictly targeted at a certain control scenario or flight phase, and it is also applicable to free route airspace.

[0086] (2) When the aircraft ahead is detected to be flying in a straight line, its heading and speed remain unchanged until it receives any adjustment instructions;

[0087] (3) In a conflict identification, the urgency of the potential conflict for each pair of aircraft (two aircraft are considered as one pair) is the same and indiscriminate;

[0088] (4) During the conflict detection and resolution process, the aircraft and the obstacle aircraft are simplified into ellipses with a certain safety radius. By expanding the safety radius of the obstacle aircraft, the aircraft can be simplified into a particle accordingly.

[0089] The present embodiment is further described below:

[0090] In step A:

[0091] A variable and precise elliptical protection zone is constructed for the speed difference of aircraft. According to air traffic control rules, in order to avoid flight conflicts, aircraft should maintain a certain safety distance between each other. Due to the large differences in performance and speed of different aircraft types, the flight protection zone established by the traditional safety interval standard is not enough to simulate the actual flight state of the aircraft, and there will be large errors in flight conflict detection. Therefore, if Figure 1 As shown, this embodiment uses an ellipse to describe the safety range of aircraft of different models, and the center of the circle is defined as P i =(x i ,y i ), the velocity vector is The velocity is v i , the velocity vector Located on the long axis and with the positive x-axis at an angle of θ i ,θ i ∈(0,2π), the aircraft safety range is determined by the precise safety interval of different aircraft types and the current cruising speed, where the short semi-axis b i Take the precise safety interval corresponding to different models, the major axis a i It is related to the cruising speed, λ is the correlation coefficient, and the above variables are calculated by formula (1):

[0092]

[0093] In step B:

[0094] Aircraft that meet the clustering conditions are clustered by constructing a clustering model based on specific rules. Define the clustering conditions: the distance between aircraft is no more than 10km, the heading difference is no more than 5°, and the speed difference is no more than 0.5km / min. The steps of the clustering model based on specific rules are: (1) Traverse the points on the ellipse and calculate the shortest distance between two points belonging to different ellipses to obtain the aircraft interval; (2) Check whether the three clustering conditions are met at the same time; (3) Those that meet the clustering conditions are a cluster, that is, the concept of "cluster" mentioned in this invention, and assign a color to each cluster to distinguish the relationship between clusters. Figure 2 As shown, aircraft P i and obstacle aircraft P j Divide into clusters S by clustering i , that is, cluster S = {S1, S2}, where S1 = {P1, P2, P3}, S2 = {P4, P5}, i = 1, 2, 3.

[0095] In step C:

[0096] Each divided cluster must satisfy the requirement that there is no conflict within it. The overlap of aircraft (ellipses) within the cluster is considered as a conflict between aircraft. The conflict area S of the aircraft is calculated by formula (2):

[0097]

[0098] Where: (x i ,y i ) is the coordinate of the center of the target aircraft ellipse; (x, y) is the coordinate point on the protection zone (ellipse) formed by the potential conflicting aircraft. i The major semi-axis of the protection zone ellipse for the target aircraft; b i Constructs the major semi-axis of the protection zone ellipse for the target aircraft.

[0099] The cluster is characterized by an ellipse, hereinafter referred to as a "cluster ellipse". A mass point, velocity vector and protected area size representing the cluster ellipse are found based on the position, velocity vector and protected area size of the aircraft in the cluster. The angle between the cluster ellipse and the x-axis adopts the combined velocity direction angle. The optimal centroid is the midpoint of the rectangle formed by all the ellipse boundary points in this direction. Considering that the angle difference of the aircraft within the cluster is small, the present invention directly adopts the mean value of each aircraft coordinate as the cluster centroid. Let the cluster centroid be P Ni =(x Ni ,y Ni ), n is the number of aircraft, and its calculation method is given by formula (3).

[0100]

[0101] Velocity vector of the cluster It is divided into velocity direction and velocity magnitude, where the angle θ between the velocity direction and the positive half axis of the x-axis Ni is the angle between the aircraft's combined velocity vector and the positive half axis of the x-axis. The velocity magnitude v Ni is the average speed of each aircraft, which is calculated by formula (4).

[0102]

[0103] The protection zone of the cluster ellipse is solved by constructing a cluster optimization model. The steps of the cluster ellipse optimization model are as follows: (1) Taking the minimum area of ​​the cluster ellipse as the objective function; (2) Traversing the points on each ellipse, making each point substituted into the cluster ellipse formula satisfy a constraint condition of less than or equal to 1, so that each ellipse (aircraft) is inside the cluster ellipse; (3) Taking the center of mass as the center of the cluster circle, θ Ni is the cluster velocity vector The angle with the x-axis is used to generate a cluster ellipse with the smallest area that contains all ellipses, and the major semi-axis a at this time is obtained. Ni and the minor axis b Ni The value of .

[0104] In step D:

[0105] In order to prevent conflicts and departures within the cluster during obstacle avoidance, the cluster control algorithm is used to adjust the aircraft within the cluster. The algorithm meets the alignment rule: aircraft i calculates the combined velocity direction of all aircraft and the average velocity of the aircraft to obtain the cluster velocity vector, and then tries to align this velocity. Since the aircraft adjustment is very small, it is a fine-tuning adjustment within the navigation accuracy error range. The method is feasible, and the adjustment heading is given by formula (5):

[0106]

[0107] In step E:

[0108] The basic idea of ​​the traditional speed obstacle method is to determine whether there is a potential flight conflict between aircraft in the encounter geometry space based on the current trajectory of the aircraft (including position, speed, etc.), without considering the influence of other factors. The release process can be regarded as the process of a particle avoiding the VO expansion ellipse. A conical obstacle interval is generated through the aircraft's velocity space. As long as the relative velocity vector of the aircraft is outside the conical area, the aircraft will not collide with other aircraft in the future. Through this method, the appropriate speed and heading of the aircraft can be selected while avoiding conflicts.

[0109] In the distributed multi-aircraft motion coordination, each aircraft regards other aircraft as motion obstacle aircraft. Assume that the square radiation distance of the aircraft is a i, the side radiation distance is b i , the square radiation distance of a dynamic obstacle aircraft is a j , the side radiation distance is b j At time t1, the positions of the aircraft and the obstacle aircraft are P i , P j , the velocity vectors are The speeds are v i and v j , the yaw angles are θ i and θ j , the aircraft's position information is The position information of the obstacle aircraft is The above variables can be calculated by formula (6):

[0110]

[0111] like Figure 3 As shown in P j is the center of the circle, a ij is the major axis, b ij As the minor axis, draw an ellipse with P i Cone and ellipse P for the vertex j Tangent, tangent to rays l1 and l2, where is the relative velocity vector of the two, l m is a vector The above variables can be calculated by formula (7):

[0112]

[0113] If the vector If the end is within the speed barrier cone, the dynamic obstacle aircraft poses a threat to the aircraft. Define the collision area set (Collision Cone, CC) in the speed space, as shown in formula (8):

[0114]

[0115] where ⊙P j It represents the safety protection area "puffed" by the obstructing aircraft, i.e. the elliptical area.

[0116] Normally, the velocity vector passing directly through the aircraft It is more intuitive to judge whether there is a flight conflict with the obstacle aircraft. Figure 4 As shown, P i and ⊙P j Along vector Direction translation distance Get Q i =(x′ i,y′ i ), Q j =(x′ j ,y′ j ), the unit time interval is Δt, the above variables can be calculated by formula (9):

[0117]

[0118] Move CC horizontally The area obtained is called the New Collision Cone (NCC), as shown in formula (10):

[0119]

[0120] in represents the Minkowski vector sum operation.

[0121] After dividing the clusters through the previous steps, the speed barrier method is improved, and a speed barrier pile is established on the basis of the clusters to realize the identification and resolution of conflicts between clusters. Figure 5 As shown, at time t1, assuming that the square radiation distance of the aircraft cluster is a Ni , the side radiation distance is b Ni , the square radiation distance of a dynamic threat obstacle aircraft is a Nj , the side radiation distance is b Nj The positions of the cluster aircraft and the cluster obstacle aircraft are P Ni , P Nj , the velocity vectors are The speeds are v Ni and v Nj , the yaw angles are θ Ni and θ Nj , and then by P Ni and ⊙P Nj Along vector Direction translation distance Get Q Ni =(x′ Ni ,y′ Ni ), Q Nj =(x′ Nj ,y′ Nj ). The aircraft situation relationship after being divided into clusters is as follows Figure 3 As shown in Q Nj is the center of the circle, a Nij is the major axis, b Nij As the minor axis, draw an ellipse with Q Ni Cone and ellipse Q for the vertex Nj Tangent, tangent to rays l3 and l4, the above variables can be calculated by formula (11),

[0122]

[0123] According to the improved speed obstacle model, the speed of the aircraft Now it falls into the NCC area and should be adjusted Make it fall outside the NCC area. Since the aircraft is subject to its own performance constraints, the adjustment range can only be taken within the constraints. Define the velocity vector The angle with the horizontal plane is θ Ni , aircraft release speed and the release heading θ′ Ni The allowable adjustment range within unit time Δt is given by formula (12):

[0124]

[0125] The key parameters of the NCC area boundary are solved through geometric relationships, and the resulting release strategy is the optimal release strategy that satisfies the minimum speed and heading change. First, let the tangent point of l3 and the optimized ellipse be P τ =(x τ ,y τ ); secondly, the optimized ellipse standard equation is differentiated with respect to x; thirdly, the tangent point P τ The derivative obtained at is equivalent to the slope of l3, and the derivative with respect to x is obtained. τ and τ Fourth, substitute the tangent point into the equation of the ellipse and combine the equations obtained in the previous step to solve for x τ ; Finally, we get y τ , The function of the line where the boundary l3 lies is The calculation method of the above variables is given by formula (13):

[0126]

[0127] like Figure 6 As shown in (a), solve for the release speed v′ Ni When the heading θ is not changed Ni , The straight line and the tangent equation The intersection point M = (x m ,y m ), through point M relative to point P Ni The horizontal and vertical components are obtained as and Solve for v′ Ni , the calculation method is given by formula (14):

[0128]

[0129] Similarly, for the NCC region boundary l4, a similar conclusion is obtained, and the release heading v″ Ni .

[0130] like Figure 6 As shown in (b), solve the release heading θ′ Ni When the velocity v does not change Ni ,vector With tangent Intersecting at point G, Relative to P Ni (x Ni ,y Ni ) is the horizontal and vertical components of and point Bring in The equation is solved by combining the relationship equation between the two components to obtain θ′ Ni , the calculation method is given by formula (15):

[0131]

[0132] Similarly, for the NCC region boundary l4, a similar conclusion is obtained, and the release heading θ″ Ni .

[0133] Example 2

[0134] According to the design and analysis of the improved speed barrier method in Example 1, the steps for aircraft cluster conflict resolution are obtained, such as: Figure 7 As shown. This embodiment is divided into two parts, namely: (1) preprocessing stage and (2) solution stage. (1) After extracting aircraft information, data preprocessing is performed through a clustering model, and the speed vector of the aircraft in the cluster is fine-tuned using a cluster control algorithm. (2) An improved speed obstacle model is established between clusters, and conflict detection is performed through geometric analysis of the model, and the optimal tangent release strategy including release speed and release heading is calculated. In order to verify the reliability of the conflict detection algorithm of this study, Python is used to perform simulation analysis on the above theory. Based on Example 1, the specific implementation method of this embodiment is as follows:

[0135] In step A:

[0136] As shown in Table 1, the initial position, heading, speed and protection zone size information of three groups of aircraft pairs of different models in the cross conflict scenario are given. In this embodiment, an accurate aircraft protection interval distance strategy is adopted to define the short semi-axis of the protection zone, and the length of the long semi-axis is determined by the speed, aiming to ensure that the long semi-axis is longer than the short semi-axis, while ensuring that its value does not exceed the traditional 15km interval, and the value of λ is 0.3. Based on this parameter setting, the improved speed obstacle method is used to calculate the required release speed and release heading.

[0137] Table 1 Aircraft posture information of three groups of different aircraft models at different accuracy intervals

[0138]

[0139] The results are shown in Table 2. In the same experimental group, the variable protection zone based on speed reduces the adjustment of the release heading and release speed by 24.07% at the minimum compared with the traditional protection zone. By comparing the three experimental groups, it can be seen that different types of aircraft have the same result, that is, the adjustment amount is reduced. It is not difficult to see that the release strategy obtained by the variable protection zone based on speed is easier to implement, and this method is more feasible in actual flight.

[0140] Table 2 Traditional interval and dynamic elliptical protection zone release strategies

[0141]

[0142] In step B:

[0143] The aircraft posture information of 4 different aircraft types are randomly selected, as shown in Table 3. The aircraft clustering model is used to cluster 10 aircraft with an aircraft interval of no more than 10 km, a heading difference of no more than 5°, and a speed difference of no more than 0.5 km / min. The results are as follows Figure 8 As shown in the figure, the 10 aircraft are divided into 6 clusters and there is no conflict within the clusters. Each cluster is represented by a color. Cluster S = {S1, S2, S3, S4, S5, S6}, where S2={P3,P4}, S1={P2}, S2={P7} S1={P8}, S2={P9}i=1,2,3.

[0144] Table 3 Aircraft posture information of four different aircraft types

[0145]

[0146] In step CD:

[0147] The cluster optimization ellipse model is used to characterize the mass points, velocity vectors, and protected area sizes of clusters S1 and S2. The cluster ellipse information obtained for clusters S1 and S2 is as follows: Fig. 9and 10 As shown in Table 4, the ellipse area of ​​S1 meets the condition when it is 453.998 km, at which a=16.55 km, b=8.73 ​​km, the angle between the major axis of the ellipse and x is 4.253°, the speed is 13.512 km, and the center coordinates are (53.75, 74.75); the ellipse area of ​​S2 meets the condition when it is 211.927 km, at which a=10.125 km, b=6.6625 km, the angle between the major axis of the ellipse x is 59.999°, the speed is 13.85 km, and the center coordinates are (80.0, 49.0). As shown in Table 4, by calculating the speed difference and angle difference between the aircraft and the cluster to which it belongs, it is found that the speed difference between the two is in the range of [0-0.2] and the angle difference is in the range of [0-3]. The adjusted speed and heading of the aircraft are within the navigation accuracy error range, so the method of making slight adjustments to the velocity vector of each aircraft is feasible. The velocity vector of each aircraft is fine-tuned to obtain a new velocity vector, namely, the cluster velocity vector.

[0148] Table 4 Difference between aircraft velocity vector and cluster velocity vector

[0149]

[0150]

[0151] In Step EF:

[0152] When there are m cluster aircraft and n cluster obstacle aircraft, the number of conflict judgments of the traditional speed obstacle method is m*n times. The reliability of this method is verified by analyzing the conflict resolution problem of aircraft in clusters S1 and S2. For the 4 aircraft given in the obstacle aircraft cluster S1 and the 2 aircraft given in the aircraft cluster S2, a total of 8 conflict judgments are required. Fig.11 In the simulation, the closest conflict time between aircraft pairs (two aircraft) is T = 1.35min. From (a)-(d), we can get that aircraft P3 and P 10 After 1.35 minutes, a collision occurred, and aircraft P4 passed by P1, P5, and P6. From (e) to (h), it can be seen that aircraft P4 and P1, P5, and P6 collided. 10 Conflicts occurred at 2.40min, 3.10min, and 2.00min respectively, and P6 passed by at the same time. Obviously, there were 4 conflicts between the two cluster aircraft within 5 minutes, and the conflict time was short and the urgency was high. The speed barrier method S1 and S2 improved by clustering only need one conflict judgment, and its calculation efficiency is improved by more than 87.5% compared with the traditional method. Fig.12 In (a), the barrier cone is formed by the improved velocity barrier method, which is composed of the velocity vector S2. Falling into the NCC area, it is known that S2 and S1 will conflict at some point in the future. After geometric analysis, the release speed and release heading are obtained, as shown in Table 5. The aircraft release strategy of the improved speed obstacle method is to obtain release by selecting an area outside the NCC boundary. i =4km / min and release heading θ i =12° to verify the reliability of this method. From (b) and (c), we can see that the velocity vector S2 As long as it does not fall within the NCC area, there will be no conflict between the two in the future.

[0153] In (d), after a step size of 0.05 and a simulation time of 8 minutes, it is found that S1 and S2 collide at T = 1.35 minutes. Compared with the traditional speed barrier method, the conflict time of cluster pairs (two clusters are a pair) obtained by the present invention is consistent with the closest conflict time of aircraft pairs obtained by the traditional speed barrier method, which verifies that the present invention is effective in aircraft conflict identification. The conflict time of cluster pairs is the red area in the simulation figure, and the closer the distance between the two, the darker the trajectory color. From the simulation of (e) and (f), it can be seen that there is no conflict between the two during flight.

[0154] In (g), the distance between the cluster pairs gradually decreases to 0, and the duration is 2.35 minutes. In (h) and (l), the distance between the cluster pairs first decreases and then increases, and the distance between the two does not fall below 0 during this period. For multi-obstacle collision avoidance, multiple areas outside the NCC area are selected to obtain the conflict resolution strategy. The principle is the same and will not be repeated here. The larger the number of clusters, the more significant the advantage of the improved speed barrier method, which has practical significance for the dynamic planning of routes. The application of this method can also reduce the operating costs of airlines, reduce the workload of controllers, and further reduce unnecessary fuel consumption and exhaust emissions in the future.

[0155] Table 5 Aircraft escape strategy

[0156]

[0157] The above-mentioned embodiments only express the specific implementation of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. An improved speed barrier method for clustering, characterized in that: The following steps are involved: Step 1: construct a variable precise elliptical protection zone according to the aircraft speed difference; Step 2: Build a clustering model based on specific rules to cluster aircraft with similar speed, heading, and close distance to form clusters; In step 2, the clustering conditions are defined: the distance, heading difference, and speed difference between aircraft are not greater than a certain threshold. The clustering model construction method based on specific rules is as follows: (1) Traverse the points on the ellipse and calculate the shortest distance between two points belonging to different ellipses to obtain the aircraft interval; (2) Check whether the three clustering conditions are met simultaneously; (3) The clusters that meet the clustering conditions are a cluster, and each cluster is distinguished; Step 3: Characterize the mass points, velocity vectors, and protected area size of the cluster, where the protected area size is determined by establishing an optimization model with the goal of minimizing the protected area of ​​the cluster and the constraints of no conflict between aircraft and the aircraft being located inside the cluster; The step 3 is as follows: Each divided cluster must satisfy the requirement that there is no internal conflict. If the aircraft within the cluster, i.e., the ellipses, overlap, it is considered that there is a conflict between aircraft. The conflict area S of the aircraft is calculated by the following formula: Where: (x i ,y i ) is the coordinate of the center of the target aircraft ellipse; (x, y) is the coordinate point on the protection zone ellipse formed by the potential conflicting aircraft; a i The major semi-axis of the protection zone ellipse for the target aircraft; b i The major semi-axis of the protection zone ellipse for the target aircraft; The cluster is characterized by an ellipse, which is called a cluster ellipse. According to the characteristics of the aircraft position, velocity vector, and protected area size in the cluster, a mass point, velocity vector, and protected area size representing the cluster ellipse are found. The angle between the cluster ellipse and the x-axis adopts the combined velocity direction angle. The optimal centroid is the midpoint of the rectangle formed by all the ellipse boundary points in this direction. The mean value of each aircraft coordinate is used as the cluster centroid. n is the number of aircraft. Let the cluster centroid be P. Ni =(x Ni ,y Ni ), which is calculated as follows: Velocity vector of the cluster Including velocity direction and velocity magnitude, where the angle θ between the velocity direction and the positive half axis of the x-axis Ni is the angle between the aircraft's combined velocity vector and the positive half axis of the x-axis. The velocity magnitude v Ni is the average speed of each aircraft and is calculated as follows: The protection zone of the cluster ellipse is solved by constructing a cluster optimization model; The method for constructing the optimization model of cluster ellipse is as follows: (1) Taking the minimum area of ​​the cluster ellipse as the objective function; (2) Traverse the points on each ellipse and make sure that each point substituted into the cluster ellipse formula satisfies a constraint of less than or equal to 1, so that each ellipse, i.e., the aircraft, is inside the cluster ellipse; (3) Taking the centroid as the cluster center, θ Ni is the cluster velocity vector The angle with the x-axis is used to generate a cluster ellipse with the smallest area that contains all ellipses, and the major semi-axis a at this time is obtained. Ni and the minor axis b Ni The value of Step 4: Use the cluster control algorithm to make slight adjustments to the velocity vectors of the aircraft in the cluster to ensure that the aircraft in the cluster will not leave the cluster and collide during the obstacle avoidance period; Step 5: Improve and optimize the speed obstacle method based on the concept of clustering, combine the constraint of the maximum change per unit time, and use plane geometry analysis to solve the obstacle avoidance strategy of the aircraft cluster including speed release and heading release.

2. The improved speed barrier method for clustering according to claim 1, characterized in that: The following steps are also included: Step 6: Use Python to simulate the conflicts between clusters and verify the reliability of the speed obstacle avoidance strategy for cluster division.

3. The improved speed barrier method for clustering according to claim 1, characterized in that: In step 1, an ellipse is used to describe the safety range of aircraft of different types, and the center of the circle is defined as P i =(x i ,y i ), the velocity vector is The velocity is v i , the velocity vector Located on the long axis and with the positive x-axis at an angle of θ i ,θ i ∈(0,2π), the aircraft safety range is determined by the precise safety interval of different aircraft types and the current cruising speed, where the short semi-axis b i Take the precise safety interval corresponding to different models, the major axis a i It is related to the cruising speed, λ is the correlation coefficient, and the above variables are calculated by the following formula:

4. The improved speed barrier method for clustering according to claim 1, characterized in that: In step 4, the cluster control algorithm satisfies the alignment rule, that is, aircraft i calculates the combined velocity direction of all aircraft and the average velocity of the aircraft to obtain the cluster velocity vector, and then tries to align it. The heading adjustment method is as follows:

5. The improved speed barrier method for cluster division according to claim 4, characterized in that: The step 5 is specifically as follows: In the distributed multi-aircraft motion coordination, each aircraft regards other aircraft as motion obstacle aircraft. Assume that the square radiation distance of the aircraft is a i , the side radiation distance is b i , the square radiation distance of a dynamic obstacle aircraft is a j , the side radiation distance is b j At time t1, the positions of the aircraft and the obstacle aircraft are P i , P j , the velocity vectors are The speeds are v i and v j , the yaw angles are θ i and θ j , the position information of the aircraft is (P i , ), the position information of the obstacle aircraft is (P j , ), the above variables are calculated by the following formula: P j is the center of the circle, a ij is the major axis, b ij As the minor axis, draw an ellipse with P i Cone and ellipse P for the vertex j Tangent, tangent to rays l1 and l2, where is the relative velocity vector of the two, l m is a vector The above variables are calculated as follows: If the vector If the end is within the speed barrier cone, the dynamic obstacle aircraft poses a threat to the aircraft; define the collision area set CC in the speed space as follows: where ⊙P j The safety protection zone that represents the expansion of the obstructing aircraft is the elliptical area; P i and ⊙P j Along vector The translation distance in the direction Get Q i =(x′ i ,y′ i ), Q j =(x′ j ,y′ j ), the unit time interval is Δt, and the above variables are calculated by the following formula: Move CC horizontally The resulting area is called the new collision area NCC, as shown below: in represents the Minkowski vector and operation; By improving the speed barrier method after dividing the clusters, a speed barrier pile is established on the basis of the clusters to realize the identification and resolution of conflicts between clusters; At time t1, assuming that the square radiation distance of the aircraft cluster is a Ni , the side radiation distance is b Ni , the square radiation distance of a dynamic threat obstacle aircraft is a Nj , the side radiation distance is b Nj ; The positions of the cluster aircraft and the cluster obstacle aircraft are P Ni , P Nj , the velocity vectors are The speeds are v Ni and v Nj , the yaw angles are θ Ni and θ Nj , and then by P Ni and ⊙P Nj Along vector The translation distance in the direction Get Q Ni =(x′ Ni ,y′ Ni ), Q Nj =(x′ Nj ,y′ Nj ) Nj is the center of the circle, a Nij is the major axis, b Nij As the minor axis, draw an ellipse with Q Ni Cone and ellipse Q for the vertex Nj Tangent, tangent to rays l3 and l4, the above variables are calculated by the following formula, According to the improved speed obstacle model, the speed of the aircraft Now it falls into the NCC area and should be adjusted Make it fall outside the NCC area; because the aircraft is subject to its own performance constraints, the adjustment range can only be within the constraints; define the velocity vector The angle with the horizontal plane is θ Ni , aircraft release speed and the release heading θ′ Ni The allowable adjustment range within unit time Δt is as follows: The key parameters of the NCC area boundary are solved through geometric relationships, and the resulting release strategy is the optimal release strategy that satisfies the minimum speed and heading change. First, let the tangent point of l3 and the optimized ellipse be P τ =(x τ ,y τ ); secondly, the optimized ellipse standard equation is differentiated with respect to x; thirdly, the tangent point P τ The derivative obtained at is equivalent to the slope of l3, and the derivative with respect to x is obtained. τ and τ Fourth, substitute the tangent point into the equation of the ellipse and combine the equations obtained in the previous step to solve for x τ ; Finally, we get y τ , The function of the line where the boundary l3 lies is The above variables are calculated as follows: Solve for the release velocity v′ Ni When the heading θ is not changed Ni , The straight line and the tangent equation The intersection point M = (x m ,y m ), through point M relative to point P Ni The horizontal and vertical components are obtained as and Solve for v′ Ni , calculated as follows: Similarly, for the NCC region boundary l4, a similar conclusion is obtained, and the release heading v″ Ni ; Solve for the release heading θ′ Ni When the velocity v does not change Ni ,vector With tangent Intersecting at point G, Relative to P Ni (x Ni ,y Ni ) is the horizontal and vertical components of and point Bring in The equation is solved by combining the relationship equation between the two components to obtain θ′ Ni , calculated as follows: Similarly, for the NCC region boundary l4, a similar conclusion is obtained, and the release heading θ″ Ni .