An aircraft positioning guidance method with adaptive adjustment of detection indicators in 2V2 scenarios
By dynamically adjusting the radar position and detection strategy of the air-based platform in 2V2 air combat scenarios, calculating the situation advantage function and optimizing the flight trajectory, the problem of detection indicator optimization in air combat is solved, the air victory rate and battlefield air supremacy are improved, and independent intelligent air combat is realized support.
Patent Information
- Application Number
- CN202211140120.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-20
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2042-09-20
AI Technical Summary
In the 2V2 air combat scenario, it is difficult for the existing technology to effectively optimize aircraft detection indicators, improve air victory rate and battlefield air supremacy, and autonomous placeholding guidance strategies are difficult to achieve intelligent autonomous decision-making in complex environments.
By dynamically adjusting the radar position and detection strategy of the air-based platform in an air combat environment, calculate the situation advantage function between aircraft, adjust the weights adaptively, and optimize the flight trajectory to improve detection capabilities and air combat advantages.
It realizes autonomous space-occupation guidance of the aircraft in 2V2 scenarios, improves the air victory rate, reduces the burden on pilots, enhances the battlefield air supremacy, and reduces risks through simulation verification.
Smart Images

Figure CN115328207B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of aviation technology, and in particular relates to an aircraft occupancy guidance method with adaptive adjustment of detection indicators in a 2V2 scenario. Background Art
[0002] The development trend of modern air combat is automation, integration, and intelligence. It is impossible to rely solely on pilots to consider many factors and then make all the planning and decision-making operations in air combat. Autonomous unmanned positional guidance strategies can effectively evaluate the combat situation of the enemy and our side in air combat scenarios, intelligently and autonomously improve the air combat advantage of our aircraft in the current scenario, and further predict the possible action strategies for enemy targets in the future based on the extracted situational elements. Through autonomous positional guidance strategies, the air victory rate can be effectively improved and air superiority on the battlefield can be obtained; at the same time, by verifying the positional guidance strategy in a simulated simulation environment, it can effectively reduce risks and provide support for autonomous and intelligent air combat. Summary of the invention
[0003] In order to improve the air combat detection capability, the present invention proposes an autonomous positioning and guidance scheme for our aircraft coordinated formation in 2V2 air combat scenarios. By changing the dynamic position and detection strategy of the air-based platform radar in the air combat environment, various detection indicators of the enemy aircraft are optimized, thereby achieving a dominant air combat situation and improving the air victory rate.
[0004] The aircraft positioning guidance method with adaptive adjustment of detection indicators in a 2V2 scenario includes the following steps:
[0005] (1) The current flight direction of the aircraft is determined based on the velocity information in the absolute coordinate system (north celestial east coordinate system), and this direction is approximately regarded as the center direction of the aircraft radar detection range. The azimuth angle (ATA) and the cosine value of the enemy aircraft approach angle (TAA) of a certain aircraft of our side relative to a certain enemy aircraft are determined based on the current position and velocity direction of the aircraft and the enemy aircraft, and then the perception situation advantage function J of the aircraft relative to the enemy aircraft is calculated.
[0006] The calculation method of the situational advantage function J of a single aircraft is as follows:
[0007]
[0008]
[0009] J angle =cos(ATA)+cos(TAA)
[0010]
[0011] J single =J angle ×Jdistance
[0012] Where: pn me is the north position of the aircraft, pu me is the celestial position of the local machine, pe me is the east position of the aircraft, vn me is the northbound speed of the aircraft, vu me is the local celestial speed, ve me is the eastward speed of the aircraft, pn enemy is the target north position, pu enemy is the target celestial position, pe enemy is the target east position, vn enemy is the target north speed, vu enemy is the target celestial velocity, ve enemy is the target eastward speed, R me-ennmy is the distance at which the enemy aircraft cannot escape, and fRtr_km is the maximum detection distance of the radar.
[0013] (2) According to the situational advantage functions between our two aircraft and the enemy's two aircraft, as well as the radar's detection of the enemy aircraft, the current situation of the enemy and our aircraft formation in the 2V2 scenario is determined, and the weights of the four advantage functions J of our aircraft over the enemy aircraft are adaptively adjusted to determine the optimization indicators and conditional constraints of our aircraft in the current 2V2 scenario.
[0014] The calculation process of the overall situation advantage function is as follows:
[0015] Let our aircraft be numbered 1 and 2, and the enemy aircraft be numbered A and B. The overall situation advantage function J all The situation advantage function J of 1 machine against A machine, 2 machines against A machine, 1 machine against B machine, 2 machines against B machine 1A ,J 1B ,J 2A ,J 2B And the penalty function J punish Composition, classified according to different detection situations in the 2V2 scenario, and determined λ 1 ,λ 2 ,λ 3 ,λ 4 ,λ 5 ,λ 6 Six weights.
[0016] J 1A =J angle (ATA 1A ,TAA 1A )J distance (R 1-A )
[0017] J2A =J angle (ATA 2A ,TAA 2A )J distance (R 2-A )
[0018] J 1B =J angle (ATA 1B ,TAA 1B )J distance (R 1-B )
[0019] J 2B =J angle (ATA 2B ,TAA 2B )J distance (R 2-B )
[0020] J all =λ 1 J 1A +λ 2 J 2A +λ 3 J 1B +λ 4 J 2B -(λ 5 J punis h1 +λ 6 J punis h2 )
[0021] Among them, J punis h1 ,J punis h2 is the penalty function for the overlap between our two aircraft. It is defined as the farthest position f(position 1 ,velocity 1 ,velocity 2 ) and the cosine value of the angle between the speed direction of the friendly aircraft, the expression is as follows:
[0022]
[0023] J punis h1 =f(postion 1 ,velocity 1 ,velocity 2 )
[0024] J punis h2 =f(postion 2 ,velocity 2 ,velocity 1 )
[0025] Taking the discrete time case as an example to determine the overall situation advantage function, the weight [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T ]The calculation process is as follows, where T is the number of time steps experienced at the current moment.
[0026]
[0027] According to our aircraft radar detection information, the battlefield situation is divided into 7 situations:
[0028] S1. Neither of our aircraft detected the enemy aircraft.
[0029] Λ 1 =diag{1,1,1,1,1,1}
[0030] In different situations, only the weight coefficient Λ changes, [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T The calculation process remains unchanged. S2. Only one of our aircraft's radars detects an enemy aircraft (for example, our aircraft 1 detects enemy aircraft A, our aircraft 2 detects A, our aircraft 1 detects enemy aircraft B, and our aircraft 2 detects enemy aircraft B in the same way).
[0031] The expression of weight coefficient Λ is:
[0032] Λ 2 =diag{l,s,s,l,0,1}
[0033] Among them: l represents the coefficient of weight increase, which means increasing the importance of a certain indicator, l>0;
[0034] s represents the coefficient of weight reduction, which means reducing the importance of a certain indicator, s<0; the same applies to the following text (l and s in each case are adjusted according to the test results).
[0035] S3. The radar of one of our aircraft detected two enemy aircraft, but the friendly aircraft did not detect any enemy aircraft (for example, our aircraft 1 detected enemy aircraft A and B, and the same applies to our aircraft 2 detecting enemy aircraft A and B).
[0036] When J 2A >J 2B When , the expression of weight coefficient Λ is:
[0037] Λ 31 =diag{1,1,l,s,0,0}
[0038] When J2A <J 2B When , the expression of weight coefficient Λ is:
[0039] Λ 32 =diag{1,1,s,l,0,0}
[0040] S4. The radars of two of our aircraft detect the same enemy aircraft (for example, our aircraft 1 and 2 detect enemy aircraft A, and the same applies to our aircraft 1 and 2 detecting enemy aircraft B).
[0041] When J 1A >J 2A When , the expression of weight coefficient Λ is:
[0042] Λ 41 =diag{l,s,l,s,0,0}
[0043] When J 1A <J 2A When , the expression of weight coefficient Λ is:
[0044] Λ 42 =diag{s,l,s,l,0,0}
[0045] S5. The radars of two friendly aircraft respectively detect a different enemy aircraft 1-A and 2-B (for example, when aircraft 1 detects enemy aircraft A and aircraft 2 detects enemy aircraft B, the same applies when aircraft 1 detects enemy aircraft B and aircraft 2 detects enemy aircraft A).
[0046] The expression of weight coefficient Λ is:
[0047] Λ 5 =diag{l,s,l,s,0,0}
[0048] S6. The radar of one of our aircraft detects two enemy aircraft, and the friendly aircraft detects one of the enemy aircraft (taking the case where our aircraft 1 detects enemy aircraft A and B and our aircraft 2 detects enemy aircraft A as an example, our aircraft 1 detects enemy aircraft A and B and our aircraft 2 detects enemy aircraft B, our aircraft 2 detects enemy aircraft A and B and our aircraft 1 detects enemy aircraft A, our aircraft 2 detects enemy aircraft A and B and our aircraft 2 detects enemy aircraft B, and the same applies).
[0049] The expression of weight coefficient Λ is:
[0050] Λ 6 =diag{1,l,l,s,0,0}
[0051] S7. Both our aircraft detected two enemy aircraft.
[0052] This situation can be divided into the following six sub-situations according to the situation advantage function of a single machine:
[0053] S71. When J 1A >J2A And J 2B >J 1B When , the weight coefficient Λ is expressed as:
[0054] Λ 71 =diag{l,1,1,l,0,0}
[0055] S72. When J 1A <J 2A And J 2B <J 1B When , the weight coefficient Λ is expressed as:
[0056] Λ 72 =diag{1,l,l,1,0,0}
[0057] S73. When J 1A >J 1B >J 2B >J 2A or J 2B >J 2A >J 1A >J 1B When , the expression of weight coefficient Λ is:
[0058] Λ 73 =diag{l,1,1,l,0,0}
[0059] S74. When J 1B >J 1A >J 2A >J 2B or J 2A >J 2B >J 1B >J 1A When , the expression of weight coefficient Λ is:
[0060] Λ 74 =diag{1,l,l,1,0,0}
[0061] S75. When the above conditions are not met, if J 1A -J 2A >J 1B -J 2B , the expression of weight coefficient Λ is:
[0062] Λ 75 =diag{l,1,1,l,0,0}
[0063] S76. When the above conditions are not met, if J 1A -J 2A <J 1B -J 2B , the expression of weight coefficient Λ is:
[0064] Λ 76 =diag{1,l,l,1,0,0}
[0065] (3) Taking the current moment as the starting point, the discretized sum of the optimization indicators within a prediction window is taken as the function to be optimized. Under the constraints corresponding to the current situation, the optimization problem with the flight state trajectory as the independent variable is solved; the flight trajectory within the control window with the current moment as the starting point and less than or equal to the length of the prediction window is selected as the real placeholder guidance solution.
[0066] The overall situation advantage function optimization problem solution expression and constraints are as follows: According to the weights calculated in step (2), the overall situation advantage function J at the current time T is calculated: all . Let P be the number of time steps in the prediction window, Keep the current weight value [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T ] all The sum of discrete time steps T+1 to T+P is used as the index function to be optimized, and the optimization problem with the flight trajectory as the independent variable is solved under the constraint of the aircraft's maneuverability. The optimization function and boundary conditions are as follows:
[0067]
[0068] Boundary conditions:
[0069] V min ≤||[V n (T+i),V u (T+i),V e (T+i)]|| 2 ≤V max ,i∈{1,…,P}||[V n (T+i)-V n (T+i-1),V u (T+i)-V u (T+i-1),V e (T+i)-V e (T+i-1)]|| 2 ≤ΔV max ,i∈{1,…,P}
[0070] |atan2(V n (T+i),V e (T+i))-atan2(V n (T+i-1),Ve (T+i-1))|≤αheading max ,i∈{1,…,P}
[0071] |atan2(V u (T+i),||V n (T+i),V e (T+i)|| 2 )-atan2(V u (T+i-1),||V n (T+i-1),V e (T+i-1)|| 2 )|≤Δpitch max ,i∈{1,…,P}
[0072] pitch min ≤atan2(V u (T+i),||V n (T+i),V e (T+i)|| 2 )≤pitch max ,i∈{1,…,P}
[0073] P n (T+i)=P n (T+i-1)+ΔtV n (T+i),i∈{1,…,P}
[0074] P u (T+i)=P u (T+i-1)+ΔtV u (T+i),i∈{1,…,P}
[0075] P e (T+i)=P e (T+i-1)+ΔtV e (T+i),i∈{1,…,P}
[0076] H min ≤P u (T+i)≤H max ,i∈{1,…,P}
[0077] Where: atan2(x,y) is a function for calculating the inverse tangent of y / x (library functions in languages such as C, Python, and Matlab can be used), and the value is (-π,π]; Δt is the discrete time step; V min ,V max Respectively represent the minimum and maximum speeds allowed for aircraft flight; ΔV maxRepresents the maximum change in speed of the aircraft within Δt time; Δheading max Represents the maximum change in the aircraft's yaw angle within Δt time; Δpitch max Represents the maximum change in the aircraft's pitch angle within Δt time; pitch min ,pitch max They represent the minimum and maximum speeds allowed for an aircraft to fly; H min ,H max Respectively represent the minimum and maximum altitudes allowed for an aircraft to fly. n ,V u ,V e Represent the aircraft's north, sky, and east speeds, respectively. n ,P u ,P e They represent the north, sky, and east positions of the aircraft respectively. The maneuverability constraints of each aircraft are common, and the subscripts (1, 2, A, B) of the aircraft numbers are omitted in the expression of the constraints. In the optimization solution, the specific values of the constraints are appropriately selected according to the maneuverability of each aircraft.
[0078] (4) Guiding the aircraft to fly according to the position-occupying guidance plan within the control window. Repeating the above steps after the control window time forms a position-occupying guidance plan generated by rolling optimization.
[0079] The method of generating a placeholder guidance solution by rolling optimization is as follows:
[0080] According to the optimization result obtained in step (3), C is the number of time steps in the control window, C≤P, The optimization results from time steps T to T+C are used as the actual output to guide the flight trajectory. At time T+C, the overall situation advantage function J is calculated based on the situation at time T+C. all , repeat this step for T+C to T+C+P, perform a new round of optimization problem calculation, and realize the online optimization generation of rolling occupancy guidance trajectory.
[0081] The present invention provides a novel aircraft positioning guidance method with adaptive adjustment of detection optimization indexes, which comprehensively utilizes battlefield situation and perception information to guide aircraft flight, can reduce the burden on pilots during combat, enable our aircraft to occupy a situational advantage during air combat, effectively improve the air victory rate, and obtain battlefield air supremacy. BRIEF DESCRIPTION OF THE DRAWINGS
[0082] Figure 1 It is an implementation flow chart of the present invention.
[0083] Figure 2 is the monomer situation advantage function in the embodiment of the present invention [J 1A ,J 2A ,J 1B,J 2B ]Trend chart over time.
[0084] Figure 3 is the overall situation advantage function J in the embodiment of the present invention all Trend graph over time. DETAILED DESCRIPTION
[0085] The present invention will be further described below in conjunction with the accompanying drawings.
[0086] The aircraft positioning guidance method with adaptive adjustment of detection indicators in a 2V2 scenario includes the following steps:
[0087] (1) The current flight direction of the aircraft is determined based on the speed information in the absolute coordinate system (North Celestial East coordinate system), and this direction is approximately regarded as the center direction of the aircraft radar detection range. The azimuth angle (ATA) and the cosine value of the enemy aircraft approach angle (TAA) of a certain aircraft of our side relative to a certain enemy aircraft are determined based on the current position and speed direction of our aircraft and the enemy aircraft, and then the advantage function J of the perception situation between our aircraft and the enemy aircraft is calculated. The calculation method of the advantage function J of the perception situation of a single aircraft is as follows:
[0088]
[0089]
[0090] J angle =cos(ATA)+cos(TAA)
[0091]
[0092] J single =J angle ×J distance
[0093] Where: pn me is the north position of the aircraft, pu me is the celestial position of the local machine, pe me is the east position of the aircraft, vn me is the northbound speed of the aircraft, vu me is the local celestial speed, ve me is the eastward speed of the aircraft, pn enemy is the target north position, pu enemy is the target celestial position, pe enemy is the target east position, vn enemy is the target north speed, vu enemy is the target celestial velocity, ve enemy is the target eastward speed, R me-ennmyis the distance at which the enemy aircraft cannot escape, and fRtr_km is the maximum detection distance of the radar.
[0094] (2) According to the situational advantage functions of our two aircraft over the enemy aircraft, as well as the detection of enemy aircraft by radar, the current situation of the enemy and our aircraft formation in the 2V2 scenario is determined, and the weights of the four advantage functions J of our aircraft over the enemy aircraft are adaptively adjusted to determine the optimization indicators and conditional constraints of our aircraft in the current 2V2 scenario.
[0095] Let our aircraft be numbered 1 and 2, and the enemy aircraft be numbered A and B. The overall situation advantage function J all The situation advantage function J of 1 machine against A machine, 2 machines against A machine, 1 machine against B machine, 2 machines against B machine 1A ,J 1B ,J 2A ,J 2B And the penalty function J punish Composition, classified according to different detection situations in the 2V2 scenario, and determined λ 1 ,λ 2 ,λ 3 ,λ 4 ,λ 5 ,λ 6 The weights of the six items.
[0096] J 1A =J angle (ATA 1A ,TAA 1A )J distance (R 1-A )
[0097] J 2A =J angle (ATA 2A ,TAA 2A )J distance (R 2-A )
[0098] J 1B =J angle (ATA 1B ,TAA 1B )J distance (R 1-B )
[0099] J 2B =J angle (ATA 2B ,TAA 2B )J distance (R 2-B )
[0100] J all =λ1 J 1A +λ 2 J 2A +λ 3 J 1B +λ 4 J 2B -(λ 5 J punish1 +λ 6 J punish2 )
[0101] Among them, J punish1 ,J punish2 is the penalty function for the overlap between our two aircraft. It is defined as the farthest position f(position 1 ,velocity 1 ,velocity 2 ) and the cosine value of the angle between the speed direction of the friendly aircraft, the expression is as follows:
[0102]
[0103] J punish1 =f(postion 1 ,velocity 1 ,velocity 2 )
[0104] J punish2 =f(postion 2 ,velocity 2 ,velocity 1 )
[0105] Taking the discrete time case as an example, the overall situation advantage function is determined, and the weight [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T ]The calculation process is as follows, where T is the number of time steps experienced at the current moment.
[0106]
[0107] According to our aircraft radar detection information, the battlefield situation is divided into 7 situations:
[0108] S1. Neither of our aircraft detected the enemy aircraft.
[0109] Λ 1 =diag{1,1,1,1,1,1}
[0110] In different situations, only the weight coefficient Λ changes, [λ1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T The calculation process remains unchanged. S2. Only one of our aircraft's radars detects an enemy aircraft (for example, our aircraft 1 detects enemy aircraft A, our aircraft 2 detects A, our aircraft 1 detects enemy aircraft B, and our aircraft 2 detects enemy aircraft B in the same way).
[0111] The expression of weight coefficient Λ is:
[0112] Λ 2 =diag{l,s,s,l,0,1}
[0113] Among them: l represents the coefficient of weight increase, which means increasing the importance of a certain indicator, l>0;
[0114] s represents the coefficient of weight reduction, which means reducing the importance of a certain indicator, s<0; the same applies to the following text (l and s in each case are adjusted according to the test results).
[0115] S3. The radar of one of our aircraft detected two enemy aircraft, but the friendly aircraft did not detect any enemy aircraft (for example, our aircraft 1 detected enemy aircraft A and B, and the same applies to our aircraft 2 detecting enemy aircraft A and B).
[0116] When J 2A >J 2B When , the expression of weight coefficient Λ is:
[0117] Λ 31 =diag{1,1,l,s,0,0}
[0118] When J 2A <J 2B When , the expression of weight coefficient Λ is:
[0119] Λ 32 =diag{1,1,s,l,0,0}
[0120] S4. The radars of two of our aircraft detect the same enemy aircraft (for example, our aircraft 1 and 2 detect enemy aircraft A, and the same applies to our aircraft 1 and 2 detecting enemy aircraft B).
[0121] When J 1A >J 2A When , the expression of weight coefficient Λ is:
[0122] Λ 41 =diag{l,s,l,s,0,0}
[0123] When J 1A <J 2A When , the expression of weight coefficient Λ is:
[0124] Λ 42 =diag{s,l,s,l,0,0}
[0125] S5. The radars of two friendly aircraft respectively detect a different enemy aircraft 1-A, 2-B (for example, when aircraft 1 detects enemy aircraft A and aircraft 2 detects enemy aircraft B, the same applies when aircraft 1 detects enemy aircraft B and aircraft 2 detects enemy aircraft A).
[0126] The expression of weight coefficient Λ is:
[0127] Λ 5 =diag{l,s,l,s,0,0}
[0128] S6. The radar of one of our aircraft detects two enemy aircraft, and the friendly aircraft detects one of the enemy aircraft (taking the case where our aircraft 1 detects enemy aircraft A and B and our aircraft 2 detects enemy aircraft A as an example, our aircraft 1 detects enemy aircraft A and B and our aircraft 2 detects enemy aircraft B, our aircraft 2 detects enemy aircraft A and B and our aircraft 1 detects enemy aircraft A, our aircraft 2 detects enemy aircraft A and B and our aircraft 2 detects enemy aircraft B, and the same applies).
[0129] The expression of weight coefficient Λ is:
[0130] Λ 6 =diag{1,l,l,s,0,0}
[0131] S7. Both our aircraft detected two enemy aircraft.
[0132] This situation can be divided into the following six sub-situations according to the situation advantage function of a single machine:
[0133] S71. When J 1A >J 2A And J 2B >J 1B When , the weight coefficient Λ is expressed as:
[0134] Λ 71 =diag{l,1,1,l,0,0}
[0135] S72. When J 1A <J 2A And J 2B <J 1B When , the weight coefficient Λ is expressed as:
[0136] Λ 72 =diag{1,l,l,1,0,0}
[0137] S73. When J 1A >J 1B >J 2B >J 2A or J2B >J 2A >J 1A >J 1B When , the expression of weight coefficient Λ is:
[0138] Λ 73 =diag{l,1,1,l,0,0}
[0139] S74. When J 1B >J 1A >J 2A >J 2B or J 2A >J 2B >J 1B >J 1A When , the expression of weight coefficient Λ is:
[0140] Λ 74 =diag{1,l,l,1,0,0}
[0141] S75. When the above conditions are not met, if J 1A -J 2A >J 1B -J 2B , the expression of weight coefficient Λ is:
[0142] Λ 75 =diag{l,1,1,l,0,0}
[0143] S76. When the above conditions are not met, if J 1A -J 2A <J 1B -J 2B , the expression of weight coefficient Λ is:
[0144] Λ 76 =diag{1,l,l,1,0,0}
[0145] (3) Taking the current moment as the starting point, the discretized sum of the optimization indicators within a prediction window is taken as the function to be optimized, and the optimization problem with the flight state trajectory as the independent variable is solved under the condition constraints corresponding to the current situation; the flight trajectory within the control window with the current moment as the starting point and less than or equal to the length of the prediction window is selected as the real occupancy guidance solution.
[0146] According to the weight calculation in step (2), the overall situation advantage function J at the current time T is obtained all . Let P be the number of time steps in the prediction window, Keep the current weight value [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ5,T ,λ 6,T ] all The sum of discrete time steps T+1 to T+P is used as the index function to be optimized, and the optimization problem with the flight trajectory as the independent variable is solved under the constraint of the aircraft's maneuverability. The optimization function and boundary conditions are as follows:
[0147]
[0148] Boundary conditions:
[0149] V min ≤||[V n (T+i),V u (T+i),V e (T+i)]|| 2 ≤V max ,i∈{1,…,P}||[V n (T+i)-V n (T+i-1),V u (T+i)-V u (T+i-1),V e (T+i)-V e (T+i-1)]|| 2 ≤ΔV max ,i∈{1,…,P}
[0150] |atan2(V n (T+i),V e (T+i))-atan2(V n (T+i-1),V e (T+i-1))|≤Δheading max ,i∈{1,…,P}
[0151] |atan2(V u (T+i),||V n (T+i),V e (T+i)|| 2 )-atan2(V u (T+i-1),||V n (T+i-1),V e (T+i-1)|| 2 )|≤Δpitch max ,i∈{1,…,P}
[0152] pitch min ≤atan2(V u (T+i),||V n (T+i),V e (T+i)||2 )≤pitch max ,i∈{1,…,P}
[0153] P n (T+i)=P n (T+i-1)+ΔtV n (T+i),i∈{1,…,P}
[0154] P u (T+i)=P u (T+i-1)+ΔtV u (T+i),i∈{1,…,P}
[0155] P e (T+i)=P e (T+i-1)+ΔtV e (T+i),i∈{1,…,P}
[0156] H min ≤P u (T+i)≤H max ,i∈{1,…,P}
[0157] Where: atan2(x,y) is a function for calculating the inverse tangent of y / x (library functions in languages such as C, Python, and Matlab can be used), and the value is (-π,π]; Δt is the discrete time step; V min ,V max Respectively represent the minimum and maximum speeds allowed for aircraft flight; ΔV max Represents the maximum change in speed of the aircraft within Δt time; Δheading max Represents the maximum change in the aircraft's yaw angle within Δt time; Δpitch max Represents the maximum change in the aircraft's pitch angle within Δt time; pitch min ,pitch max They represent the minimum and maximum speeds allowed for an aircraft to fly; H min ,H max Respectively represent the minimum and maximum altitudes allowed for an aircraft to fly. n ,V u ,V e Represent the aircraft's north, sky, and east speeds, respectively. n ,P u ,P e They represent the north, sky, and east positions of the aircraft respectively. The maneuverability constraints of each aircraft are common, and the subscripts (1, 2, A, B) of the aircraft numbers are omitted in the expression of the constraints. In the optimization solution, the specific values of the constraints should be appropriately selected according to the maneuverability of each aircraft.
[0158] The specific solution to the optimization problem can be solved by using mature nonlinear optimization solving tools in programming languages (Matlab language fmincon function, Python language SciPy library optimize.minimize function can be used).
[0159] (4) Based on the optimization result obtained in step (3), let C be the number of time steps in the control window, C ≤ P, The optimization results from time steps T to T+C are used as the actual output to guide the flight trajectory. At time T+C, the overall situation advantage function J is calculated based on the situation at time T+C. all , repeat this step for T+C to T+C+P, perform a new round of optimization problem calculation, and realize the online optimization generation of rolling occupancy guidance trajectory.
Claims
1. An aircraft occupancy guidance method with adaptive adjustment of detection indicators in a 2V2 scenario, characterized in that: The steps include: (1) Determine the current flight direction of the aircraft based on the velocity information in the absolute coordinate system, and approximately regard this direction as the center direction of the aircraft radar detection range. Determine the azimuth angle (ATA) and the cosine value of the enemy aircraft approach angle (TAA) of a certain aircraft of our side relative to a certain enemy aircraft based on the current position and velocity direction of the aircraft and the enemy aircraft, and then calculate the perception situation advantage function J between the aircraft and the enemy aircraft; The calculation method of the situational advantage function J of a single aircraft is as follows: J angle =cos(ATA)+cos(TAA) I single =J angle ×J distance Where: pn me is the north position of the aircraft, pu me is the celestial position of the local machine, pe me is the east position of the aircraft, vn me is the northbound speed of the aircraft, vu me is the local celestial speed, ve me is the eastward speed of the aircraft, pn enemy is the target north position, pu enemy is the target celestial position, pe enemy is the target east position, vn enemy is the target north speed, vu enemy is the target celestial velocity, ve enemy is the target eastward speed, R me-ennmy is the distance at which the enemy aircraft cannot escape, and fRtr_km is the maximum detection distance of the radar; (2) According to the situation advantage functions between our two aircraft and the enemy aircraft, as well as the detection of enemy aircraft by radar, the current situation of the enemy and our aircraft formation in the 2V2 scenario is determined, and the weights of the four perception situation advantage functions J of our aircraft to the enemy aircraft are adaptively adjusted to determine the optimization indicators and conditional constraints of our aircraft in the 2V2 scenario; The calculation process of the overall situation advantage function is as follows: The subscripts of our aircraft are 1 and 2, and the subscripts of the enemy aircraft are A and B. The overall perception situation advantage function J all The situation advantage function J of 1 machine against A machine, 2 machines against A machine, 1 machine against B machine, 2 machines against B machine 1A ,J 1B ,J 2A ,J 2B And the penalty function J punish Composition: Classify according to different detection situations in the 2V2 scenario and determine the weights of the six items λ1, λ2, λ3, λ4, λ5, and λ6; <h2 style=";text-align:left;direction:ltr">J<h2 style=";text-align:left;direction:ltr"> 1A <h2 style=";text-align:left;direction:ltr"> =J<h2 style=";text-align:left;direction:ltr"> angle <h2 style=";text-align:left;direction:ltr"> (ATA<h2 style=";text-align:left;direction:ltr"> 1A <h2 style=";text-align:left;direction:ltr"> ,TAA<h2 style=";text-align:left;direction:ltr"> 1A <h2 style=";text-align:left;direction:ltr"> )J<h2 style=";text-align:left;direction:ltr"> distance <h2 style=";text-align:left;direction:ltr"> (R<h2 style=";text-align:left;direction:ltr"> 1-A <h2 style=";text-align:left;direction:ltr"> ) <h2 style=";text-align:left;direction:ltr">J<h2 style=";text-align:left;direction:ltr"> 2A <h2 style=";text-align:left;direction:ltr"> =J<h2 style=";text-align:left;direction:ltr"> angl <h2 style=";text-align:left;direction:ltr"> (ATA<h2 style=";text-align:left;direction:ltr"> 2A <h2 style=";text-align:left;direction:ltr"> ,TAA<h2 style=";text-align:left;direction:ltr"> 2A <h2 style=";text-align:left;direction:ltr"> )J<h2 style=";text-align:left;direction:ltr"> distance <h2 style=";text-align:left;direction:ltr"> (R<h2 style=";text-align:left;direction:ltr"> 2-A <h2 style=";text-align:left;direction:ltr"> ) <h2 style=";text-align:left;direction:ltr">J<h2 style=";text-align:left;direction:ltr"> 1B <h2 style=";text-align:left;direction:ltr"> =J<h2 style=";text-align:left;direction:ltr"> angle <h2 style=";text-align:left;direction:ltr"> (ATA<h2 style=";text-align:left;direction:ltr"> 1B <h2 style=";text-align:left;direction:ltr"> ,TAA<h2 style=";text-align:left;direction:ltr"> 1B <h2 style=";text-align:left;direction:ltr"> )J<h2 style=";text-align:left;direction:ltr"> distance <h2 style=";text-align:left;direction:ltr"> (R<h2 style=";text-align:left;direction:ltr"> 1-B <h2 style=";text-align:left;direction:ltr"> ) <h2 style=";text-align:left;direction:ltr">J<h2 style=";text-align:left;direction:ltr"> 2B <h2 style=";text-align:left;direction:ltr"> =J<h2 style=";text-align:left;direction:ltr"> angle <h2 style=";text-align:left;direction:ltr"> (ATA<h2 style=";text-align:left;direction:ltr"> 2B <h2 style=";text-align:left;direction:ltr"> ,TAA<h2 style=";text-align:left;direction:ltr"> 2B <h2 style=";text-align:left;direction:ltr"> )J<h2 style=";text-align:left;direction:ltr"> distance <h2 style=";text-align:left;direction:ltr"> (R<h2 style=";text-align:left;direction:ltr"> 2-B <h2 style=";text-align:left;direction:ltr"> ) J all =λ1J 1A +λ2J 2A +λ3J 1B +λ4J 2B -(λ5J punis h1 +λ6J punis h2 ) Among them, J punis h1 ,J punis h2 is the penalty function for the overlap between our two aircrafts, defined as the cosine value of the angle between our aircrafts farthest position f(position1, velocity1, velocity2) and the velocity direction of the friendly aircraft. The expression is as follows: J punis h1 =f(position1,velocity1,velocity2) J punis h2 =f(postion2,velocity2,velocity1) Taking the discrete time case as an example to determine the overall situation advantage function, the weight [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T ]The calculation process is as follows, where T is the number of time steps experienced at the current moment; According to our aircraft radar detection information, the battlefield situation is divided into 7 situations: S1. Neither of our aircraft detected the enemy aircraft; The expression of weight coefficient Λ is: Λ1=diag{1,1,1,1,1,1} In different situations, only the weight coefficient Λ changes, [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T ]The calculation process remains unchanged; S2. Only one of our aircraft detected an enemy aircraft on its radar; The expression of weight coefficient Λ is: Λ2=diag{l,s,s,l,0,1} Among them: l represents the coefficient of weight increase, which means increasing the importance of a certain indicator, l>0; s represents the coefficient of weight reduction, which means reducing the importance of a certain indicator, s<0; S3. One of our aircraft's radars detected two enemy aircraft, but our friendly aircraft did not detect any enemy aircraft; When J 2A >J 2B When , the expression of weight coefficient Λ is: Λ 31 =diag{1,1,l,s,0,0} When J 2A <J 2B When , the expression of weight coefficient Λ is: Λ 32 =diag{1,1,s,l,0,0} S4. Two of our aircraft's radars detected the same enemy aircraft; When J 1A >J 2A When , the expression of weight coefficient Λ is: Λ 41 =diag{l,s,l,s,0,0} When J 1A <J 2A When , the expression of weight coefficient Λ is: Λ 42 =diag{s,l,s,l,0,0} S5. Two of our aircraft's radars detected different enemy aircraft 1-A and 2-B; The expression of weight coefficient Λ is: Λ5=diag{l,s,l,s,0,0} S6. One of our aircraft detects two enemy aircraft on its radar, and a friendly aircraft detects one of the enemy aircraft; The expression of weight coefficient Λ is: Λ6=diag{1,l,l,s,0,0} S7. Both our aircraft detected two enemy aircraft; This situation can be divided into the following six sub-situations based on the perception situation advantage function of a single machine: S71. When J 1A >J 2A And J 2B >J 1B When , the weight coefficient Λ is expressed as: Λ 71 =diag{l,1,1,l,0,0} S72. When J 1A <J 2A And J 2B <J 1B When , the weight coefficient Λ is expressed as: Λ 72 =diag{1,l,l,1,0,0} S73. When J 1A >J 1B >J 2B >J 2A or J 2B >J 2A >J 1A >J 1B When , the expression of weight coefficient Λ is: Λ 73 =diag{l,1,1,l,0,0} S74. When J 1B >J 1A >J 2A >J 2B or J 2A >J 2B >J 1B >J 1A When , the expression of weight coefficient Λ is: Λ 74 =diag{1,l,l,1,0,0} S75. When the above conditions are not met, if J 1A -J 2A >J 1B -J 2B , the expression of weight coefficient Λ is: Λ 75 diag{l,1,1,l,0,0} S76. When the above conditions are not met, if J 1A -J 2A <J 1B -J 2B , the expression of weight coefficient Λ is: Λ 76 diag{1,l,l,1,0,0} (3) Taking the current moment as the starting point, the discretized sum of the optimization index within a prediction window is taken as the function to be optimized. Under the constraints corresponding to the current situation, the optimization problem with the flight state trajectory as the independent variable is solved; the flight trajectory within the control window with the current moment as the starting point and less than or equal to the length of the prediction window is selected as the real placeholder guidance solution; The optimization problem expression and constraints of the overall perception situation advantage function are as follows: According to the weights calculated in step (2), the overall perception situation advantage function J at the current time T is obtained: all ; P is the number of time steps in the prediction window, Keep the current weight value [λ 1,T ,λ 2,T ,λ 3,T ,λ 4,T ,λ 5,T ,λ 6,T ] all The sum of discrete time steps T+1 to T+P is used as the index function to be optimized, and the optimization problem with the flight trajectory as the independent variable is solved under the constraint of the aircraft's maneuverability. The optimization function and boundary conditions are as follows: Boundary conditions: V min ≤||[V n (T+i),V u (T+i),V e (T+i)]||2≤V max ,i∈{1,...,P} ||[V n (T+i)-V n (T+i-1),V u (T+i)-V u (T+i-1),V e (T+i)-Ve(T+i-1)]||2≤ΔV max ,i∈{1,...,P} |atan2(V n (T+i),V e (T+i))-atan2(V n (T+i-1),V e (T+i-1))|≤Δheading max ,i∈{1,...,P} |atan2(V u (T+i),||V n (T+i),V e (T+i)1|2)-atan2(V u (T+i-1),||V n (T+i-1),V e (T+i-1)||2)|≤Δpitch max ,i∈{1,...,P} pitch min ≤atan2(V u (T+i),||V n (T+i),V e (T+i)||2)≤pitch max ,i∈{1,...,P} P n (T+i)=P n (T+i-1)+ΔtV n (T+i),i∈{1,...,P} P u (T+i)=P u (T+i-1)+ΔtV u (T+i),i∈{1,...,P} P e (T+i)=P e (T+i-1)+ΔtV e (T+i),i∈{1,...,P} H min ≤P u (T+i)≤H max ,i∈{1,...,P} Where: atan2(x, y) is a function for calculating the inverse tangent of y / x (library functions in languages such as C, Python, and Matlab can be used), and the value is (-π, π]; Δt is the discrete time step; V min , V max Respectively represent the minimum and maximum speeds allowed for aircraft flight; ΔV max Represents the maximum change in speed of the aircraft within Δt time; Δheading max Represents the maximum change in the aircraft's yaw angle within Δt time; Δpitch max Represents the maximum change in the aircraft's pitch angle within Δt time; pitch min , pitch max They represent the minimum and maximum speeds allowed for an aircraft to fly; H min , H max Respectively represent the minimum and maximum altitudes allowed for an aircraft to fly; V n , V u , V e Represent the aircraft's north, sky, and east speeds, respectively. n , P u , P e Represent the north, sky, and east positions of the aircraft respectively; the maneuverability constraints of each aircraft have common characteristics, and the subscripts (1, 2, A, B) of the aircraft numbers are omitted in the expression of the constraints. In the optimization solution, the specific values of the constraints are appropriately selected according to the maneuverability of each aircraft; (4) Guiding the aircraft to fly according to the occupancy guidance plan within the control window. Repeat the above steps after the control window time to form an occupancy guidance plan generated by rolling optimization. According to the optimization result obtained in step (3), C is the number of control window time steps. The optimization results within the time step from T to T+C are used as the actual output to guide the flight trajectory; at the time T+C, the overall perception situation advantage function J is calculated based on the situation at the time T+C. all , repeat this step for T+C to T+C+P, perform a new round of optimization problem calculation, and realize the online optimization generation of rolling occupancy guidance trajectory.
Citation Information
Patent Citations
Multi-unmanned aerial vehicle cooperative air combat maneuver decision-making method based on multi-agent reinforcement learning
CN112947581A
Course learning-based unmanned aerial vehicle air game confrontation solution method
CN113282061A
Expert system optimization method based on heuristic maneuvering selection algorithm
CN113625739A
Dual-aircraft air combat decision-making method based on near-end strategy optimization
CN113741186A
Air combat maneuver decision design method based on fuzzy reasoning
CN114492805A
