CN113741500B - Unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagle - Google Patents
Unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagle Download PDFInfo
- Publication number
- CN113741500B CN113741500B CN202110995706.6A CN202110995706A CN113741500B CN 113741500 B CN113741500 B CN 113741500B CN 202110995706 A CN202110995706 A CN 202110995706A CN 113741500 B CN113741500 B CN 113741500B
- Authority
- CN
- China
- Prior art keywords
- eagle
- blue
- maneuver
- red
- party
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000005457 optimization Methods 0.000 title claims abstract description 54
- 230000006870 function Effects 0.000 claims abstract description 56
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 33
- 239000011159 matrix material Substances 0.000 claims abstract description 31
- 230000008901 benefit Effects 0.000 claims abstract description 30
- 230000007246 mechanism Effects 0.000 claims abstract description 12
- 230000008569 process Effects 0.000 claims description 25
- 230000009471 action Effects 0.000 claims description 19
- 239000013598 vector Substances 0.000 claims description 17
- 238000013461 design Methods 0.000 claims description 14
- 238000005096 rolling process Methods 0.000 claims description 13
- 230000006399 behavior Effects 0.000 claims description 8
- 238000011156 evaluation Methods 0.000 claims description 5
- 230000002349 favourable effect Effects 0.000 claims description 3
- 239000002245 particle Substances 0.000 abstract description 4
- 241000272184 Falconiformes Species 0.000 description 86
- 230000009189 diving Effects 0.000 description 13
- 238000010079 rubber tapping Methods 0.000 description 13
- 238000011161 development Methods 0.000 description 8
- 238000004364 calculation method Methods 0.000 description 5
- 230000006872 improvement Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 241000282414 Homo sapiens Species 0.000 description 1
- 102100029469 WD repeat and HMG-box DNA-binding protein 1 Human genes 0.000 description 1
- 101710097421 WD repeat and HMG-box DNA-binding protein 1 Proteins 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 230000009194 climbing Effects 0.000 description 1
- 238000013211 curve analysis Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000007711 solidification Methods 0.000 description 1
- 230000008023 solidification Effects 0.000 description 1
- 238000009966 trimming Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/08—Control of attitude, i.e. control of roll, pitch, or yaw
- G05D1/0808—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
Abstract
The invention discloses an unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of a simulated Harris eagle, which comprises the following steps: step one: building a six-degree-of-freedom aircraft model and a controller; step two: designing a tactical planning maneuver instruction generator; step three: designing a red-blue game score matrix; step four: designing a maneuver decision objective function of the mixing strategy; step five: designing an optimization algorithm simulating intelligent predation of Harris eagles; step six: and updating the six-degree-of-freedom aircraft state. The invention has the advantages that: 1) The control object is a six-degree-of-freedom nonlinear aircraft model for simulating a real aircraft, and has more practical application value compared with a three-degree-of-freedom aircraft particle model; 2) Constructing a maneuvering decision objective function by utilizing a game mixing strategy, and converting constraint conditions into unconstrained optimization problems; 3) The Harris eagle intelligent predation optimization algorithm based on the multidimensional learning mechanism is designed, so that population diversity is improved, and the situation that a local optimal solution is trapped is avoided.
Description
Technical Field
The invention relates to an unmanned aerial vehicle air combat maneuver decision-making method for simulating Harris eagle intelligent predation optimization, belonging to the field of air combat autonomous decision-making.
Background
Autonomous air combat is one of the important modes of future warfare, unmanned combat aircraft (Unmanned Combat Aerial Vehicle, UCAV) can avoid casualties and cope with severe conditions which are intolerable to human beings, is the main force in the future air combat, and the strength of the air combat capability determines the dominant right of the warfare to a great extent. The core of the unmanned fight plane air combat process is maneuver decision, and the quality of maneuver decision is directly related to the win or lose of the two parties.
Motorized decision methods fall broadly into three categories: the method is based on mathematical solution and represented by differential game, and has clear mathematical concept, but complex solution and extremely high solution difficulty facing complex problems; the method is based on machine searching, and representative methods include matrix game, monte Carlo tree searching, markov decision and the like, and the method has the most wide application and strong operability; the method is based on data training, typical methods include reinforcement learning, genetic fuzzy trees and the like, and the method can derive a plurality of unexpected results through learning of a large number of samples, so that the method is an emerging air combat decision method, but the training process is very time-consuming and still faces a plurality of challenges to be overcome.
The intelligent hawk predation simulating optimization algorithm is the mathematicization of intelligent hawk predation behavior, is an emerging optimization algorithm based on population and inspired by nature, and the inspiration of the intelligent hawk predation simulating optimization algorithm is from the cooperative hunting behavior of the hawk. The harris eagle captures the hunting object, and needs to undergo a long hunting process, which is summarized into three stages: the hunting process of the three stages is mapped into an optimization algorithm in a mathematical form, namely: the algorithm has few parameters required to be set, is easy to apply and has good optimizing capability.
In summary, the invention provides the unmanned aerial vehicle air combat maneuver decision method for simulating the Harris eagle intelligent predation optimization, which combines with a game mixing strategy to optimize and obtain the optimal maneuver, thereby improving the fight efficiency of the unmanned aerial vehicle.
Disclosure of Invention
The invention aims to provide an unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagles, which aims to solve maneuver decision-making problems in the air combat process of unmanned combat aircraft so as to improve combat effectiveness and autonomous decision-making level; the invention carries out maneuvering decision based on a matrix game method, and the improvement is that an optimized objective function is designed by utilizing a game mixing strategy, and an optimization algorithm imitating Haris eagle intelligent predation is utilized for optimizing, so that after the optimal mixing strategy is obtained, the final maneuvering is determined by utilizing a roulette manner. The maneuver decision method of the mixed strategy improves the solidification of the traditional maximum and minimum algorithms of the pure strategy, and combines the optimization algorithm, so that the optimal solution is more flexible, and the effectiveness of maneuver decision is improved.
In order to achieve the purpose, the invention provides an unmanned aerial vehicle air combat maneuver decision-making method for simulating Harris eagle intelligent predation optimization, which comprises the following specific implementation steps:
step one: six-degree-of-freedom aircraft model and controller
S11, building six-degree-of-freedom aircraft model
The invention adopts a six-degree-of-freedom nonlinear model for simulating a real airplane, instead of a particle model which is usually adopted, and the motion equation of the six-degree-of-freedom airplane comprises a dynamics equation and a kinematics equation, and can be concretely divided into displacement motion of the mass center of the airplane and rotation motion around the mass center.
S12, design of controller based on attack angle and roll angle
The input of the controller is an attack angle and roll angle instruction, and the output is four control quantities of an accelerator lever, an elevator deflection angle, an aileron deflection angle and a rudder deflection angle of the aircraft, and the feedback information is a state quantity of the aircraft.
Step two: maneuvering instruction generator for tactical planning
S21, three-degree-of-freedom maneuvering instruction generator
The maneuvering instruction generator adopts a three-degree-of-freedom simplified airplane model, tangential overload, normal overload and speed rolling angle are used as inputs of the maneuvering instruction generator, and flight speed, track dip angle and course angle are used as outputs, so that control of the movement track of the airplane can be realized. The tangential overload is mainly used for adjusting the speed of the aircraft, and the normal overload and the rolling angle are mainly used for adjusting the pitch angle and the yaw angle of the aircraft.
S22, tactical planning maneuver library
The maneuver library is designed to have normal overload and speed roll angle as control commands, through which combinations of different normal overload and speed roll angles the desired maneuver can be generated. All the action combinations in the motor action library of the red and blue parties can form a game motor matrix.
The flexibility of the motion library in design is expandability, and under the condition of meeting the performance limit of the airplane, a user can set values of normal overload and speed rolling angle in the motion library according to the needs, and the strong operability can be obtained by proper value interval.
Step three: design of red and blue game scoring matrix
S31, situation assessment function design based on direct threat
The most direct threat in the air combat is reflected in the angle relation and the distance relation of the two parties, so that two components of the air combat situation assessment function can be defined: the angle threat index and the distance threat index are shown in figure 1. The overall situation assessment function is the product of the angular threat index and the distance threat index.
S32, calculating game score matrix
The game score matrix is formed by respectively calculating situation assessment functions of the two parties under each action according to the state quantity of the two parties of each step and the game maneuvering matrix in the corresponding step two.
In the invention, the red party is the my party, the blue party is the enemy party, and the larger the expected situation assessment function value of the my party is, the more favorable the situation assessment function value is, and the blue party is opposite.
Step four: designing a hybrid strategy maneuver decision objective function
The probability of selecting each maneuver is described using a hybrid strategy, which is a vector of dimensions the number of maneuvers, the sum of which is 1. The mixing strategy of the red-recording prescription is Pr= [ Pr ] 1 ,Pr 2 ,...,Pr m ] T Blue Fang Hunge strategy is pb= [ Pb ] 1 ,Pb 2 ,...,Pb n ] T 。
The maneuver decision process is divided into two steps: the first step is to predict the mixing strategy of the blue party, and the second step is to calculate the mixing strategy of the red party according to the mixing strategy of the blue party. Both red and blue are a zero and game state, each of which tries to maximize its own benefits and minimize the benefits of the other.
S41, predicting an objective function of the blue-side mixing strategy
Assuming that the blue party selects the mixing strategy Pb, the benefits achieved by the ith maneuver of red Fang Xuanze are shown in equation (1). Then all the parties are selected flexibly and the party wants to maximize his own income as shown in formula (2).
Wherein s is ij Is the element of the ith row and the jth column in the game score matrix; benefit ri Representing the benefit obtained by the ith action of red Fang Xuanze under the blue-side mixing policy Pb; benefit rmax The maximum benefit obtained by the red party under all choices under the blue party mixing strategy Pb is represented.
From the viewpoint of blue side, benefit is desirable rmax The smaller the better, the objective function of the blue side is shown as formula (3), and the constraint condition is shown as formula (4).
Wherein Pb * The optimal mixing strategy for the blue party, i.e. the predicted blue party mixing strategy.
S42, calculating an objective function of the red party mixing strategy
Based on the predicted blue-square mixing strategy, the goal of red-squareIs to find the optimal strategy Pr * So as to maximize the benefit, wherein the objective function is shown in the formula (5), and the constraint condition is shown in the formula (6).
Wherein Pr is * And (3) determining the optimal mixing strategy for the red party, wherein S is a game score matrix.
After determining the mixing strategy of the maneuver decision, how to select the final maneuver is also a problem, and the typical method is to select the maneuver corresponding to the maximum probability in the mixing strategy, and the final maneuver is determined by adopting a roulette manner in the invention. Because roulette is essentially selected based on probabilities, it is more consistent with the uncertainty in the gaming process.
Step five: optimization algorithm for designing intelligent predation imitating Harris eagle
S51, harris eagle optimization algorithm
Harris eagle optimization is a biological population heuristic optimization algorithm, and the algorithm idea is to simulate the intelligent predation mechanism of the Harris eagle, and mainly comprises the steps of exploring hunting, assault and different attack strategies. Therefore, the harris eagle optimization algorithm is divided into three stages, namely an exploration stage, a transformation stage and a development stage. The position of each halis eagle represents a candidate solution of the optimization algorithm, and the position of the prey represents an optimal solution.
(1) The exploration phase, the harris eagle observes and monitors the surrounding environment, waiting for the appearance of hunting. There are two strategies in the exploration process, and the harris eagle randomly selects one strategy according to probability. The position update formula for harris eagle is as follows:
wherein X (t) represents the position vector of the eagle at the current moment, X (t+1) represents the position vector of the eagle at the next moment, and X prey (t) represents the position of the prey, X rand (t) represents the position of random one eagle in the current eagle group, r 1 ,r 2 ,r 3 ,r 4 P is a random number between 0 and 1, and is randomly generated in each iteration, xsurrounding min And Xsurrounding max Respectively minimum and maximum positions reachable by hawks, i.e. boundary limits to be solved, X c And (t) the center position of the current eagle group is calculated as follows:
wherein X is i (t) represents the position vector of the ith eagle at time t, and N is the total number of eagles.
(2) A transition phase, in which the eagle is transitioned between the exploration phase and the development phase, depending on the escape energy change of the prey. The hunting energy calculation formula is as follows:
wherein E represents the escape energy of the prey, E 0 The initial state of energy in each iteration process is randomly generated between-1 and-1, and T is the total iteration number.
When |E| is not less than 1, an exploration phase is executed, and when |E| <1, a development phase is executed.
(3) In the development stage, hawks can launch assaults according to the position of the explored prey, and the prey always can run out to avoid attack as far as possible, so the hawks can adopt different attack strategies according to different escape behaviors of the prey, and the attack strategies are specifically divided into four types: soft tapping, hard tapping, rapid diving soft tapping, rapid diving hard tapping.
The basis of the division among different strategies is that the escape energy E of the hunting and the probability r of success of escape are random numbers between 0 and 1, and the random numbers are updated in each iteration process. The escaping energy is used for dividing soft enclosing attack and hard enclosing attack, when the E is more than or equal to 0.5, the escaping energy of the hunting is large, the soft enclosing attack is adopted, when the E is less than 0.5, the hunting is near exhaustion, and the hard enclosing attack is adopted. The probability of successful escape is used for determining whether to take a rapid dive, when r is more than or equal to 0.5, the possibility of escaping the hunting is failed, the rapid dive is not needed, and when r is less than 0.5, the possibility of escaping the hunting is indicated, and the rapid dive is adopted. The specific procedure is as follows.
1) Soft enclosing tap
When r is more than or equal to 0.5 and |E| is more than or equal to 0.5, adopting a soft-tapping strategy, and updating the position of the hawk as follows:
X(t+1)＝X prey (t)-X(t)-E|JX prey (t)-X(t)| (10)
where J represents the random jump strength of the prey during escape, j=2 (1-r 5 )，r 5 Is a random number between 0 and 1.
2) Hard girth
When r is more than or equal to 0.5 and |E| < 0.5, adopting a soft tapping strategy, and updating the position of the hawk as follows:
X(t+1)＝X prey (t)-E|X prey (t)-X(t)| (11)
3) Rapid diving soft enclosing
When r is less than 0.5 and |E| is more than or equal to 0.5, a rapid diving soft-tapping strategy is adopted, and the process is more intelligent than the previous simple soft-tapping.
To simulate the escape pattern and frog-leaping motion of a prey, the concept of Levy Flight (LF) is cited. LF is used to simulate the zig-zag fraud of a game during the escape phase, and the irregular, abrupt and rapid dive of an eagle around the escaping game.
The position update formula of the hawk is as follows:
wherein fitness () is a fitness function, Y is a position update for a non-lewy flight, and Z is a position update for a lewy flight, and the specific calculation formula is as follows:
Y＝X prey (t)-E|JX prey (t)-X(t)| (13)
Z＝Y+S×LF(D) (14)
where D represents the dimension and S is a random vector of size 1 XD. LF () is a Lewy flight function, u and v are random numbers between 0 and 1, and β is a constant.
4) Hard slam of rapid diving
When r is less than 0 . 5, and when the I E I is less than 0.5, adopting a rapid diving hard attack strategy, wherein the position updating strategy of hawk is as follows:
Y′＝X prey (t)-E|JX prey (t)-X c (t)| (18)
Z′＝Y′+S×LF(D) (19)
s52, intelligent Harris eagle predation optimization based on multidimensional learning
Aiming at the problem that the Harris eagle optimization algorithm is easy to fall into local optimum, the invention provides an improvement thought: the learning object in the exploration stage is changed, and the intelligent predation behavior of the harris eagle is fully reflected by utilizing a multidimensional learning mechanism.
In the exploration stage, the learning object in the original algorithm is a random eagle in the eagle group, the improvement idea is that the position of the eagle is updated in each dimension by utilizing a multidimensional learning mechanism, other eagles are not learned randomly in a blind way, and the learning object is determined according to the fitness function value, so that the intelligence of the eagle is reflected, the searching efficiency is improved, the population diversity is increased, and the situation of being trapped in local optimum is avoided. The update method of the search stage is changed from the expression (7) to the expression (20).
Wherein,represents the position of the ith eagle in the d-th dimension at the t-th iteration, fi represents the numbered index of the eagle, +.>Indicating the position in the d dimension of the fi. Sup. Th eagle. For each dimension update, two eagle groups are arbitrarily selected, the fitness value is calculated, the dominant eagle is taken as a learning object, and the number of the eagle is recorded as index, so that fi=index.
S53, unmanned aerial vehicle air combat maneuver decision with intelligent haustilago-simulated eagle predation optimization
For the problems, the position vector of the hawk is the mixing strategy in the step four, the fitness function is the objective function of the step four, and for the maximization problem, the reciprocal conversion is carried out to the minimization problem. It should be noted that the objective function in the fourth step is constrained and the proposed optimization algorithm cannot be directly applied, so that the invention adopts a skill process, the constraint requirement variable is between 0 and 1, and the sum is 1, the position change of hawk is set between 0 and 1, and the sum is ensured to be 1 in a normalization mode, so that the constrained optimization problem can be solved. The unmanned aerial vehicle air combat maneuver decision flow of the intelligent predation optimization of the haris-like hawk is shown in fig. 2.
Step six: updating six degree of freedom aircraft state
After the optimal maneuver is determined, the three-degree-of-freedom control instruction is input to the six-degree-of-freedom aircraft controller, and then converted into the six-degree-of-freedom aircraft control quantity, so that the aircraft state can be updated.
The invention provides an unmanned aerial vehicle air combat maneuver decision-making method for simulating Harris eagle intelligent predation optimization, which has the main advantages that: 1) The control object is a six-degree-of-freedom nonlinear aircraft model for simulating a real aircraft, and has more practical application value compared with a three-degree-of-freedom aircraft particle model; 2) Constructing a maneuvering decision objective function by utilizing a game mixing strategy, and converting constraint conditions into unconstrained optimization problems; 3) The Harris eagle intelligent predation optimization algorithm based on the multidimensional learning mechanism is designed, so that population diversity is improved, and the situation that a local optimal solution is trapped is avoided.
Drawings
FIG. 1 is a schematic diagram of red and blue two-way situation
FIG. 2 is a flow chart of a mobile decision-making method of an unmanned aerial vehicle air combat with simulated Harris eagle intelligent predation optimization
FIG. 3a, b, partial curve analysis of the air combat process of both red and blue
FIG. 4 is a diagram showing the course of the air combat process for both red and blue parties
Detailed Description
The invention relates to an unmanned aerial vehicle air combat maneuver decision-making method for simulating Harris eagle intelligent predation optimization, which comprises the following specific implementation steps:
step one: six-degree-of-freedom aircraft model and controller
S11, building six-degree-of-freedom aircraft model
The invention uses a six-degree-of-freedom nonlinear model that simulates a real aircraft, rather than the particle model that is commonly used.
The motion equation of the six-degree-of-freedom aircraft comprises a dynamics equation and a kinematics equation, and can be specifically divided into displacement motion of the mass center of the aircraft and rotation motion around the mass center. The control quantity U of the aircraft comprises: throttle lever delta T Elevator deflection delta e Aileron yaw angle delta a Steering angle delta r The method is characterized by comprising the following steps: u (U) T ＝[δ T ,δ e ,δ a ,δ r ] T . The controlled variables of the aircraft comprise 12 state variables, which are respectively: three position quantities x g ,y g H, roll angle phi, pitch angle theta, yaw angle phi, airflow velocity V, angle of attack alpha, sideslip angle beta, roll angle velocity p, pitch angle velocity q, yaw angle velocity r, are noted: x is X T ＝[x g ,y g ,h,φ,θ,ψ,V,α,β,p,q,r] T . The following equations of motion for a six degree of freedom aircraft are given without derivation:
(1) Equation of displacement motion
Kinematic equation:
kinetic equation:
(2) Equation of rotational motion
Kinematic equation:
kinetic equation:
wherein,differentiation for the corresponding variable x; x is x g ,y g H is the three-dimensional position coordinates (x g North is positive, y g East direction is positive, h upward is positive); u, v, w are the speeds of the machine body coordinate system in the x, y and z three-axis directions respectively; v is the flying speed, alpha is the attack angle, and beta is the sideslip angle; phi is the roll angle, theta is the pitch angle, and phi is the yaw angle; p is the roll angle speed, q is the pitch angle speed, and r is the yaw angle speed; i x ,I y ,I z Moment of inertia about the aircraft body axes x, y, z, respectively, I xz Is the product of inertia; />M and N are the directions of the machine body axis x, y and zIs a torque of the engine.
On the basis of the aircraft model, structural parameters and aerodynamic parameters of the aircraft are added, so that relevant functions of the real aircraft can be simulated.
S12, design of controller based on attack angle and roll angle
The elevator channel controller is designed by feeding back an attack angle and a pitch angle rate and combining the attack angle instruction to generate an elevator control instruction; the aileron channel controller is designed by feeding back the roll angle and the roll angle rate and combining the roll angle instruction to generate an aileron control instruction; and a yaw channel controller is designed to generate a yaw control instruction by feeding back an aileron control instruction, an attack angle, a roll angle rate, a y-axis direction overload and a yaw angle rate.
Based on the control quantity obtained by aircraft trimming, the final four control quantities are obtained through an aircraft control executing mechanism by combining the generated instructions of the elevator, the aileron and the yaw channel: the control of the aircraft is realized by the throttle lever, the elevator deflection angle, the aileron deflection angle and the rudder deflection angle.
The input of the controller is an attack angle and roll angle instruction, the output is four control quantities of the aircraft, and the feedback information is some state quantities of the aircraft.
Step two: maneuvering instruction generator for tactical planning
The control instructions of the six-degree-of-freedom nonlinear aircraft model are attack angle and roll angle instructions, and the to-be-selected maneuvering instructions generated by the decision layer are converted into control layer instruction forms of the six-degree-of-freedom aircraft through calculation of a tactical planning layer. The aircraft centroid kinematics equation set can describe the flight trajectory, so that the aircraft centroid kinematics equation set can be used as a maneuvering instruction generator to simplify a decision model. And designing a maneuver library according to the control quantity of the maneuver instruction generator, so as to plan all possible states of the aircraft.
S21, three-degree-of-freedom maneuvering instruction generator
Overload refers to the ratio of the resultant of aerodynamic forces acting on the aircraft and engine thrust to the weight of the aircraft. The set of aircraft centroid kinematics equations expressed by overload is as follows:
wherein v is the flight speed; n is n x Is a tangential overload; n is n f Is normal overload; mu is the track dip angle;is a course angle; gamma is the speed roll angle; x is x g ,y g H is the three-dimensional position coordinates (x g North is positive, y g East direction is positive, h upward is positive); g is gravitational acceleration.
From the above equation, it can be seen that the tangential overload n x Normal overload n f The speed rolling angle gamma can be used as the input of a maneuvering instruction generator, the flying speed, the track dip angle and the course angle are used as the output, and the control of the movement track of the airplane can be realized. The tangential overload is mainly used for adjusting the speed of the aircraft, and the normal overload and the rolling angle are mainly used for adjusting the pitch angle and the yaw angle of the aircraft.
S22, tactical planning maneuver library
According to the design thought of the typical tactical action library, some typical tactical maneuvers such as plane flight, turning, climbing, diving, etc. can be realized, and the maneuver instruction can be converted into a control layer instruction, namely, the normal overload n is used fc And a speed roll angle gamma c And a maneuvering action library is formed, and corresponding maneuvering actions can be realized through different normal overload and speed rolling angle combinations. For a six-degree-of-freedom aircraft nonlinear model, the design based on the attack angle alpha is obtained through the control law c And roll angle command phi c According to the normal overload n already generated by the motor action library fc And a speed roll angle gamma c The control command of the six-degree-of-freedom aircraft is kept unchanged, and the position of the throttle lever of the six-degree-of-freedom aircraft is changed to gamma c As phi c The normal overload instruction is converted into an attack angle instruction and is input into the attack angle autopilot loop of the airplane, so that the maneuvering control of the six-degree-of-freedom airplane is realized.
The library of maneuvers may be expressed as:
n f ＝[n f1 ,n f2 ,...,n fu ] u (6)
γ＝[γ 1 ,γ 2 ,...,γ w ] w (7)
wherein n is f Gamma represents normal overload and speed rolling angle value vectors, u and w correspond to the dimensions of the normal overload and the speed rolling angle value vectors respectively, and the normal overload and the speed rolling angle value vectors take different values, so that different maneuvering actions can be combined. L is game motor library, and is composed of n f And gamma is correspondingly combined into a value, and u multiplied by w maneuvering actions can be generated.
The flexibility of the motion library in design is expandability, and under the condition of meeting the performance limit of the airplane, a user can set values of normal overload and speed rolling angle in the motion library according to the needs, and the strong operability can be obtained by proper value interval.
All the action combinations in the red and blue two-party maneuvering action library can form the following game maneuvering matrix:
wherein L is rm L bn Representing the mth maneuver in the red Fang Xuanqu maneuver in the maneuver base, and the nth maneuver in the blue Fang Xuanqu maneuver base.
Step three: design of red and blue game scoring matrix
S31, situation assessment function design based on direct threat
The air combat situation is the comprehensive expression of the situations of the two air combat parties, namely the red and blue parties, and the most direct threat in the air combat is reflected in the angle relation and the distance relation of the two parties, so that two component parts of the air combat situation assessment function can be defined: the angle threat index and the distance threat index are shown in figure 1. The specific definition is as follows:
angular threat index:
wherein S is A Is an angular threat index; a is that R The included angle between the speed direction of the red square aircraft and the connecting line direction of the red and blue square aircraft; a is that B Is the included angle between the speed direction of the blue aircraft and the connecting line direction of the red and blue aircraft.
Distance threat index:
S R ＝e -(R-r)/k (11)
wherein S is R Is a distance threat index; r is the distance between two machines; r is the average attack range of the red and blue cannons, and r= (r) r +r b ) 2; k is the sensitivity.
The situation assessment function is the product of the two factor indexes, and is recorded as:
S＝S A S R (12)
wherein S is a situation assessment function, S A Is an angular threat index; s is S R Is a distance threat index.
The larger the value of the above-mentioned evaluation function S, the more dominant the red party, and conversely, the smaller S, the more dominant the blue Fang Yue.
S32, calculating game score matrix
The game score matrix, namely, the state quantity of each step of red and blue two sides corresponds to the game maneuvering matrix in the step two, and the situation evaluation function of each action of the two sides is calculated respectively to form the game score matrix, wherein the game score matrix is expressed as follows:
wherein SS is a game scoring matrix; s is(s) mn The evaluation function value corresponding to the maneuver in the m-th row n column in the game matrix shown in expression (9).
In the invention, the red party is the my party, the blue party is the enemy party, and the larger the expected situation assessment function value of the my party is, the more favorable the situation assessment function value is, and the blue party is opposite.
Step four: designing a hybrid strategy maneuver decision objective function
The probability of selecting each maneuver is described using a hybrid strategy, which is a vector of dimensions the number of maneuvers, the sum of which is 1. The mixing strategy of the red-recording prescription is Pr= [ Pr ] 1 ,Pr 2 ,...,Pr m ] T Blue Fang Hunge strategy is pb= [ Pb ] 1 ,Pb 2 ,...,Pb n ] T 。
The maneuver decision process is divided into two steps: the first step is to predict the mixing strategy of the blue party, and the second step is to calculate the mixing strategy of the red party according to the mixing strategy of the blue party. Both red and blue are a zero and game state, each of which tries to maximize its own benefits and minimize the benefits of the other.
S41, predicting an objective function of the blue-side mixing strategy
Assuming that the blue party selects the mixing strategy Pb, the benefits achieved by the ith maneuver of red Fang Xuanze are shown in equation (14). Then all of the red parties are selected for maneuver and the red party wants to maximize his own profit as shown in equation (15).
Wherein s is ij Is the element of the ith row and the jth column in the game score matrix; benefit ri Representing the benefit obtained by the ith action of red Fang Xuanze under the blue-side mixing policy Pb; benefit rmax The maximum benefit obtained by the red party under all choices under the blue party mixing strategy Pb is represented.
From the viewpoint of blue side, benefit is desirable rmax The smaller the better, the objective function of the blue square is shown as formula (16), and the constraint condition is shown as formula (17)As shown.
Wherein Pb * The optimal mixing strategy for the blue party, i.e. the predicted blue party mixing strategy.
S42, calculating an objective function of the red party mixing strategy
Based on the predicted blue-side mixed strategy, the red-side aims to find the optimal strategy Pr * So as to maximize the benefit, the objective function is shown as a formula (18), and the constraint condition is shown as a formula (19).
Wherein Pr is * And (3) determining the optimal mixing strategy for the red party, wherein S is a game score matrix.
After determining the mixing strategy of the maneuver decision, how to select the final maneuver is also a problem, and the typical method is to select the maneuver corresponding to the maximum probability in the mixing strategy, and the final maneuver is determined by adopting a roulette manner in the invention. Because roulette is essentially selected based on probabilities, it is more consistent with the uncertainty in the gaming process.
Step five: optimization algorithm for designing intelligent predation imitating Harris eagle
S51, harris eagle optimization algorithm
Harris eagle optimization is a biological population heuristic optimization algorithm, and the algorithm idea is to simulate the intelligent predation mechanism of the Harris eagle, and mainly comprises the steps of exploring hunting, assault and different attack strategies. Therefore, the harris eagle optimization algorithm is divided into three stages, namely an exploration stage, a transformation stage and a development stage. The position of each halis eagle represents a candidate solution of the optimization algorithm, and the position of the prey represents an optimal solution.
(1) The exploration phase, the harris eagle observes and monitors the surrounding environment, waiting for the appearance of hunting. There are two strategies in the exploration process, and the harris eagle randomly selects one strategy according to probability. The position update formula for harris eagle is as follows:
wherein X (t) represents the position vector of the eagle at the current moment, X (t+1) represents the position vector of the eagle at the next moment, and X prey (t) represents the position of the prey, X rand (t) represents the position of random one eagle in the current eagle group, r 1 ,r 2 ,r 3 ,r 4 P is a random number between 0 and 1, and is randomly generated in each iteration, xsurrounding min And Xsurrounding max Respectively minimum and maximum positions reachable by hawks, i.e. boundary limits to be solved, X c And (t) the center position of the current eagle group is calculated as follows:
wherein X is i (t) represents the position vector of the ith eagle at time t, and N is the total number of eagles.
(2) A transition phase, in which the eagle is transitioned between the exploration phase and the development phase, depending on the escape energy change of the prey. The hunting energy calculation formula is as follows:
wherein E represents the escape energy of the prey, E 0 Representing the initial state of energy during each iteration, in-1, T is the total number of iterations.
When |E| is not less than 1, an exploration phase is executed, and when |E| <1, a development phase is executed.
(3) In the development stage, hawks can launch assaults according to the position of the explored prey, and the prey always can run out to avoid attack as far as possible, so the hawks can adopt different attack strategies according to different escape behaviors of the prey, and the attack strategies are specifically divided into four types: soft tapping, hard tapping, rapid diving soft tapping, rapid diving hard tapping.
The basis of the division among different strategies is that the escape energy E of the hunting and the probability r of success of escape are random numbers between 0 and 1, and the random numbers are updated in each iteration process. The escaping energy is used for dividing soft enclosing attack and hard enclosing attack, when the E is more than or equal to 0.5, the escaping energy of the hunting is large, the soft enclosing attack is adopted, when the E is less than 0.5, the hunting is near exhaustion, and the hard enclosing attack is adopted. The probability of successful escape is used for determining whether to take a rapid dive, when r is more than or equal to 0.5, the possibility of escaping the hunting is failed, the rapid dive is not needed, and when r is less than 0.5, the possibility of escaping the hunting is indicated, and the rapid dive is adopted. The specific procedure is as follows.
1) Soft enclosing tap
When r is greater than or equal to 0 . 5, and when the I E I is more than or equal to 0.5, adopting a soft-enclosing strategy, and updating the position of the hawk as follows:
X(t+1)＝X prey (t)-X(t)-E|JX prey (t)-X(t)| (23)
where J represents the random jump strength of the prey during escape, j=2 (1-r 5 )，r 5 Is a random number between 0 and 1.
2) Hard girth
When r is more than or equal to 0.5 and |E| < 0.5, adopting a soft tapping strategy, and updating the position of the hawk as follows:
X(t+1)＝X prey (t)-E|X prey (t)-X(t)| (24)
3) Rapid diving soft enclosing
When r is less than 0.5 and |E| is more than or equal to 0 . 5, adopting a rapid diving soft-enclosing strategy, wherein the process is simpler than the previous soft-enclosingMore intelligent.
To simulate the escape pattern and frog-leaping motion of a prey, the concept of Levy Flight (LF) is cited. LF is used to simulate the zig-zag fraud of a game during the escape phase, and the irregular, abrupt and rapid dive of an eagle around the escaping game.
The position update formula of the hawk is as follows:
wherein fitness () is a fitness function, Y is a position update for a non-lewy flight, and Z is a position update for a lewy flight, and the specific calculation formula is as follows:
Y＝X prey (t)-E|JX prey (t)-X(t)| (26)
Z＝Y+S×LF(D) (27)
where D represents the dimension and S is a random vector of size 1 XD. LF () is a Lewy flight function, u and v are random numbers between 0 and 1, and β is a constant.
4) Hard slam of rapid diving
When r is less than 0.5 and |E| < 0.5, adopting a rapid diving hard attack strategy, and updating the position of the hawk by the following strategy:
Y′＝X prey (t)-E|JX prey (t)-X c (t)| (31)
Z′＝Y′+S×LF(D) (32)
s52, intelligent Harris eagle predation optimization based on multidimensional learning
Aiming at the problem that the Harris eagle optimization algorithm is easy to fall into local optimum, the invention provides an improvement thought: the learning object in the exploration stage is changed, and the intelligent predation behavior of the harris eagle is fully reflected by utilizing a multidimensional learning mechanism.
In the exploration stage, the learning object in the original algorithm is a random eagle in the eagle group, the improvement idea is that the position of the eagle is updated in each dimension by utilizing a multidimensional learning mechanism, other eagles are not learned randomly in a blind way, and the learning object is determined according to the fitness function value, so that the intelligence of the eagle is reflected, the searching efficiency is improved, the population diversity is increased, and the situation of being trapped in local optimum is avoided. The update method of the search stage is changed from the expression (20) to the expression (33).
Wherein,represents the position of the ith eagle in the d-th dimension at the t-th iteration, fi represents the numbered index of the eagle, +.>Indicating the position in the d dimension of the fi. Sup. Th eagle. For each dimension update, two eagle groups are arbitrarily selected, the fitness value is calculated, the dominant eagle is taken as a learning object, and the number of the eagle is recorded as index, so that fi=index.
S53, unmanned aerial vehicle air combat maneuver decision with intelligent haustilago-simulated eagle predation optimization
For the problems, the position vector of the hawk is the mixing strategy in the step four, the fitness function is the objective function of the step four, and for the maximization problem, the reciprocal conversion is carried out to the minimization problem. It should be noted that the objective function in the fourth step is constrained and the proposed optimization algorithm cannot be directly applied, so that the invention adopts a skill process, the constraint requirement variable is between 0 and 1, and the sum is 1, the position change of hawk is set between 0 and 1, and the sum is ensured to be 1 in a normalization mode, so that the constrained optimization problem can be solved. The unmanned aerial vehicle air combat maneuver decision flow of the intelligent predation optimization of the haris-like hawk is shown in fig. 2.
Step six: updating six degree of freedom aircraft state
After the optimal maneuver is determined, the three-degree-of-freedom control instruction is input to the six-degree-of-freedom aircraft controller, and then converted into the six-degree-of-freedom aircraft control quantity, so that the aircraft state can be updated.
Examples:
the effectiveness of the intelligent air combat maneuver decision method for optimizing the simulated Harris eagle predation provided by the invention is verified by a specific example. In the example, two F16 aircraft models are selected as both red and blue in air combat. The simulation environment of the example is configured as an intel i9-9900K processor, 3.60Ghz master frequency, 32G memory, and software is MATLAB 2018a version.
The process block diagram of the unmanned aerial vehicle air combat maneuver decision-making method for simulating Harris eagle intelligent predation optimization is shown in fig. 2, the result diagrams of the embodiment are shown in fig. 3a, b and fig. 4, and the specific practical steps of the embodiment are as follows:
step one: initializing both red and blue settings and air combat game parameters
Red warplane initial position [0,0,3300 ]](m) a flight speed of 152m/s, an initial heading angle of 15 °; blue Fang Zhanji initial position [25,1,3.3 ]](km), a flight speed of 152m/s, an initial heading angle of 180 °. Both red and blue are 10m in wingspan, 15m in fuselage length and 4.9m in radar cross section 2 Maximum flight speed 500m/s, maximum height limit 20km, minimum height limit 500m. The shot range of the red square machine is 800m, the weight of the shot is 106g, the caliber of the shot is 20mm, the maximum found target distance is 100km, the search azimuth angle is 120 degrees, and the target found probability is 0.85; the shot range of the blue square machine is 800m, the weight of the shot is 137g, the caliber of the shot is 20mm, the maximum found target distance is 74km, the search azimuth angle is 120 degrees, and the target found probability is 0.85. The simulation time is 300s, the unit maneuvering time is 2s, and the aircraft sampling period is 10ms.
Step two: maneuvering instruction generator for tactical planning
Red and blue double-way normal overload motor warehouse [0.8,1,1.2,1.4 ]]Roll angle motor warehouse [ -45 degrees, 0, 45 degrees ]]Combined motor garageThe combined maneuvers of both red and blue are 12, and then the m×n=12×12-dimensional game matrix l_rb can be obtained.
Step three: design of red and blue game scoring matrix
Setting k=1000, and calculating corresponding evaluation function values under each action of the two parties according to the formula in the step three to obtain a m=n=12×12-dimensional game score matrix.
Step four: designing a hybrid strategy maneuver decision objective function
And (3) designing an air combat maneuver decision objective function based on the mixed strategy according to the method in the step four.
Step five: unmanned aerial vehicle air combat maneuver decision for intelligent predation optimization of haris-simulated hawk
Algorithm parameter setting: the total number of hawks n=20, the search space dimension is equal to the number of maneuvers in the motion library, and the number of iterations t=100. And (5) performing maneuver decision according to the process in the step five to obtain the optimal maneuver.
Step six: updating six degree of freedom aircraft state
And D, converting the maneuvering instruction selected in the step five into an attack angle instruction and a roll angle instruction, and inputting the attack angle instruction into the six-degree-of-freedom airplane model to realize maneuvering control.
Claims (2)
1. An unmanned aerial vehicle air combat maneuver decision-making method for simulating Harris eagle intelligent predation optimization is characterized by comprising the following steps of: the method comprises the following steps:
step one: six-degree-of-freedom aircraft model and controller
Adopting a six-degree-of-freedom nonlinear model for simulating a real aircraft;
step two: maneuvering instruction generator for tactical planning
S21, three-degree-of-freedom maneuvering instruction generator
The maneuvering instruction generator adopts a three-degree-of-freedom simplified aircraft model, tangential overload, normal overload and speed rolling angle are used as inputs of the maneuvering instruction generator, and flight speed, track dip angle and course angle are used as outputs, so that control of the aircraft motion trail is realized; the tangential overload is mainly used for adjusting the speed of the aircraft, and the normal overload and the rolling angle are mainly used for adjusting the pitch angle and the yaw angle of the aircraft;
s22, tactical planning maneuver library
The maneuver library is designed to take normal overload and speed roll angle as control instructions, and generate expected maneuver through different combinations of normal overload and speed roll angle; all the action combinations in the motor action library of the red and blue parties form a game motor matrix;
the flexibility of the motion library in design is expandability, under the condition of meeting the performance limit of the airplane, a user sets the values of normal overload and speed rolling angle in the motion library according to the needs, and the values are properly spaced, so that strong operability is obtained;
step three: design of red and blue game scoring matrix
S31, situation assessment function design based on direct threat
Two components of the air combat situation assessment function are defined: the angle threat index and the distance threat index, and the total situation assessment function is the product of the angle threat index and the distance threat index;
s32, calculating game score matrix
The game score matrix is formed by respectively calculating situation assessment functions of the two parties under each action according to the state quantity of the two parties of each step and the game maneuvering matrix in the corresponding step II;
the red party is the my party, the blue party is the enemy party, the larger the expected situation evaluation function value of the my party is, the more favorable, and the blue party is opposite;
step four: designing a hybrid strategy maneuver decision objective function
Describing the probability of selecting each maneuver by using a mixing strategy, wherein the mixing strategy is a vector with dimensions of the number of maneuver, and the sum of the vector is 1; the mixing strategy of the red prescription is Pr= [ solution ]Pr 1 ,Pr 2 ,...,Pr m ] T Blue Fang Hunge strategy is pb= [ Pb ] 1 ,Pb 2 ,...,Pb n ] T ；
The maneuver decision process is divided into two steps: the first step is to predict the mixing strategy of the blue party, and the second step is to calculate the mixing strategy of the red party according to the mixing strategy of the blue party; both red and blue are a zero and game state, each trying to maximize its own benefits and minimize the benefits of the other;
after the mixed strategy of maneuver decision is determined, the final maneuver is determined by adopting a roulette manner;
step five: optimization algorithm for designing intelligent predation imitating Harris eagle
S51, harris eagle optimization algorithm
S52, intelligent Harris eagle predation optimization based on multidimensional learning
Aiming at the problem that the Harris eagle optimization algorithm is easy to fall into local optimum, a learning object in an exploration stage is changed, and the intelligent predation behavior of the Harris eagle is fully reflected by utilizing a multidimensional learning mechanism;
s53, unmanned aerial vehicle air combat maneuver decision with intelligent haustilago-simulated eagle predation optimization
Step six: updating six degree of freedom aircraft state
After the optimal maneuver is determined, the three-degree-of-freedom control instruction is input to the six-degree-of-freedom aircraft controller, and then the six-degree-of-freedom aircraft controller is converted into a six-degree-of-freedom aircraft control quantity, namely, the aircraft state is updated;
step S52, changing learning objects in the exploration stage, and fully reflecting intelligent predation behaviors of the harris eagles by utilizing a multidimensional learning mechanism, wherein the specific process is as follows:
respectively updating the position of the hawk in each dimension by utilizing a multidimensional learning mechanism, and determining a learning object according to the fitness function value; the updating mode of the exploration stage is shown in a formula (20);
wherein the method comprises the steps of，Represents the position of the ith eagle in the d-th dimension at the t-th iteration, f i The number index representing the number of the eagle,representing the position in the d dimension of the fi-th eagle; for each dimension update, two eagle groups are arbitrarily selected, the fitness value is calculated, the dominant eagle is taken as a learning object, and the number of the eagle is recorded as index, so that fi=index.
2. The haris eagle-imitated intelligent predation optimization unmanned aerial vehicle air combat maneuver decision-making method as claimed in claim 1, wherein the method comprises the following steps of: the specific process of the fourth step is as follows:
s41, predicting an objective function of the blue-side mixing strategy
Assuming that the blue party selects the mixing strategy Pb, the benefits obtained by the ith maneuver of red Fang Xuanze are shown in equation (1); then all the red parties are flexibly selected, and the red party wants to maximize the benefit of the red party, as shown in a formula (2);
wherein s is ij Is the element of the ith row and the jth column in the game score matrix; benefit ri Representing the benefit obtained by the ith action of red Fang Xuanze under the blue-side mixing policy Pb; benefit rmax Representing the maximum benefit obtained by the red party under all choices under the condition of the blue party mixing strategy Pb;
from the viewpoint of blue side, benefit is desirable rmax The smaller the better, the objective function of the blue side is as shown in formula (3)The constraint condition is shown as a formula (4);
wherein Pb * The optimal mixing strategy of the blue party is the predicted mixing strategy of the blue party;
s42, calculating an objective function of the red party mixing strategy
Based on the predicted blue-side mixed strategy, the red-side aims to find the optimal strategy Pr * The benefit is maximized, the objective function is shown as a formula (5), and the constraint condition is shown as a formula (6);
wherein Pr is * And (3) determining the optimal mixing strategy for the red party, wherein S is a game score matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110995706.6A CN113741500B (en) | 2021-08-27 | 2021-08-27 | Unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110995706.6A CN113741500B (en) | 2021-08-27 | 2021-08-27 | Unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113741500A CN113741500A (en) | 2021-12-03 |
| CN113741500B true CN113741500B (en) | 2024-03-29 |
Family
ID=78733446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110995706.6A Active CN113741500B (en) | 2021-08-27 | 2021-08-27 | Unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113741500B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114444255B (en) * | 2021-12-13 | 2023-10-03 | 西北工业大学 | General calculation method for aircraft air combat capability based on combat process |
| CN114489120A (en) * | 2021-12-31 | 2022-05-13 | 杭州电子科技大学 | Unmanned aerial vehicle deployment and tracking control method facing mobile network |
| CN116834037B (en) * | 2023-09-01 | 2023-10-31 | 广东技术师范大学 | Dynamic multi-objective optimization-based picking mechanical arm track planning method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017068224A1 (en) * | 2015-10-23 | 2017-04-27 | Consejo Superior De Investigaciones Científicas | Biomimetic and zoosemiotic unmanned aircraft guided by automatic pilot for precision and/or pursuit flights |
| CN112783209A (en) * | 2020-12-31 | 2021-05-11 | 北京航空航天大学 | Unmanned aerial vehicle cluster confrontation control method based on pigeon intelligent competition learning |
| CN113156985A (en) * | 2021-03-18 | 2021-07-23 | 南京航空航天大学 | Obstacle avoidance robust disturbance rejection flight control method of fixed-wing unmanned aerial vehicle based on preset performance |
| CN113190037A (en) * | 2021-04-08 | 2021-07-30 | 上海吞山智能科技有限公司 | Unmanned aerial vehicle optimal path searching method based on improved fluid disturbance and sparrow algorithm |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11801937B2 (en) * | 2018-07-26 | 2023-10-31 | California Institute Of Technology | Systems and methods for avian flock flight path modification using UAVs |
-
2021
- 2021-08-27 CN CN202110995706.6A patent/CN113741500B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017068224A1 (en) * | 2015-10-23 | 2017-04-27 | Consejo Superior De Investigaciones Científicas | Biomimetic and zoosemiotic unmanned aircraft guided by automatic pilot for precision and/or pursuit flights |
| CN112783209A (en) * | 2020-12-31 | 2021-05-11 | 北京航空航天大学 | Unmanned aerial vehicle cluster confrontation control method based on pigeon intelligent competition learning |
| CN113156985A (en) * | 2021-03-18 | 2021-07-23 | 南京航空航天大学 | Obstacle avoidance robust disturbance rejection flight control method of fixed-wing unmanned aerial vehicle based on preset performance |
| CN113190037A (en) * | 2021-04-08 | 2021-07-30 | 上海吞山智能科技有限公司 | Unmanned aerial vehicle optimal path searching method based on improved fluid disturbance and sparrow algorithm |
Non-Patent Citations (2)
| Title |
|---|
| 基于直觉模糊博弈的无人机空战机动决策;李世豪;丁勇;高振龙;;系统工程与电子技术(第05期);全文 * |
| 无人机集群作战指挥决策博弈分析;王宏;李建华;;军事运筹与系统工程(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113741500A (en) | 2021-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113741500B (en) | Unmanned aerial vehicle air combat maneuver decision-making method for intelligent predation optimization of simulated Harris eagle | |
| CN113625740B (en) | Unmanned aerial vehicle air combat game method based on transfer learning pigeon swarm optimization | |
| CN107390706B (en) | Unmanned aerial vehicle near combat decision method based on rehearsal maneuver rule system | |
| CN113791634B (en) | Multi-agent reinforcement learning-based multi-machine air combat decision method | |
| Yang et al. | Evasive maneuver strategy for UCAV in beyond-visual-range air combat based on hierarchical multi-objective evolutionary algorithm | |
| CN114063644B (en) | Unmanned fighter plane air combat autonomous decision-making method based on pigeon flock reverse countermeasure learning | |
| CN113050686A (en) | Combat strategy optimization method and system based on deep reinforcement learning | |
| Li et al. | Autonomous maneuver decision-making for a UCAV in short-range aerial combat based on an MS-DDQN algorithm | |
| CN113671825A (en) | Maneuvering intelligent decision missile avoidance method based on reinforcement learning | |
| CN115688268A (en) | Aircraft near-distance air combat situation assessment adaptive weight design method | |
| CN115903865A (en) | Aircraft near-distance air combat maneuver decision implementation method | |
| CN114330115A (en) | Neural network air combat maneuver decision method based on particle swarm search | |
| CN115951709A (en) | Multi-unmanned aerial vehicle air combat strategy generation method based on TD3 | |
| CN113625569A (en) | Small unmanned aerial vehicle prevention and control hybrid decision method and system based on deep reinforcement learning and rule driving | |
| Ruan et al. | Autonomous maneuver decisions via transfer learning pigeon-inspired optimization for UCAVs in dogfight engagements | |
| Chen et al. | Design and verification of UAV maneuver decision simulation system based on deep q-learning network | |
| Wu et al. | Visual range maneuver decision of unmanned combat aerial vehicle based on fuzzy reasoning | |
| Dong et al. | Trial input method and own-aircraft state prediction in autonomous air combat | |
| Duan et al. | Autonomous maneuver decision for unmanned aerial vehicle via improved pigeon-inspired optimization | |
| Xu et al. | Autonomous decision-making for dogfights based on a tactical pursuit point approach | |
| Wang et al. | Deep reinforcement learning-based air combat maneuver decision-making: literature review, implementation tutorial and future direction | |
| CN111773722B (en) | Method for generating maneuver strategy set for avoiding fighter plane in simulation environment | |
| GOODRICH et al. | Development of a tactical guidance research and evaluation system (TGRES) | |
| CN115357051A (en) | Deformation and maneuvering integrated avoidance and defense method | |
| GOODRICH et al. | An integrated environment for tactical guidance research and evaluation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |