Heterogeneous aircraft cluster cooperative hunting search method for denial environment

Abstract

The application is a heterogeneous aircraft cluster cooperative hunting search method for a repellent environment. The application relates to the field of aircraft cluster control technology. The application establishes a coordination mechanism among aircraft based on a consensus initiative mechanism, so that the aircraft influence each other's behavior and realize cooperative cooperation. In the repellent environment, indirect communication between the aircraft is realized through reading and writing of pheromone in the medium by the aircraft. In the distributed search algorithm, the pheromone map is fused and optimized with the particle swarm algorithm to obtain an optimized search and hunting result. The application solves the problems of limited communication and complex cooperation in the cooperative search and hunting of the aircraft cluster in the repellent environment. Through fusion and optimization of environmental perception and cooperative control, the search efficiency and hunting success rate are improved, and an effective distributed cooperative perception and decision mechanism is formed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft cluster control, and is a heterogeneous aircraft cluster cooperative hunting search method for denial environment. BACKGROUND

[0002] In the field of aircraft cluster cooperative control technology, traditional methods mainly adopt centralized control architecture, which requires all information to be centralized to the center node for processing and decision-making. Although this method can provide high control accuracy, it has problems such as communication bottleneck, heavy computing burden, and system vulnerability in practical application. Although the distributed control method can improve the robustness and flexibility of the system, how to realize effective cooperation and task allocation optimization among aircraft clusters is still a key problem.

[0003] In the field of communication technology under denial environment, due to the influence of electromagnetic interference, communication blockage and other factors, the aircraft cluster faces challenges such as limited communication bandwidth, unstable link and large communication delay. Existing communication schemes often fail to meet the real-time and reliability requirements under denial environment, especially in large-scale aircraft cluster cooperative operation scenarios, how to ensure the timely transmission and effective sharing of key information becomes a problem to be solved.

[0004] In the field of aircraft cluster cooperative search target technology, existing researches mainly focus on search algorithm design under ideal environment, and lack of consideration of uncertainty and dynamics under denial environment. Traditional search methods are difficult to effectively deal with situations such as strong target maneuverability, incomplete environmental information and limited perception ability. In addition, how to realize cooperative search decision of aircraft cluster under limited communication conditions, and how to balance search efficiency and system robustness are the focus and difficulty of current research.

[0005] Therefore, it is urgent to propose a heterogeneous aircraft cluster cooperative hunting search method that can effectively deal with the above challenges to improve the operation ability and task efficiency of the system under denial environment. SUMMARY

[0006] The present application provides a heterogeneous aircraft cluster cooperative hunting search method for denial environment, which optimizes the fusion of environmental perception and cooperative control, improves the search efficiency and hunting success rate, and forms an effective distributed cooperative perception and decision-making mechanism.

[0007] The present application provides the following technical solutions:

[0008] The heterogeneous aircraft cluster cooperative hunting search method for denial environment comprises the following steps:

[0009] Step 1: initialize the heterogeneous aircraft swarm and the position and velocity of the target;

[0010] Step 2: enter the first stage: the heterogeneous aircraft swarm randomly moves and explores the search area, and an initial pheromone map is established by using the pheromone map construction algorithm;

[0011] Step 3: when the upper limit of the iteration steps in the first stage is reached, the second stage is started: the heterogeneous aircraft swarm updates the motion state by using the target search algorithm, and a pheromone map is established by using the pheromone map construction algorithm; when the upper limit of the iteration steps in the second stage is reached, the second stage is ended.

[0012] Preferably, the step 2 is specifically:

[0013] The target signal value detected by the aircraft is represented by a three-dimensional function, that is:

[0014] (1)

[0015] In the formula, the function is a continuous function related to the position of the aircraft, signal representing the target signal strength that can be detected by the aircraft at the coordinates (x, y, z);

[0016] The blue aircraft moves from point A to point B along the blue dashed line and leaves a pheromone vector at point B, which is represented by the following formula:

[0017] (2)

[0018] In the formula, and represent the target signal values detected by the blue aircraft at points A and B, and represent the coordinates of points A and B, representing the distance between points A and B;

[0019] The unit vector is represented by , that is, in formula (2), and formula (2) is rewritten as:

[0020] (3)

[0021] The directional derivative of the target signal field in the direction of the unit vector is represented by , and according to the definition of the directional derivative, the following formula is obtained:

[0022] (4)

[0023] In the formula,p =||Δ x ||2,Δ x is the difference of coordinates; when the distance between A and B is close, then:

[0024] (5)

[0025] Rewrite formula (4) as:

[0026] (6)

[0027] is the directional derivative of the target signal field at point B, that is:

[0028] (7)

[0029] From the relationship between the directional derivative and the gradient, the directional derivative is the projection of the gradient in that direction, that is:

[0030] (8)

[0031] In the formula, represents the gradient of the signal field at point B. Combine the above formulas to get:

[0032] (9)

[0033] Green aircraft along the green dotted line from C to B point, leaving the vector pheromone at B point Write in the same form; one aircraft is moving along the signal gradient direction to B point, then the vector pheromone it leaves at B point is equal to the signal gradient at that point; when multiple aircraft leave their own vector pheromone at B, then the vector pheromone will be accumulated at B point, and the accumulated vector pheromone is represented as , the and in formula (9) are expressed in the Cartesian coordinate system as:

[0034] (10)

[0035] (11)

[0036] In the formula, and represent the unit vectors of the horizontal and vertical axes, respectively, and the vector pheromone generated by the aircraft at position B is represented as:

[0037] (12)

[0038]

[0039] Considering that pheromones generated from different directions need to be accumulated, the result is expressed as:

[0040] (13)

[0041]

[0042] When the aircraft passes through the same position from multiple directions and generates corresponding vector pheromones, the accumulation is obtained to obtain new vector pheromones. The accumulated generated vector pheromones should theoretically approach the gradient direction of the position, and the size is positively correlated with the gradient.

[0043] Preferably, in formula (2), is regarded as the growth rate of signal strength along the direction of aircraft movement, representing the length of vector ; and is a unit vector along the direction of aircraft movement, representing the direction of vector ; the two together constitute vector , which is regarded as the experience obtained by the aircraft in the last movement process.

[0044] Preferably, the update formula of pheromone in the cumulative update mechanism is as follows:

[0045] (14)

[0046] In the formula, represents the vector pheromone obtained by the n th update operation on the m th pheromone carrier, indicates the vector pheromone released by the i th aircraft in the pheromone carrier in this update; parameters and represent the pheromone dissipation coefficient and accumulation coefficient respectively.

[0047] Preferably, the accumulation formula of pheromone in the replacement update mechanism is as follows:

[0048] (15)

[0049] Preferably, the update formula of pheromone in the replacement update mechanism is as follows:

[0050] (16)

[0051] In the formula, It is an amplification factor. When the newly generated pheromone is significantly better than the existing pheromone at the current position, the new pheromone directly replaces the existing pheromone. When the newly generated pheromone is worse than the existing pheromone at the current position, the two are added together.

[0052] Preferably, step 3 specifically comprises:

[0053] During the search process, the spacecraft continuously detects the target signal strength and reads the pheromone carrier to obtain vector pheromones. In the basic particle swarm algorithm, particles... The velocity and position update formula is:

[0054] (17)

[0055] (18)

[0056] In the formula, and They represent the first i The particle in the first k Velocity and position in the next iteration and These represent the particle's historical best position and global best position, respectively. w It is the coefficient of inertia. and These represent self-learning factors and social learning factors, respectively. and They are two independent random numbers, Δ t It is the time step;

[0057] The particle swarm optimization algorithm employs linearly adjusted inertia weights, which in the initial stage of the algorithm... w Taking the maximum value allows the particles to search a wider space, making it less likely to miss the global optimum. This is beneficial in the later stages of the algorithm. w To accelerate the convergence speed of the algorithm, the minimum value is selected. The adjustment method is as follows:

[0058] (19)

[0059] in, and These are the initial and final values ​​of the inertia weight, respectively. t and T These are the current iteration count and the maximum iteration count, respectively.

[0060] In the early stages of the algorithm, the self-learning factor is maximized, allowing the entire population to fully search for the optimal solution within the largest possible space. In the later stages of the algorithm, the social learning factor is maximized to accelerate the convergence speed. The adjustment method is as follows:

[0061] (20)

[0062] (21)

[0063] where, and are the maximum and minimum values of the self-learning factor set before starting, and are the maximum and minimum values of the social learning factor set;

[0064] When searching for a static target, the global variable The aircraft system is required to have very good global communication capability, and the vector pheromone is used to replace the global optimal position. The speed and position update formula of the aircraft is:

[0065] (22)

[0066] (23)

[0067] where, and respectively represent the speed and position of the i th aircraft in the k th iteration, represents the historical optimal position of the aircraft i itself, and the parameter represents the vector pheromone read by the i th aircraft from the n th pheromone carrier;

[0068] For a dynamic environment, the speed and position update formula of the aircraft is modified as:

[0069] (24)

[0070] (25)

[0071] where, represents the vector pheromone generated by the i th aircraft in the k th iteration step, and the search experience of the previous step of the aircraft is still utilized to accelerate the search speed of the aircraft in the dynamic environment.

[0072] A heterogeneous aircraft cluster cooperative search system for a denial environment, the system comprising:

[0073] An initialization module, the initialization module initializes the heterogeneous aircraft cluster and the position and speed of the target;

[0074] The first stage module is a heterogeneous aircraft cluster random motion, an exploration search area, and an initial pheromone map is established by using a pheromone map construction algorithm;

[0075] The second stage module is a heterogeneous aircraft cluster using a target search algorithm to update the motion state, and a pheromone map is established by using a pheromone map construction algorithm when the upper limit of the first stage iteration step is reached.

[0076] A computer readable storage medium, which stores a computer program, the program is executed by a processor to realize a heterogeneous aircraft cluster cooperative search method for a denial environment.

[0077] A computer device, comprising a memory and a processor, the memory stores a computer program, and the processor realizes a heterogeneous aircraft cluster cooperative search method for a denial environment when executing the computer program.

[0078] The present application has the following beneficial effects:

[0079] The present application improves the search efficiency and the success rate of the aircraft cluster, and forms an effective distributed cooperative perception and decision mechanism.

[0080] The present application constructs a pheromone map, combines vector pheromone and particle swarm algorithm to update the motion state of the aircraft, so as to realize indirect communication and cooperative control of the aircraft cluster in the denial environment; the present application can meet the requirements of intelligent and networked aircraft cluster combat and monitoring in the future. BRIEF DESCRIPTION OF DRAWINGS

[0081] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0082] Figure 1 The flow chart of the heterogeneous aircraft cluster cooperative search method for a denial environment of the present application is shown;

[0083] Figure 2 The schematic diagram of the target signal function of the present application is shown;

[0084] Figure 3 The schematic diagram of the vector pheromone generation process of the present application is shown. DETAILED DESCRIPTION

[0085] The technical solutions of the present application will be described clearly and completely in connection with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, but not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0086] The present application will be described in detail below in connection with specific embodiments. Embodiment one:

[0088] According to Figures 1 to 3 The specific optimization technical solution adopted by the present application to solve the above technical problems is: the present application relates to a heterogeneous aircraft cluster cooperative hunting search method for a repellent environment.

[0089] The present application provides a heterogeneous aircraft cluster cooperative hunting search method for a repellent environment, which comprises the following steps:

[0090] Step 1: initialize the heterogeneous aircraft cluster and the position and speed of the target;

[0091] Step 2: enter the first stage: the heterogeneous aircraft cluster randomly moves to explore the search area, and an initial pheromone map is established by using a pheromone map construction algorithm;

[0092] Step 3: when the upper limit of the iteration steps of the first stage is reached, the second stage begins: the heterogeneous aircraft cluster updates the motion state by using a target search algorithm, and a pheromone map is established by using a pheromone map construction algorithm; when the upper limit of the iteration steps of the second stage is reached, the second stage ends.

[0093] Based on the consensus initiative mechanism, the present application establishes a coordination mechanism between each aircraft, so that the aircrafts influence each other's behavior and realize cooperative cooperation. In a repellent environment, indirect communication between aircrafts is realized through reading and writing of pheromones in the medium by the aircrafts. Under the distributed search algorithm, the pheromone map is fused and optimized with the particle swarm algorithm to obtain an optimized search and hunting result. The present application improves the search efficiency and hunting success rate by fusing and optimizing the environment perception and cooperative control, and forms an effective distributed cooperative perception and decision-making mechanism, which solves the problems of limited communication and complex cooperation in the cooperative search and hunting of aircraft cluster in a repellent environment. Embodiment two:

[0095] The difference between the second embodiment of the present application and the first embodiment is only that:

[0096] The step 2 is specifically:

[0097] The target signal value detected by the aircraft is represented by a three-dimensional function, that is:

[0098] (1)

[0099] where the function is a continuous function related to the position of the aircraft, signal represents the target signal strength that the aircraft can detect at the coordinates (x, y, z);

[0100] The blue aircraft moves from point A to point B along the blue dashed line and leaves a pheromone vector at point B, which is expressed by the following equation:

[0101] (2)

[0102] where and represent the target signal values detected by the blue aircraft at points A and B, and represent the coordinates of points A and B, represents the distance between points A and B;

[0103] Let represent the unit vector, i.e. in equation (2), equation (2) is rewritten as:

[0104] (3)

[0105] The directional derivative of the target signal field in the direction of the unit vector is represented by , and according to the definition of the directional derivative, we have:

[0106] (4)

[0107] where p =||Δ x ||2, Δ x is the difference in coordinates; when the distance between points A and B is close, we have:

[0108] (5)

[0109] Rewrite equation (4) as:

[0110] (6)

[0111] is the directional derivative of the target signal field at point B, i.e.:

[0112] (7)

[0113] From the relationship between directional derivative and gradient, the directional derivative is the projection of gradient in the direction, that is:

[0114] (8)

[0115] In the formula, represents the gradient of signal field at point B. Combining the above formulas, we get:

[0116] (9)

[0117] Green aircraft along the green dotted line from C to B point, leaving the vector pheromone at B point In the same form; a aircraft is moving along the signal gradient direction to B point, then it left the vector pheromone at B point is equal to the signal gradient at the point; when multiple aircrafts left their own vector pheromone at B, then the vector pheromone will be accumulated at B point, the accumulated vector pheromone is expressed as , the and in formula (9) are expressed in Cartesian coordinate system as:

[0118] (10)

[0119] (11)

[0120] In the formula, and respectively represent the unit vectors of horizontal axis and vertical axis, and the vector pheromone generated by the aircraft at position B is expressed as:

[0121] (12)

[0122]

[0123] Considering that the pheromones generated by aircrafts from different directions need to be accumulated, the result is expressed as:

[0124] (13)

[0125]

[0126] When the aircrafts pass through the same position from multiple directions and generate corresponding vector pheromones, the accumulation is obtained to obtain new vector pheromones. The accumulated vector pheromones generated in theory should approach the gradient direction of the position, and the size is positively correlated with the gradient. Specific embodiment three:

[0128] The difference between the third embodiment and the second embodiment of the present application is only that:

[0129] In formula (2), is regarded as the growth rate of signal strength along the direction of aircraft movement, representing the length of vector ; while is a unit vector along the direction of aircraft movement, representing the direction of vector ; the two together constitute vector , which is regarded as the experience obtained by the aircraft in the last movement. The fourth embodiment is as follows:

[0131] The difference between the fourth embodiment and the third embodiment of the present application is only that:

[0132] The update formula of pheromone in the cumulative update mechanism is as follows:

[0133] (14)

[0134] In formula (14), represents the vector pheromone obtained by the n th update operation on the pheromone carrier stored in the m th pheromone carrier, and represents the vector pheromone released by the i th aircraft in the pheromone carrier in this update; parameters and represent the pheromone dissipation coefficient and the accumulation coefficient, respectively. The fifth embodiment is as follows:

[0136] The difference between the fifth embodiment and the fourth embodiment of the present application is only that:

[0137] The accumulation formula of pheromone in the replacement update mechanism is as follows:

[0138] (15) The sixth embodiment is as follows:

[0140] The difference between the sixth embodiment and the fifth embodiment of the present application is only that:

[0141] The update formula of pheromone is as follows:

[0142] (16)

[0143] In formula (16), is a magnification factor, when the newly generated pheromone is obviously better than the existing pheromone at the current position, the new pheromone directly replaces the existing pheromone, and when the newly generated pheromone is not better than the existing pheromone at the current position, the two are accumulated. Specific embodiment seven:

[0145] The difference between the embodiment seven and the embodiment six is only:

[0146] The step 3 is specifically:

[0147] In the search process, the aircraft continuously detects target signal strength and reads pheromone carriers to obtain vector pheromones, in the basic particle swarm algorithm, the speed position update formula of the particle is:

[0148] (17)

[0149] (18)

[0150] In the formula, and respectively represent the speed and position of the first particle in the first iteration, i and k respectively represent the historical optimal position and the global optimal position of the particle, is an inertia coefficient, and w respectively represent a self-learning factor and a social learning factor, and are two independent random numbers, and Δ is a time step; t The particle swarm algorithm with linear adjustment of inertia weight makes take the maximum value in the initial stage of the algorithm, so that the particle searches a wider space and thus is not easy to miss the global optimal solution, and makes

[0151] take the minimum value in the later stage of the algorithm, so as to accelerate the convergence speed of the algorithm, and the adjustment mode is: w w

[0152] (19)

[0153] In the formula, and are respectively an initial value and a final value of the inertia weight, t and T are respectively a current iteration number and a maximum iteration number;

[0154] ​​​In the initial stage of the algorithm, the self-learning factor is maximized, and the entire population searches for the optimal solution in the largest space. In the later stage of the algorithm, the social learning factor is maximized to accelerate the convergence speed of the algorithm. The adjustment mode is:

[0155] (20)

[0156] (21)

[0157] In the formula, and are the maximum and minimum values of the self-learning factor set in advance before the start, and are the maximum and minimum values of the social learning factor set;

[0158] When searching for a static target, the global variable in formula (17) is required to have very good global communication capability, and the vector pheromone is used to replace the global optimal position. The speed and position update formula of the aircraft is:

[0159] (22)

[0160] (23)

[0161] In the formula, and respectively represent the speed and position of the i th aircraft in the k th iteration, represents the historical optimal position of the aircraft i itself, and the parameter represents the vector pheromone read by the i th aircraft from the n th pheromone carrier;

[0162] For a dynamic environment, the speed and position update formula of the aircraft is modified as:

[0163] (24)

[0164] (25)

[0165] In the formula, represents the vector pheromone generated by the i th aircraft in the k th iteration step. The search experience of the aircraft in the previous step is still utilized to accelerate the search speed of the aircraft in the dynamic environment. Specific embodiment eight:

[0167] The embodiment eight of the present application is only different from the embodiment seven in that:

[0168] The present application provides a heterogeneous aircraft cluster cooperative hunting search system for a denial environment, the system comprising:

[0169] An initialization module, which initializes the heterogeneous aircraft cluster and the position and speed of the target;

[0170] A first stage module, which randomly moves the heterogeneous aircraft cluster, explores the search area, and establishes an initial pheromone map by using a pheromone map construction algorithm;

[0171] A second stage module, which starts the second stage when the upper limit of the first stage iteration step is reached: the heterogeneous aircraft cluster updates the motion state by using a target search algorithm, and establishes a pheromone map by using a pheromone map construction algorithm; and ends the second stage when the upper limit of the second stage iteration step is reached. Specific embodiment nine:

[0173] The embodiment nine of the present application is only different from the embodiment eight in that:

[0174] The present application provides a computer readable storage medium, which stores a computer program, the program being executed by a processor to implement a heterogeneous aircraft cluster cooperative hunting search method for a denial environment. Specific embodiment ten:

[0176] The embodiment ten of the present application is only different from the embodiment nine in that:

[0177] The present application provides a computer device, which comprises a memory and a processor, the memory storing a computer program, and the processor implementing a heterogeneous aircraft cluster cooperative hunting search method for a denial environment when executing the computer program. Specific embodiment eleven:

[0179] The embodiment eleven of the present application is only different from the embodiment ten in that:

[0180] A "target-extreme value" model is established for the aircraft cluster target search problem, that is, the motion space of the aircraft is regarded as the independent variable space of a function, the target to be searched is regarded as the extreme value point of the function, the function value corresponding to each position in the space is related to the target signal distribution, the target signal is detected by the aircraft, the function value is calculated, and then the target search problem is converted into the problem of finding the extreme value point of the function. Figure 2 A typical target signal function is shown in the following formula.

[0181] Figure 2A two-dimensional aircraft motion space is shown in Fig. 1. In reality, the target signal intensity distribution is three-dimensional, so the target signal value detected by the aircraft can be represented by a three-dimensional function, i.e.,

[0182] (1)

[0183] where the function is a continuous function related to the aircraft position, signal represents the target signal intensity that the aircraft can detect at the coordinates (x, y, z). For an aircraft moving in a two-dimensional space, z = 0 can be set.

[0184] The local precision search adopts a heuristic search strategy. In the search process, the aircraft can only detect the target signal value and the pheromone in the environment. In each iteration step, the speed of the aircraft V is updated according to the following three sub-items: (1) the aircraft's inheritance of the speed of the previous step, specifically the speed of the previous step multiplied by an inertia coefficient; (2) the aircraft's learning of individual historical information, specifically obtaining a speed from the current position to the individual historical optimal position (i.e., the position that the aircraft has reached so far with the largest target signal value, which is continuously updated as the aircraft reaches new positions), which is multiplied by a self-learning factor for adjustment; (3) the aircraft's learning of pheromone, specifically obtaining a speed consistent with the vector pheromone direction read from the environment before each iteration, which is multiplied by a social learning factor for adjustment. Through this speed updating strategy, the aircraft constantly moves towards the target, and eventually realizes the convergence of as many aircraft as possible at the target. Since the aircraft swarm can only detect the target signal at different positions and the pheromone in the environment during the search process, the aircraft itself cannot determine whether it has "found" the target. Therefore, a group of aircraft must converge to a certain position to determine that the aircraft has found the target, thereby achieving precision search of the target.

[0185] The successful application of the consensus initiative mechanism to the target search task requires the aircraft to release its search experience in the form of pheromone into the environment for other aircraft to read, and then to obtain system-level coordination. In this process, it is important to design an appropriate pheromone model.

[0186] Inspired by the concept of vector, pheromone is modeled in the form of vector, named vector pheromone. The typical vector pheromone is shown in Fig. 2. The color of the fan area in the figure represents the target signal value detected by the aircraft. The darker the color, the larger the target signal value detected by the aircraft. Figure 3

[0187] ​Assume that the blue vehicle moves from point A to point B along the blue dashed line and leaves a pheromone vector at point B , which is expressed by the following formula:

[0188] (2)

[0189] In the formula, and represent the target signal values detected by the blue vehicle at points A and B, and represent the coordinates of points A and B, represents the distance between points A and B. In formula (2), can be regarded as the growth rate of the signal strength along the direction of the vehicle's movement, representing the length of the vector . And is a unit vector along the direction of the vehicle's movement, representing the direction of the vector . Together, they form the vector , which can be regarded as the experience obtained by the vehicle during the previous movement.

[0190] Let represent the unit vector, i.e. in formula (2), so formula (2) can also be rewritten as:

[0191] (3)

[0192] The directional derivative of the target signal field along the unit vector direction can be represented by , according to the definition of the directional derivative, we can get:

[0193] (4)

[0194] In the formula, p =||Δ x ||2, Δ x is the difference of coordinates. If points A and B are close enough, then:

[0195] (5)

[0196] Therefore, formula (4) can be rewritten as:

[0197] (6)

[0198] In practice, there will always be some distance between points A and B, so can be approximately considered as the directional derivative of the target signal field at point B, i.e.:

[0199] (7)

[0200] From the relationship between directional derivative and gradient, the directional derivative is the projection of the gradient in the direction, i.e.

[0201] (8)

[0202] where represents the gradient of the signal field at point B, as indicated by the arrow in Figure 3 . Combining the above several formulas, we can get

[0203] (9)

[0204] Similarly, the green aircraft moves along the green dashed line from point C to point B, and the vector pheromone left by the green aircraft at point B can also be written in a similar form. If an aircraft moves to point B along the direction of the signal gradient, then the vector pheromone left by the aircraft at point B is approximately equal to the gradient of the signal at that point. When multiple aircraft leave their own vector pheromones at B, the vector pheromones will be accumulated at point B. This accumulated vector pheromone is denoted as Figure 3 in . The and in formula (9) can be expressed in Cartesian coordinates as:

[0205] (10)

[0206] (11)

[0207] where and are the unit vectors of the horizontal and vertical axes, respectively. Therefore, the general form of the vector pheromone generated by the aircraft at position B can be represented as:

[0208] (12)

[0209]

[0210] Considering that the pheromones generated by the aircraft from different directions need to be accumulated, the result can be represented as:

[0211] Therefore, when the aircraft passes through the same position from multiple directions and generates corresponding vector pheromones, they can be accumulated to obtain new vector pheromones, as shown in Figure 3 the vector The accumulated pheromone vector information should theoretically approach the gradient direction of the position, and the size is positively correlated with the gradient.

[0212] (13)

[0213]

[0214] Step 3:

[0215] The accumulation and dissipation of pheromones reflect the dynamic changes of the environment. However, unlike biological pheromones such as odors and hormones, data information stored in electronic devices does not automatically accumulate or dissipate over time unless human-designed rules are designed. Therefore, in order to reflect the dynamic changes of the environment, the accumulation and dissipation of pheromones are carried out in the read-write operations of the aircraft. This method proposes three pheromone update mechanisms, namely cumulative, replacement, and comprehensive.

[0216] The update formula of pheromone in the cumulative update mechanism is as follows:

[0217] (14)

[0218] In the formula, represents the vector pheromone obtained by the n th update operation on the m th pheromone carrier, represents the vector pheromone released by the i th aircraft in the pheromone carrier in this update. Parameters and represent the pheromone dissipation coefficient and accumulation coefficient, respectively. The advantage of this update method is that it is sensitive to the dynamic changes of the environment, and the disadvantage is that the pheromone accumulation speed is slow. In order to reflect the dynamic changes of the environment or obtain a certain pheromone at a certain position, a considerable amount of time is required.

[0219] The accumulation formula of pheromone in the replacement update mechanism is as follows:

[0220] (15)

[0221] In the formula, and have the same meaning as in formula (14). This pheromone update method is easy to implement, and its advantages are fast pheromone accumulation speed, and the disadvantage is that it cannot accurately reflect the dynamic changes of the environment. It is suitable for static environments, or dynamic environments with slow changes and small changes.

[0222] The advantages and disadvantages of the two pheromone update mechanisms mentioned above are obvious, and they are complementary. Therefore, a comprehensive pheromone update mechanism is proposed, combining the two mechanisms, and its update formula is as follows:

[0223] (16)

[0224] In the formula, This is an amplification factor, usually set to 1. This means that when a newly generated pheromone is significantly better than the existing pheromone at the current location, the new pheromone directly replaces the existing one; when the newly generated pheromone is inferior to the existing pheromone at the current location, the two are added together. This comprehensive pheromone update mechanism not only improves the pheromone update speed but also effectively reflects dynamic changes in the environment.

[0225] During the search, the spacecraft continuously detects the target signal strength and reads the pheromone carrier to obtain vector pheromones. The spacecraft's velocity update mechanism is inspired by the basic particle swarm optimization algorithm. In the basic particle swarm optimization algorithm, particles... The velocity and position update formula is:

[0226] (17)

[0227] (18)

[0228] In the formula, and They represent the first i The particle in the first k Velocity and position in the next iteration and These represent the particle's historical best position and global best position, respectively. w It is the coefficient of inertia. and These represent self-learning factors and social learning factors, respectively. and They are two independent random numbers, Δ t It is the time step.

[0229] In particle swarm optimization, inertia weights w It has an impact on the performance of the algorithm, when w When the velocity is large, the proportion of the original velocity in each component of the particle's velocity is relatively large. The particle is less attracted by its individual historical best position and global best position, thus it can explore a wider space, but the overall convergence speed of the particle swarm is slower. wWhen the particle size is small, the particle is easily attracted by the individual history optimal position and the global optimal position, and converges quickly, but it is easily deceived by the local optimal position, which makes the algorithm unable to find the global optimal solution. The particle swarm optimization algorithm with linear adjustment of inertia weight makes w the inertia weight take a larger value at the beginning of the algorithm, so that the particle searches a wider space and is not easy to miss the global optimal solution, and makes the inertia weight take a smaller value at the later stage of the algorithm, so as to accelerate the convergence speed of the algorithm. The adjustment method is: w

[0230] (19)

[0231] wherein, and are the initial value and the final value of the inertia weight respectively, t and T are the current iteration number and the maximum iteration number respectively.

[0232] The values of the learning factors and have a great influence on the performance of the particle swarm optimization algorithm. When the self-learning factor is relatively large, the whole population will excessively wander in the search space, and it is difficult to converge; on the contrary, the population will converge quickly, and the algorithm is easily trapped in the local optimal solution. Therefore, at the beginning of the algorithm, the self-learning factor is made larger, so that the whole population can fully search the optimal solution in a larger space, and at the later stage of the algorithm, the social learning factor is made larger, so as to accelerate the convergence speed of the algorithm. The adjustment method is:

[0233] (20)

[0234] (21)

[0235] wherein, and are the maximum value and the minimum value of the self-learning factor set before, and are the maximum value and the minimum value of the social learning factor set.

[0236] When searching for a static target, the global variable in formula (17) requires that the aircraft system has very good global communication capability, but such a requirement is difficult to guarantee when the system scale is expanded. The vector pheromone is used to replace the global optimal position, and the speed and position update formula of the aircraft is:

[0237] (22)

[0238] (23)

[0239] In the formula,​ and denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, i k denote the historical optimal position of the i-th aerial vehicle, which is obtained from the search experience of the aerial vehicle itself. The parameter i denote the vector pheromone read by the i-th aerial vehicle from the j-th pheromone carrier i n

[0240] When searching for a dynamic target, the historical optimal position of the aerial vehicle becomes unreliable in a dynamic environment because it is determined according to the target signal value detected by the aerial vehicle, and even has a negative impact on the search task, resulting in the aerial vehicle sometimes failing to find the target. Therefore, the self-learning parameter can be set to 0. In fact, although the historical optimal position of the aerial vehicle is unreliable, the search experience of the aerial vehicle in the recent period of time can still be utilized, and for a dynamic environment, the aerial vehicle's velocity and position update formula can be modified as follows:

[0241] (24)

[0242] (25)

[0243] In the formula, φij denotes the vector pheromone generated by the i-th aerial vehicle in the j-th iteration step. In this way, the search experience of the aerial vehicle in the previous step is still utilized, which can accelerate the search speed of the aerial vehicle in a dynamic environment. i k

[0244] A typical search process can be divided into two stages: an initial mapping stage and a search and tracking stage. In the initial mapping stage, the swarm aerial vehicles move randomly, explore as wide an area as possible, and use the pheromone map construction algorithm to construct an initial pheromone map to prepare for subsequent accurate search and tracking of the target. After reaching the set upper limit of the stage iteration step, the search and tracking stage is entered. The search refers to the process of the swarm aerial vehicles gathering near the target from an arbitrary state, emphasizing the process of discovering the target; while the tracking refers to the process of the swarm aerial vehicles being able to gather near the target at all times and moving together with the target. In this stage, the aerial vehicles have the ability to update the velocity, and the aerial vehicles move under the guidance of the target search algorithm. After reaching the set motion time, the algorithm ends. The pheromone map is continuously updated using the pheromone map construction algorithm in the entire search process. DETAILED DESCRIPTION TWELVE

[0246] ​​​​​​​​A heterogeneous aircraft swarm cooperative hunting search method for a denial environment, the method comprising the following steps:

[0247] Step S1: initializing a heterogeneous aircraft swarm and a position and speed of a target.

[0248] Step S2: a first stage, the heterogeneous aircraft swarm randomly moves, explores a search area as much as possible, and establishes an initial pheromone map by using a pheromone map construction algorithm;

[0249] The step S2 specifically comprises the following steps:

[0250] Step S2.1: the aircraft moves according to an initialized speed size and direction;

[0251] Step S2.2: the aircraft detects a target signal strength of a grid where the aircraft is located during the movement;

[0252] Step S2.3: the aircraft obtains vector pheromone of the grid where the aircraft is located at the moment by using the pheromone map construction algorithm;

[0253] Step S2.4: the aircraft updates the vector pheromone of the grid where the aircraft is located according to an updating mechanism of the vector pheromone;

[0254] Step S2.5: the aircraft changes a movement state of the aircraft by using an obstacle avoidance mechanism to avoid collision with other aircraft or obstacles.

[0255] Step S3: when reaching an upper limit of iteration steps of the first stage, a second stage is started, the heterogeneous aircraft swarm updates a movement state by using a target search algorithm, and a pheromone map is established by using the pheromone map construction algorithm. When reaching an upper limit of iteration steps of the second stage, the program ends.

[0256] The step S3 specifically comprises the following steps:

[0257] Step S3.1: the aircraft reads vector pheromone of a grid where the aircraft is located;

[0258] Step S3.2: the aircraft updates the movement state of the aircraft by using the target search algorithm;

[0259] Step S3.3: the aircraft detects a target signal strength of a grid where the aircraft is located during the movement;

[0260] Step S3.4: the aircraft obtains vector pheromone of the grid where the aircraft is located at the moment by using the pheromone map construction algorithm;

[0261] Step S3.5: the aircraft updates the vector pheromone of the grid where the aircraft is located according to an updating mechanism of the vector pheromone;

[0262] Step S3.6: The aircraft adjusts the motion state by means of the obstacle avoidance algorithm to prevent collision with other aircraft or obstacles.

[0263] The application improves the search efficiency and the success rate of the aircraft cluster to the target;

[0264] The application forms an effective distributed cooperative sensing and decision mechanism;

[0265] The application constructs the pheromone map, combines the vector pheromone and the particle swarm algorithm to update the motion state of the aircraft, so as to realize the indirect communication and cooperative control of the aircraft cluster in the hostile environment;

[0266] The application can meet the requirements of the intelligent and networked aircraft cluster combat and monitoring in the future.

[0267] The above is only the preferred embodiment of the heterogeneous aircraft cluster cooperative search method facing the hostile environment, and the protection scope of the heterogeneous aircraft cluster cooperative search method facing the hostile environment is not limited to the above-mentioned embodiment, and any technical solution belonging to the idea shall belong to the protection scope of the application. It should be pointed out that, for those skilled in the art, some improvements and changes without departing from the principle of the application shall be considered as the protection scope of the application.

Claims

1. A method for cooperative pursuit search of a heterogeneous swarm of aircraft in a denial environment, characterized by: The method comprises the following steps: Step 1: initialize the heterogeneous aircraft cluster and the position and speed of the target; Step 2: enter the first stage: the heterogeneous aircraft cluster randomly moves, explores the search area, and establishes an initial pheromone map by using a pheromone map construction algorithm; The step 2 is specifically: The signal value detected by the aircraft is represented by a three-dimensional function, that is: （1） where the function is a continuous function related to the position of the aircraft, Signal represents the intensity of the target signal that can be detected by the aircraft at the coordinates (x, y, z); The blue aircraft moves from point A to point B along the blue dashed line and leaves a pheromone vector at point B is represented by the following formula: （2） wherein and denotes the target signal value detected by the blue aircraft at point A and point B, and denotes the coordinates of point A and point B, denotes the distance between point A and point B; with denotes the unit vector, i.e. , formula (2) is rewritten as: （3） The target signal field is along the unit vector The directional derivative of the direction is denoted by According to the definition of the directional derivative, we have （4） In the formula, Rewrite formula (4) as: =||Δ x ||2, Δ x is the difference of coordinates; when the distance between A and B points is close, then （5） From the relationship between the directional derivative and the gradient, the directional derivative is the projection of the gradient in the direction, that is: （6） is the directional derivative of the target signal field at point B, i.e.: （7） Considering that the pheromones generated by the aircraft from different directions need to be accumulated, the result is represented as: （8） In the formula, The gradient of the representative signal field at point B is obtained by combining the above formulas. （9） Green aircraft moves along the green dotted line from C to B, leaving vector pheromone at B In the same form; one aircraft is moving along the signal gradient direction to B, then it leaves the vector pheromone at B equal to the signal gradient at the point; when multiple aircraft leave their own vector pheromone at B, then the vector pheromone will be accumulated at B, and the accumulated vector pheromone is expressed as , the and in formula (9) are expressed in the Cartesian coordinate system as: （10） （11） where and denote the unit vectors of the horizontal and vertical axes, respectively, and the vector pheromone generated by the aerial vehicle at position B is denoted by （12） When the aircraft passes through the same position from multiple directions and generates corresponding vector pheromones, the accumulation is obtained to obtain new vector pheromones. The accumulated generated vector pheromones should theoretically tend to approach the gradient direction of the position, and the size is positively correlated with the gradient; （13） Step 3: when the upper limit of the iteration step of the first stage is reached, the second stage is started: the heterogeneous aircraft cluster updates the motion state by using a target search algorithm, and establishes a pheromone map by using a pheromone map construction algorithm; when the upper limit of the iteration step of the second stage is reached, the second stage is ended; The step 3 is specifically: ω During the search process, the aircraft constantly detects the target signal strength and reads the pheromone carrier to obtain vector pheromones. In the basic particle swarm algorithm, the position and velocity of the particles are updated according to the following formula: The position and velocity updating formula of the particles is: （17） （18） wherein, and denote the velocity and position of the i th particle in the k th iteration, respectively, and denote the historical best position and the global best position of the particle, respectively, ω is the inertia coefficient, and denote the self-learning factor and the social learning factor, respectively, and are two independent random numbers, Δ t is the time step. The particle swarm optimization algorithm employs linearly adjusted inertia weights, which in the initial stage of the algorithm... ω Taking the maximum value allows the particles to search a wider space, making it less likely to miss the global optimum. This is beneficial in the later stages of the algorithm. In the early stage of the algorithm, the self-learning factor is maximum, and the whole population fully searches for the optimal solution in the maximum space. In the later stage of the algorithm, the social learning factor is maximum, and the convergence speed of the algorithm is accelerated. The adjustment mode is: To accelerate the convergence speed of the algorithm, the minimum value is selected. The adjustment method is as follows: （19） wherein, and are an initial value and a final value of the inertia weight, respectively, t and T are a current iteration number and a maximum iteration number, respectively; For a dynamic environment, the speed and position update formula of the aircraft is modified as: （20） （21） wherein and are the maximum and minimum values of the self-learning factor set at the beginning, and are the maximum and minimum values of the social learning factor set; When searching for static targets, the global variable The aircraft system is required to have very good global communication ability, using vector pheromone to replace the global optimal position, and the speed position update formula of the aircraft is: （22） （23） wherein and denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, i denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, k denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, i denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, i denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, n denote the velocity and position of the i-th aerial vehicle in the j-th iteration, respectively, 2. The method according to claim 1, characterized in that: （24） （25） In the formula, Indicates the first i The aircraft in the k The vector pheromones generated in each iteration step allow the spacecraft to utilize its previous search experience, thus accelerating the spacecraft's search speed in dynamic environments.

3. The method according to claim 2, characterized in that: In equation (2), is considered the amplitude of the increase of the signal strength along the direction of the aircraft motion, representing the vector ; while is a unit vector along the direction of the aircraft motion, representing the direction of the vector ; both together form the vector , which is considered the experience acquired by the aircraft during the previous motion. The update formula of the pheromone in the cumulative update mechanism is as follows:

4. The method according to claim 3, characterized in that: （14） wherein represents the vector pheromone obtained after the n th update operation on the pheromone carrier of the m th information vector, represents the vector pheromone released by the i th aircraft in the pheromone carrier during the present update; the parameters and represent the pheromone dissipation coefficient and the accumulation coefficient, respectively. The accumulation formula of the pheromone in the replacement update mechanism is as follows:

5. The method according to claim 4, characterized in that: （15）。 The update formula of the pheromone in the information update mechanism is as follows: ​ （16） wherein is an amplification factor, when the newly generated pheromone is significantly better than the existing pheromone at the current location, the new pheromone directly replaces the existing pheromone, and when the newly generated pheromone is not as good as the existing pheromone at the current location, the two are added together.

Images