High-speed aircraft flight / pushing force mode conversion control optimization method based on suburb wolf intelligence

Abstract

The invention discloses a high-speed aircraft flight / pushing force mode conversion control optimization method based on suburb wolf intelligence. The method comprises the steps that 1, an environment model and a high-speed aircraft model are established; 2, designing a longitudinal control law of the high-speed aircraft; step 3, an engine mode conversion method; and step 4, performing flight / pushing force mode conversion control optimization based on suburb wolf intelligence. The invention provides a high-speed aircraft flight / thrust force mode conversion control optimization method based on suburb wolf intelligence, and aims to solve the flight / thrust force mode conversion control problem of a high-speed aircraft. A flight / thrust integrated model is constructed for a high-speed aircraft, and a control law strategy is designed for the model. Secondly, on the basis of the model of the high-speed aircraft, a dynamic mode conversion mode is constructed and adopted. And finally, a wolf pack optimization algorithm is constructed, and optimization of control parameters of the high-speed aircraft is completed aiming at the flight / thrust modal conversion control problem of the high-speed aircraft.

Description

Technical Field

[0001] The invention discloses a high-speed aircraft flight / thrust mode conversion control optimization method based on coyote intelligence, and belongs to the field of autonomous control of unmanned aerial vehicles. Background Art

[0002] High-speed aircraft is a type of aircraft that can reach a maximum flight speed of more than Mach 5, a maximum flight altitude of more than 20km, and operates in near-space. It not only has the advantages of traditional aircraft, but also has the ability to perform tasks in the aerospace field. In recent years, research on high-speed aircraft has continued to deepen, and many high-speed aircraft models have been proposed for the control of high-speed aircraft. However, the aircraft model of high-speed aircraft has stronger coupling than that of traditional aircraft, and the combined power system used brings greater challenges to control. Therefore, the design optimization method for the flight / thrust mode conversion control of high-speed aircraft has considerable research value.

[0003] The present invention proposes a high-speed aircraft flight / thrust mode conversion control optimization method based on coyote intelligence, which is used to optimize the control parameters of the high-speed aircraft flight / thrust mode conversion. Summary of the invention

[0004] The present invention proposes a high-speed aircraft flight / thrust mode conversion control optimization method based on coyote intelligence, which aims to provide a high-speed aircraft flight / thrust control method and an efficient parameter setting method, aiming to improve the flight / thrust mode conversion control effect of high-speed aircraft and enhance the stable flight capability of high-speed aircraft. First, a flight / thrust integrated model is constructed for high-speed aircraft, and a control law strategy is designed for the model. Secondly, based on the model of high-speed aircraft, a power mode conversion method is constructed. Finally, a wolf pack optimization algorithm is constructed to optimize the control parameters of the flight / thrust mode conversion control problem of high-speed aircraft.

[0005] The present invention is a high-speed aircraft flight / thrust mode conversion control optimization method based on Coyote Intelligence, and the specific implementation steps are as follows:

[0006] Step 1: Build the airplane model

[0007] The F-18 dynamics and kinematics equations are used. These nonlinear equations are based on the assumption that the aircraft is a rigid body, flying on a non-rotating flat earth, and the changes in the aircraft mass and various inertias with flight time are considered to be negligible, and the sloshing effect of the fuel in the fuel tank is also considered to be negligible. In addition, the aircraft speed in the established equations cannot be zero, and no symmetry assumptions are made for the aircraft. Considering the engine thrust factor, the default initial state engine and its thrust vector are along the x-axis of the fuselage. The derivation of the aircraft dynamics and kinematics equations uses the center of mass of the aircraft as the reference point. Based on the above assumptions, the relevant equations are given as follows:

[0008]

[0009] Where, [x, y, z] is the three-point coordinate position of the aircraft in space; [u, v, w] is the three-axis speed of the aircraft; [φ, θ, ψ] is the roll, pitch, and yaw angles of the aircraft respectively; [p, q, r] is the corresponding roll, pitch, and yaw angular velocity of the aircraft; [V, α, β] is the airspeed, angle of attack, and sideslip angle of the aircraft respectively; [I xx ,I yy ,I zz ] are the moments of inertia of the aircraft's x, y, and z axes respectively; [I xy ,I yz ,I xz ] is the product of the inertia of the corresponding three axes of the aircraft; [δ ele ,δ ail ,δ rud ,T] are the control quantities of the aircraft’s horizontal tail, aileron, vertical tail, and thrust respectively; are the average aerodynamic chord length, wingspan, wing area, mass, and dynamic pressure of the aircraft respectively; [C m ,C l ,C n ] and [C D ,C L ,C Y ] are the corresponding aerodynamic parameters of the aircraft.

[0010] Step 2: Engine mode conversion method

[0011] In the longitudinal control of high-speed aircraft, autonomous flight in the cruise phase can be decomposed into maneuvering problems such as climbing, diving, acceleration, and deceleration. In addition to the above-mentioned maneuvering requirements, autonomous flight in the low-speed startup and exit phases also requires the design of a method for mode conversion between turbine engines and ramjet engines at low Mach numbers.

[0012] The goal of the mode conversion method is to assign thrust commands to a single engine in a single mode; during the mode conversion process, the thrust changes are reasonably distributed to avoid frequent fluctuations in the engine thrust commands. The engine is divided into four working modes: turbine engine working alone, ramjet engine working alone, turbine working to ramjet working conversion, and ramjet working to turbine working conversion.

[0013] When the power mode 1 is when the turbine engine works alone, and the power mode 2 is when the ramjet engine works alone, the thrust commands of the turbine engine and the ramjet engine are respectively set to the thrust command T obtained by the speed control law. des That is, as shown below:

[0014]

[0015] In the third dynamic mode, i.e., the transition from turbine operation to ramjet operation, the ramjet engine starts from the 0 thrust state to the thrust command obtained by fully assuming speed control. The first-order inertia link is used to smooth the transition process. The thrust command is shown in the following formula:

[0016]

[0017] in, is the thrust command change rate of the ramjet engine. When entering the power mode 3, its initial value is 0; t ram is the time constant of the inertia link, and the selection principle is to ensure that the mode conversion can be completed completely and smoothly before the end of the transition process.

[0018] Power mode 4, i.e. the transition from ramjet operation to turbine operation, is similar to power mode 3. The turbine engine starts from 0 thrust state to the thrust command obtained by fully assuming speed control. The first-order inertia link is also used to smooth the transition process. The thrust command calculation formula is as follows:

[0019]

[0020] in, is the change rate of the thrust command of the turbine engine. When entering the fourth power mode, its initial value is 0; t tur is the time constant of the inertia link, and the selection principle is also to ensure that the mode conversion can be completed completely and smoothly before the end of the transition process.

[0021] The schematic diagram of mode conversion is as follows: Figure 1 shown.

[0022] Step 3: Optimization of flight / thrust mode conversion control based on Coyote Intelligence

[0023] In D-dimensional space, the number of wolves is set to N, and the position of the i-th wolf is set to X i=[X i1 ,X i2 ,X i3 ,…,X iD ], where i = 1, 2, 3, ..., N, and the wolf's position represents the independent variable of the fitness function to be solved, that is, the control parameter of the control system. The initialization formula is as follows:

[0024]

[0025] in, and Represents the upper and lower limits in the search space, and rand is a random number between 0 and 1.

[0026] The parameter optimization fitness function of the control system is obtained by modifying the conventional time-weighted integration as follows:

[0027]

[0028] After initialization, calculate the fitness values ​​of all individuals in the wolf pack, and record the best individual as the alpha wolf. Divide the wolf pack into two subgroups, and sum The individuals with the best fitness values ​​are divided into exploratory wolves, and the remaining individuals are divided into fierce wolves. sum The calculation formula is as follows:

[0029]

[0030] Among them, α is the parameter for the proportion of exploratory wolves in the total group.

[0031] In the exploration phase, the scout wolf updates its position in the search space. If the fitness value of the current scout wolf is better than the global optimum, the scout wolf is regarded as the leader wolf, and the original leader wolf becomes the scout wolf; if the fitness value of the current scout wolf is worse than the global optimum, it moves in h directions respectively, and updates the best fitness value among the fitness values ​​in the h directions as the new fitness value of the scout wolf, and updates the corresponding position to the new position. Repeat the above process until the scout wolf's fitness value is better than the global optimum or the maximum number of updates is reached. The update formula of the scout wolf's position is as follows:

[0032]

[0033] Among them, p = 1, 2, ..., h, is the pth direction explored by the wolf, l a is the exploration step length.

[0034] In the summoning phase, the alpha wolf summons all the wolves to move closer to the alpha wolf. The update formula of the wolves' positions is as follows:

[0035]

[0036] in, is the current location of the alpha wolf, l b To summon step length.

[0037] When the wolf approaches the leader, if the updated position fitness value of the wolf is better than that of the leader, the wolf becomes the leader and initiates the call; otherwise, it continues to rush towards the leader until the distance is less than the judgment distance d. near , the calculation formula is as follows:

[0038]

[0039] Among them, ω is the distance determination factor.

[0040] In the siege phase, the position of the alpha wolf is regarded as the position of the prey. The scout wolves and the fierce wolves siege the prey and the reverse direction. If the individual fitness value after the siege is better than before the siege, the position of the individual wolf is updated, otherwise it remains unchanged. The position update formula is as follows:

[0041]

[0042] Among them, l c is the siege step length, and ρ is a random number between -1 and 1.

[0043] Introduce the wolf pack update mechanism to eliminate n with poor fitness values ​​in each round of iteration low individuals and generate a corresponding number of new individuals. The specific calculation formula is as follows:.

[0044]

[0045] Among them, β is a parameter that determines the renewal ratio of the wolf pack.

[0046] The optimization steps of the high-speed aircraft flight / thrust control system are as follows. The optimization flow chart is shown in Figure 2 .

[0047] S41. Complete state balancing according to the initial flight conditions of the aircraft to obtain state quantities and control quantities of the aircraft during stable flight.

[0048] S42, initialize algorithm parameters, including: number of wolves N, exploration step length l a , Summoning step length l b , siege step length l c , distance determination factor ω, etc.

[0049] S43, initialize the specific control parameters of the system, including the height control loop parameters Speed ​​control loop parameters Angle of attack control loop parameters

[0050] S44, initialize the wolf pack position, calculate the adaptation value of the initial control parameter of the aircraft, and complete the wolf pack grouping.

[0051] S45, the wolf searches for a position in the solution space until the fitness value is optimal or the number of walks is reached.

[0052] S46. The fierce wolf moves closer to the leader wolf until the fitness value is optimal or the distance is less than the judgment distance.

[0053] S47, the scout wolves and the fierce wolves complete the siege and determine the leader; eliminate the n with poor fitness value low individuals and generate a corresponding number of new individuals.

[0054] S48. Complete the iteration and output the optimal control parameters.

[0055] Finally, the optimization of flight / propulsion mode conversion is completed. The specific results are shown in Figure 3 .

[0056] The invention provides a high-speed aircraft flight / thrust mode conversion control optimization method based on Coyote Intelligence, which has the advantages of significantly improving the thrust response speed and meeting the rapid response requirements of high-speed aircraft in different flight modes, which is particularly critical for performing rapid maneuvers or responding to emergencies in a high-dynamic environment. The optimized control method ensures smoother conversion between different thrust modes, reduces the impact and vibration caused by thrust changes, thereby improving the stability and control accuracy of the aircraft, and also improves the comfort of flight. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a schematic diagram of the mode transition of a high-speed aircraft.

[0058] Figure 2 It is a flow chart for optimizing the flight / thrust mode control of high-speed aircraft.

[0059] Figure 3 It is the comparison of parameter optimization results. DETAILED DESCRIPTION

[0060] The effectiveness of the high-speed aircraft flight / thrust integrated parameter optimization method provided by the present invention is verified by a specific example below. In this example, the initial flight conditions and initial control parameters are given, and the original pigeon grouping optimization algorithm is used for optimization. The simulation environment of this example is configured with an Intel i7-9750 processor, 2.60GHz main frequency, 8G memory, and the software is MATLAB 2019b version.

[0061] A high-speed aircraft flight / thrust mode conversion control optimization method based on coyote intelligence, the implementation process is as follows Figure 2 As shown, the specific practical steps of this example are as follows:

[0062] Step 1: Create an environment model and a high-speed aircraft model

[0063] The complete aerodynamic parameter calculation expression is as follows:

[0064]

[0065] The specific values ​​are as follows:

[0066]

[0067] Step 2: Engine mode conversion method

[0068] The goal of the mode conversion method is to assign thrust commands to a single engine in a single mode; during the mode conversion process, the thrust changes are reasonably distributed to avoid frequent fluctuations in the engine thrust commands. The engine is divided into four working modes: turbine engine working alone, ramjet engine working alone, turbine working to ramjet working conversion, and ramjet working to turbine working conversion.

[0069] The input expressions of height command and speed command are as follows:

[0070]

[0071]

[0072] The scenario is set as a high-speed aircraft starting from a low altitude and low speed, accelerating to reach the mode conversion point, and climbing after the power mode conversion, simulating the low-speed start process of a high-speed aircraft. The scenario is divided into the following stages:

[0073] (1) Stage 1: Constant speed level flight at an altitude of 10 km and a Mach number of 1, t s Take 0 seconds, t e Take 2 seconds, Δh and ΔV are both 0.

[0074] (2) Phase 2: t s Take 2 seconds, t e Take 20 seconds, Δh as 0, ΔV as 0.5 Mach, and accelerate from Mach 1 to Mach 1.5.

[0075] (3) Stage 3: t s Take 20 seconds, t e Take 30 seconds, Δh and ΔV are both 0, and the dynamic mode is transformed from mode 1 to mode 2 through mode 3.

[0076] (4) Stage 4: t s Take 30 seconds, t e Take 50 seconds, Δh as 500m, and ΔV as 0 to simulate the climbing phase in the comprehensive scenario.

[0077] (5) Stage 5: Constant speed level flight at an altitude of 10.5 km and a Mach number of 1.5, t s Take 50 seconds, t e Take 60 seconds, Δh and ΔV are both 0.

[0078] Step 3: Optimization of flight / thrust mode conversion control based on Coyote Intelligence

[0079] In D-dimensional space, the number of wolves is set to N, and the position of the i-th wolf is set to X i =[X i1 ,X i2 ,X i3 ,…,X iD ], the position of the wolf represents the independent variable of the fitness function to be solved, that is, the control parameter of the control system.

[0080] In the exploration phase, the scout wolf updates its position in the search space. If the fitness value of the current scout wolf is better than the global optimum, the scout wolf is regarded as the alpha wolf, and the alpha wolf becomes the scout wolf. If the fitness value of the current scout wolf is worse than the global optimum, it moves in h directions respectively, and updates the optimal fitness value among the fitness values ​​in the h directions as the new fitness value of the scout wolf, and updates the corresponding position to the new position. The above process is repeated until the fitness value of the scout wolf is better than the global optimum or the maximum number of updates is reached.

[0081] In the summoning phase, the alpha wolf summons all the wolves to move closer to the alpha wolf. When the wolves are moving closer to the alpha wolf, if the updated position fitness value of the alpha wolf is better than that of the alpha wolf, the alpha wolf will become the alpha wolf and initiate the summoning; otherwise, it will continue to charge towards the alpha wolf until it is less than the determined distance.

[0082] After the number of iterations reaches the maximum, the loop ends and the optimal solution is obtained, which is the optimal control parameters of the flight / propulsion integrated system.

[0083] In the siege stage, the position of the alpha wolf is regarded as the position of the prey. The scout wolves and the fierce wolves besiege the prey and the prey in reverse. If the individual fitness value after the siege is better than before the siege, the position of the individual wolf pack is updated, otherwise it remains unchanged.

[0084] Step 4: The final optimization steps for high-speed aircraft flight / thrust mode control are as follows.

[0085] S41. Complete state balancing according to the initial flight conditions of the aircraft to obtain state quantities and control quantities of the aircraft during stable flight.

[0086] S42, initialize algorithm parameters, including: number of wolves N, exploration step length l a , Summoning step length l b , siege step length l c , distance determination factor ω, etc.

[0087] S43, initialize the specific control parameters of the system, including the height control loop parameters Speed ​​control loop parameters Angle of attack control loop parameters

[0088] S44, initialize the wolf pack position, calculate the adaptation value of the initial control parameter of the aircraft, and complete the wolf pack grouping.

[0089] S45, the wolf searches for a position in the solution space until the fitness value is optimal or the number of walks is reached.

[0090] S46. The fierce wolf moves closer to the leader wolf until the fitness value is optimal or the distance is less than the judgment distance.

[0091] S47, the scout wolves and the fierce wolves complete the siege and determine the leader; eliminate the n with poor fitness value low individuals and generate a corresponding number of new individuals.

[0092] S48. Complete the iteration and output the optimal control parameters.

[0093] Finally, the optimization of high-speed aircraft flight / thrust modal control is completed.

Claims

1. A high-speed aircraft flight / propulsion mode conversion control optimization method based on coyote intelligence, characterized in that: The steps of this method are as follows: Step 1: Build the airplane model Step 2: Engine mode conversion method In the longitudinal control of high-speed aircraft, autonomous flight in the cruise phase is decomposed into the maneuvering problems of climbing, diving, accelerating, and decelerating. In addition to the above maneuvering requirements, autonomous flight in the low-speed start-up and exit phases also requires the design of a method for mode conversion between turbine engines and ramjet engines at low Mach numbers. Step 3: Optimization of flight / thrust mode conversion control based on Coyote Intelligence In D-dimensional space, the number of wolves is set to N, and the position of the i-th wolf is set to X i =[X i1 ,X i2 ,X i3 ,…,X iD ], where i = 1, 2, 3, ..., N, the wolf's position represents the independent variable of the fitness function to be solved, that is, the control parameter of the control system; the initialization formula is as follows: in, and Represents the upper and lower limits in the search space, and rand is a random number between 0 and 1; The parameter optimization fitness function of the control system is obtained by modifying the conventional time-weighted integration as follows: J=∫0 ∞ |e(t)|t 2 dt After initialization, calculate the fitness values ​​of all individuals in the wolf pack, and record the best individual as the alpha wolf; divide the wolf pack into two subgroups, and record n ​​individuals except the alpha wolf. sum The individuals with the best fitness values ​​are divided into exploratory wolves, and the remaining individuals are divided into fierce wolves. sum The calculation formula is as follows: Among them, α is the parameter of the proportion of exploratory wolves in the total group; In the exploration phase, the scout wolf updates its position in the search space. If the fitness value of the current scout wolf is better than the global optimum, the scout wolf is regarded as the leader wolf, and the original leader wolf becomes the scout wolf. If the fitness value of the current scout wolf is worse than the global optimum, it moves in h directions respectively, and updates the best fitness value among the fitness values ​​in the h directions as the new fitness value of the scout wolf, and updates the corresponding position to the new position. Repeat the above process until the scout wolf's fitness value is better than the global optimum or the maximum number of updates is reached. The update formula of the scout wolf's position is as follows: Among them, p = 1, 2, ..., h, is the pth direction explored by the wolf, l a is the exploration step length; In the summoning phase, the alpha wolf summons all the wolves to move closer to the alpha wolf. The update formula of the wolves' positions is as follows: in, is the current location of the alpha wolf, l b To summon step length; When the wolf approaches the leader, if the updated position fitness value of the wolf is better than that of the leader, the wolf becomes the leader and initiates the call; otherwise, it continues to rush towards the leader until the distance is less than the judgment distance d. near , the calculation formula is as follows: Among them, ω is the distance determination factor; In the siege stage, the position of the alpha wolf is regarded as the position of the prey. The scout wolves and the fierce wolves siege the prey and the reverse direction. If the individual fitness value after the siege is better than before the siege, the position of the individual wolf is updated, otherwise it remains unchanged; the position update formula is as follows: Among them, l c is the siege step length, ρ is a random number between -1 and 1; Introduce the wolf pack update mechanism to eliminate n with poor fitness values ​​in each round of iteration low individuals, and generate a corresponding number of new individuals. The specific calculation formula is as follows: Among them, β is a parameter that determines the renewal ratio of the wolf pack.

2. The method according to claim 1, characterized in that: The mode conversion method in step 2 divides the engine into four working modes: turbine engine working alone, ramjet engine working alone, turbine working to ramjet working conversion, and ramjet working to turbine working conversion; When the power mode 1 is when the turbine engine works alone, and the power mode 2 is when the ramjet engine works alone, the thrust commands of the turbine engine and the ramjet engine are respectively set to the thrust command T obtained by the speed control law. des That is, as shown below: In the third dynamic mode, i.e., the transition from turbine operation to ramjet operation, the ramjet engine starts from the 0 thrust state to the thrust command obtained by fully assuming speed control. The first-order inertia link is used to smooth the transition process. The thrust command is shown in the following formula: in, is the thrust command change rate of the ramjet engine. When entering the power mode 3, its initial value is 0; t ram is the time constant of the inertia link, and the selection principle is to ensure that the mode conversion can be completed completely and smoothly before the end of the transition process; Power mode 4, i.e. the transition from ramjet operation to turbine operation, is similar to power mode 3. The turbine engine starts from 0 thrust state to the thrust command obtained by fully assuming speed control. The first-order inertia link is also used to smooth the transition process. The thrust command calculation formula is as follows: in, is the change rate of the thrust command of the turbine engine. When entering the fourth power mode, its initial value is 0; t tur is the time constant of the inertia link, and the selection principle is also to ensure that the mode conversion can be completed completely and smoothly before the end of the transition process.

3. The method according to claim 2, characterized in that: The step three further comprises: Optimization steps for high-speed aircraft flight / thrust control system: S41, completing state balancing according to the initial flight conditions of the aircraft, and obtaining state quantities and control quantities of the aircraft during stable flight; S42, initialize algorithm parameters, including: number of wolves N, exploration step length l a , Summoning step length l b , siege step length l c , distance determination factor ω; S43, initialize the specific control parameters of the system, including the height control loop parameters Speed ​​control loop parameters Angle of attack control loop parameters S44, initializing the wolf pack position, calculating the adaptation value of the initial control parameter of the aircraft, and completing the wolf pack grouping; S45, the wolf explorer searches for a position in the solution space until the fitness value is optimal or the number of walks is reached; S46, the fierce wolf moves closer to the leader wolf until the fitness value is optimal or the distance is less than the judgment distance; S47, the scout wolves and the fierce wolves complete the siege and determine the leader; eliminate the n with poor fitness value low individuals, and generate a corresponding number of new individuals; S48. Complete the iteration and output the optimal control parameters.

Images