Unmanned aerial vehicle longitudinal stability optimization method based on catastrophe mechanism genetic algorithm

Abstract

The invention relates to an unmanned aerial vehicle longitudinal stability optimization method based on a catastrophe mechanism genetic algorithm, and belongs to the technical field of unmanned aerial vehicle optimization design, and the method comprises the steps: firstly, carrying out the description and modeling of an unmanned aerial vehicle longitudinal stability optimization problem, considering a static stability margin, a pitching moment and a lift margin, and selecting a proper unmanned aerial vehicle layout optimization variable; optimal unmanned aerial vehicle layout parameters are obtained through optimization based on a catastrophe mechanism genetic algorithm, so that the longitudinal stability of the unmanned aerial vehicle is improved, and the longitudinal stability of the unmanned aerial vehicle in a complex flight environment can be remarkably improved; according to the method, the optimization result is reliable, the algorithm is easy and convenient to implement, and the method can be widely applied to the urgent requirements of various unmanned aerial vehicles for longitudinal stability improvement in different flight stages and complex working conditions in unmanned aerial vehicle optimization design engineering application.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) optimization design technology, specifically to a method for optimizing the longitudinal stability of UAVs based on a catastrophic mechanism genetic algorithm. Background Technology

[0002] With the rapid development of drone technology, the application scenarios of drones continue to expand, covering many fields such as military reconnaissance, logistics delivery, agricultural plant protection, and emergency rescue. However, regardless of the application scenario, the stability of drones is a crucial performance indicator, with longitudinal stability being a key element in ensuring the safe and efficient flight of drones.

[0003] From the perspective of drone flight principles, longitudinal motion involves the aircraft's pitch motion, including nose-up and nose-down movements, as well as changes in vertical speed. When drones encounter airflow disturbances, need to adjust their attitude, or perform specific tasks, their longitudinal stability faces significant challenges. For example, when performing mapping tasks in complex terrain such as mountainous areas, drones are highly susceptible to unstable airflow between valleys, which can cause drastic fluctuations in their pitch angle. Insufficient longitudinal stability can lead to dangerous situations such as stalling, diving, or even crashing, severely impacting mission execution and drone safety.

[0004] Traditionally, to optimize the longitudinal stability of unmanned aerial vehicles (UAVs), researchers have primarily focused on improving the aircraft's aerodynamic design, such as adjusting wing sweep angles and installing suitable horizontal stabilizers. However, this approach is limited by the aircraft's physical structure, and once the design is finalized, there is little room for further adjustments. Therefore, optimizing the longitudinal layout of UAVs is necessary to improve longitudinal stability.

[0005] With the rise of intelligent algorithms, some advanced optimization algorithms have been gradually introduced into the field of UAV longitudinal stability optimization. Genetic algorithms, as optimization algorithms that simulate natural selection and genetic mechanisms, have shown potential in UAV longitudinal stability optimization due to their powerful global search capabilities. They gradually search for the optimal control parameters by performing selection, crossover, and mutation operations on "chromosomes" representing different combinations of control parameters, thereby improving the longitudinal stability of the UAV. However, traditional genetic algorithms also have some problems in practical applications, such as being prone to getting trapped in local optima and struggling to accurately find the global optimum in complex UAV longitudinal stability optimization problems. Summary of the Invention

[0006] In view of the above problems, this invention provides a method for optimizing the longitudinal stability of unmanned aerial vehicles (UAVs) based on a catastrophic genetic algorithm. This invention first describes and models the longitudinal stability optimization problem of UAVs, considering static stability margin, pitch moment, and lift margin, selecting appropriate UAV layout optimization variables, and then optimizing the optimal UAV layout parameters based on a catastrophic genetic algorithm. This improves the longitudinal stability of the UAV, and significantly enhances its longitudinal stability in complex flight environments. The optimization results of this invention are reliable, and the algorithm is easy to implement. It can be widely applied in UAV optimization design engineering to meet the urgent need for improved longitudinal stability of various UAVs in different flight stages and complex operating conditions.

[0007] This invention provides a method for optimizing the longitudinal stability of unmanned aerial vehicles (UAVs) based on a catastrophic mechanism genetic algorithm, comprising:

[0008] Step 1: Obtain the optimization variables for the longitudinal stability of the UAV and set the corresponding domain;

[0009] Step 2: Based on the optimization variables and corresponding domains of the UAV's longitudinal stability, obtain the lift margin of the UAV;

[0010] Step 3: Based on the lift margin of the UAV and the constraints of the lift margin, establish the cost function for optimizing the longitudinal stability of the UAV.

[0011] By using the cost function of UAV longitudinal stability optimization and the catastrophe mechanism genetic algorithm, multiple optimization variables of UAV longitudinal stability are optimized, and multiple optimal variables are obtained.

[0012] Optionally, the lift margin of the UAV is expressed as:

[0013]

[0014] Where ΔL is the lift margin of the UAV, m is the mass of the UAV, g is the gravitational acceleration, q is the dynamic pressure, S is the projected area of ​​the wing, and C L η is the lift coefficient. ht It is the ratio of tail fin dynamic pressure to free-flow pressure, S ht It is the wetted area of ​​the horizontal stabilizer. It is the lift coefficient C of the horizontal tail. Lht The derivative of α with respect to the angle of attack α ht It is the angle of attack of the horizontal stabilizer during stable flight.

[0015] Optionally, step 3 includes the following specific steps:

[0016] Step S31: Determine multiple individuals for the i-th iteration step, each individual including multiple optimization variables x1, x2, x3…x ni = 1, 2, 3…I, where I represents the total number of iterations;

[0017] Step S32: Based on the lift margin of the UAV and the constraints of the lift margin, construct the cost function for optimizing the longitudinal stability of the UAV, and obtain the fitness value of each individual in the i-th iteration step.

[0018] Step S33: Based on the fitness value of each individual in the i-th iteration step, perform parent selection and update for multiple individuals in the i-th iteration step to obtain multiple updated individuals in the i-th iteration.

[0019] Step S34: Perform a crossover operation on the multiple updated individuals in the i-th iteration to obtain multiple individuals after the crossover in the i-th iteration;

[0020] Step S35: Perform mutation operation on the multiple individuals after the i-th iteration crossover to obtain multiple individuals after the i-th iteration mutation;

[0021] Step S36: Obtain the fitness value of each individual after the mutation in the i-th iteration;

[0022] Determine whether the distribution of fitness values ​​of multiple individuals after the i-th iteration of mutation meets the catastrophe triggering criterion. If yes, it indicates that a catastrophe has been triggered, and proceed to the next step; otherwise, proceed to step S38.

[0023] Step S37: Eliminate some individuals from the mutated individuals of the i-th iteration and perform a second update to obtain multiple second-updated individuals of the i-th iteration;

[0024] Perform a second mutation operation on multiple individuals that have undergone secondary updates in the i-th iteration to obtain multiple individuals that have undergone tertiary updates in the i-th iteration;

[0025] Step S38: Determine if i ≥ I. If not, treat the multiple individuals updated three times in the i-th iteration as the multiple individuals in the (i+1)-th iteration, let i = i+1, and return to step S32. If yes, then treat the optimization variables corresponding to the multiple individuals updated three times in the i-th iteration as the optimal variables.

[0026] Optionally, the cost function for optimizing the longitudinal stability of the UAV is expressed as:

[0027] min J=κ1J sm +κ2J pm

[0028] Where J is the cost of optimizing the longitudinal stability of the UAV, k1 is the first weight coefficient, k2 is the second weight coefficient, and J sm It is the static stability margin, J pm It is the pitching torque.

[0029] Optionally, the disaster triggering criterion is expressed as:

[0030] J avg ≥ΞJ max

[0031] Among them, J avg J is the average fitness value of the current iteration step. max Ξ represents the maximum fitness value in the current iteration, and Ξ is the density factor characterizing the density of individuals.

[0032] Optionally, step S37, which obtains multiple individuals of the second-order update in the i-th iteration, includes the following specific steps:

[0033] A fitness threshold is preset for each individual, and individuals that exceed the fitness threshold in the i-th iteration are filtered out.

[0034] Select the top a% of individuals in the i-th iteration as the eliminated individuals in the i-th iteration, and consider them to have survived the catastrophe; the rest of the individuals are eliminated.

[0035] The missing individuals of 1-a% are replaced by replicating the individuals eliminated in the i-th iteration, resulting in multiple second-updated individuals in the i-th iteration.

[0036] Optionally, the static stability margin is expressed as:

[0037]

[0038] Among them, J sm For static stability margin; x cg It is the center of gravity, c w It is the mean aerodynamic chord of the wing. It is the lift coefficient C L The derivative with respect to the angle of attack α, It is the lift coefficient C of the horizontal stabilizer. Lht The derivative with respect to the angle of attack α, The normalized position of the wing's aerodynamic center. It is the normalized position of the aerodynamic center of the horizontal tail fin, η ht S is the ratio of tail dynamic pressure to free-flow pressure, and S is the derivative of the tail lift coefficient with respect to the angle of attack. ht It is the wetted area of ​​the horizontal tail fin, S w It is the wetted area of ​​the wing.

[0039] Optionally, the pitching moment is expressed as:

[0040]

[0041] Among them, J pm Let C be the pitching moment, q be the dynamic pressure, and C be the dynamic force.m It is the pitching moment coefficient, S is the wing's projected area, and α is the pitching moment coefficient. ht It is the angle of attack of the horizontal stabilizer during stable flight. ht ε is the tail lever arm length, ε is the wing downwash angle, and α is the angle of attack.

[0042] In this invention, the catastrophe mechanism simulates natural catastrophic phenomena. During the genetic algorithm's search process, when the algorithm gets stuck in a local optimum, a catastrophe operation is introduced to perform large-scale mutations or reinitialization of the population, breaking the current local optimum state and enabling the algorithm to escape the local optimum and perform a new global search. This mechanism effectively avoids the tendency of traditional genetic algorithms to get stuck in local optima. For complex UAV longitudinal stability optimization problems, it is expected to find better combinations of control parameters through more efficient global search, thereby significantly improving the longitudinal stability of UAVs and meeting the increasingly stringent requirements of UAV flight performance in increasingly complex application scenarios.

[0043] Compared with the prior art, the present invention has at least the following beneficial effects:

[0044] (1) The UAV longitudinal stability optimization method based on the catastrophe mechanism genetic algorithm of the present invention avoids getting trapped in local optima by introducing a catastrophe mechanism, and achieves efficient search for the optimal parameter combination to improve the longitudinal stability of the UAV, while preventing the search process from becoming a simple or completely random search.

[0045] (2) The UAV longitudinal stability optimization method based on the catastrophic mechanism genetic algorithm of the present invention does not require repeated large-scale parameter debugging for different environments, saving the preparation time in the early stage of optimization and having high execution efficiency.

[0046] (3) The longitudinal stability optimization method of UAV based on catastrophic mechanism genetic algorithm of the present invention can effectively improve flight stability and safety through catastrophic mechanism genetic algorithm, is easy to implement in engineering, and has broad engineering application prospects. Attached Figure Description

[0047] The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of the invention.

[0048] Figure 1 This is a schematic diagram of a flowchart illustrating a method for optimizing the longitudinal stability of unmanned aerial vehicles based on a catastrophic mechanism genetic algorithm, according to an embodiment of the present invention.

[0049] Figure 2 This is a schematic diagram showing the change curve of the static stability margin during the optimization process in an embodiment of the present invention and its comparison with the standard genetic algorithm;

[0050] Figure 3This is a schematic diagram showing the change curve of pitch moment during the optimization process in an embodiment of the present invention and its comparison with the standard genetic algorithm;

[0051] Figure 4 This is a schematic diagram showing the change curve of lift margin during the optimization process in an embodiment of the present invention and its comparison with the standard genetic algorithm. Detailed Implementation

[0052] To better understand the above-described objectives, features, and advantages of the present invention, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other. Furthermore, the present invention can be implemented in other ways different from those described herein; therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0053] A specific embodiment of the present invention, such as Figure 1-4 A method for optimizing the longitudinal stability of unmanned aerial vehicles (UAVs) based on a catastrophic mechanism genetic algorithm is disclosed. The specific implementation steps are as follows:

[0054] Step 1: Obtain the optimization variables for the longitudinal stability of the UAV and set the corresponding domain;

[0055] It is understood that the longitudinal stability of the UAV is the pitch stability of the UAV about the horizontal axis;

[0056] Optionally, the optimization variables are the wing projected area S and the center of gravity position x. cg Tail lever arm length l ht and horizontal tail fin mounting angle I ht ;

[0057] Step 2: Based on the optimization variables and corresponding domain of the UAV's longitudinal stability, obtain the UAV's lift margin ΔL, expressed as:

[0058]

[0059] Where ΔL is the lift margin of the UAV, m is the mass of the UAV, g is the gravitational acceleration, q is the dynamic pressure, S is the projected area of ​​the wing, and C L η is the lift coefficient. ht It is the ratio of tail fin dynamic pressure to free-flow pressure, S ht It is the wetted area of ​​the horizontal stabilizer. It is the lift coefficient C of the horizontal tail. Lht The derivative of α with respect to the angle of attack α ht It is the angle of attack of the horizontal stabilizer during stable flight.

[0060] Set constraints on lift margin to ensure stable flight of the UAV;

[0061] Optionally, the expression for the lift margin constraint is:

[0062] ΔL≥0

[0063] Step 3: Based on the lift margin of the UAV and the constraints of the lift margin, establish the cost function for optimizing the longitudinal stability of the UAV.

[0064] By using the cost function of UAV longitudinal stability optimization and the catastrophe mechanism genetic algorithm, multiple optimization variables of UAV longitudinal stability are optimized, and multiple optimal variables are obtained.

[0065] Optionally, step 3 includes the following specific steps:

[0066] Step S31: Determine multiple individuals for the i-th iteration step, each individual including multiple optimization variables x1, x2, x3…x n i = 1, 2, 3…I, where I represents the total number of iterations;

[0067] Optionally, each individual includes multiple optimization variables x1, x2, x3…x n Each individual [x1,x2,x3…x] n [] is a real number vector, corresponding to a possible solution to the optimization problem;

[0068] The catastrophic mechanism genetic algorithm proposed in this invention encodes individuals using real numbers, and the optimization problem is performed in the solution space of the real number field.

[0069] Step S32: Based on the lift margin of the UAV and the constraints of the lift margin, construct the cost function for optimizing the longitudinal stability of the UAV.

[0070] Based on the cost function for optimizing the longitudinal stability of the UAV, the fitness value of each individual in the i-th iteration step is obtained;

[0071] It is understandable that the fitness value reflects the degree of excellence or poorness of an individual in the corresponding optimization problem;

[0072] Optionally, the cost function for optimizing the longitudinal stability of the UAV is expressed as:

[0073] minJ=κ1J sm +κ2J pm

[0074] Where J is the cost of optimizing the longitudinal stability of the UAV, κ1 is the first weight coefficient, κ2 is the second weight coefficient, and J sm It is the static stability margin, J pm It is the pitching torque.

[0075] Optionally, the static stability margin is expressed as:

[0076]

[0077] Among them, J sm Static stability margin, representing the degree of stability of the UAV, is defined as the dimensionless quantity based on the wing chord length, representing the distance between the center of gravity and the focal point; x cg It is the center of gravity, c w It is the mean aerodynamic chord of the wing. It is the lift coefficient C L The derivative with respect to the angle of attack α, It is the lift coefficient C of the horizontal tail. Lht The derivative with respect to the angle of attack α, The normalized position of the wing's aerodynamic center. It is the normalized position of the aerodynamic center of the horizontal tail fin, η ht S is the ratio of tail dynamic pressure to free-flow pressure, and S is the derivative of the tail lift coefficient with respect to the angle of attack. ht It is the wetted area of ​​the horizontal tail fin, S w It is the wetted area of ​​the wing.

[0078] The pitching moment is expressed as follows:

[0079]

[0080] Among them, J pm Let q be the pitching moment, and C be the dynamic pressure. m It is the pitching moment coefficient, S is the wing's projected area, and α is the pitching moment coefficient. ht It is the angle of attack of the horizontal stabilizer during stable flight. ht It is the length of the tail lever arm. It is the derivative of the wing downwash angle with respect to the angle of attack, where ε is the wing downwash angle and α is the angle of attack.

[0081] Furthermore, the expression for the derivative of the wing downwash angle with respect to the angle of attack is:

[0082]

[0083] Where AR is the aspect ratio and π is the value of a circle.

[0084] Step S33: Based on the fitness value of each individual in the i-th iteration step, select multiple individuals in the i-th iteration step to obtain multiple parent individuals in the i-th iteration; obtain multiple updated individuals in the i-th iteration based on the multiple parent individuals in the i-th iteration.

[0085] Optionally, the specific steps for obtaining the updated individual in the i-th iteration as described in step S33 include:

[0086] Obtain the total fitness A of multiple individuals in the i-th iteration step. i ;

[0087] Obtain the ratio of each individual's fitness to the total fitness A in the i-th iteration step. i The probability of being selected is represented by the probability of being chosen.

[0088] In the i-th iteration step, after sorting each individual by fitness value, the corresponding cumulative probability is obtained based on the selection probability of each individual, which represents the area it occupies on the roulette wheel;

[0089] Based on total fitness A i Set the random number selection range B i Within the random number selection range B i Select the p-th element i A random number;

[0090] Iterate through the cumulative probability of each individual in the i-th iteration step, and find the first individual whose cumulative probability is greater than or equal to that of the p-th individual. i An individual with a random number of , as the p-th... i Select individuals and record them; the number of random numbers is the same as the number of individuals; if this condition is not met, let p i =p i +1;

[0091] Traversing P i Filter using random numbers to obtain the filtered O i Individuals, of which O i Less than or equal to P i ;

[0092] Determine O i Is it equal to P? i If so, the filtered O i If a group of individuals is used as the updated population in the i-th iteration step, then sort the selected individuals in ascending order based on their fitness and select the top P individuals with the highest fitness. i -O i Individuals on the screened M i Each individual is supplemented to obtain P. i An updated individual is used as the updated population for the i-th iteration step.

[0093] Optionally, the method for obtaining the updated individual in the i-th iteration described in step S33 also includes the tournament selection method.

[0094] Step S34: Perform a crossover operation on the updated individuals in the i-th iteration to obtain multiple individuals after the i-th iteration crossover;

[0095] Step S35: Perform mutation operation on the multiple individuals after the i-th iteration crossover to obtain multiple individuals after the i-th iteration mutation;

[0096] Optionally, based on the fitness value of each individual, the top 20% of the best individuals with the highest fitness are selected directly as parents, and the number of missing individuals is generated by regeneration and replication of the selected parents.

[0097] It is understandable that crossover simulates the gene exchange process in biological evolution, producing offspring by exchanging some genes from parent individuals.

[0098] Optionally, the crossing operation can be a single-point crossing or a two-point crossing.

[0099] It is understandable that mutation operations simulate the gene mutation process in biological evolution, randomly changing certain genes of an individual with a certain probability; performing mutation operations on newly generated individuals can introduce new genetic information and increase the diversity of the population.

[0100] Step S36: Obtain the fitness value of each individual after the mutation in the i-th iteration;

[0101] Determine whether the distribution of fitness values ​​of multiple individuals after the i-th iteration of mutation meets the catastrophe triggering criterion. If yes, it indicates that a catastrophe has been triggered, and proceed to the next step; otherwise, proceed to step S38.

[0102] This invention designs a catastrophe-mechanism genetic algorithm that addresses the issue of mutation preventing the emergence of superoptimal individuals, thereby avoiding premature convergence to local optima. In practical applications, to ensure algorithm stability, the mutation probability is typically low. Once a superoptimal individual emerges, it requires several generations of mutation to produce new individuals, which can easily lead to local optima.

[0103] Step S37: Eliminate some individuals from the mutated individuals of the i-th iteration to obtain the individuals eliminated in the i-th iteration;

[0104] Perform a second update on the individuals eliminated in the i-th iteration to obtain multiple individuals with second updates in the i-th iteration;

[0105] Perform a second mutation operation on the multiple individuals that have undergone secondary updates in the i-th iteration to obtain multiple individuals that have undergone tertiary updates in the i-th iteration;

[0106] Optionally, the second mutation is a mutation operation with a high mutation probability;

[0107] Furthermore, the probability of the second mutation is m times the probability of the first mutation;

[0108] Optionally, m = 2;

[0109] Optionally, the expression for each variable of each individual in the population that has undergone three updates is:

[0110] x n '=x n +σ×N(0,1)

[0111] Where, x n ' is the value of the nth variable of an individual after the second mutation, x n σ is the original value of the nth variable of an individual before the second variation, σ is the maximum second variation amplitude, N(0,1) is the standard normal distribution, n=1,2,3…N, and N represents the total number of individuals.

[0112] Optionally, the disaster triggering criterion is expressed as:

[0113] J avg ≥ΞJ max

[0114] Among them, J avg J is the average fitness value of the current iteration step. max Ξ represents the maximum fitness value of the current generation, and Ξ is the density factor characterizing the density of individuals. The value of Ξ ranges from 0.5 to Ξ.

[0115] Optionally, step S37, which obtains the individuals of multiple secondary updates in the i-th iteration, includes the following specific steps:

[0116] Preset individual fitness threshold ΞJ max In the i-th iteration, select the individuals that exceed the fitness threshold of the individual.

[0117] Select the top a% of individuals in the i-th iteration as the eliminated individuals in the i-th iteration, and consider them to have survived the catastrophe; the rest of the individuals are eliminated.

[0118] The missing individuals of 1-a% are replaced by replicating the individuals eliminated in the i-th iteration, resulting in the individuals updated twice in the i-th iteration.

[0119] In this invention, the individuals eliminated in the current iteration step are those with good or suboptimal fitness that survive. They continue to provide good fitness values ​​in subsequent population reproduction. These individuals guide the population in the next iteration step to avoid getting trapped in local optima by using a catastrophe mechanism, to reproduce and evolve in a convergent direction, and to prevent the search process from becoming a simple or completely random search.

[0120] In this invention, the probability of individual mutation is relatively high, which can increase population diversity, make the search for solutions more diverse, and avoid getting trapped in local optima.

[0121] Step S38: Determine if i ≥ I. If not, treat the multiple individuals updated three times in the i-th iteration as the multiple individuals in the (i+1)-th iteration, let i = i+1, and return to step S32. If yes, then treat the optimization variables corresponding to the multiple individuals updated three times in the i-th iteration as the optimal variables.

[0122] Optionally, the specific steps to obtain multiple individuals after the i-th iteration crossover include:

[0123] The average fitness variance of each individual in the population is obtained and represented as individual diversity.

[0124] The crossover rate for each individual is obtained based on individual diversity;

[0125] Based on the crossover rate of each individual, crossover operations are performed on pairs of individuals to obtain multiple updated individuals. These updated individuals are then arranged according to their corresponding fitness to obtain multiple individuals after the i-th iteration crossover.

[0126] Optionally, the specific steps to obtain multiple individuals after the i-th iteration mutation include:

[0127] The variation rate of each individual is obtained based on individual diversity;

[0128] Based on the mutation rate of each population, mutation operations are performed on multiple individuals after the i-th iteration crossover to obtain multiple mutated individuals. These mutated individuals are then arranged according to their corresponding fitness to obtain multiple individuals after the i-th iteration mutation. The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the longitudinal stability of unmanned aerial vehicles (UAVs) based on a catastrophic mechanism genetic algorithm, characterized in that, include: Step 1: Obtain the optimization variables for the longitudinal stability of the UAV and set the corresponding domain; Step 2: Based on the optimization variables and corresponding domains of the UAV's longitudinal stability, obtain the lift margin of the UAV; Step 3: Based on the lift margin of the UAV and the constraints of the lift margin, establish the cost function for optimizing the longitudinal stability of the UAV. By using the cost function of UAV longitudinal stability optimization and the catastrophe mechanism genetic algorithm, multiple optimization variables of UAV longitudinal stability are optimized, and multiple optimal variables are obtained.

2. The method for optimizing the longitudinal stability of unmanned aerial vehicles based on a catastrophic mechanism genetic algorithm according to claim 1, characterized in that, The lift margin of a drone is expressed as: Where ΔL is the lift margin of the UAV, m is the mass of the UAV, g is the gravitational acceleration, q is the dynamic pressure, S is the projected area of ​​the wing, and C L η is the lift coefficient. ht It is the ratio of tail fin dynamic pressure to free-flow pressure, S ht It is the wetted area of ​​the horizontal stabilizer. It is the lift coefficient C of the horizontal stabilizer. Lht The derivative of α with respect to the angle of attack α ht It is the angle of attack of the horizontal stabilizer during stable flight.

3. The method for optimizing the longitudinal stability of unmanned aerial vehicles based on a catastrophic mechanism genetic algorithm according to claim 1, characterized in that, Step 3 includes the following specific steps: Step S31: Determine multiple individuals for the i-th iteration step, each individual including multiple optimization variables x1, x2, x3…x n i = 1, 2, 3…I, where I represents the total number of iterations; Step S32: Based on the lift margin of the UAV and the constraints of the lift margin, construct the cost function for optimizing the longitudinal stability of the UAV, and obtain the fitness value of each individual in the i-th iteration step. Step S33: Based on the fitness value of each individual in the i-th iteration step, perform parent selection and update for multiple individuals in the i-th iteration step to obtain multiple updated individuals in the i-th iteration. Step S34: Perform a crossover operation on the multiple updated individuals in the i-th iteration to obtain multiple individuals after the crossover in the i-th iteration; Step S35: Perform mutation operation on the multiple individuals after the i-th iteration crossover to obtain multiple individuals after the i-th iteration mutation; Step S36: Obtain the fitness value of each individual after the mutation in the i-th iteration; Determine whether the distribution of fitness values ​​of multiple individuals after the i-th iteration of mutation meets the catastrophe triggering criterion. If yes, it indicates that a catastrophe has been triggered, and proceed to the next step; otherwise, proceed to step S38. Step S37: Eliminate some individuals from the mutated individuals of the i-th iteration and perform a second update to obtain multiple second-updated individuals of the i-th iteration; Perform a second mutation operation on multiple individuals that have undergone secondary updates in the i-th iteration to obtain multiple individuals that have undergone tertiary updates in the i-th iteration; Step S38: Determine if i ≥ I. If not, treat the multiple individuals updated three times in the i-th iteration as the multiple individuals in the (i+1)-th iteration, let i = i+1, and return to step S32. If yes, then treat the optimization variables corresponding to the multiple individuals updated three times in the i-th iteration as the optimal variables.

4. The method for optimizing the longitudinal stability of unmanned aerial vehicles based on a catastrophic mechanism genetic algorithm according to claim 1, characterized in that, The cost function for optimizing the longitudinal stability of the UAV is expressed as: my J=κ1J sm +κ2J pm Where J is the cost of optimizing the longitudinal stability of the UAV, κ1 is the first weight coefficient, κ2 is the second weight coefficient, and J sm It is the static stability margin, J pm It is the pitching torque.

5. The method for optimizing the longitudinal stability of unmanned aerial vehicles based on a catastrophic mechanism genetic algorithm according to claim 3, characterized in that, The disaster triggering criterion is expressed as follows: J avg ≥ΞJ max Among them, J avg J is the average fitness value of the current iteration step. max Ξ represents the maximum fitness value in the current iteration, and Ξ is the density factor characterizing the density of individuals.

6. The method for optimizing the longitudinal stability of unmanned aerial vehicles based on a catastrophic mechanism genetic algorithm according to claim 3, characterized in that, Step S37, which obtains multiple individuals for the i-th iteration of the quadratic update, includes the following specific steps: A fitness threshold is preset for each individual, and individuals that exceed the fitness threshold in the i-th iteration are filtered out. Select the top a% of individuals in the i-th iteration as the eliminated individuals in the i-th iteration, and consider them to have survived the catastrophe; the rest of the individuals are eliminated. The missing individuals of 1-a% are replaced by replicating the individuals eliminated in the i-th iteration, resulting in multiple second-updated individuals in the i-th iteration.

7. The UAV longitudinal stability optimization method based on catastrophe mechanism genetic algorithm according to claim 4, wherein the static stability margin is expressed as: in, J sm For static stability margin; x cg It is the center of gravity, c w It is the mean aerodynamic chord of the wing. It is the lift coefficient C L The derivative with respect to the angle of attack α, It is the lift coefficient C of the horizontal stabilizer. Lht The derivative with respect to the angle of attack α, The normalized position of the wing's aerodynamic center. It is the normalized position of the aerodynamic center of the horizontal tail fin, η ht S is the ratio of tail dynamic pressure to free-flow pressure, and S is the derivative of the tail lift coefficient with respect to the angle of attack. ht It is the wetted area of ​​the horizontal tail fin, S w It is the wetted area of ​​the wing.

8. The UAV longitudinal stability optimization method based on catastrophic mechanism genetic algorithm according to claim 7, wherein the pitching moment is expressed as: in, J pm Let C be the pitching moment, q be the dynamic pressure, and C be the dynamic force. m It is the pitching moment coefficient, S is the wing's projected area, and α is the pitching moment coefficient. ht It is the angle of attack of the horizontal stabilizer during stable flight. ht ε is the tail lever arm length, ε is the wing downwash angle, and α is the angle of attack.

9. The method for optimizing the longitudinal stability of a UAV based on a catastrophic mechanism genetic algorithm according to claim 1, wherein the optimization variables are the wing projection area, the center of gravity position, the tail arm length, and the horizontal tail fin installation angle.

Images