Trial flight subject planning method based on series cutting variation genetic algorithm
By introducing tandem cutting variation operation in the test flight subject planning, the genetic algorithm is improved, the local optimal solution problem is solved, the efficiency and performance of the test flight subject allocation are improved, and the global optimal solution is found.
Patent Information
- Application Number
- CN202411351611.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-26
- Publication Date
- 2024-12-27
AI Technical Summary
Existing genetic algorithms are prone to fall into local optimal solutions in the test flight subject planning, and the exploration of genetic evolution is not in-depth enough, which affects the global optimization efficiency of the algorithm.
Genetic algorithm based on tandem cleavage mutation is adopted, and the test flight subject planning scheme is iteratively optimized through selection, crossover and tandem cleavage mutation operations to improve the efficiency of test flight subject allocation. The specific steps include setting the initial test flight subject planning scheme, calculating the required time, executing the selection operation of the roulette algorithm and elite retention strategy, using sequential crossing methods and tandem cutting and mutation operations until the maximum number of iterations is reached.
It effectively avoids the algorithm from falling into local optimal solutions, improves the efficiency and performance of the assignment of test flight subjects, and ensures that the global optimal solution is found.
Smart Images
Figure CN119204567A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of flight test, and in particular to a flight test subject planning method based on a tandem cutting mutation genetic algorithm. Background Art
[0002] Flight tests, or "flight tests" for short, are crucial in the aircraft development process. Their purpose is to ensure that the aircraft meets established design specifications and regulatory standards in terms of safety, performance, and reliability through a systematic and comprehensive testing and verification process. Efficient and reasonable planning of flight test subjects is a decisive factor in improving flight test efficiency and saving flight test costs.
[0003] In the existing technology, domestic and foreign experts have conducted some research on the flight test subject planning problem, mainly using genetic algorithms for optimization. Genetic algorithms simulate the process of biological evolution, randomly generate initial solutions and continuously optimize the quality of solutions through selection, crossover and mutation operations to find the optimal solution or near-optimal solution to the problem. It has a wide range of applications in global optimization of complex problems, engineering design, scheduling planning and machine learning, emphasizing its ability to effectively explore the solution space and adapt to diverse problems.
[0004] However, the existing genetic algorithms ignore the complexity of multi-aircraft subject allocation when solving the flight test subject planning problem, which may lead to falling into the local optimal solution and affect the global optimization efficiency of the algorithm. In addition, the current exploration of the genetic evolution part of the genetic algorithm is not deep enough, lacking a comprehensive understanding and application of different crossover and mutation operations to explore the solution space.
[0005] Therefore, there is an urgent need for a flight test subject planning method based on a tandem cutting mutation genetic algorithm to address the shortcomings of the existing technology. Summary of the invention
[0006] The purpose of the present invention is to provide a flight test subject planning method based on a tandem cutting mutation genetic algorithm, so as to solve the problems in the prior art that the improved genetic algorithm is prone to fall into a local optimal solution and the exploration of the genetic evolution part in the genetic algorithm is not deep enough, thereby improving the efficiency of flight test subject allocation.
[0007] To achieve the above object, the present invention provides the following solutions:
[0008] A flight test subject planning method based on a serial cutting mutation genetic algorithm, comprising:
[0009] Step 1: Set the total number of flight test subjects, the number of test aircraft and execution parameters, construct an initial flight test subject planning solution set, and set the initial flight test subject planning solution set as the current flight test subject planning solution set;
[0010] Step 2: Calculate the time required to complete each flight test subject planning scheme in the current flight test subject planning scheme set;
[0011] Step 3: Based on the required duration, sequentially perform a selection operation, a crossover operation, and a series cutting and mutation operation on the current flight test subject planning scheme set to obtain a new flight test subject planning scheme set;
[0012] Step 4: Iterate steps 2 and 3 until the maximum number of iterations is reached, and output the final flight test subject planning plan and the time required to complete it.
[0013] Optionally, the test flight subject in step 1 is represented by {R i , H i , {Q i}, Z i}, where R i represents the i-th test flight subject, H i The time required to complete the i-th test flight subject, Q i represents the set of prerequisite subjects required to complete the i-th flight test subject, Z i Indicates the execution status of the i-th flight test subject.
[0014] Optionally, the time required for completion of calculation in step 2 adopts a fitness function, and the fitness function is:
[0015] J = time (T P )
[0016] Where time(T p ) is T p The time required to complete the test flight subject allocation plan is represented, and J represents the fitness function jssj.
[0017] Optionally, the selection operation in step 3 is performed using a roulette algorithm and an elite retention strategy, wherein both the roulette algorithm and the elite retention strategy are selected according to the fitness value of the individual calculated using a fitness function, and the selection rule of the roulette algorithm is: the smaller the fitness value of the individual, the greater the probability of the individual being selected;
[0018] The selection rule of the elite retention strategy is: sort the fitness values of the individuals in ascending order, and select the individuals whose fitness values are ranked in the top m.
[0019] Optionally, the crossover operation in step 3 is performed using a sequential crossover method.
[0020] Optionally, performing the tandem cutting mutation operation in step 3 includes:
[0021] Determine the initial flight test subject for each test aircraft based on the flight test subject planning scheme set after cross operation;
[0022] Combine the initial test flight subjects in the order of the test aircraft serial numbers, with four test aircraft as a group, to obtain the test flight subjects to be allocated;
[0023] The flight test subjects to be assigned are randomly cut, and the cut flight test subjects to be assigned are reallocated to the test aircraft to complete the serial cutting variation operation.
[0024] The beneficial effects of the present invention are:
[0025] The present invention proposes a flight test subject planning method based on a tandem cutting mutation genetic algorithm, which can solve the problems in the prior art that the improved genetic algorithm is prone to fall into a local optimal solution and the exploration of the genetic evolution part of the genetic algorithm is not deep enough, thereby improving the performance of the improved genetic algorithm and improving the efficiency of flight test subject allocation. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0027] Figure 1 A flow chart of a flight test subject planning method based on a tandem cutting mutation genetic algorithm according to an embodiment of the present invention;
[0028] Figure 2 The parent generation and gene fragments randomly selected for the sequential crossover operation of the embodiment of the present invention;
[0029] Figure 3 A schematic diagram of a method for sequential crossover operation according to an embodiment of the present invention;
[0030] Figure 4 It is the initial test flight subject that the test aircraft undertakes when the tandem cutting variation of the embodiment of the present invention begins;
[0031] Figure 5 It is the test flight subjects to be assigned after merging the initial subjects undertaken by the test aircraft when the serial cutting variation of the embodiment of the present invention is carried out;
[0032] Figure 6 To undertake the test flight subjects for the test aircraft after reallocation at the end of the series cutting variation of the embodiment of the present invention;
[0033] Figure 7 A line graph comparing the tandem cutting variation, multi-point variation, uniform variation, and displacement variation of an embodiment of the present invention;
[0034] Figure 8 This is a Gantt chart of the flight test subject tasks in an embodiment of the present invention. DETAILED DESCRIPTION
[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0036] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0037] This embodiment provides a flight test subject planning method based on a serial cutting mutation genetic algorithm, including:
[0038] Step 1: Set the total number of flight test subjects, the number of test aircraft and execution parameters, build an initial flight test subject planning scheme set, and set the initial flight test subject planning scheme set as the current flight test subject planning scheme set;
[0039] Step 2, calculating the time required to complete each flight test subject planning scheme in the current flight test subject planning scheme set;
[0040] Step 3: Based on the required duration, select, cross, and serial cut mutation operations are sequentially performed on the current flight test subject planning scheme set to obtain a new flight test subject planning scheme set;
[0041] Step 4: Iterate steps 2 and 3 until the maximum number of iterations is reached, and output the final flight test subject planning plan and the time required to complete it.
[0042] Furthermore, the test flight subject in step 1 is represented as {R i , H i , {Q i}, Z i}, where R i represents the i-th test flight subject, H i The time required to complete the i-th test flight subject, Q i represents the set of prerequisite subjects required to complete the i-th flight test subject, Z i Indicates the execution status of the i-th flight test subject.
[0043] Furthermore, the time required to complete the calculation in step 2 is calculated using the jssj fitness function.
[0044] In step 3, the selection operation is performed using the roulette algorithm and the elite retention strategy, wherein both the roulette algorithm and the elite retention strategy are selected according to the fitness value of the individual calculated by the fitness function. The selection rule of the roulette algorithm is: when the fitness value of the individual is smaller, the probability of the individual being selected is greater;
[0045] The selection rule of the elite retention strategy is: sort the fitness values of the individuals in ascending order, and select the individuals with the top m fitness values.
[0046] The crossover operation in step 3 adopts a sequential crossover method.
[0047] The tandem cutting mutation operation performed in step 3 includes:
[0048] Determine the initial flight test subject for each test aircraft based on the flight test subject planning scheme set after cross operation;
[0049] Combine the initial test flight subjects in the order of the test aircraft serial numbers, with four test aircraft as a group, to obtain the test flight subjects to be allocated;
[0050] The flight test subjects to be assigned are randomly cut, and the cut flight test subjects to be assigned are reallocated to the test aircraft to complete the serial cutting mutation operation.
[0051] Combine the following Figure 1-Figure 6 A flight test subject planning method based on a tandem cutting mutation genetic algorithm proposed in this embodiment is described in detail, and specifically includes the following steps:
[0052] Step 1: Initialize the population to form different flight test subject planning schemes and determine parameters such as population size, operator parameters, and maximum number of iterations.
[0053] Coding is to express the problem to be solved as a chromosome or individual in the genetic space, so as to directly or indirectly process the parameters of the problem. The solution to the flight test subject planning problem is the mission sequence number and execution time undertaken by the pilot, so there is no need for transcoding and decoding operations, so the direct encoding method is used to encode the individual. In order to simplify the complexity of the task, the flight test subjects are encoded in the form of task sequences based on the characteristics of the flight test subjects. The encoding process is as follows:
[0054] (1) Assume that the total number of flight test subjects to be completed is N, R i represents the i-th test flight subject, where i = 1, 2, ..., N;
[0055] (2) Suppose you need to complete subject R i The duration is H, H i represents the time required to complete the i-th subject, where i = 1, 2, ..., N;
[0056] (3) Set subject Ri The prerequisite subject is the set Q, Q i Indicates completion of subject R i The set of pre-requisites to be completed, where i = 1, 2, ..., N. For the first batch of flight test subjects, in the model proposed in this embodiment, the set of pre-requisites can be considered to be an empty set;
[0057] (4) Set subject R i The execution status is Z, Z i Indicates subject R i Execution state of Z, where i = 1, 2, ... N. i =1 means the flight test subject is completed, Z i =0 means the test flight subject has not been executed.
[0058] The set of all the test flight subjects described above is the subject set R that needs to be executed, R = (R1, R2, ..., R n ); the subject duration set is H = (H1, H2, ..., H n ); the set of pre-condition subjects is Q = (Q1, Q2, ..., Q n ); the subject execution status is Z = (Z1, Z2, ..., Z n ); then each specific flight test subject can be described as {R i , H i , {Q i}, Z i}.
[0059] Step 2: Use the prepared fitness function to calculate the time required for each test flight subject planning plan, that is, to evaluate the fitness of each individual in the population.
[0060] The fitness function is a function used to measure the quality of individuals in the genetic algorithm. Its role is to quantify the fitness of individuals so that individuals can be selected, crossed and mutated during the evolution process, and finally the optimal solution to the problem can be found. The form of the fitness function will vary according to the nature of the problem and the optimization goal. The goal of the flight test subject planning problem in this embodiment is to determine the allocation plan with the shortest total flight test subject time, so the total time is used as the criterion for judging the quality of individuals. J is a function that judges the time required for the flight test subject allocation plan, and its expression is:
[0061] J = time (T P )
[0062] Where: time(T p ) is T p The time required to complete the flight test subject allocation plan represented by jssj is used as the fitness function, and the smaller the value, the better the individual.
[0063] Step 3: Based on the result of function calculation, a selection operation combining the elite retention strategy and roulette is performed to form a new population.
[0064] The selection operation is to select individuals from the population for crossover and mutation operations. This embodiment combines the roulette algorithm with the elite retention strategy to perform crossover and mutation operations, aiming to pass on superior genes to offspring, thereby increasing the probability of generating higher quality offspring. The roulette algorithm selects based on the fitness value of the individual. The smaller the fitness value of the individual, the greater the probability of being selected, making it more likely that excellent parent genes will be passed on to offspring. At the same time, the elite retention strategy directly selects the best parent individuals and passes them on to offspring, solving the problem that the roulette algorithm may miss the best parent.
[0065] The probability of an individual being selected is denoted by X a :
[0066]
[0067] Where: f a is the fitness value of each individual in the population; F is the total fitness of the population.
[0068] Step 4: Perform a crossover operation using the sequential crossover method based on the new population obtained in step 3.
[0069] The crossover operation obtains offspring by crossing two partial gene fragments of the parent generation, and inherits some of its own characteristics to the offspring, increasing the diversity of the population, thereby improving the global search capability. The flight test subject planning problem involves complex task dependencies and execution order requirements. Therefore, this embodiment uses a sequential crossover method to ensure that the generated solution can meet the logical relationship between subjects, while improving the diversity and effectiveness of the solution. The specific steps of the sequential crossover method are as follows:
[0070] (1) Randomly select the segments that need to be crossed, such as Figure 2 The shaded part represents the fragments that need to be crossed with the corresponding parent.
[0071] (2) Figure 3 As shown, one offspring inherits the color-coded gene fragment from its parent, and then the missing genes are filled in order in another parent, and another offspring is obtained in the same way.
[0072] Step 5: Perform a tandem cutting mutation operation based on the result of step 4.
[0073] The mutation operation generates offspring by changing some genes of the parent generation, which is used to maintain and introduce diversity in the population and prevent the algorithm from prematurely converging to the local optimal solution. Since the number of subjects undertaken by the test machine gradually becomes fixed during the iteration of the genetic algorithm, the algorithm converges slowly and is prone to fall into the local optimal solution.
[0074] This embodiment proposes a serial cutting mutation method to solve this problem. Serial cutting mutation is a mutation operation for the flight test subject allocation plan. First, multiple chromosomes are merged into one chromosome, and cutting points are randomly selected in this individual. The task allocation plan is cut into multiple parts according to these cutting points, and then these parts are recombined to create a new task allocation plan, thereby solving the problem of the algorithm falling into a local optimum. The specific steps of the designed mutation operator are as follows:
[0075] (1) After cross operation, determine the subjects to be undertaken on each test aircraft, such as Figure 4 shown.
[0076] (2) Merge the subjects on the four test machines in order of their serial numbers to form a single chromosome, such as Figure 5 shown.
[0077] (3) Starting from the first subject of the chromosome, a gene is randomly selected and assigned to the first test machine, and so on, until all subjects are assigned to the test machines. Figure 6 shown.
[0078] Step 6: Determine whether the termination condition is met. When the algorithm reaches the maximum number of iterations, output the current plan and the shortest flight test subject planning time, otherwise go to step 2.
[0079] verify:
[0080] In order to verify the effectiveness of the flight test subject planning method based on the tandem cutting mutation genetic algorithm proposed in this embodiment, the algorithm is tested by data analysis. Four test aircraft and 23 flight test subjects are set. The execution time of the flight test subjects and the information of the pre-condition subjects are shown in Table 1.
[0081] Table 1
[0082]
[0083]
[0084] In the improved genetic algorithm used, other parts are kept unchanged, and only the mutation part is compared and verified. Three traditional mutation operators are selected, including multipoint mutation (MPM), uniform mutation (UM) and shift mutation (SM), and compared with the designed serial cutting mutation. Then, the code is run ten times for each algorithm, and the average value is taken to draw a comparison line chart of the mutation operator, as shown in the figure. Figure 7As shown in the figure, the results show that in 20 iterations, the improved genetic algorithm using serial cutting mutation first found the global optimal solution, and other mutation operators did not find the global optimal solution within the iteration range. It can be seen that the serial cutting mutation operator can greatly improve the performance of the improved genetic algorithm, thereby effectively avoiding the algorithm from falling into the situation of local optimal solution.
[0085] In order to present the final flight test subject allocation plan more intuitively, this embodiment adopts the task Gantt chart method, such as Figure 8 , and the execution status and task arrangement of each test flight subject on different test aircraft are given. The results show that the improved genetic algorithm proposed in this embodiment can optimize the time required for the test flight subject allocation plan. Figure 8 The horizontal axis represents time, and the vertical axis represents the test aircraft number. Each bar represents the execution period of a test flight subject, and different serial numbers represent test flight subjects on different test aircraft. By comparing the relative positions between the bars, you can intuitively judge the relationship between the test flight subjects and the execution status on different test aircraft.
[0086] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.
Claims
1. A flight test subject planning method based on a tandem cutting mutation genetic algorithm, characterized in that: include: Step 1: Set the total number of flight test subjects, the number of test aircraft and execution parameters, construct an initial flight test subject planning solution set, and set the initial flight test subject planning solution set as the current flight test subject planning solution set; Step 2: Calculate the time required to complete each flight test subject planning scheme in the current flight test subject planning scheme set; Step 3: Based on the required duration, sequentially perform a selection operation, a crossover operation, and a series cutting and mutation operation on the current flight test subject planning scheme set to obtain a new flight test subject planning scheme set; Step 4: Iterate steps 2 and 3 until the maximum number of iterations is reached, and output the final flight test subject planning plan and the time required to complete it.
2. The flight test subject planning method based on the tandem cutting mutation genetic algorithm according to claim 1 is characterized in that: The test flight subject in step 1 is represented by {R i , H i , {Q i }, Z i }, where R i represents the i-th test flight subject, H i The time required to complete the i-th test flight subject, Q i represents the set of prerequisite subjects required to complete the i-th flight test subject, Z i Indicates the execution status of the i-th flight test subject.
3. The flight test subject planning method based on the tandem cutting mutation genetic algorithm according to claim 1 is characterized in that: The time required for calculation in step 2 is calculated using a fitness function, which is: J=time(T P ) Where time(T p ) is T p The time required to complete the test flight subject allocation plan is represented, and J represents the fitness function jssj.
4. The flight test subject planning method based on the tandem cutting mutation genetic algorithm according to claim 1 is characterized in that: The selection operation in step 3 is performed using a roulette algorithm and an elite retention strategy, wherein both the roulette algorithm and the elite retention strategy are selected according to the fitness value of the individual calculated using a fitness function, and the selection rule of the roulette algorithm is: the smaller the fitness value of the individual, the greater the probability of the individual being selected; The selection rule of the elite retention strategy is: sort the fitness values of the individuals in ascending order, and select the individuals whose fitness values are ranked in the top m.
5. The flight test subject planning method based on the tandem cutting mutation genetic algorithm according to claim 1 is characterized in that: The crossover operation in step 3 is performed using a sequential crossover method.
6. The flight test subject planning method based on the tandem cutting mutation genetic algorithm according to claim 1 is characterized in that: The step 3 of performing the tandem cutting mutation operation includes: Determine the initial flight test subject for each test aircraft based on the flight test subject planning scheme set after cross operation; Combine the initial test flight subjects in the order of the test aircraft serial numbers, with four test aircraft as a group, to obtain the test flight subjects to be allocated; The flight test subjects to be assigned are randomly cut, and the cut flight test subjects to be assigned are reallocated to the test aircraft to complete the serial cutting variation operation.