Rotor blade bionic arrangement method and device based on improved genetic algorithm

Abstract

The invention discloses a rotor blade bionic arrangement method and device based on an improved genetic algorithm, belongs to the technical field of blade static unbalance measurement, and particularly relates to the rotor blade bionic arrangement method based on the improved genetic algorithm. The problem that the optimization result has no convergence when a population screening system based on a wheel disc selection algorithm is adopted in a traditional genetic algorithm is solved, and the problem that the traditional genetic algorithm cannot guarantee the expected performance of the algorithm is avoided. The method comprises the following steps: setting a fitness function inversely proportional to the unbalance amount; and performing iterative calculation of an improved genetic algorithm according to the sizes of the initial population and the parent population, the number of iterations, the fitness function, the crossover probability and the mutation probability to obtain an optimized blade array arrangement sequence. The rotor blade bionic arrangement method and device based on the improved genetic algorithm are suitable for adjusting the assembly sequence of blades of a turbine rotor of an aero-engine disc separation structure.

Description

Technical Field

[0001] The present invention relates to the technical field of blade static unbalance measurement, and particularly to a bionic arrangement method for rotor blades based on an improved genetic algorithm. Background Art

[0002] When an aeroengine turbine rotor rotates at high speed, vibration occurs. The fundamental reason is that the overall unbalance of the turbine rotor exceeds the tolerance. In actual assembly, measures such as grinding the balance surface or balance holes, adding mass blocks, spraying, and pasting patches are often used to reduce the rotor unbalance so that it meets the required standards. However, machining methods such as grinding the balance surface or balance holes and adding mass blocks will cause residual stress inside the turbine rotor, thereby affecting the fatigue strength and service life of the rotor; spraying and pasting patch methods have extremely high requirements for the rotor material properties and have limitations in the process of reducing the rotor unbalance. Therefore, there is an urgent need for a regulation process that neither damages the relevant components of the rotor nor can be widely applied.

[0003] For a turbine rotor with a disk-blade separation structure in an aeroengine, adjusting and optimizing the assembly sequence of rotor blades (i.e., rotor blade sorting) has a certain effect on reducing or even eliminating the rotor unbalance. However, the turbine rotor has a large number of blades, and the mass moment values of some blades are large, which poses a great challenge to adjusting and optimizing the assembly sequence to reduce the rotor unbalance.

[0004] The existing rotor blade sorting methods generally adopt a traditional genetic algorithm based on the roulette wheel selection algorithm as the population screening system. The optimization results of the traditional genetic algorithm based on the roulette wheel selection algorithm as the population screening system have a non-convergent situation. Summary of the Invention

[0005] The present invention proposes a bionic arrangement method for rotor blades based on an improved genetic algorithm, which solves the problem of non-convergence of the optimization results of the traditional genetic algorithm using the roulette wheel selection algorithm as the population screening system, and avoids the problem that the traditional genetic algorithm cannot guarantee the expected performance of the algorithm.

[0006] The bionic arrangement method for rotor blades based on an improved genetic algorithm according to the present invention obtains an optimized blade array arrangement sequence based on the improved genetic algorithm;

[0007] The method includes the following steps:

[0008] Step 1: Consider a set of blade arrangement sequences with different arrangement orders as a species population, and regard each blade arrangement sequence as a chromosome;

[0009] Step 2: Encode the blade arrangement sequence using real number encoding;

[0010] Step 3: Set the sizes of the initial population and the parental population, the number of iterations, the crossover probability, and the mutation probability; the crossover probability and the mutation probability are self-adjusting probabilities for avoiding local optimality;

[0011] Step 4: Set a fitness function that is inversely proportional to the unbalance amount;

[0012] Step 5: According to the sizes of the initial population and the parental population, the number of iterations, the fitness function, the crossover probability, and the mutation probability, perform iterative calculations of the improved genetic algorithm to obtain an optimized arrangement order of the blade array.

[0013] Further, a preferred implementation manner is provided, where the initial population is generated by randomly extracting random numbers.

[0014] Further, a preferred implementation manner is provided, where the fitness function is:

[0015] f(w) = u(w);

[0016] where u is the unbalance amount of the disk separation structure turbine rotor; w is the arrangement order of the blade array;

[0017] In the fitness function, the fitness value is inversely proportional to the overall unbalance amount after the assembly of the disk separation structure turbine rotor.

[0018] Further, a preferred implementation manner is provided, where the iterative calculation of the improved genetic algorithm includes the following steps:

[0019] Crossover and mutation operation steps: For the set of parental blade arrangement sequences, perform individual mutation operations based on the mutation probability and gene crossover operations based on the crossover probability to generate a set of offspring blade arrangement sequences that meet the set scale requirements;

[0020] Population screening step based on the elitist strategy: Combine the generated set of offspring blade arrangement sequences with the set of parental blade arrangement sequences; sort all the blade arrangement sequences in the combined set in ascending order according to their respective corresponding unbalance amount values; select the individuals in the top given proportion from the sorted set for retention, and these retained individuals will serve as the set of parental blade arrangement sequences for the next iteration.

[0021] Further, a preferred implementation manner is provided, where the crossover probability is:

[0022]

[0023] The mutation probability is:

[0024]

[0025] where:

[0026] p c is the crossover probability; p m is the mutation probability; q max is the maximum value of the individual fitness function value in the population; q avg is the average fitness function value of the individuals in the population; q' is the larger one of the fitness function values of the two gene crossover individuals; q is the fitness function value of the mutated individual; k1, k2, k3, k4 are constants.

[0027] Furthermore, a preferred embodiment is provided, where the number of iterations is 100 times.

[0028] The present invention also proposes a rotor blade bionic arrangement device based on an improved genetic algorithm. The device is based on the improved genetic algorithm to obtain an optimized blade array arrangement order;

[0029] The device includes the following modules:

[0030] Module 1: Regarding the set of blade arrangement sequences with different arrangement orders as a species population, and regarding each blade arrangement sequence as a chromosome;

[0031] Module 2: Encoding the blade arrangement sequence using real number encoding;

[0032] Module 3: Setting the sizes of the initial population and the parental population, the number of iterations, the crossover probability, and the mutation probability; the crossover probability and the mutation probability are self-adjusting probabilities for avoiding local optima;

[0033] Module 4: Setting a fitness function inversely proportional to the unbalance;

[0034] Module 5: Performing iterative calculations of the improved genetic algorithm according to the sizes of the initial population and the parental population, the number of iterations, the fitness function, the crossover probability, and the mutation probability to obtain an optimized blade array arrangement order.

[0035] The present invention also proposes a computer device, including: a processor and a memory. The memory is used to store the executable instructions of the processor, and the processor is configured to execute the rotor blade bionic arrangement method based on the improved genetic algorithm as described in any one of the above through executing the executable instructions.

[0036] The present invention also proposes a computer storage medium, in which a computer program is stored. When the computer program runs, it executes the rotor blade bionic arrangement method based on the improved genetic algorithm as described in any one of the above.

[0037] The present invention also provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the method for bionic arrangement of rotor blades based on an improved genetic algorithm described in any one of the above.

[0038] The present invention has the following beneficial effects:

[0039] 1. The method for bionic arrangement of rotor blades based on an improved genetic algorithm according to the present invention takes into account the problem of poor convergence of the traditional population screening system. By adopting the elitist strategy, individuals with smaller imbalance amounts in the parent and offspring are obtained, effectively controlling the evolutionary and convergence trends of the set. Then, iterative optimization of gene crossover and individual mutation is continuously carried out to explore a better blade arrangement scheme and achieve the purpose of reducing the imbalance amount.

[0040] 2. The method for bionic arrangement of rotor blades based on an improved genetic algorithm according to the present invention takes into account the problem that the crossover and mutation probabilities are difficult to control. By improving the original adaptive genetic algorithm, the global search probability of the algorithm is enhanced. The crossover and mutation probabilities are increased by reducing the variance of the population fitness function value, and the population stability is maintained according to the increased variance, inheriting the advantages of the original formula and enhancing the global search ability and avoiding the risk of local optimum.

[0041] 2. The method for bionic arrangement of rotor blades based on an improved genetic algorithm according to the present invention solves the problem of non-convergence of the optimization results in the population screening system of the traditional genetic algorithm using the roulette wheel selection algorithm, and avoids the problem that the traditional genetic algorithm cannot guarantee the expected performance of the algorithm.

[0042] The method and device for bionic arrangement of rotor blades based on an improved genetic algorithm according to the present invention are applicable to adjusting the assembly sequence of the blades of the turbine rotor of the aeroengine disk separation structure, can be used to optimize the imbalance amount after the assembly of the engine turbine rotor of the disk separation structure, and can further be used to guide the stacking of the engine rotor. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0044] Figure 1 It is a schematic diagram of the blade array arrangement under the improved genetic algorithm in an embodiment of the present invention;

[0045] Figure 2Schematic diagram of the combination of the offspring blade arrangement sequence set and the parent blade arrangement sequence set in an embodiment of the present invention;

[0046] Figure 3 Schematic diagram of selecting the top 50% of individuals from the sorted set for retention (i.e., retaining excellent individuals in the blade arrangement sequence set) in an embodiment of the present invention;

[0047] Figure 4 Schematic flow chart of the bionic arrangement method of rotor blades based on the improved genetic algorithm in an embodiment of the present invention. Detailed implementation manners

[0048] To make the technical solutions and advantages of the present invention more clearly expressed, the following will further describe in detail and completely the specific implementation manners of the present invention with reference to the accompanying drawings. The described embodiments are only some preferred embodiments of the present invention, rather than all the implementation manners; the described embodiments are intended to explain the present invention and should not be construed as a limitation to the present invention; the reasonable combination of the technical features defined in the embodiments of the present invention, and all other implementation manners obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention, fall within the scope of protection of the present invention.

[0049] In one embodiment, a bionic arrangement method of rotor blades based on an improved genetic algorithm is provided. The method is based on the improved genetic algorithm to obtain an optimized arrangement order of the blade array;

[0050] The method includes the following steps:

[0051] Step 1: Consider the blade arrangement sequence sets with different arrangement orders as species populations, and regard each blade arrangement sequence as a chromosome;

[0052] Step 2: Encode the blade arrangement sequence using real number coding;

[0053] Step 3: Set the size of the initial population and the parent population, the number of iterations, the crossover probability, and the mutation probability;

[0054] Step 4: Set a fitness function inversely proportional to the unbalance;

[0055] Step 5: Perform iterative calculations of the improved genetic algorithm according to the size of the initial population and the parent population, the number of iterations, the fitness function, the crossover probability, and the mutation probability to obtain an optimized arrangement order of the blade array.

[0056] It should be noted that since the blades are arranged in an array on the disk, a real number coding is used as the coding method for optimizing the rotor unbalance amount for the blade array arrangement. At the same time, the real number arrangement from 1 to N is output as the optimal solution sorting for optimizing the rotor unbalance amount of the blade array arrangement. Such a method can achieve the most accurate effect.

[0057] In addition, in one embodiment, the initial population is generated by random number extraction.

[0058] In this embodiment, the blade numbers in the blade arrangement sequence in the initial population are not repeated.

[0059] In this embodiment, to ensure that the blade arrangement sequence set has sufficient randomness at the initial stage, a method of random number extraction is used to generate a set of blade arrangement sequences with non-repeated blade numbers, and iterative operations are continuously performed until the blade arrangement sequence set meets the established requirements.

[0060] In addition, in one embodiment, the fitness function is:

[0061] f(w) = u(w);

[0062] where u is the unbalance amount of the disk-separated structure turbine rotor; w is the blade array arrangement order;

[0063] In the fitness function, the fitness value is inversely proportional to the overall unbalance amount after the assembly of the disk-separated structure turbine rotor.

[0064] In this embodiment, in order to make the overall unbalance amount after the assembly of the disk-separated structure turbine rotor as small as possible, the following formula is selected as the fitness function of the improved genetic algorithm:

[0065] f(w) = u(w);

[0066] where u is the unbalance amount of the disk-separated structure turbine rotor; w is the blade array arrangement order;

[0067] In this fitness function, the fitness value is inversely proportional to the overall unbalance amount after assembly, that is, the smaller the fitness value, the lower the overall unbalance amount after assembly. This is in line with the optimization goal pursued by the improved genetic algorithm, and can effectively guide the improved genetic algorithm to search and optimize in the direction of reducing the unbalance amount, thereby improving the balance performance of the turbine rotor after assembly.

[0068] In addition, in one embodiment, the iterative calculation of the improved genetic algorithm includes the following steps:

[0069] Crossing and mutation operation steps: For the set of parental leaf arrangement sequences, individual mutation operations are performed based on the mutation probability, and gene crossing operations are performed based on the crossing probability to generate a set of filial leaf arrangement sequences that meet the set scale requirements;

[0070] Population screening steps based on elitist strategy: Merge the generated set of filial leaf arrangement sequences with the set of parental leaf arrangement sequences; for all leaf arrangement sequences in the merged set, sort them in ascending order according to their respective corresponding imbalance values; select the top given percentage of individuals from the sorted set for retention, and these retained individuals will serve as the set of parental leaf arrangement sequences for the next iteration.

[0071] In this embodiment, in the population screening steps based on elitist strategy, the given percentage is 50%, that is, the top 50% of individuals are selected from the sorted set for retention.

[0072] In this embodiment, in terms of the population screening system, an elitist strategy is introduced to replace the traditional method based on roulette wheel selection algorithm. Specifically: Use crossing and mutation operations to generate a set of filial leaf arrangement sequences that reach the set scale, and then fuse it with the corresponding set of parental generation. In the fused set, each leaf arrangement sequence is arranged in ascending order according to its corresponding imbalance value, and then the top 50% of individuals are selected for retention to make them the parental leaf arrangement sequences of the next generation set. In this process, by selecting individuals with relatively smaller imbalance values in the parental and filial generations, the evolutionary direction and convergence trend of the set can be accurately controlled. On this basis, continue to carry out the iterative optimization process of subsequent gene crossing and individual mutation, continuously explore better leaf arrangement schemes, and gradually realize the optimization and improvement of the overall performance, and finally achieve the expected optimization goal of reducing the imbalance.

[0073] In this embodiment, an improved genetic algorithm based on elitist strategy is applied to retain the sequences with small imbalance values in each generation of leaf arrangement sequence sets to the next generation, prevent the loss of the optimal solution, and let it guide the new sequences to evolve in the direction of smaller imbalance, so that the algorithm results converge.

[0074] In addition, in one embodiment, the crossing probability is:

[0075]

[0076] The mutation probability is:

[0077]

[0078] Where:

[0079] p c is the crossing probability; pm is the mutation probability; q max is the maximum value of the individual fitness function value in the population; q avg is the average fitness function value of individuals in the population; q' is the larger one of the fitness function values of two gene-crossed individuals; q is the fitness function value of the mutated individual; k1, k2, k3, k4 are constants.

[0080] In this embodiment, the crossover probability and the mutation probability are self-adjusting probabilities (or formulas) for avoiding local optima.

[0081] In this embodiment, by self-adjusting the crossover probability and the mutation probability, the problem that it is difficult to regulate the crossover probability and the mutation probability is solved. The above self-adjusting crossover probability and mutation probability (formulas) are adopted:

[0082] When the variance of the population fitness function value decreases, the probabilities of crossover and mutation are correspondingly increased;

[0083] Conversely, if the variance of the population fitness function value increases, this will help to maintain the stability of the population. In the iterative process, the stability of the population enables excellent individuals to better retain their excellent characteristics, thereby strengthening the global search ability, effectively reducing the risk of falling into local optimal solutions, and providing a strong guarantee for obtaining better results.

[0084] In this embodiment, to enhance the search ability of the algorithm solution space, a gene crossover and individual mutation method with self-adjusting probability is proposed to optimize and improve the genetic algorithm, ensuring the accuracy of the imbalance of the blade array layout optimization.

[0085] It should be noted that those skilled in the art have proposed an adaptive genetic algorithm to solve the problem that it is difficult to regulate the crossover and mutation probabilities, as shown in the following formula:

[0086]

[0087] In the formula: p c is the crossover probability; p m is the mutation probability; q max is the maximum value of the individual fitness function value in the population; q avg is the average fitness function value of individuals in the population; q' is the larger one of the fitness function values of two gene-crossed individuals; q is the fitness function value of the mutated individual; k1, k2, k3, k4 are constants.

[0088] Through in-depth analysis of the above formula, it can be seen that there is a close relationship between the genetic operations of individuals, such as crossover and mutation, and the fitness function values of individuals. When the fitness of an individual approaches the lowest level of the population, the probabilities of crossover and mutation of this individual will correspondingly decrease; if these two probabilities drop to zero, it means that the fitness of this individual is at the minimum state. In previous research discussions, the reduction of the crossover probability has a positive effect on maintaining the stability of the population, while a lower mutation probability can make the convergence process of the algorithm easier to control. Especially in the later stage of algorithm iteration, in order to ensure the stable state of the population, this adaptive genetic algorithm can effectively meet this key requirement.

[0089] However, it should be noted that in the initial stage of population development, lower crossover and mutation probabilities not only cannot promote the generation and evolution of excellent individuals in the population, but on the contrary, due to excessive retention of the characteristics of the parent generation, it is difficult for the algorithm to break free when it falls into the dilemma of local optimal solutions. In view of this, it is particularly necessary to improve the original formula.

[0090] In this embodiment, the improved genetic algorithm improves the formulas of the crossover probability and mutation probability of the above adaptive genetic algorithm, that is, the formulas of the crossover probability and mutation probability of this embodiment. The improved formulas increase the probability of the algorithm for global search, enabling the algorithm to explore the solution space at a faster speed. Specifically:

[0091] The improved formulas adopt the means of reducing the variance of the population fitness function values to increase the probabilities of crossover and mutation. On the other hand, if the variance of the population fitness function values is increased, the stability of the population can be maintained, and thus excellent individuals have a stronger ability to maintain their own characteristics during the iteration process. By comparing with the original formula, it can be clearly seen that the improved formulas not only inherit the advantageous characteristics of maintaining the population stability in the original formula, but also can effectively enhance the global search ability by flexibly regulating the variance of the population fitness values, thereby successfully avoiding the risk of falling into local optimal solutions and providing a more powerful guarantee for the efficient operation of the algorithm.

[0092] In addition, in one embodiment, the number of iterations is 100 times.

[0093] In this embodiment, to improve the operation efficiency and avoid redundant calculations, the improved genetic algorithm implements an iteration termination mechanism, and the set iteration termination condition is that the algorithm iterates and loops until 100 times. At that time, the current lowest fitness function value and the corresponding blade array arrangement sequence (or called the blade array arrangement order) will be output as the final solution of the algorithm (that is, the final output optimization result).

[0094] In addition, in one embodiment, a bionic arrangement device for rotor blades based on an improved genetic algorithm is provided. The device obtains an optimized arrangement order of the blade array based on the improved genetic algorithm.

[0095] The device includes the following modules:

[0096] Module 1: Consider a set of blade arrangement sequences with different arrangement orders as a species population, where each blade arrangement sequence is regarded as a chromosome.

[0097] Module 2: Encode the blade arrangement sequence using real number encoding.

[0098] Module 3: Set the sizes of the initial population and the parental population, the number of iterations, the crossover probability, and the mutation probability.

[0099] Module 4: Set a fitness function inversely proportional to the unbalance.

[0100] Module 5: Perform iterative calculations of the improved genetic algorithm according to the sizes of the initial population and the parental population, the number of iterations, the fitness function, the crossover probability, and the mutation probability, and obtain an optimized arrangement order of the blade array.

[0101] The above further describes the technical solutions provided by the present invention in several specific embodiments to highlight the advantages and beneficial effects of the technical solutions provided by the present invention. However, the above several specific embodiments are not used as limitations on the present invention. Any reasonable modifications and improvements to the present invention, reasonable combinations of implementation manners, and equivalent replacements within the spirit and principle of the present invention should be included within the protection scope of the present invention.

Claims

1. A bionic arrangement method of rotor blades based on an improved genetic algorithm, characterized in that: The method is based on an improved genetic algorithm to obtain an optimized blade array arrangement order; The method comprises the following steps: Step 1: The set of leaf arrangement sequences with different arrangement orders is regarded as a species population, where each leaf arrangement sequence is regarded as a chromosome; Step 2: Use real number coding to encode the leaf arrangement sequence; Step 3: Set the size of the initial population and the parent population, the number of iterations, the crossover probability and the mutation probability; the crossover probability and the mutation probability are self-adjusting probabilities to avoid local optimality; Step 4: Set a fitness function that is inversely proportional to the imbalance; Step 5: According to the size of the initial population and the parent population, the number of iterations, the fitness function, the crossover probability and the mutation probability, an iterative calculation of the improved genetic algorithm is performed to obtain an optimized blade array arrangement order.

2. The bionic arrangement method of rotor blades based on improved genetic algorithm according to claim 1, characterized in that: The initial population is generated by random number extraction.

3. The bionic arrangement method of rotor blades based on improved genetic algorithm according to claim 1, characterized in that: The fitness function is: f(w)=u(w); Where u is the unbalance of the turbine rotor with disc separation structure; w is the arrangement order of the blade array; In the fitness function, the fitness value is inversely proportional to the overall unbalance amount of the turbine rotor with a disc separation structure after assembly.

4. The bionic arrangement method of rotor blades based on improved genetic algorithm according to claim 1, characterized in that: The iterative calculation of the improved genetic algorithm comprises the following steps: Crossover and mutation operation steps: for the parent generation leaf arrangement sequence set, perform individual mutation operation based on mutation probability, and perform gene crossover operation based on crossover probability to generate a descendant leaf arrangement sequence set that meets the set scale requirements; The steps of population screening based on the elitist strategy are as follows: merge the generated offspring leaf arrangement sequence set with the parent leaf arrangement sequence set; sort all leaf arrangement sequences in the merged set from small to large according to their corresponding imbalance values; select a given proportion of individuals from the sorted set and retain them. These retained individuals will serve as the parent leaf arrangement sequence set for the next iteration.

5. The bionic arrangement method of rotor blades based on improved genetic algorithm according to claim 1, characterized in that: The crossover probability is: The mutation probability is: in: p c is the crossover probability; p m is the mutation probability; q max is the maximum value of the individual fitness function in the population; q avg is the average fitness function value of individuals in the population; q' is the larger fitness function value of two gene crossover individuals; q is the fitness function value of the mutant individual; k1, k2, k3, k4 are constants.

6. The bionic arrangement method of rotor blades based on improved genetic algorithm according to claim 1, characterized in that: The number of iterations is 100.

7. A bionic arrangement device for rotor blades based on an improved genetic algorithm, characterized in that: The device obtains an optimized blade array arrangement sequence based on an improved genetic algorithm; The device comprises the following modules: Module 1: The set of leaf arrangement sequences with different arrangement orders is regarded as a species population, where each leaf arrangement sequence is regarded as a chromosome; Module 2: Use real number coding to encode the leaf arrangement sequence; Module 3: Setting the size of the initial population and the parent population, the number of iterations, the crossover probability and the mutation probability; the crossover probability and the mutation probability are self-adjusting probabilities to avoid local optimality; Module 4: Setting a fitness function inversely proportional to the amount of imbalance; Module 5: According to the size of the initial population and the parent population, the number of iterations, the fitness function, the crossover probability and the mutation probability, an iterative calculation of the improved genetic algorithm is performed to obtain an optimized blade array arrangement order.

8. A computer device comprising: A processor and a memory, characterized in that the memory is used to store executable instructions of the processor, and the processor is configured to execute the bionic arrangement method of rotor blades based on an improved genetic algorithm as described in any one of claims 1 to 6 by executing the executable instructions.

9. A computer storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is run, the bionic arrangement method of rotor blades based on an improved genetic algorithm according to any one of claims 1 to 6 is executed.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps of the bionic arrangement method of rotor blades based on an improved genetic algorithm as described in any one of claims 1 to 6 are implemented.

Images