A Phase Change Launch Intelligent Control Method Based on Genetic Algorithm

Abstract

The application discloses a phase change ejection intelligent control method based on a genetic algorithm, and steps are as follows: obtaining initial parameters of a phase change ejection system and an ejection velocity of an ejected object; calculating an ejection velocity range under a current mass specification; if a target ejection velocity is within the ejection velocity range, parameter optimization is performed on a phase change number, otherwise, parameter optimization is ended; judging whether a ratio of the target ejection velocity and a minimum value of the ejection velocity is less than a set threshold value; if yes, an initial phase change number is saved as an optimal phase change number, then parameter optimization is performed on a time interval, otherwise, the phase change number is optimized based on the genetic algorithm to obtain an optimal phase change number; the time interval is optimized based on the genetic algorithm to obtain an optimal time interval; and the optimal phase change number and the optimal time interval are output. The application performs multi-target intelligent optimization control according to a priority of parameters based on the genetic algorithm, and performs hierarchical and step-by-step optimization control, so that the precision and efficiency of the phase change ejection intelligent control are improved.

Description

Technical Field

[0001] This invention belongs to the field of powered launch technology, and in particular relates to a phase change ejection intelligent control method based on genetic algorithm. Background Technology

[0002] Phase change ejection (PCE) technology, represented by supercritical carbon dioxide, boasts advantages such as clean working fluid, rapid reaction, low product temperature, low launch characteristics, and strong adaptability, making it a hot research topic in weapon system ejection power source research. In PCE control, the ejection velocity of the projectile is typically used as the target. However, the internal ballistic characteristics and ejection dynamic characteristics of a PCE system are influenced by numerous factors. Therefore, establishing a multi-objective control model using initial parameters of the PCE and the state parameters of the projectile as input variables to achieve precise and efficient control of PCE is a pressing problem that needs to be solved. Summary of the Invention

[0003] The purpose of this invention is to provide a phase change ejection intelligent control method based on genetic algorithms that can improve the accuracy and efficiency of phase change ejection control.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A phase change ejection intelligent control method based on genetic algorithm includes the following steps:

[0006] Step S1: Obtain the initial parameters of the phase change ejection system and the derailment velocity of the ejected object. The initial parameters include the initial mass specification, the initial number of phase changes, and the initial time interval. The derailment velocity of the ejected object is used as the target derailment velocity.

[0007] Step S2: Quality specification determination; Calculate the derailment speed range under the current quality specification. If the target derailment speed is within the derailment speed range, further optimize the number of phase transitions; if the target derailment speed is less than the minimum derailment speed or greater than the maximum derailment speed, end the parameter optimization.

[0008] Step S3: Phase transition number optimization; Determine whether the ratio of the target deorbit speed to the minimum deorbit speed is less than a set threshold. If so, save the initial number of phase transitions as the optimal value of the number of phase transitions, and then further optimize the time interval parameters. Otherwise, optimize the number of phase transitions based on the genetic algorithm to obtain the optimal value of the number of phase transitions.

[0009] Step S4: Time interval optimization; The time interval is optimized based on a genetic algorithm to obtain the optimal value of the time interval;

[0010] Step S5: Output the optimal value for the number of phase transitions and the optimal value for the time interval, and perform phase transition ejection control according to the optimal value for the number of phase transitions and the optimal value for the time interval.

[0011] In some embodiments, the step of optimizing the number of phase transitions using a genetic algorithm in step S3 is as follows:

[0012] S01. Randomly generate M individuals as the initial population P1(0) for the number of phase transitions;

[0013] S02. Calculate the fitness of each individual in the population P1(t) with the number of phase transitions, where t is the generation number, t = 0, 1, 2, ..., and the population at t = 0 is the initial population P1(0) with the number of phase transitions.

[0014] S03. Perform selection, crossover, and mutation operations sequentially on the population P1(t) after fitness evaluation to obtain the next generation population P1(t+1), and update the generation number t to t+1.

[0015] S04. If the number of generations t is less than the maximum number of generations T, then return to S02; otherwise, output the individual corresponding to the maximum fitness value obtained during the evolution process as the optimal value for the number of phase transitions.

[0016] In some embodiments, the step of optimizing the time interval using a genetic algorithm in step S4 is as follows:

[0017] S01. Randomly generate M individuals as the initial time interval group P2(0), in which the later time interval is greater than the previous time interval;

[0018] S02, calculate the fitness of each individual in the time interval population P2(t);

[0019] S03. Perform selection, crossover, and mutation operations sequentially on the population P2(t) after fitness evaluation to obtain the next generation population P2(t+1), and update the generation number t to t+1.

[0020] S04. If the number of generations t is less than the maximum number of generations T, then return to S02; otherwise, output the individual corresponding to the maximum fitness value obtained during the evolution process as the optimal time interval.

[0021] In some embodiments, steps S3 and S4 employ an improved genetic algorithm to optimize the number of phase transitions and the time interval, as follows:

[0022] S001, Population initialization; Based on the upper and lower limits of the parameter to be optimized, randomly generate Np individuals as the initial population f(0); The parameter to be optimized is the number of phase transitions or the time interval;

[0023] S002, Initial fitness evaluation; Calculate the fitness value of each individual in the population f(g), the evolutionary generations g = 0, 1, 2, ..., the population when g = 0 is the initial population f(0);

[0024] S003, Selection operation: Sort the individuals in the population f(g) in ascending order of their calculated fitness values, select the individual with the smallest fitness value as the optimal individual, and save it.

[0025] S004, Crossover operation: Cross the best individual with the other even-numbered individuals after ascending sorting. Each crossover produces two new individuals, forming a new population f'(g).

[0026] S005, Mutation operation; Perform multi-point mutation on f'(g) to generate offspring population ff(g);

[0027] S006. Secondary fitness evaluation: Calculate the fitness value of each individual in the offspring population ff(g) and sort them in ascending order according to the fitness values ​​calculated for each individual in the population ff(g);

[0028] S007, cross-generational optimization: merge the reordered offspring population ff(g) with the initial population f(0), sort the merged individuals in ascending order according to their fitness values, take the first Np individuals as the next generation population f(g+1), and update the generation number g to g+1.

[0029] S008. If g > G, then the individual with the smallest fitness value in each generation is output as the optimal individual, and the calculation is terminated; otherwise, return to step S002.

[0030] Furthermore, the fitness value is the absolute value of the difference between the calculated off-orbit velocity and the target parameter value.

[0031] As can be seen from the above technical solutions, this invention uses a genetic algorithm to optimize the control parameters in the phase change catapult control process. Addressing the problem of precise control of phase change catapult parameters, and taking the derailment velocity of the launched object as the objective, it proposes methods for quality specification determination, intelligent optimization of the number of phase changes based on the genetic algorithm, and intelligent optimization of the interval time. This achieves intelligent control of control parameters such as phase change quality, number of phase changes, and interval time in phase change catapults. Furthermore, the hierarchical and step-by-step control improves the accuracy and efficiency of intelligent phase change catapult control. In some embodiments, an improved genetic algorithm is used for parameter optimization. By improving the standard genetic algorithm through parameter design strategies, fitness encoding mechanisms, selection strategies, crossover operators, secondary evaluation, and parent-child generational optimization, it features dynamic parameter adjustment, efficient optimal solution determination, focused selection and crossover operations, and parent-child generational doubling optimization. This effectively avoids getting trapped in local mechanisms, improves the efficiency, accuracy, and stability of the algorithm, and achieves intelligent optimization of multi-objective parameters for different launch velocities. Attached Figure Description

[0032] To more clearly illustrate the embodiments of the present invention, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 The flowchart is for a standard genetic algorithm;

[0034] Figure 2 This is a flowchart of the method of the present invention;

[0035] Figure 3 This is a flowchart of the improved genetic algorithm of the present invention;

[0036] Figure 4 The graph shows the change in fitness with the number of generations when using a phase change ejection intelligent control method based on an improved genetic algorithm for simulation.

[0037] Figure 5 The graph shows the change in fitness with the number of generations when the phase change ejection intelligent control method based on the improved genetic algorithm is used for simulation at different target de-orbit velocities.

[0038] Figure 6 The error distribution diagram of the calculated derailment velocity and the target derailment velocity when the phase change ejection intelligent control method based on the improved genetic algorithm is used in the simulation.

[0039] Figure 7 The time interval distribution is obtained from multiple simulations using a phase change ejection intelligent control method based on an improved genetic algorithm.

[0040] Figure 8 This is a comparison of the ejection velocity curves of the projectile during multiple simulations using a phase change ejection intelligent control method based on an improved genetic algorithm.

[0041] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings. Detailed Implementation

[0042] The present invention will now be described in detail with reference to the accompanying drawings. In the detailed description of the embodiments of the present invention, for ease of explanation, the drawings illustrating the device structure will be partially enlarged without adhering to the general scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of the present invention. It should be noted that the drawings are in a simplified form and use non-precise scales, solely for the purpose of conveniently and clearly illustrating the embodiments of the present invention. Additionally, in the description of this application, terms such as "first" and "second" are used only to distinguish descriptions and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Terms such as "positive," "negative," "bottom," "upper," and "lower" indicate orientation or positional relationships based on the orientation or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0043] The control objective of the phase change ejection control system is the ejection velocity (also known as the ejection speed) of the ejected object. The ejection velocity of the ejected object is related to the internal ballistic characteristics and ejection dynamic characteristics of the phase change ejection. The internal ballistic characteristics and ejection dynamic characteristics of the phase change ejection are affected by many factors, such as the initial parameters of the phase change ejection and the state parameters of the ejected object. These parameters are all variables in the phase change ejection control process. How to achieve multi-parameter intelligent control is the key to phase change ejection control.

[0044] Genetic algorithms possess powerful global optimal solution search capabilities, the duality of the problem domain, parallelism in information processing, robustness in application, and simplicity of operation, making them highly effective in multi-parameter intelligent control. Based on this, this invention proposes a phase-change catapult intelligent control method based on a genetic algorithm. The method uses a genetic algorithm to optimize control parameters, achieving intelligent optimization of multi-objective parameters at different launch velocities. Through hierarchical and step-by-step control, the accuracy and efficiency of phase-change catapult intelligent control are improved.

[0045] Figure 1 The flowchart of the standard genetic algorithm is shown below. Figure 1 This section provides a brief explanation of the steps involved in a standard genetic algorithm. For a detailed explanation of genetic algorithms, please refer to "Intelligent Optimization Algorithms and Their MATLAB Examples" published by the Electronic Industry Press. Figure 1As shown, the standard genetic algorithm includes the following steps:

[0046] S01, Population initialization; Set the generation counter t = 0, set the maximum generation T, and randomly generate M individuals as the initial population P(0);

[0047] S02, Fitness Evaluation; Calculate the fitness FIT of each individual in the population P(t), where t is the generation number, t = 0, 1, 2, ..., the population P(0) at t = 0 is the first generation population, that is, the initial population P(0) is the first generation population, the population P(1) at t = 1 is the second generation population, and so on; In this invention, the absolute value of the difference between the calculated deorbit velocity and the target parameter value is taken as the fitness value FIT;

[0048] S03, Operations: After fitness evaluation, the population P(t) is subjected to selection, crossover, and mutation operations in sequence to obtain the next generation population P(t+1), and the generation number t is updated to t+1. The purpose of selection and crossover operations is to directly pass on the optimized individuals to the next generation or to generate new individuals through pairing and crossover and then pass them on to the next generation. Mutation operations are used to change the gene values ​​at certain loci in the individual strings of the population.

[0049] Specifically, the selection operation selects the individual with the highest fitness in the population if the goal is to maximize the fitness, and selects the individual with the lowest fitness if the goal is to minimize the fitness. The crossover operation randomly crosses all individuals in the population according to probability. This random crossover process does not change the sample set, only the positions change. The mutation operation performs mutation according to the mutation probability. For details on how to implement the selection, crossover and mutation operations of the standard genetic algorithm, please refer to the book "Intelligent Optimization Algorithms and Their MATLAB Examples".

[0050] S04. Termination condition judgment: Determine whether the number of generations t is less than the maximum number of generations T. If it is, return to step S02. Otherwise, output the individual corresponding to the maximum fitness value obtained during the evolution process as the optimal value and terminate the calculation.

[0051] To address the problem of precise control of phase change ejection parameters, this invention proposes an intelligent control method for phase change ejection based on a genetic algorithm, taking the derailment velocity of the ejected object as the objective. This method optimizes control parameters such as phase change mass, number of phase changes, and interval time. A hierarchical and step-by-step optimization strategy is employed: first optimizing the mass specifications, then the number of phase changes, and finally the interval time. A genetic algorithm is used to optimize the number of phase changes and the interval time, thus achieving intelligent control of the phase change ejection. Figure 2 As shown, the phase change ejection intelligent control method of the present invention includes the following steps:

[0052] Step S1: Obtain the initial parameters of the phase change ejection system and the derailment speed of the ejected object. The initial parameters include the initial mass specification, the initial number of phase changes, and the initial time interval. The derailment speed of the ejected object is taken as the target derailment speed, and the derailment speed of the ejected object is the controlled object.

[0053] Step S2: Quality Specification Determination; Calculate the derailment speed range under the current quality specification, i.e., calculate the maximum and minimum derailment speed under the current quality specification. If the target derailment speed is within the range of the maximum and minimum derailment speed, further optimize the number of phase transitions. If the target derailment speed is less than the minimum derailment speed or greater than the maximum derailment speed, end the parameter optimization. In specific applications, if the target derailment speed is greater than the maximum derailment speed, prompt "Quality specification exceeds the maximum value, reduce quality specification" and then end. If the target derailment speed is less than the minimum derailment speed, prompt "Quality specification is lower than the minimum value, increase quality specification" and then end, avoiding unnecessary calculations.

[0054] Specifically, the maximum and minimum derailment velocities under a given mass specification are calculated based on the range of the number of phase transitions and the time interval of the phase change catapult system. The maximum derailment velocity is calculated based on the maximum number of phase transitions and the minimum time interval of the phase change catapult system under that mass specification, while the minimum derailment velocity is calculated based on the minimum number of phase transitions of the phase change catapult system under that mass specification. For example, if the phase change catapult system has 7 launchers, then the range of the number of phase transitions is 1 to 7, and the range of the time interval is 0 to 0.2 seconds. The maximum derailment velocity is the velocity when all 7 launchers are activated simultaneously (with a time interval of 0 seconds), and the minimum derailment velocity is the velocity when only 1 launcher is activated. The time interval is the time interval between launches from different launchers.

[0055] The deorbit velocity is calculated based on the internal ballistic mathematical model; it is a calculated value. Specifically, the calculated deorbit velocity value... Where m is the mass of the projectile, l is the trajectory length of the launcher, and F(t) is the internal thrust of the launcher. The internal thrust of the launcher is related to the number of phase transitions and the time interval of the phase transition catapult system, and it is a function that varies with time t.

[0056] Step S3: Phase transition number optimization; Determine whether the ratio of the target de-orbit speed to the minimum de-orbit speed is less than the set threshold. If so, there is no need to optimize the number of phase transitions, and the initial number of phase transitions is saved as the optimal value. Then, further optimize the time interval parameters to improve operating efficiency. Otherwise, optimize the number of phase transitions based on the genetic algorithm to obtain the optimal value of the number of phase transitions.

[0057] The threshold is set as an empirical value, which is related to the quality specifications and the number of transmitters. It can be adjusted accordingly after the maximum and minimum deorbit velocities are calculated to improve search efficiency. In this embodiment, the threshold is set to 1.15. In this step, the number of phase transitions is dynamically adjusted according to the ratio of the target deorbit velocity to the minimum deorbit velocity. This allows for adjustment of the upper and lower limits of random number selection when performing parameter optimization using a genetic algorithm. Then, an initial population of phase transitions is randomly generated to further focus the search range, avoid searching for unnecessary points, and improve the efficiency and quality of searching for the optimal solution by searching more points in the possible domain.

[0058] Step S4: Time interval optimization; The time interval is optimized based on a genetic algorithm to obtain the optimal value of the time interval;

[0059] Step S5: Output the optimal value for the number of phase transitions and the optimal value for the time interval, and perform phase transition ejection control according to the optimal value for the number of phase transitions and the optimal value for the time interval.

[0060] Specifically, in step S3, when the ratio of the target deorbit velocity to the minimum deorbit velocity is greater than or equal to a set threshold, the specific steps for optimizing the number of phase transitions using a genetic algorithm are as follows:

[0061] S01. Set the generation counter t=0 and the maximum generation T, and randomly generate M individuals as the initial population P1(0) for the number of phase transitions;

[0062] S02. Calculate the fitness of each individual in the population P1(t) with the number of phase transitions;

[0063] S03. Perform selection, crossover, and mutation operations sequentially on the population P1(t) after fitness evaluation to obtain the next generation population P1(t+1), and update the generation number t to t+1.

[0064] S04. If the number of generations t is less than the maximum number of generations T, return to step S02 and recalculate the fitness of each individual in the population P1(t) with the number of phase transitions. Otherwise, output the individual corresponding to the maximum fitness value obtained during the evolution process as the optimal value of the number of phase transitions.

[0065] In step S3, the steps for optimizing the time interval using a genetic algorithm are the same as those for optimizing the number of phase transitions. After generating the initial time interval population P2(0), fitness is calculated, followed by selection, crossover, and mutation operations to obtain the next generation population P2(t+1). Each step can be referenced from the steps for optimizing the number of phase transitions, and will not be repeated here. When using a genetic algorithm to optimize the time interval, during the random generation of the initial time interval population, due to the irreversibility of time, the later time interval must be longer than the previous time interval.

[0066] While genetic algorithms possess numerous advantages, including powerful global optimum search capabilities, the duality of the problem domain, parallelism in information processing, robustness in application, and simplicity of operation, the inventors, through extensive practice and research, discovered that standard genetic algorithms suffer from drawbacks such as poor search ability and premature convergence in practical applications. They cannot guarantee rapid convergence and still have room for improvement in the accuracy and efficiency of phase change ejection intelligent control. Therefore, the inventors have improved the standard genetic algorithm to enhance its efficiency, accuracy, and stability when applied to phase change ejection control.

[0067] like Figure 3 As shown, the improved genetic algorithm in this embodiment includes the following steps:

[0068] S001. Population initialization; Set the generation counter g = 0, set the maximum generation number G, and randomly generate Np individuals as the initial population f(0) based on the upper and lower limits of the parameters to be optimized; the parameters to be optimized are the number of phase transitions or the time interval; in specific applications, the number of individuals Np can be randomly generated using a random number function based on the upper and lower limits of the parameters to be optimized, such as running in MATLAB software:

[0069] For i = 1:1:NP

[0070] f(i) = rand() * (upper bound - lower bound) + lower bound

[0071] end

[0072] Here, rand() represents a random number between 0 and 1. Each loop can obtain one arbitrary number between the lower and upper limits. After Np loops, Np individuals are obtained. This is the conventional operation method for random numbers and is not an innovation of this invention, so it will not be elaborated on here.

[0073] S002, Initial Fitness Evaluation; Calculate the fitness of each individual in the population f(g), with evolutionary generations g = 0, 1, 2, ..., the population when g = 0 is the initial population f(0). This invention uses the calculated deorbit velocity V and the target parameter value V MB The absolute value of the difference between V and V is taken as the fitness value FIT, FIT = |VV MB The fitness is calculated by taking the absolute value of the difference between the calculated deorbit velocity and the target parameter value. A fitness value closer to 0 is better. The target parameter value is V. MB It is related to the design speed of the projectile, and is an empirical value that is set by the user; no limit is set here.

[0074] S003, Selection Operation: Sort individuals in the population f(g) in ascending order of their calculated fitness values, and save the individual with the smallest fitness value as the optimal individual; the standard genetic algorithm does not sort individuals in the population according to their fitness values.

[0075] S004, Crossover operation; Cross the best individual with other even-numbered individuals after ascending sorting. Each crossover produces two new individuals, forming a new population f'(g); Unlike the crossover operation step in the standard genetic algorithm, this step crosses the best individual with other even-numbered individuals after sorting, so that even-numbered individuals become the optimal solution. Removing samples from even-numbered individuals can focus the range more.

[0076] S005, Mutation Operation; Perform multi-point mutation on the new population f'(g) to generate the offspring population ff(g), and the parent population of ff(g) is f(g); This step is the same as the mutation operation step in the standard genetic algorithm;

[0077] S006. Secondary fitness evaluation: Calculate the fitness of each individual in the offspring population ff(g). The fitness calculation is the same as the fitness calculation in step S002. Sort the individuals in the population ff(g) in ascending order of the fitness values ​​calculated for each individual.

[0078] S007. Cross-generational optimization: merge the offspring population ff(g) after reordering in step S006 with the initial population f(0), sort the merged individuals in ascending order according to their fitness values, take the first Np individuals as the next generation population f(g+1), and update the generation number g to g+1.

[0079] S008. If g > G (termination criterion), then the individual with the smallest fitness value in all generations is output as the optimal individual, and the calculation is terminated; otherwise, return to step S002. When outputting the optimal individual, the fitness values ​​of the number of phase transitions or time intervals in all generations are sorted in ascending order, and the individual ranked first is output.

[0080] This embodiment improves the standard genetic algorithm by modifying parameter design strategies, fitness encoding mechanisms, selection strategies, crossover operators, secondary evaluation, and parent-child generational optimization. In the phase transition number optimization step, a dynamic parameter adjustment strategy is adopted to dynamically adjust the upper and lower limits, narrowing the search range and improving search accuracy. When optimizing the number of UI phase transitions and time intervals, fitness is determined by the deviation value (the absolute value of the difference between the calculated objective function value and the objective parameter value), with the smallest value being the optimal solution for that iteration, improving the accuracy and efficiency of the determination. Selection and crossover operations are performed using the optimal parent solution; ascending order selection is more efficient and accurate, and the selected optimal solution is crossed with even-numbered positions during crossover to further focus the range and improve the algorithm's convergence. The addition of secondary evaluation evaluates the offspring, calculating their fitness values, and the addition of generational optimization allows for the ranking of parent and offspring fitness values ​​together to select the optimal solution, avoiding getting trapped in local dynamics and improving the algorithm's efficiency, accuracy, and stability.

[0081] The phase change ejection intelligent control method based on genetic algorithm of the present invention can realize multi-objective intelligent optimization control according to different ejection speed requirements and parameter priority, hierarchical and step-by-step optimization control, and give the optimal solution of optimal phase change mass, number of phase changes, and interval time parameters, thereby improving the accuracy and efficiency of phase change ejection intelligent control.

[0082] To verify the effectiveness of the method of this invention, a simulation comparison was conducted between the control method using a standard genetic algorithm and the control method using an improved genetic algorithm. The simulation tests were performed using MATLAB software. The basic parameters of the phase change ejection system are shown in Table 1.

[0083] Table 1 Basic Parameters Table

[0084]

[0085] The number of phase transitions and the interval time are optimized based on the target deorbit velocity. The improved genetic algorithm has a population size of Np = 200, a population size of chromosomes and a gene length of D = 3, a crossover rate of Pc of 0.8, a mutation rate of Pm of 0.1, and a maximum number of generations G of 100.

[0086] I. Convergence Verification

[0087] When the target deorbit velocity is 30 m / s, the fitness obtained by the intelligent control method based on the improved genetic algorithm changes with the number of generations as follows: Figure 4 As shown. From Figure 4 It can be seen that after iterative optimization, the error between the calculated derailment speed and the target derailment speed approaches zero.

[0088] Figure 5This describes the variation of fitness with the number of generations obtained using an intelligent control method based on an improved genetic algorithm when the target deorbit velocity is 20–80 m / s. Figure 5 It can be seen that the fitness evolution curves converge rapidly at different speeds. When the number of evolutions exceeds 78, the fitness approaches 0, which verifies the rapid convergence of the intelligent control method based on the improved genetic algorithm.

[0089] The intelligent control method using the standard genetic algorithm requires more than 160 evolutions to reduce the error to 0.1%, which is not as fast as the intelligent control method based on the improved genetic algorithm.

[0090] II. Accuracy Verification

[0091] The error distribution patterns between the deorbit velocities and the target deorbit velocities calculated by the intelligent control method based on the improved genetic algorithm under different target deorbit velocities are statistically analyzed as follows: Figure 6 As shown. From Figure 6 It can be seen that the maximum error of the intelligent control for different target de-orbit speeds does not exceed 0.1%, which verifies the accuracy of the intelligent control method.

[0092] III. Stability Verification

[0093] Taking a target deorbit speed of 30 m / s as an example, an intelligent control method based on an improved genetic algorithm was used to conduct intelligent optimization control simulations for multiple phase change ejection. When the mass specification is 3 kg and the ejection is controlled in three stages, the optimal number of phase changes obtained from multiple simulations is [1,1,1], and the optimal time interval distribution is as follows. Figure 7 As shown.

[0094] Based on the mathematical model of the internal ballistics of phase change ejection, and through multiple intelligent control calculations, the velocity curve of the ejected object is obtained, for example... Figure 8 As shown. From Figure 8 It can be seen that, after multiple simulations and optimizations, the optimal number of phase transitions recommended by the intelligent control method based on the improved genetic algorithm exhibits high stability, with small deviations in the phase transition intervals. The velocity curve of the launched object is smooth and regular, reflecting a smaller overload on the launched object and improving the efficiency of phase transition ejection. The errors between the deorbit velocity and the target velocity calculated in multiple simulations do not exceed 0.013%, which not only verifies the accuracy of the intelligent control method based on the improved genetic algorithm but also its stability.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A genetic algorithm-based phase change elastic intelligent control method, characterized in that, The method comprises the following steps: Step S1, obtaining initial parameters of the phase-change ejection system and an orbit departure speed of the ejected object, the initial parameters including initial mass specifications, initial phase-change number, and initial time interval, and taking the orbit departure speed of the ejected object as a target orbit departure speed; Step S2, mass specification determination; calculating an orbit departure speed range under a current mass specification, if the target orbit departure speed is within the orbit departure speed range, further optimizing the phase-change number; if the target orbit departure speed is less than a minimum orbit departure speed or greater than a maximum orbit departure speed, ending the parameter optimization; Step S3, phase-change number optimization; judging whether a ratio of the target orbit departure speed to the minimum orbit departure speed is less than a set threshold value, if yes, saving the initial phase-change number as a phase-change number optimal value, and then further optimizing the time interval, otherwise, optimizing the phase-change number based on a genetic algorithm to obtain the phase-change number optimal value; Step S4, time interval optimization; optimizing the time interval based on the genetic algorithm to obtain a time interval optimal value; Step S5, outputting the phase-change number optimal value and the time interval optimal value, and performing phase-change ejection control according to the phase-change number optimal value and the time interval optimal value.

2. The phase-change ejection intelligent control method of claim 1, wherein: In the step S3, the step of optimizing the phase-change number based on the genetic algorithm is as follows: S01, randomly generating M individuals as an initial population P1(0) of the phase-change number; S02, calculating an adaptability of each individual in a population P1(t) of the phase-change number, t being an evolution generation number, t=0, 1, 2, …, the population at t=0 being the initial population P1(0) of the phase-change number; S03, sequentially performing selection operation, crossover operation and mutation operation on the population P1(t) after adaptability evaluation to obtain a next generation population P1(t+1), and updating the evolution generation number t to t+1; S04, if the evolution generation number t is less than a maximum evolution generation number T, returning to S02, otherwise outputting an individual corresponding to a maximum adaptability value in an evolution process as the phase-change number optimal value.

3. The method of claim 1, wherein: In the step S4, the step of optimizing the time interval based on the genetic algorithm is as follows: S01, randomly generating M individuals as an initial population P2(0) of the time interval, the initial population P2(0) of the time interval being such that a later time is greater than an earlier time; S02, calculating an adaptability of each individual in a population P2(t) of the time interval; S03, sequentially performing selection operation, crossover operation and mutation operation on the population P2(t) after adaptability evaluation to obtain a next generation population P2(t+1), and updating the evolution generation number t to t+1; S04, if the evolution generation number t is less than a maximum evolution generation number T, returning to S02, otherwise outputting an individual corresponding to a maximum adaptability value in an evolution process as the time interval optimal value.

4. The phase-change ejection intelligent control method of claim 1, wherein: The improved genetic algorithm is used to optimize the phase-change number and the time interval, and the steps are as follows: S001, population initialization; randomly generating Np individuals as an initial population f(0) according to upper and lower limits of a to-be-optimized parameter; the to-be-optimized parameter being the phase-change number or the time interval; S002, first fitness evaluation; calculate the fitness value of each individual in the population f(g), evolution generation g=0, 1, 2, …, g=0 is the initial population f(0); S003, selection operation; sort the fitness values of each individual in the population f(g) in ascending order, and save the individual with the minimum fitness value as the optimal individual; S004, crossover operation; cross the optimal individual with the even-numbered individuals in the ascending order, and generate two new individuals each time to form a new population f'(g); S005, mutation operation; multiple-point mutation is performed on f'(g) to generate a child population ff(g); S006, second fitness evaluation; calculate the fitness value of each individual in the child population ff(g), and sort the fitness values of each individual in the population ff(g) in ascending order; S007, cross-generation selection; combine the reordered child population ff(g) and the initial population f(0), and sort the combined individuals in ascending order according to the fitness values; take the first Np individuals as the next generation population f(g+1), and update the evolution generation g to g+1; S008, if g>G, output the individual with the minimum fitness value in the history as the optimal individual and terminate the calculation, otherwise return to step S002.

5. The phase-change ejection intelligent control method according to claim 2 or 3 or 4, characterized in that: The fitness value is the absolute value of the difference between the off-orbit speed calculation value and the target parameter value.

Images