Robust design optimization method and system for spacecraft thermal control system

Abstract

This invention belongs to the field of spacecraft thermal design and provides a robust design optimization method and system for spacecraft thermal control systems. The method includes: transforming the unknown distribution of multi-source uncertain coupled parameters into a range scalar of coupled coexisting uncertain parameters; performing iterative inversion optimization on the range scalar of coupled coexisting uncertain parameters based on a cognitive uncertainty objective function to correct the cognitive uncertainty of the parameters; performing iterative inversion optimization on the inherent random uncertainty after correcting the cognitive uncertainty of the parameters based on a random uncertainty objective function to correct the random uncertainty of the parameters; and determining the robust optimization result of the thermal design parameter order based on the two-stage uncertainty correction results. This invention can solve the problem of mixed coexistence of two types of uncertainty to improve the robustness of thermal control system design optimization.

Description

Technical Field

[0001] This invention belongs to the field of spacecraft thermal design technology, specifically relating to a robust design optimization method and system for spacecraft thermal control systems. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] In the parameter design of spacecraft thermal control systems, multi-source uncertainties are generally classified into two categories: cognitive uncertainty and random uncertainty. The former stems from errors in heat transfer numerical modeling and can be corrected using thermal balance test data, while the latter originates from inherent random fluctuations in the real physical environment and requires robust optimization to reduce temperature response sensitivity. However, in practical engineering, these two types of uncertainties often coexist and are coupled, and their statistical models are unknown, posing a severe challenge to parameter optimization.

[0004] Traditional thermal control parameter optimization often relies on single probabilistic models or single-factor sensitivity analysis, which struggles to effectively handle multi-source coupled uncertainties involving unknown distributions and the coexistence of cognitive and accidental factors. When prior information about parameters is lacking, traditional probabilistic model assumptions are prone to introducing significant biases, leading to unrobust optimization results. While some studies employ maximum likelihood estimation or non-probabilistic methods such as evidence theory and interval theory, these often treat coupled uncertainties uniformly, tending towards conservatism due to a lack of prior knowledge, making it difficult to achieve cost-effective designs. Secondly, spacecraft thermal system parameters have high dimensionality, and direct optimization based on finite element models consumes substantial computational resources. While existing surrogate models partially alleviate the computational burden, they fail to specifically address the need for double-layered nested modeling of parameter interval scalars, resulting in limited inversion optimization efficiency. Furthermore, most methods fail to fundamentally distinguish between cognitive and accidental uncertainties. The former can be gradually reduced through data correction, while the latter is irremovable. This confusion blurs the optimization objective, making it impossible to establish a closed-loop process of uncertainty identification, parameter correction, and robustness verification, thus failing to simultaneously meet the design requirements of accuracy and reliability. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a robust design optimization method and system for spacecraft thermal control systems. This invention can comprehensively consider the uncertainties of multi-source coupling in spacecraft thermal design, and avoid conservative design results caused by a lack of prior knowledge such as parameter probability distribution types, thus better aligning with the robustness objective of precise thermal management in spacecraft.

[0006] According to some embodiments, the first aspect of the present invention provides a robust design optimization method for a spacecraft thermal control system, employing the following technical solution: Robust design optimization methods for spacecraft thermal control systems include: The unknown distribution of multi-source uncertain coupled coexistence parameters is transformed into an interval scalar of coupled coexistence uncertain parameters; Based on the objective function of cognitive uncertainty, iterative inversion optimization is performed on the range scalar of coupled coexisting uncertainty parameters to complete the correction of cognitive uncertainty of parameters. Based on the objective function of random uncertainty, the inherent random uncertainty after the correction of parameter cognitive uncertainty is iteratively inverted and optimized to complete the correction of parameter random uncertainty; Based on the two-stage uncertainty correction results, the robust optimization results of the thermal design parameter sequence are determined.

[0007] Furthermore, the transformation of the unknown distribution of multi-source uncertain coupled coexistence parameters into an interval scalar of coupled coexistence uncertainty parameters specifically involves: Based on the nonprobabilistic interval theory, the unknown distribution of multi-source uncertain coupling coexistence parameters is transformed into scalars of the midpoint interval and the radius interval. The midpoint interval scalar and the radius interval scalar are used as interval scalars of coupled and coexisting uncertainty parameters.

[0008] Furthermore, the step of iteratively inverting and optimizing the scalar range of coupled coexisting uncertainty parameters based on the cognitive uncertainty objective function to complete the correction of parameter cognitive uncertainty is as follows: For the midpoint interval scalar of the coupled coexisting uncertainty parameters, a Gaussian process regression model is used to model the midpoint Gaussian surrogate model. Based on the objective function of cognitive uncertainty, the midpoint Gaussian surrogate model is used to iteratively invert and optimize the midpoint of the uncertainty interval. For the radius interval scalar of the coupled coexisting uncertainty parameter, a Gaussian process regression model is used to model it, resulting in a radius double-nested Gaussian surrogate model; Based on the objective function of cognitive uncertainty, the radius of the uncertainty interval is iteratively inverted and optimized using a radius-double-nested Gaussian surrogate model to complete the correction of cognitive uncertainty of parameters and obtain the corrected uncertainty parameter interval.

[0009] Furthermore, for the radius interval scalar of the coupled coexisting uncertainty parameter, a Gaussian process regression model is used for modeling, resulting in a radius-based double-nested Gaussian surrogate model, specifically: Based on the uncertainty range of each thermal system parameter, radius sampling is performed on each thermal system parameter to generate a radius sample; Each radius sample is combined with the set of midpoint values ​​optimized by midpoint inversion to obtain the radius sample interval; Within each radius sample interval, uniform sampling is performed and the sample evaluation is performed using the midpoint Gaussian surrogate model to obtain the thermal system temperature response value. The radius of the temperature response interval corresponding to each group of radius samples is obtained by numerical statistics based on the temperature response values ​​of the thermal system. Using the radius interval scalar as input and the temperature response interval radius as output, a radius-based double-nested Gaussian surrogate model is constructed.

[0010] Furthermore, the objective function for cognitive uncertainty is to minimize the difference between the output temperature dynamic multi-feature quantity and the midpoint scalar or radius scalar of the interval of the thermal equilibrium test data.

[0011] Furthermore, the objective function for random uncertainty is to minimize the mean and standard deviation between the output temperature dynamic multi-feature quantity and the thermal design temperature standard.

[0012] According to some embodiments, a second aspect of the present invention provides a robust design optimization system for a spacecraft thermal control system, employing the following technical solution: Robust design optimization systems for spacecraft thermal control systems include: The parameter transformation module is configured to transform the unknown distribution of multi-source uncertain coupled coexisting parameters into a range scalar of coupled coexisting uncertain parameters. The cognitive uncertainty optimization module is configured to perform iterative inversion optimization on the range scalar of coupled coexisting uncertainty parameters based on the cognitive uncertainty objective function, thereby completing the correction of parameter cognitive uncertainty. The random uncertainty optimization module is configured to perform iterative inversion optimization on the inherent random uncertainty after the correction of parameter cognitive uncertainty based on the random uncertainty objective function, thereby completing the correction of parameter random uncertainty. The robustness optimization module is configured to determine the robustness optimization results of the thermal design parameter order based on the two-stage uncertainty correction results.

[0013] According to some embodiments, a third aspect of the present invention provides a computer-readable storage medium.

[0014] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the robust design optimization method for a spacecraft thermal control system as described in the first aspect above.

[0015] According to some embodiments, a fourth aspect of the present invention provides a computer device.

[0016] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps in the robust design optimization method for a spacecraft thermal control system as described in the first aspect above.

[0017] According to some embodiments, a fifth aspect of the present invention provides a computer program product or computer program.

[0018] This invention provides a computer program product or computer program comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps in the robust design optimization method for a spacecraft thermal control system as described in the first aspect above.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention establishes a two-stage sequential robust design optimization method for spacecraft thermal system parameters with multiple sources of uncertainty coexisting. Compared with the current practice of performing one-time co-optimization of multi-source coupling uncertainties of parameters, it can not only take into account the multi-source coupling uncertainty factors of spacecraft thermal design, but also avoid conservative design results caused by the lack of prior knowledge such as parameter probability distribution types, which is more in line with the robustness goal of precise thermal management of spacecraft. Attached Figure Description

[0020] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0021] Figure 1 This is the numerical calculation process for two-stage sequential robustness optimization in this embodiment of the invention; Figure 2 This is a schematic diagram of the results of the non-probabilistic interval method based on interval scalar operations in an embodiment of the present invention; Figure 3 This is a diagram showing the effective reduction of cognitive uncertainty in coupled uncertainty in an embodiment of the present invention; Figure 4 This is a robust optimization process with embedded uncertainty quantification in the embodiments of the present invention; Figure 5 This is the most robust optimization result in terms of output response in this embodiment of the invention. Detailed Implementation

[0022] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0023] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0024] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0025] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0026] As mentioned in the background section, existing technologies still have many shortcomings. First, they lack sufficient handling of uncertainties. Traditional thermal control parameter optimization is often based on single probabilistic models (such as Monte Carlo simulations) or single-factor sensitivity analyses (such as stepwise optimization methods), making it difficult to effectively handle multi-source coupled uncertainties (such as unknown parameter distributions and the coexistence of cognitive and accidental uncertainties). Especially when prior parameter information is lacking, assumptions in traditional probabilistic models (such as normal distribution) may introduce biases, leading to unsustainable optimization results. Second, there are issues with computational efficiency and the curse of dimensionality. Spacecraft thermal system parameters have high dimensionality, and direct optimization using traditional finite element models requires substantial computational resources. While existing surrogate models are used, they do not address the need for double-layer nested modeling of parameter range scalars, limiting the efficiency of inversion optimization. Third, cognitive and accidental uncertainties are not separated. Most methods fail to distinguish between cognitive uncertainty (unknown parameter ranges) and accidental uncertainty (inherent parameter fluctuations), resulting in ambiguous optimization objectives and a lack of a closed-loop process for uncertainty correction. Based on this, the present invention proposes the following embodiments to address the aforementioned shortcomings.

[0027] Example 1 like Figure 1As shown, this embodiment provides a robust design optimization method for a spacecraft thermal control system. This embodiment uses the application of this method to a server as an example for illustration. It is understood that this method can also be applied to terminals, and can also be applied to systems including terminals, servers, and other components, and implemented through interaction between the terminal and the server. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network servers, cloud communication, middleware services, domain name services, CDN security services, and big data and artificial intelligence platforms. The terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and server can be directly or indirectly connected via wired or wireless communication, which is not limited herein. In this embodiment, the method includes the following steps: Step 1: Transform the unknown distribution of multi-source uncertain coupled coexistence parameters into an interval scalar of coupled coexistence uncertain parameters; Step 2: Based on the objective function of cognitive uncertainty, perform iterative inversion optimization on the scalar range of coupled coexisting uncertainty parameters to complete the correction of cognitive uncertainty of parameters; Step 3: Based on the objective function of random uncertainty, perform iterative inversion optimization on the inherent random uncertainty after the correction of parameter cognitive uncertainty to complete the correction of parameter random uncertainty; Step 4: Based on the two-stage uncertainty correction results, determine the robust optimization results of the thermal design parameter sequence.

[0028] This embodiment proposes a sequential decoupling design optimization method for the coexistence of multiple sources of uncertainty. First, the unknown distribution of parameters with coexisting uncertainties is transformed into a large number of intervals represented by the midpoint and radius, and inversion optimization of the cognitive uncertainty of thermal system parameters is performed based on non-probabilistic interval theory. Second, for the inherent random uncertainty of the remaining parameters, the designable parameters of the thermal system are selected, and robust optimization research on thermal design parameters based on probability and statistics theory is carried out, realizing a two-stage sequential decoupling optimization for the coexistence of multiple sources of uncertainty.

[0029] The design optimization of spacecraft thermal control systems is an iterative process involving prototype design, performance testing, and design improvement. The primary method is to conduct extreme-condition thermal analysis based on a spacecraft system-level heat transfer model and optimize the model's design parameters based on the temperature response results. However, spacecraft system-level heat transfer models inevitably possess uncertainties compared to real physics. On one hand, the nodal thermal network method, widely used in spacecraft thermal analysis, simplifies and makes numerous assumptions based on the lumped parameter concept. On the other hand, real spacecraft components inevitably suffer from manufacturing and assembly errors, as well as on-orbit property degradation. According to the classification of uncertainties, the uncertainties arising from the simplifications and assumptions in the aforementioned spacecraft heat transfer model can be categorized as cognitive uncertainties, while the uncertainties caused by physical variables in the real environment can be categorized as accidental uncertainties. These two types of uncertainties together constitute the main source of uncertainty in heat transfer model parameters. Due to the objective coupling of these two types of uncertainties, the thermal control system parameters exhibit random variables with unknown distributions, referred to as parameters with coexisting uncertainties.

[0030] In step 1, the transformation of the unknown distribution of multi-source uncertain coupled coexistence parameters into an interval scalar of coupled coexistence uncertainty parameters specifically involves: Based on the nonprobabilistic interval theory, the unknown distribution of multi-source uncertain coupling coexistence parameters is transformed into scalars of the midpoint interval and the radius interval. Specifically, for the first A coupling coexistence uncertainty parameter When the lower and upper limits of the coupled coexistence uncertainty parameters are determined based on prior experience, experimental results, or engineering boundaries, respectively... and At this time, based on the non-probabilistic interval theory, the coupled coexisting uncertainty parameters can be transformed into scalars at the midpoint of the interval. Scalar of interval radius ,Right now (1); (2); And have (3); For including For a thermal system with coupled and coexisting uncertain parameters, a midpoint vector can be further constructed. and radius vector , respectively represented as: (4); (5); The midpoint vector and radius vector Together they constitute a range of scalar parameters for coupled and coexisting uncertainties, which are used for subsequent cognitive uncertainty inversion optimization and accidental uncertainty robustness optimization.

[0031] The distribution of uncertain parameters in coupled coexistence is unknown, but a rough range of random variations can be determined based on prior experience. This range can be appropriately expanded, thus clarifying the upper and lower limits of the range, such as... Figure 2 As shown, the midpoint and radius of the range of values ​​are thus clearly defined. Therefore, the uncertainty parameter of an unknown distribution is transformed into a large interval number represented by two scalars: the midpoint value and the radius value.

[0032] This embodiment transforms the parameters of the unknown distribution into a midpoint-radius scalar, avoiding the limitations of traditional probability distribution assumptions. It is suitable for scenarios with scarce information. By expanding the parameter range and scalarizing it, it is compatible with engineering experience data and enhances the applicability of the model.

[0033] In step 2, based on the objective function of cognitive uncertainty, iterative inversion optimization is performed on the scalar range of coupled coexisting uncertainty parameters to complete the correction of parameter cognitive uncertainty, specifically as follows: For the cognitive uncertainty of parameter coupling and coexistence uncertainty, since it is difficult to accurately know the prior probability distribution type of the parameters, the statistical characteristics of the mean and standard deviation of the temperature data of the thermal equilibrium test are transformed into interval scalars of the midpoint and radius based on the non-probability interval theory. According to the objective function of formula (6), combined with the genetic algorithm, the optimization inverse problem of the interval scalar of the coupled and coexisting uncertainty parameters is constructed, such as... Figure 3 As shown.

[0034] Since the scalar at the midpoint of the parameter interval and the corresponding output quantity are temperature values, they can be determined directly by using the Kriging surrogate model of the spacecraft heat transfer model combined with the genetic algorithm for inversion optimization. Since the parameter interval radius scalar and the corresponding output quantity are not temperature values, the Kriging surrogate model of the spacecraft heat transfer model cannot be directly used for inversion optimization. This invention continues to use this surrogate model as a solver and combines it with Monte Carlo statistical methods to obtain the output interval radius response value based on the input parameter interval radius sample (sampling calculation is performed within the parameter interval radius range and the output statistical standard deviation is converted into the radius response value). Then, a double-layer nested Kriging model of the interval radius scalar is established to complete the inversion optimization determination of the interval radius scalar and obtain the parameter interval after sufficient correction of cognitive uncertainty.

[0035] Understandably, firstly, the Kriging surrogate model has the advantage of high efficiency in optimizing parameters of spacecraft thermal systems with many parameters, such as those here, without causing the curse of dimensionality. In addition, it can also be achieved using a neural network model.

[0036] Step 2.1: For the midpoint interval scalar of the coupled coexisting uncertainty parameters, a Gaussian process regression model is used to model the midpoint Gaussian surrogate model, specifically: After determining the uncertainty range of each thermal system parameter based on prior experience, the midpoint Kriging surrogate model of the midpoint value of the thermal system parameter range is first modeled. Assume the spacecraft thermal system includes There are several cognitive uncertainty parameters, and the midpoint vector is... Midpoint sampling is performed, and Latin hypercube sampling is used to generate a midpoint sample set. The number of midpoint samples Covering the midpoint parameter space, for each midpoint sample Finite-difference thermal analysis was performed using the original spacecraft numerical heat transfer model to obtain the thermal system temperature response value for each sample, with the midpoint sample as an example. Using temperature response as the input and temperature response as the output, a midpoint Kriging surrogate model, i.e. a midpoint Gaussian surrogate model, is constructed based on the Gaussian process principle to construct the midpoint value of the thermal system parameters corresponding to the temperature response of the spacecraft thermal system. Midpoint optimization: A midpoint Kriging surrogate model is constructed based on Latin hypercube sampling, and combined with genetic algorithm inversion optimization to quickly determine the midpoint of the parameter interval. This process solves the curse of dimensionality problem in high-dimensional parameter optimization and significantly improves efficiency compared to traditional finite element models.

[0037] Step 2.2: Based on the cognitive uncertainty objective function, the midpoint Gaussian surrogate model is used to iteratively invert and optimize the midpoint of the uncertainty interval, specifically as follows: First, a cognitive uncertainty objective function is constructed, which minimizes the difference between the output temperature dynamic multi-feature quantity and the midpoint scalar or radius scalar of the interval of the thermal equilibrium test data. Specifically, the objective of cognitive uncertainty optimization is to minimize the dynamic multi-characteristics of the output temperature. The difference between the data and the thermal equilibrium test data is transformed into minimizing the difference between the midpoint scalar or radius scalar of the interval between the two data based on the non-probability interval theory. The objective function for recognizing uncertainty is as follows: (6); in, This represents a scalar matrix representing the midpoint or radius of a multidimensional uncertainty parameter interval. Represents multiple dynamic temperature characteristics The difference between the scalar value at the midpoint or radius of the thermal equilibrium test data interval. This represents the prediction function of the surrogate model; here, the midpoint Gaussian surrogate model is used. Represents cognitive constraints. and This indicates the upper and lower limits of the range of values ​​for a scalar parameter interval.

[0038] The midpoints of the uncertainty intervals of the spacecraft's thermal system are optimized sequentially using the same method: a combination of the midpoint Kriging surrogate model and a genetic algorithm. The program steps are described below: Genetic algorithm encoding is performed, with chromosomes directly representing parameter vectors. Randomly generated within the feasible region of parameters. =50-100 individuals to ensure diversity; Genetic manipulation is performed using a tournament selection strategy to select individuals with high fitness for the mating pool. Simulated binary crossover (SBX) is used to maintain population diversity, and polynomial mutation is performed with probabilistic probability. Disturbance parameter values; The objective function of formula (6) is iteratively optimized at the midpoint of the interval for each individual. The Kriging model is called to predict the temperature response, calculate the fitness, and output the optimization results that satisfy the objective function according to the convergence criteria, that is, to determine the set of midpoint values ​​of all parameters. .

[0039] The convergence criteria are: 1. The rate of change of the optimal fitness is <0.1% for 10 generations; 2. The optimal number of iterations is reached. =200.

[0040] Step 2.3: For the radius interval scalar of the coupled coexisting uncertainty parameter, a Gaussian process regression model is used to model it, resulting in a radius-based double-nested Gaussian surrogate model, specifically: Based on the uncertainty range of each thermal system parameter, radius sampling is performed on each thermal system parameter to generate a radius sample; Each radius sample is combined with the set of midpoint values ​​optimized by midpoint inversion to obtain the radius sample interval; Within each radius sample interval, uniform sampling is performed and the sample evaluation is performed using the midpoint Gaussian surrogate model to obtain the thermal system temperature response value. The radius of the temperature response interval corresponding to each group of radius samples is obtained by numerical statistics based on the temperature response values ​​of the thermal system. Using the radius interval scalar as input and the temperature response interval radius as output, a radius-double-nested Gaussian surrogate model is constructed. Kriging surrogate modeling of the radius values ​​of thermal system parameters is used. Similarly, the spacecraft thermal system contains n cognitive uncertainty parameters, and the radius vector is... Radius sampling is performed, and a radius sample set is generated using Latin hypercube sampling. The number of radius samples The radius value sampling does not exceed the radius of the maximum uncertainty interval of this parameter, for each radius sample , and the already optimized set of midpoint values Together forming the radius sample interval Within each radius sample interval, uniform sampling is performed, and the thermal system temperature response value is obtained by using the Kriging surrogate model of the midpoint value. Numerical statistics are then performed to obtain the temperature response value of each radius sample. The corresponding temperature response range radius is also represented by the radius sample. Using the temperature response range radius as input and the radius of the temperature response range as output, a radius sample is established. The Kriging surrogate model corresponding to the radius of the temperature response interval is called the radius double-nested Gaussian surrogate model because the midpoint value Kriging surrogate model is called an additional time during the establishment process.

[0041] Step 2.4: Based on the cognitive uncertainty objective function, the radius of the uncertainty interval is iteratively inverted and optimized using a radius-double-nested Gaussian surrogate model to complete the correction of the cognitive uncertainty of the parameters and obtain the corrected uncertainty parameter interval.

[0042] The radius of the uncertainty interval of the spacecraft's thermal system is optimized sequentially. A combination of a double-nested Kriging surrogate model and a genetic algorithm is used to achieve this. The program steps are described below: Genetic algorithm encoding is performed, with chromosomes directly representing parameter vectors. Randomly generated within the feasible region of parameters. =50-100 individuals to ensure diversity; Genetic manipulation is performed using a tournament selection strategy to select individuals with high fitness for the mating pool. Simulated binary crossover (SBX) is used to maintain population diversity, and polynomial mutation is performed with probabilistic probability. Disturbance parameter values; The objective function of formula (6) is iteratively optimized for the interval radius for each individual x. j The Kriging model is called to predict the temperature response, calculate the fitness, and output the optimization results that satisfy the objective function according to the convergence criteria. It can be understood that when solving the objective function with cognitive uncertainty, the surrogate model prediction function adopts a radius-double-nested Gaussian surrogate model.

[0043] The convergence criteria are: 1. The rate of change of the optimal fitness is <0.1% for 10 generations; 2. The optimal number of iterations G is reached. max =200, thus obtaining the corrected uncertainty parameter range.

[0044] This embodiment, for the radius scalar, employs a combination of Monte Carlo sampling and a surrogate model to establish a mapping between the input radius and the output response radius. This method overcomes the limitations of traditional single surrogate models, achieving efficient interval radius inversion through nested calls to the midpoint model, avoiding the computational burden of directly calling the original heat transfer model.

[0045] In step 3, based on the objective function of random uncertainty, iterative inversion optimization is performed on the inherent random uncertainty after the correction of parameter cognitive uncertainty to complete the correction of parameter random uncertainty. Specifically: First, we construct a random uncertainty objective function, which minimizes the mean and standard deviation between the output temperature dynamic multi-feature quantity and the thermal design temperature standard.

[0046] Specifically, the optimization objective of inherent random uncertainty is to optimize the dynamic multi-characteristics of the output temperature. While aiming to closely approximate the thermal design temperature standard, it also strives for greater robustness (i.e., less sensitive to random uncertainties in parameter response). Based on probability and statistics theory, this is transformed into minimizing the mean and standard deviation of both. The objective function for random uncertainty is as follows: (7); in, Represents a multidimensional uncertainty parameter matrix. Represents the output temperature dynamic multi-characteristic quantity The difference between the statistical mean and the standard thermal design temperature. Indicates output Statistical standard deviation This represents the prediction function of the surrogate model. This represents a random constraint. and This represents the upper and lower limits of the range of values ​​for the multidimensional uncertainty parameter matrix.

[0047] For the inherent random uncertainty after the parameter cognitive uncertainty is fully corrected, it usually exhibits a normal distribution within the confidence interval. After confirming the designable parameters and their design feasible regions in the uncertain parameters of the thermal system according to actual requirements, the parameter uncertainty quantification method based on the surrogate model combined with Monte Carlo simulation established in this project is embedded into each step of the optimization iteration of the genetic algorithm to obtain the forward modeling results of its output response statistical characteristics. Then, a robust optimization method for the random uncertainty parameters is established according to the objective function of formula (7). Finally, the above optimization steps are compiled into a numerical calculation program to complete the two-stage sequential robust optimization of the parameters of the spacecraft thermal control system with uncertainties coupled together.

[0048] This embodiment integrates Monte Carlo sampling into the genetic algorithm, sampling the corrected parameter range using a probability distribution (such as a normal distribution), and quickly predicts the statistical characteristics of the temperature response through a surrogate model. This method combines the advantages of both non-probabilistic and probabilistic models. By minimizing the combined deviation of the mean and variance of the temperature response (the objective function of formula (7),) the stability of the thermal control system under parameter fluctuations is ensured. Compared to simply minimizing the temperature error, this method is more suitable for robustness requirements under complex operating conditions.

[0049] After performing cognitive uncertainty inversion optimization, the uncertainty interval of the thermal system parameters is determined. Affected by the remaining inherent random uncertainty, it is generally assumed that the parameters are normally distributed within their uncertainty interval. Next, the robustness optimization of random uncertainty is performed with formula (7) as the objective function. The overall process is similar to the cognitive uncertainty inversion process. The only difference is that this inversion optimization needs to perform the forward quantification process of the uncertainty of the embedded parameters. In the genetic algorithm evaluation stage, Monte Carlo sampling is performed on each individual to generate M=500 groups of perturbation samples within the parameter uncertainty range. The temperature perturbation response value is quickly predicted by the Kriging model, and the statistical characteristic values ​​of temperature (mean, variance, probability density function, etc.) are obtained. The robustness optimization process is as follows: Figure 4 and Figure 5 As shown.

[0050] The specific steps are as follows: First, based on step 2, the parameter range after cognitive uncertainty correction is obtained, and the designable parameters and their design feasible regions participating in robust optimization are determined. Secondly, in each iteration of the genetic algorithm, for each candidate design individual Monte Carlo sampling is performed within the corrected parameter range according to a preset probability distribution to obtain... Group of perturbation samples; Next, the perturbation samples from each group are input into the surrogate model to calculate the corresponding temperature dynamic multi-features, and candidate design individuals are obtained based on these statistics. Corresponding output mean and standard deviation ; Then, and Substitute into formula (7) to calculate the fitness value; finally, after selection, crossover, mutation and convergence determination, output the robust optimization result of the thermal design parameters that minimizes formula (7).

[0051] Step 4 specifically includes: The design optimization of a spacecraft thermal control system is an iterative process involving prototype design, performance testing, and design improvement. The optimization process mentioned in this embodiment refers to one cycle of this iterative optimization, and each cycle includes a thermal balance test. The purpose of optimizing the spacecraft thermal control system is to control the temperature response of specific components within a acceptable fluctuation range. The thermal balance test involves placing the spacecraft thermal control system test sample in a vacuum environment simulation chamber to obtain temperature data from temperature sensors at various locations under simulated space conditions.

[0052] Understandably, based on the composition of the spacecraft's thermal control system, and according to the principles of heat transfer, the material structure of the thermal system components—primarily passive thermal control hardware (including heat sinks, passive heat pipes, insulation materials, radiation coatings, etc.) and supplemented by active thermal control hardware (including pump-driven two-phase circuits, cryogenic refrigerators, electric heaters, etc.)—is converted into the basic model parameters of a lumped-parameter thermal network numerical heat transfer model. These parameters mainly consist of the specific heat capacity of the numerous thermal nodes themselves or between them, contact thermal resistance, thermal conductivity, absorptivity, emissivity, etc. However, due to constraints such as overall satellite power consumption, mass, and lifespan during the spacecraft's overall design process, some of these parameters cannot be thermally designed and are therefore considered undesignable parameters; the others are designable parameters.

[0053] This embodiment embeds the optimization process into a loop of spacecraft prototype design, thermal equilibrium testing (such as vacuum environment simulation), and improvement. It dynamically corrects surrogate model parameters using experimental data to enhance the engineering applicability of the optimization results. The thermal equilibrium test verifies whether temperature limits are met under various operating conditions; if not, the process returns to the optimization loop to ensure the global optimality of the design.

[0054] This embodiment combines the Kriging proxy model with a genetic algorithm, reducing the number of original model calls by more than 90% and improving optimization speed by 3-5 times compared to traditional methods. The double-nested model design avoids redundant modeling and reduces computational resource consumption by more than 50%.

[0055] After multi-source uncertainty decoupling correction, the temperature response fluctuation range of this embodiment is reduced by 30%-50%. The pass rate of thermal equilibrium test is increased to over 95% (compared to about 70%-80% for traditional methods).

[0056] This embodiment is compatible with scenarios where parameter distribution is unknown and is suitable for complex missions such as deep space exploration. It can be extended to other aerospace subsystems or fluid-structure interaction problems in industrial fields.

[0057] This invention significantly improves the robustness and design efficiency of spacecraft thermal control systems through multi-stage uncertainty handling, integration of efficient surrogate models and optimization algorithms, and an experiment-design closed-loop mechanism, providing reliable technical support for complex space missions.

[0058] Example 2 This embodiment provides a robust design optimization system for spacecraft thermal control systems, including: The parameter transformation module is configured to transform the unknown distribution of multi-source uncertain coupled coexisting parameters into a range scalar of coupled coexisting uncertain parameters. The cognitive uncertainty optimization module is configured to perform iterative inversion optimization on the range scalar of coupled coexisting uncertainty parameters based on the cognitive uncertainty objective function, thereby completing the correction of parameter cognitive uncertainty. The random uncertainty optimization module is configured to perform iterative inversion optimization on the inherent random uncertainty after the correction of parameter cognitive uncertainty based on the random uncertainty objective function, thereby completing the correction of parameter random uncertainty. The robustness optimization module is configured to determine the robustness optimization results of the thermal design parameter order based on the two-stage uncertainty correction results.

[0059] The examples and application scenarios implemented by the above modules and corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1 above. It should be noted that the above modules, as part of the system, can be executed in a computer system such as a set of computer-executable instructions.

[0060] The descriptions of each embodiment in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0061] The proposed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and the division of modules described above is only a logical functional division. In actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed.

[0062] Example 3 This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the robust design optimization method for a spacecraft thermal control system as described in Embodiment 1 above.

[0063] Example 4 This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the robust design optimization method for a spacecraft thermal control system as described in Embodiment 1 above.

[0064] Example 5 This embodiment provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps in the robust design optimization method for a spacecraft thermal control system described in Embodiment 1.

[0065] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0066] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0067] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0068] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0069] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0070] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A robust design optimization method for spacecraft thermal control systems, characterized by, include: The unknown distribution of multi-source uncertain coupled coexistence parameters is transformed into an interval scalar of coupled coexistence uncertain parameters; Based on the objective function of cognitive uncertainty, iterative inversion optimization is performed on the range scalar of coupled coexisting uncertainty parameters to complete the correction of cognitive uncertainty of parameters. Based on the objective function of random uncertainty, the inherent random uncertainty after the correction of parameter cognitive uncertainty is iteratively inverted and optimized to complete the correction of parameter random uncertainty; Based on the two-stage uncertainty correction results, the robust optimization results of the thermal design parameter sequence are determined.

2. The robust design optimization method for a spacecraft thermal control system of claim 1, wherein, The process of transforming the unknown distribution of multi-source uncertain coupled coexistence parameters into an interval scalar of coupled coexistence uncertain parameters specifically involves: Based on the nonprobabilistic interval theory, the unknown distribution of multi-source uncertain coupling coexistence parameters is transformed into scalars of the midpoint interval and the radius interval. The midpoint interval scalar and the radius interval scalar are used as interval scalars of coupled and coexisting uncertainty parameters.

3. The method for robust design optimization of a spacecraft thermal control system of claim 1, wherein, The objective function based on cognitive uncertainty involves iterative inversion optimization of the scalar range of coupled coexisting uncertainty parameters to complete the correction of cognitive uncertainty in the parameters. Specifically: For the midpoint interval scalar of the coupled coexisting uncertainty parameters, a Gaussian process regression model is used to model the midpoint Gaussian surrogate model. Based on the objective function of cognitive uncertainty, the midpoint Gaussian surrogate model is used to iteratively invert and optimize the midpoint of the uncertainty interval. For the radius interval scalar of the coupled coexisting uncertainty parameter, a Gaussian process regression model is used to model it, resulting in a radius double-nested Gaussian surrogate model; Based on the objective function of cognitive uncertainty, the radius of the uncertainty interval is iteratively inverted and optimized using a radius-double-nested Gaussian surrogate model to complete the correction of cognitive uncertainty of parameters and obtain the corrected uncertainty parameter interval.

4. The method for robust design optimization of a spacecraft thermal control system of claim 3, wherein, The radius interval scalar for the coupled coexisting uncertainty parameter is modeled using a Gaussian process regression model, resulting in a radius-based double-nested Gaussian surrogate model, specifically: Based on the uncertainty range of each thermal system parameter, radius sampling is performed on each thermal system parameter to generate a radius sample; Each radius sample is combined with the set of midpoint values ​​optimized by midpoint inversion to obtain the radius sample interval; Within each radius sample interval, uniform sampling is performed and the sample evaluation is performed using the midpoint Gaussian surrogate model to obtain the thermal system temperature response value. The radius of the temperature response interval corresponding to each group of radius samples is obtained by numerical statistics based on the temperature response values ​​of the thermal system. Using the radius interval scalar as input and the temperature response interval radius as output, a radius-based double-nested Gaussian surrogate model is constructed.

5. The method for robust design optimization of a spacecraft thermal control system of claim 1, wherein, The objective function for cognitive uncertainty is to minimize the difference between the output temperature dynamic multi-feature quantity and the midpoint scalar or radius scalar of the interval of the thermal equilibrium test data.

6. The method for robust design optimization of a spacecraft thermal control system of claim 1, wherein, The objective function for random uncertainty is to minimize the mean and standard deviation between the output temperature dynamic multi-feature quantity and the thermal design temperature standard.

7. A robust design optimization system for a spacecraft thermal control system, characterized by, include: The parameter transformation module is configured to transform the unknown distribution of multi-source uncertain coupled coexisting parameters into a range scalar of coupled coexisting uncertain parameters. The cognitive uncertainty optimization module is configured to perform iterative inversion optimization on the range scalar of coupled coexisting uncertainty parameters based on the cognitive uncertainty objective function, thereby completing the correction of parameter cognitive uncertainty. The random uncertainty optimization module is configured to perform iterative inversion optimization on the inherent random uncertainty after the correction of parameter cognitive uncertainty based on the random uncertainty objective function, thereby completing the correction of parameter random uncertainty. The robustness optimization module is configured to determine the robustness optimization results of the thermal design parameter order based on the two-stage uncertainty correction results.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the steps in the robust design optimization method for a spacecraft thermal control system as described in any one of claims 1-6.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the robust design optimization method for a spacecraft thermal control system as described in any one of claims 1-6.

10. A computer program product, characterised in that, The computer program product includes a computer program that, when executed by a processor, implements the steps in the robust design optimization method for a spacecraft thermal control system as described in any one of claims 1-6.

Images