A space solid rocket engine system reliability analysis method

Abstract

The application discloses a spaceflight solid rocket engine system reliability analysis method and relates to the technical field of solid rocket engines. The method adopts a Kriging-assisted Monte Carlo method to calculate the system reliability of the solid rocket engine. If the Kriging model does not meet the convergence criterion, the response values of the candidate sample set are predicted by using the Kriging model, the probability that each sample in the positive sample prediction set and the negative sample prediction set is wrongly predicted is calculated, and the probability is defined as a first probability and a second probability respectively; the best candidate sample and the corresponding response value are determined according to the probability, and then the Kriging model is sequentially updated. After the Kriging model meets the convergence criterion, the to-be-analyzed parameters related to the performance of the solid rocket engine are input into the converged Kriging model, and the time and thrust values predicted by the model are obtained. Finally, the Monte Carlo method is adopted to obtain the system reliability of the solid rocket engine.

Description

Technical Field

[0001] This application relates to the field of aerospace solid rocket engine technology, and in particular to a reliability analysis method for aerospace solid rocket engine systems. Background Technology

[0002] Solid rocket motors are chemical rocket propulsion devices that use solid propellants. They have many advantages, such as simple structure, ease of use, ability to maintain combat readiness for extended periods, and high mass-to-weight ratio. They are widely used in military fields such as missile weapons, launch vehicles, and spacecraft.

[0003] Operating time and thrust are crucial performance indicators for solid rocket motors, and their reliability significantly impacts their operational reliability. However, due to multiple uncertainties in design, manufacturing, and service, the actual operating time and thrust performance parameters deviate considerably from the design values, ultimately leading to a sharp increase in the probability of performance failures in solid rocket motors.

[0004] Once a failure occurs, it can easily lead to accidents during ground testing and flight. Therefore, considering the impact of uncertainties on operating time and thrust, conducting reliability analysis of the performance failure system of solid rocket motors is an effective way to achieve accurate performance reliability assessment of solid rocket motors and provides an important reference for the high-reliability design of domestically produced solid rocket motors. Summary of the Invention

[0005] This application provides a reliability analysis method for aerospace solid rocket motor systems, which can efficiently and accurately analyze the reliability of aerospace solid rocket motor performance.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] Firstly, this application provides a reliability analysis method for aerospace solid rocket motor systems, the method comprising:

[0008] A proxy model is constructed based on the initial sample set. The t-th round candidate sample prediction set is obtained by the Kriging model outputting the candidate sample set in the t-th round. The t-th round candidate sample prediction set can be divided into positive sample prediction set and negative sample prediction set according to the sign of the response value.

[0009] Based on the predicted set of candidate samples in round t, determine whether the accuracy of the Kriging model in round t meets the requirements;

[0010] If the accuracy of the Kriging model in round t does not meet the requirements, calculate the probability that each positive sample in the positive sample prediction set in round t may be a negative sample, and define it as the first probability; calculate the probability that each negative sample in the negative sample prediction set in round t may be a positive sample, and define it as the second probability.

[0011] Based on the first and second probabilities of round t, determine the best candidate sample for round t from the candidate sample set, and calculate the response value corresponding to the best candidate sample;

[0012] Based on the best candidate sample in round t and the response value corresponding to the best candidate sample, the Kriging model in round t is updated to obtain the Kriging model in round t+1, where t is an integer greater than 0;

[0013] If the accuracy of the Kriging model in round t meets the requirements, failure analysis of the solid rocket motor performance can be performed based on the predicted time and thrust values ​​in round t.

[0014] In the above technical solution, the criteria for determining whether the accuracy meets the requirements based on the predicted set of candidate samples in the t-th round are as follows:

[0015]

[0016] In the formula, Let t be the accuracy threshold of the Kriging model in round t. Let be the number of samples in the negative sample prediction set for round t. for The upper boundary of the confidence interval, for The upper boundary of the confidence interval, Let the number of positive samples predicted as negative samples in the candidate sample prediction set for round t be denoted by . The number of samples that should be negative and the number of samples that are predicted to be positive in the candidate sample prediction set for round t;

[0017] if If the accuracy is less than or equal to the preset accuracy threshold, then the accuracy of the Kriging model in round t meets the requirements; if If the accuracy exceeds the preset accuracy threshold, then the accuracy of the Kriging model in round t does not meet the requirements.

[0018] In the above technical solution, the probability that each positive sample in the positive sample prediction subset of the t-th round may be a negative sample is calculated and defined as the first probability, and the formula is:

[0019]

[0020] In the formula, Let be the probability that a positive sample in the prediction set of the positive samples in the t-th round is a negative sample, and define it as the first probability. Let be the variance of the Kriging model predictions in round t. Let be the response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

[0021] Calculate the probability that each negative sample in the negative sample prediction set of round t could be a positive sample, and define it as the second probability, as follows:

[0022]

[0023] In the formula, Let the probability that a negative sample in the subset of negative samples in round t is a positive sample be defined as the second probability. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

[0024] In the above technical solution, based on the first and second probabilities of round t, the best candidate sample for round t and the corresponding response value of the best candidate sample are determined from the candidate sample set. Specifically:

[0025] The learning function value for round t is determined based on the first and second probabilities of round t.

[0026] Determine the maximum learning function value in round t from the learning function values ​​in round t;

[0027] The sample corresponding to the maximum learning function value in round t in the candidate sample set is determined as the best candidate sample in round t, and the response value corresponding to the best candidate sample is calculated.

[0028] The learning function expression is:

[0029]

[0030]

[0031] In the formula, Let t be the value of the learning function in the t-th round. Let be the first probability in round t. The second probability in round t. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The cumulative distribution function of the normal distribution, For about The cumulative distribution function of the normal distribution, Let be the difference between the first probability and the second probability in round t. Based on the predicted time and thrust in round t, a reliability analysis of the solid rocket motor performance is performed, including:

[0032] If the predicted time value in round t represents the failure of the working time performance of the solid rocket motor, or the predicted thrust value in round t represents the failure of the working thrust performance of the solid rocket motor, then the solid rocket motor is determined to have failed.

[0033] If the predicted time value in round t represents the normal operating time performance of the solid rocket motor and the predicted thrust value in round t represents the normal operating thrust performance of the solid rocket motor, then it is determined that the solid rocket motor has not failed.

[0034] Secondly, this application provides a failure analysis device for the performance of aerospace solid rocket engines, the device comprising:

[0035] The update module is used to update the Kriging model based on the candidate samples in round t and the corresponding response values ​​of the candidate samples, to obtain the Kriging model in round t+1, where t is an integer greater than 0;

[0036] The acquisition module is used to acquire the predicted set of candidate sample response values ​​for the (t+1)th round output by the Kriging model for the candidate sample set, including the positive sample prediction set and the negative sample prediction set;

[0037] The judgment module is used to determine whether the accuracy of the Kriging model in the (t+1)th round meets the requirements based on the candidate sample prediction set in the (t+1)th round.

[0038] The probability module is used to determine the probability that each positive sample in the positive sample prediction set of the (t+1)th round may be a negative sample if the accuracy of the Kriging model in the (t+1)th round does not meet the requirements, and defines it as the first probability; and to determine the probability that each negative sample in the negative sample prediction set of the (t+1)th round may be a positive sample, and defines it as the second probability.

[0039] The filtering module is used to determine the candidate sample for the (t+1)th round from the candidate sample set based on the first probability and the second probability of the (t+1)th round, and to determine the response value corresponding to the candidate sample for the (t+1)th round.

[0040] The prediction module is used to input the parameters to be analyzed of the solid rocket motor into the Kriging model of the (t+1)th round if the accuracy of the Kriging model of the (t+1)th round meets the requirements, so as to obtain the predicted time value and the predicted thrust value of the (t+1)th round output by the Kriging model of the (t+1)th round.

[0041] The analysis module is used to perform failure analysis on the performance of the solid rocket motor based on the predicted time value and the predicted thrust value of the (t+1)th round.

[0042] Thirdly, this application provides a computing device, including a memory and a processor;

[0043] The memory stores one or more computer programs, the one or more computer programs including instructions; when the instructions are executed by the processor, the computing device performs the method as described in any one of the first aspects.

[0044] Fourthly, this application provides a computer-readable storage medium for storing a computer program for performing the method as described in any one of the first aspects.

[0045] As can be seen from the above technical solution, this application has at least the following beneficial effects:

[0046] This application provides a reliability analysis method for aerospace solid rocket motor systems. In this method, if the accuracy of the Kriging model does not meet requirements during training, the Kriging model is updated again using the candidate sample prediction set output by the candidate sample set. Specifically, the candidate sample prediction set includes a positive sample prediction set and a negative sample prediction set. The probability that each positive sample in the positive sample prediction set may be a negative sample is determined and defined as a first probability. The probability that each negative sample in the negative sample prediction set may be a positive sample is determined and defined as a second probability. Based on the first and second probabilities, the optimal candidate sample and its corresponding response value are determined from the candidate sample set. The Kriging model is updated using the optimal candidate sample and response value until the accuracy of the Kriging model meets the requirements. After the accuracy of the Kriging model meets the requirements, the parameters to be analyzed related to the performance of the aerospace solid rocket motor can be input into the Kriging model to obtain the time prediction value and thrust prediction value output by the Kriging model. Based on the time prediction value and thrust prediction value, the reliability analysis of the aerospace solid rocket motor performance is performed. It is evident that this method can analyze the performance failure of solid rocket motors, providing an effective assessment method for the performance reliability of solid rocket motors and offering a valuable reference for the high-reliability design of solid rocket motors.

[0047] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0048] Figure 1 A schematic diagram of a solid rocket motor for aerospace applications provided in this application;

[0049] Figure 2 A flowchart illustrating an analysis method for performance failure of a solid rocket motor in aerospace, provided as an embodiment of this application;

[0050] Figure 3 A schematic diagram of an analysis device for performance failure of a solid rocket motor provided in this application embodiment;

[0051] Figure 4 This is a schematic diagram of a computing device provided in an embodiment of this application. Detailed Implementation

[0052] The terms "first," "second," and "third," etc., used in this application specification and accompanying drawings are used to distinguish different objects, not to limit a specific order.

[0053] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0054] like Figure 1 As shown in the figure, this is a schematic diagram of a solid rocket motor for aerospace provided in an embodiment of this application. From the figure, it can be seen that the solid rocket motor for aerospace includes a dataset of parameters such as canard length 11, canard circumscribed circle diameter 12, aft wing length 13, aft wing circumscribed circle diameter 14, maximum pressure of the inner ballistic-combustion chamber, canard tilt angle 15, aft wing tilt angle 16, wing width 17, reference burning rate of the first-stage propellant, and expansion ratio.

[0055] The design of these parameters affects the performance of solid rocket motors (SPLs), such as their operational time and thrust. Failure to meet these requirements will lead to SPL failure. SPL failure can result in unforeseen accidents; therefore, it is necessary to assess the reliability of SPLs to provide effective guidance for their design and reduce the likelihood of failure.

[0056] This application addresses the problems of reduced reliability of solid rocket motors (SPLMs) due to deviations between their room-temperature operating time and average room-temperature thrust under the influence of multiple uncertainties and their corresponding design values, as well as the lack of reliability assessment methods. A reliability analysis method for SPLM systems is proposed. This method first establishes a cascaded system reliability model for operating time and thrust. Second, initial Kriging models for operating time and thrust performance are established separately. Then, considering the possibility of prediction errors in the Kriging model during the prediction process, an improved learning function is proposed. This learning function is used to select sample points from a candidate sample pool. Based on the selected sample points and their corresponding response values, the initial (or previously updated) Kriging model is updated until the accuracy of the Kriging model meets the requirements. Finally, based on the established cascaded system Kriging model, the Monte Carlo method is used to calculate the performance reliability of the SPLM. Compared to other implementation schemes, this invention significantly improves computational efficiency while achieving accurate assessment of the performance reliability of SPLMs.

[0057] To make the technical solution of this application clearer and easier to understand, the technical solution of this application will be described in detail below with reference to the accompanying drawings, such as... Figure 2 As shown, this figure is a flowchart of a system reliability analysis method for the thrust performance and operating time of a solid rocket motor provided in this application. The method includes:

[0058] S201. Construct the initial sample set and the candidate sample set respectively.

[0059] The sample set includes multiple samples and their corresponding response values. The initial sample set is a small sample library generated by Latin hypercube sampling based on the distribution followed by parameters such as the forewing length of the samples; the candidate sample set is a large sample library generated by Latin hypercube sampling based on the distribution followed by parameters such as the forewing length of the samples.

[0060] In the embodiments of this application, the samples used include: sample canard length, sample canard circumscribed circle diameter, sample aft wing length, sample aft wing circumscribed circle diameter, sample canard tilt angle, sample aft tilt angle, sample wing width, sample primary charge reference burning rate, and sample internal ballistic maximum combustion chamber pressure. Response values ​​include sample time response values ​​and sample thrust response values.

[0061] The sample time response value is obtained by subtracting a time threshold from the corresponding operational time of the sample. If the resulting value is greater than or equal to 0, it indicates that the solid rocket motor corresponding to that sample has not failed in terms of operational time performance. Similarly, the sample thrust response value is obtained by subtracting a thrust threshold from the corresponding operational thrust of the sample. If the resulting value is greater than or equal to 0, it indicates that the solid rocket motor corresponding to that sample feature has not failed in terms of operational thrust performance. Thus, based on the sign of the output result of the Kriging model obtained after training, it is possible to determine whether the performance of the solid rocket motor corresponding to the parameters input to the model is failed or not.

[0062] In this scheme, the sample set can be constructed using Latin hypercube sampling.

[0063] S202. Construct a proxy model based on the initial sample set.

[0064] In this embodiment, the surrogate model can be a Kriging model. The initial Kriging model includes two types: the first type is a Kriging model for working time, used to predict working time; the second type is a Kriging model for working thrust, used to predict working thrust. The construction method and principle of the Kriging model in these two methods are similar. For ease of description, we will take a Kriging model whose output parameters include working time and working thrust as an example. The initial Kriging model is constructed based on the samples in the initial sample set and their corresponding response values.

[0065] S203. Predict the candidate sample set using a surrogate model and classify it based on the sign of the predicted response value.

[0066] After the proxy model for working time and working push is constructed, the candidate sample pool prediction set corresponding to the candidate sample is calculated. According to the positive or negative of the predicted response value, it can be divided into two sets: positive sample set and negative sample set.

[0067] S204. Determine whether the accuracy of the Kriging model in round t meets the requirements.

[0068] Determine whether the accuracy of the Kriging model in round t meets the requirements. If the accuracy of the Kriging model in round t does not meet the requirements, execute S205. If the accuracy of the Kriging model in round t meets the requirements, execute S208.

[0069] Where t is an integer greater than 0, when t=2, the Kriging model of the tth round is the Kriging model of the 2nd round, and the initial Kriging model is the Kriging model of the 1st round. The parameter t is introduced here only to more clearly explain the loop judgment process, and thus more clearly explain the technical solution of this application.

[0070] The following describes the process of determining whether the accuracy of the Kriging model in round t meets the requirements.

[0071] First, based on the predicted set of candidate samples for the t-th round output by the Kriging model for the candidate sample set, the following formula is used for judgment:

[0072]

[0073] In the formula, Let t be the accuracy threshold of the Kriging model in round t. Let be the number of samples in the negative sample prediction set for round t. for The upper boundary of the confidence interval, for The upper boundary of the confidence interval, Let be the number of samples that should be positive in the candidate sample prediction set for round t, and the number of samples predicted as negative. Let be the number of samples that should be negative in the candidate sample prediction set for round t, and predict them as positive.

[0074] If that If the accuracy is less than or equal to the preset accuracy threshold, then the accuracy of the Kriging model in round t meets the requirements; if If the accuracy exceeds the preset accuracy threshold, then the accuracy of the Kriging model in round t does not meet the requirements.

[0075] The following is an introduction and The process of determining it.

[0076] The prediction error of the Kriging model is caused by samples that should have been positive but were predicted as negative, and samples that should have been negative but were predicted as positive. Uncertain samples are defined as follows:

[0077]

[0078] In the formula, This means that the sample is in the fault domain but also belongs to the safe domain predicted by the Kriging model. That is, the sample should be a negative sample but is predicted as a positive sample. This means that the sample is in the safe domain but also belongs to the fault domain predicted by the Kriging model. That is, the sample should be a positive sample but is predicted as a negative sample. These are the time and thrust response values ​​predicted by the Kriging model; The variance corresponding to the time and thrust response values ​​predicted by the Kriging model. Factors for constructing confidence intervals, such as The confidence interval is 95%. Considering the statistical characteristics of the Kriging model, and Each sample in the dataset has a probability of being predicted with an incorrect symbol, which can be calculated as follows:

[0079]

[0080] In the formula, In response to The probability of incorrect prediction, where, Represents the fault domain. Indicates a security domain. The time and thrust response values ​​predicted by the Kriging model for a given sample; The variance corresponding to the time and thrust response values ​​predicted by the Kriging model for a given sample.

[0081] according to , and The definition can be calculated using the bootstrap confidence estimation method. and Specifically, from Generated in Samples (correspondingly, from) Generated in (Samples) and replace them, then calculate the value of each sample. and calculate mean This process needs to be repeated. Second-rate, The values ​​are sorted in ascending order. and The confidence intervals are as follows:

[0082]

[0083]

[0084] In the formula, express The confidence interval, express The confidence interval. and Sort by value The order can be calculated as follows:

[0085]

[0086] In the formula, It is the number of repetitions that guide the confidence estimate; It is used to determine and The confidence level. In some examples, and The values ​​can be set to 1000 and 0.05 respectively.

[0087] It should be noted that the method for determining whether the accuracy of the initial Kriging model meets the requirements is similar to the method described above, and will not be repeated here.

[0088] S205. In round t, the probability of incorrect prediction of response values ​​in the two sets is calculated respectively.

[0089] Calculate the probability that each positive sample in the positive sample prediction subset of round t could be a negative sample, and define it as the first probability, using the formula:

[0090]

[0091] In the formula, Let be the probability that a positive sample in the prediction set of the positive samples in the t-th round is a negative sample, and define it as the first probability. Let be the variance of the Kriging model predictions in round t. Let be the response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

[0092] Calculate the probability that each negative sample in the negative sample prediction set of round t could be a positive sample, and define it as the second probability, as follows:

[0093]

[0094] In the formula, Let be the probability that each negative sample in the negative sample prediction set of round t is a positive sample, and define it as the second probability. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

[0095] S206. In the t-th round, the learning function value is calculated based on the probability of incorrect prediction corresponding to the response value in the two sets. The best candidate sample is selected from the candidate sample pool, and the response value corresponding to the best candidate sample is calculated.

[0096] The learning function value for round t is determined based on the first and second probabilities of round t.

[0097] Determine the maximum learning function value in round t from the learning function values ​​in round t;

[0098] The sample corresponding to the maximum learning function value in round t in the candidate sample set is determined as the best candidate sample in round t, and the response value corresponding to the best candidate sample is calculated.

[0099] The learning function expression is:

[0100]

[0101]

[0102] In the formula, Let t be the learning value of the function in the t-th round. Let be the first probability in round t. The second probability in round t. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The cumulative distribution function of the normal distribution, For about The cumulative distribution function of the normal distribution, Let be the difference between the first probability and the second probability in round t.

[0103] S207. Add the best candidate sample and the response value corresponding to the best candidate sample obtained in round t to the initial sample set, update the Kriging model in round t, and obtain the Kriging model in round t+1.

[0104] After obtaining the best candidate sample in round t and its corresponding response value, the Kriging model in round t can be updated using the candidate sample and its corresponding response value to obtain the Kriging model in round t+1. Then, return to S204 to re-evaluate whether the accuracy of the Kriging model in round t+1 meets the requirements.

[0105] S208. Based on the predicted time and thrust values ​​of round t, a reliability analysis of the performance of the solid rocket motor is conducted.

[0106] If the accuracy of the Kriging model in round t meets the requirements, the parameters to be analyzed related to the performance of the solid rocket motor are input into the Kriging model in round t to obtain the predicted time and thrust values ​​for round t output by the Kriging model in round t. The parameters to be analyzed include: canard length, canard circumscribed circle diameter, aft wing length, aft circumscribed circle diameter, canard tilt angle, aft tilt angle, wing width, first-stage propellant reference burning rate, and expansion ratio.

[0107] If the predicted time value in round t indicates that the operational time performance of the solid rocket motor is in failure, or the predicted thrust value in round t indicates that the operational thrust performance of the solid rocket motor is in failure, then the solid rocket motor is determined to be in failure. If the predicted time value in round t indicates that the operational time performance of the solid rocket motor is normal, and the predicted thrust value in round t indicates that the operational thrust performance of the solid rocket motor is normal, then the solid rocket motor is determined to be in failure.

[0108] Specifically, it can be determined whether the predicted time and thrust values ​​for round t are less than 0. If they are, it indicates that the solid rocket motor corresponding to the parameter being analyzed has failed and requires further optimization. If they are not, it indicates that the solid rocket motor corresponding to the parameter has not failed and can provide a valid reference for subsequent design.

[0109] Examples of parameters to be analyzed are shown in Table 1 below.

[0110] Table 1:

[0111]

[0112] In the embodiments of this application, for a series system of performance failure events of aerospace solid rocket motors, the failure probability can be expressed by the following formula:

[0113]

[0114] in, These could be parameters to be analyzed in aerospace solid rocket motors. Represents with respect to x k The state function for each failure event. Pr Indicates an event The probability of occurrence, as expressed in this formula, is shown in the embodiments of this application. k=2, corresponding to the operating time performance and the operating thrust performance, respectively. In this application, failure of the operating thrust or operating time to meet the specified requirements constitutes a failure event. Once either failure event occurs, the series system consisting of the operating thrust performance and the operating time performance will fail.

[0115] The following table compares the number of function calls for parameters in the case study of aerospace solid rocket engines with the results of the Monte Carlo method (MCS).

[0116] Table 2:

[0117]

[0118] As can be seen, in the technical solution of this application, the number of function calls is significantly less than that of the MCS method, which is reduced by 98.8%. In the process of calculating the failure probability of a series system, the solution provided by the embodiment of this application improves the calculation efficiency while ensuring the calculation accuracy.

[0119] Based on the above description, the technical solution of this application fully considers the impact of uncertainties such as geometric errors of aerospace solid rocket motors on operating time and thrust. It establishes a performance reliability model for a series system of aerospace solid rocket motors and proposes a reliability analysis method for aerospace solid rocket motor systems. This method comprehensively considers prediction errors using known information from the model construction process, constructing an improved expected risk learning function. This achieves efficient and accurate evaluation of the performance reliability of the series system, which comprises the operating thrust performance and operating time performance of aerospace solid rocket motors. Furthermore, this method is also applicable to performance functions of aerospace solid rocket motors with high nonlinearity. Compared to component reliability analysis methods based on the Monte Carlo method, this invention conducts performance reliability analysis from the perspective of a series system, which better reflects actual conditions. Simultaneously, this method has lower computational cost and higher accuracy.

[0120] The above text combined Figures 1 to 2 The method for analyzing the performance failure of a fixed engine provided in the embodiments of this application has been described in detail. The device provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0121] like Figure 3 As shown in the figure, this is a schematic diagram of an analysis device for fixed engine performance failure provided in an embodiment of this application. The device includes:

[0122] The update module 301 is used to update the Kriging model in the t-th round based on the candidate samples in the t-th round and the response values ​​corresponding to the candidate samples, so as to obtain the Kriging model in the (t+1)-th round, where t is an integer greater than 0;

[0123] The acquisition module 302 is used to acquire the candidate sample prediction set for the (t+1)th round output by the Kriging model for the candidate sample set, wherein the candidate sample prediction set for the (t+1)th round includes the positive sample prediction set for the (t+1)th round and the negative sample prediction set for the (t+1)th round.

[0124] The judgment module 303 is used to determine whether the accuracy of the Kriging model in the (t+1)th round meets the requirements based on the candidate sample prediction set in the (t+1)th round.

[0125] The probability module 304 is used to determine the probability that each positive sample in the positive sample prediction set of the t-th round may be a negative sample if the accuracy of the Kriging model in the (t+1)-th round does not meet the requirements, and define it as a first probability; and to determine the probability that each negative sample in the negative sample prediction set of the t-th round may be a positive sample, and define it as a second probability.

[0126] The filtering module 305 is used to determine the best candidate sample for the (t+1)th round from the candidate sample set based on the first probability and the second probability for the (t+1)th round, and to determine the response value corresponding to the best candidate sample for the (t)th round.

[0127] The prediction module 306 is used to input the parameters to be analyzed of the solid rocket motor into the Kriging model of the (t+1)th round if the accuracy of the Kriging model of the (t+1)th round meets the requirements, so as to obtain the predicted time value and the predicted thrust value of the (t+1)th round output by the Kriging model of the (t+1)th round.

[0128] Analysis module 307 is used to perform performance failure analysis on the aerospace solid rocket motor based on the predicted time value and the predicted thrust value of the (t+1)th round.

[0129] Optionally, the judgment module 303 can determine the convergence value using the following formula:

[0130]

[0131] in, Let t be the accuracy threshold of the Kriging model in round t. Let be the number of samples in the negative sample prediction set for round t. for The upper boundary of the confidence interval, for The upper boundary of the confidence interval, Let be the number of samples that should be positive in the candidate sample prediction set for round t, and the number of samples predicted as negative. For the candidate sample prediction set in round t, the number of samples that should be negative and are predicted as positive; if If the accuracy is less than or equal to the preset accuracy threshold, then the accuracy of the Kriging model in the t-th round meets the requirements; if If the accuracy exceeds the preset accuracy threshold, the result is that the accuracy of the Kriging model in the t-th round does not meet the requirements.

[0132] Optionally, probability module 304 is specifically used to determine the probability that a positive sample in the positive sample prediction set of round t may be a negative sample using the following formula, and defines it as the first probability:

[0133]

[0134] in, Let be the probability that a positive sample in the prediction set of the positive samples in the t-th round is a negative sample, and define it as the first probability. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

[0135] Optionally, probability module 304 is specifically used to determine the probability that each negative sample in the negative sample prediction set of round t may be a positive sample using the following formula, and defines it as a second probability:

[0136]

[0137] in, Let be the probability that each negative sample in the negative sample prediction set of round t is a positive sample, and define it as the second probability. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

[0138] The filtering module 305 is specifically used to determine the learning function value of round t based on the first probability of round t and the second probability of round t; determine the maximum learning function value of round t from the learning function values ​​of round t; determine the sample corresponding to the maximum learning value of round t in the candidate sample set as the best candidate sample of round t, and then determine the response value corresponding to the best candidate sample of round t.

[0139] Filtering module 305 is specifically used to determine the learning function value for the t-th round using the following formula:

[0140]

[0141]

[0142] in, Let t be the value of the learning function in the t-th round. Let be the first probability in round t. The second probability in round t. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The cumulative distribution function of the normal distribution, For about The cumulative distribution function of a normal distribution.

[0143] Optionally, the analysis module 307 is specifically used to determine that the solid rocket motor performance is failed if the predicted time value of the t-th round indicates that the working time performance of the solid rocket motor is failed or the predicted thrust value of the t-th round indicates that the working thrust performance of the solid rocket motor is failed; and to determine that the solid rocket motor performance is not failed if the predicted time value of the t-th round indicates that the working time performance of the solid rocket motor is normal and the predicted thrust value of the t-th round indicates that the working thrust performance of the solid rocket motor is normal.

[0144] The fixed engine failure analysis apparatus according to the embodiments of this application can correspond to the execution of the method described in the embodiments of this application, and the other operations and / or functions of each module / unit of the fixed engine failure analysis apparatus are respectively for implementing Figure 2 For the sake of brevity, the corresponding processes of each method in the illustrated embodiments will not be described in detail here.

[0145] This application also provides a computing device, such as... Figure 4 As shown, this figure is a schematic diagram of a computing device provided in an embodiment of this application. Figure 4 As shown, the computing device 400 includes a bus 401, a processor 402, a communication interface 403, and a memory 404. The processor 402, the memory 404, and the communication interface 403 communicate with each other via the bus 401.

[0146] Bus 401 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0147] Processor 402 can be any one or more of the following processors: central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP), or digital signal processor (DSP).

[0148] Communication interface 403 is used for communication with external devices.

[0149] Memory 404 may include volatile memory, such as random access memory (RAM). Memory 404 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0150] The memory 404 stores executable code, and the processor 402 executes the executable code to perform the aforementioned fixed engine failure analysis method.

[0151] Specifically, in achieving Figure 3 In the case of the illustrated embodiment, and Figure 3 When the modules or units of the fixed engine failure analysis device described in the embodiment are implemented by software, the following steps are performed: Figure 3 The software or program code required for the functions of each module / unit can be partially or entirely stored in memory 404. Processor 402 executes the program code corresponding to each unit stored in memory 404, and performs the aforementioned fixed engine failure analysis method.

[0152] This application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium capable of being stored by a computing device, or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium includes instructions that instruct a computing device to execute the aforementioned fixed engine failure analysis method.

[0153] This application also provides a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this application are generated.

[0154] The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another website, computer, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0155] When the computer program product is executed by a computer, the computer performs any of the aforementioned methods for analyzing fixed engine failures. The computer program product can be a software installation package; when any of the aforementioned methods for analyzing fixed engine failures is required, the computer program product can be downloaded and executed on the computer.

[0156] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0157] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be covered within the scope of protection of this application.

Claims

1. A reliability analysis method for aerospace solid rocket motor systems, characterized in that, The method includes: An initial sample set and a candidate sample set were constructed respectively. The samples used included: sample forewing length, sample forewing circumscribed circle diameter, sample aft wing length, sample aft wing circumscribed circle diameter, sample forewing tilt angle, sample aft wing tilt angle, sample wing width, sample first-stage charge reference burning rate, and sample internal ballistic maximum combustion chamber pressure. The response values ​​included sample time response value and sample thrust response value. Construct a Kriging model based on the initial sample set; Using the Kriging model in round t, the predicted output results corresponding to the candidate samples are calculated; the obtained predicted output results are divided into two categories according to their signs: the prediction set of positive samples and the prediction set of negative samples. Based on the prediction output of the candidate samples in round t, determine whether the accuracy of the Kriging model in round t meets the requirements; If the accuracy of the Kriging model in round t does not meet the requirements, calculate the probability that each positive sample in the positive sample prediction set in round t may be a negative sample, and define it as the first probability; calculate the probability that each negative sample in the negative sample prediction set in round t may be a positive sample, and define it as the second probability. Based on the first and second probabilities of round t, determine the best candidate sample for round t from the candidate sample set, and the response value corresponding to the best candidate sample; Based on the best candidate sample in round t and the response value corresponding to the best candidate sample, the Kriging model in round t is updated to obtain the Kriging model in round t+1, where t is an integer greater than 0; If the accuracy of the Kriging model in round t meets the requirements, a reliability analysis of the solid rocket motor performance is performed based on the predicted time and thrust values ​​in round t.

2. The method according to claim 1, characterized in that, Based on the predicted set of candidate sample response values ​​from round t, determine whether the accuracy of the Kriging model in round t meets the requirements: In the formula, Let t be the accuracy threshold of the Kriging model in round t. Let be the number of samples in the negative sample prediction set for round t. for The upper boundary of the confidence interval, for The upper boundary of the confidence interval, Let be the number of samples that should be positive in the candidate sample prediction set for round t, and the number of samples predicted as negative. The number of samples that should be negative in the candidate sample prediction set for round t is predicted as positive. if If the accuracy is less than or equal to the preset accuracy threshold, then the accuracy of the Kriging model in round t meets the requirements; if If the accuracy exceeds the preset accuracy threshold, then the accuracy of the Kriging model in round t does not meet the requirements.

3. The method according to claim 1, characterized in that, Calculate the probability that each positive sample in the positive sample prediction subset of round t could be a negative sample, and define it as the first probability. The calculation formula is as follows: in, Let be the probability that a positive sample in the subset of positive samples predicted in round t is a negative sample, and define it as the first probability. Let be the variance of the Kriging model predictions in round t. Let be the response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

4. The method according to claim 1, characterized in that, Calculate the probability that each negative sample in the negative sample prediction subset of round t could be a positive sample, and define it as the second probability. The calculation formula is as follows: in, Let be the probability that a negative sample in the negative sample prediction subset of round t is a positive sample, and define it as the second probability. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The probability density function of the normal distribution. For about The cumulative distribution function of a normal distribution.

5. The method according to claim 1, characterized in that, Based on the first and second probabilities of round t, the best candidate sample for round t and the corresponding response value of the best candidate sample are determined from the candidate sample set, including: The learning function value for round t is determined based on the first and second probabilities of round t. Determine the maximum learning function value in round t from the learning function values ​​in round t; The sample corresponding to the maximum learning function value in round t in the candidate sample set is determined as the best candidate sample in round t, and the response value corresponding to the best candidate sample is calculated.

6. The method according to claim 5, characterized in that, The learning function expression is: In the formula, Let t be the learning value of the function in the t-th round. Let be the first probability in round t. The second probability in round t. Let Variance be the variance of the Kriging model prediction in the t-th round. The response value predicted by the Kriging model in the t-th round. For about The cumulative distribution function of the normal distribution, For about The cumulative distribution function of the normal distribution, Let be the difference between the first probability and the second probability in round t.

7. The method according to claim 1, characterized in that, Based on the predicted time and thrust values ​​for round t, a reliability analysis of the solid rocket motor performance is conducted, including: If the predicted time value in round t represents the failure of the working time performance of the solid rocket motor, or the predicted thrust value in round t represents the failure of the working thrust performance of the solid rocket motor, then the solid rocket motor is determined to have failed. If the predicted time value in round t represents the normal operating time performance of the solid rocket motor and the predicted thrust value in round t represents the normal operating thrust performance of the solid rocket motor, then it is determined that the solid rocket motor has not failed.

Images