Self-adaptive weighted chaotic polynomial reliability analysis method, device and equipment for pressure intensity performance of solid engine and medium

Abstract

The invention discloses a self-adaptive weighted chaotic polynomial reliability analysis method, device and equipment for the pressure intensity performance of a solid engine and a medium, and relates to the technical field of solid engines, and the method comprises the steps: obtaining a training sample according to a variable influencing the pressure intensity performance of the solid engine and an inner ballistic performance calculation model; determining a weight matrix of the training sample by adopting the provided weight definition function, calculating a chaotic polynomial coefficient through a weighted least square method, and constructing a proxy model; selecting an optimal sample point by applying a model prediction error evaluation method and a global search learning function provided by the invention; the training sample set and the weight are dynamically updated according to the provided self-adaptive weighting strategy, and it is worthy of being explained that the updating of the weight does not need to additionally call an inner ballistic calculation model. The method has better robustness for noise in data, the accuracy and calculation efficiency of reliability analysis of the pressure intensity performance of the solid engine can be improved, and it is ensured that the reliability risk is predicted and avoided in the design stage.

Description

Technical Field

[0001] The present application relates to the technical field of solid motors, and in particular to an adaptive weighted chaotic polynomial reliability analysis method, device, equipment and medium for solid motor pressure performance. Background Art

[0002] Solid engines are power equipment in aerospace propulsion engineering and are widely used in the aerospace field. However, in the manufacturing, operation and maintenance of solid engines, there are widespread uncertainties in material properties, geometric dimensions, loads, etc. The failure probability of fixed engines often increases due to the existence of uncertainties, and the reliability is significantly reduced.

[0003] At present, engineers mostly rely on existing experience to design and optimize the reliability of the relevant parameters that affect the combustion chamber pressure. However, the combustion chamber is the main component that bears the internal pressure, and the force of each part in the combustion chamber depends on the pressure of the combustion chamber. Due to the influence of uncertain factors such as environmental factors, material properties, and geometric dimensions, the actual combustion chamber pressure may exceed the shell bearing threshold, resulting in engine failure and safety accidents.

[0004] Therefore, the industry needs a reliability analysis method for the pressure performance of solid rocket motors. Summary of the invention

[0005] The present application provides an adaptive weighted chaotic polynomial reliability analysis method, device, equipment and medium for the pressure performance of solid motors, which can analyze the reliability of the pressure performance of solid motors, thereby providing effective guidance for the reliability optimization of solid motors.

[0006] In order to achieve the above objectives, this application adopts the following technical solutions: In a first aspect, the present application provides an adaptive weighted chaotic polynomial reliability analysis method for solid rocket motor pressure performance, the method comprising: Obtaining a K-th round of training sample set and a training sample pressure corresponding to each training sample group in the K-th round of training sample set, wherein each training sample group in the training sample set includes a charge reference burning rate dimension, a burning rate pressure index dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension, and a charge density dimension; Estimating the density of each training sample group in the K-th round training sample set; Determining a weight matrix of the K-th round training sample set according to the density of each training sample group in the K-th round training sample set; Constructing a K-round chaotic polynomial model by using the K-round training sample set, the training sample pressure corresponding to each training sample group in the K-round training sample set, and the weight matrix corresponding to the K-round training sample set; Obtaining a candidate sample set, inputting the candidate sample set into the K-th round chaotic polynomial model, and obtaining a candidate predicted pressure value corresponding to each candidate sample group; Determining the K-th round failure probability of the solid motor according to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than a preset pressure threshold and the number of candidate sample groups in the candidate sample set; Determine the K-th round learning value corresponding to each candidate sample group in the candidate sample set according to the candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set, and determine the candidate sample group with the smallest learning value in the candidate sample set as the K-th round target candidate sample group; Calculate the distance between the K-th round target candidate sample group and each training sample group in the K-th round training sample set, and determine the training sample group with the smallest distance to the K-th round target candidate sample value as the K-th round target training sample group; If the target sample distance between the K-th round target candidate sample group and the K-th round target training sample group is less than the first distance, the weight of the K-th round target training sample group is updated to obtain the K+1-th round training sample set; if the target distance is greater than or equal to the first distance, the K-th round target candidate sample group is added to the K-th round training sample set to obtain the K+1-th round training sample set; If the failure probability from the Kkth round to the Kth round meets the condition for stopping iteration, the reliability analysis result of the pressure performance of the solid motor is output, and the reliability analysis result is the failure probability of the Kth round.

[0007] Optionally, determining the K-th round learning value corresponding to each candidate sample group in the candidate sample set according to the candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set includes: If K is equal to 1, the first round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula:

[0008] Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding first round learning value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate prediction error; If K is greater than 1, the K-th round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula:

[0009] Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding K-th round learning value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, represents the number of training sample groups in the K-th round of training sample set, represents the number of training sample groups in the first round of training sample set, represents the jth training sample group in the Kth round training sample set in the Kth round iteration.

[0010] Optionally, the candidate prediction error corresponding to each candidate sample group in the candidate sample set is determined by the following formula:

[0011]

[0012] Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, Indicates the prediction model of the lth type in the Kth iteration for the i-th candidate sample group in the candidate sample set The prediction results, Indicates the prediction model of L types in the Kth iteration for the i-th candidate sample group in the candidate sample set The average of the predicted results.

[0013] Optionally, the first distance is determined by the following formula:

[0014] in, represents the first distance, Represents the standard deviation of the dth dimension, where D is the number of dimensions, D=5.

[0015] Optionally, updating the weight of the K-th round target training sample group includes:

[0016] in, is the Kth round of target training sample group The updated weights, is the Kth round of target training sample group The corresponding maximum pressure is Represents the target training sample group in the training sample set in the Kth iteration The corresponding target predicted pressure value, The Kth round of target training sample group Weight before update.

[0017] Optionally, the failure probability from the Kkth round to the Kth round satisfies the condition for stopping iteration, including:

[0018]

[0019]

[0020] in, represents the standard deviation of the failure probability from round Kk to round K, represents the average failure probability from round Kk to round K, represents the failure probability in the mth iteration.

[0021] Optionally, the training sample pressure corresponding to each training sample group in the K-th round training sample set is calculated by a pre-built simulation model.

[0022] In a second aspect, the present application provides an adaptive weighted chaotic polynomial reliability analysis device for solid rocket motor pressure performance, the device comprising: an acquisition module, used to acquire a K-th round of training sample sets and a training sample pressure corresponding to each training sample group in the K-th round of training sample sets, wherein each training sample group in the training sample sets includes a charge reference burning rate dimension, a burning rate pressure index dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension, and a charge density dimension; An estimation module, used for estimating the density of each training sample group in the K-th round training sample set; determining a weight matrix of the K-th round training sample set according to the density of each training sample group in the K-th round training sample set; A training module, used to construct a K-th round chaotic polynomial model by using the K-th round training sample set, the training sample pressure corresponding to each training sample group in the K-th round training sample set, and the weight matrix corresponding to the K-th round training sample set; The acquisition module is also used to acquire a candidate sample set; A verification module is used to input the candidate sample set into the K-th round chaotic polynomial model to obtain a candidate predicted pressure value corresponding to each candidate sample group; according to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than a preset pressure threshold and the number of candidate sample groups in the candidate sample set, determine the K-th round failure probability of the solid motor; A learning module, configured to determine a K-th round learning value corresponding to each candidate sample group in the candidate sample set according to a candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and a candidate prediction error corresponding to each candidate sample group in the candidate sample set, and determine the candidate sample group with the smallest learning value in the candidate sample set as the K-th round target candidate sample group; A calculation module is used to calculate the distance between the K-th round target candidate sample group and each training sample group in the K-th round training sample set, and determine the training sample group with the smallest distance to the K-th round target candidate sample value as the K-th round target training sample group; An updating module, configured to update the weight of the K-th round target training sample group to obtain a K+1-th round training sample set if the target sample distance between the K-th round target candidate sample group and the K-th round target training sample group is less than a first distance, and to add the K-th round target candidate sample group to the K-th round training sample set to obtain a K+1-th round training sample set if the target distance is greater than or equal to the first distance; The analysis module is used to output the reliability analysis result of the pressure performance of the solid engine if the failure probability from the Kkth round to the Kth round meets the stopping iteration condition, and the reliability analysis result is the Kth round failure probability.

[0023] In a third aspect, the present application provides a computing device, including a memory and a processor; One or more computer programs are stored in the memory, and the one or more computer programs include instructions; when the instructions are executed by the processor, the computing device executes the method as described in any one of the first aspects.

[0024] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store a computer program, and the computer program is used to execute the method as described in any one of the first aspects.

[0025] It can be seen from the above technical solution that the present application has at least the following beneficial effects: The present application provides an adaptive weighted chaotic polynomial reliability analysis method for the pressure performance of a solid rocket engine, the method comprising: obtaining a K-th round training sample set and a training sample pressure corresponding to each training sample group in the K-th round training sample set, each training sample group in the training sample set comprising a charge reference burning rate dimension, a burning rate pressure exponent dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension and a charge density dimension, estimating the density of each training sample group in the K-th round training sample set, and then determining a weight matrix of the K-th round training sample set according to the density of each training sample group, and constructing a K-th round chaotic polynomial model using the K-th round training sample set, the training sample pressure corresponding to each training sample group in the K-th round training sample set and the weight matrix corresponding to the K-th round training sample set. Then, a candidate sample set is obtained, and the candidate sample set is input into the K-th round chaotic polynomial model to obtain the candidate predicted pressure value corresponding to each candidate sample group. According to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than the preset pressure threshold and the number of candidate sample groups in the candidate sample set, the K-th round failure probability of the solid motor is determined. According to the candidate predicted pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set, the K-th round learning value corresponding to each candidate sample group in the candidate sample set is determined. The candidate sample group with the smallest learning value in the candidate sample set is determined as the K-th round target candidate sample group; the K-th round target candidate sample group and the K-th round training sample are calculated. The distance between each training sample group is concentrated, and the training sample group with the smallest distance to the target candidate sample value of the Kth round is determined as the target training sample group of the Kth round; if the target sample distance between the target candidate sample group of the Kth round and the target training sample group of the Kth round is less than the first distance, the weight of the target training sample group of the Kth round is updated to obtain the K+1th round training sample set, and if the target distance is greater than or equal to the first distance, the target candidate sample group of the Kth round is added to the Kth round training sample set to obtain the K+1th round training sample set; if the failure probability from the Kkth round to the Kth round meets the condition for stopping iteration, the reliability analysis result of the solid engine pressure performance is output, and the reliability analysis result is the failure probability of the Kth round. It can be seen that this method can analyze the reliability of the solid engine pressure performance, and the reliability analysis result can provide effective guidance for the reliability optimization design of the solid engine, helping users to more conveniently design a solid engine that meets their needs.

[0026] It should be understood that the description of technical features, technical solutions, beneficial effects or similar language in this application does not imply that all features and advantages can be realized in any single embodiment. On the contrary, it is understood that the description of features or beneficial effects means that specific technical features, technical solutions or beneficial effects are included in at least one embodiment. Therefore, the description of technical features, technical solutions or beneficial effects in this specification does not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions and beneficial effects described in the present embodiment can also be combined in any appropriate manner. Those skilled in the art will understand that the embodiment can be realized without one or more specific technical features, technical solutions or beneficial effects of a specific embodiment. In other embodiments, additional technical features and beneficial effects can also be identified in a specific embodiment that does not embody all embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A flow chart of an adaptive weighted chaotic polynomial reliability analysis method for solid rocket motor pressure performance provided in an embodiment of the present application; Figure 2 A schematic diagram of a solid engine provided in an embodiment of the present application; Figure 3 A schematic diagram of an adaptive weighted chaotic polynomial reliability analysis device for solid rocket motor pressure performance provided in an embodiment of the present application; Figure 4 A schematic diagram of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The terms "first", "second", "third", etc. in the specification of this application and the accompanying drawings are used to distinguish different objects rather than to limit a specific order.

[0029] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0030] In order to make the description of the following embodiments clear and concise, a brief introduction to the related technology is first given: Solid engines use solid propellants as the power source for launch vehicles and other equipment. The propellants burn in the combustion chamber to produce high-temperature and high-pressure gas, which is ejected outward through the nozzle to generate thrust to provide sufficient power for the aircraft. The thrust directly depends on the change in the combustion chamber pressure, so the combustion chamber pressure is an important parameter for evaluating the working performance of the engine.

[0031] In addition, the combustion chamber is the main component that bears the internal pressure, and the force applied to each part of the combustion chamber depends on the pressure in the combustion chamber. Due to the influence of uncertain factors such as environmental factors, material properties, and geometric dimensions, the actual combustion chamber pressure may exceed the shell bearing threshold, causing engine failure or even major safety accidents.

[0032] Therefore, evaluating the reliability of solid motor combustion chamber pressure while considering uncertain variables is of great significance for improving the working safety of solid motors.

[0033] In view of this, an embodiment of the present application provides an adaptive weighted chaotic polynomial reliability analysis method for the pressure performance of a solid engine. The method can be applied to a processing device, which can be a terminal or a server. Terminals include but are not limited to smart phones, tablet computers, laptops, personal digital assistants, or smart wearable devices. The server can be a cloud server, such as a central server in a central cloud computing cluster, or an edge server in an edge cloud computing cluster. Of course, the server can also be a server in a local data center. A local data center refers to a data center that is directly controlled by the user. The method can analyze the reliability of the pressure performance of a solid engine, and the reliability analysis results can provide effective guidance for the reliability optimization design of a solid engine, helping users to more conveniently design a solid engine that meets their needs.

[0034] In order to make the technical solution of the present application clearer and easier to understand, the following describes the adaptive weighted chaotic polynomial reliability analysis method for the pressure performance of solid rocket motors provided by the present application in conjunction with the accompanying drawings. Figure 1 As shown in FIG. 1 , the figure is a flow chart of an adaptive weighted chaotic polynomial reliability analysis method for solid rocket motor pressure performance provided by an embodiment of the present application, the method comprising: S101: A processing device obtains a K-th round training sample set and a training sample pressure corresponding to each training sample group in the K-th round training sample set.

[0035] Each training sample group of the training sample set includes a charge reference burning rate dimension, a burning rate pressure exponent dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension, and a charge density dimension.

[0036] In some embodiments, the processing device may first obtain the uncertainty variable information of the given charge reference burning rate, burning rate pressure index, characteristic velocity, nozzle throat ablation rate, charge density, and deterministic variable information such as the combustion chamber cylinder section diameter and nozzle throat diameter. The mean value of the uncertainty variable is , the standard deviation is , where d=1,2,3…D, D=5, represents the mean of the uncertainty variable in the dth dimension, Represents the standard deviation of the uncertainty variable of the dth dimension. For example, the first dimension may be the charge reference burning rate dimension, the second dimension may be the burning rate pressure exponent dimension, the third dimension may be the characteristic velocity dimension, the fourth dimension may be the nozzle throat ablation rate dimension, and the fifth dimension may be the charge density dimension.

[0037] like Figure 2 As shown, this figure is a schematic diagram of a solid engine provided by an embodiment of the present application. The solid engine mainly includes an engine casing 201, a charge 202, an insulation layer 203 and a nozzle 204. The distribution information of the factors affecting the combustion chamber pressure is shown in Table 1 below.

[0038] Table 1:

[0039] After initializing the above uncertainty variables, the processing device can use the Latin hypercube sampling method to obtain the first round of training sample sets.

[0040] Next, the processing device may also calculate the training sample pressure corresponding to each training sample group in the first round of training sample set through a pre-built simulation model.

[0041] Among them, the training sample set of the Kth round is obtained based on the training sample set of the K-1th round, which is introduced below.

[0042] S102: The processing device estimates the density of each training sample group in the K-th round training sample set.

[0043] In some embodiments, the processing device may use a kernel density estimation algorithm to calculate the density of each training sample group in the training sample set. Specifically, the processing device may determine the density of the training sample set by the following formula:

[0044] in, represents the nth training sample group in the training sample set, Represents the nth training sample group in the Kth round training sample set in the Kth round iteration The density of Represents the nth training sample group in the Kth round of training sample set The variable of the dth dimension in represents the first The variable of the dth dimension of the training sample group, It represents the bandwidth of the dth dimension in the Kth round of training sample set, D=5, and D represents the total number of dimensions of the training sample set. represents the Gaussian kernel function.

[0045] Among them, ,but: .

[0046] bandwidth It can be obtained through the empirical formula:

[0047] in, Represents the sample standard deviation of the dth dimension in the Kth round of training sample set: . Indicates the first K round of training sample set The average value of the dimensions.

[0048] S103: The processing device determines a weight matrix of the K-th round training sample set according to the density of each training sample group in the K-th round training sample set.

[0049] In some embodiments, the processing device determines the weight of the K-th round of training sample set by the following formula:

[0050] in, represents the weight matrix of the training sample group in the K-th round of training sample set, Indicates the number of samples in the K-th round of training samples in the K-th round of iteration. Training sample groups density.

[0051] S104: The processing device constructs a K-th round chaotic polynomial model by using the K-th round training sample set, the training sample pressure corresponding to each training sample group in the K-th round training sample set, and the weight matrix corresponding to the K-th round training sample set.

[0052] In some embodiments, the K-th round chaotic polynomial model can be expressed by the following formula:

[0053] In some embodiments, the processing device may be based on the truncation degree By truncating the above formula, we can get:

[0054] in, , is the total number of chaotic polynomials, represents the cutoff degree, =5.

[0055]

[0056]

[0057] in, Represents the input variable in the Kth iteration The predicted pressure value of represents the coefficient of the pth polynomial, Represents input variables The multivariate orthogonal polynomials of Represents the multidimensional index conformity vector of the p-th polynomial, Representation vector The first dimension scalar in , Representation vector The second dimension scalar in , Representation vector The third dimension scalar in , Representation vector The fourth-dimensional scalar in , Representation vector The fifth dimension scalar in . Indicates the d-th dimension random variable corresponding to The orthogonal polynomial of one variable, Represents input variables The d-th dimension scalar of .

[0058] In some embodiments, the orthogonal polynomial follows the Askey scheme, that is, a matching orthogonal polynomial is selected according to the parameter distribution type. For example, when the charge reference burning rate is a normal distribution, the basis function selects the Hermite polynomial. Table 2 is a table of correspondence between parameter distribution types and orthogonal polynomials.

[0059] Table 2:

[0060] Among them, the coefficients of the chaotic polynomial model are solved using the weighted least squares method:

[0061]

[0062] in, represents the coefficient vector of the chaotic polynomial model, represents the weight matrix of the training sample group in the K-th round of training sample set, Represents the vector composed of the training sample pressures (referring to the maximum pressure) corresponding to all training sample groups in the K-th round of training sample set. Indicates that the input variable is The 0th orthogonal polynomial at , Indicates that the input variable is The first orthogonal polynomial when , Indicates that the input variable is The Pth orthogonal polynomial at , for Similarly, I will not go into details here.

[0063] After constructing the above chaotic polynomial model, the processing device can train the above chaotic polynomial model using the K-th round training sample set, the training sample pressure corresponding to each training sample group in the K-th round training sample set, and the weight matrix corresponding to the K-th round training sample set.

[0064] S105: The processing device obtains a candidate sample set.

[0065] In some embodiments, the processing device acquires the candidate sample set in a manner similar to the manner in which the training sample set is acquired, and will not be described in detail herein.

[0066] S106: The processing device inputs the candidate sample set to the K-th round chaotic polynomial model to obtain the candidate predicted pressure value corresponding to each candidate sample group.

[0067] In some embodiments, after the processing device obtains the candidate sample set, it can input the candidate sample set into the K-th round chaotic polynomial model to obtain the candidate predicted pressure value corresponding to each candidate sample group.

[0068] S107. The processing device determines the Kth round failure probability of the solid motor according to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than the preset pressure threshold and the number of candidate sample groups in the candidate sample set.

[0069] After obtaining the candidate predicted pressure value corresponding to each candidate sample group, the processing device compares it with the preset pressure threshold respectively to obtain the number of candidate samples whose candidate predicted pressure values ​​are greater than the preset pressure threshold, and then determines the Kth round failure probability of the solid motor according to the number of candidate samples in the candidate sample set and the number of candidate samples whose candidate predicted pressure values ​​are greater than the preset pressure threshold.

[0070] Specifically, the processing device can determine the failure probability of the Kth round by the following formula:

[0071] in, represents the failure probability of the Kth round, Indicates the number of candidate samples whose candidate prediction pressure value is greater than the preset pressure threshold. Represents the number of candidate samples in the candidate sample set.

[0072] S108. The processing device determines the K-th round learning value corresponding to each candidate sample group in the candidate sample set according to the candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set, and determines the candidate sample group with the smallest learning value in the candidate sample set as the K-th round target candidate sample group.

[0073] In some embodiments, the processing device may first determine the size of K, specifically: If K is equal to 1, the first round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula:

[0074] Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding first round learning value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate prediction error; If K is greater than 1, the K-th round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula:

[0075] Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding K-th round learning value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, represents the number of training sample groups in the K-th round of training sample set, represents the number of training sample groups in the first round of training sample set, represents the jth training sample group in the Kth round training sample set in the Kth round iteration.

[0076] It should be noted that in the embodiment of the present application, the variable K representing the round and other variables representing the number (such as d, j, i, etc.) are all positive integers.

[0077] In some embodiments, the candidate prediction error corresponding to each candidate sample group in the candidate sample set is determined by the following formula:

[0078]

[0079] Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, Indicates the prediction model of the lth type in the Kth iteration for the i-th candidate sample group in the candidate sample set The prediction results, Indicates the prediction model of L types in the Kth iteration for the i-th candidate sample group in the candidate sample set The average of the predicted results.

[0080] In some embodiments, L=4, the processing device can use Hermite, Laguerre, Jacobi and Legendre orthogonal polynomials as basis functions to construct four chaotic polynomial verification models: . Specifically:

[0081]

[0082]

[0083]

[0084] in, Indicates the chaotic polynomial verification model with Hermite as the basis function for the input variable in the Kth iteration The predicted value of pressure, Represents input variables The multivariate Hermite orthogonal polynomials of , represents the coefficient of the pth Hermite polynomial; It represents the chaotic polynomial verification model with Laguerre as the basis function for the input variable in the Kth iteration The predicted value of pressure, Represents input variables The multivariate Laguerre orthogonal polynomials of , represents the coefficient of the pth Laguerre polynomial; Indicates the chaotic polynomial verification model with Jacobi as the basis function for the input variable in the Kth iteration The predicted pressure value of Represents input variables The multivariate Jacobi orthogonal polynomials of represents the coefficients of the p-th Jacobi polynomial; Indicates the chaotic polynomial verification model with Legendre as the basis function for the input variable in the Kth iteration The predicted pressure value of Represents input variables The multivariate Legendre orthogonal polynomials of , represents the coefficients of the p-th Legendre polynomial.

[0085] S109: The processing device calculates the distance between the K-th round target candidate sample group and each training sample group in the K-th round training sample set, and determines the training sample group with the smallest distance to the K-th round target candidate sample value as the K-th round target training sample group.

[0086] In some embodiments, the processing device may calculate the distance between the Kth round target candidate sample group and each training sample group in the Kth round training sample set, and then determine the training sample group with the smallest distance to the Kth round target candidate sample value as the Kth round target training sample group. The above distance is a Euclidean distance.

[0087] S110: The processing device determines whether the target sample distance is less than the first distance.

[0088] The target sample distance is the distance between the Kth round target candidate sample group and the Kth round target training sample group, such as the Euclidean distance. If the target sample distance is less than the first distance, S111 is executed; if the target sample distance is not less than the first distance, S112 is executed.

[0089] The first distance is determined by the following formula:

[0090] in, represents the first distance, Represents the standard deviation of the dth dimension, where D is the number of dimensions, D=5.

[0091] S111. The processing device updates the weight of the K-th round target training sample group to obtain the K+1-th round training sample set.

[0092] In some embodiments, the target sample distance is less than the first distance, and the processing device may update the weight of the target training sample group in the Kth round by the following formula:

[0093] in, is the Kth round of target training sample group The updated weights, is the Kth round of target training sample group The corresponding maximum pressure is Represents the target training sample group in the training sample set in the Kth iteration The corresponding target predicted pressure value, The Kth round of target training sample group Weight before update.

[0094] It should be noted that the processing device updates the weight of the K-1th round target training sample group. After obtaining the weight of the Kth round target training sample group, the Kth round training sample set is obtained. The specific updating process is similar to the above embodiment and will not be repeated here.

[0095] S112: The processing device adds the K-th round target candidate sample group to the K-th round training sample set to obtain the K+1-th round training sample set.

[0096] If the target sample distance is not less than the first distance, the processing device may add the K-th round target candidate sample group to the K-th round training sample set to obtain the K+1-th round training sample set. It should be noted that the process of the processing device processing the K-1-th round training sample set to obtain the K-th round training sample set is similar and will not be repeated here.

[0097] S113. If the failure probability from the Kkth round to the Kth round meets the conditions for stopping iteration, the reliability analysis result of the solid rocket engine pressure performance is output.

[0098] The reliability analysis result is the K-th round failure probability.

[0099] In some embodiments, the failure probability from the Kkth round to the Kth round satisfies the condition for stopping iteration, including:

[0100]

[0101]

[0102] in, represents the standard deviation of the failure probability from round Kk to round K, represents the average failure probability from round Kk to round K, represents the failure probability in the mth iteration.

[0103] Table 3 is a schematic diagram of the maximum pressure reliability analysis results of the solid rocket motor provided in the embodiments of the present application.

[0104] Table 3:

[0105] From the reliability analysis results shown in Table 3, we can see that due to the introduction of the adaptive weighting strategy and active learning algorithm, the interior ballistic calculation model was only called 28 times, and the final calculated failure probability is , with an error of only 0.59% compared with the Monte Carlo simulation.

[0106] Table 4 is a comparison of the reliability analysis results of the maximum pressure of the solid rocket engine considering noise provided in the embodiments of the present application.

[0107] Table 4:

[0108] The results show that the proposed method still has high computational efficiency when noisy data is introduced. The zero-dimensional interior ballistic calculation model is called 28 and 35 times, respectively, and the prediction errors of the failure probability are 0.29% and 0.59%, respectively.

[0109] Based on the above description, the adaptive weighted chaotic polynomial reliability analysis method judges the prediction error of the chaotic polynomial model through the difference of basis functions, proposes an adaptive weighting strategy, updates the training sample set and sample weight of the chaotic polynomial model according to the model prediction error and the mutual correlation information of the training samples, and improves the solution efficiency and anti-noise ability of the maximum pressure reliability analysis of the solid motor. (1) Compared with the reliability analysis method based on sampling, the present invention uses a small number of samples to construct a high-precision proxy model, and uses the proxy model to replace the solid motor internal ballistic calculation model for reliability analysis, thereby improving the calculation efficiency; (2) Compared with the existing reliability analysis method based on the proxy model, the present invention proposes a prediction error evaluation method for the chaotic polynomial model, and combines the adaptive weighting strategy to select sample points and update sample weights, so as to achieve a rapid prediction of the maximum pressure failure probability of the solid motor at a lower calculation cost.

[0110] Combination of the above Figure 1 to Figure 2 The adaptive weighted chaotic polynomial reliability analysis method for the pressure performance of a solid rocket engine provided in an embodiment of the present application is introduced in detail. The apparatus and device provided in the embodiment of the present application will be introduced in conjunction with the accompanying drawings.

[0111] like Figure 3 As shown in FIG. 1 , this figure is a schematic diagram of an adaptive weighted chaotic polynomial reliability analysis device for solid rocket engine pressure performance provided by an embodiment of the present application, and the device includes: An acquisition module 301 is used to acquire a K-th round of training sample sets and a training sample pressure corresponding to each training sample group in the K-th round of training sample sets, wherein each training sample group in the training sample sets includes a charge reference burning rate dimension, a burning rate pressure index dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension, and a charge density dimension; An estimation module 302 is used to estimate the density of each training sample group in the K-th round training sample set; and determine a weight matrix of the K-th round training sample set according to the density of each training sample group in the K-th round training sample set; A training module 303 is used to construct a K-th round chaotic polynomial model by using the K-th round training sample set, the training sample pressure corresponding to each training sample group in the K-th round training sample set, and the weight matrix corresponding to the K-th round training sample set; The acquisition module 301 is also used to acquire a candidate sample set; The verification module 304 is used to input the candidate sample set into the K-th round chaotic polynomial model to obtain a candidate predicted pressure value corresponding to each candidate sample group; according to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than a preset pressure threshold and the number of candidate sample groups in the candidate sample set, determine the K-th round failure probability of the solid motor; A learning module 305 is used to determine the K-th round learning value corresponding to each candidate sample group in the candidate sample set according to the candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set, and determine the candidate sample group with the smallest learning value in the candidate sample set as the K-th round target candidate sample group; A calculation module 306 is used to calculate the distance between the K-th round target candidate sample group and each training sample group in the K-th round training sample set, and determine the training sample group with the smallest distance to the K-th round target candidate sample value as the K-th round target training sample group; An updating module 307 is configured to update the weight of the K-th round target training sample group to obtain a K+1-th round training sample set if the target sample distance between the K-th round target candidate sample group and the K-th round target training sample group is less than a first distance, and to add the K-th round target candidate sample group to the K-th round training sample set to obtain a K+1-th round training sample set if the target distance is greater than or equal to the first distance; The analysis module 308 is used to output the reliability analysis result of the pressure performance of the solid motor if the failure probability from the Kkth round to the Kth round meets the condition for stopping iteration, and the reliability analysis result is the Kth round failure probability.

[0112] Optionally, the learning module 305 is specifically used for: If K is equal to 1, the first round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula:

[0113] Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding first round learning value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate prediction error; If K is greater than 1, the K-th round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula:

[0114] Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding K-th round learning value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, represents the number of training sample groups in the K-th round of training sample set, represents the number of training sample groups in the first round of training sample set, represents the jth training sample group in the Kth round training sample set in the Kth round iteration.

[0115] Optionally, the candidate prediction error corresponding to each candidate sample group in the candidate sample set is determined by the following formula:

[0116]

[0117] Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, Indicates the prediction model of the lth type in the Kth iteration for the i-th candidate sample group in the candidate sample set The prediction results, Indicates the prediction model of L types in the Kth iteration for the i-th candidate sample group in the candidate sample set The average of the predicted results.

[0118] Optionally, the first distance is determined by the following formula:

[0119] in, represents the first distance, Represents the standard deviation of the dth dimension, where D is the number of dimensions, D=5.

[0120] Optionally, the updating module 307 is specifically configured to update the weight using the following formula:

[0121] in, is the Kth round of target training sample group The updated weights, is the Kth round of target training sample group The corresponding maximum pressure of the target training sample, Represents the target training sample group in the training sample set in the Kth iteration The corresponding target predicted maximum pressure value, The Kth round of target training sample group Weight before update.

[0122] Optionally, the failure probability from the Kkth round to the Kth round satisfies the condition for stopping iteration, including:

[0123]

[0124]

[0125] in, represents the standard deviation of the failure probability from round Kk to round K, represents the average failure probability from round Kk to round K, represents the failure probability in the mth iteration.

[0126] The adaptive weighted chaotic polynomial reliability analysis device for solid engine pressure performance according to the embodiment of the present application can correspond to the method described in the embodiment of the present application, and the above-mentioned other operations and / or functions of each module / unit of the reliability analysis device for solid engine pressure performance are respectively to achieve Figure 1 For the sake of brevity, the corresponding processes of each method in the illustrated embodiment are not described in detail here.

[0127] The present application also provides a computing device. Figure 4 As shown, this figure is a schematic diagram of a computing device provided in an embodiment of the present application, and 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 through the bus 401.

[0128] The bus 401 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0129] The processor 402 may be any one or more of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MP), or a digital signal processor (DSP).

[0130] The communication interface 403 is used for communicating with the outside.

[0131] The memory 404 may include a volatile memory, such as a random access memory (RAM). The memory 404 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid state drive (SSD).

[0132] The memory 404 stores executable codes, and the processor 402 executes the executable codes to perform the reliability analysis method of the pressure performance of the solid rocket engine.

[0133] Specifically, in implementing Figure 3 In the case of the embodiment shown, and Figure 3 When each module or unit of the reliability analysis device for the pressure performance of a solid rocket engine described in the embodiment is implemented by software, the execution Figure 3 The software or program code required for the functions of each module / unit in the system may be partially or completely stored in the memory 404. The processor 402 executes the program code corresponding to each unit stored in the memory 404 to execute the reliability analysis method of the pressure performance of the solid rocket engine.

[0134] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium can be any available medium that can be 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 (such as a floppy disk, a hard disk, a tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid-state hard disk). The computer-readable storage medium includes instructions that instruct the computing device to execute the reliability analysis method of the pressure performance of the solid engine.

[0135] The embodiment of the present application further provides a computer program product, which includes one or more computer instructions. When the computer instructions are loaded and executed on a computing device, the process or function described in the embodiment of the present application is generated in whole or in part.

[0136] The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. 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, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0137] When the computer program product is executed by a computer, the computer executes any of the aforementioned methods for analyzing the reliability of pressure performance of a solid rocket engine. The computer program product may be a software installation package, and when any of the aforementioned methods for analyzing the reliability of pressure performance of a solid rocket engine is required, the computer program product may be downloaded and executed on a computer.

[0138] The descriptions of the processes or structures corresponding to the above-mentioned figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0139] The above description is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application.

Claims

1. An adaptive weighted chaotic polynomial reliability analysis method for solid rocket motor pressure performance, characterized in that: The method comprises: Obtaining a K-th round of training sample set and a training sample pressure corresponding to each training sample group in the K-th round of training sample set, wherein each training sample group in the training sample set includes a charge reference burning rate dimension, a burning rate pressure index dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension, and a charge density dimension; Estimating the density of each training sample group in the K-th round training sample set; Determining a weight matrix of the K-th round training sample set according to the density of each training sample group in the K-th round training sample set; Constructing a K-round chaotic polynomial model by using the K-round training sample set, the training sample pressure corresponding to each training sample group in the K-round training sample set, and the weight matrix corresponding to the K-round training sample set; Obtaining a candidate sample set; inputting the candidate sample set into the K-th round chaotic polynomial model to obtain a candidate predicted pressure value corresponding to each candidate sample group; Determining the K-th round failure probability of the solid motor according to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than a preset pressure threshold and the number of candidate sample groups in the candidate sample set; Determine the K-th round learning value corresponding to each candidate sample group in the candidate sample set according to the candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set, and determine the candidate sample group with the smallest learning value in the candidate sample set as the K-th round target candidate sample group; Calculate the distance between the K-th round target candidate sample group and each training sample group in the K-th round training sample set, and determine the training sample group with the smallest distance to the K-th round target candidate sample value as the K-th round target training sample group; If the target sample distance between the K-th round target candidate sample group and the K-th round target training sample group is less than the first distance, the weight of the K-th round target training sample group is updated to obtain the K+1-th round training sample set; if the target distance is greater than or equal to the first distance, the K-th round target candidate sample group is added to the K-th round training sample set to obtain the K+1-th round training sample set; If the failure probability from the Kkth round to the Kth round meets the condition for stopping iteration, the reliability analysis result of the pressure performance of the solid motor is output, and the reliability analysis result is the failure probability of the Kth round.

2. The method according to claim 1, characterized in that The determining, according to the candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and the candidate prediction error corresponding to each candidate sample group in the candidate sample set, the K-th round learning value corresponding to each candidate sample group in the candidate sample set comprises: If K is equal to 1, the first round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula: Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding first round learning value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate predicted pressure value, Represents the i-th candidate sample group in the candidate sample set in the first iteration The corresponding candidate prediction error; If K is greater than 1, the K-th round learning value corresponding to each candidate sample group in the candidate sample set is determined according to the following formula: Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding K-th round learning value, Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate predicted pressure value, represents the candidate prediction error corresponding to the i-th candidate sample group in the candidate sample set in the K-th iteration, represents the number of training sample groups in the K-th round of training sample set, represents the number of training sample groups in the first round of training sample set, represents the jth training sample group in the Kth round training sample set in the Kth round iteration.

3. The method according to claim 2, characterized in that The candidate prediction error corresponding to each candidate sample group in the candidate sample set is determined by the following formula: Represents the i-th candidate sample group in the candidate sample set in the K-th iteration The corresponding candidate prediction error, Indicates the prediction model of the lth type in the Kth iteration for the i-th candidate sample group in the candidate sample set The prediction results, Indicates the prediction model of L types in the Kth iteration for the i-th candidate sample group in the candidate sample set The average of the predicted results.

4. The method according to claim 1, characterized in that: The first distance is determined by the following formula: in, represents the first distance, Represents the standard deviation of the dth dimension, where D is the number of dimensions, D=5.

5. The method according to claim 1, characterized in that The updating of the weight of the K-th round target training sample group includes: in, is the Kth round of target training sample group The updated weights, is the Kth round of target training sample group The corresponding maximum pressure is Represents the target training sample group in the training sample set in the Kth iteration The corresponding target predicted pressure value, The Kth round of target training sample group Weight before update.

6. The method according to claim 1, characterized in that The failure probability from the Kkth round to the Kth round satisfies the conditions for stopping iteration, including: in, represents the standard deviation of the failure probability from round Kk to round K, represents the average failure probability from round Kk to round K, represents the failure probability in the mth iteration.

7. The method according to claim 1, characterized in that The training sample pressure corresponding to each training sample group in the K-th round training sample set is calculated by a pre-built simulation model.

8. An adaptive weighted chaotic polynomial reliability analysis device for solid rocket engine pressure performance, characterized in that: The device comprises: an acquisition module, used to acquire a K-th round of training sample sets and a training sample pressure corresponding to each training sample group in the K-th round of training sample sets, wherein each training sample group in the training sample sets includes a charge reference burning rate dimension, a burning rate pressure index dimension, a characteristic velocity dimension, a nozzle throat ablation rate dimension, and a charge density dimension; An estimation module, used for estimating the density of each training sample group in the K-th round training sample set; determining a weight matrix of the K-th round training sample set according to the density of each training sample group in the K-th round training sample set; A training module, used to construct a K-th round chaotic polynomial model by using the K-th round training sample set, the training sample pressure corresponding to each training sample group in the K-th round training sample set, and the weight matrix corresponding to the K-th round training sample set; The acquisition module is also used to acquire a candidate sample set; A verification module is used to input the candidate sample set into the K-th round chaotic polynomial model to obtain a candidate predicted pressure value corresponding to each candidate sample group; according to the number of candidate sample groups whose candidate predicted pressure values ​​are greater than a preset pressure threshold and the number of candidate sample groups in the candidate sample set, determine the K-th round failure probability of the solid motor; A learning module, configured to determine a K-th round learning value corresponding to each candidate sample group in the candidate sample set according to a candidate prediction pressure value corresponding to each candidate sample group in the candidate sample set and a candidate prediction error corresponding to each candidate sample group in the candidate sample set, and determine the candidate sample group with the smallest learning value in the candidate sample set as the K-th round target candidate sample group; A calculation module is used to calculate the distance between the K-th round target candidate sample group and each training sample group in the K-th round training sample set, and determine the training sample group with the smallest distance to the K-th round target candidate sample value as the K-th round target training sample group; An updating module, configured to update the weight of the K-th round target training sample group to obtain a K+1-th round training sample set if the target sample distance between the K-th round target candidate sample group and the K-th round target training sample group is less than a first distance, and to add the K-th round target candidate sample group to the K-th round training sample set to obtain a K+1-th round training sample set if the target distance is greater than or equal to the first distance; The analysis module is used to output the reliability analysis result of the pressure performance of the solid engine if the failure probability from the Kkth round to the Kth round meets the stopping iteration condition, and the reliability analysis result is the Kth round failure probability.

9. A computing device, characterized in that including memory and processor; One or more computer programs are stored in the memory, and the one or more computer programs include instructions; when the instructions are executed by the processor, the computing device executes the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store a computer program, and the computer program is used to execute the method according to any one of claims 1 to 7.

Images