CN117875094B

CN117875094B - Precision evaluation method and device for engine simulation model and electronic equipment

Info

Publication number: CN117875094B
Application number: CN202410282665.XA
Authority: CN
Inventors: 李斌; 高玉闪; 胡海峰; 周晨初; 陈嘉智; 曹国彦; 李舒欣; 李晨沛; 谢豫
Original assignee: Xian Aerospace Propulsion Institute
Current assignee: Xian Aerospace Propulsion Institute
Priority date: 2024-03-13
Filing date: 2024-03-13
Publication date: 2024-06-21
Anticipated expiration: 2044-03-13
Also published as: CN117875094A

Abstract

The invention discloses a precision evaluation method and device of an engine simulation model and electronic equipment, relates to the technical field of engine simulation, and aims to provide a technical scheme for improving the precision of the engine simulation model so as to better support the design, optimization and application of an engine. Comprising the following steps: and obtaining target parameter simulation data and target parameter test data of the engine. And determining the target parameter simulation data and the target parameter test data as target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data. And determining the steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter. And determining the fit degree evaluation result of the target parameter and the similarity evaluation result of the target parameter under each prediction algorithm by utilizing a plurality of prediction algorithms. And determining the transient precision of the target parameter by using a gray correlation analysis method.

Description

Precision evaluation method and device for engine simulation model and electronic equipment

Technical Field

The invention relates to the technical field of rocket engine simulation, in particular to an accuracy evaluation method and device for an engine simulation model and electronic equipment.

Background

The modeling and simulation of the high-precision and high-reliability space rocket engine play an extremely important role in any important space rocket launching and running activities.

The space engine has the characteristics of various data types, huge quantity, strong time-varying property, high coupling property and nonlinearity, so that the traditional model identification and parameter identification technology cannot meet the high-precision understanding. Therefore, the efficiency and the accuracy of the operation reliability analysis of the current space engine still need to be improved. The modeling method based on the multisource information fusion technology can realize high-precision, high-real-time and high-reliability liquid rocket engine analysis.

Through high fidelity simulation experiments, the operation rule and the working mode in the space engine can be deeply and systematically known, possible faults can be obtained through early insight, design defects can be found early, and therefore development efficiency and quality are greatly improved. In addition, the requirement of physical tests can be reduced through simulation, so that the research and development cost and the failure risk are obviously reduced. Meanwhile, the simulation technology enables the space motor to rapidly perform technology upgrading and iteration, and the space motor is promoted to continuously develop.

The accuracy can influence the reliability and the credibility of the simulation model, and the application value of the simulation model can be influenced. Although the rocket engine model in China is built, the simulation is still mainly numerical simulation, and the precision is still improved. The accuracy not only can influence the reliability and the credibility of the simulation model, but also can influence the application value of the simulation model.

Disclosure of Invention

The invention aims to provide a precision evaluation method and device for an engine simulation model and electronic equipment, which are used for providing a technical scheme for improving the precision of the engine simulation model so as to better support the design, optimization and application of an engine.

In a first aspect, the present invention provides a method for evaluating accuracy of an engine simulation model, the method for evaluating accuracy of an engine simulation model comprising:

and obtaining target parameter simulation data and target parameter test data of the engine.

And determining the target parameter simulation data and the target parameter test data as target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data by using a first preset mode.

And according to a second preset mode, determining the steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter.

And determining a fit degree evaluation result of the target parameter and a similarity evaluation result of the target parameter under each prediction algorithm by utilizing a plurality of prediction algorithms based on the target parameter transient simulation data and the target parameter transient test data.

And determining the transient precision of the target parameter by using a gray correlation analysis method based on the fitting degree evaluation result of the target parameter and the similarity evaluation result of the target parameter corresponding to the multiple prediction algorithms.

Under the condition of adopting the technical scheme, the precision evaluation method of the engine simulation model provides the evaluation of the precision of the engine simulation model based on steady-state precision and transient precision. Based on the method, the accuracy of the engine simulation model can be evaluated from different dimensions, and the accuracy of the accuracy evaluation of the engine simulation model is improved, so that the design, optimization and application of the engine are better supported.

In the prior art, a single index is generally used to evaluate the simulation accuracy of an engine. However, since a single index evaluation cannot comprehensively consider a plurality of performance indexes and requirements of an engine, an accurate evaluation on some important performance parameters may be lacked, so that simulation accuracy is affected. Secondly, as various performance indexes of the engine are mutually influenced, the error evaluation indexes are selected to cause deviation on simulation precision. In addition, for some complex problems, the conventional single-index evaluation is difficult to meet the requirement of multi-objective optimization, and cannot provide comprehensive and multi-dimensional simulation precision evaluation. Finally, the traditional single index evaluation method is insufficient for evaluating the performance of the space engine under different working conditions, and the performance of the engine under different working conditions cannot be comprehensively reflected, so that the deviation of the evaluation result is caused. The invention carries out precision evaluation on the engine simulation model based on a combined prediction precision evaluation system of gray correlation analysis. Based on the method, the fitting degree evaluation result of the target parameter and the similarity evaluation result of the target parameter under each prediction algorithm can be determined based on a plurality of prediction algorithms, so that the performance of the engine can be more comprehensively evaluated. By combining the fit degree evaluation results of the target parameters corresponding to the plurality of prediction algorithms with the similarity evaluation results of the target parameters, the limitation of a single algorithm can be reduced, and the evaluation accuracy can be further improved. Secondly, the combined prediction based on gray correlation analysis can effectively reveal the correlation and interaction between indexes corresponding to different algorithms. There is often a certain correlation between different indexes of the engine, and changing one of the indexes may have an influence on the other indexes. According to the method, the contribution and the influence of each parameter can be accurately estimated by analyzing the relevance among indexes, so that the estimation accuracy is improved. In addition, the combined prediction method based on gray correlation analysis can flexibly consider the weights and the importance of different indexes. Different performance indicators may have different importance and weights in the design and optimization of the engine. The method can reasonably set weights according to actual demands, so that the quality and the importance degree of each index can be accurately evaluated, and the design, optimization and application of the engine are better supported.

Further, the first preset mode includes a sliding window algorithm.

The target parameter steady-state simulation data comprises a plurality of target parameter steady-state simulation data segments; the target parameter steady-state test data comprises a plurality of target parameter steady-state test data segments, the target parameter transient simulation data comprises a plurality of target parameter transient simulation data segments, and the target parameter transient test data comprises a plurality of target parameter transient test data segments; the target parameter steady-state simulation data segments, the target parameter transient simulation data segments and the target parameter transient test data segments are in one-to-one correspondence.

Further, the second preset mode includes determining the steady-state accuracy of the target parameter by using the difference between the steady-state simulation data of the target parameter and the steady-state test data of the target parameter.

Further, the plurality of prediction algorithms include a correlation coefficient method, a root mean square error method, a relative deviation method and a tayer inconsistency coefficient analysis method for determining a fitness evaluation result of the target parameter, and a dynamic time-warping similarity algorithm, a puloc Lu Si s similarity algorithm, a power spectral density algorithm and a fourier similarity algorithm for determining a similarity evaluation result of the target parameter.

Further, the determining, by using a gray correlation analysis method, the transient precision of the target parameter based on the fit evaluation result of the target parameter and the similarity evaluation result of the target parameter corresponding to the plurality of prediction algorithms includes:

Normalizing the fit degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm;

Determining a target evaluation result from the fit evaluation result of the target parameters and the similarity evaluation result of the target parameters corresponding to the multiple prediction algorithms;

Determining a correlation coefficient of each remaining evaluation result and the target evaluation result based on an absolute difference value between the target evaluation result and the remaining evaluation result;

And determining the transient precision of the target parameter based on the association coefficients of the rest evaluation results and the target evaluation result.

Further, normalizing the fit evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm includes:

Initializing a data sequence meeting a first preset condition in the fit degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm;

And carrying out averaging treatment on the data sequences meeting the second preset conditions in the fit degree evaluation results of the target parameters corresponding to each prediction algorithm and the similarity evaluation results of the target parameters corresponding to each prediction algorithm.

Further, the first preset condition includes: the difference between at least one data in the sequence of data and adjacent data is greater than a first value.

And/or, the second preset condition includes: the difference between the data in the data sequence and the adjacent data is smaller than the second value.

The association coefficients of the rest evaluation results and the target evaluation result satisfy the following conditions:

；

Wherein, ；/>；，/>As the resolution coefficient, the value range is [0,1], and P ₀ (k) represents the target evaluation result; p _i (k) represents the rest of the evaluation results; /(I)Representing the absolute value of the difference between the target evaluation result and the rest of the evaluation results,/>Representing the minimum value of the absolute value of the difference between the target evaluation result and the rest of the evaluation results,/>And the maximum value of the absolute value of the difference value between the target evaluation result and the rest evaluation results is represented.

In a second aspect, the present invention also provides a precision evaluation device of an engine simulation model, the precision evaluation device of the engine simulation model including:

and the acquisition module is used for acquiring the target parameter simulation data and the target parameter test data of the engine.

The data determining module is used for determining the target parameter simulation data and the target parameter test data into target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data by utilizing a first preset mode.

The steady-state precision determining module is used for determining the steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter according to a second preset mode.

The transient evaluation determining module is used for determining a fitting degree evaluation result of the target parameter and a similarity evaluation result of the target parameter under each prediction algorithm by utilizing a plurality of prediction algorithms based on the target parameter transient simulation data and the target parameter transient test data.

The transient precision determining module is used for determining the transient precision of the target parameter by using a gray correlation analysis method based on the fitting degree evaluation result of the target parameter and the similarity evaluation result of the target parameter corresponding to the plurality of prediction algorithms.

In a third aspect, the present invention also provides an electronic device, including: one or more processors; and one or more machine readable media having instructions stored thereon that, when executed by the one or more processors, cause performance of the method of precision assessment of an engine simulation model of the first aspect.

The technical effects achieved by the solutions provided in the second aspect and the third aspect are the same as those of the method aspects provided in the first aspect, and are not described herein.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention and do not constitute a limitation on the invention. In the drawings:

FIG. 1 is a schematic flow chart of steps of a precision evaluation method of an engine simulation model provided by the invention;

FIG. 2 is a schematic diagram of a device for evaluating the accuracy of an engine simulation model according to the present invention;

fig. 3 is a schematic structural diagram of an electronic device according to the present invention;

fig. 4 is a schematic structural diagram of a chip according to the present invention.

Detailed Description

In order to clearly describe the technical solution of the embodiments of the present invention, in the embodiments of the present invention, the words "first", "second", etc. are used to distinguish the same item or similar items having substantially the same function and effect. For example, the first threshold and the second threshold are merely for distinguishing between different thresholds, and are not limited in order. It will be appreciated by those of skill in the art that the words "first," "second," and the like do not limit the amount and order of execution, and that the words "first," "second," and the like do not necessarily differ.

In the present invention, the words "exemplary" or "such as" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "for example" should not be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.

Based on this, the present invention provides a precision evaluation method of an engine simulation model, referring to fig. 1, the precision evaluation method of the engine simulation model includes:

s101, acquiring target parameter simulation data and target parameter test data of the engine.

In an embodiment of the invention, the target parameters of the engine include, but are not limited to, thrust, fuel flow, oxidant flow, combustion chamber pressure, combustion chamber temperature.

The target parameter test data are data obtained when the engine is tested. The target parameter simulation data are simulation data corresponding to the target parameter test data. The simulation data may be obtained using simulation software.

S102, determining the target parameter simulation data and the target parameter test data as target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data by using a first preset mode.

The first preset mode may be a sliding window algorithm. The basic idea of the sliding window algorithm is to divide the data into a plurality of windows with fixed sizes, and analyze or process the data in each window. Specifically, the sliding window algorithm may be divided into the following steps:

1) Interference and noise removal: noise and interference are one of the main factors affecting the judgment in transient and steady state analysis of the parameters of the engine. Therefore, the target parameter steady-state simulation data, the target parameter steady-state test data, the target parameter transient-state simulation data and the target parameter transient-state test data need to be subjected to the processes of filtering, denoising, noise reduction and the like, so that the quality of signals is improved.

2) Determining the window size: the size of each window first needs to be determined. The window size is generally fixed and may be chosen according to the circumstances.

3) Sliding window: and sliding the window from the starting position of the data, wherein each time the window is slid by a fixed step length, repeating the data analysis or processing in the window until the window is slid to the end position of the data.

4) Window data processing: the data within each window is analyzed or processed, such as by averaging, standard deviation, maximum, minimum, etc. of the data within the window.

5) It is assumed that there is a data sequence x= { X ₁,X₂,…,X_n } of length n, which is to be subjected to a sliding window treatment, where the window has a size of k and the sliding step has a size of s. Then the data of the window at time i can be formulated using the following formula. The data sequence may be a certain segment of the target parameter simulation data and the target parameter trial data.

Wherein, [ X _i,X_i+1,…,X_i+k−1 ] represents a window consisting of k data points from X _i to X _i+k−1. The process of sliding the window can be expressed using the following formula:

；

suppose the standard deviation of the data within window Xi is:

；

If the standard deviation within a certain window Wherein/>If the threshold is set by human, the interval is regarded as a steady state (STEADY STATE); otherwise, it is recorded as a transient (TRANSIENT STATE). Specifically, the start of the first section that reaches the steady state is taken as the start of the transient section, and the end of the section that is the first transient section after the steady state minus the step size is taken as the end of the steady state section, and the start of the transient section is taken as the end of the steady state section.

In the embodiment of the invention, the target parameter simulation data and the target parameter test data can be determined as the target parameter steady-state simulation data, the target parameter steady-state test data, the target parameter transient simulation data and the target parameter transient test data based on the sliding window algorithm.

S103, determining the steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter according to a second preset mode.

The second preset mode includes: and determining the steady-state precision of the target parameter by utilizing the difference value between the steady-state simulation data of the target parameter and the steady-state test data of the target parameter.

In practice, in steady state estimation, errors in engine parameters in steady state operation are of greater concern. The difference between the simulation result and the actual result is calculated by comparison with the test data to evaluate the accuracy of the steady state. Such evaluations typically involve comparisons of engine key parameters, such as thrust, combustion chamber pressure, nozzle temperature, etc., to evaluate steady state accuracy by calculating deviations between simulation results and test data.

S104, determining a fit degree evaluation result of the target parameter and a similarity evaluation result of the target parameter under each prediction algorithm by utilizing a plurality of prediction algorithms based on the target parameter transient simulation data and the target parameter transient test data.

In an embodiment of the present invention, the plurality of prediction algorithms includes: correlation coefficient method (correlation), root mean square error method (rmse-value), relative deviation method (BIAS), and tayer non-uniform coefficient analysis method (theil's U) for determining the fit evaluation result of the target parameter, and dynamic time-warping similarity algorithm (DTW), prolog Lu Si s similarity algorithm (Procrustes), power spectral density algorithm (PSD), and Fourier similarity algorithm (Fourier) for determining the similarity evaluation result of the target parameter.

The following explains each algorithm in detail:

the core idea of the dynamic time warping similarity algorithm (DTW) is dynamic planning. The DTW algorithm describes the alignment between time series by means of an alignment path pi, which is made up of n tuples, each tuple being made up of two data points from two different time series respectively, under three rules of borderline (the start and stop points of the two time series must correspond to each other), monotonicity (the data point after the data point that has been corresponding cannot correspond to the data point before the data point that has been corresponding, i.e. it cannot "cross-correspond"), continuity (all data points on one time series must have data points on the other time series corresponding to it). The alignment path pi between them can be defined as follows:

given the time series x= { X1, X2, … XR }, y= { Y1, Y2, … YC }, the alignment path between X and Y is:

Because the different time series lengths are not uniform, there may be a large number of different "one-to-many" alignments between them, and thus the alignment path is not unique, the goal of the DTW algorithm is to find an alignment path that minimizes the sum of the distance values between the corresponding data points, and take that sum as the distance value between the series. The distance of the DTW algorithm is defined as follows:

Given the time series x= { X1, X2, … Xm }, y= { Y1, Y2, … Yn }, let a denote the set of all aligned paths, the dynamic time-planning distance between X and Y is as follows:

The DTW algorithm adopts a dynamic programming theory, and the whole solving process is regarded as a multi-stage decision problem. The algorithm starts with the first data point of the two time sequences, and takes the alignment operation between each data point as a decision stage, so that the calculation of the DTW distance value of the original sequence is converted into the calculation of the distance value of the subsequence. The core of the DTW algorithm is to build state transition equations to select the best alignment in each decision stage. The construction method of the state transition equation is as follows:

Where DTW (Xi, yj) represents the DTW distance value required to align the first i data points of time series X with the first j data points of time series Y, and dist (Xi, yj) represents the value of euclidean distance or other distance metric method (e.g., manhattan distance, chebyshev distance, etc.) between the i data points of time series X and the j data points of time series Y.

The state transfer equation may be used to calculate each sub-sequence problem, but since each sub-sequence problem depends on the other three sub-sequence problems, and some sub-sequence problems may be relied on by more than one sub-sequence problem, a large number of sub-sequence problems may be repeatedly calculated by directly recursively calculating the state transfer equation. Thus, the DTW algorithm uses the cumulative distance matrix D to store the values of the sub-sequence problem, thereby avoiding duplicate calculations.

The registration result can be obtained by using the accumulated distance matrix: let N be the number of numerical pairs in the registration result, then P can be denoted/> From the obtained registration result, the DTW distance can be calculated according to the following equation:

wherein N is the number of numerical pairs in the registration result.

After calculating the DTW distance, the evaluation result of the similarity of the target parameters can be calculated according to the following formula:

。

The Procrustes problem generally refers to the problem of matching one object (typically a set of points or a set of shapes) to another object in correspondence in mathematics, using the Procrustes Lu Si s similarity (Procrustes) algorithm. Specifically, given two sets of data points or shapes, the Procludes method aligns one object to the other object by translation, rotation, scaling, etc. (Rigid Body Transformation) to best match them. The term Procrustes includes all classical rigid body movements, as well as the possibility of uniform scaling (stretching or shrinking).

Specifically, the goal of the Procrustes algorithm is to align the two sets of data points so that their average error is minimized. Let X be a matrix of n X d representing n d-dimensional data points and Y be a matrix of n X d representing another set of n d-dimensional data points. The Procrustes algorithm will try to find a rotation matrix R, scaling factor s and translation vector t such that

；

Normalizing X' and Y to obtain X_norm and Y_norm, and lettingAnd then calculating a similarity evaluation result of the target parameters according to the following formula:

wherein +..

The power spectrum density algorithm is to convert time sequences into frequency domains and then calculate the power spectrum density of the time sequences, so that the similarity between the two time sequences can be obtained. In particular, the power spectral densities of two time series may be calculated and then compared for similarity. If the power spectral densities of the two time series are similar, the similarity between them is also relatively high.

The mathematical realization of the power spectral density may be realized using a fourier transform. In particular, converting the time series to the frequency domain may be achieved by fourier transforming it. Then, by calculating the power spectral density of the frequency domain representation, frequency domain features of the time sequence can be obtained.

For a time series x (t) of length N, its Fourier transform is given by:

Where f is the frequency, X (f) is the frequency domain representation after fourier transform and i is the imaginary unit.

Using the frequency domain representation X (f) obtained by fourier transformation, the power spectral density P (f) can be calculated by the following formula:

where X (f) is the amplitude of the frequency domain representation and N is the length of the time sequence.

Assuming that their power spectral density functions are X and Y, respectively, after completion of fourier transform, the similarity evaluation result of the target parameter can be calculated by the following formula.

Wherein: < X, Y > represents the dot product of X and Y, where Y represents the complex conjugate of Y, X represents the norm of X, i.e. the length of vector X, and Y represents the norm of Y, i.e. the length of vector Y.

Smaller similarity values indicate more similarity between the curves, and larger values indicate greater differences between the curves.

The fourier similarity algorithm is an index for comparing two signals or sequences of similarity, based on the principle of fourier transformation. The fourier transform converts the signal from the time domain to the frequency domain, decomposing the signal into a series of combinations of sine and cosine functions, revealing the frequency content of the signal.

In computing the fourier similarity we use the result of the fourier transform, i.e. the spectral representation of the two sequences. Let us assume that we have two sequences X and Y, the spectra obtained after fourier transformation of which are X and Y, respectively.

；

The correlation coefficient method is to calculate a correlation coefficient between two time series, and can be used to evaluate the similarity between them. The correlation coefficient has a value ranging from-1 to 1, where the closer the value is to 1, the more similar the two time series.

In statistics, the correlation coefficient is a measure of the degree of linear correlation between two variables by calculating the covariance. Covariance is a statistic that measures the relationship between two variables and can represent whether the trends in the two variables are similar. Covariance can be calculated by:

Wherein EX represents the expected value of variable X and EY represents the expected value of variable Y. If the covariance is positive, then X and Y are positive correlations; if the covariance is negative, then X and Y are described as being negatively correlated; if the covariance is 0, then it is stated that there is no linear relationship between X and Y.

However, the range of values for covariance is large and depends on the unit and range of variables. Therefore, to compare the degree of correlation between different variables, the covariance is typically normalized to a correlation coefficient, which can be calculated by:

Wherein cov (X, Y) is the covariance between the two real random variables X and Y; std (X) and std (Y) represent standard deviations of variables X and Y, respectively. The formula can ensure that the value range of the correlation coefficient is between-1 and 1, and when the correlation coefficient is 1, the two variables are completely and positively correlated; when the correlation coefficient is-1, the two variables are completely inversely correlated; when the correlation coefficient is 0, it means that there is no linear relationship between the two variables.

Root mean square error (Root Mean Squared Error, RMSE) is a commonly used indicator of the prediction error of a regression model. The calculation formula is as follows:

Wherein n is the number of samples, S is the data obtained by analog simulation, and T is the data obtained by experimental test.

RMSE represents the standard deviation of the difference between the predicted value and the true value. The smaller its value, the better the predictive power of the model. In general, the smaller the RMSE value, the higher the prediction accuracy of the model.

The Relative Bias (Relative Bias) is an index that measures the difference between the simulation data and the experimental data, and can be used to evaluate the accuracy of the simulation data. The relative deviation reflects the average degree of deviation of the simulation data from the experimental data.

The formula for calculating the relative deviation is as follows:

；

Wherein, the BIAS has a value range of (- ≡infinity), ++ infinity A kind of electronic device. If the value of the relative deviation is 0, the predicted value of the simulation data is completely consistent with the true value of the test data; if the value of the relative deviation is larger, the larger the difference between the predicted value of the simulation data and the true value of the test data is, the lower the precision of the simulation data is.

4) The tayer's patch analysis (Theil's U) is a statistical indicator that measures the degree of disagreement between two variables. It is often used to compare differences between predicted and actual values, and can also be used to compare data from different sets or at different points in time.

The calculation of the tayer's patch analysis is based on normalization of the mean square error (Mean Squared Deviation, MSD). The calculation formula of the Taer inconsistency coefficient analysis method is as follows:

；

Wherein the method comprises the steps of For simulation result data, A _i is test data, N is the number of data, and U is more than or equal to 0 and less than or equal to 1. The tel index analysis reflects the consistency degree of the simulation result and the test data, u=0 indicates that the simulation result is completely consistent with the test data, and u=1 indicates that the simulation result and the test difference are the largest.

S105, determining the transient precision of the target parameter by using a gray correlation analysis method based on the fitting degree evaluation result of the target parameter and the similarity evaluation result of the target parameter corresponding to the multiple prediction algorithms.

Gray correlation analysis is a method for comparing correlations between data sequences that can be analyzed with less or incomplete data. The grey association degree is measured according to the relation between the data sequences, and the association degree is calculated by normalizing the sequences and calculating the difference value between each rest evaluation result and the reference sequence. And forming an evaluation matrix by the calculated fitting degree evaluation result of the target parameter and the similarity evaluation result of the target parameter, determining a target evaluation result, then initializing the evaluation matrix, calculating the absolute difference value of each other evaluation result and the target evaluation result one by one to obtain two-stage maximum difference and two-stage minimum difference, then calculating the association coefficient of each other evaluation result and the corresponding element of the target evaluation result respectively, and calculating the weighted average value of the association coefficient of each data in the other evaluation results and the corresponding data of the target evaluation result respectively so as to reflect the association relationship between each other evaluation result and the target evaluation result, namely gray association degree.

In the embodiment of the present invention, the determining, by using a gray correlation analysis method, the transient precision of the target parameter based on the fit evaluation result of the target parameter and the similarity evaluation result of the target parameter corresponding to the plurality of prediction algorithms includes:

and normalizing the fit degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm.

Specifically, normalizing the fit evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm may include:

And initializing a data sequence meeting a first preset condition in the fit degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm.

Wherein the first preset condition includes: the difference between at least one data in the sequence of data and adjacent data is greater than a first value. The first value is a value that characterizes the data sequence as having a significantly increasing and a significantly decreasing data sequence. Specific numerical values may be determined according to actual conditions, and embodiments of the present invention are not limited thereto in particular.

And carrying out averaging treatment on the data sequences meeting the second preset conditions in the fit degree evaluation results of the target parameters corresponding to each prediction algorithm and the similarity evaluation results of the target parameters corresponding to each prediction algorithm. Wherein the second preset condition includes: the difference between the data in the data sequence and the adjacent data is smaller than the second value. The second data is used for representing that the data sequence is a sequence with unobvious ascending and descending trend. Specific numerical values may be determined according to actual conditions, and embodiments of the present invention are not limited thereto in particular.

And determining a target evaluation result from the fit evaluation results of the target parameters and the similarity evaluation results of the target parameters corresponding to the multiple prediction algorithms.

It should be understood that the target evaluation result may be any one of a fit evaluation result of the target parameter and a similarity evaluation result of the target parameter, which is not particularly limited in the embodiment of the present invention. For example, the fitting degree evaluation result of the target parameter determined by the root mean square error method may be a similarity evaluation result of the target parameter determined by the fourier similarity algorithm.

And determining the association coefficient of each remaining evaluation result based on the absolute difference value between the target evaluation result and the remaining evaluation result.

Specifically, when the target evaluation result is the fit evaluation result of the target parameter determined by the root mean square error method, the rest evaluation results are: the method comprises the following steps of determining a fitting degree evaluation result of a target parameter by a correlation coefficient method, determining a fitting degree evaluation result of the target parameter by a relative deviation method, determining a fitting degree evaluation result of the target parameter by a Taer index analysis method, determining a fitting degree evaluation result of the target parameter by a dynamic time-warping similarity algorithm, determining a fitting degree evaluation result of the target parameter by a Program Lu Si S similarity algorithm, determining a fitting degree evaluation result of the target parameter by a power spectral density algorithm and determining a fitting degree evaluation result of the target parameter by a Fourier similarity algorithm.

Based on this, the correlation coefficient of each of the remaining evaluation results includes: the method comprises the following steps of determining a fitting degree evaluation result of a target parameter by a correlation coefficient method, determining a fitting degree evaluation result of the target parameter by a relative deviation method, determining a fitting degree evaluation result of the target parameter by a Taer index analysis method, determining a fitting degree evaluation result of the target parameter by a dynamic time-ordered similarity algorithm, determining a fitting degree evaluation result of the target parameter by a Program Lu Si S similarity algorithm, determining a fitting degree evaluation result of the target parameter by a power spectral density algorithm and determining absolute values of differences between the fitting degree evaluation result of the target parameter by a Fourier similarity algorithm and the fitting degree evaluation result of the target parameter by a root mean square error method.

It should be understood that the fit evaluation result of the target parameter determined by the correlation coefficient method, the fit evaluation result of the target parameter determined by the relative deviation method, the fit evaluation result of the target parameter determined by the tayer inconsistency coefficient analysis method, the fit evaluation result of the target parameter determined by the dynamic time-warping similarity algorithm, the fit evaluation result of the target parameter determined by the puloc Lu Si s similarity algorithm, the fit evaluation result of the target parameter determined by the power spectral density algorithm, the fit evaluation result of the target parameter determined by the fourier similarity algorithm, and the fit evaluation result of the target parameter determined by the root mean square error method are all data sequences, and the number of data in the plurality of data sequences is the same and corresponds to one.

And determining the transient precision of the target parameter based on the association coefficients of the rest evaluation results and the target evaluation result. The association coefficients of the rest evaluation results and the target evaluation result satisfy the following conditions:

；

Wherein, ；/>；，/>As the resolution coefficient, the value range is [0,1], and P ₀ (k) represents the target evaluation result; p _i (k) represents the rest of the evaluation results,/>Representing the absolute value of the difference between the target evaluation result and the rest of the evaluation results,/>Representing the minimum value of the absolute value of the difference between the target evaluation result and the rest of the evaluation results,/>And the maximum value of the absolute value of the difference value between the target evaluation result and the rest evaluation results is represented.

As a possible implementation, the known time sequence { x _t, t=1, 2, …, n } is the original time sequence, and the prediction value at the t-th moment by using the single prediction algorithm is x _it, and m single prediction models (i=1, 2, … m) are used in total. Using m single predictive models and different evaluation criteria, s different combined predictive models are constructed:

；

Wherein, And the weight coefficient vector of the i-th combined prediction model. Generally, the weight coefficient vectors of the s combined prediction models are different, and the prediction results are not identical. And the corresponding predictive accuracy evaluation index values are also generally different. In order to compare the s combined prediction models, one of the optimal combined prediction models is selected, and a prediction accuracy evaluation index system needs to be established.

And the 8 precision evaluation indexes obtained by calculation according to the fit degree evaluation result of the target parameter and the similarity evaluation result of the target parameter reflect the effect of the combined prediction model from different angles. For each combined prediction model, the 8 precision indexes form an evaluation index vector (Pi 1, pi2, pi3, pi4, pi5, pi6, pi7, pi 8), wherein i=1, 2, … s. The vector records the prediction effect of different single-term prediction algorithms. Thus, the evaluation index values of the s different combined prediction models form an evaluation matrix:

。

The fit evaluation result of the target parameter or the similarity evaluation result of the target parameter is as follows: 。

since the index dimensions may be inconsistent, normalization processing of the index data is required. For sequences with a significant increasing, decreasing trend, it can be initialized, i.e.:

where P (k) represents each data in the sequence and P (1) represents the first data in the sequence.

For sequences with insignificant upward and downward trends, a means of averaging may be used, i.e.:

wherein P (k) represents the individual data in the sequence,/> Represents the average of all data in the sequence.

After the normalization of the data is completed, the absolute difference value between each remaining evaluation result and the target evaluation result needs to be calculated one by one, namely:

a two-stage maximum difference and a two-stage minimum difference are then determined. The calculation method comprises the following steps:

Two-stage minimum difference:

Two-stage maximum difference:

And then respectively calculating association coefficients of the rest evaluation results and the target evaluation results:

；

wherein ρ is a resolution coefficient, the range of the value is [0,1], and the smaller the resolution coefficient is, the larger the difference between the correlation coefficients is, and the stronger the distinguishing capability is. Typically, ρ is typically 0.5.

And respectively calculating weighted average values of the association coefficients of the other evaluation results and the corresponding elements of the target evaluation results so as to reflect the association relationship between the other evaluation results and the target evaluation result sequence, and the association relationship is called as association degree. The calculation formula is as follows:

；

And establishing the association sequence degree of each other evaluation result according to the gray weighted association degree. The higher the association degree, the higher the importance degree of the rest of the evaluation results to the evaluation standard.

In the embodiment of the invention, taking a certain group of time sequences generated in the simulation of the space engine and time sequences obtained by the actual physical process as examples, a data set containing a large number of data samples is obtained. Some of the samples are shown below:

Table 1 partial test data and simulation data

In this set of data, the time series generated from the space engine simulation process contains about 2 tens of thousands of samples due to the difference in sampling frequency, and the time series from the actual physical process record contains about 10 tens of thousands of samples. Each sample recorded data about the space motor.

Because negative time exists in the simulation process, samples of the negative time are deleted when the data is processed, so that the accuracy and consistency of the data are ensured.

Since the simulation data generally contains some noise, and the simulation data itself may have large fluctuations or abrupt changes, the simulation data needs to be subjected to filtering processing. Filtering can smooth data, remove noise, and make the trend of data change more obvious and visible. Common filtering methods include moving average filtering, median filtering, low pass filtering, high pass filtering, and the like. In the embodiment of the invention, a moving average filtering method is adopted to process the simulation data. Moving average filtering is a simple and common filtering method that smoothes data by calculating the average of the data within a window. Specifically, the embodiment of the invention uses a moving average filter with a window size of 21, and the filtered data is obtained by applying the filter to the simulation data. The following is information of the filter:

table 2 information of filter

The embodiment of the invention divides the time sequence into a plurality of steady-state data segments and transient data segments by using a sliding window method for the simulation data subjected to the filtering processing. By calculating the standard deviation of the observations within the window, the start and end points of the steady-state data segment can be determined. The function adopts a sliding window mode, slides on the time sequence by taking step length as a unit, and calculates the standard deviation of the observed value in the window. When the standard deviation of the observed values in the window is smaller than the threshold value of 0.001 and is not in a steady state before, marking the current position as a starting point of a steady-state data segment; when the standard deviation of the observations within the window is equal to or greater than the threshold value and is previously in steady state, the current location is marked as the end point of the transient data segment. Finally, the function returns a list of starting and ending points for all steady-state data segments.

Through detection by a sliding window method, the time sequence is divided into a steady-state data segment and a transient data segment. In steady state data segments, we observe the following data segments: [1150,1260], [1410,50290], [50780,99985]. These data segments represent that the time series values within these ranges are relatively stable with no significant changes or mutations. In the transient data segment, the following intervals can be observed: [1,1150], [1260, 1410], [50290, 50780]. These data segments represent time series that have undergone large changes or mutations within these ranges, possibly due to external factors or system characteristics. By identifying and separating steady-state intervals and transient intervals, we can better understand the different features and behavior of time series data. This also corresponds well to the three phases of start-up, water hammer and shut-down of the space motor.

Because the sampling interval of the real data is inconsistent with that of the simulation data, the time range corresponding to the transient data segment in the real time sequence can be found, and then the simulation result in the time range is found. The lengths of the two sequences are almost impossible to agree with each other, and thus interpolation processing is required for the simulation sequences. Common interpolation methods include linear interpolation, polynomial interpolation, spline interpolation, and the like. Spline interpolation is selected herein to fill in missing data points in the simulation sequence. Spline interpolation approximates missing data points using polynomial functions between local data segments to fit the entire sequence smoothly. By using spline interpolation methods, missing values in the simulation sequence can be supplemented as accurately as possible while maintaining data continuity and smoothness.

It is noted that since the DTW algorithm is itself applicable to cases where the two time series are not equal in length, there is no need to interpolate the data before calculating the DTW similarity of the two time series.

After interpolation is finished, two equal-length time sequences can be obtained, and then the simulation precision of each transient interval can be estimated according to various prediction algorithms.

The evaluation matrix obtained by the calculation is as follows:

after each column of the evaluation matrix is averaged, the obtained dimensionless evaluation matrix is as follows:

a column is selected as a parent sequence of the evaluation matrix, wherein the fourier similarity is taken as the parent sequence, and the absolute difference between each evaluation object and the parent sequence is calculated one by one, and the calculation result is as follows:

Two-stage maximum difference Δmax= 1.247 and two-stage minimum difference Δmin=0.004 are determined. The correlation coefficient of each comparison sequence and the corresponding element of the parent sequence is calculated as follows:

and removing the parameter with the highest grey correlation degree for each transient interval, and comprehensively evaluating the rest parameters.

All precision evaluation parameters are divided into two sequences, m and n, according to the correlation of the values and the precision. Namely the sequence m: comprising m evaluation parameters positively correlated with the accuracy. The higher the values of these evaluation parameters, the higher the corresponding accuracy. Such as: fourier similarity algorithms, correlation coefficient methods, and the like.

Sequence n: comprising n evaluation parameters inversely related to the accuracy. The lower the values of these evaluation parameters, the higher the corresponding accuracy. Such as: root mean square error method, relative deviation method, and tayer's Path index analysis method.

The calculation is performed as follows:

；

The comprehensive precision a1=0.914 of the interval [1,1150] corresponding to the starting-up stage, the comprehensive precision a2=0.814 of the interval [1260,1410] corresponding to the water attack stage and the comprehensive precision a3=0.771 of the interval [50290,50780] corresponding to the shutdown stage are obtained through calculation. Because A1> A2> A3, the comprehensive precision of the simulation model corresponding to the start-up stage can be considered to be highest, and the comprehensive precision of the simulation model corresponding to the shut-down stage is lower, the main reason is probably that the shut-down time in the simulation model is advanced by tens of milliseconds, so that the simulation sequence and the numerical value of the real sequence have larger difference. Based on these results, when performing model modification and parameter improvement, modification of the simulation model corresponding to the shutdown phase should be considered first. The overall accuracy of the shutdown phase is low, and there may be inconsistencies between the model and the actual data. Therefore, it is a key to improve the overall simulation accuracy to adjust the simulation model for the section. In addition, the exploration of time domain errors should be particularly enhanced, because time domain errors are one of the main reasons for low simulation accuracy in the shutdown phase. By deeply analyzing the time domain error of the shutdown phase, the difference between the model and the actual data can be identified, and the model modification and the parameter improvement can be performed in a targeted manner. By modifying the simulation model at the shutdown stage and exploring the time domain error, the accuracy and the reliability of the model at the shutdown stage can be improved, so that the real situation can be better reflected. Such adjustments help to improve the performance of the model in steady-state intervals and enhance the performance and reliability of the overall simulation system. Meanwhile, the improvement of the shutdown stage can also positively influence the overall simulation precision, and the application value of the model in actual operation is improved.

By adaptively adjusting the simulation model of the interval, the characteristics and time domain errors of the shutdown stage can be optimized. The time domain error can be reduced by adjusting the simulation mode of the shutdown time, optimizing the model parameters of the shutdown transition period, enhancing the error compensation in the shutdown process and the like.

In a word, the performance of the simulation model in the shutdown stage is improved by a proper method, and the overall reliability and accuracy of the engine simulation model are improved by improving the comprehensive precision of the shutdown stage.

In a second aspect, referring to fig. 2, the present invention provides an accuracy evaluation device of an engine simulation model, the accuracy evaluation device of an engine simulation model including:

an acquisition module 201, configured to acquire target parameter simulation data and target parameter test data of the engine;

The data determining module 202 is configured to determine the target parameter simulation data and the target parameter test data as target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data by using a first preset manner;

The steady-state precision determining module 203 is configured to determine, according to a second preset manner, a steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter;

the transient evaluation determining module 204 is configured to determine, based on the target parameter transient simulation data and the target parameter transient test data, a fit evaluation result of the target parameter and a similarity evaluation result of the target parameter under each of the prediction algorithms by using a plurality of prediction algorithms;

the transient precision determining module 205 is configured to determine, according to a gray correlation analysis method, a transient precision of the target parameter based on a fit degree evaluation result of the target parameter and a similarity evaluation result of the target parameter corresponding to the multiple prediction algorithms.

The first preset mode comprises a sliding window algorithm;

The second preset mode comprises the following steps: and determining the steady-state precision of the target parameter by utilizing the difference value between the steady-state simulation data of the target parameter and the steady-state test data of the target parameter.

Further, the plurality of prediction algorithms include a correlation coefficient method, a root mean square error method, a relative deviation method and a tayer index analysis method for determining the fit degree evaluation result of the target parameter, and a dynamic time-ordered similarity algorithm, a puloc Lu Si s similarity algorithm, a power spectral density algorithm and a fourier similarity algorithm for determining the similarity evaluation result of the target parameter.

The transient evaluation determination module 204 includes:

and the normalization processing unit is used for carrying out normalization processing on the fitting degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm.

And the target evaluation result determining unit is used for determining one target evaluation result from the fit evaluation results of the target parameters and the similarity evaluation results of the target parameters corresponding to the plurality of prediction algorithms.

And the association coefficient determining unit is used for determining the association coefficient of each other evaluation result and the target evaluation result based on the absolute difference value between the target evaluation result and the other evaluation results.

And the transient precision determining unit is used for determining the transient precision of the target parameter based on the association coefficients of the rest evaluation results and the target evaluation result.

The normalization processing unit includes:

and the initialization unit is used for initializing the data sequence meeting the first preset condition in the fit degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm.

And the averaging unit is used for carrying out averaging processing on the data sequences meeting the second preset conditions in the fitting degree evaluation result of the target parameters corresponding to each prediction algorithm and the similarity evaluation result of the target parameters corresponding to each prediction algorithm.

In a third aspect, an embodiment of the present invention further provides an electronic device, including: one or more processors; and one or more machine readable media having instructions stored thereon that, when executed by the one or more processors, cause performance of the method of precision assessment of an engine simulation model of the first aspect.

The electronic device in the embodiment of the invention can be a device, a component in a terminal, an integrated circuit, or a chip. The device may be a mobile electronic device or a non-mobile electronic device. By way of example, the mobile electronic device may be a mobile phone, a tablet computer, a notebook computer, a palm computer, a vehicle-mounted electronic device, a wearable device, an ultra-mobile personal computer (UMPC), a netbook or a Personal Digital Assistant (PDA), etc., and the non-mobile electronic device may be a server, a network attached storage (Network Attached Storage, NAS), a personal computer (personal computer, PC), a Television (TV), a teller machine, a self-service machine, etc., and the embodiments of the present invention are not limited in particular.

The electronic device in the embodiment of the invention can be a device with an operating system. The operating system may be an Android operating system, an ios operating system, or other possible operating systems, and the embodiment of the present invention is not limited specifically.

Fig. 3 shows a schematic hardware structure of an electronic device according to an embodiment of the present invention. As shown in fig. 3, the electronic device 400 includes a processor 410.

As shown in FIG. 3, the processor 410 may be a general purpose central processing unit (central processing unit, CPU), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present invention.

As shown in fig. 3, the electronic device 400 may also include a communication line 440. Communication line 440 may include a path to communicate information between the above-described components.

Optionally, as shown in fig. 3, the electronic device may further include a communication interface 420. The communication interface 420 may be one or more. Communication interface 420 may use any transceiver-like device for communicating with other devices or communication networks.

Optionally, as shown in fig. 3, the electronic device may also include a memory 430. Memory 430 is used to store computer-executable instructions for performing aspects of the present invention and is controlled by the processor for execution. The processor is configured to execute computer-executable instructions stored in the memory, thereby implementing the method provided by the embodiment of the invention.

As shown in fig. 3, the memory 430 may be, but is not limited to, a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (random access memory, RAM) or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (ELECTRICALLY ERASABLE PROGRAMMABLE READ-only memory, EEPROM), a compact disc read-only memory (compact disc read-only memory) or other optical disc storage, optical disc storage (including compact disc, laser disc, optical disc, digital versatile disc, blu-ray disc, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory 430 may be stand alone and be coupled to the processor 410 via a communication line 440. Memory 430 may also be integrated with processor 410.

Alternatively, the computer-executable instructions in the embodiments of the present invention may be referred to as application program codes, which are not particularly limited in the embodiments of the present invention.

In a particular implementation, as one embodiment, as shown in FIG. 3, processor 410 may include one or more CPUs, such as CPU0 and CPU1 in FIG. 3.

In a specific implementation, as an embodiment, as shown in fig. 3, the terminal device may include a plurality of processors, such as the first processor 4101 and the second processor 4102 in fig. 3. Each of these processors may be a single-core processor or a multi-core processor.

Fig. 4 is a schematic structural diagram of a chip according to an embodiment of the present invention. As shown in fig. 4, the chip 500 includes one or more (including two) processors 410.

Optionally, as shown in fig. 4, the chip further includes a communication interface 420 and a memory 430, and the memory 430 may include a read-only memory and a random access memory, and provides operation instructions and data to the processor. A portion of the memory may also include non-volatile random access memory (non-volatile random access memory, NVRAM).

In some implementations, as shown in FIG. 4, memory 430 stores elements, execution modules or data structures, or a subset thereof, or an extended set thereof.

In the embodiment of the present invention, as shown in fig. 4, by calling the operation instruction stored in the memory (the operation instruction may be stored in the operating system), the corresponding operation is performed.

As shown in fig. 4, the processor 410 controls the processing operations of any one of the terminal devices, and the processor 410 may also be referred to as a central processing unit (central processing unit, CPU).

As shown in fig. 4, memory 430 may include read only memory and random access memory, and provides instructions and data to the processor. A portion of the memory 430 may also include NVRAM. Such as a memory, a communication interface, and a memory coupled together by a bus system that may include a power bus, a control bus, a status signal bus, etc., in addition to a data bus. The various buses are labeled as bus system 540 in fig. 4 for clarity of illustration.

As shown in fig. 4, the method disclosed in the above embodiment of the present invention may be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or by instructions in the form of software. The processor may be a general purpose processor, a digital signal processor (DIGITAL SIGNAL processing, DSP), an ASIC, an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be embodied directly in the execution of a hardware decoding processor, or in the execution of a combination of hardware and software modules in a decoding processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory, and the processor reads the information in the memory and, in combination with its hardware, performs the steps of the above method.

In one aspect, a computer readable storage medium is provided, in which instructions are stored, which when executed, implement the functions performed by the terminal device in the above embodiments.

In one aspect, a chip is provided, where the chip is applied to a terminal device, and the chip includes at least one processor and a communication interface, where the communication interface is coupled to the at least one processor, and the processor is configured to execute instructions to implement the functions performed by the clock tree generating method in the foregoing embodiment.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present invention are performed in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, a terminal, a user equipment, or other programmable apparatus. The computer program or 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 program or instructions may be transmitted from one website site, computer, server, or data center of gravity to another website site, computer, server, or data center of gravity by wired or wireless means. The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center of gravity, etc. that integrates one or more available media. The usable medium may be a magnetic medium, e.g., floppy disk, hard disk, tape; but also optical media such as digital video discs (digital video disc, DVD); but also semiconductor media such as Solid State Drives (SSDs) STATE DRIVE.

Although the invention is described herein in connection with various embodiments, other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the "a" or "an" does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

1. The method for evaluating the precision of the engine simulation model is characterized by comprising the following steps of:

acquiring target parameter simulation data and target parameter test data of the engine;

Determining the target parameter simulation data and the target parameter test data as target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data by using a first preset mode; the first preset mode comprises a sliding window algorithm; the target parameter steady-state simulation data comprises a plurality of target parameter steady-state simulation data segments; the target parameter steady-state test data comprises a plurality of target parameter steady-state test data segments, the target parameter transient simulation data comprises a plurality of target parameter transient simulation data segments, and the target parameter transient test data comprises a plurality of target parameter transient test data segments; the target parameter steady-state simulation data segments, the target parameter transient simulation data segments and the target parameter transient test data segments are in one-to-one correspondence;

According to a second preset mode, determining the steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter; the second preset mode comprises the following steps: determining the steady-state precision of the target parameter by utilizing the difference value between the steady-state simulation data of the target parameter and the steady-state test data of the target parameter;

Determining a fitting degree evaluation result of the target parameter and a similarity evaluation result of the target parameter under each prediction algorithm by utilizing a plurality of prediction algorithms based on the target parameter transient simulation data and the target parameter transient test data; the plurality of prediction algorithms comprise a correlation coefficient method, a root mean square error method, a relative deviation method and a Taer inconsistency coefficient method for determining the fit degree evaluation result of the target parameter, and a dynamic time regular similarity algorithm, a Programm Lu Si s similarity algorithm, a power spectrum density algorithm and a Fourier similarity algorithm for determining the similarity evaluation result of the target parameter;

2. The method for evaluating the accuracy of an engine simulation model according to claim 1, wherein the determining the transient accuracy of the target parameter by using a gray correlation analysis method based on the fit evaluation result of the target parameter and the similarity evaluation result of the target parameter corresponding to the plurality of prediction algorithms comprises:

3. The accuracy evaluation method of an engine simulation model according to claim 2, characterized in that: the normalizing the fit degree evaluation result of the target parameter corresponding to each prediction algorithm and the similarity evaluation result of the target parameter corresponding to each prediction algorithm comprises the following steps:

4. The method for evaluating the accuracy of an engine simulation model according to claim 3, wherein the first preset condition includes: the difference between at least one data in the data sequence and adjacent data is greater than a first value;

5. The method for evaluating the accuracy of an engine simulation model according to claim 2, wherein the correlation coefficient of the remaining evaluation results and the target evaluation result satisfies:

；

6. An accuracy evaluation device of an engine simulation model, characterized in that the accuracy evaluation device of an engine simulation model comprises:

The acquisition module is used for acquiring target parameter simulation data and target parameter test data of the engine;

The data determining module is used for determining the target parameter simulation data and the target parameter test data into target parameter steady-state simulation data, target parameter steady-state test data, target parameter transient simulation data and target parameter transient test data by utilizing a first preset mode; the first preset mode comprises a sliding window algorithm; the target parameter steady-state simulation data comprises a plurality of target parameter steady-state simulation data segments; the target parameter steady-state test data comprises a plurality of target parameter steady-state test data segments, the target parameter transient simulation data comprises a plurality of target parameter transient simulation data segments, and the target parameter transient test data comprises a plurality of target parameter transient test data segments; the target parameter steady-state simulation data segments, the target parameter transient simulation data segments and the target parameter transient test data segments are in one-to-one correspondence;

the steady-state precision determining module is used for determining the steady-state precision of the target parameter based on the steady-state simulation data of the target parameter and the steady-state test data of the target parameter according to a second preset mode; the second preset mode comprises the following steps: determining the steady-state precision of the target parameter by utilizing the difference value between the steady-state simulation data of the target parameter and the steady-state test data of the target parameter;

The transient evaluation determining module is used for determining a fitting degree evaluation result of the target parameter and a similarity evaluation result of the target parameter under each prediction algorithm by utilizing a plurality of prediction algorithms based on the target parameter transient simulation data and the target parameter transient test data; the plurality of prediction algorithms comprise a correlation coefficient method, a root mean square error method, a relative deviation method and a Taer inconsistency coefficient method for determining the fit degree evaluation result of the target parameter, and a dynamic time regular similarity algorithm, a Programm Lu Si s similarity algorithm, a power spectrum density algorithm and a Fourier similarity algorithm for determining the similarity evaluation result of the target parameter;

7. An electronic device, comprising: one or more processors; and one or more machine readable media having instructions stored thereon that, when executed by the one or more processors, cause performance of the accuracy assessment method of the engine simulation model of any of claims 1-5.