A method, system, device, and computer-readable storage medium for predicting engine remaining life based on KNN.

Abstract

This specification discloses a method, system, device, and computer-readable storage medium for predicting engine remaining life based on KNN (Knowledge, Neural Networks) technology. It relates to the field of engine remaining life and prediction, aiming to solve the problem of hyperparameter configuration optimization in existing KNN prediction models. The method includes: acquiring and preprocessing data; calculating correlation coefficients between each preprocessed data point and its corresponding remaining life data, and selecting parameters; defining a hyperparameter grid; optimizing the grid parameters to establish an optimal KNN model; and predicting the engine's remaining life. This invention combines distance correlation coefficients with an automated hyperparameter optimization algorithm based on grid search to determine the optimal K value and distance metric of the KNN model. This enables faster finding of the optimal sensor parameter combination, achieving automated hyperparameter optimization of the KNN model, thereby improving the performance and reliability of the prediction model.

Description

TECHNICAL FIELD

[0001] The present document belongs to the field of engine life prediction, and particularly relates to an engine residual life prediction method, system, device and computer readable storage medium based on KNN. BACKGROUND

[0002] In the industrial field such as aviation, the performance and reliability of the engine are key factors to ensure safety, and engine health management and predictive maintenance are of great significance to reduce operating costs and improve safety.

[0003] KNN (K-Nearest Neighbors) algorithm is an instance-based learning algorithm, which has been widely used in many fields, especially in classification and regression problems. Its basic principle is to measure the distance between different data points to predict the class or attribute value of new data points. For engine residual life prediction problem, KNN algorithm can effectively analyze historical data, identify patterns and trends, and thus predict the residual service life of the engine. Therefore, it is of great engineering significance to study the KNN algorithm in engine residual life prediction. By optimizing the hyperparameter configuration of KNN, the prediction accuracy can be improved, the maintenance cost can be reduced, and unexpected failures can be avoided, thereby improving the operating efficiency and safety of the engine.

[0004] In practical applications, the simplicity and intuitiveness of KNN algorithm make it easy to implement and understand, but KNN algorithm is very sensitive to the quality and scale of data, so when using KNN algorithm for engine residual life prediction, how to preprocess the data and select the hyperparameters to get more accurate residual life prediction value is a key problem of the prediction model. SUMMARY

[0005] The present specification provides an engine residual life prediction method, system, device and computer readable storage medium based on KNN, to solve the problem of how to preprocess the data and select the hyperparameters to get more accurate residual life prediction value in the prior art, i.e. the problem of optimizing the hyperparameter configuration of the existing KNN prediction model.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0007] The first aspect of the present specification provides an engine residual life prediction method based on KNN, which comprises the following steps:

[0008] Step S1, acquiring sensor parameter data of the engine as training data to construct a training set; preprocessing the sensor parameter data of the engine in the training set to obtain preprocessed data;

[0009] Step S2, correlation coefficient calculation is performed on each pre-processed data and its corresponding residual life data, and pre-processed data with a correlation coefficient greater than a set correlation coefficient threshold is added to the pre-constructed optimal parameter set;

[0010] Step S3, obtaining the hyperparameters corresponding to each pre-processed data in the optimal parameter set, and constructing a hyperparameter grid; the hyperparameters include K value, distance metric and weight function;

[0011] Step S4, traversing the hyperparameter grid, substituting each group of hyperparameters in the hyperparameter grid into the pre-constructed KNN model for grid parameter optimization to obtain an optimal hyperparameter combination; and substituting the optimal hyperparameter combination into the KNN model as an optimal KNN model;

[0012] Step S5, obtaining monitoring data of an engine to be used for residual life prediction; inputting the monitoring data of the engine to be used for residual life prediction into the optimal KNN model to obtain a prediction result of the engine residual life.

[0013] In some preferred embodiments, the pre-processing includes normalization processing on the training set, and mapping the data values of the engine sensor parameters to between [0, 1].

[0014] In some preferred embodiments, the correlation coefficient calculation is performed on each pre-processed data and its corresponding residual life data by the following method:

[0015] The distance correlation coefficient is used to calculate the correlation coefficient of each pre-processed data and its corresponding residual life data.

[0016] In some preferred embodiments, the distance correlation coefficient is calculated by the following method:

[0017] The sensor parameter data X and the residual life data Y are defined as pairs of continuous variables with a length of n, and the distance between all elements is:

[0018] a j,k =‖X j -X k ‖,j,k=1,2,...,n

[0019] b j,k =‖Y j -Y k ‖,j,k=1,2,...,n

[0020] Wherein, a j,k , b j,k respectively represent the distance matrix between each element of X and Y;

[0021] Calculate the center distance matrix:

[0022]

[0023] where A j,k , B j,k are the center distance matrices of X and Y, respectively, is the average of the jth row of the distance matrix of X, is the average of the kth column of the distance matrix of X, is the overall average of the distance matrix of X; is the average of the jth row of the distance matrix of Y, is the average of the kth column of the distance matrix of Y, is the overall average of the distance matrix of Y;

[0024] The distance covariance is calculated as

[0025]

[0026] The distance variances of X and Y are calculated as

[0027]

[0028]

[0029] where, are the distance variances of X and Y, respectively;

[0030] Finally, the distance correlation coefficient is calculated based on the distance variances and the distance covariance.

[0031] In some preferred embodiments, the distance correlation coefficient is calculated by:

[0032]

[0033] where, is the distance correlation coefficient, is the distance covariance between X and Y, are the distance variances of X and Y, respectively.

[0034] In some preferred embodiments, the weight function includes uniform weight or distance inverse weight, where the distance inverse weight selects Euclidean distance, Manhattan distance, or Chebyshev distance as the distance metric.

[0035] In some preferred embodiments, the grid search is performed by:

[0036] Traverse the defined hyperparameter grid, for each set of hyperparameters, use the cross-validation method to divide the training set into multiple subsets, train the KNN model, and record the performance indicators of each model, get the best performance hyperparameter combination, determine the optimal K value, and establish the optimal KNN model.

[0037] In a second aspect of the present specification, a KNN-based engine remaining life prediction system comprises the following modules:

[0038] The acquisition module is configured to obtain sensor parameter data of the engine as training data to construct a training set; and pre-process the sensor parameter data of the engine in the training set to obtain pre-processed data.

[0039] The construction module is configured to calculate the correlation coefficient of each pre-processed data and its corresponding remaining life data, and add the pre-processed data with a correlation coefficient greater than a set correlation coefficient threshold to the pre-constructed optimal parameter set.

[0040] The optimization module is configured to obtain the hyperparameters corresponding to each pre-processed data in the optimal parameter set, construct a hyperparameter grid, traverse the hyperparameter grid, and substitute each set of hyperparameters in the hyperparameter grid into the pre-constructed KNN model for grid parameter optimization to obtain the optimal hyperparameter combination and substitute it into the KNN model as an optimal KNN model.

[0041] The prediction module is configured to obtain monitoring data of an engine to be predicted for remaining life and input the monitoring data into the optimal KNN model to obtain a prediction result of the remaining life of the engine.

[0042] The verification module is configured to evaluate the prediction result.

[0043] In a third aspect of the present specification, a KNN-based engine remaining life prediction device comprises:

[0044] At least one processor; and

[0045] A memory communicatively connected to the at least one processor; wherein

[0046] The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the KNN-based engine remaining life prediction method.

[0047] In a fourth aspect of the present specification, a computer readable storage medium stores computer instructions, and the computer instructions are used to be executed by the processor to implement the KNN-based engine remaining life prediction method.

[0048] The at least one technical scheme adopted by the embodiments of the present specification can achieve the following beneficial effects:

[0049] 1) By combining the distance correlation coefficient with the KNN model, the remaining life of the engine is more accurately predicted, the classification accuracy and calculation efficiency of the KNN model are improved, the optimal sensor parameter combination can be found more quickly, and the automatic hyperparameter optimization of the KNN model is realized, thereby improving the performance and reliability of the prediction model in practical applications.

[0050] 2) By combining the distance correlation coefficient and the automatic hyperparameter optimization algorithm based on grid search, the optimal K value and distance measure of the KNN model are determined, the hyperparameter configuration of the KNN is optimized, and the selection efficiency and prediction accuracy of the KNN algorithm are improved. BRIEF DESCRIPTION OF DRAWINGS

[0051] The accompanying drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings:

[0052] Figure 1 A flowchart of the KNN-based engine remaining life prediction method provided by the embodiments of the present specification. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described in detail below in combination with specific embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0054] The technical scheme provided by the embodiments of the present application will be described in detail below in combination with the drawings.

[0055] The present specification combines the distance correlation coefficient with the KNN model to find the optimal sensor parameter combination more quickly, realizes the automatic hyperparameter optimization of the KNN model, and thereby improves the performance and reliability of the prediction model in practical applications. Specifically as follows:

[0056] The first embodiment of the present specification provides a KNN-based engine remaining life prediction method, which includes steps S1-S5, as shown in Figure 1 The steps are described in detail as follows:

[0057] Step S1, obtaining sensor parameter data of the engine as training data to construct a training set; preprocessing the sensor parameter data of the engine in the training set to obtain preprocessed data.

[0058] Preferably, the preprocessing method is:

[0059] The engine sensor parameter data in the training set is normalized, that is, standardized by deviation, linearly transformed, and mapped to [0, 1].

[0060] Further, in the embodiment, the specific calculation formula of the normalization processing is as follows:

[0061]

[0062] Wherein, x i is the i-th sensor parameter, max is the maximum value of the sensor parameter x i , and min is the minimum value of the sensor parameter x i .

[0063] Step S2, calculating the correlation coefficient of each preprocessed data and its corresponding remaining life data, and adding the preprocessed data with a correlation coefficient greater than a set correlation coefficient threshold to the pre-constructed optimal parameter set;

[0064] The distance correlation coefficient is used to calculate each engine sensor parameter data and the remaining life data, and the method is:

[0065] Define the sensor parameter data X and the remaining life data Y as pairs of continuous variables with a length of n, and the distance of all elements is:

[0066] a j,k =‖X j -X k ‖,j,k=1,2,...,n

[0067] b j,k =‖Y j -Y k ‖,j,k=1,2,...,n

[0068] Wherein, a j,k , b j,k respectively represent the distance matrix between the elements of X and Y;

[0069] Calculate the center distance matrix:

[0070]

[0071] Wherein, A j,k , B j,krespectively the center distance matrix of X and Y, is the average of the jth row of the distance matrix of X, is the average of the kth column of the distance matrix of X, is the total average of the distance matrix of X; is the average of the jth row of the distance matrix of Y, is the average of the kth column of the distance matrix of Y, is the total average of the distance matrix of Y;

[0072] Calculate the distance covariance

[0073]

[0074] Calculate the distance variance of X and Y respectively:

[0075]

[0076] wherein, is the distance variance of X and Y respectively;

[0077] Finally, based on the distance variance, the distance covariance, the distance correlation coefficient is calculated, and the distance correlation coefficient is:

[0078]

[0079] wherein, is the distance correlation coefficient, is the distance covariance between X and Y, (Y) is the distance variance of X and Y respectively.

[0080] In this embodiment, it is preferred to calculate the 20 sensor parameter data of certain engine Mach number, throttle lever angle, fan inlet temperature and the corresponding residual life value thereof, and the optimal parameter set obtained according to the distance correlation coefficient calculation result is {Mach number, throttle lever angle, fan inlet temperature, low pressure turbine outlet temperature, fan inlet pressure, outer duct total pressure, engine pressure ratio, high pressure compressor outlet static pressure, fuel flow and P30 ratio, core engine corrected speed, bleed air enthalpy value}.

[0081] Step S3, obtaining the hyperparameters corresponding to each preprocessed data in the optimal parameter set, constructing a hyperparameter grid; the hyperparameters include K value, distance measure and weight function;

[0082] In this embodiment, it is preferred that the K value is in the range of [1, 6].

[0083] Preferably, the weight function comprises uniform weight or distance reciprocal weight, wherein the distance metric selected by the distance reciprocal weight is Euclidean distance, Manhattan distance or Chebyshev distance.

[0084] Step S4, traverse the hyperparameter grid, and substitute each set of hyperparameters in the hyperparameter grid into the pre-constructed KNN model for grid parameter optimization to obtain an optimal hyperparameter combination; and substitute the optimal hyperparameter combination into the KNN model as an optimal KNN model.

[0085] Preferably, the grid optimization method is as follows:

[0086] Traverse the defined hyperparameter grid, for each set of hyperparameters, use the cross-validation method to divide the training set into multiple subsets, train the KNN model, and record the performance indicators of each model to obtain the best-performing hyperparameter combination, determine the optimal K value, and establish an optimal KNN model.

[0087] Preferably, in this embodiment, the performance indicators are accuracy, recall rate and precision.

[0088] In this embodiment, according to the optimal parameter set described above, preferably K = 2, and the weight function is uniform weight.

[0089] Step S5, based on the monitoring data of the engine to be used for remaining life prediction, a test set is constructed; and the test set is input into the optimal KNN model to obtain a predicted value of the remaining life of the engine.

[0090] Preferably, after obtaining the predicted value, the prediction effect is evaluated based on the root mean square error RMSE.

[0091] The second embodiment of the present specification provides a KNN-based engine remaining life prediction system, comprising the following modules:

[0092] The acquisition module is configured to obtain sensor parameter data of the engine as training data to construct a training set; and pre-process the sensor parameter data of the engine in the training set to obtain pre-processed data.

[0093] The construction module is configured to calculate the correlation coefficient of each pre-processed data and its corresponding remaining life data using a distance correlation coefficient, and add the pre-processed data with a correlation coefficient greater than a set correlation coefficient threshold to the pre-constructed optimal parameter set.

[0094] The optimization module is configured to obtain the hyperparameters corresponding to each pre-processed data in the optimal parameter set, construct a hyperparameter grid, traverse the hyperparameter grid, substitute each set of hyperparameters in the hyperparameter grid into the pre-constructed KNN model for grid parameter optimization to obtain an optimal hyperparameter combination, and substitute the optimal hyperparameter combination into the KNN model as an optimal KNN model.

[0095] a prediction module configured to acquire monitoring data of an engine to be subjected to remaining life prediction and input the monitoring data to the optimal KNN model to obtain a prediction result of the remaining life of the engine;

[0096] a verification module configured to evaluate the prediction result.

[0097] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related description of the system described above can refer to the corresponding process in the method embodiments, which will not be repeated here.

[0098] It should be noted that the above embodiments provide a KNN-based engine remaining life prediction system, and only the division of the above functional modules is exemplified, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the modules or steps in the embodiments of the present application are further decomposed or combined, for example, the modules of the above embodiments can be combined into one module, or can be further split into multiple sub-modules to complete all or part of the functions described above. The names of the modules involved in the embodiments of the present application are only for distinguishing each module, and are not considered as improper limitation of the present application.

[0099] The third embodiment of the present specification provides a KNN-based engine remaining life prediction device, comprising:

[0100] at least one processor; and

[0101] a memory in communication connection with the at least one processor; wherein

[0102] the memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned KNN-based engine remaining life prediction method.

[0103] The fourth embodiment of the present specification provides a computer readable storage medium, the computer readable storage medium stores computer instructions, and the computer instructions are used to be executed by the processor to implement the above-mentioned KNN-based engine remaining life prediction method.

[0104] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related description of the above-mentioned device and computer readable storage medium can refer to the corresponding process in the foregoing method examples, which will not be repeated here.

[0105] The above merely provides an example of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application should be included in the scope of claims of the present application.

Claims

1. A KNN-based engine remaining life prediction method, characterized by, The method comprises the following steps: Step S1, obtaining sensor parameter data of an engine as training data to construct a training set; The sensor parameter data of the engine in the training set is preprocessed to obtain preprocessed data; Step S2, calculating the correlation coefficient of each preprocessed data and its corresponding remaining life data, and adding the preprocessed data with a correlation coefficient greater than a set correlation coefficient threshold to a pre-constructed optimal parameter set; The correlation coefficient of each preprocessed data and its corresponding remaining life data is calculated by the following method: The correlation coefficient of each preprocessed data and its corresponding remaining life data is calculated by the following method: Step S3, obtaining the hyperparameters corresponding to each preprocessed data in the optimal parameter set to construct a hyperparameter grid; the hyperparameters include K value, distance metric and weight function; Step S4, traversing the hyperparameter grid, substituting each group of hyperparameters in the hyperparameter grid into the pre-constructed KNN model for grid parameter optimization to obtain an optimal hyperparameter combination; and substituting the optimal hyperparameter combination into the KNN model as an optimal KNN model; Step S5, obtaining monitoring data of an engine to be predicted for remaining life; and inputting the monitoring data of the engine to be predicted for remaining life into the optimal KNN model to obtain a prediction result of the remaining life of the engine.

2. The KNN-based engine remaining useful life prediction method of claim 1, wherein, The preprocessing includes normalizing the training set to map the data values of the sensor parameter data of the engine to between 0 and 1.

3. The KNN-based engine remaining useful life prediction method of claim 1, wherein, The correlation coefficient is calculated by the following method: The sensor parameter data X and the remaining life data Y are defined as pairs of continuous variables with a length of n, and the distance between all elements is: a j,k =‖X j -X k ‖,j,k=1,2,...,n b j,k =‖Y j -Y k ‖,j,k=1,2,...,n where a j,k , b j,k denote the distance matrix between the elements of X and Y, respectively. The center distance matrix is calculated: where A j,k , B j,k denote the center distance matrix of X and Y, respectively, is the average of the jth row of the distance matrix of X, is the average of the kth column of the distance matrix of X, is the overall average of the distance matrix of X; is the average of the jth row of the distance matrix of Y, is the average of the kth column of the distance matrix of Y, is the overall average of the distance matrix of Y; Computing distance covariance The distance variance of X and Y is calculated respectively: wherein respectively the distance variance of X and Y; Finally, the distance correlation coefficient is calculated based on the distance variance and distance covariance.

4. The KNN-based engine remaining useful life prediction method of claim 3, wherein, The method for calculating the distance correlation coefficient is as follows: wherein, is the distance correlation coefficient, is the distance covariance between X and Y, is the distance variance of X and Y, respectively.

5. The KNN-based engine remaining useful life prediction method of claim 1, wherein, The weight function includes uniform weight or distance reciprocal weight, wherein the distance reciprocal weight selects Euclidean distance, Manhattan distance or Chebyshev distance as the distance metric.

6. The KNN-based engine remaining useful life prediction method of claim 1, wherein, The method for grid parameter optimization is as follows: Traverse the defined hyperparameter grid, for each group of hyperparameters, use the cross-validation method to divide the training set into multiple subsets, train the KNN model, and record the performance indicators of each model to obtain the best hyperparameter combination and determine the optimal K value.

7. A KNN-based engine remaining life prediction system, characterized by, The method comprises the following modules: The acquisition module is configured to obtain sensor parameter data of an engine as training data to construct a training set; and preprocess the sensor parameter data of the engine in the training set to obtain preprocessed data; The construction module is configured to calculate the correlation coefficient of each preprocessed data and its corresponding remaining life data, and add the preprocessed data with a correlation coefficient greater than a set correlation coefficient threshold to a pre-constructed optimal parameter set; The construction module is configured to calculate the correlation coefficient of each preprocessed data and its corresponding remaining life data, and add the preprocessed data with a correlation coefficient greater than a set correlation coefficient threshold to a pre-constructed optimal parameter set; An optimization module is configured to obtain the hyperparameters corresponding to each preprocessed data in the optimal parameter set, construct a hyperparameter grid, traverse the hyperparameter grid, and input each group of hyperparameters in the hyperparameter grid into a pre-constructed KNN model for grid parameter optimization to obtain an optimal hyperparameter combination and input the optimal hyperparameter combination into the KNN model as an optimal KNN model. A prediction module is configured to obtain monitoring data of an engine to be subjected to remaining life prediction and input the monitoring data into the optimal KNN model to obtain a prediction result of the remaining life of the engine. A verification module is configured to evaluate the prediction result. The correlation coefficient calculation of each preprocessed data and the corresponding remaining life data is performed by using a distance correlation coefficient. The method comprises:

8. A KNN-based engine remaining life prediction device characterized by comprising: at least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the KNN-based engine remaining life prediction method according to any one of claims 1-6. The computer readable storage medium stores computer instructions, and the computer instructions are used to be executed by the processor to implement the KNN-based engine remaining life prediction method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, ​

Images