Magnetic levitation traffic system health monitoring-oriented missing data recovery method

Abstract

The invention relates to the technical field of operation and maintenance of a maglev traffic system, and discloses a missing data recovery method for health monitoring of the maglev traffic system. The method comprises the following steps: acquiring data detected by each sensor of the maglev traffic system to obtain a data table with missing data; extracting data from the data table with the missing data, and constructing a two-dimensional data matrix with the missing data; constructing a hydrophilic-sparse matrix of the two-dimensional data matrix; constructing an optimization objective function based on the stability of the hydrophilic-hydrophobic matrix; iteratively updating the affinity-sparseness matrix by using the optimized objective function to obtain an updated affinity-sparseness matrix; and carrying out missing data recovery by utilizing the updated intimacy-sparseness matrix to obtain a data matrix after missing data recovery. Therefore, the missing data can be accurately recovered, the availability of the data is greatly improved, and data analysts can conveniently and effectively analyze test results.

Description

Technical Field

[0001] The present invention relates to the field of maglev transportation system operation and maintenance technology, and in particular to a missing data recovery method for maglev transportation system health monitoring. Background Art

[0002] The ultra-high-speed low-vacuum maglev transportation system adopts superconducting electric suspension technology. Through the interaction between the superconducting magnet installed on the vehicle and the ground coil module, the suspension force, guiding force and propulsion force required for train operation are generated. During the health monitoring process of the ultra-high-speed low-vacuum maglev transportation system, possible phenomena such as sensor detachment, data acquisition board damage, data transmission cable breakpoints, network interruption and abnormal data removal will cause data loss. The loss is generally stored in the database in the form of "Null Value", "NaN" or approximate 0 value. These data lose the health information of the maglev transportation system, damage the integrity and accuracy of the data, and reduce the quality of the monitoring data. Therefore, a missing data recovery method is urgently needed to provide technical support for improving the level of health monitoring of ultra-high-speed low-vacuum maglev transportation system.

[0003] The existing missing data recovery methods mainly include the zero value method, the statistical value method and the statistical model filling method.

[0004] Among them, the 0-value method is to fill all missing values ​​with 0, but the 0-value method cannot be regarded as a data recovery method in essence. It only converts the missing value into a numerical value to avoid the system from performing additional processing on the missing data during data management; the statistical value method is to calculate the average value of all data in the channel and regard the missing value as the average value. This method has improved accuracy compared to the 0-value method, but replacing all missing data with the same value is still not accurate enough; the statistical model filling method is to use statistical models such as Gaussian distribution and Poisson distribution to estimate the missing data, which greatly improves the accuracy, but it is impossible to establish an effective statistical model when there are many missing data. Summary of the invention

[0005] The present invention provides a missing data recovery method for health monitoring of a maglev transportation system, which can solve the problems in the prior art.

[0006] The present invention provides a missing data recovery method for health monitoring of a maglev transportation system, wherein the method comprises:

[0007] Obtain the data detected by each sensor of the maglev transportation system and obtain a data table with missing data;

[0008] Extract data from a data table with missing data and construct a two-dimensional data matrix with missing data;

[0009] Construct the affinity matrix of the two-dimensional data matrix;

[0010] Construct an optimization objective function based on the stability of affinity matrix;

[0011] The affinity matrix is ​​iteratively updated using the optimization objective function to obtain an updated affinity matrix;

[0012] The updated affinity matrix is ​​used to recover the missing data, and a data matrix after the missing data is recovered is obtained.

[0013] Preferably, the affinity matrix constructed is:

[0014] γ=[y p,q ] M×M ,

[0015] Initially, element y p,q The value of is:

[0016]

[0017] Among them, γ is the affinity matrix, M is the number of data, y p,q is the qth affinity weight of the pth row.

[0018] Preferably, the optimization objective function is:

[0019]

[0020] A=[x i,j ] M×N ,

[0021] Among them, S is the optimization objective function, A is the two-dimensional data matrix with missing data, F is the Frobenius norm, N is the dimension of the data, and x i,j is the jth data in the i-th data.

[0022] Preferably, the affinity matrix is ​​iteratively updated by using a gradient descent method and an optimization objective function to obtain an updated affinity matrix.

[0023] Preferably, the updated affinity matrix is ​​obtained by the following formula:

[0024]

[0025] Among them, t is the number of iterations, γ t is the affinity matrix after t iterations, γ t-1 is the affinity matrix after t-1 iterations, and η is the gradient descent update step size.

[0026] Preferably, the data matrix after missing data recovery is obtained by the following formula:

[0027] A 恢复 =γt A,

[0028] Among them, A 恢复 is the data matrix after the missing data is restored.

[0029] Through the above technical solution, the stability of the affinity matrix can be utilized to eliminate the noise influence caused by correlation in general methods, so that the missing data can be accurately restored, which greatly improves the availability of data and facilitates data analysts to effectively analyze the test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The included drawings are used to provide a further understanding of the embodiments of the present invention, which constitute a part of the specification, are used to illustrate the embodiments of the present invention, and together with the text description, explain the principles of the present invention. Obviously, the drawings in the following description are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 A flow chart of a missing data recovery method for health monitoring of a maglev transportation system according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0032] It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is by no means intended to limit the present invention and its application or use. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0034] Unless otherwise specifically stated, the relative arrangement of the parts and steps described in these embodiments, numerical expressions and numerical values ​​do not limit the scope of the present invention. At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the accompanying drawings are not drawn according to the actual proportional relationship. The technology, method and equipment known to ordinary technicians in the relevant field may not be discussed in detail, but in appropriate cases, the technology, method and equipment should be regarded as a part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as being merely exemplary, rather than as a limitation. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once a certain item is defined in an accompanying drawing, it does not need to be further discussed in subsequent drawings.

[0035] Figure 1 A flow chart of a missing data recovery method for health monitoring of a maglev transportation system according to an embodiment of the present invention is shown.

[0036] Among them, the method described in the present invention is suitable for recovering missing data in health monitoring of ultra-high-speed low-vacuum maglev transportation systems.

[0037] like Figure 1 As shown, an embodiment of the present invention provides a missing data recovery method for health monitoring of a maglev transportation system, wherein the method comprises:

[0038] Obtain the data detected by each sensor of the maglev transportation system and obtain a data table with missing data;

[0039] For example, the structured data output by each sensor of the ultra-high-speed low-vacuum maglev transportation system can be stored in a relational database in the form of a data form (i.e., a data table).

[0040] Extract data from a data table with missing data and construct a two-dimensional data matrix with missing data;

[0041] Construct the affinity matrix of the two-dimensional data matrix;

[0042] Construct an optimization objective function based on the stability of affinity matrix;

[0043] The affinity matrix is ​​iteratively updated using the optimization objective function to obtain an updated affinity matrix;

[0044] The updated affinity matrix is ​​used to recover the missing data, and a data matrix after the missing data is recovered is obtained.

[0045] Through the above technical solution, the stability of the affinity matrix can be utilized to eliminate the noise influence caused by correlation in general methods, so that the missing data can be accurately restored, which greatly improves the availability of data and facilitates data analysts to effectively analyze the test results.

[0046] According to an embodiment of the present invention, the affinity matrix constructed is:

[0047] γ=[y p,q ] M×M ,

[0048] Initially, element y p,q The value of is:

[0049]

[0050] Among them, γ is the affinity matrix, M is the number of data, y p,q is the qth affinity weight in the pth row (that is, any affinity weight in the affinity matrix).

[0051] There is a closeness relationship between each piece of data. The similarity of data with close relationship is high, and the similarity of data with distant relationship is low. Therefore, the matrix composed of the closeness weights between M pieces of data can be recorded as γ.

[0052] Initially, each piece of data in the two-dimensional data matrix is ​​represented as the weighted sum of all other data. Data with close relationships have a large weight and dominate, while data with distant relationships have a small weight and cannot dominate the value of the data. Initially, all closeness weights are assumed to be the same.

[0053] According to one embodiment of the present invention, the optimization objective function is:

[0054]

[0055] A=[x i,j ] M×N ,

[0056] Among them, S is the optimization objective function, A is the two-dimensional data matrix with missing data, F is the Frobenius norm, N is the dimension of the data (collected N-dimensional data), x i,j is the jth data in the i-th data (i.e., any element in the two-dimensional data matrix with missing data).

[0057] Specifically, for the affinity matrix γ and the two-dimensional data matrix A, if γ can completely and correctly describe the affinity relationship of the data in A, then the product of the affinity matrix and the data matrix (γ·A) should be very close to A, and then there can be a relationship: If (γ·A) is regarded as a data matrix, then since (γ·A) is very close to A, the affinity matrices corresponding to the two should also be very close. Therefore, the affinity matrix of (γ·A) can be regarded as γ. This property is called the stability of the affinity matrix, and the relationship can be obtained. Therefore, the above optimization objective function S can be established.

[0058] According to an embodiment of the present invention, the affinity matrix is ​​iteratively updated by using a gradient descent method and an optimization objective function to obtain an updated affinity matrix.

[0059] That is, the affinity matrix can be iteratively updated by the gradient descent method.

[0060] According to an embodiment of the present invention, the updated affinity matrix is ​​obtained by the following formula:

[0061]

[0062] Among them, t is the number of iterations, γ t is the affinity matrix after t iterations, γ t-1 is the affinity matrix after t-1 iterations, and η is the gradient descent update step size.

[0063] According to an embodiment of the present invention, the data matrix after the missing data is restored is obtained by the following formula:

[0064] A 恢复 =γ t A,

[0065] Among them, A 恢复 is the data matrix after the missing data is restored.

[0066] That is, after the iterations converge, the affinity matrix updated after t iterations can be used to recover the missing data.

[0067] Further, in order to verify the effectiveness of the method described in the present invention, simulated data are generated for experiments. In this experiment, five data sets simu1, simu2, simu3, simu4 and simu5 were obtained, each of which contained 500 data, each of which contained data from 2000 channels. For each data set, 28.38%, 63.14%, 78.00%, 88.06% and 97.22% of the non-zero values ​​were artificially replaced with zeros. In this way, five data matrices containing missing data can be obtained, and the sources of missing values ​​in the matrices include both true zeros with true expression values ​​of 0 and false zeros obtained by replacing non-zero values. Finally, the zero value ratios of the five data sets were 42.86%, 70.59%, 82.45%, 90.47% and 97.79% respectively. Using RMSE as the criterion, the comparison before and after recovery can be obtained as shown in Table 1 below:

[0068]

[0069] Table 1 Comparison results before and after data recovery

[0070] It can be seen that after recovery using the missing data recovery method of the present invention, the accuracy of the data is greatly improved.

[0071] In the description of the present invention, it is necessary to understand that the directions or positional relationships indicated by directional words such as "front, back, up, down, left, right", "lateral, vertical, perpendicular, horizontal" and "top, bottom" are usually based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description. Unless otherwise specified, these directional words do not indicate or imply that the devices or elements referred to must have a specific direction or be constructed and operated in a specific direction. Therefore, they cannot be understood as limiting the scope of protection of the present invention. The directional words "inside and outside" refer to the inside and outside relative to the contours of each component itself.

[0072] For ease of description, spatially relative terms such as "above", "above", "on the upper surface of", "above", etc. may be used here to describe the spatial positional relationship between a device or feature and other devices or features as shown in the figure. It should be understood that spatially relative terms are intended to include different orientations of the device in use or operation in addition to the orientation described in the figure. For example, if the device in the accompanying drawings is inverted, the device described as "above other devices or structures" or "above other devices or structures" will be positioned as "below other devices or structures" or "below other devices or structures". Thus, the exemplary term "above" can include both "above" and "below". The device can also be positioned in other different ways (rotated 90 degrees or in other orientations), and the spatially relative descriptions used here are interpreted accordingly.

[0073] In addition, it should be noted that the use of terms such as "first" and "second" to limit components is only for the convenience of distinguishing the corresponding components. If not otherwise stated, the above terms have no special meaning and therefore cannot be understood as limiting the scope of protection of the present invention.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A missing data recovery method for health monitoring of maglev transportation systems, It is characterized in that The method includes: Obtain the data detected by each sensor of the maglev transportation system and obtain a data table with missing data; Extract data from a data table with missing data and construct a two-dimensional data matrix with missing data; Construct the affinity matrix of the two-dimensional data matrix; Construct an optimization objective function based on the stability of affinity matrix; The affinity matrix is ​​iteratively updated using the optimization objective function to obtain an updated affinity matrix; The updated affinity matrix is ​​used to recover the missing data, and a data matrix after the missing data is recovered is obtained.

2. The method according to claim 1, It is characterized in that The affinity matrix constructed is: γ=[y p,q ] M×M , Initially, element y p,q The value of is: Among them, γ is the affinity matrix, M is the number of data, y p,q is the qth affinity weight of the pth row.

3. The method according to claim 2, It is characterized in that The optimization objective function is: A=[x i,j ] M×N , Among them, S is the optimization objective function, A is the two-dimensional data matrix with missing data, F is the Frobenius norm, N is the dimension of the data, and x i,j is the jth data in the i-th data.

4. The method according to claim 3, It is characterized in that Through the gradient descent method, the affinity matrix is ​​iteratively updated by optimizing the objective function to obtain the updated affinity matrix.

5. The method according to claim 4, It is characterized in that The updated affinity matrix is ​​obtained by the following formula: Among them, t is the number of iterations, γ t is the affinity matrix after t iterations, γ t-1 is the affinity matrix after t-1 iterations, and η is the gradient descent update step size.

6. The method according to claim 5, It is characterized in that The data matrix after missing data recovery is obtained by the following formula: A 恢复 =c t ·A, Among them, A 恢复 is the data matrix after the missing data is restored.

Images