WO2024109315A1 - Device supervision method and system, and device and computer-readable storage medium - Google Patents

Abstract

A device supervision method and system, and a device and a computer-readable storage medium. The method comprises: acquiring a group of real-time monitoring data of a target device at the current moment; acquiring a preset number of groups of historical monitoring data of the target device; calculating a Euclidean distance value between the real-time monitoring data and each piece of the historical monitoring data; vectorizing all the Euclidean distance values as weight bases; determining estimated monitoring data of the target device at the current moment on the basis of the weight bases and the historical monitoring data; calculating a residual value between the estimated monitoring data and the real-time monitoring data; and on the basis of the residual value, determining whether the target device is abnormal, so as to obtain a corresponding monitoring result. In the present application, it is necessary to calculate a Euclidean distance value between each piece of historical monitoring data and real-time monitoring data; and since the Euclidean distance value may reflect the degree of correlation between the historical monitoring data and the real-time monitoring data, the present application realizes supervision on a target device by means of the degree of correlation between the historical monitoring data and the real-time monitoring data, and has high accuracy.

Description

Device supervision method, system, device and computer-readable storage medium

This application claims priority to the Chinese patent application filed with the China Patent Office on November 23, 2022, with application number 202211475013.5 and invention name “A device supervision method, system, device and computer-readable storage medium”, the entire contents of which are incorporated by reference in this application.

Technical Field

The present application relates to the technical field of equipment monitoring, and more specifically, to an equipment monitoring method, system, equipment and computer-readable storage medium.

Background technique

With the rapid development and progress of industry, industrial equipment is constantly moving towards automation, informationization and intelligence. During the operation of these equipment, abnormalities/failures may occur from time to time, which first damages the equipment and causes property losses, and secondly affects the efficiency and progress of industrial production, and also causes personal harm to operators around the equipment. In short, in order to ensure the stable operation of machinery, reduce safety hazards, and ensure the smooth progress of production, it is necessary to monitor and evaluate the status of mechanical equipment.

Generally, the condition monitoring method for mechanical equipment sets a fixed threshold during the operation of the equipment to deal with the trend rise, fall and fluctuation of the steady-state data. When the monitoring data exceeds the limit set by the fixed threshold, a corresponding alarm will be generated according to the equipment mechanism. This has a clear directionality in the process of equipment condition monitoring, but in order to reduce the occurrence of false reports in the actual process, the fixed threshold is generally set more loosely, which leads to the abnormality of the equipment only being discovered when the equipment state decays to a certain extent. At this time, it is necessary to shut down the relevant equipment, arrange maintenance personnel, prepare materials, allocate repair windows, and repair equipment in a short time. Therefore, there are two defects in the fixed threshold method for condition monitoring: first, the equipment abnormality found by threshold monitoring is already serious, and in order to avoid further deterioration, it is often necessary to shut down the equipment urgently and then arrange manpower and material resources for equipment maintenance; second, this method ignores the correlation between equipment and ignores the in-depth analysis of relevant data.

In addition, in order to deal with non-steady-state data, such as periodically changing temperature data, which will produce periodic fluctuations due to the seasonal changes in ambient temperature, fixed thresholds are Such data has limited monitoring and evaluation effects, so it is necessary to use multidimensional data monitoring methods to ensure the supervision and evaluation of such data. When the fluctuating data produces anomalies/failures in a reasonable trend, this method can be used for better monitoring and evaluation. However, multidimensional data monitoring and evaluation methods are prone to large errors for non-steady-state conditions in the data, that is, the existing multidimensional data monitoring methods are capable of processing data with small fluctuations or periodic fluctuations, but are unable to process non-steady-state/non-periodic data, or the supervision and evaluation accuracy of such data is poor, resulting in poor accuracy of equipment supervision.

In summary, how to accurately monitor the equipment is a problem that needs to be solved urgently by those skilled in the art.

Summary of the invention

The purpose of this application is to provide a device supervision method, which can solve the technical problem of how to accurately supervise the device to a certain extent. This application also provides a device supervision system, a device and a computer-readable storage medium.

In order to achieve the above objectives, this application provides the following technical solutions:

A device supervision method, comprising:

Acquire a set of real-time monitoring data of the target device at the current moment, wherein the real-time monitoring data includes current data values of a preset number of data types;

Acquire a preset number of groups of historical monitoring data of the target device, the historical monitoring data including historical data values of the preset number of data types;

Calculating the Euclidean distance between the real-time monitoring data and each of the historical monitoring data;

Vectorizing all of the Euclidean distance values into weight bases;

Determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data;

Calculating a residual value between the estimated monitoring data and the real-time monitoring data;

Whether the target device is abnormal is determined based on the residual value to obtain a corresponding monitoring result.

Preferably, before vectorizing all the Euclidean distance values into weight bases, the method further includes:

Calculating a similarity density matrix of the historical monitoring data;

The Euclidean distance value between the real-time monitoring data and each of the historical monitoring data is optimized based on the similarity density matrix.

Preferably, the calculating of the similarity density matrix of the historical monitoring data comprises:

Calculate the similarity value between any two of the historical monitoring data of the preset number of groups through a first calculation formula;

The first calculation formula includes:

X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)… _Xj (m)] ^T ；
X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)… _Xk (m)] ^T ；

Wherein, X'(j) represents the historical data value of the j-th data type; X'(k) represents the historical data value of the k-th data type; Sim(j,k) represents the similarity value between the historical data value of the j-th data type and the historical data value of the k-th data type; _Xj (a) represents the a-th historical data value of the j-th data type, _Xk (a) represents the a-th historical data value of the k-th data type, 1≤a≤m, m represents the value of the preset number group; T represents transposition; |||| represents taking the absolute value;

The similarity values between all pairs of historical monitoring data are calculated to obtain the similarity density matrix of the historical monitoring data. The similarity density matrix is shown in the following formula:

Among them, Sim represents the similarity density matrix composed of the similarity values; X′(b) represents the historical data value of the b-th data type, X′'(d) represents the historical data value of the d-th data type, 1≤b≤n, 1≤d≤n; n represents the value of the preset number.

Preferably, the optimizing the Euclidean distance value between the real-time monitoring data and each of the historical monitoring data based on the similarity density matrix includes:

By using a second calculation formula, optimizing the Euclidean distance value between the real-time monitoring data and each of the historical monitoring data based on the similarity density matrix;

The second calculation formula includes:

Among them, X(obs) represents the real-time monitoring data; X(i) represents the i-th group of historical monitoring data, 1≤i≤m, and m represents the value of the preset number group; _Xf (i) represents the historical data value of the f-th data type in the i-th group of historical monitoring data, and _Xf (obs) represents the current data value of the f-th data type in the real-time monitoring data, 1≤f≤n; _disi (X(i),X(obs)) represents the optimized Euclidean distance value between the i-th group of historical monitoring data and the real-time monitoring data.

Preferably, vectorizing all the Euclidean distance values into weight bases includes:

Based on Gaussian kernel function transformation, all the Euclidean distance values are vectorized into the weight base.

Preferably, the formula for the Gaussian kernel function transformation includes:

Wherein, W represents the weight base; _wi represents the weight value of the i-th group of historical monitoring data; h represents the bandwidth of the kernel function; and T represents transposition.

Preferably, the determining the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data includes:

Determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data through a third calculation formula;

The third calculation formula includes:

Wherein, X _est represents the estimated monitoring data.

Preferably, judging whether the target device is abnormal based on the residual value to obtain a corresponding monitoring result includes:

It is determined whether the residual value is within a preset range. If so, the monitoring result indicating that the target device is normal is obtained. If not, the monitoring result indicating that the target device is abnormal is obtained.

A device monitoring system, comprising:

A first acquisition module, used to acquire a set of real-time monitoring data of the target device at the current moment, wherein the real-time monitoring data includes current data values of a preset number of data types;

A second acquisition module is used to acquire a preset number of groups of historical monitoring data of the target device, wherein the historical monitoring data includes historical data values of the preset number of data types;

A first calculation module, used to calculate the similarity value between the historical monitoring data and the real-time monitoring data;

A second calculation module, configured to calculate, for each group of the historical monitoring data, a Euclidean distance value between the historical monitoring data and the real-time monitoring data based on the similarity value;

A first conversion module, used for vectorizing all the Euclidean distance values into weight bases;

A first determination module, configured to determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data;

A third calculation module, used to calculate the residual value between the estimated monitoring data and the real-time monitoring data;

The first judgment module is used to judge whether the target device is abnormal based on the residual value to obtain a corresponding monitoring result.

An electronic device, comprising:

Memory for storing computer programs;

A processor is used to implement the steps of any of the above-mentioned device supervision methods when executing the computer program.

A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any of the above-mentioned device supervision methods.

The present application provides a device supervision method, which obtains a set of real-time monitoring data of the target device at the current moment, the real-time monitoring data includes the current data values of a preset number of data types; obtains a preset number of groups of historical monitoring data of the target device, the historical monitoring data includes the historical data values of a preset number of data types; calculates the Euclidean distance value between the real-time monitoring data and each historical monitoring data; vectorizes all Euclidean distance values into a weight base; determines the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data; calculates the residual value between the estimated monitoring data and the real-time monitoring data; and determines whether the target device is abnormal based on the residual value to obtain the corresponding monitoring result. In the present application, in the process of supervising the target device, the Euclidean distance value between each historical monitoring data and the real-time monitoring data can be calculated. Since the Euclidean distance value can reflect the historical monitoring data, the Euclidean distance value can reflect the historical monitoring data. The correlation between historical monitoring data and real-time monitoring data is calculated, so the present application realizes monitoring the target device with the help of the correlation between historical monitoring data and real-time monitoring data, with high accuracy. The present application provides a device monitoring system, device and computer-readable storage medium that also solves the corresponding technical problems.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

FIG1 is a flow chart of a device monitoring method provided by an embodiment of the present application;

FIG2 is a schematic diagram of the structure of a device monitoring system provided in an embodiment of the present application;

FIG3 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

FIG. 4 is another schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

Detailed ways

The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

Please refer to FIG. 1 , which is a flow chart of a device monitoring method provided in an embodiment of the present application.

An apparatus monitoring method provided in an embodiment of the present application may include the following steps:

Step S101: obtaining a set of real-time monitoring data of the target device at the current moment, where the real-time monitoring data includes current data values of a preset number of data types.

In practical applications, a set of real-time monitoring data of the target device at the current moment can be obtained first, and the real-time monitoring data includes current data values of a preset number of data types. It should be noted that the corresponding information of the target device, the preset number and the data type can be flexibly determined according to actual needs. For example, the target device may be a cooling water pump, and the data types may include temperature, pressure, current, etc., which are not specifically limited in this application.

In specific application scenarios, in the process of obtaining the corresponding data of the target device, the collected data of the target device can be normalized first to ensure the quality of the data; missing values and abnormal numerical points can also be deleted; the abnormal state of the cleaned data is then removed to ensure that the data used for subsequent modeling is completely under normal working conditions, thereby ensuring that the established model can learn the historical information of normal working conditions without being disturbed by abnormal working conditions, thereby ensuring the accuracy of equipment supervision.

Step S102: Acquire a preset number of groups of historical monitoring data of the target device, where the historical monitoring data includes historical data values of a preset number of data types.

In actual applications, after obtaining a set of real-time monitoring data of the target device at the current moment, a preset number of groups of historical monitoring data of the target device can be obtained, such as 5 groups or 10 groups of historical monitoring data, etc., and each group of historical monitoring data includes a preset number of historical data values of the data type, so that the accuracy of the real-time monitoring data can be evaluated based on the historical monitoring data in the future.

In specific application scenarios, real-time monitoring data can be saved in the form of X(i) = [X ₁ (i) X ₂ (i) X ₃ (i) ... X _n (i)] ^T. The historical monitoring data is saved in the form of, etc., and this application does not make any specific limitations here.

Step S103: Calculate the Euclidean distance between the real-time monitoring data and each historical monitoring data.

In practical applications, after obtaining the real-time monitoring data and historical monitoring data of the target device, the Euclidean distance value between the real-time monitoring data and each historical monitoring data can be calculated, so as to characterize the correlation between the historical monitoring data and the real-time monitoring data based on the Euclidean distance value.

In practical applications, after calculating the Euclidean distance value between the real-time monitoring data and each historical monitoring data, the similarity density matrix of the historical monitoring data can also be calculated; based on the similarity density matrix, the Euclidean distance value between the real-time monitoring data and each historical monitoring data is optimized to further supervise the equipment based on the similarity between the historical detection data.

In specific application scenarios, in the process of calculating the similarity density matrix of historical monitoring data, the similarity density matrix of historical monitoring data can be calculated based on the cosine similarity method. Assuming that there are m data points of time series length for the jth measurement point:
X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)... _Xj (m)] ^T ；

For the kth measurement point, there are also data points with a time series length of m:
X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)... _Xk (m)] ^T ；

Then the cosine value between the two measuring points is:

The numerator can be expressed as:

The denominator can be expressed as:

The final similarity density matrix is:

That is, the similarity value between any two historical monitoring data of the preset number of groups of historical monitoring data can be calculated by the first calculation formula;

The first calculation formula includes:

X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)… _Xj (m)] ^T ；
X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)… _Xk (m)] ^T ；

Among them, X'(j) represents the historical data value of the jth data type; X'(k) represents the historical data value of the kth data type; Sim(j,k) represents the similarity value between the historical data value of the jth data type and the historical data value of the kth data type; _Xj (a) represents the historical data value of the ath data type of the jth data type. Historical data value, X _k (a) represents the a-th historical data value of the k-th data type, 1≤a≤m, m represents the value of the preset number group, which can be determined by the number of selected time series; T represents transposition; |||| represents taking the absolute value;

Calculate the similarity values between all pairwise historical monitoring data to obtain the similarity density matrix of historical monitoring data. The similarity density matrix is shown in the following formula:

Among them, Sim represents the similarity density matrix composed of similarity values; X′(b) represents the historical data value of the b-th data type, X′'(d) represents the historical data value of the d-th data type, 1≤b≤n, 1≤d≤n; n represents a preset number of values.

In a specific application scenario, in the process of optimizing the Euclidean distance value between the real-time monitoring data and each historical monitoring data based on the similarity density matrix, the Euclidean distance value between the real-time monitoring data and each historical monitoring data can be optimized based on the similarity density matrix by the second calculation formula;

The second calculation formula includes:

Among them, X(obs) represents real-time monitoring data; X(i) represents the i-th group of historical monitoring data, 1≤i≤m, and m represents the value of a preset number of groups; _Xf (i) represents the historical data value of the f-th data type in the i-th group of historical monitoring data, and _Xf (obs) represents the current data value of the f-th data type in the real-time monitoring data, 1≤f≤n; _disi (X(i),X(obs)) represents the optimized Euclidean distance value between the i-th group of historical monitoring data and the real-time monitoring data.

Step S104: vectorize all Euclidean distance values into weight bases.

Step S105: Determine the estimated monitoring data of the target device at the current moment based on the weight base and historical monitoring data.

In practical applications, after calculating the Euclidean distance value between the real-time monitoring data and each historical monitoring data, all the Euclidean distance values can be vectorized into a weighted base; finally, the estimated monitoring data of the target device at the current moment is determined based on the weighted base and the historical monitoring data, so as to predict the estimated monitoring data of the target device at the current moment based on the historical monitoring data.

In a specific application scenario, in the process of vectorizing all Euclidean distance values into weight bases, all Euclidean distance values can be vectorized into weight bases based on Gaussian kernel function transformation;

Among them, the formula of Gaussian kernel function transformation may include:

Among them, W represents the weight base; _wi represents the weight value of the i-th group of historical monitoring data; h represents the bandwidth of the kernel function; and T represents transpose.

In a specific application scenario, in the process of determining the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data, the estimated monitoring data of the target device at the current moment can be determined based on the weight base and the historical monitoring data by a third calculation formula;

The third calculation formula includes:

Among them, X _est represents the estimated monitoring data.

Step S106: Calculate the residual value between the estimated monitoring data and the real-time monitoring data.

Step S107: determine whether the target device is abnormal based on the residual value and obtain corresponding monitoring results.

In actual applications, after determining the estimated monitoring data of the target device at the current moment based on the weight base and historical monitoring data, the residual value between the estimated monitoring data and the real-time monitoring data can be calculated, and based on the residual value, it can be judged whether the target device is abnormal to obtain the corresponding monitoring results.

In a specific application scenario, in the process of judging whether the target device is abnormal based on the residual value and obtaining the corresponding monitoring result, it can be judged whether the residual value is within the preset range. If so, a monitoring result characterizing that the target device is normal is obtained; if not, a monitoring result characterizing that the target device is abnormal is obtained.

The present application provides a device supervision method, which obtains a set of real-time monitoring data of a target device at the current moment, wherein the real-time monitoring data includes current data values of a preset number of data types; obtains a preset number of historical monitoring data of the target device, wherein the historical monitoring data includes historical data values of a preset number of data types; calculates the Euclidean distance value between the real-time monitoring data and each historical monitoring data; vectorizes all the Euclidean distance values into a weighted basis; determines the estimated monitoring data of the target device at the current moment based on the weighted basis and the historical monitoring data; calculates the residual value between the estimated monitoring data and the real-time monitoring data; determines whether the target device is abnormal based on the residual value, and obtains the corresponding Monitoring results. In the present application, in the process of monitoring the target device, the Euclidean distance value between each historical monitoring data and the real-time monitoring data can be calculated. Since the Euclidean distance value can reflect the degree of correlation between the historical monitoring data and the real-time monitoring data, the present application realizes the supervision of the target device with the help of the degree of correlation between the historical monitoring data and the real-time monitoring data, with high accuracy.

Please refer to FIG. 2 , which is a schematic diagram of the structure of a device monitoring system provided in an embodiment of the present application.

An equipment monitoring system provided in an embodiment of the present application may include:

A first acquisition module 101 is used to acquire a set of real-time monitoring data of the target device at the current moment, where the real-time monitoring data includes current data values of a preset number of data types;

A second acquisition module 102 is used to acquire a preset number of groups of historical monitoring data of the target device, where the historical monitoring data includes historical data values of a preset number of data types;

A first calculation module 103, used to calculate the Euclidean distance value between the real-time monitoring data and each historical monitoring data;

A first conversion module 104, used for vectorizing all Euclidean distance values into weight bases;

A first determination module 105, configured to determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data;

A second calculation module 106, used to calculate the residual value between the estimated monitoring data and the real-time monitoring data;

The first judgment module 107 is used to judge whether the target device is abnormal based on the residual value and obtain corresponding monitoring results.

The device monitoring system provided in the embodiment of the present application may further include:

A third calculation module is used to calculate the similarity density matrix of the historical monitoring data before the first conversion module vectorizes all the Euclidean distance values into weight bases;

The first optimization module is used to optimize the Euclidean distance value between the real-time monitoring data and each historical monitoring data based on the similarity density matrix.

In an equipment monitoring system provided by an embodiment of the present application, the third computing module can be used for:

Calculate the similarity value between any two historical monitoring data of the preset number of groups of historical monitoring data by using the first calculation formula;

The first calculation formula includes:

X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)… _Xj (m)] ^T ；
X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)… _Xk (m)] ^T ；

Wherein, X'(j) represents the historical data value of the jth data type; X'(k) represents the historical data value of the kth data type; Sim(j,k) represents the similarity value between the historical data value of the jth data type and the historical data value of the kth data type; _Xj (a) represents the ath historical data value of the jth data type, _Xk (a) represents the ath historical data value of the kth data type, 1≤a≤m, m represents the value of the preset number group; T represents transposition; |||| represents taking the absolute value;

Calculate the similarity values between all pairwise historical monitoring data to obtain the similarity density matrix of historical monitoring data. The similarity density matrix is shown in the following formula:

Among them, Sim represents the similarity density matrix composed of similarity values; X′(b) represents the historical data value of the b-th data type, X′'(d) represents the historical data value of the d-th data type, 1≤b≤n, 1≤d≤n; n represents a preset number of values.

In an equipment monitoring system provided by an embodiment of the present application, the first optimization module can be used for:

By using a second calculation formula, the Euclidean distance value between the real-time monitoring data and each historical monitoring data is optimized based on a similarity density matrix;

The second calculation formula includes:

Among them, X(obs) represents real-time monitoring data; X(i) represents the i-th group of historical monitoring data, 1≤i≤m, and m represents the value of the preset number group; _Xf (i) represents the historical data value of the f-th data type in the i-th group of historical monitoring data, and _Xf (obs) represents the current data value of the f-th data type in the real-time monitoring data, 1≤f≤n; _disi (X(i),X(obs)) represents the optimized Euclidean distance value between the i-th group of historical monitoring data and the real-time monitoring data.

In an equipment monitoring system provided by an embodiment of the present application, the first conversion module can be used for:

Based on the Gaussian kernel function transformation, all Euclidean distance values are vectorized into weight bases.

In an equipment monitoring system provided by an embodiment of the present application, the formula for Gaussian kernel function transformation includes:

Among them, W represents the weight base; _wi represents the weight value of the i-th group of historical monitoring data; h represents the bandwidth of the kernel function; and T represents transpose.

In an equipment monitoring system provided by an embodiment of the present application, the first determination module may be used to:

Determine the estimated monitoring data of the target device at the current moment based on the weight base and historical monitoring data through a third calculation formula;

The third calculation formula includes:

Among them, X _est represents the estimated monitoring data.

In an equipment monitoring system provided by an embodiment of the present application, the first judgment module can be used for:

It is determined whether the residual value is within a preset range. If so, a monitoring result indicating that the target device is normal is obtained. If not, a monitoring result indicating that the target device is abnormal is obtained.

The present application also provides an electronic device and a computer-readable storage medium, both of which have the corresponding effects of the device supervision method provided in the embodiment of the present application. Please refer to Figure 3, which is a schematic diagram of the structure of an electronic device provided in the embodiment of the present application.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the following steps are implemented:

Obtain a set of real-time monitoring data of the target device at the current moment, the real-time monitoring data including current data values of a preset number of data types;

Acquire a preset number of groups of historical monitoring data of the target device, the historical monitoring data including historical data values of a preset number of data types;

Calculate the Euclidean distance between the real-time monitoring data and each historical monitoring data;

Vectorize all Euclidean distance values into weighted bases;

Determine the estimated monitoring data of the target device at the current moment based on the weight base and historical monitoring data;

Calculate the residual value between the estimated monitoring data and the real-time monitoring data;

Based on the residual value, it is determined whether the target device is abnormal and the corresponding monitoring result is obtained.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202, wherein a computer program is stored in the memory 201, and the processor 202 implements the following steps when executing the computer program: before all Euclidean distance values are vectorized into weight bases, a similarity density matrix of historical monitoring data is calculated; and based on the similarity density matrix, the Euclidean distance value between the real-time monitoring data and each historical monitoring data is optimized.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the following steps are implemented: calculating the similarity value between any two historical monitoring data of a preset number of groups of historical monitoring data by a first calculation formula;

The first calculation formula includes:

X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)… _Xj (m)] ^T ；
X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)… _Xk (m)] ^T ；

Wherein, X'(j) represents the historical data value of the j-th data type; X'(k) represents the historical data value of the k-th data type; Sim(j,k) represents the similarity value between the historical data value of the j-th data type and the historical data value of the k-th data type; _Xj (a) represents the a-th historical data value of the j-th data type, _Xk (a) represents the a-th historical data value of the k-th data type, 1≤a≤m, m represents the value of the preset number group; T represents transposition; |||| represents taking the absolute value;

Calculate the similarity values between all pairwise historical monitoring data to obtain the similarity density matrix of the historical monitoring data. The similarity density matrix is shown in the following formula:

Among them, Sim represents the similarity density matrix composed of similarity values; X′(b) represents the historical data value of the b-th data type, X′'(d) represents the historical data value of the d-th data type, 1≤b≤n, 1≤d≤n; n represents a preset number of values.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the following steps are implemented: optimizing the Euclidean distance value between the real-time monitoring data and each historical monitoring data based on the similarity density matrix by using a second calculation formula;

The second calculation formula includes:

Among them, X(obs) represents real-time monitoring data; X(i) represents the i-th group of historical monitoring data, 1≤i≤m, and m represents the value of the preset number group; _Xf (i) represents the historical data value of the f-th data type in the i-th group of historical monitoring data, and _Xf (obs) represents the current data value of the f-th data type in the real-time monitoring data, 1≤f≤n; _disi (X(i),X(obs)) represents the optimized Euclidean distance value between the i-th group of historical monitoring data and the real-time monitoring data.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the following steps are implemented: based on Gaussian kernel function transformation, all Euclidean distance values are vectorized into weight bases.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the following steps are implemented: the formula of Gaussian kernel function transformation includes:

Among them, W represents the weight base; _wi represents the weight value of the i-th group of historical monitoring data; h represents the bandwidth of the kernel function; and T represents transpose.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the following steps are implemented: determining the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data by using a third calculation formula;

The third calculation formula includes:

Among them, X _est represents the estimated monitoring data.

An electronic device provided in an embodiment of the present application includes a memory 201 and a processor 202. A computer program is stored in the memory 201. When the processor 202 executes the computer program, the following steps are implemented: determining whether the residual value is within a preset range, and if so, obtaining a monitoring result characterizing that the target device is normal; if not, obtaining a monitoring result characterizing that the target device is abnormal.

Please refer to Figure 4. Another electronic device provided in the embodiment of the present application may also include: an input port 203 connected to the processor 202, used to transmit commands input from the outside to the processor 202; a display unit 204 connected to the processor 202, used to display the processing results of the processor 202 to the outside; a communication module 205 connected to the processor 202, used to realize the communication between the electronic device and the outside. The display unit 204 can be a display panel, a laser scanning display, etc.; the communication mode adopted by the communication module 205 includes but is not limited to mobile high-definition link technology (HML), universal serial bus (USB), high-definition multimedia interface (HDMI), wireless connection: wireless fidelity technology (WiFi), Bluetooth communication technology, low-power Bluetooth communication technology, and communication technology based on IEEE802.11s.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

Obtain a set of real-time monitoring data of the target device at the current moment, the real-time monitoring data including current data values of a preset number of data types;

Acquire a preset number of groups of historical monitoring data of the target device, the historical monitoring data including historical data values of a preset number of data types;

Calculate the Euclidean distance between the real-time monitoring data and each historical monitoring data;

Vectorize all Euclidean distance values into weighted bases;

Determine the estimated monitoring data of the target device at the current moment based on the weight base and historical monitoring data;

Calculate the residual value between the estimated monitoring data and the real-time monitoring data;

Based on the residual value, it is determined whether the target device is abnormal and the corresponding monitoring result is obtained.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: before all Euclidean distance values are vectorized into weight bases, a similarity density matrix of historical monitoring data is calculated; based on the similarity density matrix, the Euclidean distance value between the real-time monitoring data and each historical monitoring data is optimized.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: calculating the similarity value between any two historical monitoring data of a preset number of groups of historical monitoring data by a first calculation formula;

The first calculation formula includes:

X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)… _Xj (m)] ^T ；
X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)… _Xk (m)] ^T ；

Wherein, X'(j) represents the historical data value of the jth data type; X'(k) represents the historical data value of the kth data type; Sim(j,k) represents the similarity value between the historical data value of the jth data type and the historical data value of the kth data type; _Xj (a) represents the ath historical data value of the jth data type, _Xk (a) represents the ath historical data value of the kth data type, 1≤a≤m, m represents the value of the preset number group; T represents transposition; |||| represents taking the absolute value;

Calculate the similarity values between all pairwise historical monitoring data to obtain the similarity density matrix of historical monitoring data. The similarity density matrix is shown in the following formula:

Among them, Sim represents the similarity density matrix composed of similarity values; X′(b) represents the historical data value of the b-th data type, X′'(d) represents the historical data value of the d-th data type, 1≤b≤n, 1≤d≤n; n represents a preset number of values.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: optimizing the Euclidean distance value between the real-time monitoring data and each historical monitoring data based on the similarity density matrix by using a second calculation formula;

The second calculation formula includes:

Among them, X(obs) represents real-time monitoring data; X(i) represents the i-th group of historical monitoring data, 1≤i≤m, and m represents the value of a preset number of groups; _Xf (i) represents the historical data value of the f-th data type in the i-th group of historical monitoring data, and _Xf (obs) represents the current data value of the f-th data type in the real-time monitoring data, 1≤f≤n; _disi (X(i),X(obs)) represents the optimized Euclidean distance value between the i-th group of historical monitoring data and the real-time monitoring data.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: based on a Gaussian kernel function transformation, all Euclidean distance values are vectorized into a weight base.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: the formula of Gaussian kernel function transformation includes:

Among them, W represents the weight base; _wi represents the weight value of the i-th group of historical monitoring data; h represents the bandwidth of the kernel function; and T represents transpose.

An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: determining the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data by using a third calculation formula;

The third calculation formula includes:

Among them, X _est represents the estimated monitoring data.

A computer-readable storage medium is provided in an embodiment of the present application, and a computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, the following steps are implemented: determining whether the residual value is within a preset range, and if so, obtaining a monitoring result characterizing that the target device is normal; if not, obtaining a monitoring result characterizing that the target device is abnormal.

The computer-readable storage medium involved in this application includes random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, Hard disk, removable disk, CD-ROM, or any other form of storage medium known in the technical field.

For the description of the relevant parts of the device monitoring system, device and computer-readable storage medium provided in the embodiments of the present application, please refer to the detailed description of the corresponding parts in the device monitoring method provided in the embodiments of the present application, which will not be repeated here. In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the corresponding technical solutions in the prior art are not described in detail to avoid excessive elaboration.

It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims (11)

1. A device supervision method, characterized by comprising:

   Acquire a set of real-time monitoring data of the target device at the current moment, wherein the real-time monitoring data includes current data values of a preset number of data types;

   Acquire a preset number of groups of historical monitoring data of the target device, the historical monitoring data including historical data values of the preset number of data types;

   Calculating the Euclidean distance between the real-time monitoring data and each of the historical monitoring data;

   Vectorizing all of the Euclidean distance values into weight bases;

   Determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data;

   Calculating a residual value between the estimated monitoring data and the real-time monitoring data;

   Whether the target device is abnormal is determined based on the residual value to obtain a corresponding monitoring result.
2. The method according to claim 1, characterized in that before vectorizing all the Euclidean distance values into weight bases, it also includes:

   Calculating a similarity density matrix of the historical monitoring data;

   The Euclidean distance value between the real-time monitoring data and each of the historical monitoring data is optimized based on the similarity density matrix.
3. The method according to claim 2, characterized in that the calculating the similarity density matrix of the historical monitoring data comprises:

   Calculate the similarity value between any two of the historical monitoring data of the preset number of groups through a first calculation formula;

   The first calculation formula includes:
   
   X'(j)＝[ _Xj (1) _Xj (2) _Xj (3)… _Xj (m)] ^T ；
   X'(k)＝[ _Xk (1) _Xk (2) _Xk (3)… _Xk (m)] ^T ；

   Wherein, X'(j) represents the historical data value of the j-th data type; X'(k) represents the historical data value of the k-th data type; Sim(j,k) represents the similarity value between the historical data value of the j-th data type and the historical data value of the k-th data type; _Xj (a) represents the historical data value of the a-th data type of the j-th data type. historical data values, X _k (a) represents the a-th historical data value of the k-th data type, 1≤a≤m, m represents the value of the preset number group; T represents transposition; || | _| represents taking the absolute value;

   The similarity values between all pairs of historical monitoring data are calculated to obtain the similarity density matrix of the historical monitoring data. The similarity density matrix is shown in the following formula:
   

   Among them, Sim represents the similarity density matrix composed of the similarity values; X′(b) represents the historical data value of the b-th data type, X′'(d) represents the historical data value of the d-th data type, 1≤b≤n, 1≤d≤n; n represents the value of the preset number.
4. The method according to claim 3, characterized in that the step of optimizing the Euclidean distance value between the real-time monitoring data and each of the historical monitoring data based on the similarity density matrix comprises:

   By using a second calculation formula, optimizing the Euclidean distance value between the real-time monitoring data and each of the historical monitoring data based on the similarity density matrix;

   The second calculation formula includes:
   

   Among them, X(obs) represents the real-time monitoring data; X(i) represents the i-th group of historical monitoring data, 1≤i≤m, and m represents the value of the preset number group; _Xf (i) represents the historical data value of the f-th data type in the i-th group of historical monitoring data, and _Xf (obs) represents the current data value of the f-th data type in the real-time monitoring data, 1≤f≤n; _disi (X(i),X(obs)) represents the optimized Euclidean distance value between the i-th group of historical monitoring data and the real-time monitoring data.
5. The method according to claim 4, characterized in that the step of vectorizing all the Euclidean distance values into a weighted basis comprises:

   Based on Gaussian kernel function transformation, all the Euclidean distance values are vectorized into the weight base.
6. The method according to claim 5, characterized in that the formula of the Gaussian kernel function transformation includes:
   

   Wherein, W represents the weight base; _wi represents the weight value of the i-th group of historical monitoring data; h represents the bandwidth of the kernel function; and T represents transposition.
7. The method according to claim 5, characterized in that the step of determining the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data comprises:

   Determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data through a third calculation formula;

   The third calculation formula includes:
   

   Wherein, X _est represents the estimated monitoring data.
8. The method according to any one of claims 1 to 7, characterized in that judging whether the target device is abnormal based on the residual value and obtaining the corresponding monitoring result comprises:

   It is determined whether the residual value is within a preset range. If so, the monitoring result indicating that the target device is normal is obtained. If not, the monitoring result indicating that the target device is abnormal is obtained.
9. A device monitoring system, characterized in that it comprises:

   A first acquisition module, used to acquire a set of real-time monitoring data of the target device at the current moment, wherein the real-time monitoring data includes current data values of a preset number of data types;

   A second acquisition module is used to acquire a preset number of groups of historical monitoring data of the target device, wherein the historical monitoring data includes historical data values of the preset number of data types;

   A first calculation module, used for calculating the Euclidean distance value between the real-time monitoring data and each of the historical monitoring data;

   A first conversion module, used for vectorizing all the Euclidean distance values into weight bases;

   A first determination module, configured to determine the estimated monitoring data of the target device at the current moment based on the weight base and the historical monitoring data;

   A second calculation module, used to calculate the residual value between the estimated monitoring data and the real-time monitoring data;

   The first judgment module is used to judge whether the target device is abnormal based on the residual value to obtain a corresponding monitoring result.
10. An electronic device, comprising:

    Memory for storing computer programs;

    A processor, configured to implement the steps of the device supervision method according to any one of claims 1 to 8 when executing the computer program.
11. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the device supervision method according to any one of claims 1 to 8 are implemented.