CN118913483A - Sensor fault judging method, device and storage medium - Google Patents

Abstract

The present invention discloses a sensor fault determination method, device and storage medium, the method comprising: determining a target device, selecting sensor information related to the target device according to an expert knowledge base; respectively calculating correlation coefficient thresholds H1 and W1 of the sensor information in a healthy state and a deviation state; constructing a correlation coefficient model; calculating an average coefficient for a sensor to be determined according to the constructed correlation coefficient model; and determining whether the sensor is faulty.

Description

A sensor fault determination method, device and storage medium

Technical Field

The present invention relates to sensor technology, and in particular to a sensor fault determination method, device and storage medium.

Background Art

At present, sensor fault judgment methods are mainly divided into two categories: Method 1 is a fault diagnosis method that relies on mathematical models, and Method 2 is a fault diagnosis method that does not rely on mathematical models. In the prior art, Method 1 requires physical modeling in advance based on the controller and the sensors corresponding to the controller. Method 1 can be subdivided into: state evaluation method, parameter evaluation method, and equivalent space method. Method 2 can be subdivided into: data-driven method, support-driven method, and discrete event-based method. Now, as control systems become more and more complex and there are more and more sensors, the problem that follows is that the above-mentioned method 1 will have problems such as increased complexity in mathematical model construction, poor reliability, and increased calculations. In addition, the data-driven method of the above-mentioned method 2 will also become more complicated and difficult to implement due to the increased complexity of the waveform data method.

Summary of the invention

The invention provides a sensor fault determination method, device and storage medium.

In a first aspect, the present invention provides a sensor fault determination method, the method comprising:

Determine the target device, and select the sensor information related to the target device according to the expert knowledge base; calculate the correlation coefficient threshold of the sensor information in the healthy state and the deviation state of the sensor respectively , ;

Construct correlation coefficient model;

According to the constructed correlation coefficient model, the average coefficient is calculated for the sensor to be judged;

Determine whether the sensor is faulty.

Furthermore, the correlation coefficient threshold of the sensor information under the sensor health state is calculated , specifically: sensor set , is the number of sensors, The specific expression is:

;

in, for and Pearson correlation coefficient of sensors in healthy state;

Furthermore, the calculation of the correlation coefficient threshold of the sensor information in the sensor deviation state is specifically:

Based on the historical data of sensor anomalies , calculate the correlation coefficient, and select the minimum value of the correlation matrix as the correlation coefficient threshold , The specific expression is:

;

in, For the Historical data points in sensors and Pearson correlation coefficient for sensor bias conditions.

Furthermore, the calculating of the average coefficient for the sensor to be determined includes:

Before calculation, the data is preprocessed as follows:

The data within 2 hours after the unit was screened are:

;

in, Indicates that the sensor is at time The measured value of is the start time;

The average value in the sliding window is used for frequency reduction, and the data after frequency reduction is:

;

in, is the sliding window size;

For time points where there are still null values after the frequency reduction, the previous value is selected for patching.

Further, the determining whether the sensor is faulty includes:

The output average coefficient is less than the correlation coefficient threshold The sensor name.

In a second aspect, the present invention provides a sensor fault determination device, characterized in that the device comprises:

Calculate threshold module for:

Determine the target device, and select the sensor information related to the target device according to the expert knowledge base; calculate the correlation coefficient threshold of the sensor information in the healthy state and the deviation state of the sensor respectively , ;

Build models and calculate modules for:

Construct correlation coefficient model;

According to the constructed correlation coefficient model, the average coefficient is calculated for the sensor to be judged;

The judgment module is used to:

Determine whether the sensor is faulty.

Furthermore, the device comprises:

Memory, used to store programs;

A processor, configured to execute the program stored in the memory; when the program stored in the memory is executed, the processor is configured to execute the steps of the method as claimed in any one of claims 1 to 4.

In a third aspect, the present invention provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, causes the one or more processors to perform the steps of the method as described in any one of claims 1 to 4.

Compared with the prior art, the present invention has the following advantages:

By digitizing waveform data, driving based on expert knowledge, and combining historical fault samples and healthy samples to screen out highly correlated measurement points, and then building a model, the problems existing in the existing technology are solved.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of a sensor fault determination method provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram showing the composition of a sensor fault determination device provided in an embodiment of the present invention.

DETAILED DESCRIPTION

In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

Explanation of terms involved in the embodiments of the present invention:

Spearman's correlation coefficient is a non-parametric indicator used in statistics to measure the correlation between two variables. It is based on the relationship between the rankings (grades) of the two variables rather than the specific values between the variables.

The Spearman correlation coefficient ranges from -1 to +1, where:

1 indicates a perfect positive correlation, meaning that when one variable increases, the other also increases.

-1 indicates a perfect negative correlation, meaning that when one variable increases, the other decreases.

0 means there is no correlation, that is, changes in one variable do not predict changes in the other variable.

The Spearman correlation coefficient is applicable to data sets that are not completely linear or do not meet the assumptions of the Pearson correlation coefficient. It is often used to analyze rank data or when the data does not meet the normal distribution. Compared with the Pearson correlation coefficient, the Spearman correlation coefficient is less sensitive to outliers and does not require the assumption that the data follows a specific distribution.

The Pearson correlation coefficient, also known as the Pearson product-moment correlation coefficient (PPMCC or PCCs), is a statistic used to measure the degree of linear correlation between two variables X and Y, and its value is between -1 and 1. The larger the absolute value of the Pearson correlation coefficient, the stronger the linear correlation between the two variables.

When the Pearson correlation coefficient is 1 or -1, it means that the two variables are completely linearly correlated.

When the Pearson correlation coefficient is 0, it means that there is no linear relationship between the two variables.

The Pearson correlation coefficient is calculated as the quotient of the covariance and the standard deviation of the two variables.

FIG1 is a schematic diagram of a sensor fault determination method provided by an embodiment of the present invention, the method comprising:

Determine the target device, and select the sensor information related to the target device according to the expert knowledge base; calculate the correlation coefficient threshold of the sensor information in the healthy state and the deviation state of the sensor respectively , ;

Construct correlation coefficient model;

According to the constructed correlation coefficient model, the average coefficient is calculated for the sensor to be judged;

Determine whether the sensor is faulty.

Furthermore, the correlation coefficient threshold of the sensor information under the sensor health state is calculated , specifically: sensor set , is the number of sensors, The specific expression is:

;

in, for and Pearson correlation coefficient of sensors in healthy state;

Furthermore, the calculation of the correlation coefficient threshold of the sensor information in the sensor deviation state is specifically:

Based on the historical data of sensor anomalies , calculate the correlation coefficient, and select the minimum value of the correlation matrix as the correlation coefficient threshold , The specific expression is:

;

in, For the Historical data points in sensors and Pearson correlation coefficient for sensor bias conditions.

Furthermore, the calculating of the average coefficient for the sensor to be determined includes:

Before calculation, the data is preprocessed as follows:

The data within 2 hours after the unit was screened are:

;

in, Indicates that the sensor is at time The measured value of is the start time;

The average value in the sliding window is used for frequency reduction, and the data after frequency reduction is:

;

in, is the sliding window size;

For time points where there are still null values after the frequency reduction, the previous value is selected for patching.

In one embodiment, for the upper guide tile, according to the expert knowledge base, the sensor information related to the device (that is, the specific measuring point, 1# upper guide tile temperature, 2# upper guide tile temperature, ..., 12# upper guide tile temperature) and the correlation coefficient thresholds in the healthy state and the deviation state are selected. and .

Calculated from the historical data of the target measurement point and confirmed by the business experts. If the data in September is the data of abnormal sensor, calculate its correlation coefficient and select the minimum value of the correlation matrix as

Furthermore, the calculating of the average coefficient for the sensor to be determined includes:

Before calculation, the data is preprocessed as follows:

Filter the data within 2 hours after the unit starts;

Use the mean value for frequency reduction;

For time points where there are still null values after the frequency reduction, the previous value is selected for patching.

In one embodiment, data of the upper guide shoe for the past one day is selected and necessary data preprocessing is performed on the data.

In one embodiment, to meet the time alignment requirement, a 60-second frequency reduction process is performed.

Further, the determining whether the sensor is faulty includes:

The output average coefficient is less than the correlation coefficient threshold The sensor name.

In one embodiment, according to the alarm threshold , the output average coefficient is less than The name of the measuring point.

An embodiment of the present invention provides a sensor fault determination device, the device comprising:

Calculate threshold module for:

Determine the target device, and select the sensor information related to the target device according to the expert knowledge base; calculate the correlation coefficient threshold of the sensor information in the healthy state and the deviation state of the sensor respectively , ;

Build models and calculate modules for:

Construct correlation coefficient model;

According to the constructed correlation coefficient model, the average coefficient is calculated for the sensor to be judged;

The judgment module is used to:

Determine whether the sensor is faulty.

Furthermore, the threshold calculation module includes:

The submodule is calculated based on historical data, and is used to calculate the correlation coefficient based on the historical data of sensor abnormalities, and select the minimum value of the correlation matrix as the correlation coefficient threshold W1.

Furthermore, the model building and calculation module includes:

Data preprocessing submodule, used to:

Before calculation, preprocess the data:

Filter the data within 2 hours after the unit starts;

Use the mean value for frequency reduction;

For time points where there are still null values after the frequency reduction, the previous value is selected for patching.

Furthermore, the judging module comprises:

The output submodule is used to: output the name of the sensor whose average coefficient is less than the correlation coefficient threshold W1.

FIG2 is a schematic diagram of a sensor fault determination device provided by an embodiment of the present invention; the device comprises:

Memory, used to store programs;

The processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute the above method.

In addition, an embodiment of the present invention provides a storage medium storing computer-readable instructions, and when the computer-readable instructions are executed by one or more processors, the one or more processors execute the above method.

Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. 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 (8)

1. A method for determining a sensor failure, the method comprising:

Determining target equipment, and selecting sensor information related to the target equipment according to an expert knowledge base; calculating correlation coefficient threshold values of sensor information under the sensor health state and the deviation state respectively 、；

Constructing a correlation coefficient model;

calculating an average coefficient for a sensor to be judged according to the constructed correlation coefficient model;

And judging whether the sensor is faulty or not.

2. The sensor malfunction determination method according to claim 1, wherein a correlation coefficient threshold value of the sensor information in the sensor health state is calculatedThe method specifically comprises the following steps: sensor set，In order to provide for the number of sensors,The specific expression is:

；

wherein, Is thatAndPearson correlation coefficient of the sensor in a healthy state.

3. The method for determining a sensor failure according to claim 1, wherein the calculating the correlation coefficient threshold of the sensor information in the sensor deviation state is specifically:

based on historical data of sensor anomalies Calculating a correlation coefficient, and selecting the minimum value of the correlation matrix as a correlation coefficient threshold value，The specific expression is:

;

wherein, Is the firstSensor in historical data pointsAndPearson correlation coefficient in sensor bias state.

4. The sensor malfunction determination method according to claim 1, wherein the calculating an average coefficient for the sensor to be determined includes:

before calculation, preprocessing the data, specifically:

screening the data within 2 hours from the start of the unit, wherein the screened data are as follows:

；

wherein, Indicating that the sensor is at timeIs used for the measurement of (a),Is the start time;

The average value in the sliding window is adopted for frequency reduction, and the data after frequency reduction are as follows:

；

wherein, Is the sliding window size;

and (5) selecting a previous value patch for the time point with a null value after the frequency reduction.

5. The sensor malfunction determination method according to claim 1, wherein said determining whether the sensor malfunctions comprises:

The output average coefficient is less than the correlation coefficient threshold Is a sensor name of (c).

6. A sensor failure determination apparatus, characterized in that the apparatus comprises:

A calculation threshold module for:

Determining target equipment, and selecting sensor information related to the target equipment according to an expert knowledge base; calculating correlation coefficient threshold values of sensor information under the sensor health state and the deviation state respectively 、；

A model building and calculating module for:

Constructing a correlation coefficient model;

calculating an average coefficient for a sensor to be judged according to the constructed correlation coefficient model;

the judging module is used for:

And judging whether the sensor is faulty or not.

7. A sensor failure determination apparatus, characterized in that the apparatus comprises:

A memory for storing a program;

A processor for executing a program stored in the memory, which processor is adapted to perform the steps of the method according to any one of claims 1 to 4 when the program stored in the memory is executed.

8. A storage medium storing computer readable instructions which, when executed by one or more processors, cause the one or more processors to perform the steps of the method of any of claims 1 to 4.