HPLC (High Performance Liquid Chromatography) field fault diagnosis test method and system

By calculating the similarity and easy segmentation of data points in the HPLC system, screening effective isolated trees for fault diagnosis, solving the problems of lag and inaccurate in fault diagnosis of traditional HPLC systems, and achieving more accurate abnormality detection and early warning.

CN118921082AActive Publication Date: 2024-11-08LIAONING HEZHITUO ELECTRIC POWER TECH CO LTD

Patent Information

Application Number
CN202411411832.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-11
Publication Date
2024-11-08
Estimated Expiration
2044-10-11

AI Technical Summary

Technical Problem

Traditional HPLC systems have problems such as lag in fault identification and low maintenance efficiency in fault diagnosis, and the isolated forest algorithm is inaccurate in the process of segmenting abnormal data points, resulting in inaccurate abnormal detection results.

Method used

By calculating the similarity and easy segmentation of data points in different isolated trees, we can filter effective isolated trees for abnormal detection, improve the accuracy of abnormal detection, and use the filtered isolated trees for fault diagnosis of power carrier data.

Benefits of technology

It improves the accuracy of abnormal detection results and the accuracy of diagnostic warnings, reduces noise interference, and can provide timely and accurately warning of faults to staff.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118921082A_ABST
    Figure CN118921082A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of wired transmission fault diagnosis, in particular to a field fault diagnosis test method and system for HPLC (High Performance Liquid Chromatography). The method comprises the following steps: acquiring power line carrier data, and constructing an isolated tree according to the power line carrier data; for any data point of any isolated tree, calculating the easy segmentation degree of the data point, and calculating the similarity degree of the node of the data point in the isolated tree relative to the nodes of the data point in other isolated trees; for any isolated tree, the effectiveness of the isolated tree is calculated according to the similarity degree of the nodes of all the data points in the isolated tree relative to the nodes of all the data points in other isolated trees; according to the method, the isolation trees are screened according to the validity of the isolation trees, the screened isolation trees are utilized to complete anomaly detection of the power line carrier data, diagnosis and early warning are made to workers, and the anomaly detection result of the power line carrier data and the diagnosis and early warning made to the workers are more accurate.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of wired transmission fault diagnosis, and in particular to an on-site fault diagnosis method and system for HPLC. Background Art

[0002] High Speed ​​Power Line Communication (HPLC) technology is a communication method that uses power lines to transmit data and is widely used in the monitoring and diagnosis of power systems. The HPLC system transmits information in power lines through high-frequency carrier signals to achieve remote monitoring and data collection of power equipment. Due to the complex power line environment and numerous interference factors, the HPLC system may encounter various faults during operation. Traditional diagnostic methods usually rely on regular maintenance and manual inspections, and there are problems such as delayed fault identification and low maintenance efficiency. Therefore, a field fault diagnosis test method and system for HPLC is needed to monitor and diagnose power carrier data in real time. Since power carrier data contains data in multiple dimensions, the isolation forest algorithm has a good effect on anomaly detection of multi-dimensional data. Therefore, the isolation forest algorithm can be used to detect anomalies in power carrier data.

[0003] For example, a Chinese patent application document with publication number CN111784392A discloses an abnormal user group detection method, apparatus, and device based on an isolation forest. The application document points out that for any data point, the average path length of the data point is calculated by calculating the path length of the data point in all isolated trees, thereby obtaining the degree of abnormality of the data point.

[0004] For abnormal data points, since the data samples are segmented by randomly selecting the segmentation threshold in the process of building an isolated tree, the abnormal data points cannot be effectively segmented in some isolated trees, which makes the path length of the abnormal data points in the isolated tree longer, and further makes the average path of the data point longer, thereby reducing the abnormality of the data point, and further making the anomaly detection result inaccurate. Summary of the invention

[0005] The purpose of the present invention is to provide a field fault diagnosis test method and system for HPLC, so as to solve the technical problem of inaccurate detection results of abnormal power carrier data and inaccurate early warning results for staff.

[0006] In a first aspect, the present invention provides an on-site fault diagnosis test method for HPLC, which adopts the following technical scheme: Acquire power carrier data, and construct an isolation tree according to the power carrier data; For any data point in any isolated tree, calculate the similarity between the nodes of the data point in the isolated tree and the nodes in other isolated trees. For the The data point in The deepest level in an isolated tree; For the The mean of the deepest level of all isolated trees for a data point; For the The data point in The degree of splitting in an isolated tree; For the The average degree of segmentation of a data point in all isolated trees; is the total number of isolated trees; is the isolated tree ordinal number; The natural constant is an exponential function with a base; the method for obtaining the degree of easy segmentation is: calculating the difference between the number of data points of the nodes to which the data point belongs in any layer in the isolated tree and the number of data points of the nodes to which the data point belongs in all layers, and calculating the degree of easy segmentation of the data point according to the difference, and the degree of easy segmentation is positively correlated with the absolute value of the difference; For any isolated tree, the effectiveness of the isolated tree is calculated based on the similarity of all data points in the isolated tree relative to the nodes in other isolated trees, and the effectiveness is positively correlated with the similarity. The isolated trees are screened according to their effectiveness, and the screened isolated trees are used to complete the abnormal detection of power carrier data, and make diagnostic warnings to the staff.

[0007] For any data point, the present invention further calculates the validity of any isolated tree by calculating the similarity between the nodes of the data point in the isolated tree and the nodes in other isolated trees, and screens the isolated trees according to their validity, so as to make the anomaly detection results more accurate, and further make the diagnosis and early warning results more accurate.

[0008] Preferably, the power line carrier data includes: Signal strength, current, voltage.

[0009] Preferably, the degree of easy splitting satisfies the expression: ; In the formula, Indicates The data point in The degree of ease of splitting in an isolated tree; Indicates The data point in The deepest level in an isolated tree; Indicates the ordinal number of the level in the isolated tree; Indicates The data point in The first The number of data points of the node to which the layer belongs; Indicates The data point in The mean number of data points of all nodes in an isolated tree; Represents the standard normalization function.

[0010] The total number of data points in the nodes to which the data point belongs at any layer of the isolated tree is obtained, and the degree of segmentation of the data point in the isolated tree is calculated according to the change in the total number of data points in the nodes to which the data point belongs during the segmentation process of the isolated tree, so as to calculate the effectiveness of the isolated tree.

[0011] Preferably, the belonging node includes: For any data point of any isolated tree, the node where the data point is located in any layer of the isolated tree is recorded as the node to which the data point belongs in the layer.

[0012] Preferably, the validity satisfies the expression: ; In the formula, Indicates The validity of an isolated tree; Indicates the total number of data points; Indicates the ordinal number of the data point; Indicates The data point in The degree of similarity between nodes in an isolated tree and nodes in other isolated trees; Represents the standard normalization function.

[0013] The effectiveness of the isolated tree is calculated by the degree of segmentation of all data points in the isolated tree, and the average level of segmentation of all data points in the isolated tree is analyzed, so that the calculation result of the effectiveness of the isolated tree is more accurate.

[0014] Preferably, the screening of isolated trees according to the effectiveness of the isolated trees comprises: If the validity of the isolated tree is less than the preset validity threshold , then the isolated tree is removed and does not participate in subsequent anomaly detection.

[0015] By removing isolated trees with less effectiveness, the accuracy of anomaly detection of power carrier data by the isolation forest algorithm is improved, and the impact of randomly constructing isolated trees by the traditional isolation forest algorithm is reduced.

[0016] Preferably, the method of using the screened isolated tree to complete the abnormal detection of the power line carrier data and make a diagnosis warning to the staff includes: For any data point, the average path of the data point is calculated based on the deepest layer in the remaining isolated tree after removal, and the anomaly score of the data point is further calculated. The average anomaly score of all data points is obtained. If the average anomaly score is greater than the anomaly threshold , it is judged that there is a fault and an early warning is given to the staff.

[0017] By using the mean of the abnormal scores of data points for diagnosis and avoiding the interference of noise data, early warnings can be given to staff in a timely and accurate manner.

[0018] In a second aspect, the present invention provides an on-site fault diagnosis and testing system for HPLC, which adopts the following technical solution: A field fault diagnosis and testing system for HPLC includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the field fault diagnosis and testing method for HPLC is implemented.

[0019] By adopting the above technical solution, the above-mentioned HPLC on-site fault diagnosis test method is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is made according to the memory and the processor for easy use.

[0020] The present invention has the following technical effects: The present invention calculates the validity of any isolated tree, screens the isolated trees according to their validity, obtains the abnormality score of each data point using the screened isolated trees, completes the fault diagnosis of power carrier data, and warns the staff, making the abnormality detection result more accurate.

[0021] Furthermore, the present invention calculates the similarity between the nodes of any data point in any isolated tree and the nodes in other isolated trees, and calculates the validity of the isolated tree based on the similarity of all data points in any isolated tree, thereby making the calculation result of the validity of the isolated tree more accurate, and further making the anomaly detection result more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0023] Figure 1 The present invention is a flowchart of a method for on-site fault diagnosis and testing of HPLC according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0025] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0026] The present invention discloses a method for on-site fault diagnosis and testing of HPLC, referring to Figure 1 , comprising steps S1 to S4: S1: Collect power carrier data and build an isolation tree based on the power carrier data.

[0027] It should be noted that power carrier data refers to signals transmitted through power lines, which are superimposed on the alternating current of the power lines. Power carrier technology uses power lines as a communication medium to transmit load data, thereby achieving real-time monitoring. By performing on-site fault diagnosis on power carrier data, anomalies in signal transmission can be quickly discovered so that problems can be diagnosed and repaired in a timely manner, thereby reducing equipment failures and extending equipment service life.

[0028] In one embodiment, a HPLC on-site fault diagnosis tester is used to collect power carrier data, wherein the power carrier data includes: signal strength, current, voltage, and the preset collection period is 1 second.

[0029] The power carrier data is used to construct an isolated tree using the isolation forest algorithm, where the location of the root node of the isolated tree is recorded as the first node of the isolated tree. The number of layers increases by one from top to bottom.

[0030] It should be noted that the isolation forest algorithm is an existing technology and will not be described in detail here.

[0031] At this point, the acquisition of power carrier data is completed and the construction of the isolation tree is completed.

[0032] S2: For any data point in any isolated tree, calculate the similarity of the data point in the isolated tree relative to the nodes in other isolated trees.

[0033] It should be noted that the isolation forest algorithm isolates data points by constructing multiple isolated trees, and each isolated tree is independently and randomly generated, which means that different isolated trees may have different segmentation results for the same data point. Therefore, the present invention first calculates the degree of segmentation of any data point in any isolated tree by analyzing the change in the number of data points of the nodes to which it belongs during the continuous segmentation process, so as to facilitate the calculation of the similarity.

[0034] It should be further explained that, in the process of continuously segmenting any data point, the data point will appear multiple times in the isolated tree until the data point is independently segmented or the isolated tree reaches a preset segmentation depth. In the process of continuously segmenting the data point, the change in the number of data points of the node to which it belongs can reflect whether the data point is easy to be segmented. If the data sample of the node to which the data point belongs is continuously decreasing during the process of continuously segmenting the data point, it means that the difference between the data point and other data points is large, that is, the data point is easy to be segmented. Therefore, the present invention calculates the degree of segmentation ease of the data point in the isolated tree according to the total number of data points of the node to which the data point belongs during the process of continuously segmenting.

[0035] In one embodiment, for any data point in any isolated tree, the total number of data points of the nodes to which it belongs during the continuous segmentation process is obtained.

[0036] It should be noted that in the process of constructing an isolated tree, first, all data points are taken as root nodes, and the root node is recorded as the 0th layer of the isolated tree. All data points of the root node are segmented using a preset segmentation threshold to obtain two child nodes in the 1st layer of the isolated tree. For any data point, the node where it is located in the 0th layer is the root node, and the root node is recorded as the node to which the data point belongs in the 0th layer. In the 1st layer of the isolated tree, it will be divided into a node in the 1st layer of the isolated tree, and the node is recorded as the node to which the data point belongs in the 1st layer. By analogy, the node to which any data point belongs in any layer of the isolated tree can be obtained.

[0037] In one embodiment, the degree of segmentation of the data point in the isolated tree is calculated using the following calculation formula: ; In the formula, Indicates The data point in The degree of ease of splitting in an isolated tree; Indicates The data point in The deepest level in an isolated tree; Indicates the ordinal number of the level in the isolated tree; Indicates The data point in The first The number of data points of the node to which the layer belongs; Indicates The data point in The mean number of data points of all nodes in an isolated tree; Represents the standard normalization function.

[0038] in, Indicates The data point in The first The difference between the number of data points of the nodes belonging to the layer and the average level, the larger the value is, the easier it is for the data point to be segmented, that is, The larger the value, the easier it is to split the data point in the isolated tree.

[0039] It should be noted that, due to random factors, a single isolated tree may mistakenly divide normal data points into nodes with a larger number of isolated tree layers, and divide abnormal data points into nodes with a smaller number of isolated tree layers. Therefore, the present invention calculates the similarity of the nodes of the data point in the isolated tree relative to the nodes in other isolated trees by comparing the node features of the data point in multiple isolated trees.

[0040] In one embodiment, the similarity calculation formula of the data point in the node of the isolated tree relative to the nodes in other isolated trees is as follows: ; In the formula, Indicates The data point in The degree of similarity between nodes in an isolated tree and nodes in other isolated trees; Indicates The data point in The deepest level in an isolated tree; Indicates The mean of the deepest level of all isolated trees for a data point; Indicates The data point in The degree of ease of splitting in an isolated tree; Indicates The average degree of segmentation of a data point in all isolated trees; represents the total number of isolated trees; represents the isolated tree ordinal number; Expressed as a natural constant The exponential function of base .

[0041] in, Indicates The data point in The difference between the deepest level in an isolated tree and the average level of the deepest level in all isolated trees, the smaller the value, the greater the similarity; Indicates The data point in The smaller the difference between the degree of segmentation in an isolated tree and the average level of segmentation in all isolated trees, the greater the similarity. express The smaller the confidence level, the more concentrated the easy segmentation degree of the data point in all isolated trees, that is, the data point is in the first The higher the confidence level is, the more likely it is that the splittability of an isolated tree is different from the average level of splittability of all isolated trees.

[0042] At this point, the similarity of any data point in the node of this isolated tree relative to its nodes in other isolated trees has been calculated.

[0043] S3: Calculate the validity of any isolated tree.

[0044] It should be noted that, for any isolated tree, the higher the similarity between the nodes of any data point in the isolated tree and its nodes in other isolated trees, the better the segmentation effect of the isolated tree on the data point, that is, the higher the effectiveness of the isolated tree. Therefore, the present invention calculates the effectiveness of the isolated tree by the similarity between the nodes of the data point in the isolated tree and its nodes in other isolated trees.

[0045] In one embodiment, the calculation formula for the validity of any isolated tree is as follows: ; In the formula, Indicates The validity of an isolated tree; Indicates the total number of data points; Indicates the ordinal number of the data point; Indicates The data point in The degree of similarity between nodes in an isolated tree and nodes in other isolated trees; Represents the standard normalization function.

[0046] in, It represents the average similarity between the nodes to which all data points in the isolated tree belong and the nodes to which they belong in other isolated trees. The larger the value, the more similar they are, and the higher the validity of the isolated tree. Conversely, the lower the validity of the isolated tree.

[0047] At this point, the validity calculation of any isolated tree is completed.

[0048] S4: Screening the isolated trees according to their validity, and using the screened isolated trees to complete abnormality detection of the power carrier data, thereby completing on-site fault diagnosis.

[0049] In one embodiment, if the validity of the isolated tree is less than a preset validity threshold , then the isolated tree is removed and does not participate in subsequent anomaly detection. The effective threshold is preset, and the present invention does not specifically limit it. The implementer can adjust it according to the specific implementation situation, for example .

[0050] For any data point, the average path of the data point is calculated based on the deepest layer in the remaining isolated tree after removal, and the anomaly score of the data point is further calculated. The average anomaly score of all data points is obtained. If the average anomaly score is greater than the anomaly threshold , it is determined that there is a fault and an early warning is issued to the staff. The abnormal threshold is preset, and the present invention does not specifically limit it. The implementer can adjust it according to the specific implementation situation, for example .

[0051] The calculation of the average path and the anomaly score are explained in detail in the isolation forest algorithm. In the present invention, the average path of any data point refers to the average of the deepest layers of the data point in all the remaining isolated trees. The anomaly score of the data point is calculated based on the average path of the data point in all the isolated trees. The calculation formula of the anomaly score is as follows: ; In the formula, Representing multidimensional data points The abnormal score of Representing multidimensional data points The average path length among all isolated trees; represents the total number of samples in the isolated tree; is a constant representing the expected value of the path length.

[0052] It should be noted that represents the expected value of the path length, which is not specifically limited in the present invention and can be determined by the implementer according to the actual situation. For example, ,in, represents the total number of samples in the isolated tree; The first character representing the harmonic sequence The sum of items.

[0053] At this point, the on-site fault diagnosis of the power carrier data has been completed.

[0054] Through the above steps, the on-site fault diagnosis of power line carrier data is completed.

[0055] The embodiment of the present invention further discloses an on-site fault diagnosis and testing system for HPLC, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an on-site fault diagnosis and testing method for HPLC according to the present invention is implemented.

[0056] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0057] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, an enhanced dynamic random access memory, a high bandwidth memory, a hybrid storage cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device.

[0058] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0059] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for on-site fault diagnosis and testing of HPLC, characterized in that: Includes steps: Acquire power carrier data, and construct an isolation tree according to the power carrier data; For any data point in any isolated tree, calculate the similarity between the nodes of the data point in the isolated tree and the nodes in other isolated trees. For the The data point in The deepest level in an isolated tree; For the The mean of the deepest level of all isolated trees for a data point; For the The data point in The degree of ease of splitting in an isolated tree; For the The average degree of segmentation of a data point in all isolated trees; is the total number of isolated trees; is the isolated tree ordinal number; The natural constant is an exponential function with a base; the method for obtaining the degree of easy segmentation is: calculating the difference between the number of data points of the nodes to which the data point belongs in any layer in the isolated tree and the number of data points of the nodes to which the data point belongs in all layers, and calculating the degree of easy segmentation of the data point according to the difference, and the degree of easy segmentation is positively correlated with the absolute value of the difference; For any isolated tree, the effectiveness of the isolated tree is calculated based on the similarity of all data points in the isolated tree relative to the nodes in other isolated trees, and the effectiveness is positively correlated with the similarity. The isolated trees are screened according to their effectiveness, and the screened isolated trees are used to complete the abnormal detection of power carrier data, and make diagnostic warnings to the staff.

2. A method for on-site fault diagnosis and testing of HPLC according to claim 1, characterized in that: The power carrier data includes: Signal strength, current, voltage.

3. A method for on-site fault diagnosis and testing of HPLC according to claim 1, characterized in that: The degree of easy splitting satisfies the expression: ; In the formula, Indicates The data point in The degree of ease of splitting in an isolated tree; Indicates The data point in The deepest level in an isolated tree; Indicates the ordinal number of the level in the isolated tree; Indicates The data point in The first The number of data points of the node to which the layer belongs; Indicates The data point in The mean number of data points of all nodes in an isolated tree; Represents the standard normalization function.

4. A method for on-site fault diagnosis and testing of HPLC according to claim 1, characterized in that: The belonging nodes include: For any data point of any isolated tree, the node where the data point is located in any layer of the isolated tree is recorded as the node to which the data point belongs in the layer.

5. A method for on-site fault diagnosis and testing of HPLC according to claim 1, characterized in that: The validity satisfies the expression: ; In the formula, Indicates The validity of an isolated tree; Indicates the total number of data points; Indicates the ordinal number of the data point; Indicates The data point in The degree of similarity between nodes in an isolated tree and nodes in other isolated trees; Represents the standard normalization function.

6. A method for on-site fault diagnosis and testing of HPLC according to claim 1, characterized in that: The screening of isolated trees according to the validity of the isolated trees includes: If the validity of the isolated tree is less than the preset validity threshold , then the isolated tree is removed and does not participate in subsequent anomaly detection.

7. A method for on-site fault diagnosis and testing of HPLC according to claim 1, characterized in that: The method of using the isolated tree after screening to complete the abnormal detection of the power carrier data and make a diagnosis and warning to the staff includes: For any data point, the average path of the data point is calculated according to the deepest layer number in the remaining isolated tree after removal, and the anomaly score of the data point is further calculated; the average anomaly score of all data points is obtained, and if the average anomaly score is greater than the anomaly threshold , it is judged that there is a fault and an early warning is given to the staff.

8. A field fault diagnosis test system for HPLC, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an on-site fault diagnosis test method for HPLC according to any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Power operation and maintenance data cleaning method based on isolation forest algorithm and neural network

    CN108776683A

  • High-precision spectral measurement data optimization processing method

    CN117095771A

  • Distribution network abnormal line change relation identification method based on cosine similarity and isolated forest

    CN117540317A

  • Power grid data quality abnormity diagnosis method based on data analysis

    CN117786587A

  • Coal bunker safety monitoring and early warning method and system based on artificial intelligence

    CN118447666A

Cited By

  • Data security transmission method and system of digital energy air compression station

    CN119226986A

  • Underground information data acquisition method and system

    CN119415841A

  • An underground information data acquisition method and system

    CN119415841B

  • Multi-parameter comprehensive monitoring method and system for wet desulphurization

    CN119622601A

  • Substation secondary circuit fault monitoring method

    CN119864946A