CN111881177A - Power Internet of things data flow anomaly detection system and method - Google Patents

Abstract

The invention discloses a data flow abnormality detection system of the power Internet of things, comprising a metadata acquisition module, a data mining module and a power data monitoring platform. The metadata acquisition module is connected with the data mining module through a data communication module I, and the data mining module Connected to the power data monitoring platform through the data communication module II, the metadata collection module is used to collect the working status data of each power distribution equipment in the power system, and send the collected information to the data mining module through the data communication module I, The data mining module is used to receive the working state data, and analyze the working state data abnormally according to the outlier detection algorithm model, so as to obtain the abnormal data in the power system. The invention quickly and accurately finds out abnormal data in the power system, and analyzes the cause of abnormal data generation, so as to ensure the stable operation of the power system.

Description

A system and method for detecting abnormality in data flow of power internet of things

technical field

The invention relates to the technical field of power data analysis, in particular to a system and method for detecting abnormality of data flow of the power Internet of things.

Background technique

The power system is an electric energy production and consumption system composed of power generation, substation, transmission, distribution and consumption. Its function is to convert primary energy in nature into electric energy through power generation devices (mainly including boilers, steam turbines, generators and auxiliary production systems of power plants, etc.), and then supply electric energy to various loads through transmission, transformation and distribution systems. center. Since most of the power points and load centers are located in different regions and cannot be stored in large quantities, the electricity production must always be kept in balance with the consumption. Therefore, the centralized development and decentralized use of electric energy, as well as the continuous supply of electric energy and the random change of load, restrict the structure and operation of the power system.

Outliers are also known as outliers and outliers. The most influential definition in the academic world is the definition proposed by Hawkins: “Outliers are data points that are distinctive in a data set, and their performance is so different from other points that it makes people It is suspected that these data are not random biases, but are produced by a completely different mechanism." In addition, each type of outlier detection algorithm gives the corresponding definition of outliers. Outlier detection, also known as anomaly detection, deviation detection or outlier mining, is to detect outliers in the data set according to a certain algorithm, such as detecting TOP-n outliers, or all outliers that meet the requirements. group point. In other words, outlier detection is to mine very few points in the massive data that are significantly different from the mainstream data.

Distance-based outlier detection algorithms are general. It does not require the user to have relevant domain knowledge, nor does it assume that the dataset satisfies any particular probability distribution model. In general, distance-based detection algorithms usually only need to give distance metrics between objects without additional information. Under the challenge of today's big data Variety (type), these inherent advantages have greatly improved its research and application value, and become a hot research field in academia.

Therefore, in view of the above status quo, it is urgent to develop a system and method for abnormality detection of data flow in the power Internet of Things to overcome the deficiencies in current practical applications.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a system and method for detecting abnormality of data flow of the Internet of Things in electric power, so as to solve the problems raised in the above-mentioned background art.

To achieve the above object, the present invention provides the following technical solutions:

An abnormal detection system for data flow of the Internet of Things in electric power, including a metadata collection module, a data mining module and a power data monitoring platform, the metadata collection module is connected with the data mining module through the data communication module I, and the data mining module is connected through the data communication module. II is connected to the power data monitoring platform, the metadata collection module is used to collect the working status data of each power distribution equipment in the power system, and sends the collected information to the data mining module through the data communication module I, and the data mining module uses It receives the working state data, and analyzes the abnormality of the working state data according to the outlier detection algorithm model to obtain the abnormal data in the power system.

As a further solution of the present invention: the metadata collection module includes a power data collection unit, a power data processing unit and a power data storage unit, and the power data collection unit, the power data processing unit and the power data storage unit are connected in sequence.

As a further solution of the present invention: the electric power data acquisition unit is arranged on the electric power equipment, and is used for real-time collection of the working state data of the equipment.

As a further solution of the present invention: the working state data is divided into power data and non-power data.

As a further solution of the present invention: the power data includes but is not limited to current information data, voltage information data and power information data, and the non-power data includes but is not limited to temperature information data and humidity information data.

As a further solution of the present invention: the power data processing unit is used for screening, sorting and classifying the collected working state data.

As a further solution of the present invention: the data mining module includes a power data receiving unit, a model establishing unit and a data analyzing unit, and the power data receiving unit, the model establishing unit and the data analyzing unit are connected in sequence.

As a further solution of the present invention: the power data receiving unit is used to receive the processed working state data, the model building unit establishes an outlier detection algorithm model to analyze the working state data, obtains abnormal data, and will obtain The abnormal data is sent to the data analysis unit.

A method for detecting abnormality in data flow of the Internet of Things in electric power, comprising the following steps:

S1. Collect the working status data of each power distribution equipment in the power system through the power data collection unit, and send the collected information to the data mining module through the data communication module I;

S2, the model establishment unit establishes an outlier detection algorithm model to analyze the working state data, and obtains abnormal data;

S3. The model establishment unit sends the obtained abnormal data to the data analysis unit, and analyzes the cause of the abnormal data according to the type and numerical value of the abnormal data.

As a further scheme of the present invention: in step S2, the establishment method of the outlier detection algorithm model is:

S21, the collected working state data is detected in a block manner, and each piece of detected data is called a data block;

S22. Calculate the local reachability density lrd _k (p) of each object p in the read data block:

Where: N _k (p) is the k-th distance field of point p;

reach-dist _k (p,o) is the reachable distance from point o to point p;

S23. Calculate the local outlier factor LOF _{k(p) of p} :

The closer the ratio of the local outlier factor LOF _k(p) is to 1, it means that p has a similar density of neighboring points, and p may belong to the same cluster as its neighbors; the smaller the ratio is, it means that the density of p is higher than that of its neighbors. , p is a dense point; if the ratio is greater than 1, it means that the density of p is less than the density of its neighbors, and p is an abnormal point.

Compared with the prior art, the beneficial effects of the present invention are: the present invention obtains multiple data points under one data feature of a multi-segment power Internet of Things data stream to establish a data point set, and finds the data point set through an outlier detection algorithm. and mark a section of power IoT data flow corresponding to the abnormal point as abnormal flow, quickly and accurately find abnormal data in the power system, and analyze the causes of abnormal data to ensure the stable operation of the power system.

Description of drawings

Figure 1 is a block diagram of the structure of the data flow anomaly detection system of the power Internet of things.

Figure 2 is a structural block diagram of the metadata acquisition module in the power Internet of Things data flow anomaly detection system.

Fig. 3 is a structural block diagram of the data mining module in the data flow anomaly detection system of the power Internet of things.

FIG. 4 is a flowchart of a method for detecting abnormality in data flow of the Internet of Things in electric power.

In the figure: 1-metadata acquisition module, 11-power data acquisition unit, 12-power data processing unit, 13-power data storage unit, 2-data communication module I, 3-data mining module, 31-power data receiving unit , 32 - model establishment unit, 33 - data analysis unit 33, 4 - data communication module II, 5 - power data monitoring platform.

Detailed ways

The technical solution of the present patent will be described in further detail below in conjunction with specific embodiments.

Embodiments of the present patent are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present patent, but should not be construed as a limitation on the present patent.

Example 1

Referring to FIGS. 1 to 3 , in an embodiment of the present invention, a system for detecting abnormality in data flow of the Internet of Things in electric power includes a metadata collection module 1 , a data mining module 3 and a power data monitoring platform 5 . The metadata collection module 1 passes through The data communication module I2 is connected with the data mining module 3, and the data mining module 3 is connected with the power data monitoring platform 5 through the data communication module II4;

The metadata collection module 1 is used to collect the working status data of each power distribution equipment in the power system, and send the collected information to the data mining module 3 through the data communication module I2;

The data mining module 3 is used to receive the working state data, and perform anomaly analysis on the working state data according to the outlier detection algorithm model, so as to obtain the anomalous data in the power system;

The metadata collection module 1 includes a power data collection unit 11, a power data processing unit 12 and a power data storage unit 13, and the power data collection unit 11, the power data processing unit 12 and the power data storage unit 13 are connected in sequence;

The power data collection unit 11 is arranged on the power equipment, and is used to collect the working state data of the equipment in real time. In this embodiment, the working state data is divided into power data and non-power data;

The power data includes but is not limited to current information data, voltage information data and power information data, and the non-power data includes but is not limited to temperature information data and humidity information data;

The power data processing unit 12 is used for screening, sorting and classifying the collected working state data, removing duplicate, redundant and invalid data in the data, and then sending it to the power data storage unit 13 for storage;

The data mining module 3 includes a power data receiving unit 31, a model establishing unit 32 and a data analyzing unit 33, and the power data receiving unit 31, the model establishing unit 32 and the data analyzing unit 33 are connected in sequence;

The power data receiving unit 31 is used to receive the processed working state data, and the model building unit 32 establishes an outlier detection algorithm model to analyze the working state data, obtains abnormal data, and sends the obtained abnormal data to The data analysis unit 33 analyzes the cause of the abnormal data according to the type and numerical value of the abnormal data.

Example 2

Referring to FIGS. 1 to 3 , in an embodiment of the present invention, a system for detecting abnormality in data flow of the Internet of Things in electric power includes a metadata collection module 1 , a data mining module 3 and a power data monitoring platform 5 . The metadata collection module 1 passes through The data communication module I2 is connected with the data mining module 3, and the data mining module 3 is connected with the power data monitoring platform 5 through the data communication module II4;

The metadata collection module 1 is used to collect the working status data of each power distribution equipment in the power system, and send the collected information to the data mining module 3 through the data communication module I2;

The data mining module 3 is used to receive the working state data, and perform anomaly analysis on the working state data according to the outlier detection algorithm model, so as to obtain the anomalous data in the power system;

The metadata collection module 1 includes a power data collection unit 11, a power data processing unit 12 and a power data storage unit 13, and the power data collection unit 11, the power data processing unit 12 and the power data storage unit 13 are connected in sequence;

The power data collection unit 11 is arranged on the power equipment, and is used to collect the working state data of the equipment in real time. In this embodiment, the working state data is divided into power data and non-power data;

The power data includes but is not limited to current information data, voltage information data and power information data, and the non-power data includes but is not limited to temperature information data and humidity information data;

The power data processing unit 12 is used for screening, sorting and classifying the collected working state data, removing duplicate, redundant and invalid data in the data, and then sending it to the power data storage unit 13 for storage;

The data mining module 3 includes a power data receiving unit 31, a model establishing unit 32 and a data analyzing unit 33, and the power data receiving unit 31, the model establishing unit 32 and the data analyzing unit 33 are connected in sequence;

The power data receiving unit 31 is used to receive the processed working state data, and the model building unit 32 establishes an outlier detection algorithm model to analyze the working state data, obtains abnormal data, and sends the obtained abnormal data to The data analysis unit 33 analyzes the cause of the abnormal data according to the type and numerical value of the abnormal data.

Referring to Figure 4, a method for detecting abnormality in data flow of the Internet of Things in electric power includes the following steps:

S1, collect the working state data of each power distribution equipment in the power system through the power data collection unit 11, and send the collected information to the data mining module 3 through the data communication module I2;

S2, the model establishment unit 32 establishes an outlier detection algorithm model to analyze the working state data, and obtains abnormal data;

S3, the model establishment unit 32 sends the obtained abnormal data to the data analysis unit 33, and analyzes the cause of the abnormal data generation according to the type and numerical value of the abnormal data;

Specifically, in step S2 of this embodiment, the method for establishing the outlier detection algorithm model is as follows:

S21, the collected working state data is detected in a block manner, and each piece of detected data is called a data block;

S22. Calculate the local reachability density lrd _k (p) of each object p in the read data block:

Where: N _k (p) is the k-th distance field of point p;

reach-dist _k (p,o) is the reachable distance from point o to point p;

S23. Calculate the local outlier factor LOF _{k(p) of p} :

The closer the ratio of the local outlier factor LOF _k(p) is to 1, it means that the density of points in the neighborhood of p is similar, and p may belong to the same cluster as the neighborhood; if the ratio is smaller than 1, it means that the density of p is higher than that of its neighborhood. Point density, p is a dense point; if this ratio is greater than 1, it means that the density of p is less than the density of its neighbors, and p is an abnormal point.

The above are only the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, some modifications and improvements can be made without departing from the concept of the present invention, and these should also be regarded as the protection of the present invention. scope, these will not affect the effect of the implementation of the present invention and the practicability of the patent.

Claims (10)

1. An electric power Internet of things data flow abnormity detection system is characterized by comprising a metadata acquisition module (1), a data mining module (3) and an electric power data monitoring platform (5), the metadata acquisition module (1) is connected with the data mining module (3) through the data communication module I (2), the data mining module (3) is connected with the electric power data monitoring platform (5) through the data communication module II (4), the metadata acquisition module (1) is used for acquiring the working state data of each power distribution device in the power system, and sends the acquired information to a data mining module (3) through a data communication module I (2), the data mining module (3) is used for receiving working state data, and performing anomaly analysis on the working state data according to the outlier detection algorithm model to obtain the anomaly data in the power system.

2. The power internet of things data flow abnormality detection system according to claim 1, wherein the metadata collection module (1) comprises a power data collection unit (11), a power data processing unit (12) and a power data storage unit (13), and the power data collection unit (11), the power data processing unit (12) and the power data storage unit (13) are sequentially connected.

3. The power internet of things data flow abnormality detection system according to claim 2, wherein the power data acquisition unit (11) is arranged on power equipment and used for acquiring working state data of the equipment in real time.

4. The power internet of things data flow anomaly detection system according to claim 3, wherein the working state data is divided into power data and non-power data.

5. The power internet of things data flow anomaly detection system as claimed in claim 4, wherein the power data includes, but is not limited to, current information data, voltage information data, and power information data, and the non-power data includes, but is not limited to, temperature information data and humidity information data.

6. The power internet of things data flow anomaly detection system according to claim 2, wherein the power data processing unit (12) is used for screening, sorting and classifying the acquired working state data.

7. The power internet of things data flow abnormality detection system according to claim 1, wherein the data mining module (3) includes a power data receiving unit (31), a model establishing unit (32) and a data analyzing unit (33), and the power data receiving unit (31), the model establishing unit (32) and the data analyzing unit (33) are connected in sequence.

8. The power internet of things data flow anomaly detection system according to claim 7, wherein the power data receiving unit (31) is configured to receive the processed working state data, the model establishing unit (32) establishes an outlier detection algorithm model to analyze the working state data to obtain anomaly data, and sends the obtained anomaly data to the data analysis unit (33).

9. The method for detecting the data flow abnormality of the power internet of things as claimed in any one of claims 1 to 8, characterized by comprising the following steps:

s1, acquiring working state data of each power distribution device in the power system through the power data acquisition unit (11), and sending acquired information to the data mining module (3) through the data communication module I (2);

s2, the model establishing unit (32) establishes an outlier detection algorithm model to analyze the working state data to obtain abnormal data;

s3, the model establishing unit (32) sends the obtained abnormal data to the data analysis unit (33), and the reason for generating the abnormal data is analyzed according to the type and the numerical value of the abnormal data.

10. The method for detecting the abnormality of the data flow of the internet of things in the power system according to claim 9, wherein in the step S2, the method for establishing the outlier detection algorithm model includes:

s21, detecting the acquired working state data in a block mode, wherein each detected block data is called a data block;

s22, calculating local reachable density lrd of each object p in the read data block_k(p)：

Wherein: n is a radical of_k(p) a kth distance domain for point p;

reach-dist_k(p, o) is the reachable distance from point o to point p;

s23, calculating local outlier LOF of p_k(p)：

Local outlier LOF_k(p)The closer the ratio is to 1, the more the density of the neighborhood points of p is, the more the p possibly belongs to a cluster with the neighborhood; the more the ratio is less than 1, the density of p is higher than that of the neighborhood points, and p is a dense point; the ratio is larger than 1, which indicates that the density of p is smaller than that of its neighborhood points, and p is an outlier.

Images