CN118232835A

CN118232835A - Abnormality detection method, abnormality detection system, abnormality detection device and abnormality detection storage medium for distributed photovoltaic system

Info

Publication number: CN118232835A
Application number: CN202410631222.7A
Authority: CN
Inventors: 陈静锋; 董昭阳; 欧阳晔
Original assignee: Beijing Yaxin Xingyuan Technology Co ltd
Current assignee: Beijing Yaxin Xingyuan Technology Co ltd
Priority date: 2024-05-21
Filing date: 2024-05-21
Publication date: 2024-06-21

Abstract

The application discloses an anomaly detection method, an anomaly detection system, anomaly detection equipment and an anomaly detection storage medium for a distributed photovoltaic system, and relates to the field of photovoltaic power generation, wherein the anomaly detection method comprises the following steps of: obtaining a string current parameter set of each photovoltaic matrix connected with the inverter from the inverter, inputting the string current parameter sets into a preset multi-algorithm abnormality detection engine, obtaining a plurality of initial detection results output by the preset multi-algorithm abnormality detection engine, determining a first current abnormality detection result of each string current parameter set according to a detection result label in each initial detection result by using a preset hard voting algorithm, extracting a first current abnormality parameter set in a target string current parameter set, and generating a first abnormality detection result of the distributed photovoltaic system based on the identification of the target string current parameter set and a fault type and a fault reason corresponding to a first fault characteristic parameter set with highest similarity to the first current abnormality parameter set. The application improves the operation and maintenance efficiency and the power generation efficiency of the distributed photovoltaic system.

Description

Abnormality detection method, abnormality detection system, abnormality detection device and abnormality detection storage medium for distributed photovoltaic system

Technical Field

The application relates to the technical field of photovoltaic power generation, in particular to an anomaly detection method, an anomaly detection system, an anomaly detection device and an anomaly detection storage medium for a distributed photovoltaic system.

Background

With the continuous increase of energy demand, distributed photovoltaic systems are rapidly popularized and developed due to their clean and low cost characteristics. How to accurately detect the fault position, the fault type and the fault reason of the distributed photovoltaic system, so that the operation and maintenance efficiency and the power generation efficiency of the distributed photovoltaic system are improved, and the method becomes a key link in the construction and operation processes of the distributed photovoltaic system.

Disclosure of Invention

In view of the above problems, the present application provides a method, a system, an apparatus, and a storage medium for detecting anomalies in a distributed photovoltaic system, so as to achieve the purpose of improving the operation and maintenance efficiency and the power generation efficiency of the distributed photovoltaic system. The specific scheme is as follows:

an anomaly detection method for a distributed photovoltaic system, the distributed photovoltaic system comprising: at least one inverter and at least one photovoltaic matrix electrically connected to the inverter, the method comprising:

Obtaining, from the inverter, a set of string current parameters for each of the photovoltaic matrices connected to the inverter;

Inputting each group of string current parameter groups into a preset multi-algorithm abnormality detection engine to obtain a plurality of initial detection results output by the preset multi-algorithm abnormality detection engine, wherein the initial detection results comprise identification, abnormal time and detection result labels of the group of string current parameter groups; determining a first current abnormality detection result of each group of string current parameter groups according to the detection result labels in each initial detection result by using a preset hard voting algorithm, wherein the content of the first current abnormality detection result comprises current abnormality and current normal;

And extracting a first current abnormality parameter set from a target group string current parameter set, generating a first abnormality detection result of the distributed photovoltaic system based on the identification of the target group string current parameter set and a fault type and a fault reason corresponding to a first fault characteristic parameter set with highest similarity to the first current abnormality parameter set, wherein the target group string current parameter set is the string current parameter set with the content of the first current abnormality detection result being the current abnormality, and the first current abnormality parameter set is a parameter set corresponding to the abnormality moment in the target group string current parameter set.

Optionally, the method further comprises:

Obtaining from the inverter a set of power parameters and irradiance for each of the photovoltaic matrices connected to the inverter;

For each of the power parameter sets: under the condition that the dynamic power generation power range between the power parameter set and the irradiance is within a preset power generation power range, outputting a power abnormality detection result with normal content; judging whether the duration of the power abnormality parameter in the power parameter set is not less than a preset threshold value when the upper limit value of the preset power generation power range exists in the dynamic power generation power range or the parameter smaller than the lower limit value of the preset power generation power range exists in the dynamic power generation power range, if so, outputting a power abnormality detection result with abnormal content, and if not, outputting a power abnormality detection result with normal content;

And generating a second abnormality detection result of the distributed photovoltaic system based on the identification of a target power parameter set and a fault type and a fault reason corresponding to a second fault characteristic parameter set with highest similarity of the power abnormality parameters, wherein the target power parameter set is the power parameter set with abnormal content of the power abnormality detection result.

Optionally, before the generating the second anomaly detection result of the distributed photovoltaic system, the method further includes:

obtaining a comparison group string current parameter group which is in the same sampling period as the target power parameter group and belongs to the same photovoltaic matrix as the target power parameter group;

The abnormality detection engine is used for carrying out abnormality detection on the comparison group string current parameter set, and a second current abnormality detection result is output according to the abnormality detection result by using the preset hard voting algorithm; and executing the steps when the content of the second current abnormality detection result is that the current is normal: generating a second abnormality detection result of the distributed photovoltaic system based on the identification of the target power parameter set, and a fault type and a fault cause corresponding to a second fault characteristic parameter set with the highest similarity of the power abnormality parameters;

And under the condition that the content of the second current abnormality detection result is current abnormality, extracting a second current abnormality parameter set from the comparison group string current parameter set, and generating a third abnormality detection result of the distributed photovoltaic system based on the identification of the comparison group string current parameter set, the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity to the second current abnormality parameter set, wherein the second current abnormality parameter set is a parameter set at the abnormality moment in the comparison group string current parameter set.

Optionally, the preset multi-algorithm anomaly detection engine includes a plurality of types of current anomaly detection algorithms, and the training process of the preset multi-algorithm anomaly detection engine includes:

Determining a plurality of history group string current parameter groups added with type labels in a history sampling period as training data, wherein the content of the type labels is normal or abnormal;

training each initial current abnormality detection algorithm in the initial multi-algorithm abnormality detection engine by utilizing each training data to obtain a preset multi-algorithm abnormality detection engine comprising a plurality of current abnormality detection algorithms, wherein the input of the preset multi-algorithm abnormality detection engine is a string current parameter set, and the output of the preset multi-algorithm abnormality detection engine is a plurality of initial detection results, and the initial detection results comprise identification of the string current parameter set, abnormal time and detection result labels.

An anomaly detection system for a distributed photovoltaic system, the distributed photovoltaic system comprising: at least one inverter and at least one photovoltaic matrix electrically connected to the inverter, the anomaly detection system of the distributed photovoltaic system comprising:

a first parameter obtaining module for obtaining, from the inverter, a string current parameter set for each of the photovoltaic matrices connected to the inverter;

The current anomaly detection module is used for inputting each string current parameter group into a preset multi-algorithm anomaly detection engine to obtain a plurality of initial detection results output by the preset multi-algorithm anomaly detection engine, wherein the initial detection results comprise identification, anomaly time and detection result labels of the string current parameter groups; determining a first current abnormality detection result of each group of string current parameter groups according to the detection result labels in each initial detection result by using a preset hard voting algorithm, wherein the content of the first current abnormality detection result comprises current abnormality and current normal;

The first result generating module is used for extracting a first current abnormal parameter set from a target set of string current parameter sets, generating a first abnormal detection result of the distributed photovoltaic system based on the identification of the target set of string current parameter sets and the fault type and the fault reason corresponding to a first fault characteristic parameter set with highest similarity to the first current abnormal parameter set, wherein the target set of string current parameter sets are the string current parameter sets with the content of the first current abnormal detection result being the current abnormality, and the first current abnormal parameter sets are parameter sets corresponding to the abnormality moment in the target set of string current parameter sets.

Optionally, the anomaly detection system of the distributed photovoltaic system further includes:

a second parameter obtaining module for obtaining, from the inverter, a set of power parameters and irradiance of each of the photovoltaic matrices connected to the inverter;

The power abnormality detection module is used for detecting each power parameter group: under the condition that the dynamic power generation power range between the power parameter set and the irradiance is within a preset power generation power range, outputting a power abnormality detection result with normal content; judging whether the duration of the power abnormality parameter in the power parameter set is not less than a preset threshold value when the upper limit value of the preset power generation power range exists in the dynamic power generation power range or the parameter smaller than the lower limit value of the preset power generation power range exists in the dynamic power generation power range, if so, outputting a power abnormality detection result with abnormal content, and if not, outputting a power abnormality detection result with normal content;

And the second result generation module is used for generating a second abnormality detection result of the distributed photovoltaic system based on the identification of a target power parameter set and a fault type and a fault reason corresponding to a second fault characteristic parameter set with the highest similarity of the power abnormality parameters, wherein the target power parameter set is the power parameter set with abnormal content of the power abnormality detection result.

The comprehensive detection module is used for obtaining a comparison group string current parameter group which is in the same sampling period as the target power parameter group and belongs to the same photovoltaic matrix as the target power parameter group before the second abnormal detection result of the distributed photovoltaic system is generated; the abnormality detection engine is used for carrying out abnormality detection on the comparison group string current parameter set, and a second current abnormality detection result is output according to the abnormality detection result by using the preset hard voting algorithm; and executing the steps when the content of the second current abnormality detection result is that the current is normal: generating a second abnormality detection result of the distributed photovoltaic system based on the identification of the target power parameter set, and a fault type and a fault cause corresponding to a second fault characteristic parameter set with the highest similarity of the power abnormality parameters; and under the condition that the content of the second current abnormality detection result is current abnormality, extracting a second current abnormality parameter set from the comparison group string current parameter set, and generating a third abnormality detection result of the distributed photovoltaic system based on the identification of the comparison group string current parameter set, the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity to the second current abnormality parameter set, wherein the second current abnormality parameter set is a parameter set at the abnormality moment in the comparison group string current parameter set.

Optionally, the anomaly detection system of the distributed photovoltaic system further includes: the training module comprises a plurality of types of current abnormality detection algorithms, and is set to:

An anomaly detection device for a distributed photovoltaic system comprising at least one processor and a memory coupled to the processor, wherein:

the memory is used for storing a computer program;

The processor is configured to execute the computer program to enable the abnormality detection apparatus of the distributed photovoltaic system to implement the abnormality detection method of the distributed photovoltaic system as described in any one of the above.

A computer storage medium carrying one or more computer programs which, when executed by an anomaly detection device of a distributed photovoltaic system, enable the anomaly detection device of the distributed photovoltaic system to implement the anomaly detection method of the distributed photovoltaic system as described in any one of the above.

By means of the technical scheme, the obtained serial current parameter sets are input into the preset multi-algorithm anomaly detection engine to obtain a plurality of initial detection results output by the preset multi-algorithm anomaly detection engine, and the first current anomaly detection result of each serial current parameter set is determined according to the detection result label in each initial detection result by using the preset hard voting algorithm, so that compared with a single model detection mode, the anomaly detection precision of the distributed photovoltaic system is improved. And the first abnormality detection result of the distributed photovoltaic system is generated based on the identification of the target group string current parameter group and the fault type and the fault reason corresponding to the first fault characteristic parameter group with the highest similarity to the first current abnormality parameter group, so that the determination precision of the fault position, the fault type and the fault reason is improved, and operation and maintenance personnel can timely operate and maintain according to the first abnormality detection result, and the operation and maintenance efficiency and the power generation efficiency of the distributed photovoltaic system are improved.

Drawings

The above and other features, advantages, and aspects of embodiments of the present disclosure will become more apparent by reference to the following detailed description when taken in conjunction with the accompanying drawings. The same or similar reference numbers will be used throughout the drawings to refer to the same or like elements. It should be understood that the figures are schematic and that elements and components are not necessarily drawn to scale.

FIG. 1 is a flow chart of an anomaly detection method for a distributed photovoltaic system provided by the application;

fig. 2 is a schematic structural diagram of a distributed photovoltaic system according to the present application;

FIG. 3 is a flowchart of a first current anomaly detection result provided by the present application;

FIG. 4 is a schematic diagram of a dynamic generated power range provided by the present application;

FIG. 5 is a schematic diagram of a "partial string current mutability anomaly" provided by the present application;

FIG. 6 is a schematic diagram of a "partial string current persistence anomaly" provided by the present application;

FIG. 7 is a schematic diagram of a "partial string current periodic anomaly" provided by the present application;

FIG. 8 is a flowchart of an anomaly detection method for a distributed photovoltaic system provided by the present application;

FIG. 9 is a schematic diagram of a "partial string current mutability anomaly" fault cause provided by the present application;

FIG. 10 is a schematic diagram of a "partial string current persistent anomaly" fault cause provided by the present application;

FIG. 11 is a schematic diagram of a "partial string current periodic anomaly" fault cause provided by the present application;

FIG. 12 is a block diagram of an anomaly detection system for a distributed photovoltaic system provided by the present application;

fig. 13 is a block diagram of an abnormality detection apparatus of a distributed photovoltaic system provided by the present application;

fig. 14 is a schematic diagram of a fault cause of "each group of string current parameter sets is 0" in an illumination period according to the present application.

Detailed Description

Embodiments of the present application will be described below with reference to the accompanying drawings in the embodiments of the present application. The terminology used in the description of the embodiments of the application herein is for the purpose of describing particular embodiments of the application only and is not intended to be limiting of the application.

Embodiments of the present application are described below with reference to the accompanying drawings. As one of ordinary skill in the art can know, with the development of technology and the appearance of new scenes, the technical scheme provided by the embodiment of the application is also applicable to similar technical problems.

The terms first, second and the like in the description and in the claims and in the above-described figures, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances and are merely illustrative of the manner in which embodiments of the application have been described in connection with the description of the objects having the same attributes. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of elements is not necessarily limited to those elements, but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.

The application provides an anomaly detection method of a distributed photovoltaic system, which comprises the following steps: at least one inverter and at least one photovoltaic matrix electrically connected to the inverter, as shown in fig. 1, the abnormality detection method of the distributed photovoltaic system includes:

S101, obtaining a string current parameter set of each photovoltaic matrix connected with the inverter from the inverter.

It should be noted that, in an actual application scenario, a schematic structural diagram of the above-mentioned distributed photovoltaic system is shown in fig. 2. The distributed photovoltaic system consists of a photovoltaic matrix, a combiner box, an inverter, a grid-connected cabinet, a transformer and electric equipment. The plurality of photovoltaic matrixes 201 transmit the generated current to the inverter 203 through the corresponding combiner boxes 202, the inverter 203 transmits the current subjected to frequency modulation and voltage regulation to the transformer 205 through the grid-connected cabinet 204, and the transformer 205 boosts the current and transmits the boosted current to the power consumer 206. The power utilization party comprises, but is not limited to, a power grid, a station, electric equipment and the like.

As will be appreciated by those skilled in the art, in a practical application scenario, the String current parameter set is the current output by a String (String) constructed by connecting a plurality of solar panels in series. The photovoltaic matrix is a group string formed by connecting a plurality of solar panels in series. The magnitude of the string current directly influences the generated energy of the photovoltaic matrix. Therefore, the application obtains the series current parameter group through configuration, thereby realizing the abnormality detection of the distributed photovoltaic system through the subsequent steps.

S102, inputting each group of string current parameter groups into a preset multi-algorithm abnormality detection engine to obtain a plurality of initial detection results output by the preset multi-algorithm abnormality detection engine, wherein the initial detection results comprise identification, abnormal time and detection result labels of the string current parameter groups; and determining a first current abnormality detection result of each group of string current parameter groups according to the detection result labels in each initial detection result by using a preset hard voting algorithm, wherein the content of the first current abnormality detection result comprises current abnormality and current normality.

It should be noted that, in an actual application scenario, the preset multi-algorithm anomaly detection engine is composed of a plurality of types of current anomaly detection algorithms. The existing current anomaly detection mode is mostly based on a single model, but the sensitivity and the robustness of different algorithms to data are greatly different, so that the detection accuracy of the single model is reduced. Compared with the existing single model detection mode, the accuracy of the finally obtained detection result is improved by configuring the preset multi-algorithm anomaly detection engine.

It should be noted that, in the actual application scenario, the types of algorithms in the preset multi-algorithm anomaly detection engine may be various, including but not limited to: quarter bit-spacing (Interquartile Range, IQR), standard-fraction (Z-Score) algorithms, extreme student bias (Extreme Studentized Deviate, ESD) algorithms, kernel Density Estimation (KDE) algorithms, K-means clustering (K-means clustering algorithm, K-means) algorithms, and the like. The application does not excessively limit and repeat the construction process of the preset multi-algorithm anomaly detection engine and the types of the algorithms.

In the actual application scenario, the preset hard voting algorithm (Hard Voting Classifier) is an algorithm for determining the first current anomaly detection result by voting from a plurality of initial detection results according to the detection result label in each initial detection result. Because the plurality of initial detection results are respectively generated by a plurality of algorithms in the preset multi-algorithm abnormal detection engine, in order to improve the accuracy of the finally output first current abnormal detection result and avoid the problem of result misalignment caused by single model detection, the initial detection result with the highest vote number is determined to be the final first current abnormal detection result by configuring the preset hard voting algorithm.

It should be noted that, in a practical application scenario, there are various embodiments of step S102 shown in fig. 1, and an exemplary embodiment is provided herein:

As shown in fig. 3, a flowchart for obtaining the first current anomaly detection result is provided. And inputting each group of string current parameter groups into a preset multi-algorithm anomaly detection engine. The types of the abnormality detection algorithm in the preset multi-algorithm abnormality detection engine comprise: quarter bit-spacing (Interquartile Range, IQR), standard-fraction (Z-Score) algorithms, extreme student bias (Extreme Studentized Deviate, ESD) algorithms, kernel Density Estimation (KDE) algorithms, and K-means clustering (K-means clustering algorithm, K-means) algorithms.

Based on the thought of multi-object outlier anomaly detection, each anomaly detection algorithm outputs the initial detection result according to each group of string current parameter groups. And inputting each initial detection result into a preset hard voting algorithm, so that the preset hard voting algorithm outputs a first current abnormality detection result of the string current parameter group. The content of the initial detection result may be: a (identification of group string current parameter set), 7:10 AM-8: 10AM (abnormal time), abnormality (detection result label); B. 0, normal; C. 7:10 AM-7: 15AM, normal.

As can be appreciated by those skilled in the art, in the practical application scenario, the implementation of inputting each initial detection result into the preset hard voting algorithm to make the preset hard voting algorithm output the first current anomaly detection result of the series current parameter set may be: counting according to the content of the monitoring result label of the A group string current parameter group in each initial result, if the content is that the number of the 'normal' is larger than the number of the 'abnormal' content, the content of the first current abnormality detection result of the A group string current parameter group is that the 'current is normal', otherwise, the content is that the 'current is abnormal'. The specific implementation process of the preset hard voting algorithm is not excessively limited and repeated.

Optionally, in an optional embodiment of the present application, in order to improve accuracy of the initial detection result, before each set of string current parameter sets is input to a preset multi-algorithm anomaly detection engine, each set of string current parameter sets may be preprocessed, where the preprocessing process may be:

Extracting illumination starting time and illumination ending time in the sampling period of each group of string current parameter groups, wherein the illumination starting time is the first time when the current in the group of string current parameter groups is greater than 0, and the illumination starting time is the last time when the current in the group of string current parameter groups is less than 0.

And determining the average value of the illumination starting time as the starting time, determining the average value of the illumination ending time as the ending time, and intercepting each group of serial current parameter groups based on the starting time and the ending time.

And judging whether the data quantity in each intercepted string current parameter set is not smaller than a preset threshold value, if so, inputting each string current parameter set into a preset multi-algorithm abnormality detection engine, and if not, considering that the quality of the string current parameter set is poor, and deleting.

S103, a first current abnormality parameter set is extracted from a target string current parameter set, a first abnormality detection result of the distributed photovoltaic system is generated based on the identification of the target string current parameter set and the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity to the first current abnormality parameter set, the target string current parameter set is a string current parameter set with current abnormality content, and the first current abnormality parameter set is a parameter set corresponding to an abnormality moment in the target string current parameter set.

It should be noted that, in an actual application scenario, the first fault characteristic parameter set may be a parameter set that is pre-constructed based on historical operation and maintenance data and a fault type research result.

According to the application, the obtained serial current parameter sets are input into the preset multi-algorithm anomaly detection engine to obtain a plurality of initial detection results output by the preset multi-algorithm anomaly detection engine, and the first current anomaly detection result of each serial current parameter set is determined according to the detection result label in each initial detection result by utilizing the preset hard voting algorithm, so that compared with a single model detection mode, the anomaly detection precision of the distributed photovoltaic system is improved. And the first abnormality detection result of the distributed photovoltaic system is generated based on the identification of the target group string current parameter group and the fault type and the fault reason corresponding to the first fault characteristic parameter group with the highest similarity to the first current abnormality parameter group, so that the determination precision of the fault position, the fault type and the fault reason is improved, and operation and maintenance personnel can timely operate and maintain according to the first abnormality detection result, and the operation and maintenance efficiency and the power generation efficiency of the distributed photovoltaic system are improved.

Optionally, the method for detecting an abnormality of a distributed photovoltaic system as shown in fig. 1 further includes:

obtaining from the inverter a set of power parameters and irradiance for each photovoltaic matrix connected to the inverter;

For each set of power parameters: under the condition that the dynamic power generation power range between the power parameter set and irradiance is within the preset power generation power range, outputting a power abnormality detection result with normal content; judging whether the duration of the power abnormality parameter in the power parameter set is not less than a preset threshold value when the upper limit value of the power range of the dynamic power generation is greater than the preset upper limit value of the power range of the dynamic power generation or the parameter less than the lower limit value of the power range of the dynamic power generation is present, if so, outputting a power abnormality detection result with abnormal content, and if not, outputting a power abnormality detection result with normal content;

And generating a second abnormality detection result of the distributed photovoltaic system based on the identification of the target power parameter set and the fault type and the fault reason corresponding to the second fault characteristic parameter set with the highest similarity of the power abnormality parameters, wherein the target power parameter set is the power parameter set with abnormal content of the power abnormality detection result.

It should be noted that, in a practical application scenario, the irradiance (IRRADIANCE) may be a parameter collected by a meteorological device in a deployment site of the distributed photovoltaic system. The magnitude of the irradiance directly affects the output power of the photovoltaic matrix.

It should be noted that, in an actual application scenario, the power parameter set includes output power of a photovoltaic matrix at each moment in a certain period of time.

The dynamic generated power range is a variation range of the output power of the photovoltaic matrix without abnormality under each irradiance. The upper limit value is the maximum value of the output power of the photovoltaic matrix without abnormality under each irradiance, and the lower limit value is the minimum value of the output power of the photovoltaic matrix without abnormality under each irradiance. Since irradiance is proportional to output power, i.e., the higher the irradiance value, the higher the output power over a range of times. For example, as shown in fig. 4, a schematic diagram of the dynamic power generation range is shown in fig. 4, wherein the black dots are the values of the dynamic power generation range at each time. Curve 401 in fig. 4 represents the trend of the above-described preset generation power range upper limit value. Curve 402 in fig. 4 represents the trend of the lower limit value of the preset generated power range.

In the practical application scene, the method and the device realize the preliminary detection of whether the photovoltaic matrix has power abnormality by configuring and judging whether the dynamic power generation power range is within the preset power generation power range. And then, calculating the similarity between the power abnormality parameter and a plurality of second fault characteristic parameter sets, so that a second abnormality detection result of the distributed photovoltaic system is generated based on the fault type and the fault reason corresponding to the second fault characteristic parameter set with the highest similarity, thereby realizing the fine determination of the specific abnormality reason of the photovoltaic matrix and improving the abnormality detection precision.

Optionally, before generating the second anomaly detection result of the distributed photovoltaic system, the anomaly detection method of the distributed photovoltaic system shown in fig. 1 further includes:

performing anomaly detection on the string current parameter sets by utilizing a preset multi-algorithm anomaly detection engine, and outputting a second current anomaly detection result according to the anomaly detection result by utilizing a preset hard voting algorithm; and under the condition that the content of the second current abnormality detection result is that the current is normal, executing the steps of: generating a second abnormality detection result of the distributed photovoltaic system based on the identification of the target power parameter set and the fault type and the fault cause corresponding to the second fault characteristic parameter set with the highest similarity of the power abnormality parameters;

And under the condition that the content of the second current abnormality detection result is current abnormality, extracting a second current abnormality parameter set from the comparison string current parameter set, and generating a third abnormality detection result of the distributed photovoltaic system based on the identification of the comparison string current parameter set and the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity to the second current abnormality parameter set, wherein the second current abnormality parameter set is a parameter set at an abnormality moment in the comparison string current parameter set.

It should be noted that, in the practical application scenario, the power abnormality may be caused at the same time due to the partial fault that causes the string current, such as an illumination problem, the photovoltaic panel being damaged, the photovoltaic panel being covered by a foreign object, the photovoltaic panel being blocked by an object from illumination, a junction box fault, and the like. Power faults may also be initiated when the string current produces a periodic or persistent fault due to the fault described above, but such faults may be confused with the inverter itself. Therefore, the comparison group string current parameter group which is in the same sampling period as the target power parameter group and belongs to the same photovoltaic matrix as the target power parameter group is obtained by configuration before the second abnormality detection result of the distributed photovoltaic system is generated. And comparing the group string current parameter groups by using a preset multi-algorithm abnormality detection engine to perform abnormality detection, thereby realizing accurate identification of faults which cause abnormality of power and group string current at the same time and improving the accuracy of abnormality detection.

It should be noted that, in a practical application scenario, the specific types of the first fault characteristic parameters are various, and 3 types are provided herein by way of example:

FIG. 5 is a schematic diagram showing that the fault type is "partial string current mutability anomaly". In FIG. 5, the horizontal axis represents the time axis from 12:00 on 4/30 days to 12:00 on 5/6 days, the vertical axis represents the string current axis, the black point represents the normal parameter in the string current parameter sets, and the white point represents the abnormal parameter. From 5/1 day 00 in FIG. 5: as can be seen from the string current distribution of 00 to 5/5 days 00:00, the normal string current parameters are 12: the current distribution of the strings is uniform in the process of 00 to 00:00 in a descending trend. However, at day 5/5, 12:00 to day 5/6, 00:00, a few significant deviations of the string current parameters occur, and this type of string current parameter does not occur in the period preceding day 5/5. Thus, the fault type is "partial string current mutability anomaly".

The causes of the failure that caused the "partial string current mutability anomaly" are analyzed as shown in fig. 9, including: illumination effects, photovoltaic panels damaged, photovoltaic panels covered by foreign matter, and header failure. However, since the illumination effect affects most of the group string current parameter sets, the probability that the abnormality of a small group string current parameter set is affected by illumination is low, the fault of the combiner box affects all of the group string current parameter sets, and the probability that the abnormality of a small group string current parameter set is affected by the fault of the combiner box is low. A few outliers as shown in fig. 5 do not occur. While damage to the photovoltaic panel and coverage of the photovoltaic panel by foreign matter may cause current mutability abnormality and affect only a single photovoltaic panel, it can be determined that the cause of the failure type, which causes "partial string current mutability abnormality", is that the photovoltaic panel is damaged or the photovoltaic panel is covered by foreign matter.

Fig. 6 is a schematic diagram showing that the fault type is "partial string current persistence abnormality". The horizontal and vertical axes in fig. 6 represent the same meaning as those in fig. 5, and the black point is a normal parameter in the plurality of sets of series current parameters and the white point is an abnormal parameter. As can be seen from fig. 6, in each period, there are a plurality of abnormal parameters, and the time period, the number of the abnormal parameters and the corresponding string current values are similar, so that the abnormal parameters are persistent in the fault type. Thus, the fault type is "partial string current persistence abnormality".

The causes of the failure causing the "partial string current persistence abnormality" are analyzed as shown in fig. 10, including: illumination effects, photovoltaic panels damaged, photovoltaic panels covered by foreign matter, photovoltaic panel installation angle errors, and header failure. However, since the illumination effect affects most of the group string current parameter sets, the probability that the abnormality of a small group string current parameter set is affected by illumination is low, the fault of the combiner box affects all of the group string current parameter sets, and the probability that the abnormality of a small group string current parameter set is affected by the fault of the combiner box is low. A few outliers as shown in fig. 6 do not occur. And the damage to the photovoltaic panel, the covering of the photovoltaic panel by the foreign matter, and the error of the installation angle of the photovoltaic panel may cause the sustainability abnormality of the string current parameter group of the minority photovoltaic panel group, so it can be determined that the failure cause of the failure type of "partial string current sustainability abnormality" is that the photovoltaic panel is damaged, the photovoltaic panel is covered by the foreign matter, or the installation angle of the photovoltaic panel is wrong.

As shown in fig. 7, the fault type is a schematic diagram of "partial string current periodic anomaly". The horizontal and vertical axes in fig. 7 represent the same meaning as those in fig. 5, and the black point is a normal parameter in the plurality of sets of series current parameters and the white point is an abnormal parameter. As can be seen from fig. 7, in each cycle, there are a plurality of abnormal parameters that have an upward trend, and the abnormal parameters do not last for the whole cycle. Thus, the fault type is "partial string current periodic anomaly".

The causes of the failure causing the "partial string current periodicity abnormality" are analyzed as shown in fig. 11, including: illumination effect, photovoltaic panel is damaged, photovoltaic panel is covered by the foreign matter, and photovoltaic panel is sheltered from illumination and collection flow box trouble by the object. But affects most of the series current parameter sets due to illumination and is transient; damage to the photovoltaic panel can continuously affect the string current parameter set; covering the photovoltaic panel with foreign matter can continuously affect the string current parameter set; a combiner box fault affects all string current parameter sets. Not a small number of periodic anomalies as shown in fig. 7. The light shielding of the photovoltaic panel by the object can influence a few series current parameter groups and show periodicity. It can thus be determined that the cause of the failure type that causes "the partial string current to be periodically abnormal" is that the photovoltaic panel is shielded from light by the object.

It should be noted that, if the fault type is "the string current parameter sets in the illumination period are all 0", the schematic diagram of the fault cause is shown in fig. 14. The horizontal and vertical axes in fig. 14 represent the same meaning as those in fig. 5, and the black point is a normal parameter in the plurality of sets of series current parameters and the white point is an abnormal parameter. As can be seen from fig. 14, at 5/5 day 12:00 to 5/6 day 00:00, a plurality of consecutive white points appear, indicating that all the group string currents are 0 in the illumination time range, and the failure is "combiner box failure". The fault may be detected by setting 0 as a preset threshold, and monitoring whether each group of string current parameter sets is not greater than the preset threshold in the illumination period, if so, indicating that there is an abnormality in the fault type.

The fault type and the fault cause of the first fault characteristic parameter are finely divided, so that the abnormality detection precision of the distributed photovoltaic system is improved.

Optionally, the preset multi-algorithm anomaly detection engine includes a plurality of types of current anomaly detection algorithms, and a training process of the preset multi-algorithm anomaly detection engine includes:

Training each initial current abnormality detection algorithm in the initial multi-algorithm abnormality detection engine by utilizing each training data to obtain a preset multi-algorithm abnormality detection engine comprising a plurality of current abnormality detection algorithms, wherein the input of the preset multi-algorithm abnormality detection engine is a string current parameter set, and the output of the preset multi-algorithm abnormality detection engine is a plurality of initial detection results, and the initial detection results comprise identification of the string current parameter set, abnormality time and detection result labels.

In the practical application scenario, various embodiments of the foregoing abnormality detection method of the distributed photovoltaic system shown in fig. 1 are provided, and an example is as follows:

As shown in fig. 8, a flowchart of an anomaly detection method of a distributed photovoltaic system is shown, and specific operation steps are as follows:

Step S801, irradiance, a string current parameter set and a power parameter set of each photovoltaic matrix are obtained. And triggers step S802 and step S803.

In the practical application scenario, the irradiance and the power parameter set of each photovoltaic matrix in step S801 shown in fig. 8 may be obtained simultaneously with the string current parameter set of each photovoltaic matrix, or may be obtained step by step.

Step S802, inputting each group of string current parameter groups into a preset multi-algorithm anomaly detection engine, and determining a first current anomaly detection result of each group of string current parameter groups by utilizing a preset hard voting algorithm according to a detection result label in a plurality of initial detection results output by the preset multi-algorithm anomaly detection engine. And triggers step S804.

Step S803, for each power parameter set: a dynamic generated power range between the set of power parameters and irradiance is obtained. And triggers step S805.

It should be noted that, in the actual application scenario, the step S802 and the step S803 shown in fig. 8 may be executed simultaneously, or may be executed sequentially, and the execution sequence of the step S802 and the step S803 is not limited too much in the present application.

Step S804, for each group of string current parameter sets: and judging whether the content of the first current abnormality detection result of the series current parameter set is current abnormality or not. If yes, step S806 is triggered, and if no, step S807 is triggered.

Step S805, for each dynamic generated power range: and judging whether the dynamic power generation power range is within a preset power generation power range. If yes, step S808 is triggered, and if no, step S809 is triggered.

Step S806, for each group of string current parameter sets whose contents of the first current abnormality detection result are current abnormality: and extracting a first current abnormal parameter set from the series current parameter set, and generating a first abnormal detection result based on the identification of the series current parameter set, the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity to the first current abnormal parameter set.

Step S807, for each group of string current parameter groups whose content of the first current abnormality detection result is not a current abnormality: and judging whether parameters smaller than a preset threshold value exist in the series current parameter set. If yes, step S810 is triggered, and if no, step S811 is triggered.

Step S808, outputting the power abnormality detection result whose content is normal.

Step S809, for each power parameter set whose dynamic generated power range is not within the preset generated power range: and judging whether the duration time of the power abnormality parameter in the power parameter set is not less than a preset threshold value. If yes, step S808 is triggered, and if no, step S812 is triggered.

Step S810, outputting a static threshold detection result with current abnormality.

In step S811, the output content is the static threshold detection result with normal current.

Step S812, for each power parameter set whose content of the power abnormality detection result is abnormal: and extracting comparison group string current parameter groups which are in the same sampling period as the power parameter groups and belong to the same photovoltaic matrix from the string current parameter groups. And triggers step S813.

Step S813, the comparison string current parameter set is input to the preset multi-algorithm anomaly detection engine, and a second current anomaly detection result of the comparison string current parameter set is determined according to the detection result labels in the plurality of initial detection results output by the preset multi-algorithm anomaly detection engine by using the preset hard voting algorithm. And triggers step S814.

Step S814, it is determined whether the content of the second current anomaly detection result of the comparison string current parameter set is a current anomaly. If yes, step S815 is triggered, and if not, step S816 is triggered.

Step S815, the content of the second current anomaly detection result is each comparison set of current parameter sets of current anomalies: and extracting a second current abnormal parameter set from the comparison group string current parameter set, and generating a third abnormal detection result based on the identification of the comparison group string current parameter set, the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity with the second current abnormal parameter set.

Step S816, for each power parameter set whose content of the power abnormality detection result is abnormal: and generating a second abnormality detection result based on the identification of the power parameter set and the fault type and the fault reason corresponding to the second fault characteristic parameter set with the highest similarity of the power abnormality parameters.

Corresponding to the embodiment of the method, the application also provides an abnormality detection system of the distributed photovoltaic system, and the distributed photovoltaic system comprises: at least one inverter and at least one photovoltaic matrix electrically connected to the inverter, as shown in fig. 12, the abnormality detection system of the distributed photovoltaic system includes:

A first parameter obtaining module 1201, configured to obtain, from the inverter, a string current parameter set of each photovoltaic matrix connected to the inverter;

The current anomaly detection module 1202 is configured to input each group of string current parameter sets to a preset multi-algorithm anomaly detection engine, and obtain a plurality of initial detection results output by the preset multi-algorithm anomaly detection engine, where the initial detection results include an identifier of the group string current parameter set, an anomaly time and a detection result label; determining a first current abnormality detection result of each group of serial current parameter groups according to detection result labels in each initial detection result by using a preset hard voting algorithm, wherein the content of the first current abnormality detection result comprises current abnormality and current normality;

the first result generating module 1203 is configured to extract a first current anomaly parameter set from a target string current parameter set, and generate a first anomaly detection result of the distributed photovoltaic system based on an identifier of the target string current parameter set and a fault type and a fault cause corresponding to a first fault feature parameter set with a highest similarity to the first current anomaly parameter set, where the target string current parameter set is a string current parameter set with a content of the first current anomaly detection result being a current anomaly, and the first current anomaly parameter set is a parameter set corresponding to an anomaly time in the target string current parameter set.

Optionally, the abnormality detection system of the distributed photovoltaic system shown in fig. 12 further includes:

a second parameter obtaining module for obtaining a power parameter set and irradiance of each photovoltaic matrix connected with the inverter from the inverter;

The power anomaly detection module is used for detecting each power parameter group: under the condition that the dynamic power generation power range between the power parameter set and irradiance is within the preset power generation power range, outputting a power abnormality detection result with normal content; judging whether the duration of the power abnormality parameter in the power parameter set is not less than a preset threshold value when the upper limit value of the power range of the dynamic power generation is greater than the preset upper limit value of the power range of the dynamic power generation or the parameter less than the lower limit value of the power range of the dynamic power generation is present, if so, outputting a power abnormality detection result with abnormal content, and if not, outputting a power abnormality detection result with normal content;

the second result generating module is used for generating a second abnormality detection result of the distributed photovoltaic system based on the identification of the target power parameter set, the fault type and the fault reason corresponding to the second fault characteristic parameter set with the highest similarity of the power abnormality parameters, wherein the target power parameter set is the power parameter set with abnormal content of the power abnormality detection result.

The comprehensive detection module is used for obtaining a comparison group string current parameter group which is in the same sampling period as the target power parameter group and belongs to the same photovoltaic matrix as the target power parameter group before generating a second abnormal detection result of the distributed photovoltaic system; performing anomaly detection on the string current parameter sets by utilizing a preset multi-algorithm anomaly detection engine, and outputting a second current anomaly detection result according to the anomaly detection result by utilizing a preset hard voting algorithm; and under the condition that the content of the second current abnormality detection result is that the current is normal, executing the steps of: generating a second abnormality detection result of the distributed photovoltaic system based on the identification of the target power parameter set and the fault type and the fault cause corresponding to the second fault characteristic parameter set with the highest similarity of the power abnormality parameters; and under the condition that the content of the second current abnormality detection result is current abnormality, extracting a second current abnormality parameter set from the comparison string current parameter set, and generating a third abnormality detection result of the distributed photovoltaic system based on the identification of the comparison string current parameter set and the fault type and the fault reason corresponding to the first fault characteristic parameter set with the highest similarity to the second current abnormality parameter set, wherein the second current abnormality parameter set is a parameter set at an abnormality moment in the comparison string current parameter set.

Optionally, the abnormality detection system of the distributed photovoltaic system shown in fig. 12 further includes: the training module is used for training the preset multi-algorithm abnormality detection engine and is set to:

The embodiment of the application also provides an abnormality detection device of the distributed photovoltaic system, as shown in fig. 13, including at least one processor 1301 and a memory 1302 connected with the processor 1301, wherein:

The memory 1302 is for storing a computer program;

the processor 1301 is configured to execute a computer program to enable the abnormality detection apparatus of the distributed photovoltaic system to implement the abnormality detection method of the distributed photovoltaic system as any one of the above.

The embodiment of the application also provides a computer storage medium, which carries one or more computer programs, and when the one or more computer programs are executed by the abnormality detection equipment of the distributed photovoltaic system, the abnormality detection equipment of the distributed photovoltaic system can realize the abnormality detection method of the distributed photovoltaic system.

It should be further noted that the above-described apparatus embodiments are merely illustrative, and that the units described as separate units may or may not be physically separate, and that units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the drawings of the embodiment of the device provided by the application, the connection relation between the modules represents that the modules have communication connection, and can be specifically implemented as one or more communication buses or signal lines.

From the above description of the embodiments, it will be apparent to those skilled in the art that the present application may be implemented by means of software plus necessary general purpose hardware, or of course by means of special purpose hardware including application specific integrated circuits, special purpose CPUs, special purpose memories, special purpose components, etc. Generally, functions performed by computer programs can be easily implemented by corresponding hardware, and specific hardware structures for implementing the same functions can be varied, such as analog circuits, digital circuits, or dedicated circuits. But a software program implementation is a preferred embodiment for many more of the cases of the present application. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a readable storage medium, such as a floppy disk, a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk or an optical disk of a computer, etc., comprising several instructions for causing a computer device (which may be a personal computer, a training device, a network device, etc.) to perform the method according to the embodiments of the present application.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product.

The computer program product includes one or more computer instructions. When loaded and executed on a computer, produces a flow or function in accordance with embodiments of the present application, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via a wired (e.g., coaxial cable, optical fiber, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be stored by a computer or a data storage device such as a training device, a data center, or the like that contains an integration of one or more available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (Solid STATE DISK, SSD)), etc.

Claims

1. An anomaly detection method for a distributed photovoltaic system, the distributed photovoltaic system comprising: at least one inverter and at least one photovoltaic matrix electrically connected to the inverter, the method comprising:

2. The method according to claim 1, wherein the method further comprises:

3. The method of claim 2, wherein prior to the generating the second anomaly detection result for the distributed photovoltaic system, the method further comprises:

4. The method of claim 1, wherein the pre-set multi-algorithm anomaly detection engine comprises a plurality of types of current anomaly detection algorithms, a training process of the pre-set multi-algorithm anomaly detection engine comprising:

5. An anomaly detection system for a distributed photovoltaic system, the distributed photovoltaic system comprising: at least one inverter and at least one photovoltaic matrix electrically connected to the inverter, the anomaly detection system of the distributed photovoltaic system comprising:

6. The anomaly detection system of a distributed photovoltaic system of claim 5, further comprising:

7. The anomaly detection system of a distributed photovoltaic system of claim 6, further comprising:

8. The anomaly detection system of a distributed photovoltaic system of claim 5, further comprising: the training module comprises a plurality of types of current abnormality detection algorithms, and is set to:

9. An anomaly detection device for a distributed photovoltaic system, comprising at least one processor and a memory coupled to the processor, wherein:

the memory is used for storing a computer program;

The processor is configured to execute the computer program to enable the abnormality detection apparatus of a distributed photovoltaic system to implement the abnormality detection method of a distributed photovoltaic system according to any one of claims 1 to 4.

10. A computer storage medium carrying one or more computer programs which, when executed by an anomaly detection device of a distributed photovoltaic system, enable the anomaly detection device of the distributed photovoltaic system to implement the anomaly detection method of the distributed photovoltaic system of any one of claims 1 to 4.