CN115800916A - Intelligent I-V diagnosis system for photovoltaic power station - Google Patents

Abstract

The invention relates to an intelligent I-V diagnosis system of a photovoltaic power station, which comprises a data acquisition, screening and processing module, a fault type mismatching judgment module, a fault diagnosis analysis module and a fault diagnosis result comparison and verification module. The method and the system can realize real-time diagnosis of the photovoltaic power station string equipment, support fault detail analysis of the alarm equipment, judge fault types and give fault reason analysis and processing suggestions. The inverter, the combiner box and the photovoltaic branch power generation abnormity can be analyzed and diagnosed in real time, and low-efficiency operation equipment is rapidly screened out, so that an auxiliary manager can perform routing inspection and defect elimination work in a planned way, and the routing inspection and defect elimination work efficiency is improved. The fault diagnosis and analysis module can realize the full detection and fault identification of the photovoltaic power station string and the inverter. The data acquisition, screening and processing module effectively improves the reliability and authenticity of the data, and the fault diagnosis result comparison and verification module ensures the accuracy and reliability of the fault diagnosis result.

Description

Intelligent I-V diagnosis system for photovoltaic power station

Technical Field

The invention belongs to the technical field of photovoltaics, and particularly relates to an intelligent I-V diagnosis system for a photovoltaic power station.

Background

The intelligent I-V diagnosis system is an I-V diagnosis algorithm based on big data, and realizes diagnosis of core key equipment such as a photovoltaic array, a combiner box and an inverter through fault accurate positioning. The system extracts the characteristics of the inverter, the combiner box and the photovoltaic branch circuit under abnormal conditions, establishes a fault diagnosis model, confirms the fault type of the photovoltaic string by combining a big data excavator AI identification algorithm based on machine learning, analyzes and diagnoses the inverter, the combiner box and the photovoltaic branch circuit power generation abnormity in real time, and rapidly screens out low-efficiency operation equipment, so that a manager is assisted to carry out routing inspection and defect elimination work in a planned way, and the routing inspection and defect elimination work efficiency is improved. The real-time diagnosis of the photovoltaic power station string equipment is realized, the fault detail analysis of the alarm equipment is supported, the fault type is judged, and the fault reason analysis and processing suggestion are given.

The conventional intelligent I-V diagnosis system can judge whether the photovoltaic system has a fault or not by comparing an I-V curve of the photovoltaic system obtained through measurement with a theoretical curve.

Firstly, scanning to obtain I-V data of a photovoltaic module to be diagnosed, and reading the open-circuit voltage of a nameplate

Ambient temperature T during fault diagnosis _e . T is determined because the open circuit voltage of the photovoltaic module is reduced along with the increase of the temperature _e Open circuit voltage of lower normal photovoltaic module

And

the relationship of (1) is:

wherein K _T The photovoltaic module K is the open-circuit voltage temperature coefficient of the photovoltaic module and made of different materials with different service lives _T The sizes are different, and after a photovoltaic module sample to be diagnosed is determined, K can be measured by a temperature coefficient measuring method specified in ASTME 1036-2002 Standard test method for non-energy-gathering ground photoelectric module and array Electrical Performance Using Standard Battery _T The value is obtained.

Comparison U _oc And with

The comparison error margin is improved by setting the voltage interval. When the temperature is higher than the set temperature

When the voltage is larger than the set voltage, the I-V curve of the component is not a voltage loss curve, and the next diagnosis process is started; when the temperature is higher than the set temperature

When the voltage is larger than the voltage of the substring, the I-V curve of the component is of a voltage loss type, and one substring voltage is reduced; when in use

When the voltage is not 0, the curve of the component is also of a voltage loss type, and two substring voltages are reduced; when in use

At 0, the component has no I-V curve and its output power is completely lost.

To U _dN >2/3U _OC The photovoltaic modules perform I-V data analysis to obtain d of the photovoltaic modules _maxN -U _dN And (4) respectively diagnosing a single-step curve, a double-step curve and a current collapse curve according to the distinguishing standard provided in the previous section. The fault diagnosis process is shown in fig. 1.

The existing intelligent I-V diagnosis system has the following technical defects:

1) The prior art scheme lacks the process of screening and processing the acquired data such as voltage, current and the like, and can influence the accuracy and reliability of fault diagnosis.

2) The prior art scheme does not obviously divide the fault types of the photovoltaic module, which can cause the fault diagnosis of the photovoltaic module to lack pertinence and influence the efficiency of the fault diagnosis of the photovoltaic module and the customized generation of a solution.

3) The fault diagnosis method in the prior art is single, and the diagnosis result is lack of comparison, calibration and verification, so that the accuracy and reliability of the diagnosis result are lack of powerful support.

Disclosure of Invention

The invention aims to provide an intelligent I-V diagnosis system of a photovoltaic power station, which solves the problems of insufficient accuracy and reliability of I-V intelligent diagnosis of the photovoltaic power station and improves the fault diagnosis precision by carrying out conformity judgment and smoothness processing on photovoltaic I-V data; the method comprises the steps that photovoltaic module faults are sorted and divided in an intelligent I-V diagnosis system of a photovoltaic power station, a fault knowledge base is created, and customized diagnosis and solution schemes are provided for various photovoltaic module faults in a targeted mode; by operating different photovoltaic module fault diagnosis methods, fault diagnosis results are compared and calibrated, the problem that the reliability of the photovoltaic module fault diagnosis results is insufficient is solved, and the accuracy and precision of photovoltaic fault diagnosis are improved.

The invention provides an intelligent I-V diagnosis system of a photovoltaic power station, which comprises a data acquisition, screening and processing module, a fault type mismatching judgment module, a fault diagnosis analysis module and a fault diagnosis result comparison and verification module;

the data acquisition, screening and processing module is used for data acquisition, data conformity judgment and data smoothing; the data conformity judgment comprises the following steps: in the normal case of component I-V data (U) ₀ ,I ₀ ) Corresponding to the open circuit voltage point, judging the current at the theoretical open circuit voltage point to judge whether the I-V data is missing, if so, judging whether the I-V data is missing ₀ If the value is less than or equal to 0.05A, the I-V curve meets the requirement, otherwise, the data of the I-V curve is abnormal; for the condition that the open-circuit voltage lacks I-V data, the group of data is no longer used for fault diagnosis, the I-V data is not matched, and the output data is abnormal; the data smoothing process includes: voltage, current data (U) for photovoltaic modules _i ,I _i ) Wherein, i =0,1,2,3 …,31, the corresponding voltage data remains unchanged, when the current data Ii satisfies the following condition:

or is provided with

The current data is smoothed as follows:

I _i ＝(I _i-1 +I _i+1 )/2

the above process is circulated for 30 times to obtain smoothed data, and the data are arranged according to the sequence of voltage from large to small (U) _n ,I _n ) Where n =0,1,2,3 …,31;

the fault type mismatch judging module is used for judging whether the assembly is mismatched or not according to the concavity and convexity of the I-V curve of the assembly, according to the direction from low to high of the I-V curve, a point of concavity and convexity change is specified to be a lower inflection point, namely a low voltage point, and a point of a stage is specified to be an upper inflection point, namely a high voltage point, through inflection point detection of the I-V curve, a voltage interval of the stage is determined, and fault characteristics of shadow, hot spot and glass breakage are further decoupled according to data of the stage of the I-V curve;

the fault diagnosis and analysis module is used for mismatch fault diagnosis and non-mismatch fault diagnosis;

the mismatch fault diagnosis comprises a mismatch fault diagnosis method based on I-V curve concavity and convexity and a mismatch fault diagnosis method based on I-V curve interval division and characteristic points;

the mismatch fault diagnosis method based on the I-V curve concavity and convexity adopts an I-V curve fixed line variable step length detection method, which comprises the following steps:

by short-circuit current point (U) on the I-V curve ₃₁ ,I ₃₁ ) And open circuit voltage point (U) ₀ ,I ₀ ) The slope constructs a detection line, a plurality of groups of balanced line clusters are obtained, and the slope and the expression of a line equation are as follows:

I _N ＝kU _N +b _I

wherein b is _I For detecting the step length of the change of the line, also the intersection of the line and the current axis, b _I Is the short-circuit current I ₃₁ ，b _I Does not exceed twice the short-circuit current value, i.e. 2I ₃₁ When the straight line passes through the virtual maximum power point, the I-V curve is completely below the detection straight line, i.e. I _N <k*U _N +b _I ；

The two points adjacent to the point are also positioned below the straight line, i.e. I _N-1 <k*U _N-1 +b _I ，I _N+1 <k*U _N+1 +b _I ，I _N-1 >0；

The other two points adjacent to the point are located above the straight line, I _N-4 <k*U _N-4 +b _I ，I _N+4 <k*U _N+4 +b _I ，I _N-4 >0；

If the I-V curve has a lower inflection point, the comparison is stopped, the photovoltaic module has mismatch, and the coordinates (U) of the points meeting the condition are recorded _n ，I _n ) Namely the voltage and the current at the lower inflection point; if not, change b _I A value of (b) _I ＝b _I + Step _ b, compare with next straight line, repeat the above judgment until b _I If the upper limit value of (c) still does not satisfy the above condition, then no mismatch exists; determining the lower inflection point of the stage according to the above method, and determining the lower inflection point (U) according to the second determined line _n ，I _n ) And open circuit voltage point (U) ₀ ，I ₀ ) Calculating the maximum value of the I-V data from the open-circuit voltage to the upper inflection point, and the coordinate (U) corresponding to the maximum value _m ，I _m ) That is, the voltage and current values corresponding to the end points of the step sections, and the step interval is (U) _n ，U _m ) (ii) a The equation of the step end point detection straight line and the distance expression from the data point on the I-V curve to the detection straight line are as follows:

I＝kU+b

detecting the coordinate of a lower inflection point under the condition of mismatch, constructing a stage end point of a fixed linear detection platform by using the lower inflection point and an open-circuit voltage point, and determining the stage end point according to the maximum value of the distance from I-V data in the range from the lower inflection point to the open-circuit voltage point to a detected linear distance;

the non-mismatch fault diagnosis comprises an optimal component comparison method and a key parameter discrimination method;

the key parameter discrimination method comprises the following steps: judging the fault type through the combination of the key parameters;

for a power station application scene, the volt-ampere characteristic of a photovoltaic string is similar to that of a classical single-diode model, and the relationship between current and voltage is represented by the following classical formula:

in the formula, I: working current of the photovoltaic string; i is _L : generating current by the photovoltaic string; io: diode reverse saturation current; n: a diode ideality factor; v: operating voltage of the photovoltaic string; rs is a series resistor of the photovoltaic group string; q: single electron electric quantity; k: boltzmann constant; t: a thermodynamic temperature;

for mismatch fault diagnosis, the fault diagnosis and analysis module adopts a mismatch fault diagnosis method based on I-V curve concavity and convexity and a mismatch fault diagnosis method based on I-V curve interval division and characteristic points to model, simulate and analyze, and the two methods are synchronously carried out to obtain an analysis result; for non-mismatch fault diagnosis, the fault diagnosis analysis module adopts a key parameter discrimination method and an optimal component comparison method to model, simulate and analyze, and synchronously carries out the two methods to obtain an analysis result;

the fault diagnosis result comparison and verification module is used for comparing and verifying fault diagnosis analysis results obtained by adopting two methods of mismatch fault diagnosis or two methods of non-mismatch fault diagnosis, and outputting a fault diagnosis result if the result comparison is consistent or smaller than a reasonable error interval; and if the result comparison is different, feeding back the analysis result to have difference, and recalculating and analyzing until an accurate optimal analysis result is obtained.

By means of the scheme, the intelligent I-V diagnosis system for the photovoltaic power station has the following technical effects:

1) The diagnosis system can realize real-time diagnosis of the photovoltaic power station string equipment, support fault detail analysis of the alarm equipment, judge fault types and give fault reason analysis and processing suggestions.

2) The diagnosis system can analyze and diagnose the power generation abnormity of the inverter, the header box and the photovoltaic branch in real time and quickly screen out low-efficiency operation equipment, so that an auxiliary manager can perform routing inspection and defect elimination work in a planned way, and the routing inspection and defect elimination work efficiency is improved.

3) The fault diagnosis analysis module of the diagnosis system can realize the full detection and fault recognition of the photovoltaic power station string and the inverter, automatically generate an IV diagnosis operation and maintenance report after the completion, and provide different operation and maintenance suggestions aiming at different fault types.

4) The diagnosis system effectively improves the reliability and authenticity of data through the data acquisition, screening and processing module, and ensures the accuracy and reliability of the fault diagnosis result through the fault diagnosis result comparison and verification module.

The foregoing description is only an overview of the technical solutions of the present invention, and in order to make the technical solutions of the present invention more clearly understood and to implement them in accordance with the contents of the description, the following detailed description is given with reference to the preferred embodiments of the present invention and the accompanying drawings.

Drawings

FIG. 1 is a flow chart of I-V diagnosis of a photovoltaic module of a conventional intelligent I-V diagnosis system;

FIG. 2 is a block diagram of an intelligent I-V diagnostic system for a photovoltaic power plant according to the present invention;

FIG. 3 is a flow chart of the component I-V data conformance determination algorithm of the present invention;

FIG. 4 is a flow chart of the component I-V data smoothing algorithm of the present invention;

FIG. 5 is a schematic diagram of a curve fixed-line variable-step-size detection method according to the present invention;

FIG. 6 is a flow chart of a non-mismatch diagnostic algorithm of the key parameter discrimination method of the present invention.

Detailed Description

The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.

Referring to fig. 2 to 6, the present embodiment provides an intelligent I-V diagnosis system applicable to a photovoltaic station, and the system can implement diagnosis of core key devices such as a photovoltaic array, a combiner box, and an inverter by fault accurate positioning based on an I-V diagnosis algorithm of big data. The system comprises a data acquisition module, a screening and processing module, a fault type mismatching judgment module, a fault diagnosis analysis module and a fault diagnosis result comparison and verification module. The system module architecture and flow chart are shown in fig. 2. Wherein:

the data acquisition, screening and processing module is used for data acquisition, data conformity judgment and data smoothing processing. In order to avoid misjudgment in the subsequent diagnosis process, the I-V data conformity needs to be judged. In the normal case of component I-V data (U) ₀ ,I ₀ ) And judging the current at the theoretical open-circuit voltage point corresponding to the open-circuit voltage point to judge whether the I-V data is missing. If open circuit current I ₀ If the value is less than or equal to 0.05A, the I-V curve meets the requirement, otherwise, the data of the I-V curve is abnormal. According to the judgment, for the condition that the open-circuit voltage lacks in I-V data, the group of data is not used for fault diagnosis any more, the I-V data is not matched, and the output data is abnormal. The flow chart of the whole I-V data conformity judgment algorithm is shown in the attached figure 3. The set of data of the I-V curve judged abnormal is no longer used for the fault diagnosis analysis.

Due to the problem of data acquisition, the obtained I-V data may have large jitter, which affects subsequent fault diagnosis, and in order to eliminate the jitter of the I-V data and improve the dataAnd quality, smoothing the current data by using a smoothing algorithm. Voltage, current data (U) of photovoltaic module _i ,I _i ) Wherein I =0,1,2,3 …,31, the corresponding voltage data is kept unchanged when the current data I _i When the following conditions are satisfied:

or is provided with

The current data is smoothed as follows:

I _i ＝(I _i-1 +I _i+1 )/2

the above process is circulated for 30 times to obtain smoothed data, and the data are arranged according to the sequence of voltage from large to small (U) _n ,I _n ) Where n =0,1,2,3 …,31. The flow chart of the I-V data smoothing algorithm of the above components is shown in the attached FIG. 4.

The fault type mismatch judging module is used for judging the fault type mismatch, and the principle is that when the battery units in the photovoltaic module work in a reverse bias state in a mismatched mode, bypass diodes of substrings where the battery units are located are conducted, so that inflection points and step characteristics can appear on an I-V curve of the mismatched module, and the concavity and convexity of the I-V curve are changed. Judging whether the assembly is mismatched according to the unevenness of the assembly, detecting the inflection point of the I-V curve, and according to the direction from low to high of the I-V curve, setting the point of the unevenness change as a lower inflection point, namely a low voltage point, and the point of the stage as an upper inflection point, namely a high voltage point. And the voltage interval of the stage can be determined through inflection point detection of the I-V curve, and the fault characteristics of shadow, hot spot and glass breakage are further decoupled according to data of the stage of the I-V curve.

The fault diagnosis analysis module is used for mismatch fault diagnosis and non-mismatch fault diagnosis.

The mismatch fault diagnosis mainly comprises a mismatch fault diagnosis method based on the concavity and the convexity of an I-V curve and a mismatch fault diagnosis method based on the division of an I-V curve interval and a characteristic point. Taking the former as an example, the scheme adopts an I-V curve fixed-line variable-step-size detection method. According to the characteristics of the I-V curve, a method for constructing a straight line can be adopted to detect inflection points and step intervals, and the inflection point detection method is shown in figure 5.

By short-circuit current point (U) on the I-V curve ₃₁ ,I ₃₁ ) And open circuit voltage point (U) ₀ ,I ₀ ) The slope constructs a detection straight line, a plurality of groups of balanced straight line clusters are obtained, and the slope and the expression of the straight line equation are as follows:

I _N ＝kU _N +b _I

wherein b is _I For detecting the step length of the change of the line, also the intersection of the line and the current axis, b _I Is the short-circuit current I ₃₁ To ensure that the detection line envelopes the entire I-V curve, so theoretically b _I Does not exceed twice the short-circuit current value, i.e. 2I ₃₁ Because the I-V curve is completely below the detection straight line when the straight line passes through the virtual maximum power point, i.e. I _N <k*U _N +b _I ；

The two points adjacent to the point are also positioned below the straight line, i.e. I _N-1 <k*U _N-1 +b _I ，I _N+1 <k*U _N+1 +b _I ，I _N-1 >0；

The other two points adjacent to the point are located above the straight line, I _N-4 <k*U _N-4 +b _I ，I _N+4 <k*U _N+4 +b _I ，I _N-4 >0；

The I-V curve has a lower inflection point, the comparison is stopped, the photovoltaic module has mismatch, and the coordinates (U) of the points meeting the condition are recorded _n ，I _n ) Namely the voltage and the current at the lower inflection point; if not, change b _I A value of (b) _I ＝b _I + Step _ b, comparing with the next straight line, repeating the above judgment until straightTo b _I If the upper limit value of (b) still does not satisfy the above condition, then there is no mismatch. Determining the lower inflection point of the step section according to the above method, and determining the lower inflection point (U) according to the second determined straight line _n ，I _n ) And open circuit voltage point (U) ₀ ，I ₀ ) Forming a detection straight line, and calculating the I-V data in the range from the open-circuit voltage to the upper inflection point to the maximum value of the detection straight line, wherein the maximum value corresponds to the coordinate (U) _m ，I _m ) That is, the voltage and current values corresponding to the end points of the step sections, and the step interval is (U) _n ，U _m ). The equation of the step end point detection straight line and the distance expression from the data point on the I-V curve to the detection straight line are as follows:

I＝kU+b

the straight line can be detected through the fixed oblique line variable step length at the open-circuit voltage and the short-circuit point, whether mismatch occurs or not can be judged, and the coordinate of a lower inflection point at the position is detected under the condition of mismatch. And further constructing a fixed straight line detection platform stage end point by using a lower inflection point and an open circuit voltage point, and determining a stage end point according to the maximum value of the distance from the I-V data in the range from the lower inflection point to the open circuit voltage point to the detected straight line.

The non-mismatch fault diagnosis mainly comprises an optimal component comparison method and a key parameter discrimination method. Taking a key parameter discrimination method as an example, according to the fault characteristics, the concavity and the convexity of the I-V curve of the non-mismatch fault component are obtained without changing, and only the key parameters are changed. Therefore, for the fault, the fault type can be judged through the combination of the key parameters. The specific diagnostic algorithm flow chart is shown in fig. 6.

In a power station application scene, the volt-ampere characteristic of a photovoltaic string is similar to that of a classical single-diode model, and the relationship between current and voltage can be represented by the following classical formula:

in the above formula, the performance parameters of each key point are explained as follows:

i: working current of the photovoltaic string; i is _L : generating current by the photovoltaic string;

io: diode reverse saturation current; n: a diode ideality factor;

v: operating voltage of the photovoltaic string; rs is a series resistor of the photovoltaic group string;

q: single electron electric quantity; k: boltzmann constant;

t: thermodynamic temperature

Iteration and evolution of intelligent IV diagnosis functional characteristics are realized by taking the data of a massive photovoltaic string in a photovoltaic power station as a basis, combining a classical diode model, deeply learning possible failure modes of the photovoltaic string in a power station application scene, establishing corresponding failure recognition and diagnosis models aiming at different failure modes of the photovoltaic string and realizing iterative upgrade of the failure recognition models; and the inverter reports the acquired IV curve to an IV fault identification algorithm module of the management system, and the IV fault identification algorithm module judges whether the string has faults or not according to the current fault identification and diagnosis model. For mismatching fault diagnosis, the system adopts a mismatching fault diagnosis method based on the concavity and the convexity of an I-V curve and a mismatching fault diagnosis method based on the division of an I-V curve interval and a characteristic point to model, simulate and analyze, and the two methods are synchronously operated by algorithms to obtain an analysis result. For non-mismatch fault diagnosis, the system adopts a key parameter discrimination method and an optimal component comparison method to model, simulate and analyze, and the algorithms of the two methods are synchronously carried out to obtain an analysis result.

The fault diagnosis result comparison and verification module mainly compares and verifies the fault diagnosis analysis results obtained by the two methods of mismatch fault diagnosis or the two methods of non-mismatch fault diagnosis. If the comparison result is consistent or smaller than a certain reasonable error interval, the system outputs a fault diagnosis result; and if the results are different in comparison, feeding back the difference of the analysis results, returning to the initial process of the system for recalculating and analyzing until an accurate optimal analysis result is obtained.

The intelligent I-V diagnosis system of the photovoltaic power station has the following technical effects:

1) The diagnosis system can realize real-time diagnosis of the photovoltaic power station string equipment, support fault detail analysis of the alarm equipment, judge fault types and give fault reason analysis and processing suggestions.

2) The diagnosis system can analyze and diagnose the power generation abnormity of the inverter, the header box and the photovoltaic branch in real time and quickly screen out low-efficiency operation equipment, so that an auxiliary manager can perform routing inspection and defect elimination work in a planned way, and the routing inspection and defect elimination work efficiency is improved.

3) The fault diagnosis analysis module of the diagnosis system establishes a fault expert model base, and continuously corrects the database through model training; the fault diagnosis and analysis module can realize the full detection and fault identification of the photovoltaic power station string and the inverter, automatically generate an IV diagnosis operation and maintenance report after the completion, and provide different operation and maintenance suggestions aiming at different fault types.

4) The diagnosis system effectively improves the reliability and authenticity of data through the data acquisition, screening and processing module, and ensures the accuracy and reliability of the fault diagnosis result through the fault diagnosis result comparison and verification module.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, it should be noted that, for those skilled in the art, many modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (1)

1. An intelligent I-V diagnosis system of a photovoltaic power station is characterized by comprising a data acquisition, screening and processing module, a fault type mismatching judgment module, a fault diagnosis analysis module and a fault diagnosis result comparison and verification module;

the data acquisition, screening and processing module is used for data acquisition, data conformity judgment and data smoothing processing; the data conformity judgment comprises the following steps: in the normal case of component I-V data (U) ₀ ,I ₀ ) Corresponding to the open circuit voltage point, judging the current at the theoretical open circuit voltage point to judge whether the I-V data is missing, if so, judging whether the I-V data is missing ₀ If the value is less than or equal to 0.05A, the I-V curve meets the requirement, otherwise, the data of the I-V curve is abnormal; for the condition that the open-circuit voltage lacks I-V data, the group of data is no longer used for fault diagnosis, the I-V data is not matched, and the output data is abnormal; the data smoothing process includes: voltage, current data (U) for photovoltaic modules _i ,I _i ) Wherein, I =0,1,2,3 …,31, the corresponding voltage data is kept unchanged when the current data I _i When the following conditions are satisfied:

or is provided with

The current data is smoothed as follows:

I _i ＝(I _i-1 +I _i+1 )/2

the above process is circulated for 30 times to obtain smoothed data, and the data are arranged according to the sequence of voltage from large to small (U) _n ,I _n ) Where n =0,1,2,3 …,31;

the fault type mismatch judging module is used for judging whether the assembly is mismatched or not according to the concavity and convexity of the I-V curve of the assembly, according to the direction from low to high of the I-V curve, a point of concavity and convexity change is specified to be a lower inflection point, namely a low voltage point, and a point of a stage is specified to be an upper inflection point, namely a high voltage point, through inflection point detection of the I-V curve, a voltage interval of the stage is determined, and fault characteristics of shadow, hot spot and glass breakage are further decoupled according to data of the stage of the I-V curve;

the fault diagnosis and analysis module is used for mismatch fault diagnosis and non-mismatch fault diagnosis;

the mismatch fault diagnosis comprises a mismatch fault diagnosis method based on I-V curve concavity and convexity and a mismatch fault diagnosis method based on I-V curve interval division and characteristic points;

the mismatch fault diagnosis method based on the I-V curve concavity and convexity adopts an I-V curve fixed line variable step length detection method, which comprises the following steps:

by short-circuit current point (U) on the I-V curve ₃₁ ,I ₃₁ ) And open circuit voltage point (U) ₀ ,I ₀ ) The slope constructs a detection straight line, a plurality of groups of balanced straight line clusters are obtained, and the slope and the expression of the straight line equation are as follows:

I _N ＝kU _N +b _I

wherein b is _I For detecting the step length of the change of the line, also the intersection of the line and the current axis, b _I Is the short-circuit current I ₃₁ ，b _I Does not exceed twice the short-circuit current value, i.e. 2I ₃₁ When the straight line passes through the virtual maximum power point, the I-V curve is completely below the detection straight line, i.e. I _N <k*U _N +b _I ；

The two points adjacent to the point are also positioned below the straight line, i.e. I _N-1 <k*U _N-1 +b _I ，I _N+1 <k*U _N+1 +b _I ，I _N-1 >0；

The other two points adjacent to the point are located above the straight line, I _N-4 <k*U _N-4 +b _I ，I _N+4 <k*U _N+4 +b _I ，I _N-4 >0；

The I-V curve has a lower inflection point, the comparison is stopped, the photovoltaic module has mismatch, and the record meets the requirementCoordinates (U) of points of the condition _n ，I _n ) Namely the voltage and the current at the lower inflection point; if not, change b _I A value of (b) _I ＝b _I + Step _ b, compare with next straight line, repeat the above judgment until b _I If the upper limit value of (c) still does not satisfy the above condition, then no mismatch exists; determining the lower inflection point of the stage according to the above method, and determining the lower inflection point (U) according to the second determined line _n ，I _n ) And open circuit voltage point (U) ₀ ，I ₀ ) Calculating the maximum value of the I-V data from the open-circuit voltage to the upper inflection point, and the coordinate (U) corresponding to the maximum value _m ，I _m ) That is, the voltage and current values corresponding to the end points of the step sections, and the step interval is (U) _n ，U _m ) (ii) a The equation of the step end point detection straight line and the distance expression from the data point on the I-V curve to the detection straight line are as follows:

I＝kU+b

detecting the coordinate of a lower inflection point under the condition of mismatch, constructing a stage end point of a fixed linear detection platform by using the lower inflection point and an open-circuit voltage point, and determining the stage end point according to the maximum value of the distance from I-V data in the range from the lower inflection point to the open-circuit voltage point to a detected linear distance;

the non-mismatch fault diagnosis comprises an optimal component comparison method and a key parameter discrimination method;

the key parameter discrimination method comprises the following steps: judging the fault type through the combination of the key parameters;

for a power station application scene, the volt-ampere characteristic of a photovoltaic string is similar to that of a classical single-diode model, and the relationship between current and voltage is represented by the following classical formula:

in the formula, I: working current of the photovoltaic string; i is _L : generating current by the photovoltaic string; io: diode reverse saturation current; n: a diode ideality factor; v: operating voltage of the photovoltaic string; rs is a series resistor of the photovoltaic group string; q: single electron electric quantity; k: boltzmann constant; t: a thermodynamic temperature;

for mismatch fault diagnosis, the fault diagnosis and analysis module adopts a mismatch fault diagnosis method based on I-V curve concavity and convexity and a mismatch fault diagnosis method based on I-V curve interval division and characteristic points to model, simulate and analyze, and the two methods are synchronously carried out to obtain an analysis result; for non-mismatch fault diagnosis, the fault diagnosis analysis module adopts a key parameter discrimination method and an optimal component comparison method to model, simulate and analyze, and synchronously carries out the two methods to obtain an analysis result;

the fault diagnosis result comparison and verification module is used for comparing and verifying fault diagnosis analysis results obtained by adopting two methods of mismatch fault diagnosis or two methods of non-mismatch fault diagnosis, and outputting a fault diagnosis result if the result comparison is consistent or smaller than a reasonable error interval; and if the result comparison is different, feeding back the analysis result to have difference, and recalculating and analyzing until an accurate optimal analysis result is obtained.

Images