CN109409420B - Photovoltaic string fault diagnosis method under non-uniform irradiance - Google Patents

Photovoltaic string fault diagnosis method under non-uniform irradiance Download PDF

Info

Publication number
CN109409420B
CN109409420B CN201811168547.7A CN201811168547A CN109409420B CN 109409420 B CN109409420 B CN 109409420B CN 201811168547 A CN201811168547 A CN 201811168547A CN 109409420 B CN109409420 B CN 109409420B
Authority
CN
China
Prior art keywords
photovoltaic
fault
string
fireworks
circuit
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811168547.7A
Other languages
Chinese (zh)
Other versions
CN109409420A (en
Inventor
王靖程
陈仓
姚玲玲
牛瑞杰
敖海
王建峰
许国泽
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huaneng Jinchang Photovoltaic Power Generation Co ltd
Xian Thermal Power Research Institute Co Ltd
Huaneng Group Technology Innovation Center Co Ltd
Original Assignee
Huaneng Jinchang Photovoltaic Power Generation Co ltd
Xian Thermal Power Research Institute Co Ltd
Huaneng Group Technology Innovation Center Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huaneng Jinchang Photovoltaic Power Generation Co ltd, Xian Thermal Power Research Institute Co Ltd, Huaneng Group Technology Innovation Center Co Ltd filed Critical Huaneng Jinchang Photovoltaic Power Generation Co ltd
Priority to CN201811168547.7A priority Critical patent/CN109409420B/en
Publication of CN109409420A publication Critical patent/CN109409420A/en
Application granted granted Critical
Publication of CN109409420B publication Critical patent/CN109409420B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Photovoltaic Devices (AREA)
  • Testing Of Individual Semiconductor Devices (AREA)

Abstract

The invention relates to a photovoltaic string fault diagnosis method under non-uniform irradiance, which is characterized in that under the condition that irradiance in a photovoltaic string is inconsistent due to variable weather, environmental parameters of photovoltaic components at different positions are measured through irradiance sensors and temperature sensors which are arranged at multiple points, and photovoltaic current, voltage and output power are theoretically estimated by utilizing a photovoltaic cell equivalent circuit model. At the same time, the actual photovoltaic current, voltage and output power are measured at the combiner box of the photovoltaic string. And then carrying out fault coding on the open-circuit fault and the short-circuit fault to obtain a theoretical output power function related to the fault coding. And finally, optimizing a target function by using an enhanced firework algorithm with rapid mining and exploration performances by taking the absolute value of the difference between the theoretical output power and the actual output power of the photovoltaic string as an optimization target, so as to realize fault diagnosis and positioning of the photovoltaic string. The invention can enable the staff for the operation and maintenance of the photovoltaic power station to arrange and overhaul the failed photovoltaic module in time.

Description

Photovoltaic string fault diagnosis method under non-uniform irradiance
Technical Field
The invention relates to a photovoltaic string fault diagnosis method, in particular to a photovoltaic string fault diagnosis method under non-uniform irradiance.
Background
With the increasing exhaustion of traditional energy sources such as fossil fuels and the more prominent problem of environmental pollution, the demand of human beings for new energy sources is more urgent. The solar energy has the advantages of cleanness, environmental protection, safety, reliability, sustainable development and the like, and is an ideal renewable new energy source. At present, photovoltaic power generation is an effective technology for converting solar energy into electric energy, and can effectively relieve energy crisis and improve the problem of environmental pollution. The photovoltaic array is taken as an important component of a photovoltaic power generation system, the average service life of the photovoltaic array can reach 20-30 years theoretically, but in practical engineering application, a photovoltaic module can suffer from the problems of packaging technology, severe use environment, hot spot phenomenon, mechanical damage, circuit faults (open circuit and short circuit) and the like, so that the normal service life of the photovoltaic module is greatly shortened. The Short Circuit (SC) fault and the Open Circuit (OC) fault of the photovoltaic string have serious influence on the normal operation of photovoltaic power generation, and even threaten the safety of the whole photovoltaic power station. Therefore, it is necessary to monitor the operation state of the photovoltaic array in real time, find faults in time and locate the faults.
In a traditional photovoltaic array fault detection mode, offline detection and online detection are available. The off-line detection can be implemented only after the whole photovoltaic power generation system stops running, but the photovoltaic power station needs to be continuously operated to ensure the stability and continuity of power supply. Therefore, offline fault detection is rarely used on a large scale in practice. The online detection generally comprises a photovoltaic array fault detection method based on a mathematical model and an intelligent photovoltaic array fault diagnosis system, the methods are complex in mathematical model, large in calculation amount and poor in real-time performance, and under the influence of variable environmental factors, the defects of low fault diagnosis accuracy and high fault false alarm rate exist, so that the working intensity of field operation and maintenance personnel is increased, even serious fault false alarm situations occur, and the benefit and the safety stability of a photovoltaic power generation system are directly influenced.
In the practical engineering, how to efficiently and timely analyze the working state of the photovoltaic string, and quickly determine the type and the position of the photovoltaic fault is an urgent problem to be solved, and is also an important content of the operation and maintenance of the photovoltaic power station in the future.
Disclosure of Invention
In order to overcome the defects of the existing fault diagnosis technology, the invention aims to provide a photovoltaic string fault diagnosis method under non-uniform irradiance based on a reinforced firework algorithm.
The invention is realized by adopting the following technical scheme:
a photovoltaic string fault diagnosis method under non-uniform irradiance comprises the following steps:
step 1: aiming at the photovoltaic string with N assemblies, collecting the output power P of the photovoltaic stringMeasuredVoltage VMeasuredAnd current IMeasuredAnd irradiance G of component i in photovoltaic stringiAnd the temperature T of the assemblyi；
Step 2: irradiance G according to photovoltaic equivalent circuit model and collectioniAnd the temperature T of the assemblyiCalculating to obtain the theoretical maximum output power of the component i in the photovoltaic group string;
and step 3: calculating theoretical output power P of photovoltaic stringArray(ii) a Namely, the output power of the photovoltaic group string is equal to the sum of the output power of the photovoltaic components in the photovoltaic group string;
and 4, step 4: determining a fault threshold epsilon and judging whether the photovoltaic string is abnormal or not; if P isMeasured+ε＜PArrayIf so, indicating that the string operation is abnormal, and performing step 5; otherwise, the group string is not abnormal;
and 5: establishing a fault equivalent model; aiming at short-circuit faults and open-circuit faults in the photovoltaic power generation system, the short-circuit faults and the open-circuit faults are similar to a switch model, when the short-circuit faults occur, the short-circuit faults are equivalent to short-circuit of diodes connected in parallel with a photovoltaic panel, namely the short-circuit faults are equivalent to the fact that the photovoltaic panel is connected in parallel with a closed switch K1; when the photovoltaic panel is in an open circuit, the photovoltaic panel is equivalently connected with a disconnected switch K2 in series, and an external output path of the photovoltaic panel is cut off; the normal operation condition is equivalent to that the switch K1 is opened and the switch K2 is closed;
step 6: constructing a fitness function for representing the fitness of each firework, wherein the value of the fitness determines the number of sparks emitted by the fireworks and the explosion radius;
Fitness＝f(NSC,NOC,SC,OC)＝|PMeasured-PArray|
then the goal of optimization with the enhanced firework algorithm is:
minf(NSC,NOC,SC,OC)＝|PMeasured-PArray|
Figure BDA0001821796140000031
wherein N isSCNumber of short-circuit faults, NOCThe number of open circuit faults;
and 7: initializing a firework population, randomly selecting M' fireworks in a feasible region space as a preliminary iteration firework population, and initializing a maximum iteration number iter;
and 8: calculating Fitness of Firework i in Firework populationi(ii) a Then carrying out normalization treatment to obtain
Figure BDA0001821796140000032
Will remain in the population
Figure BDA0001821796140000033
Performing the next operation on the M groups;
and step 9: designing explosion operator, and setting (N) according to each positionSC,NOC,SC,OC)iCalculating the individual Fitness of the fireworks, and calculating the initial explosion radius R of the fireworks, wherein rho is an expansion coefficient of the explosion radius, the coefficient is determined by the specific size range of a feasible domain, and the formula shows that the smaller the Fitness is the better fireworks, the explosion range of the fireworks is reduced, and the detailed search is carried out; the fireworks with large adaptability have large explosion radius and are exploited in a wide range; calculating the number of explosion sparks according to the individual fitness of the fireworks
Figure BDA0001821796140000034
Figure BDA0001821796140000035
Wherein theta is an expansion coefficient of the number of explosion sparks, the coefficient is determined by the size of the firework population, and meanwhile, a normal random number with the mean value of 0 is added; so that each spark has a radius of Ri＝R+N(0,σR) The random number is added to increase the diversity of the spark while producing a uniformly distributed generated directional angle thetaiU (0, 2); the position at which spark i is obtained is:
(NSC,NOC,SC,OC)′i＝(NSC,NOC,SC,OC)i
＋(NSC,NOC,(Ricosθi)bin,(Risinθi)bin)i
wherein (R)icosθi)binDenotes a real number RicosθiConverting into an N-bit binary form, and extracting bit numbers of '1' to form a set SC; same principle (R)isinθi)binTo make a real number RisinθiConverting the binary form into an N-bit binary form, and extracting bit numbers of '1' to form a set OC;
step 10: designing a Gaussian mutation operator, and carrying out Gaussian mutation operation on sparks emitted by fireworks in order to increase the longitudinal diversity of the sparks, namely selecting n' < n sparks as a mutation population to carry out Gaussian mutation
Ri＝N(R,σR) (ii) a Wherein here it isi in the population of spark variants; whereby the location of the altered spark is
(NSC,NOC,SC,OC)′i＝N(R,σR)×(NSC,NOC,SC,OC)i；
Step 11: designing a spark collision operator to strengthen the diversity of sparks in the firework explosion process, namely, different spark collisions generate new sparks in the firework explosion process; i.e. parent spark (N)SC,NOC,SC,OC)1And (N)SC,NOC,SC,OC)2Collision generating daughter sparks (N)SC,NOC，SC,OC)′1And (N)SC,NOC,SC,OC)′2(ii) a Firstly, a random number n between 0 and 1 is randomly generated, and then two random numbers gamma are generated by utilizing polynomial probability distribution1And gamma2，
Figure BDA0001821796140000041
Wherein
Figure BDA0001821796140000043
Figure BDA0001821796140000042
ηcIs a non-negative number;
the children spark is thus:
(NSC,NOC,SC,OC)p1
＝0.5[((NSC,NOC,SC,OC)2+(NSC,NOC,SC,OC)1)
-γ1(|(NSC,NOC,SC,OC)2-(NSC,NOC,SC,OC)1|)]
(NSC,NOC,SC,OC)p2
＝0.5[((NSC,NOC,SC,OC)2+(NSC,NOC,SC,OC)1)
-γ2(|(NSC,NOC,SC,OC)2-(NSC,NOC,SC,OC)1|)]
step 12: selecting the next generation of sparks as the fireworks for the next iteration;
Figure BDA0001821796140000051
step 13: judging a termination condition; if f ((N)SC,NOC,SC,OC)k) Less than or equal to epsilon, obtaining VArrayAnd IArrayIf V isMeasure＝VArrayAnd IMeasure＝IArrayIf so, terminating, giving fault codes and decoding, namely, under the condition that the same power, current and voltage are generated under the non-uniform irradiance, indicating that the fault type calculated by the model is consistent with the actual condition, namely, diagnosing the fault as an open-circuit fault or a short-circuit fault; otherwise, carrying out the next iteration, and turning to the step 8; if the iteration exceeds the maximum iteration number limit iter, the iteration is terminated, and the fault type is a non-open-circuit fault or a short-circuit fault, namely the high-loss abnormity is determined.
The photovoltaic string fault method is further improved in that the Si-01TC-T irradiance sensor and the photovoltaic monitoring system are adopted to collect relevant data, binary equivalent coding is carried out on open-circuit faults and short-circuit faults, and optimization solution is carried out by using an enhanced firework algorithm.
The invention is further improved in that, in step 2, the calculation formula is as follows:
Figure BDA0001821796140000061
Figure BDA0001821796140000062
Figure BDA0001821796140000063
wherein the content of the first and second substances,
Figure BDA0001821796140000064
is the output current of the component i,
Figure BDA0001821796140000065
is the output power of the component i,
Figure BDA0001821796140000066
the number of modules connected in parallel in the component i,
Figure BDA0001821796140000067
the number of modules connected in series in the component i,
Figure BDA0001821796140000068
for the generation of the photo-generated current for the modules in the module i,
Figure BDA0001821796140000069
for the reverse saturation current of the module parallel diode in component i,
Figure BDA00018217961400000610
is the equivalent series resistance of the component i,
Figure BDA00018217961400000611
is the equivalent parallel resistance of component i;
Figure BDA00018217961400000612
ambient temperature at component i and NOCT is the nominal operating temperature of the battery.
The invention is further improved in that, in step 3, the calculation formula is as follows:
Figure BDA00018217961400000613
wherein, Pi(Ti,Gi) Represents the ith component theoretical output maximum power, which is a function of irradiance and temperature; SC represents the set of short-circuit faults in the string, and OC represents the set of open-circuit faults;
Figure BDA00018217961400000614
the losses due to the parallel diodes of the open component i,
Figure BDA00018217961400000615
Figure BDA00018217961400000616
is the branch current of the string of the group in which component i is located,
Figure BDA00018217961400000617
the parallel diode of component i conducts the voltage.
The invention has the following beneficial technical effects:
the method comprises the steps of measuring and analyzing electrical parameters and environmental parameters of actual current, voltage, power, photovoltaic panel temperature and irradiance of a photovoltaic array, establishing a fault equivalent model, and performing photovoltaic string fault diagnosis by adopting an optimized model based on a reinforced firework algorithm. The method can be used for diagnosing the short-circuit fault and the open-circuit fault of the photovoltaic string, can be used for positioning and distinguishing the short-circuit fault and the open-circuit fault under the non-uniform irradiance, and has the advantages of simple model, good robustness, economy, practicality and good real-time performance; the photovoltaic power generation system can guarantee the optimal operation state on the premise of safety and stability, provide reliable fault early warning for operation and maintenance personnel, and reduce the situations of missing report and false report.
In conclusion, the photovoltaic power plant fault detection method and the photovoltaic power plant fault detection system can enable workers of photovoltaic power plant operation maintenance to arrange and overhaul the faulted photovoltaic modules in time, so that the photovoltaic modules can operate in a continuous optimal state, and the economic benefit of the photovoltaic power plant is guaranteed.
Detailed Description
The present invention is described in more detail below with reference to the diagnosis of faults in photovoltaic strings under non-uniform irradiance as an example.
The invention provides a photovoltaic string fault diagnosis method under non-uniform irradiance, which comprises the following steps:
step 1: aiming at a photovoltaic string with 10 assemblies in total, collecting the output power P of the photovoltaic stringMeasured1070.50W, voltage VMeasured155.20V and current IMeasured6.90A, and irradiance G of component i in photovoltaic stringiAnd the temperature T of the assemblyi. As in table 1.
TABLE 1 environmental parameters
Figure BDA0001821796140000071
Step 2: and calculating the theoretical output maximum power of the ith photovoltaic assembly according to the photovoltaic equivalent circuit model and the collected irradiance and temperature.
Figure BDA0001821796140000081
Figure BDA0001821796140000082
Figure BDA0001821796140000083
Wherein the content of the first and second substances,
Figure BDA0001821796140000084
is the output current of the component i,
Figure BDA0001821796140000085
is the output power of the component i,
Figure BDA0001821796140000086
the number of modules connected in parallel in the component i,
Figure BDA0001821796140000087
the number of modules connected in series in the component i,
Figure BDA0001821796140000088
for the generation of the photo-generated current for the modules in the module i,
Figure BDA0001821796140000089
for the reverse saturation current of the module parallel diode in component i,
Figure BDA00018217961400000810
is the equivalent series resistance of the component i,
Figure BDA00018217961400000811
is the equivalent parallel resistance of component i.
Figure BDA00018217961400000812
Ambient temperature at component i and NOCT is the nominal operating temperature of the battery.
Here, the NOCT is 48 deg.c,
Figure BDA00018217961400000813
and step 3: calculating theoretical output power P of photovoltaic stringArray. That is, the output power of the photovoltaic string is equal to the sum of the output powers of the photovoltaic modules in the photovoltaic string.
Figure BDA00018217961400000814
Wherein, Pi(Ti,Gi) Represents the theoretical maximum output power of the ith component as a function of irradiance and temperature. SC represents a set of short faults in the string of groups and OC represents a set of open faults.
Figure BDA00018217961400000815
The losses due to the parallel diodes of the open component i,
Figure BDA00018217961400000816
Figure BDA00018217961400000817
is the branch current of the string of the group in which component i is located,
Figure BDA00018217961400000818
the parallel diode of component i conducts the voltage.
And 4, step 4: determining the fault threshold ε 0.1% PMeasuredAnd judging whether the photovoltaic string is abnormal or not. If P isMeasured+ε＜PArrayIf so, indicating that the string operation is abnormal, and performing step 5; otherwise, the group string is not abnormal.
And 5: and establishing a fault equivalent model. Aiming at short-circuit faults and open-circuit faults in the photovoltaic power generation system, the short-circuit faults and the open-circuit faults can be approximated to be a switch model, and when the short-circuit faults occur, the short-circuit faults are equivalent to that the photovoltaic panel parallel diodes are short-circuited, namely that the photovoltaic panel is equivalent to that the photovoltaic panel is connected with a closed switch K1 in parallel; when the photovoltaic panel is opened, the photovoltaic panel is equivalently connected with a disconnected switch K2 in series, and an external output path of the photovoltaic panel is cut off. In the normal operation, the switch K1 is opened and the switch K2 is closed.
Step 6: and constructing a fitness function for representing the fitness of each firework, wherein the value of the fitness determines the number of sparks emitted by the fireworks and the explosion radius.
Fitness＝f(NSC,NOC,SC,OC)＝|PMeasured-PArray|
Then the optimized objective function based on the enhanced firework algorithm is:
minf(NSC,NOC,SC,OC)＝|PMeasured-PArray|
Figure BDA0001821796140000091
wherein N isSCNumber of short-circuit faults, NOCThe number of open circuit faults.
And 7: initializing a firework population, randomly selecting M' to 20 fireworks in a feasible region space as a preliminary iteration firework population, and initializing a maximum iteration number iter to 50.
And 8: calculating Fitness of Firework i in Firework populationi. Then carrying out normalization treatment to obtain
Figure BDA0001821796140000092
Will remain in the population
Figure BDA0001821796140000093
M populations (λ ═ 0.6) were taken for the next step.
And step 9: designing explosion operator, and setting (N) according to each positionSC,NOC,SC,OC)iCalculating the individual adaptability of the fireworks, calculating the initial explosion radius R of the fireworks, wherein rho is 5 and is an expansion coefficient of the explosion radius, the coefficient is determined by the specific size range of a feasible region, and the smaller the adaptability is, the better the fireworks is, the explosion range is reduced, and the detailed search is carried out; and fireworks with large adaptability have large explosion radius and are exploited in a wide range. Calculating the number of explosion sparks according to the individual fitness of the fireworks
Figure BDA0001821796140000101
Take sigmanAnd 30, wherein theta is 8, the expansion coefficient of the explosion spark number is determined by the size of the firework population, and a normal random number with the mean value of 0 is added. So that each spark has a radius of Ri＝R+N(0,σR) Taking σRThe random number is increased 12 to increase the spark diversity and to produce a uniformly distributed generated directional angle θiU (0,2 pi). Thus, the position of the spark i can be found as follows:
(NSC,NOC,SC,OC)ii＝(NSC,NOC,SC,OC)i
+(NSC,NOC,(Ricosθi)bin,(Risinθi)bin)i
wherein (R)icosθi)binDenotes a real number RicosθiConverting into an N-bit binary form, and extracting bit numbers of '1' to form a set SC; same principle (R)isinθi)binTo make a real number RisinθiConverted into an N-bit binary form, and the bit numbers extracted as '1' form a set OC.
Step 10: designing a Gaussian mutation operator, and carrying out Gaussian mutation operation on sparks emitted by fireworks in order to increase the longitudinal diversity of the sparks, namely selecting n' < n sparks as a mutation population to carry out Gaussian mutation, namely Ri＝N(R,σR). Where the ith is in the spark variant population. The varied spark positions are thus:
(NSC,NOC,SC,OC)′i＝N(R,σR)×(NSC,NOC,SC,OC)i。
step 11: and designing a spark collision operator to strengthen the diversity of sparks in the firework explosion process, namely, different spark collisions can generate new sparks in the firework explosion process. I.e. parent spark (N)SC,NOC,SC,OC)1And (N)SC,NOC,SC,OC)2Collision generating daughter sparks (N)SC,NOC,SC,OC)′1And (N)SC,NOC,SC,OC)′2。
Firstly, a random number n between 0 and 1 is randomly generated, and then two random numbers gamma are generated by utilizing polynomial probability distribution1And gamma2，
Figure BDA0001821796140000111
Wherein
Figure BDA0001821796140000114
Figure BDA0001821796140000112
ηcIs a non-negative number, and takes 0.8.
The children spark is thus:
(NSC,NOC,SC,OC)p1
＝0.5[((NSC,NOC,SC,OC)2+(NSC,NOC,SC,OC)1)
-γ1(|(NSC,NOC,SC,OC)2-(NSC,NOC,SC,OC)1|）]
(NSC,NOC，SC,OC)p2
＝0.5[((NSC,NOC,SC,OC)2+(NSC,NOC,SC,OC)1)
-γ2(|(NSC,NOC,SC,OC)2-(NSC,NOC,SC,OC)1|)]
step 12: the next generation of sparks is selected as the fireworks for the next iteration.
Figure BDA0001821796140000113
Step 13: and judging a termination condition. If f ((N)SC,NOC,SC,OC)k) Less than or equal to epsilon, obtaining VArrayAnd IArrayIf V isMeasure＝VArrayAnd IMeasure＝IArrayAnd ending, giving fault codes and decoding, namely, under the condition that the same power, current and voltage are generated under the condition of non-uniform irradiance, indicating that the fault type calculated by the model is consistent with the actual condition, namely, diagnosing the fault as an open-circuit fault or a short-circuit fault. Otherwise, the next iteration is carried out, and step 8 is carried out. If the iteration exceeds the maximum iteration number limit iter, the iteration is terminated, and the fault type is a non-open-circuit fault or a short-circuit fault, namely the high-loss abnormity is determined.
The photovoltaic string fault diagnosis is carried out through the process, and 31 times of iteration is carried out, so that the fault information is as follows: open circuit faults occur in the No. 1, No. 8 and No. 9 components; the short circuit failure occurred in the 2 nd and 5 th components. The diagnosis result can be provided for operating and maintaining workers of the photovoltaic power station, and the photovoltaic module with faults is arranged and overhauled in time, so that the photovoltaic module can operate in a continuous optimal state, and the economic benefit of the photovoltaic power station is guaranteed.

Claims (4)

1. A method for diagnosing faults of a photovoltaic string under non-uniform irradiance is characterized by comprising the following steps:
step 1: aiming at the photovoltaic string with N assemblies, collecting the output power P of the photovoltaic stringMeasuredVoltage VMeasuredAnd current IMeasuredAnd irradiance G of component i in photovoltaic stringiAnd the temperature T of the assemblyi；
Step 2: irradiance G according to photovoltaic equivalent circuit model and collectioniAnd the temperature T of the assemblyiCalculating to obtain the theoretical maximum output power of the component i in the photovoltaic group string;
and step 3: calculating theoretical output power P of photovoltaic stringArray(ii) a Namely, the output power of the photovoltaic group string is equal to the sum of the output power of the photovoltaic components in the photovoltaic group string;
and 4, step 4: determining a fault threshold epsilon and judging whether the photovoltaic string is abnormal or not; if P isMeasured+ε＜PArrayIf so, indicating that the string operation is abnormal, and performing step 5; otherwise, the group string is not abnormal;
and 5: establishing a fault equivalent model; aiming at short-circuit faults and open-circuit faults in the photovoltaic power generation system, the short-circuit faults and the open-circuit faults are similar to a switch model, when the short-circuit faults occur, the short-circuit faults are equivalent to short-circuit of diodes connected in parallel with a photovoltaic panel, namely the short-circuit faults are equivalent to the fact that the photovoltaic panel is connected in parallel with a closed switch K1; when the photovoltaic panel is in an open circuit, the photovoltaic panel is equivalently connected with a disconnected switch K2 in series, and an external output path of the photovoltaic panel is cut off; the normal operation condition is equivalent to that the switch K1 is opened and the switch K2 is closed;
step 6: constructing a fitness function for representing the fitness of each firework, wherein the value of the fitness determines the number of sparks emitted by the fireworks and the explosion radius;
Fitness＝f(NSC，NOC，SC，OC)＝|PMeasured-PArray|
then the goal of optimization with the enhanced firework algorithm is:
min f(NSC，NOC，SC，OC)＝|PMeasured-PArray|
Figure FDA0003558409760000021
wherein N isSCNumber of short-circuit faults, NOCFor the number of open-circuit faults, SC represents a set of short-circuit faults in the string group, and OC represents a set of open-circuit faults;
and 7: initializing a firework population, randomly selecting M' fireworks in a feasible region space as a preliminary iteration firework population, and initializing a maximum iteration number iter;
and 8: calculating Fitness of Firework z in Firework populationz(ii) a Then carrying out normalization treatment to obtain
Figure FDA0003558409760000022
Will remain in the population
Figure FDA0003558409760000023
Performing the next operation on the M groups;
and step 9: designing explosion operator, and setting (N) according to each positionSC，NOC，SC，OC)jCalculating the individual Fitness of the fireworks, and calculating the initial explosion radius R of the fireworks, wherein rho is an expansion coefficient of the explosion radius, the coefficient is determined by the specific size range of a feasible domain, and the formula shows that the smaller the Fitness is the better fireworks, the explosion range of the fireworks is reduced, and the detailed search is carried out; the fireworks with large adaptability have large explosion radius and are exploited in a wide range; calculating the number of explosion sparks according to the individual fitness of the fireworks
Figure FDA0003558409760000024
Figure FDA0003558409760000025
Wherein theta is the expansion coefficient of the explosive spark number, the coefficient is determined by the size of the firework population, and a normal random number phi (0, sigma) with the mean value of 0 is addedm) (ii) a So that each spark has a radius of Rj＝R+Φ(0，σR) The random number is added to increase the diversity of the spark while producing a uniformly distributed generated directional angle thetajU (0,2 pi); to this end, the position at which spark j is obtained is:
(NSC，NOC，SC，OC)′j＝(NSC，NOC，SC，OC)j+(NSC，NOC，(Rjcosθj)bin，(Rjsinθj)bin)
wherein (R)jcosθj)binDenotes a real number RjcosθjConverting the binary number into an N-bit binary form, and extracting the serial number of the bit of '1' in the binary number to form a set SC; same principle (R)jsinθj)binTo make a real number RjsinθjConverting the binary system into an N-bit binary system form, wherein the serial numbers of the bits of '1' in the binary system form a set OC; n is the number of the photovoltaic group strings; j represents the spark sequence number of the firework population;
step 10: designing a Gaussian mutation operator, and carrying out Gaussian mutation operation on sparks emitted by fireworks in order to increase the longitudinal diversity of the sparks, namely selecting m' sparks less than m sparks as a mutation population to carry out Gaussian mutation, namely Rj＝Φ(R，σR) (ii) a Where j is in the spark variant population; whereby the location of the altered spark is
(NSC，NOC，SC，OC)′j＝Φ(R，σR)×(NSC，NOC，SC，OC)j；
Step 11: designing a spark collision operator to strengthen the diversity of sparks in the firework explosion process, namely fireworksDuring the explosion, different spark collisions will produce new sparks; i.e. parent spark (N)SC，NOC，SC，OC)1And (N)SC，NOC，SC，OC)2Collision generating daughter sparks (N)SC，NOC，SC，OC)′1And (N)SC，NOC，SC，OC)′2(ii) a Firstly, a random number n between 0 and 1 is randomly generated, and then two random numbers gamma are generated by utilizing polynomial probability distribution1And gamma2，
Figure FDA0003558409760000031
Wherein
Figure FDA0003558409760000032
Figure FDA0003558409760000033
ηcIs a non-negative number;
the children spark is thus:
(NSC，NOC，SC，OC)′1
＝0.5[((NSC，NOC，SC，OC)2+(NSC，NOC，SC，OC)1)-γ1(|(NSC，NOC，SC，OC)2-(NSC，NOC，SC，OC)1|)]
(NSC，NOC，SC，OC)′2
＝0.5[((NSC，NOC，SC，OC)2+(NSC，NOC，SC，OC)1)-γ2(|(NSC，NOC，SC，OC)2-(NSC，NOC，SC，OC)1|)]
step 12: selecting the next generation of sparks as the fireworks for the next iteration;
Figure FDA0003558409760000041
step 13: judging a termination condition; if f ((N)SC，NOC，SC，OC)t) Less than or equal to epsilon, obtaining VArrayAnd IArrayIf V isMeasure＝VArrayAnd IMeasure＝IArrayIf so, terminating, giving fault codes and decoding, namely, under the condition that the same power, current and voltage are generated under the non-uniform irradiance, indicating that the fault type calculated by the model is consistent with the actual condition, namely, diagnosing the fault as an open-circuit fault or a short-circuit fault; otherwise, carrying out the next iteration, and turning to the step 8; if the iteration exceeds the maximum iteration time limit iter, the iteration is terminated, and the fault type is neither short-circuit fault nor open-circuit fault but high loss; wherein (N)SC，NOC，SC，OC)tIs the parent fireworks, t is the number of parent fireworks, (N)SC，NOC，SC，OC)tThe children fireworks obtained for the parent fireworks.
2. The method for diagnosing the faults of the photovoltaic string under the non-uniform irradiance is characterized in that the method for diagnosing the faults of the photovoltaic string adopts a Si-01TC-T irradiance sensor and a photovoltaic monitoring system to collect relevant data, carries out binary equivalent coding on open-circuit faults and short-circuit faults and carries out optimization solution by using an enhanced firework algorithm.
3. The method for diagnosing the fault of the photovoltaic string under the non-uniform irradiance as recited in claim 1 or 2, wherein in the step 2, the calculation formula is as follows:
Figure FDA0003558409760000051
Figure FDA0003558409760000052
Figure FDA0003558409760000053
wherein the content of the first and second substances,
Figure FDA0003558409760000054
is the output current of the component i,
Figure FDA0003558409760000055
is the output power of the component i,
Figure FDA0003558409760000056
the number of modules connected in parallel in the component i,
Figure FDA0003558409760000057
the number of modules connected in series in the component i,
Figure FDA0003558409760000058
for the generation of the photo-generated current for the modules in the module i,
Figure FDA0003558409760000059
for the reverse saturation current of the module parallel diode in component i,
Figure FDA00035584097600000510
is the equivalent series resistance of the component i,
Figure FDA00035584097600000511
is the equivalent parallel resistance of component i;
Figure FDA00035584097600000512
ambient temperature at component i and NOCT is the nominal operating temperature of the battery.
4. The method for diagnosing the fault of the photovoltaic string under the nonuniform irradiance as recited in claim 3, wherein in the step 3, the calculation formula is as follows:
Figure FDA00035584097600000513
wherein, Pi(Ti，Gi) Represents the ith component theoretical output maximum power, which is a function of irradiance and temperature; SC represents the set of short-circuit faults in the string, and OC represents the set of open-circuit faults;
Figure FDA00035584097600000514
the losses due to the parallel diodes of the open component i,
Figure FDA00035584097600000515
Figure FDA00035584097600000516
is the branch current of the string of the group in which component i is located,
Figure FDA00035584097600000517
the parallel diode of component i conducts the voltage.
CN201811168547.7A 2018-10-08 2018-10-08 Photovoltaic string fault diagnosis method under non-uniform irradiance Active CN109409420B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811168547.7A CN109409420B (en) 2018-10-08 2018-10-08 Photovoltaic string fault diagnosis method under non-uniform irradiance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811168547.7A CN109409420B (en) 2018-10-08 2018-10-08 Photovoltaic string fault diagnosis method under non-uniform irradiance

Publications (2)

Publication Number Publication Date
CN109409420A CN109409420A (en) 2019-03-01
CN109409420B true CN109409420B (en) 2022-05-03

Family

ID=65466089

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811168547.7A Active CN109409420B (en) 2018-10-08 2018-10-08 Photovoltaic string fault diagnosis method under non-uniform irradiance

Country Status (1)

Country Link
CN (1) CN109409420B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104767482A (en) * 2014-01-02 2015-07-08 上海岩芯电子科技有限公司 Method for diagnosing aging and short circuit of photovoltaic module online
CN105652952A (en) * 2016-04-18 2016-06-08 中国矿业大学 Maximum power point tracking method for photovoltaic power generation system based on fireworks algorithm
CN106776469A (en) * 2017-03-17 2017-05-31 沈阳航空航天大学 A kind of fireworks algorithm based on simulated annealing Yu Gauss disturbance
CN107102223A (en) * 2017-03-29 2017-08-29 江苏大学 NPC photovoltaic DC-to-AC converter method for diagnosing faults based on improved hidden Markov model GHMM
CN107579707A (en) * 2017-10-13 2018-01-12 江苏大学 A kind of diagnosing failure of photovoltaic array method based on parameter identification
CN107658840A (en) * 2017-09-30 2018-02-02 山东科技大学 Distribution network failure restoration methods based on A* algorithms Yu fireworks algorithm
CN108062571A (en) * 2017-12-27 2018-05-22 福州大学 Diagnosing failure of photovoltaic array method based on differential evolution random forest grader

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6209412B2 (en) * 2013-09-27 2017-10-04 株式会社日立製作所 Fault diagnosis system and fault diagnosis method for photovoltaic power generation system
CN106443297B (en) * 2016-09-07 2018-05-22 江南大学 The decision tree SVM method for diagnosing faults of photovoltaic diode Clamp three-level inverter

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104767482A (en) * 2014-01-02 2015-07-08 上海岩芯电子科技有限公司 Method for diagnosing aging and short circuit of photovoltaic module online
CN105652952A (en) * 2016-04-18 2016-06-08 中国矿业大学 Maximum power point tracking method for photovoltaic power generation system based on fireworks algorithm
CN106776469A (en) * 2017-03-17 2017-05-31 沈阳航空航天大学 A kind of fireworks algorithm based on simulated annealing Yu Gauss disturbance
CN107102223A (en) * 2017-03-29 2017-08-29 江苏大学 NPC photovoltaic DC-to-AC converter method for diagnosing faults based on improved hidden Markov model GHMM
CN107658840A (en) * 2017-09-30 2018-02-02 山东科技大学 Distribution network failure restoration methods based on A* algorithms Yu fireworks algorithm
CN107579707A (en) * 2017-10-13 2018-01-12 江苏大学 A kind of diagnosing failure of photovoltaic array method based on parameter identification
CN108062571A (en) * 2017-12-27 2018-05-22 福州大学 Diagnosing failure of photovoltaic array method based on differential evolution random forest grader

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
Parameter extraction of solar photovoltaic models using an improved whale optimization algorithm;Guojiang Xiong等;《Energy Conversion and Management》;20180820;第174卷;第388-405页 *
Real Time Fault Detection in Photovoltaic Systems;Mohamed Hassan Ali等;《Energy Procedia》;20171231;第111卷;第914-923页 *
烟花算法研究进展;谭营等;《智能系统学报》;20141031;第9卷(第5期);第515-528页 *
考虑降低暂态电压失稳风险的动态无功优化配置方法;周仕豪;《电力系统保护与控制》;20180401;第46卷(第7期);第68-75页 *
自适应爆炸半径和差分变异的增强烟花算法;叶雯雯等;《应用数学进展》;20180228;第7卷(第2期);第152-162页 *

Also Published As

Publication number Publication date
CN109409420A (en) 2019-03-01

Similar Documents

Publication Publication Date Title
CN105134510A (en) State monitoring and failure diagnosis method for wind generating set variable pitch system
CN111293683B (en) Distribution network self-healing optimization method considering safety and economy
Djalab et al. Robust method for diagnosis and detection of faults in photovoltaic systems using artificial neural networks
CN112083270B (en) Wind power plant current collection line single-phase earth fault line selection method based on correlation coefficient
Yun et al. Research on fault diagnosis of photovoltaic array based on random forest algorithm
CN106846436B (en) Method for dividing multi-peak P-U curve of series photovoltaic module into regions
CN115208310A (en) Monitoring method of photovoltaic power generation system
CN109409420B (en) Photovoltaic string fault diagnosis method under non-uniform irradiance
Li et al. A line-fault cause analysis method for distribution network based on decision-making tree and machine learning
Adouni et al. DC-DC converter fault diagnostic in PV system
Jadin et al. A sizing tool for PV standalone system
WO2023103455A1 (en) Topological structure optimization method and system for power collection system of large-scale offshore wind farm
CN107271916B (en) Battery plate group string health state detection method
CN115117944A (en) Power control method for wind power and photovoltaic combined power generation
Ghazali et al. A multi-scale dual-stage model for PV array fault detection, classification, and monitoring technique
CN113417809A (en) Visual lightning stroke monitoring method and system
Liu et al. Fault diagnosis of photovoltaic array based on XGBoost method
Wang et al. Fault diagnosis method based on an improved KNN algorithm for PV strings
Dhoundiyal et al. Fault detection and classification in solar photovoltaic array
Zhang et al. Analysis of the factors affecting the fault of photovoltaic power generation equipment
Lin et al. Recognition and Reconstruction of Photovoltaic Output Abnormal Data Based on Geographic Correlation
CN104779907B (en) A kind of work state monitoring method of solar photovoltaic power plant and system
CN109725674A (en) A kind of optimization algorithm of the photovoltaic system maximal power tracing based on SA+PSO hybrid algorithm
Amiri et al. Faults detection of PV systems based on extracted parameters using Modified Grey Wolf algorithm
Jing et al. On-line Diagnosis Technology of Distributed Photovoltaic Abnormal Generation Based on Output Prediction

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant