Photovoltaic module fault classification system and method based on genetic algorithm

Abstract

This invention proposes a photovoltaic module fault classification system and method based on a genetic algorithm, belonging to the field of photovoltaic technology. The method involves acquiring standard sample data under normal system conditions and calculating the temperature and power values ​​of the standard samples for each time period under normal conditions; then acquiring test sample data under current conditions, calculating the temperature and power values ​​of the test samples for each time period, comparing the power and temperature values ​​of the test samples with those of the standard samples, and performing fault classification. The method can quickly determine whether a photovoltaic module has a fault and the type of fault, thereby providing appropriate treatment measures. This saves technicians time in troubleshooting, reduces the rate of misjudgment of faults, reduces downtime caused by troubleshooting and locating faults, and improves the system's power generation efficiency.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaics, and in particular relates to a photovoltaic component fault classification system and method based on genetic algorithms. Background Art

[0002] During use, photovoltaic modules are often covered by bird droppings, dust, shadows, fallen leaves, snow, etc. Due to the existence of local obstruction, the current and voltage of some cells change, resulting in a local temperature rise on these cell modules, that is, a hot spot effect, which greatly reduces the power generation. In addition to hot spot failures, short circuits or open circuits may also occur due to serious situations such as lightning strikes and burning of modules (including junction boxes), as well as shadow obstruction failures. All failures will lead to reduced module power. Currently, after a failure occurs, it needs to be manually inspected, and it relies on manual experience, and there is no unified standard. In addition, manual inspection will consume a lot of time, resulting in reduced power generation. Therefore, there is an urgent need for an intelligent classification system and method for photovoltaic module failures. Summary of the Invention

[0003] In response to the problems existing in the background technology, the present invention provides a photovoltaic module fault classification system and method based on genetic algorithm.

[0004] The technical solution of the present invention is:

[0005] A photovoltaic module fault classification method based on a genetic algorithm comprises the following steps:

[0006] S1. Under normal system conditions, select a sunny day and obtain the radiation intensity, ambient temperature, current, and voltage values ​​at different time periods every day within a period as standard samples;

[0007] S2. Under the current conditions, select a sunny day and obtain the radiation intensity value, ambient temperature value, current value, and voltage value at different time periods every day within a period as samples to be tested;

[0008] S3. In the sample to be tested, if the current value is zero, it is an open circuit fault; if the voltage value is zero, it is a short circuit fault; otherwise, proceed to the next step;

[0009] S4, selecting sample data of the sample to be tested that has the same irradiation intensity and ambient temperature as the standard sample in any time period, and calculating the power value and temperature value of the selected standard sample data, as well as the power value and temperature value of the selected sample data to be tested;

[0010] S5. Compare the power values ​​and temperature values ​​of the selected standard sample data obtained in S4 with the selected sample data to be tested, and classify the faults according to the comparison results.

[0011] Furthermore, in steps S1 and S2, sunny weather is selected to obtain the radiation intensity values, ambient temperature values, current values ​​and voltage values ​​in different time periods between 9:00 am and 3:00 pm.

[0012] Furthermore, in step S4, if the irradiation intensity and ambient temperature values ​​of the sample to be tested in a certain time period are different from the irradiation intensity and ambient temperature values ​​of the selected standard sample data, they are standardized and converted into sample data with the same irradiation intensity and ambient temperature as the selected standard sample data.

[0013] Furthermore, in step S4, the power value and temperature value of the selected standard sample data are calculated as follows:

[0014] The photocurrent of the photovoltaic module at the time corresponding to the selected standard sample data is calculated as:

[0015]

[0016] Among them, I pv0 =I ph0 +I0,I ph0 is the photocurrent under standard conditions, I0 is the reverse saturation current; Q0 is the standard irradiation intensity; is the irradiance intensity value at the corresponding moment of the selected standard sample data;

[0017] Calculate the temperature value T of the photovoltaic module at the corresponding moment of the selected standard sample data 0 (t):

[0018]

[0019] Among them, R s is the series resistance; G = nk / q, n is the characteristic coefficient of the PN junction material, k is the Boltzmann constant, and q is the electron charge; and are the current and voltage values ​​of the photovoltaic modules at the corresponding moments of the selected standard sample data;

[0020] Calculate the power value of the photovoltaic module at the corresponding moment of the selected standard sample data:

[0021]

[0022] Among them, W 0 (t) is the power value of the photovoltaic module at the corresponding moment of the selected standard sample data.

[0023] Furthermore, in step S4, the power value and temperature value of the selected sample data to be detected are calculated as follows:

[0024] Calculate the photocurrent of the photovoltaic module at the time corresponding to the selected sample data to be tested:

[0025]

[0026] Among them, Q s (t) is the irradiance value at the corresponding moment of the selected sample data to be tested;

[0027] Calculate the temperature value T(t) of the photovoltaic module at the corresponding moment of the selected sample data to be tested:

[0028]

[0029] Among them, I s (t) and U s (t) are the current and voltage values ​​of the photovoltaic module at the corresponding moment of the selected sample data to be tested;

[0030] Calculate the power value of the photovoltaic module at the corresponding moment of the selected sample data to be tested:

[0031] W(t)=U s (t)*I s (t) (6)

[0032] Wherein, W(t) is the power value of the photovoltaic module at the corresponding moment of the selected sample data to be tested.

[0033] Furthermore, the G, R s , I pv0 The solutions for I0 include:

[0034] Obtain the system calibration value of the photovoltaic module, including: the maximum output power P of the photovoltaic module under standard sunshine and temperature conditions m , Maximum operating voltage U m , Maximum operating current I m , open circuit voltage U oc and short-circuit current I sc ;

[0035] According to the volt-ampere characteristic curve equation:

[0036]

[0037] Where: U is the voltage value of the photovoltaic module; I is the current value of the photovoltaic module; T is the absolute temperature of the photovoltaic module; I ph is the photocurrent, I0 is the reverse saturation current, let I pv =I ph +I0;R s is the series resistance; n is the characteristic coefficient of the PN junction material; k is the Boltzmann constant; q is the electron charge;

[0038] The maximum operating voltage U m and the maximum operating current I m Substituting the value of into formula (7), we get:

[0039]

[0040] Where: G = nk / q, T0 is the absolute temperature of the photovoltaic module under standard conditions, I pv0 =I ph0 +I0,I ph0 is the photocurrent of the photovoltaic module under standard conditions;

[0041] The open circuit voltage U oc Substituting into formula (7), we get:

[0042]

[0043] The short-circuit current I sc Substituting into formula (7), we get:

[0044]

[0045] Taking the partial derivative of current with respect to voltage in formula (7), and then using the calculated relationship between current and voltage at the maximum power point, we can obtain:

[0046]

[0047] Eliminate the unknown number I in formulas (8) to (11) pv0 and I0, we have only two unknowns G and R s The system of equations:

[0048]

[0049] Genetic algorithm is used to solve formula (12) and (13) to obtain G and R s The value of G and R s Substitute the value into formula (9) and formula (10) to obtain and the value of I0.

[0050] Furthermore, the solving steps of the genetic algorithm are as follows:

[0051] 1) Randomly generate the initial population

[0052] Parameters G and R s The encoding method adopts binary multi-parameter cascade encoding method, and the initial population is randomly generated according to this encoding method;

[0053] 2) Calculation of fitness

[0054] set up:

[0055]

[0056] The multi-objective optimization function of the genetic algorithm is obtained:

[0057]

[0058] The weight coefficient change method is used to transform the above multi-objective optimization problem into a single-objective optimization problem:

[0059]

[0060] Since the objective function in formula (17) is a positive value, the solution of the objective function is the fitness value of the individual;

[0061] 3) Design genetic operators

[0062] Set the population size M, the termination number T, and the crossover probability p c , mutual difference probability p m , generation gap G;

[0063] The selection operation in the genetic algorithm uses the proportional selection operator combined with the optimal preservation strategy. First, the fitness value of each individual is calculated based on the proportional selection operator, and the corresponding selection probability is generated accordingly. The number of selections for each individual is then determined based on the selection probability. If the number of random selections is less than the number of individuals in the population M, the optimal preservation strategy is used to preserve the best individuals. These individuals do not participate in crossover and mutation operations and are directly inherited to the next generation of the population. The remaining individuals undergo normal crossover and mutation operations.

[0064] 4) When the objective function shown in formula (17) is less than the specified threshold or reaches the termination generation T, the genetic operation is terminated. At this time, G and R s The value is the final calculation result.

[0065] Furthermore, the specific process of step S5 is as follows:

[0066] Set ZW j is the average power difference between the sample data to be tested under the current condition and the standard sample data selected under normal conditions, where j = 1, 2, ..., m, represents the number of sample data to be tested, then:

[0067]

[0068] Where t1 and t2 are the moments when the selected sample data to be tested and the selected standard sample data are at the same irradiation intensity and ambient temperature respectively;

[0069] Let A and B be preset thresholds greater than zero. For any value of j among j = 1, 2, …, m, if |ZW j | > A, it is considered that this photovoltaic module has a fault, and then continue to the next step of judgment; otherwise, it is considered that the photovoltaic module is normal;

[0070] When |ZW j | > A, that is, when the photovoltaic module has a fault, let ZT i be the temperature average difference between all selected sample data to be detected with |ZW j | > A (i.e., the sample data with faults to be detected) and the selected standard sample data under normal conditions. Here, i = 1, 2, …, n represents the number of sample data with faults to be detected. Then:

[0071]

[0072] If for all i = 1, 2, …, n, |ZT i | < B, then the photovoltaic module has an abnormal aging fault; if for i = 1, 2, …, n, at least one i satisfies |ZT i | > B, and ZT i > 0, then the photovoltaic module has a hot spot fault; if for all i = 1, 2, …, n, |ZT i | > B, and ZT i < 0, then the photovoltaic module has a real shadow occlusion fault; if for i = 1, 2, …, n, at least one i satisfies |ZT i | > B, and ZT i < 0, and at the same time there is at least one i value that satisfies |ZT i | < B, then the photovoltaic module has a virtual shadow occlusion fault.

[0073] The present invention also proposes a photovoltaic module fault classification system based on a genetic algorithm, including a component data collector, a wireless gateway, a host computer, an illuminance meter, and an environmental illuminance meter;

[0074] The component data collector includes a sampling module, a CPU control module, and a communication module; the input end and output end of the sampling module are respectively connected to the output end of the photovoltaic module and the input end of the CPU control module. The sampling module includes a voltage sampling module and a current sampling module, and is used to collect the output voltage and output current of the photovoltaic array in real time; the CPU control module is used to convert and encode the voltage and current values;

[0075] The output ends of the ambient illuminance meter and the illuminance meter are connected to the input end of the CPU control module; the input end of the communication module is connected to the output end of the CPU control module, and is used to wirelessly transmit the data of the CPU control module to the wireless gateway; the wireless gateway transmits the data to the host computer through the cloud; the host computer executes the above-mentioned fault classification method to classify the photovoltaic component faults on the collected data.

[0076] Compared with the prior art, the present invention has the following beneficial effects: the present invention can quickly determine whether a photovoltaic module has a fault and the type of fault, and thus provide corresponding treatment measures, saving technicians' troubleshooting time, reducing the misjudgment rate of faults, reducing downtime caused by troubleshooting and locating faults, and improving the system's power generation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0078] Figure 2 A schematic diagram of the system structure provided by an embodiment of the present invention;

[0079] Figure 3 This is the convergence process of the genetic algorithm provided by the embodiment of the present invention. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0081] See also Figure 2 A photovoltaic module fault classification system based on a genetic algorithm includes a module data collector, a wireless gateway, a host computer, an illuminance meter, and an ambient illuminance meter; the module data collector measures a photovoltaic array composed of a plurality of photovoltaic modules connected in series; the output end of the photovoltaic array is connected to the input end of the module data collector;

[0082] The component data collector includes a sampling module, a CPU control module, and a communication module; the input end of the sampling module is connected to the output end of the photovoltaic component, and the output end of the sampling module is connected to the input end of the CPU control module. The sampling module includes a voltage sampling module and a current sampling module for real-time acquisition of the output voltage and output current of the photovoltaic array; the CPU control module is used to convert and encode the voltage and current values;

[0083] The output ends of the ambient illuminance measuring instrument and the illuminance measuring instrument are connected to the input end of the CPU control module; the input end of the communication module is connected to the output end of the CPU control module, and is used to wirelessly transmit the data of the CPU control module to the wireless gateway; the wireless gateway transmits the data to the host computer through the cloud.

[0084] The host computer is used to classify component faults based on the collected data. This includes obtaining standard sample data under normal system conditions (meaning when the system has no faults) and calculating the temperature and power values ​​of the standard samples in each time period under normal conditions. Then, under the current conditions (during the monitoring process), it obtains the sample data to be tested and calculates the temperature and power values ​​of the sample to be tested in each time period. The power and temperature values ​​of the sample to be tested are compared with those of the standard sample and the fault classification is performed. The specific steps include the following:

[0085] Step 1) Under normal system conditions, obtain standard samples

[0086] Select sunny weather, where sunny weather refers to abundant sunshine. During the period between 9:00 AM and 3:00 PM, obtain the irradiance intensity, ambient temperature, current, and voltage values ​​for different time periods. Repeat this for several days as standard samples.

[0087] Step 2) Under normal system conditions, calculate the temperature and power values ​​of the photovoltaic modules at each time period as a standard for comparison.

[0088] In the standard sample, all the irradiation intensity values, current values ​​and voltage values ​​are arranged into a sequence in chronological order to obtain the irradiation intensity sequence value. Current sequence values ​​for standard PV modules And standard voltage sequence value Where t is the time variable; the photocurrent sequence value of the standard sample is calculated as:

[0089]

[0090] Among them, I pv0 =I ph0 +I0,I ph0 is the photocurrent under standard conditions, I0 is the reverse saturation current; Q0 is the standard irradiation intensity;

[0091] Calculate the temperature value T of the standard sample in each time period under normal circumstances 0 (t):

[0092]

[0093] Among them, R sis the series resistance; G = nk / q, n is the characteristic coefficient of the PN junction material, k is the Boltzmann constant, and q is the electron charge;

[0094] According to the formula that power equals current multiplied by voltage, the power series value of the standard sample is:

[0095]

[0096] Step 3) Under the current situation, that is, under monitoring, obtain the sample to be tested

[0097] Still choose a sunny day, between 9 am and 3 pm, and obtain the radiation intensity value, ambient temperature value, current value and voltage value of different time periods as the samples to be tested;

[0098] If the current value is equal to zero, it is an open circuit fault. If the voltage value is equal to zero, it is a short circuit fault. If neither is the case, proceed to the next step.

[0099] Step 4) Under the current conditions, calculate the temperature and power values ​​of the photovoltaic components in each time period

[0100] Sequence all the irradiation intensity values, current values ​​and voltage values ​​of the sample to be tested in chronological order to obtain the irradiation intensity sequence value Q s (t), the current sequence value I of the sample photovoltaic module to be tested s (t) and voltage sequence value U s (t), where t is the time variable. Since the photocurrent is directly proportional to the irradiation intensity, the photocurrent sequence value of the sample to be tested is:

[0101]

[0102] Calculate the temperature value T(t) of the test sample in each time period:

[0103]

[0104] According to the principle that power equals current multiplied by voltage, the power sequence value of the test sample is:

[0105] W(t)=U s (t)*I s (t) (6)

[0106] Step 5) Select sample data with the same irradiation intensity and ambient temperature as the standard sample in any time period from the sample to be tested, calculate the power value and temperature value of the selected standard sample data and the selected sample data to be tested, and compare them. Fault classification is performed based on the comparison results; the details are as follows:

[0107] Randomly select the sample data of any time period in the standard sample, and select the sample data in the sample to be detected with the same irradiation intensity and environmental temperature as the selected standard sample data. If the irradiation intensity and environmental temperature values of the sample to be detected in a certain time period are different from those of the selected standard sample data, perform standardization processing on it to convert it into sample data with the same irradiation intensity and environmental temperature as the selected standard sample data; Let ZW j be the average power difference between the selected sample data to be detected in the current situation and the selected standard sample data in the normal situation, where j = 1, 2,..., m, representing the number of selected sample data to be detected, then:

[0108]

[0109] where t1 and t2 are the moments of the selected sample data to be detected and the selected standard sample data under the same irradiation intensity and environmental temperature respectively;

[0110] Let A and B be preset thresholds greater than zero. For any j value in j = 1, 2,..., m, if |ZW j | > A, it is considered that this photovoltaic module has a fault, and then continue to the next judgment, otherwise it is considered that the photovoltaic module is normal;

[0111] When |ZW j | > A, that is, when the photovoltaic module has a fault, let ZT i be the average temperature difference between all selected sample data to be detected with |ZW j | > A in the current situation, that is, the fault sample data to be detected and the selected standard sample data in the normal situation, where i = 1, 2,..., n, representing the number of fault sample data to be detected, then:

[0112]

[0113] If for all i = 1, 2,..., n, |ZT i | < B is satisfied, it is considered that the photovoltaic module has an abnormal aging type of fault; if for i = 1, 2,..., n, at least one i satisfies |ZT i | > B, and ZT i > 0, it is considered that the photovoltaic module has a hot spot type of fault; if for all i = 1, 2,..., n, |ZT i | > B is satisfied, and ZT i < 0, it is considered that the photovoltaic module has a real shadow occlusion type of fault; if for i = 1, 2,..., n, at least one i satisfies |ZT i | > B, and ZT i < 0, and at the same time there is at least one i value that satisfies |ZT iIf < B, it is considered that the photovoltaic module has a virtual shadow occlusion fault, as follows:

[0114]

[0115] The host computer classifies the photovoltaic module faults into five categories according to the following rules, including open - circuit faults, short - circuit faults, shadow occlusion faults, hot - spot faults, and aging faults. All faults will cause the module power to decrease. Open - circuit faults are generally serious situations such as the module (including the junction box) being struck by lightning or burned, and at this time the current is equal to zero. In the case of a short - circuit fault, the voltage is equal to zero. In the case of a shadow occlusion fault, the module temperature decreases, including real shadow occlusion and virtual shadow occlusion. In the case of a hot - spot fault, the module temperature increases, and in the case of an aging fault, the module temperature remains unchanged.

[0116] During the above - mentioned calculation process by the host computer, it is necessary to calculate the temperature values for each time period, which is calculated based on the volt - ampere characteristic curve equation:

[0117]

[0118] [[ID=1 ]]5In the formula: U is the voltage value of the photovoltaic module; I is the current value of the photovoltaic module; T is the absolute temperature of the photovoltaic module; I ph is the photo - generated current, I0 is the reverse saturation current, let I pv = I ph + I0; R s is the series resistance; n is the P - N junction material characteristic coefficient; k is the Boltzmann constant; q is the electron charge;

[0119] Obtain the current value I s 、voltage value U s and the current irradiance intensity value Q s , then:

[0120]

[0121] Among them: G = nk / q; I ph is the current photo - generated current, Q0 is the standard irradiance intensity.

[0122] In the above formula, the solution methods of G, R s 、I pv0 and I0 include:

[0123] Assign the system calibration values of the photovoltaic module to the host computer, including: under the conditions of standard sunlight and temperature, the maximum output power P m 、maximum working voltage U[[ID=S5]] m 、maximum working current I m 、open - circuit voltage U oc and short - circuit current Isc ;

[0124] The maximum operating voltage U m and the maximum operating current I m Substituting the value of into formula (9), we get:

[0125]

[0126] Where: T0 is the absolute temperature of the photovoltaic module under standard conditions, I pv0 =I ph0 +I0,I ph0 is the photocurrent of the photovoltaic module under standard conditions;

[0127] The open circuit voltage U oc Substituting into formula (9), we get:

[0128]

[0129] The short-circuit current I sc Substituting into formula (9), we get:

[0130]

[0131] Taking the partial derivative of the current with respect to the voltage in formula (9), and then using the calculated relationship between the current and voltage at the maximum power point, we can obtain:

[0132]

[0133] Eliminate the unknown number I in the above formulas (11) to (13) pv0 and I0, we have only two unknowns G and R s The system of equations:

[0134]

[0135] Genetic algorithm is used to solve formula (15) and (16) to obtain G and R s The value of G and R s Substitute the value into formula (12) and formula (13) to obtain and the value of I0; the specific process is as follows:

[0136] 1) Randomly generate the initial population

[0137] Parameters G and R s The encoding method adopts binary multi-parameter cascade encoding method, and the initial population is randomly generated according to this encoding method;

[0138] 2) Calculation of fitness

[0139] set up:

[0140]

[0141] The multi-objective optimization function of the genetic algorithm is obtained:

[0142]

[0143] The weight coefficient change method is used to transform the above multi-objective optimization problem into a single-objective optimization problem:

[0144]

[0145] Since the objective function in formula (20) is a positive value, the solution of the objective function is the fitness value of the individual;

[0146] 3) Design genetic operators

[0147] Setting parameters include: population size M, termination generation number T, crossover probability p c , mutual difference probability p m , generation gap G;

[0148] The selection operation in the genetic algorithm uses a proportional selection operator combined with an optimal preservation strategy. First, the fitness value of each individual is calculated based on the proportional selection operator, and the corresponding selection probability is generated accordingly. The number of selections for each individual is then determined based on the selection probability. If the number of random selections is less than the number of individuals in the population M, the optimal preservation strategy is used to preserve the best individuals. These individuals do not participate in crossover and mutation operations and are directly inherited to the next generation of the population. The remaining individuals undergo normal crossover and mutation operations.

[0149] 4) When the objective function shown in formula (20) is less than the specified threshold or reaches the termination generation T, the genetic operation is terminated. At this time, G and R s The value is the final calculation result. The convergence process is as follows Figure 3 shown.

[0150] Finally, it should be noted that the above embodiments are intended to illustrate the technical solutions of the present invention and do not constitute any form of limitation of the present invention. Those skilled in the art should fully understand that it is entirely feasible to modify the technical solutions described in the above embodiments or to replace any or all of the technical features with equivalents. Such modifications or replacements, as long as they do not deviate from the scope of protection defined by the claims of the present invention, should be considered reasonable extensions of the present invention.

Claims

1. A photovoltaic module fault classification method based on genetic algorithm, characterized in that: The following steps are involved: S1. Under normal system conditions, select a sunny day and obtain the radiation intensity, ambient temperature, current, and voltage values ​​at different time periods every day within a period as standard samples; S2. Under the current conditions, select a sunny day and obtain the radiation intensity value, ambient temperature value, current value, and voltage value at different time periods every day within a period as samples to be tested; S3. In the sample to be tested, if the current value is zero, it is an open circuit fault; if the voltage value is zero, it is a short circuit fault; otherwise, proceed to the next step; S4. Select sample data with the same irradiation intensity and ambient temperature as the standard sample in any time period from the sample to be tested, and calculate the power value and temperature value of the selected standard sample data, as well as the power value and temperature value of the selected sample data to be tested; the calculation method of the power value and temperature value of the selected standard sample data is as follows: The photocurrent of the photovoltaic module at the time corresponding to the selected standard sample data is calculated as: (1) in, , is the photocurrent under standard conditions, is the reverse saturation current; is the standard irradiation intensity; is the irradiance intensity value at the corresponding moment of the selected standard sample data; Calculate the temperature value of the photovoltaic module at the corresponding moment of the selected standard sample data : (2) in, is the series resistance; G=nk / q , is the PN junction material characteristic coefficient, is the Boltzmann constant, q is the electron charge; and are the current and voltage values ​​of the photovoltaic modules at the corresponding moments of the selected standard sample data; Calculate the power value of the photovoltaic module at the corresponding moment of the selected standard sample data: (3) in, is the power value of the photovoltaic module at the corresponding moment of the selected standard sample data; described G, R s 、I pv0 and I The solutions to 0 include: Obtain the system calibration value of the photovoltaic module, including: the maximum output power of the photovoltaic module under standard sunshine and temperature conditions P m , Maximum operating voltage U m , Maximum operating current I m , open circuit voltage U oc and short-circuit current I sc ; According to the volt-ampere characteristic curve equation: (4) Where: U is the voltage value of the photovoltaic module; I is the current value of the photovoltaic module; T is the absolute temperature of the PV module; I ph is the photogenerated current, is the reverse saturation current, let I pv =I ph +I 0; The maximum operating voltage U m and maximum operating current I m Substituting the value of into formula (4), we get: (5) in: T 0 is the absolute temperature value of the photovoltaic module under standard conditions, = + ; The open circuit voltage Substituting into formula (4), we get: (6) The short-circuit current Substituting into formula (4), we get: (7) Taking the partial derivative of current with respect to voltage in formula (4), and then using the calculated relationship between current and voltage at the maximum power point, we can obtain: (8) Eliminate the unknowns in formulas (5) to (8) and , we have only two unknowns G and R s The system of equations: (9) (10) Genetic algorithm is used to solve formula (9) and (10), and we get G and R s , and G and R s Substitute the value into formula (6) and formula (7) to obtain and The solution steps of the genetic algorithm are as follows: 1) Randomly generate the initial population parameter G and R s The encoding method adopts binary multi-parameter cascade encoding method, and the initial population is randomly generated according to this encoding method; 2) Calculation of fitness set up: (11) (12) The multi-objective optimization function of the genetic algorithm is obtained: (13) The weight coefficient change method is used to transform the above multi-objective optimization problem into a single-objective optimization problem: (14) The solution of the objective function in formula (14) is the fitness value of the individual; 3) Design genetic operators Setting the group size M , terminates the algebra T , crossover probability p c , mutually different probabilities p m , generation gap G ; First, the fitness value of each individual is calculated based on the proportional selection operator, and the corresponding selection probability is generated accordingly; then the number of selections for each individual is obtained based on the selection probability. If the number of random selections is less than the number of individuals in the group, M , we combine the optimal preservation strategy to preserve the best individuals, without participating in crossover and mutation operations, and directly pass them on to the next generation; the remaining individuals undergo normal crossover and mutation operations; 4) When the objective function shown in formula (14) is less than the specified threshold or reaches the termination algebra T When , the genetic operation is terminated. G and R s The value is the final calculation result; S5. Compare the power values ​​and temperature values ​​of the selected standard sample data obtained in S4 with the selected sample data to be tested, and classify the faults according to the comparison results.

2. A photovoltaic module fault classification method based on genetic algorithm according to claim 1, characterized in that: In step S4, the power value and temperature value of the selected sample data to be detected are calculated as follows: Calculate the photocurrent of the photovoltaic module at the time corresponding to the selected sample data to be tested: (15) in, is the irradiance intensity value at the corresponding moment of the selected sample data to be tested; Calculate the temperature value of the photovoltaic module at the corresponding moment of the selected sample data to be tested : (16) in, and are the current value and voltage value of the photovoltaic module at the corresponding moment of the selected sample data to be tested; Calculate the power value of the photovoltaic module at the corresponding moment of the selected sample data to be tested: (17) in, is the power value of the photovoltaic module at the corresponding moment of the selected sample data to be tested.

3. A photovoltaic module fault classification method based on genetic algorithm according to claim 1 or 2, characterized in that: The specific process of step S5 is as follows: set up ZW j is the power difference between the sample data to be tested under the current situation and the standard sample data selected under normal circumstances, where j=1,2,…,m , represents the number of sample data selected to be tested, then: (18) in t 1, t 2 are the moments when the selected sample data to be tested and the selected standard sample data are at the same irradiation intensity and ambient temperature; set up A and B is a pre-set threshold greater than zero. j=1,2,…,m Any of j Value, if , then it is considered that the PV module is faulty and the next step is to be judged; otherwise, it is considered that the PV module is normal; when When a PV module fails, For all current conditions The selected sample data to be tested is the average temperature difference between the fault sample to be tested and the standard sample data selected under normal conditions, where i=1, 2,…,n , represents the number of fault samples to be detected, then: (19) If for all i=1,2,…,n , all satisfied , then the PV module is abnormally aged; if i= 1,2,…,n , at least one i satisfy ,and , then the PV module is a hot spot failure; if for all i=1,2,…,n , all satisfied ,and , then the PV module is a real shadow blocking fault; if i=1,2,…,n , at least one i satisfy ,and , and there is at least one i value, satisfy , then the PV module is a virtual shadow blocking type fault.

4. The photovoltaic module fault classification method based on genetic algorithm according to claim 1, characterized in that: In step S4, if the radiation intensity and ambient temperature values ​​of the sample to be tested in a certain time period are different from those of the selected standard sample data, they are standardized and converted into sample data with the same radiation intensity and ambient temperature as the selected standard sample data.

5. The photovoltaic module fault classification method based on genetic algorithm according to claim 1, characterized in that: In the steps S1 and S2, sunny weather is selected to obtain the radiation intensity values, ambient temperature values, current values ​​and voltage values ​​in different time periods between 9:00 am and 3:00 pm.

6. A photovoltaic module fault classification system based on genetic algorithm, characterized in that: Including component data collector, wireless gateway, host computer, illuminance meter and ambient illuminance meter; The component data collector includes a sampling module, a CPU control module and a communication module; the input and output ends of the sampling module are respectively connected to the output end of the photovoltaic component and the input end of the CPU control module. The sampling module includes a voltage sampling module and a current sampling module for real-time acquisition of the output voltage and output current of the photovoltaic array; the CPU control module is used to convert and encode the voltage and current values; The output ends of the ambient illuminance measuring instrument and the illuminance measuring instrument are connected to the input end of the CPU control module; the input end of the communication module is connected to the output end of the CPU control module, and is used to wirelessly transmit the data of the CPU control module to the wireless gateway; the wireless gateway transmits the data to the host computer through the cloud; the host computer executes the fault classification method described in any one of claims 1 to 5 to classify the faults of the photovoltaic components.

Images