Photovoltaic array fault detection and classification method based on fuzzy logic and I-V trend line

Abstract

The invention provides a photovoltaic array fault detection and classification method based on fuzzy logic and an I-V trend line. The method comprises the steps that S1, a data set used for training and verification is constructed under series-parallel SP and TCT full-crosslinking configuration; s2, preprocessing data in the obtained data set, and defining a fault index based on a photovoltaic array I-V trend line and P-V maximum power linear interpolation; s3, according to the change of the I-V trend line, a triangular membership function is selected to represent different numerical ranges of the fault index fi and the maximum power point voltage Vmpp; s4, according to the boundary range and number of the membership function, combining the fault characteristics of the photovoltaic array, and defining an'if-thin 'rule in fuzzy logic reasoning; and S5, performing defuzzification processing by using a maximum membership degree method and outputting. According to the method, the maximum membership degree method is adopted to carry out anti-fuzzy processing, so that accurate detection and classification of photovoltaic array faults are realized.

Description

Technical Field

[0001] The invention relates to the technical field of photovoltaic power generation, and in particular to a photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line. Background Art

[0002] Photovoltaic power generation is an important form of renewable energy. It converts sunlight into electrical energy and has the characteristics of being clean, environmentally friendly and renewable. In recent years, with the continuous advancement of photovoltaic technology and the gradual reduction of costs, photovoltaic arrays have been widely promoted and applied in my country and have become an important part of my country's new energy field. However, due to the inherent defects of photovoltaic modules, environmental factors and long-term use, photovoltaic arrays are prone to failures during operation. For example, shadow, short circuit faults and open circuit faults, etc., these faults will cause the power generation efficiency of photovoltaic arrays to decrease, or even stop generating electricity, seriously affecting the economic benefits and operating life of photovoltaic systems. The research and development of photovoltaic array fault diagnosis technology can not only effectively improve the operating efficiency and economic benefits of photovoltaic arrays, but also reduce operation and maintenance costs and increase the service life of photovoltaic modules. In addition, the research results of photovoltaic array fault diagnosis technology can also provide technical support for the standardization and large-scale development of the photovoltaic industry. At present, the automatic identification and classification of photovoltaic array faults have not yet been achieved.

[0003] Therefore, studying the fault diagnosis technology of photovoltaic arrays and realizing timely diagnosis and processing of faults are of great significance to improving the power generation efficiency and economic benefits of photovoltaic arrays. Some commonly used methods, such as fault diagnosis methods based on electrical parameters, are simple and easy to use, but are easily affected by environmental factors and system noise, and it is difficult to distinguish different types of faults. Summary of the invention

[0004] The present invention provides a photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line, which detects and classifies common faults in photovoltaic arrays, including open circuit, short circuit and shielding faults, covering the main fault types in photovoltaic arrays.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line includes the following steps: S1. In the series-parallel SP and TCT full cross-link configuration, different fault scenarios are set in the PV array through simulation experiments, and the corresponding open circuit voltage is recorded. V oc , short circuit current I sc , power generation P, to construct a dataset for training and validation; the data includes the I-V trend line and PV trend line collected under each fault scenario. The IV trend line refers to the linear trend of the IV curve approximately near the short-circuit current; S2, preprocessing the acquired data in the data set, and defining the fault index based on the PV array IV trend line and PV maximum power linear interpolation; S3. According to the change of IV trend line, the triangle membership function is selected to characterize the fault index. fi and maximum power point voltage V mpp Different value ranges; S4. According to the boundary range and quantity of the membership function and the fault characteristics of the photovoltaic array, define the “if-then” rules in fuzzy logic reasoning; S5. Use the maximum membership method to perform defuzzification processing and output.

[0006] The specific process of S3 mentioned above is: First determine the fault index and maximum power point voltage V mpp The fault index and the maximum power point voltage are used to construct the fuzzy input membership function; then, the fault index is converted into fi It is divided into four fuzzy sets corresponding to different fault severity levels: very low VL, low L, medium M and high H, the maximum power point voltage V mpp It is divided into three fuzzy sets corresponding to different voltage ranges: low L, medium M and high H. Finally, the left and right endpoints and the midpoint of each range are selected to represent the three endpoints of the triangular membership function to calculate the fault index. fi And the maximum power point voltage V mpp The membership degree of each fuzzy set is used to judge the degree of belonging to different fuzzy sets; the output membership function type is determined according to the fault type, including four fault types: no fault NF, open circuit fault OCF, short circuit fault SCF and shadow SF.

[0007] The above-mentioned "if-then" rules in S4 specifically include: Rule 1: If (fault index is medium) and (maximum power point voltage is high), then (PV fault is no fault) Rule 2: If (fault index is very low) and (maximum power point voltage is low), then (PV fault is an open circuit fault) Rule 3: If (fault index is high) and (maximum power point voltage is medium), then (PV fault is a short circuit fault) Rule 4: If (fault index is low) and (maximum power point voltage is high), then (the PV fault is a short circuit fault).

[0008] The specific process of S5 mentioned above is: First, the minimum value of the membership of each fuzzy subset is extracted from each fuzzy rule as the membership of the output fuzzy subset; then, the maximum membership is selected by comparing the size of the output membership corresponding to each rule. The fault type corresponding to the membership degree is the final derivation result.

[0009] The specific steps of S2 above include: S2.1. Take the voltage at the maximum power point as one of the parameters; S2.2. Based on the IV trend line slope and PV maximum power linear interpolation calculation of the PV array under various fault conditions, a newly defined parameter, namely the “fault index”, is obtained; the trend line slope under different fault scenarios can be calculated by (1), and the fault index value can be calculated by (2): ; (1) ; (2) In the formula, is the instantaneous current change, is the total voltage change, is the failure index, is the slope of the trend line, The maximum power of the photovoltaic array.

[0010] In the above S3, the failure index fi The intervals of the four different fault severity levels ranged from very low VL [19.16, 49.93], low L [49.93, 73.66], medium M [73.66, 75.02], and high H [75.02, 130.82]; Maximum power point voltage V mpp The three different voltage intervals range: low L [178.54, 195.29], medium M [195.29, 211.34] and high H [211.34, 231.17].

[0011] In the above S3, the fault index is calculated fi And the maximum power point voltage V mpp The membership process belonging to each fuzzy set is: Failure Index fi The attribute formula is: ; (3) ; (4) ; (5) ; (6) The corresponding maximum power point voltage V mpp The membership formula is: ; (7) ; (8) ; (9) Finally, the output membership function type is determined according to the fault type, including four fault types: no fault NF, open circuit fault OCF, short circuit fault SCF and shadow SF.

[0012] In the above S5, the calculation formula for the final derivation result is: ; (10).

[0013] The present invention provides a photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line, and newly defines the parameter "fault index" to quantify the degree of fault. In addition, a simple and easy-to-understand triangular membership function is selected to characterize the fault index. fi and maximum power point voltage V mpp Different numerical ranges can effectively handle the uncertainty in fuzzy logic reasoning, improve the accuracy of detection and classification, and provide a basis for taking corresponding protective measures; the effectiveness of the proposed technology was evaluated in the simulation and experiment of a 5KW photovoltaic array. The results show that this method can diagnose the occurrence of different faults with an accuracy rate of more than 98%. The present invention defines a new parameter fault index by further linear interpolation calculation based on the maximum power of the photovoltaic array and the slope of the I-V trend line fi , quantifies the degree of fault, is easy to understand, and does not require complex algorithms and models. It provides a basis for taking appropriate protection measures, which is easy to implement and maintain. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 is a flow chart of a photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line provided by an embodiment of the present invention; Figure 2 The impact of the fault on the TCT configuration in the present invention (a) Open circuit fault Figure 3 The impact of the fault on the TCT configuration in the present invention (b) short circuit fault Figure 4 It is the impact of the fault on the TCT configuration in the present invention (c) shadow fault; Figure 5 It is the impact of the fault on the SP configuration in the present invention (a) open circuit fault; Figure 6 It is the impact of the fault on the SP configuration in the present invention (b) short circuit fault; Figure 7 This is the impact of the fault on the SP configuration in the present invention (c) Shadow fault. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solutions and advantages of the present invention clearer, the following content will systematically and completely describe the specific technical solutions of the present invention in combination with the drawings provided according to the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0016] Embodiment 1: The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line includes the following steps: S1. In the series-parallel SP and TCT full cross-link configuration, different fault scenarios are set in the PV array through simulation experiments, and the corresponding open circuit voltage is recorded. V oc , short circuit current I sc , power generation P , to construct a dataset for training and validation; the data includes the I-V trend line and PV trend line collected under each fault scenario. The IV trend line refers to the linear trend of the IV curve approximately near the short-circuit current; S2, preprocessing the acquired data in the data set, and defining the fault index based on the PV array IV trend line and PV maximum power linear interpolation; S3. According to the change of IV trend line, the triangle membership function is selected to characterize the fault index. fi and maximum power point voltage V mpp Different numerical ranges for input into fuzzy logic reasoning; S4. According to the boundary range and quantity of the membership function and the fault characteristics of the photovoltaic array, define the “if-then” rules in fuzzy logic reasoning; S5. Use the maximum membership method to perform defuzzification processing and output.

[0017] The specific process of S3 mentioned above is: First determine the fault index and maximum power point voltage V mpp The fuzzy input membership function is constructed using the fault index (ranging from 19.16 to 130.82) and the maximum power point voltage (ranging from 178.54 to 231.17); then, the fault index is assigned to the fi It is divided into four fuzzy sets corresponding to different fault severity levels: very low VL, low L, medium M and high H, the maximum power point voltage V mpp It is divided into three fuzzy sets corresponding to different voltage ranges: low L, medium M and high H. Finally, the left and right endpoints and the midpoint of each range are selected to represent the three endpoints of the triangular membership function to calculate the fault index. fi And the maximum power point voltage V mpp The membership degree of each fuzzy set is used to judge the degree of belonging to different fuzzy sets; the output membership function type is determined according to the fault type, including four fault types: no fault NF, open circuit fault OCF, short circuit fault SCF and shadow SF.

[0018] The above-mentioned "if-then" rules in S4 specifically include: Rule 1: If (fault index is medium) and (maximum power point voltage is high), then (PV fault is no fault) Rule 2: If (fault index is very low) and (maximum power point voltage is low), then (PV fault is an open circuit fault) Rule 3: If (fault index is high) and (maximum power point voltage is medium), then (PV fault is a short circuit fault) Rule 4: If (fault index is low) and (maximum power point voltage is high), then (the PV fault is a short circuit fault).

[0019] The specific process of S5 mentioned above is: First, the minimum value of the membership of each fuzzy subset is extracted from each fuzzy rule as the membership of the output fuzzy subset; then, the maximum membership is selected by comparing the size of the output membership corresponding to each rule. The fault type corresponding to the membership degree is the final derivation result.

[0020] The specific steps of S2 above include: S2.1. Take the voltage at the maximum power point as one of the parameters, because each fault will cause different severity. V mpp value; S2.2. Based on the IV trend line slope and PV maximum power linear interpolation calculation of the PV array under various fault conditions, a newly defined parameter, namely the “fault index”, is obtained; the trend line slope under different fault scenarios can be calculated by (1), and the fault index value can be calculated by (2): ; (1) ; (2) In the formula, is the instantaneous current change, is the total voltage change, is the failure index, is the slope of the trend line, The maximum power of the photovoltaic array.

[0021] In the above S3, the failure index fi The intervals of the four different fault severity levels ranged from very low VL [19.16, 49.93], low L [49.93, 73.66], medium M [73.66, 75.02], and high H [75.02, 130.82]; Maximum power point voltage V mpp The three different voltage intervals range: low L [178.54, 195.29], medium M [195.29, 211.34] and high H [211.34, 231.17].

[0022] In the above S3, the fault index is calculated fi And the maximum power point voltage V mpp The membership process belonging to each fuzzy set is: Failure Index fi The attribute formula is: ; (3) ; (4) ; (5) ; (6) The corresponding maximum power point voltage V mpp The membership formula is: ; (7) ; (8) ; (9) Finally, the output membership function type is determined according to the fault type, including four fault types: no fault NF, open circuit fault OCF, short circuit fault SCF and shadow SF.

[0023] In the above S5, the calculation formula for the final derivation result is: ; (10).

[0024] Embodiment 2: The present invention establishes fuzzy indicators by collecting simulated fault data of photovoltaic arrays to verify the effectiveness of the method of the present invention.

[0025] Step 1: Set up different fault scenarios in the PV array through simulation experiments under SP and TCT configurations and record the corresponding open circuit voltages V oc , short circuit current I sc , power generation P , to build a dataset for training and validation. The data includes the I-V trend line collected under each fault scenario. The IV trend line refers to the linear trend of the IV curve near the short-circuit current. The simulation experiment uses a 6×6 photovoltaic array, in which the rated power of each module is 150 W. The specific parameters are shown in the table below;

[0026] The same operation steps were repeated on an experimental setup of a 5 kW PV system connected to an inverter, where the shadow fault was applied by covering the PV array with sunshades of different transparency. The open and short circuit faults were applied by intentionally cutting and shorting the connections at multiple locations.

[0027] Step 2: Preprocess the acquired data and define the fault index based on the PV array IV trend line and maximum power linear interpolation. Step 2 includes the following specific sub-steps: Step 2-1: Take the voltage at the maximum power point as one of the parameters, because each fault will cause a different severity. V mpp value; Step 2-2: A newly defined parameter, namely the “fault index”, is obtained based on the IV trend line slope and PV maximum power linear interpolation calculation of the PV array under various fault conditions. The trend line slope under different fault scenarios can be calculated by (1), and the fault index value can be calculated by (2).

[0028] ; (1) ; (2) In the formula, is the instantaneous current change, is the total voltage change, i is the failure index, m is the slope of the trend line, P max The maximum power of the photovoltaic array.

[0029] Step 3: Determine the type of membership function and generate input and output membership functions. Step 3 includes the following specific steps: Step 3-1: Determine the failure index fi and V mpp The fuzzy input membership function is constructed using the fault index (ranging from 19.16 to 130.82) and the maximum power point voltage (ranging from 178.54 to 231.17). Then, according to the numerical range, the fault index is fi It is divided into four fuzzy sets corresponding to different fault severity levels: very low (VL), low (L), medium (M) and high (H). The corresponding intervals are: VL: [19.16, 49.93], L: [49.93,73.66], M: [73.66, 75.02], H: [75.02, 130.82], V mpp Divided into three fuzzy sets corresponding to different voltage ranges: low (L), medium (M) and high (H); the corresponding interval ranges are: L: [178.54, 195.29], M: [195.29, 211.34], H: [211.34, 231.17]; Step 3-2: According to the change of IV trend line, choose to use triangle membership function to characterize the fault index fi and maximum power point voltage V mpp Different numerical ranges are used to input fuzzy logic reasoning. First, the left and right endpoints and the midpoint of each fuzzy set interval are selected to represent the three endpoints of the triangular membership function, and the fault index is calculated by the formula fi and the maximum power point voltage V mpp The membership degree of each fuzzy set is used to judge the degree of belonging to different fuzzy sets. The corresponding fault index membership formula is: ; (3) ; (4) ; (5) ; (6) The corresponding maximum power point voltage V mpp The membership formula is: ; (7) ; (8) ; (9) Finally, the output membership function type is determined according to the fault type, including four fault types: no fault (NF), open circuit fault (OCF), short circuit fault (SCF) and shadow shading (SF); Step 4: Based on the boundary range and quantity of the membership function and the fault characteristics of the photovoltaic array, define the "if-then" rules in fuzzy logic reasoning, as shown in the following table:

[0030] Step 5: Defuzzification is performed using the maximum membership method and output; first, the minimum value of the membership of each fuzzy subset is extracted from each fuzzy rule as the membership of the output fuzzy subset; then, the maximum membership is selected by comparing the size of the output membership corresponding to each rule. ; The output result is obtained by multiplying the membership degree by the midpoint of the fuzzy subset interval range corresponding to the membership degree. The fault type corresponding to the membership degree is the final derivation result: ; (10).

[0031] According to the failure index fi and V mpp The fuzzy output membership function values ​​obtained by different combinations can classify the operating status of the photovoltaic array. The following table gives the range of each fuzzy subset under different working conditions of the photovoltaic array.

[0032]

[0033] The output values ​​of the fuzzy algorithm for classifying the operating status of the photovoltaic array are shown in the table: Fault type Fuzzy results SP Configuration No error 11.3 Open circuit error 33.5 Short circuit error 50 Shadow Error 83.5 TCT Configuration No error 5.9 Open circuit error 39.2 Short circuit error 58.6 Shadow Error 88.3 The effect of faults in SP and TCT configurations is observed. In TCT configuration, open circuit faults lead to an uneven pattern of power reduction, while due to short circuits, a sharp drop in voltage can be observed. In addition, shadow faults lead to a sudden reduction in current and power, such as Figure 2As shown in Figure 2. In general, during the shadowing process, the TCT configuration is less affected by the shadow fault than the SP configuration. In the SP configuration, the short-circuit fault in the SP has less impact on the power and current values ​​than the open-circuit fault. However, the shadow fault shows uneven changes due to the sudden drop in the power and current values. At the open-circuit fault, the short-circuit current decreases sharply, while at the short-circuit fault, the open-circuit voltage drops sharply. The SP configuration is more susceptible to open-circuit faults than short-circuit faults, as shown in Figure 2. Figure 3 shown.

[0034] The present invention detects and classifies the most common faults occurring in photovoltaic arrays by using fuzzy logic algorithms. The required parameters along with the I-V trend line are obtained through the photovoltaic array and further used to calculate the defined parameter "Fault Index". Compared with other techniques, this method is efficient and easy to implement.

[0035] The above-described scheme is only a preferred implementation mode of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line, characterized in that: The following steps are involved: S1. In the series-parallel SP and TCT full cross-link configuration, different fault scenarios are set in the PV array through simulation experiments, and the corresponding open circuit voltage is recorded. V oc , short circuit current I sc , power generation P , to construct a dataset for training and validation; the data includes the I-V trend line and PV trend line collected under each fault scenario. The IV trend line refers to the linear trend of the IV curve approximately near the short-circuit current; S2, preprocessing the acquired data in the data set, and defining the fault index based on the PV array IV trend line and PV maximum power linear interpolation; S3. According to the change of IV trend line, the triangle membership function is selected to characterize the fault index. fi and maximum power point voltage V mpp Different value ranges; S4. Based on the boundary range and quantity of the membership function and the fault characteristics of the photovoltaic array, define "if-then" rules in fuzzy logic reasoning; S5. Use the maximum membership method to perform defuzzification processing and output.

2. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 1, characterized in that: The specific process of S3 is as follows: First determine the fault index and maximum power point voltage V mpp The fault index and the maximum power point voltage are used to construct the fuzzy input membership function; then, the fault index is converted into fi It is divided into four fuzzy sets corresponding to different fault severity levels: very low VL, low L, medium M and high H, the maximum power point voltage V mpp It is divided into three fuzzy sets corresponding to different voltage ranges: low L, medium M and high H. Finally, the left and right endpoints and the midpoint of each range are selected to represent the three endpoints of the triangular membership function to calculate the fault index. fi And the maximum power point voltage V mpp The membership degree of each fuzzy set is used to judge the degree of belonging to different fuzzy sets; the output membership function type is determined according to the fault type, including four fault types: no fault NF, open circuit fault OCF, short circuit fault SCF and shadow SF.

3. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 2, characterized in that: The "if-then" rules in S4 specifically include: Rule 1: If (fault index is medium) and (maximum power point voltage is high), then (PV fault is no fault) Rule 2: If (fault index is very low) and (maximum power point voltage is low), then (PV fault is an open circuit fault) Rule 3: If (fault index is high) and (maximum power point voltage is medium), then (PV fault is a short circuit fault) Rule 4: If (fault index is low) and (maximum power point voltage is high), then (the PV fault is a short circuit fault).

4. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 3, characterized in that: The specific process of S5 is as follows: First, the minimum value of the membership of each fuzzy subset is extracted from each fuzzy rule as the membership of the output fuzzy subset; then, the maximum membership is selected by comparing the size of the output membership corresponding to each rule. The fault type corresponding to the membership degree is the final derivation result.

5. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 4, characterized in that: The specific steps of S2 include: S2.

1. Take the voltage at the maximum power point as one of the parameters; S2.

2. Based on the IV trend line slope and PV maximum power linear interpolation calculation of the PV array under various fault conditions, a newly defined parameter, namely "fault index", is obtained; the trend line slope under different fault scenarios can be calculated by (1), and the fault index value can be calculated by (2): ；(1) ；(2) In the formula, is the instantaneous current change, is the total voltage change, is the failure index, is the slope of the trend line, The maximum power of the photovoltaic array.

6. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 5, characterized in that: In S3, the fault index fi The intervals of the four different fault severity levels ranged from very low VL [19.16, 49.93], low L [49.93, 73.66], medium M [73.66, 75.02], and high H [75.02, 130.82]; Maximum power point voltage V mpp The three different voltage intervals range: low L [178.54, 195.29], medium M [195.29, 211.34] and high H [211.34, 231.17].

7. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 6, characterized in that: In S3, the fault index is calculated fi And the maximum power point voltage V mpp The membership process belonging to each fuzzy set is: Failure Index fi The attribute formula is: ；(3) ；(4) ；(5) ； (6) The corresponding maximum power point voltage V mpp The membership formula is: ；(7) ；(8) ； (9) Finally, the output membership function type is determined according to the fault type, including four fault types: no fault NF, open circuit fault OCF, short circuit fault SCF and shadow SF.

8. The photovoltaic array fault detection and classification method based on fuzzy logic and IV trend line according to claim 7, characterized in that: In the above-mentioned S5, the calculation formula for the final derivation result is: ；（10）。

Images