Photovoltaic array fault detection and location method based on bipartite graph matching
By optimizing the voltage sensor configuration using a bipartite graph matching method, the constraints of sensor quantity and cost are resolved, enabling accurate detection and location of photovoltaic array faults. This reduces costs and simplifies operation, making it suitable for large-scale photovoltaic arrays.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SOUTH CHINA UNIV OF TECH
- Filing Date
- 2025-07-07
- Publication Date
- 2026-07-24
AI Technical Summary
In existing photovoltaic array fault detection methods, the number of sensors, the difficulty of configuration, and the cost are mutually restrictive, resulting in a mismatch between the accuracy, operability, and economy of detection and positioning, and making large-scale application complex.
A bipartite graph matching method is adopted to optimize the voltage sensor configuration. By establishing a bipartite graph of the photovoltaic array, edge matching schemes are selected and the edges are transformed into voltage sensors, thereby realizing the detection and location of photovoltaic module faults, reducing the number of sensors and simplifying the operation.
It enables accurate detection and location of photovoltaic array faults, reduces costs, simplifies operation, is applicable to large-scale photovoltaic arrays, and has versatility and high accuracy.
Smart Images
Figure CN120934456B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the technical field of photovoltaic array fault detection and location, and in particular to a photovoltaic array fault detection and location method based on bipartite graph matching. Background Technology
[0002] With the global green transformation of the energy structure, photovoltaic (PV) power generation, as a large-scale application of renewable energy, has seen its installed capacity grow rapidly. However, because PV arrays operate in complex outdoor environments, they are prone to faults such as open circuits, short circuits, aging, and shading, threatening the safe operation of PV power generation systems. Therefore, accurate detection and location of PV array faults are crucial for improving the operation and maintenance capabilities of PV power generation systems.
[0003] Because there is a significant difference between fault operation data and normal operation data of photovoltaic (PV) modules, fault detection and location can be effectively achieved by configuring sensors in the PV array to collect PV module operation data. However, existing methods typically increase the fault location accuracy by increasing the number of sensors, thereby increasing the sensor configuration cost. In addition, the large number of sensor configuration rules leads to complex wiring, making it difficult to apply to large-scale PV arrays.
[0004] Therefore, in existing photovoltaic array fault detection and location methods based on sensor configuration, the number of sensors, configuration difficulty, and configuration cost are mutually restrictive, resulting in a lack of good compatibility between the accuracy, operability, and economy of photovoltaic array fault detection and location. Summary of the Invention
[0005] The purpose of this invention is to overcome the shortcomings and deficiencies of existing technologies and provide a high-accuracy, low-cost photovoltaic array fault detection and location method based on bipartite graph matching. This method optimizes the configuration of voltage sensors in the photovoltaic array, enabling the detection and location of open-circuit, short-circuit, aging, and shading faults in photovoltaic modules, thus ensuring the safe operation of the photovoltaic array. Compared to traditional photovoltaic array fault detection and location methods, this invention uses fewer voltage sensors to achieve accurate detection and location of faults in each photovoltaic module, reducing costs. Furthermore, the optimized voltage sensor configuration scheme provided by this invention only requires connecting voltage sensors between adjacent photovoltaic module strings, making it easy to operate.
[0006] To achieve the above objectives, the technical solution provided by this invention is: a photovoltaic array fault detection and location method based on bipartite graph matching, comprising the following steps:
[0007] 1) Based on the connection structure of the photovoltaic array, a bipartite graph of the photovoltaic array is established, and the photovoltaic array is represented as a bipartite graph set composed of several sub-bipartite graphs. The component connection points on the same photovoltaic module string constitute an independent node set in the sub-bipartite graph.
[0008] 2) Under the constraints of the complete matching rule, the minimum edge coverage rule, and the non-connection of equal weight points in the bipartite graph, the edge matching scheme of the bipartite graph is selected, the edges are transformed into voltage sensors, and the voltage sensor optimization configuration scheme of the photovoltaic array is obtained.
[0009] 3) Using the output voltage at the maximum power point of the photovoltaic module and the open circuit voltage as the main fault characteristic quantities, calculate the fault voltage threshold when a single photovoltaic module at different locations experiences open circuit, short circuit, aging, or shading faults. Collect the measured voltage values of the photovoltaic module through voltage sensors, and compare the measured voltage values with the fault voltage thresholds to realize the detection and location of photovoltaic array faults.
[0010] Furthermore, in step 1), when the number of photovoltaic module strings is even, every two adjacent photovoltaic module strings form a sub-bipartite graph. Each sub-bipartite graph has two independent node sets, which are respectively composed of the component connection points on the two photovoltaic module strings. When the number of photovoltaic module strings is odd, three adjacent photovoltaic module strings form a special sub-bipartite graph, and the remaining two adjacent photovoltaic module strings form a sub-bipartite graph. The special sub-bipartite graph has three independent node sets, which are respectively composed of the component connection points on the three photovoltaic module strings.
[0011] Further, in step 2), the configuration of the voltage sensor of the photovoltaic array is equivalent to edge matching in a bipartite graph, where matching refers to a set of edges in the bipartite graph where no two edges share a common node; the complete matching rule means that each node in the graph is associated with an edge in the graph, and no two edges connect to the same node; the minimum edge coverage rule means that the fewest edges are used to cover all nodes, and each node is covered by only one edge; a weight is assigned to each node in the bipartite graph, and the connection point of two photovoltaic modules connected in series at the same potential is regarded as an equal-weight point in the bipartite graph, and a constraint is set that equal-weight points in the bipartite graph cannot be connected; in the edge matching filtered by the above complete matching rule, minimum edge coverage rule and the constraint that equal-weight points in the bipartite graph cannot be connected, the edges are converted into voltage sensors, and the optimized configuration scheme of the voltage sensor is obtained.
[0012] Further, in step 3), the voltage sensor obtains the measured voltage value by monitoring the voltage difference between non-equipotential connection points of adjacent photovoltaic module strings in the sub-bipart diagram of the same photovoltaic array. This measured voltage value changes with the voltage at the terminals of each photovoltaic module, thereby constructing a relationship model between the measured voltage value and the terminal voltage of the photovoltaic module, and then obtaining the photovoltaic module terminal voltage operating data through the measured voltage value. Assuming that only one photovoltaic module in the sub-bipart diagram of a photovoltaic array fails at the same time, and only one of the following faults occurs: open circuit, short circuit, aging, and shading, the voltage change patterns of the photovoltaic module terminals caused by different faults are as follows:
[0013] ① In a photovoltaic module string that has an open-circuit fault, the terminal voltage of the photovoltaic module is equal to the open-circuit voltage of the photovoltaic module;
[0014] ② The terminal voltage of the photovoltaic module that experiences a short circuit fault is 0;
[0015] ③ The terminal voltage of the photovoltaic module that has experienced aging failure is lower than the output voltage at the maximum power point of the photovoltaic module;
[0016] ④ The terminal voltage of the photovoltaic module experiencing a shading fault is lower than the open-circuit voltage of the photovoltaic module;
[0017] Using the output voltage and open-circuit voltage at the maximum power point of the photovoltaic module as the main fault characteristic quantities, and combining the voltage change pattern of the photovoltaic module terminals caused by the fault, the fault voltage threshold corresponding to different fault locations and fault types is calculated. The measured voltage value is compared with the fault voltage threshold. If the measured voltage value is within the fault voltage threshold range, the fault location of the photovoltaic array can be realized.
[0018] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0019] 1. This invention introduces the concept of bipartite graphs from graph theory for the first time, establishes a bipartite graph of photovoltaic arrays, and equates the sensor configuration of photovoltaic arrays to edge matching of bipartite graphs, simplifying the analysis process, providing a reliable theoretical method for sensor configuration of photovoltaic arrays, and is applicable to the analysis of sensor configuration of photovoltaic arrays of any scale, thus possessing universality.
[0020] 2. The optimized configuration scheme of voltage sensors for photovoltaic arrays provided by this invention can accurately detect and locate open circuit, short circuit, aging, and shading faults of each photovoltaic module while optimizing the number of voltage sensors. This ensures the accuracy of fault detection and location of the photovoltaic array, reduces costs, and the voltage sensors only need to be connected between adjacent photovoltaic module strings, making operation simple and easy to apply to large-scale photovoltaic arrays. Attached Figure Description
[0021] Figure 1 This is a flowchart illustrating the implementation of the method of the present invention.
[0022] Figure 2 This is a diagram showing the connection structure of a photovoltaic array.
[0023] Figures 3a-3b This is a schematic diagram of a bipartite graph of a photovoltaic array.
[0024] Figures 4a-4c Diagram showing an optimized configuration scheme for voltage sensors. Detailed Implementation
[0025] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.
[0026] like Figures 1 to 4c As shown in the figure, this embodiment discloses a photovoltaic array fault detection and localization method based on bipartite graph matching, the specific details of which are as follows:
[0027] 1) Based on the connection structure of the photovoltaic array, a bipartite graph of the photovoltaic array is established, and the photovoltaic array is represented as a bipartite graph set composed of several sub-bipartite graphs. The component connection points on the same photovoltaic module string constitute an independent node set in the sub-bipartite graph.
[0028] The connection structure of the photovoltaic array provided in this embodiment is as follows: Figure 2 As shown, there are m photovoltaic module strings, each containing n+1 photovoltaic modules and n module connection points. The j-th photovoltaic module in the i-th string is defined as a. i,j .
[0029] A photovoltaic array can be represented as a set of bipartite graphs consisting of several sub-bipartite graphs, where the connection points of the same photovoltaic module string are considered as an independent set of nodes in the sub-bipartite graph. Based on the parity of the number of photovoltaic module strings m, the bipartite graph of the photovoltaic array can be represented as follows:
[0030] ① When the number of photovoltaic module strings m is even, the bipartite graph of the photovoltaic array is as follows: Figure 3a As shown, every two adjacent photovoltaic (PV) module strings form a sub-bipartite graph, and the connection points of the modules on the same PV module string form an independent node set. Each sub-bipartite graph contains two independent node sets. In this case, a PV array containing m PV module strings can be represented as a bipartite graph set consisting of m / 2 sub-bipartite graphs.
[0031] ② When the number of photovoltaic module strings m is odd, the bipartite graph of the photovoltaic array is as follows: Figure 3bAs shown, three adjacent photovoltaic (PV) module strings are considered as a special sub-bipartite graph, and each of the remaining two adjacent PV module strings forms a sub-bipartite graph. The connection points of the modules on the same PV module string form an independent set of nodes. Each sub-bipartite graph has two independent sets of nodes; in the special sub-bipartite graph formed by three adjacent PV module strings, there are three independent sets of nodes. In this case, a PV array containing m PV module strings can be represented as a set of bipartite graphs consisting of (m-3) / 2 sub-bipartite graphs and 1 special sub-bipartite graph.
[0032] 2) Under the constraints of the complete matching rule, the minimum edge coverage rule, and the non-connected equal weight points in the bipartite graph, the edge matching scheme of the bipartite graph is selected, the edges are transformed into voltage sensors, and the voltage sensor optimization configuration scheme of the photovoltaic array is obtained.
[0033] To detect and locate faults in photovoltaic (PV) arrays, voltage sensors can be used to collect the terminal voltage data of the PV modules within the array. Traditional voltage sensor configurations require a large number of sensors to detect and locate PV array faults, resulting in complex configurations and significant costs. However, reducing the number of voltage sensors may not accurately obtain the terminal voltage data for each PV module. Therefore, an optimized voltage sensor configuration scheme must meet the following conditions:
[0034] ① Each component connection point in the photovoltaic array needs to be covered by a voltage sensor to obtain the terminal voltage data of each photovoltaic module;
[0035] ② The two ends of the voltage sensor need to be connected to the connection points of the two photovoltaic module strings at different potentials to ensure the validity of the acquired voltage data;
[0036] ③ Each component connection point does not need to repeatedly connect to the voltage sensor, thereby reducing the redundancy of the voltage sensor configuration.
[0037] In a bipartite graph, a matching edge is a set of edges where no two edges share a common node. Therefore, edge matching in a bipartite graph can simplify the analysis process for configuring voltage sensors in a photovoltaic array. Furthermore, the complete matching rule and the minimum edge coverage rule of a bipartite graph are introduced as constraints, while the constraint that equally weighted points in a bipartite graph cannot be connected is set to achieve optimized voltage sensor configuration. Specific constraints and operations are as follows:
[0038] ① The complete matching rule of a bipartite graph means that every node in the graph is associated with some edge in the graph, and no two edges connect to the same node. The voltage sensor configuration obtained under the constraint of the complete matching rule of a bipartite graph can realize the detection and location of faults in each photovoltaic module;
[0039] ② Assign weights to each node in the bipartite graph, set the connection points of the equipotential components on the two photovoltaic module strings as the equiweighted points in the bipartite graph, and set the constraint that the equiweighted points in the bipartite graph cannot be connected. Under this constraint, the voltage sensor can be connected to the connection points of the components at different potentials on the two photovoltaic module strings.
[0040] ③ The minimum edge coverage rule for bipartite graphs refers to covering all nodes with the fewest possible edges, and each node is covered by only one edge. The voltage sensor configuration obtained under the constraint of the minimum edge coverage rule for bipartite graphs greatly reduces the number of voltage sensors required without reducing positioning accuracy.
[0041] When the number of photovoltaic module strings m is even, the photovoltaic array bipartite graph model includes m / 2 sub-bipartite graphs. Each sub-bipartite graph can be filtered for edge matching schemes through the above constraints. By converting the edges into voltage sensors, an optimized voltage sensor configuration scheme can be obtained, such as... Figure 4a As shown, only mn / 2 voltage sensors are needed.
[0042] When the number of photovoltaic module strings m is odd, the photovoltaic array bipartite graph model includes (m-3) / 2 sub-bipartite graphs and 1 special sub-bipartite graph. In the (m-3) / 2 sub-bipartite graphs, edge matching schemes can be selected through the constraints mentioned above. Furthermore, in the special sub-bipartite graph, the connection points of each row of equipotential modules (i.e., equal-weighted points) can be treated as an independent node set, further dividing the sub-bipartite graph to obtain edge matching schemes. Therefore, based on the parity of the module connection points n, it can be divided into two cases for operation:
[0043] When the number of component connection points n is even, the special sub-bipartite graph is horizontally divided into n / 2 sub-bipartite graphs. Each sub-bipartite graph can also be filtered for edge matching schemes using the above constraints. Converting the edges into voltage sensors yields optimized voltage sensor configuration schemes, such as... Figure 4b As shown, only mn / 2 voltage sensors are needed.
[0044] When the number of component connection points n is odd, the special sub-bipartite graph is horizontally divided into (n-3) / 2 sub-bipartite graphs and 1 special sub-bipartite graph. Each independent node set of this special sub-bipartite graph contains only 3 nodes, making it an undivisible minimum special sub-bipartite graph. By filtering the sub-bipartite graphs and the minimum special sub-bipartite graph using the above constraints to find edge matching schemes, and converting the edges into voltage sensors, an optimized voltage sensor configuration scheme can be obtained, as shown below. Figure 4c As shown, only A voltage sensor.
[0045] The voltage sensor optimization configuration scheme provided by this invention uses as few voltage sensors as possible to achieve the positioning accuracy of a single photovoltaic module, ensuring the detection and location of faults in each photovoltaic module in the photovoltaic array. Moreover, the voltage sensors only need to be connected between adjacent photovoltaic module strings, making the operation simple.
[0046] When the photovoltaic array has m photovoltaic module strings, n+1 photovoltaic modules, and n module connection points, the voltage sensor optimized configuration scheme provided by this invention requires at most only A voltage sensor is compared with a traditional voltage sensor configuration scheme, as shown in Table 1.
[0047] Table 1 Comparison of Optimized Configuration Schemes for Voltage Sensors
[0048]
[0049] Among the proposed solutions, Scheme 1 involves configuring a voltage sensor for each photovoltaic module; Scheme 2 involves configuring a voltage sensor every other photovoltaic module; and Scheme 3 involves configuring a voltage sensor between the connection points of two modules in adjacent photovoltaic module strings. As shown in Table 1, the number of voltage sensors required by this invention is consistently less than that of Scheme 1, less than that of Scheme 2 when n≥2, and less than that of Scheme 3 when m≥2. Therefore, the optimized voltage sensor configuration scheme provided by this invention significantly reduces the number of voltage sensors and lowers costs while ensuring that faults in each photovoltaic module can be located.
[0050] 3) Using the output voltage at the maximum power point of the photovoltaic module and the open circuit voltage as the main fault characteristic quantities, calculate the fault voltage threshold when a single photovoltaic module at different locations experiences open circuit, short circuit, aging, or shading faults. Collect the measured voltage values of the photovoltaic module through voltage sensors, and compare the measured voltage values with the fault voltage thresholds to realize the detection and location of photovoltaic array faults.
[0051] for Figure 2 The photovoltaic array shown has m photovoltaic module strings, n+1 photovoltaic modules, and n module connection points. The output voltage at the maximum power point of the photovoltaic modules is V. m The open-circuit voltage of the photovoltaic module is V. oc Under normal operation, the total voltage of each photovoltaic module string is expressed as V = V m (n+1), the i-th and j-th photovoltaic modules a i,j With the bottom photovoltaic module a i,1 The sum of the terminal voltages of the photovoltaic modules is expressed as V. ij .
[0052] The invention established such as Figures 3a-3bThe illustrated bipartite diagram of a photovoltaic array assumes that at any given moment, only one photovoltaic module in one string of a sub-array within the array experiences a fault, specifically one of the following fault types: open circuit, short circuit, aging, or shading. Different fault locations and types will cause changes in the terminal voltage of the faulty photovoltaic module compared to its normal operating state, thus affecting V. ij The value. According to... Figures 4a-4c The voltage sensor optimization configuration scheme shown assumes that the i-th string and the i'-th string belong to the same photovoltaic array sub-bipartite graph, where the component connection points connected to each voltage sensor are j and j', respectively, and the measured voltage value collected by the voltage sensor is V. jj' Establish V jj' The relationship model between the photovoltaic module terminal voltage and the voltage is as follows:
[0053] V jj' =V i'j' -V ij (1)
[0054] In equation (1), V ij For photovoltaic module a i,j With a i,1 The sum of the terminal voltages of the photovoltaic modules, V i'j' For photovoltaic module a i',j' With a i',1 The sum of the terminal voltages of the photovoltaic modules in the i-th string. When a photovoltaic module in the i-th string fails, the i'-th string is a non-faulty string, V i'j' = [j' / (n+1)]V. According to equation (1), the measured voltage value V can be obtained through the voltage sensor. jj' The system acquires the terminal voltage operating data of the photovoltaic module.
[0055] Therefore, using the output voltage and open-circuit voltage at the maximum power point of the photovoltaic module as the main fault characteristic quantities, and combining the voltage change pattern of the photovoltaic module terminals caused by the fault, the fault voltage threshold corresponding to different fault locations and fault types is calculated, and then the measured voltage value V is... jj' By comparing with the fault voltage threshold, fault detection and location can be achieved.
[0056] ① Open circuit fault
[0057] The terminal voltage of a photovoltaic module string experiencing an open-circuit fault is equal to V. oc When the voltage sensor measures V jj' When the fault voltage threshold shown in equation (2) is equal to the fault voltage threshold, the open-circuit fault is located in photovoltaic module a. i,j With a i,1 between:
[0058]
[0059] When the voltage sensor measures V jj'When the fault voltage threshold shown in equation (3) is equal to the fault voltage threshold, the open-circuit fault is located in photovoltaic module a. i,j+1 With a i,n+1 between:
[0060]
[0061] ② Short circuit fault
[0062] The voltage at the photovoltaic module terminals where a short circuit fault has occurred is 0. When the voltage sensor measures V... jj' When the fault voltage threshold shown in equation (4) is equal to the fault voltage threshold, the short-circuit fault is located in photovoltaic module a. i,j With a i,1 between:
[0063]
[0064] When the voltage sensor measures V jj' When the fault voltage threshold shown in equation (5) is equal to the fault voltage threshold, the short-circuit fault is located in photovoltaic module a. i,j+1 With a i,n+1 between:
[0065]
[0066] ③Aging fault
[0067] The terminal voltage of a photovoltaic module experiencing aging failure is lower than the output voltage V at the maximum power point of the photovoltaic module. m When the voltage sensor measures V jj' When the fault voltage threshold range shown in equation (6) is within the range, the aging fault is located in photovoltaic module a. i,j With a i,1 between:
[0068]
[0069] When the voltage sensor measures V jj' When the fault voltage threshold range shown in equation (7) is within the range, the aging fault is located in photovoltaic module a. i,j+1 With a i,n+1 between:
[0070]
[0071] ④ Obstruction fault
[0072] The terminal voltage of the photovoltaic module experiencing a shading fault is less than the open-circuit voltage V. oc When the voltage sensor measures V jj' When the fault voltage threshold range shown in equation (8) is within the range, the shading fault is located in photovoltaic module a. i,j With a i,1 between:
[0073]
[0074] When the voltage sensor measures V jj' When the fault voltage threshold range shown in equation (9) is within the range, the shading fault is located in photovoltaic module a. i,j+1 With a i,n+1 between:
[0075]
[0076] According to equations (1)-(9), in the voltage sensor optimization configuration scheme provided by the present invention, the measured voltage value of the photovoltaic module collected by the voltage sensor can significantly distinguish the fault types of open circuit, short circuit, aging, and shading, and realize the location of the fault.
[0077] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A photovoltaic array fault detection and localization method based on bipartite graph matching, characterized in that, Includes the following steps: 1) Based on the connection structure of the photovoltaic array, a bipartite graph of the photovoltaic array is established, and the photovoltaic array is represented as a bipartite graph set composed of several sub-bipartite graphs. The component connection points on the same photovoltaic module string constitute an independent node set in the sub-bipartite graph. 2) Under the constraints of the complete matching rule, the minimum edge coverage rule, and the non-connected equal weight points in the bipartite graph, the edge matching scheme of the bipartite graph is selected, the edges are transformed into voltage sensors, and the voltage sensor optimization configuration scheme of the photovoltaic array is obtained. The configuration of the voltage sensor in the photovoltaic array is equivalent to edge matching in a bipartite graph. Matching refers to a set of edges in the bipartite graph where no two edges share a common node. The complete matching rule means that every node in the graph is associated with an edge, and no two edges connect to the same node. The minimum edge coverage rule means that all nodes are covered by the fewest possible edges, and each node is covered by only one edge. A weight is assigned to each node in the bipartite graph. The connection point of two photovoltaic modules connected in series at the same potential is considered an equal-weight point in the bipartite graph, and a constraint is set that equal-weight points in the bipartite graph cannot be connected. From the edge matching selected by the complete matching rule, the minimum edge coverage rule, and the constraint that equal-weight points in the bipartite graph cannot be connected, the edges are converted into voltage sensors, thus obtaining the optimized configuration scheme for the voltage sensors. 3) Using the output voltage at the maximum power point of the photovoltaic module and the open circuit voltage as the main fault characteristic quantities, calculate the fault voltage threshold when a single photovoltaic module at different locations experiences open circuit, short circuit, aging, or shading faults. Collect the measured voltage values of the photovoltaic module through voltage sensors, and compare the measured voltage values with the fault voltage thresholds to realize the detection and location of photovoltaic array faults.
2. The photovoltaic array fault detection and location method based on bipartite graph matching according to claim 1, characterized in that, In step 1), when the number of photovoltaic module strings is even, every two adjacent photovoltaic module strings form a sub-bipartite graph. Each sub-bipartite graph has two independent node sets, which are respectively composed of the component connection points on the two photovoltaic module strings. When the number of photovoltaic module strings is odd, three adjacent photovoltaic module strings form a special sub-bipartite graph, and the remaining two adjacent photovoltaic module strings form a sub-bipartite graph. The special sub-bipartite graph has three independent node sets, which are respectively composed of the component connection points on the three photovoltaic module strings.
3. The photovoltaic array fault detection and location method based on bipartite graph matching according to claim 2, characterized in that, In step 3), the voltage sensor obtains the measured voltage value by monitoring the voltage difference between non-equipotential connection points of adjacent photovoltaic module strings in the sub-bipart diagram of the same photovoltaic array. This measured voltage value changes with the voltage at the terminals of each photovoltaic module, thereby constructing a relationship model between the measured voltage value and the terminal voltage of the photovoltaic module, and then obtaining the photovoltaic module terminal voltage operating data through the measured voltage value. Assuming that only one photovoltaic module in the sub-bipart diagram of a photovoltaic array fails at the same time, and only one of the following faults occurs: open circuit, short circuit, aging, and shading, the voltage change patterns of the photovoltaic module terminals caused by different faults are as follows: In a photovoltaic module string that experiences an open-circuit fault, the terminal voltage of the photovoltaic module is equal to the open-circuit voltage of the photovoltaic module; The terminal voltage of the photovoltaic module that has experienced a short circuit fault is 0. The terminal voltage of a photovoltaic module that has experienced aging failure is lower than the output voltage at the maximum power point of the photovoltaic module. The terminal voltage of the photovoltaic module experiencing a shading fault is lower than the open-circuit voltage of the photovoltaic module; Using the output voltage and open-circuit voltage at the maximum power point of the photovoltaic module as the main fault characteristic quantities, and combining the voltage change pattern of the photovoltaic module terminals caused by the fault, the fault voltage threshold corresponding to different fault locations and fault types is calculated. The measured voltage value is compared with the fault voltage threshold. If the measured voltage value is within the fault voltage threshold range, the fault location of the photovoltaic array can be realized.