Power generation anomaly diagnosis system suitable for the transformation of inefficient photovoltaic power plants

Abstract

The present invention relates to the technical field of photovoltaic power station diagnosis, and in particular to a power generation abnormality diagnosis system suitable for the transformation of inefficient photovoltaic power stations, comprising a power generation diagnosis center of a power station, a hierarchical fault location unit, a correlation analysis unit, a fault diagnosis unit and a diagnosis response unit; the present invention initially analyzes from the perspective of historical potential interference to reduce the impact of historical interference on current power generation, and further analyzes from the perspective of hierarchical fault location, dividing the fault diagnosis process of the inefficient photovoltaic power station into different levels, thereby improving the power generation fault diagnosis and location efficiency of the inefficient photovoltaic power station, and obtaining abnormal photovoltaic components based on the correlation analysis, and performing in-depth fault analysis on the abnormal photovoltaic components, so as to intuitively understand whether the abnormal photovoltaic components are caused by their own faults or other factors, thereby accurately locating the fault point, and facilitating in-depth excavation of the root cause of the fault, while determining whether the abnormal photovoltaic components are hot spot faults, thereby improving operation and maintenance efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station diagnosis, and in particular to a power generation anomaly diagnosis system suitable for reconstruction of inefficient photovoltaic power stations. Background Art

[0002] As the global demand for renewable energy continues to grow, photovoltaic power stations have been widely used as an important form of clean energy generation. However, in actual operation, many photovoltaic power stations have experienced low power generation efficiency due to equipment aging, environmental factors, improper installation, and other reasons. Transforming "inefficient photovoltaic power stations" to improve power generation efficiency has become an important issue in the new energy field.

[0003] However, existing power generation anomaly diagnosis systems still have the following shortcomings: They are difficult to analyze in conjunction with historical data, leading to potential risks affecting the power generation efficiency of inefficient PV power plants. Furthermore, they are unable to achieve hierarchical fault location at the "plant level, array level, and string level," often resulting in "false alarms of concentrated regional faults when they are actually single-point failures." Furthermore, they are unable to analyze component-level anomalies from a correlation perspective, making it difficult to determine whether component anomalies are single-point anomalies or overall anomalies, which in turn reduces operation and maintenance management efficiency.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a power generation abnormality diagnosis system suitable for the transformation of inefficient photovoltaic power stations to solve the technical defects mentioned above. The present invention initially analyzes from the perspective of historical potential interference to reduce the impact of historical interference on current power generation, and further analyzes from the perspective of hierarchical fault location, dividing the fault diagnosis process of inefficient photovoltaic power stations into different levels, and gradually narrowing the fault scope from macro to micro, so as to achieve the effect of improving the power generation fault diagnosis and positioning efficiency of inefficient photovoltaic power stations, and obtain abnormal photovoltaic components based on correlation analysis, and conduct in-depth fault analysis of the abnormal photovoltaic components, so as to intuitively understand whether the abnormal photovoltaic components are their own faults or the overall abnormality of the string photovoltaic components caused by other factors, and realize accurate positioning of the fault point, which is conducive to in-depth excavation of the root cause of the fault, and further determine whether the abnormal photovoltaic components are hot spot faults, so as to carry out targeted fault management of the abnormal photovoltaic components, so as to improve operation and maintenance efficiency and reduce operation and maintenance costs.

[0006] The object of the present invention can be achieved by the following technical solutions: a power generation anomaly diagnosis system suitable for the reconstruction of an inefficient photovoltaic power station, comprising a power station power generation diagnosis center, a hierarchical fault location unit, a correlation analysis unit, a fault diagnosis unit, and a diagnosis response unit;

[0007] The power generation diagnosis center of the power station is used to retrieve the operating data of the inefficient photovoltaic power station after transformation, and send the operating data to the layered fault location unit for layered power generation fault diagnosis and analysis to obtain a stable signal or a normal signal of the array or an abnormal signal of the string or a normal signal of the string;

[0008] When a string abnormality signal is generated, the correlation analysis unit is used to perform string-level correlation evaluation and analysis on the collected radiation and output power, and interactively analyze the obtained working condition stability signal, working condition risk signal, correlation signal, and correlation risk signal to obtain a correlation normal signal or a correlation abnormal signal;

[0009] When an associated abnormal signal is generated, the fault diagnosis unit is used to perform fault location and division feedback analysis on the output power characteristic curve of the abnormal photovoltaic component, and to perform discrimination processing on the obtained diagnostic coincidence to obtain a single-point fault signal or an overall fault signal. When a single-point fault signal is generated, the obtained distribution offset value is discriminated and processed to obtain a hot spot fault signal or a conventional signal.

[0010] Preferably, the hierarchical power generation fault diagnosis and analysis process is as follows:

[0011] The operating time period of the inefficient photovoltaic power station is collected and set as a time threshold. The operating data of the inefficient photovoltaic power station after transformation within the time threshold is obtained. The operating data includes solar radiation, output power, operating voltage and temperature of photovoltaic modules, and power generation efficiency.

[0012] Preprocess the collected operating data, including data cleaning, denoising, and normalization;

[0013] Based on the operating data, the power generation efficiency characteristic curve of the inefficient photovoltaic power station after transformation within the time threshold is obtained, and the power generation efficiency characteristic curve is compared and analyzed with the preset power generation efficiency characteristic curve. The difference value between the power generation efficiency characteristic curve and the preset power generation efficiency characteristic curve is set as the power generation offset, and the power generation offset is discriminated and processed to obtain a stable signal or a power generation abnormality signal.

[0014] Preferably, when an abnormal power generation signal is generated, the array power deviation rate of each photovoltaic component array after the transformation of the inefficient photovoltaic power station within the time threshold is obtained. The array power deviation rate represents the ratio between the actual power of the array and the average array power of the entire station, and the array power deviation rate is judged and processed. If the array power deviation rate is less than the preset array power deviation rate threshold, a normal array signal is generated; if the array power deviation rate is greater than or equal to the preset array power deviation rate threshold, an abnormal array signal is generated.

[0015] Preferably, when an array abnormality signal is generated, the current-voltage curve of each string of photovoltaic components in the photovoltaic component array corresponding to the array abnormality signal within the time threshold is obtained, the current-voltage curve is compared and analyzed with the standard current-voltage curve, the cosine similarity of the current-voltage curve and the standard current-voltage curve is obtained, and the cosine similarity is judged and processed. If the cosine similarity is less than the preset cosine similarity threshold, a string abnormality signal is generated. If the cosine similarity is greater than or equal to the preset cosine similarity threshold, a string normal signal is generated.

[0016] Preferably, the string-level correlation evaluation and analysis process is as follows: the string photovoltaic component corresponding to the string abnormal signal is set as an abnormal photovoltaic string, and the radiation Xi and output power Yi of each photovoltaic component in the i group of abnormal photovoltaic strings in the same period are obtained based on the operation data, i=1, 2, 3, ..., n, and the radiation mean XP and output power mean YP of each photovoltaic component in the abnormal photovoltaic string are obtained at the same time. The radiation Xi, output power Yi, radiation mean XP and output power mean YP are substituted into the Pearson correlation coefficient formula to calculate the data correlation coefficient R of the radiation X and output power Y of each photovoltaic component in the abnormal photovoltaic string, the absolute value of the data correlation coefficient R is set as the data correlation coefficient, and the data correlation coefficient is discriminated: if the data correlation coefficient is greater than the preset data correlation coefficient threshold, a working condition stability signal is generated; if the data correlation coefficient is less than or equal to the preset data correlation coefficient threshold, a working condition risk signal is generated.

[0017] Preferably, the operating voltage and temperature of each photovoltaic module in the i group of abnormal photovoltaic strings are obtained based on the operating data, and the pressure-temperature correlation coefficient of the operating voltage and temperature of each photovoltaic module in the abnormal photovoltaic string is obtained based on the Spearman rank correlation coefficient. If the pressure-temperature correlation coefficient is less than a preset pressure-temperature correlation coefficient threshold, an association signal is generated; if the pressure-temperature correlation coefficient is greater than or equal to the preset pressure-temperature correlation coefficient threshold, an association risk signal is generated.

[0018] Preferably, the fault location and classification feedback analysis process is as follows: the photovoltaic component corresponding to the associated abnormal signal is set as the abnormal photovoltaic component, the overlap between the output power characteristic curve of the abnormal photovoltaic component and the output power characteristic curve of the photovoltaic component in the same array is obtained, and the overlap between the output power characteristic curve of the abnormal photovoltaic component and the output power characteristic curve of the photovoltaic component in the same array is set as the diagnostic overlap, and the diagnostic overlap is judged and processed. If the diagnostic overlap is less than the preset diagnostic overlap threshold, a single point fault signal is generated; if the diagnostic overlap is greater than or equal to the preset diagnostic overlap threshold, an overall fault signal is generated.

[0019] Preferably, when a single-point fault signal is generated, a surface infrared thermal characteristic image of the abnormal photovoltaic component within the time threshold is further obtained, and the surface maximum temperature value and the surface minimum temperature value of the abnormal photovoltaic component are obtained based on the surface infrared thermal characteristic image. The difference between the surface maximum temperature value and the surface minimum temperature value is set as the distribution offset value, and the distribution offset value is discriminated and processed. If the distribution offset value is less than the preset distribution offset value threshold, a conventional signal is generated. If the distribution offset value is greater than or equal to the preset distribution offset value threshold, a hot spot fault signal is generated.

[0020] The beneficial effects of the present invention are as follows:

[0021] 1. This invention initially analyzes historical potential interference to reduce its impact on current power generation. It then conducts a hierarchical fault location analysis, dividing the fault diagnosis process for inefficient PV power plants into different levels. This gradually narrows the scope of faults from a macro to micro perspective, improving the efficiency of diagnosing and locating power generation faults in inefficient PV power plants. Furthermore, through information progression, it analyzes the correlation of PV panels at the string level, helping to identify any correlation anomalies among PV panels.

[0022] 2. The present invention obtains abnormal photovoltaic modules based on correlation analysis, and conducts in-depth fault analysis on the abnormal photovoltaic modules, so as to intuitively understand whether the abnormal photovoltaic modules are self-faulty or the overall abnormality of the photovoltaic modules in the string is caused by other factors, and accurately locate the fault point, which can reduce manpower input, avoid "over-maintenance" or "missed inspection", and help to achieve in-depth excavation of the root cause of the fault. At the same time, it further determines whether the abnormal photovoltaic modules are hot spot faults, so as to carry out targeted fault management of the abnormal photovoltaic modules, thereby improving operation and maintenance efficiency and reducing operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present invention will be further described below with reference to the accompanying drawings;

[0024] Figure 1 It is a flow chart of the system of the present invention;

[0025] Figure 2 This is a reference diagram for local analysis of Example 1 of the present invention. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0028] Example 1

[0029] See also Figures 1 to 2 As shown, the present invention is a power generation abnormality diagnosis system suitable for the transformation of low-efficiency photovoltaic power stations, including a power generation diagnosis center of a power station, a hierarchical fault location unit, a correlation analysis unit, a fault diagnosis unit and a diagnosis response unit. The power generation diagnosis center of the power station is communicatively connected to the hierarchical fault location unit, the correlation analysis unit, the fault diagnosis unit and the diagnosis response unit. The hierarchical fault location unit is in a one-way communication connection with the correlation analysis unit and the diagnosis response unit. The correlation analysis unit is in a one-way communication connection with the fault diagnosis unit and the diagnosis response unit. The fault diagnosis unit is in a one-way communication connection with the diagnosis response unit.

[0030] The power generation diagnosis center is used to retrieve the operating data of the inefficient photovoltaic power station after transformation, and send the operating data to the hierarchical fault location unit for hierarchical power generation fault diagnosis and analysis. That is, from the perspective of hierarchical fault location, the fault diagnosis process of the inefficient photovoltaic power station is divided into different levels, and the fault scope is gradually narrowed from macro to micro, thereby improving the efficiency of power generation fault diagnosis and location of the inefficient photovoltaic power station. The specific hierarchical power generation fault diagnosis and analysis process is as follows:

[0031] The operating time period of the inefficient photovoltaic power station is collected and set as a time threshold. The operating data and historical operating data of the inefficient photovoltaic power station within the time threshold are obtained. The operating data includes solar radiation, output power, operating voltage and temperature of photovoltaic modules, power generation efficiency, etc.

[0032] Preprocess the collected operation data and historical operation data, including data cleaning, denoising, normalization, etc.

[0033] Based on historical operating data, the operating quality coefficient of the inefficient photovoltaic power station is obtained and the operating quality coefficient is judged and processed. If the operating quality coefficient is equal to 1, a potential safety signal is generated. If the operating quality coefficient is not equal to 1, a potential risk signal is generated. The diagnostic response unit is used to respond to the potential risk signal and immediately display the preset warning text corresponding to the potential risk signal, so as to manage the various working equipment of the inefficient photovoltaic power station and reduce the impact of the working equipment on power generation;

[0034] Analysis process of the operation quality coefficient: call the preset state health scoring model of each working equipment of the inefficient photovoltaic power station, obtain the state health score of each working equipment of the inefficient photovoltaic power station based on the preset state health scoring model, the working equipment includes the combiner box, photovoltaic panels, etc., obtain the state health score of each working equipment within the historical m time threshold, where m is a natural number greater than zero, construct a state health score characteristic curve, obtain the change trend value of the state health score characteristic curve, and set it as the state health coefficient. The ratio of the number of working equipment corresponding to the state health coefficient greater than or equal to the preset state health coefficient threshold to the total number of working equipment is set as the operation quality coefficient;

[0035] When a potential safety signal is generated, a power generation efficiency characteristic curve of the transformed inefficient photovoltaic power station within a time threshold is obtained based on the operating data, and the power generation efficiency characteristic curve is compared and analyzed with a preset power generation efficiency characteristic curve. The difference between the power generation efficiency characteristic curve and the preset power generation efficiency characteristic curve is set as the power generation deviation, and the power generation deviation is judged and processed. If the power generation deviation is less than the preset power generation deviation threshold, a stable signal is generated; if the power generation deviation is greater than or equal to the preset power generation deviation threshold, a power generation abnormality signal is generated;

[0036] When an abnormal power generation signal is generated, the array power deviation rate of each photovoltaic module array after the transformation of the inefficient photovoltaic power station within the time threshold is obtained. The array power deviation rate represents the ratio between the actual array power and the average array power of the entire station, and the array power deviation rate is judged and processed. If the array power deviation rate is less than the preset array power deviation rate threshold, a normal array signal is generated; if the array power deviation rate is greater than or equal to the preset array power deviation rate threshold, an abnormal array signal is generated;

[0037] When an array abnormality signal is generated, the current-voltage curve of each string of photovoltaic modules in the photovoltaic module array corresponding to the array abnormality signal within the time threshold is obtained, the current-voltage curve is compared and analyzed with the standard current-voltage curve, the cosine similarity of the current-voltage curve and the standard current-voltage curve is obtained, and the cosine similarity is judged. If the cosine similarity is less than a preset cosine similarity threshold, a string abnormality signal is generated; if the cosine similarity is greater than or equal to the preset cosine similarity threshold, a string normal signal is generated;

[0038] The diagnostic response unit is used to respond to a stable signal or a normal array signal or a string abnormal signal or a normal string signal, and immediately display the preset warning text corresponding to the stable signal or the normal array signal or the string abnormal signal or the normal string signal. That is, from the perspective of hierarchical fault location, the fault diagnosis process of the inefficient photovoltaic power station is divided into different levels, and the fault scope is gradually narrowed from macro to micro, so as to achieve precise positioning of "power station level → array level → string level". Precise positioning of the fault point can reduce manpower input, avoid "over-maintenance" or "missed detection", help to achieve in-depth excavation of the root cause of the fault, and at the same time improve the efficiency of power generation fault diagnosis and positioning of inefficient photovoltaic power stations.

[0039] Example 2

[0040] When a string abnormality signal is generated, the correlation analysis unit is used to perform string-level correlation evaluation and analysis on the collected radiation and output power. This helps to understand whether there is any correlation abnormality in the PV modules and provides data support for subsequent analysis. The specific string-level correlation evaluation and analysis process is as follows:

[0041] The photovoltaic component of the string corresponding to the string abnormality signal is set as an abnormal photovoltaic string. Based on the operating data, the radiation amount Xi and output power Yi of each photovoltaic component in the i-th group of abnormal photovoltaic strings in the same period are obtained, where i=1, 2, 3, ..., n, and n is a natural number greater than zero. At the same time, the radiation amount mean XP and output power mean YP of each photovoltaic component in the abnormal photovoltaic string are obtained. The radiation amount Xi, output power Yi, radiation amount mean XP, and output power mean YP are substituted into the Pearson correlation coefficient formula to calculate the data correlation coefficient R of the radiation amount X and output power Y of each photovoltaic component in the abnormal photovoltaic string. The absolute value of the data correlation coefficient R is set as the data correlation coefficient, and the data correlation coefficient is discriminated: if the data correlation coefficient is greater than a preset data correlation coefficient threshold, a working condition stability signal is generated; if the data correlation coefficient is less than or equal to the preset data correlation coefficient threshold, a working condition risk signal is generated;

[0042] Among them, the Pearson correlation coefficient formula is: ;

[0043] At the same time, the operating voltage and temperature of each photovoltaic module in the i-group abnormal photovoltaic string are obtained based on the operating data, and the pressure-temperature correlation coefficient of the operating voltage and temperature of each photovoltaic module in the abnormal photovoltaic string is obtained based on the Spearman rank correlation coefficient. If the pressure-temperature correlation coefficient is less than a preset pressure-temperature correlation coefficient threshold, an association signal is generated; if the pressure-temperature correlation coefficient is greater than or equal to the preset pressure-temperature correlation coefficient threshold, an association risk signal is generated;

[0044] Interactive analysis of operating condition stability signals, operating condition risk signals, correlation signals, and correlation risk signals:

[0045] If the working condition stable signal and the associated signal are obtained, the associated normal signal is generated;

[0046] If an operating condition stability signal and an associated risk signal or an operating condition risk signal and an associated signal or an operating condition risk signal and an associated risk signal are obtained, a associated abnormality signal is generated;

[0047] The diagnostic response unit is used to respond to correlated normal signals or correlated abnormal signals, and immediately display the preset warning text corresponding to the correlated normal signal or correlated abnormal signal, so as to analyze the component-level fault accurately from the perspective of correlation, help to understand whether there is any correlation abnormality in the photovoltaic module, and provide data support for subsequent analysis;

[0048] When an associated abnormal signal is generated, the fault diagnosis unit is used to perform fault location and classification feedback analysis on the output power characteristic curve of the abnormal PV module, so as to intuitively understand whether the abnormal PV module is a fault of its own or other factors causing the overall abnormality of the string PV modules, so as to carry out targeted management. The specific fault location and classification feedback analysis process is as follows:

[0049] The photovoltaic module corresponding to the associated abnormal signal is set as the abnormal photovoltaic module, the overlap between the output power characteristic curve of the abnormal photovoltaic module and the output power characteristic curve of the photovoltaic modules in the same array is obtained, and the overlap between the output power characteristic curve of the abnormal photovoltaic module and the output power characteristic curve of the photovoltaic modules in the same array is set as the diagnostic overlap, and the diagnostic overlap is judged and processed. If the diagnostic overlap is less than a preset diagnostic overlap threshold, a single point fault signal is generated; if the diagnostic overlap is greater than or equal to the preset diagnostic overlap threshold, an overall fault signal is generated;

[0050] The diagnostic response unit is used to respond to single-point fault signals or overall fault signals, and immediately display the preset warning text corresponding to the single-point fault signal or the overall fault signal, so as to intuitively understand whether the abnormal photovoltaic module is a fault of the itself or other factors causing the overall abnormality of the photovoltaic modules in the string, so as to carry out targeted management and improve the power generation efficiency and operational stability of low-efficiency photovoltaic power stations;

[0051] When a single-point fault signal is generated, a surface infrared thermal characteristic image of the abnormal photovoltaic module within a time threshold is further obtained, and a surface maximum temperature value and a surface minimum temperature value of the abnormal photovoltaic module are obtained based on the surface infrared thermal characteristic image. The difference between the surface maximum temperature value and the surface minimum temperature value is set as a distribution offset value, and the distribution offset value is discriminated and processed. If the distribution offset value is less than a preset distribution offset value threshold, a normal signal is generated; if the distribution offset value is greater than or equal to the preset distribution offset value threshold, a hot spot fault signal is generated;

[0052] The diagnostic response unit is used to respond to hot spot fault signals or normal signals and immediately display the preset warning text corresponding to the hot spot fault signal or normal signal, so as to analyze whether the abnormal photovoltaic module is a hot spot fault based on a single point fault, so as to carry out targeted fault management of the abnormal photovoltaic module, thereby improving operation and maintenance efficiency and reducing operation and maintenance costs;

[0053] In summary, the present invention conducts preliminary analysis from the perspective of historical potential interference to reduce the impact of historical interference on current power generation, and further analyzes from the perspective of hierarchical fault location, dividing the fault diagnosis process of inefficient photovoltaic power stations into different levels, and gradually narrowing the fault scope from macro to micro, so as to achieve the effect of improving the power generation fault diagnosis and positioning efficiency of inefficient photovoltaic power stations. The correlation of the string-level photovoltaic components is further analyzed in an information progressive manner, which helps to understand whether there are correlation anomalies in the photovoltaic components. Based on the correlation analysis, the abnormal photovoltaic components are obtained, and the fault analysis of the abnormal photovoltaic components is carried out in depth, so as to intuitively understand whether the abnormal photovoltaic components are self-faults or other factors cause the overall abnormality of the string photovoltaic components, and accurately locate the fault point, which can reduce manpower investment, avoid "over-maintenance" or "missed detection", and help to achieve in-depth excavation of the root cause of the fault. At the same time, it is further determined whether the abnormal photovoltaic components are hot spot faults, so as to carry out targeted fault management of the abnormal photovoltaic components, so as to improve operation and maintenance efficiency and reduce operation and maintenance costs.

[0054] The threshold is set for result comparison and analysis in order to determine whether it is good or bad. The value of the threshold is set based on a combination of large-scale model analysis of sample data and manual experience to enter and store it. It can also be appropriately adjusted based on seasonal or common sense influencing conditions.

[0055] The size of the coefficient is to quantify each parameter to obtain a specific numerical value, which is convenient for subsequent comparison. The size of the coefficient depends on the amount of sample data and the preliminary setting of the corresponding operating coefficient for each set of sample data by technical personnel in this field; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0056] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A power generation abnormality diagnosis system suitable for the transformation of inefficient photovoltaic power stations, characterized by: It includes power plant power generation diagnosis center, hierarchical fault location unit, correlation analysis unit, fault diagnosis unit and diagnosis response unit; The power generation diagnosis center of the power station is used to retrieve the operating data of the inefficient photovoltaic power station after transformation, and send the operating data to the layered fault location unit for layered power generation fault diagnosis and analysis to obtain a stable signal or a normal signal of the array or an abnormal signal of the string or a normal signal of the string; When a string abnormality signal is generated, the correlation analysis unit is used to perform string-level correlation evaluation and analysis on the collected radiation and output power, and interactively analyze the obtained working condition stability signal, working condition risk signal, correlation signal, and correlation risk signal to obtain a correlation normal signal or a correlation abnormal signal; When an associated abnormal signal is generated, the fault diagnosis unit is used to perform fault location and classification feedback analysis on the output power characteristic curve of the abnormal photovoltaic module, and to perform discrimination processing on the obtained diagnostic coincidence to obtain a single-point fault signal or an overall fault signal. When a single-point fault signal is generated, the obtained distribution offset value is discriminated and processed to obtain a hot spot fault signal or a normal signal. The hierarchical power generation fault diagnosis and analysis process is as follows: The operating time period of the inefficient photovoltaic power station is collected and set as a time threshold. The operating data of the inefficient photovoltaic power station after transformation within the time threshold is obtained. The operating data includes solar radiation, output power, operating voltage and temperature of photovoltaic modules, and power generation efficiency. Based on historical operating data, the operating quality coefficient of the inefficient photovoltaic power station is obtained, and the operating quality coefficient is judged and processed to obtain potential safety signals or potential risk signals; Based on the operating data, a power generation efficiency characteristic curve of the transformed inefficient photovoltaic power station within a time threshold is obtained, and the power generation efficiency characteristic curve is compared and analyzed with a preset power generation efficiency characteristic curve. The difference between the power generation efficiency characteristic curve and the preset power generation efficiency characteristic curve is set as the power generation deviation, and the power generation deviation is discriminated and processed to obtain a stable signal or a power generation abnormality signal; When an abnormal power generation signal is generated, the array power deviation rate of each photovoltaic module array after the transformation of the inefficient photovoltaic power station within the time threshold is obtained. The array power deviation rate represents the ratio between the actual array power and the average array power of the entire station, and the array power deviation rate is judged and processed. If the array power deviation rate is less than the preset array power deviation rate threshold, a normal array signal is generated; if the array power deviation rate is greater than or equal to the preset array power deviation rate threshold, an abnormal array signal is generated; When an array abnormality signal is generated, the current-voltage curve of each string of photovoltaic modules in the photovoltaic module array corresponding to the array abnormality signal within the time threshold is obtained, the current-voltage curve is compared and analyzed with the standard current-voltage curve, the cosine similarity of the current-voltage curve and the standard current-voltage curve is obtained, and the cosine similarity is judged. If the cosine similarity is less than a preset cosine similarity threshold, a string abnormality signal is generated; if the cosine similarity is greater than or equal to the preset cosine similarity threshold, a string normal signal is generated; The fault location and classification feedback analysis process is as follows: the photovoltaic component corresponding to the associated abnormal signal is set as the abnormal photovoltaic component, the overlap between the output power characteristic curve of the abnormal photovoltaic component and the output power characteristic curve of the photovoltaic components in the same array is obtained, and the overlap between the output power characteristic curve of the abnormal photovoltaic component and the output power characteristic curve of the photovoltaic components in the same array is set as the diagnostic overlap, and the diagnostic overlap is judged and processed. If the diagnostic overlap is less than a preset diagnostic overlap threshold, a single point fault signal is generated; if the diagnostic overlap is greater than or equal to the preset diagnostic overlap threshold, an overall fault signal is generated; When a single-point fault signal is generated, a surface infrared thermal characteristic image of the abnormal photovoltaic component within the time threshold is further obtained, and the surface maximum temperature value and the surface minimum temperature value of the abnormal photovoltaic component are obtained based on the surface infrared thermal characteristic image. The difference between the surface maximum temperature value and the surface minimum temperature value is set as the distribution offset value, and the distribution offset value is discriminated and processed. If the distribution offset value is less than the preset distribution offset value threshold, a normal signal is generated. If the distribution offset value is greater than or equal to the preset distribution offset value threshold, a hot spot fault signal is generated.

2. The power generation abnormality diagnosis system suitable for the transformation of low-efficiency photovoltaic power stations according to claim 1 is characterized in that: The string-level correlation evaluation and analysis process is as follows: the string photovoltaic module corresponding to the string abnormal signal is set as an abnormal photovoltaic string, and the radiation amount Xi and output power Yi of each photovoltaic module in the i-th group of abnormal photovoltaic strings in the same period are obtained based on the operating data, where i=1, 2, 3, ..., n. At the same time, the radiation amount mean XP and output power mean YP of each photovoltaic module in the abnormal photovoltaic string are obtained. The radiation amount Xi, output power Yi, radiation amount mean XP, and output power mean YP are substituted into the Pearson correlation coefficient formula to calculate the data correlation coefficient R of the radiation amount X and output power Y of each photovoltaic module in the abnormal photovoltaic string. The absolute value of the data correlation coefficient R is set as the data correlation coefficient, and the data correlation coefficient is discriminated: if the data correlation coefficient is greater than a preset data correlation coefficient threshold, a working condition stability signal is generated; if the data correlation coefficient is less than or equal to the preset data correlation coefficient threshold, a working condition risk signal is generated.

3. The power generation abnormality diagnosis system suitable for the transformation of low-efficiency photovoltaic power stations according to claim 2 is characterized in that: Based on the operating data, the operating voltage and temperature of each photovoltaic module in the i-group abnormal photovoltaic string are obtained. Based on the Spearman rank correlation coefficient, the pressure-temperature correlation coefficient of the operating voltage and temperature of each photovoltaic module in the abnormal photovoltaic string is obtained. If the pressure-temperature correlation coefficient is less than the preset pressure-temperature correlation coefficient threshold, an association signal is generated; if the pressure-temperature correlation coefficient is greater than or equal to the preset pressure-temperature correlation coefficient threshold, an association risk signal is generated.

Images