A photovoltaic power station intelligent inspection method

Abstract

The present application relates to the technical field of photovoltaic monitoring, and discloses a method for intelligent inspection of photovoltaic power stations, including: obtaining an environmental data set and a power generation data set corresponding to each pre-divided photovoltaic area; integrating time period information into an effective working time period, and integrating each power generation data set into a photovoltaic data segment; determining standard data for each photovoltaic area, and determining doubtful data in the photovoltaic data segment of the corresponding photovoltaic area based on the standard data; eliminating abnormal weather data in the doubtful data to convert the doubtful data into abnormal data; determining the inspection priority of the corresponding photovoltaic area based on the abnormal data; and determining the inspection area of ​​the current round based on the inspection priority. The data anomalies of each photovoltaic area are analyzed in combination with the environmental data and the power generation data, and then the inspection priority is determined based on the data anomalies. In the case of limited manpower and material resources, priority is given to inspecting the photovoltaic areas that are more in need, so that faults can be discovered in a timely manner.

Description

Technical Field

[0001] The present application relates to the field of photovoltaic monitoring, and in particular to an intelligent inspection method for photovoltaic power stations. Background Art

[0002] As the global demand for renewable energy continues to grow, photovoltaic power generation, as a clean and renewable energy form, has received widespread attention and application. However, due to the complexity of photovoltaic power generation systems and the diversity of operating environments, various failures are inevitable in the long-term use of photovoltaic equipment, affecting the stability of the system and power generation efficiency. Therefore, inspection of photovoltaic equipment is essential. And with the development of technology, most photovoltaic power stations currently use drones for inspection.

[0003] Due to the limitation of cost and the endurance of drones, power stations divide the areas where photovoltaic systems are deployed into zones and use drones to inspect them in a certain order. However, photovoltaic systems often cover a large area and are divided into many zones, which means that it takes a long time to rotate all zones. This may cause the photovoltaic equipment in a certain area to have a hidden fault and it will take a long time for drones to rotate to that area, causing the hidden danger to become a real fault. Summary of the invention

[0004] In order to improve the timeliness of fault discovery, the present application provides a photovoltaic power station intelligent inspection method.

[0005] This application provides a photovoltaic power station intelligent inspection method, which adopts the following technical solutions:

[0006] A photovoltaic power station intelligent inspection method comprises the following steps:

[0007] At the end of each round of inspection, the environmental data set collected during the round of inspection and the power generation data set corresponding to each pre-divided photovoltaic area are obtained, wherein the environmental data set includes meteorological data and time period information corresponding to the meteorological data;

[0008] Integrate the time period information into effective working time periods, and integrate each power generation data set into photovoltaic data segments according to the effective working time periods;

[0009] Determine standard data of each photovoltaic area according to the environmental data set, and determine doubtful data in the photovoltaic data segment of the corresponding photovoltaic area according to the standard data; wherein the doubtful data includes abnormal equipment data and abnormal weather data;

[0010] Eliminate abnormal weather data in questionable data according to meteorological data to convert the questionable data into abnormal data;

[0011] Determine the inspection priority of the corresponding photovoltaic area based on abnormal data;

[0012] The inspection area of ​​the current round is determined based on the inspection priority.

[0013] Furthermore, the step of integrating the time period information into the effective working time period includes the following steps:

[0014] Dividing the time period information into a plurality of theoretical working time periods and theoretical non-working time periods according to the first preset time and the second preset time;

[0015] It is determined whether the meteorological data corresponding to each theoretical working time period contains specific weather, where the specific weather is suitable for the photovoltaic system to work.

[0016] If so, the time period corresponding to the specific weather is extracted as the valid period, and all valid periods are sorted in chronological order to form a valid working time period.

[0017] Further, determining the standard data of each photovoltaic area according to the environmental data set includes the following steps:

[0018] Obtain the pre-stored equipment parameters and corresponding storage time of each photovoltaic area;

[0019] Determine whether the storage time is within the preset time.

[0020] If so, standard data is simulated based on the corresponding equipment parameters and meteorological data;

[0021] If not, then the corresponding historical data is matched as standard data from the corresponding relationship between the preset meteorological data and the historical data according to the meteorological data.

[0022] Further, the determining of the doubtful data in the photovoltaic data segment corresponding to the photovoltaic area according to the standard data comprises the following steps:

[0023] Determine an adjustment coefficient according to the meteorological data corresponding to each valid segment, and adjust the standard data according to the adjustment coefficient to form a standard interval corresponding to the valid segment;

[0024] Determine whether the photovoltaic data corresponding to each valid segment is within the corresponding standard range.

[0025] If so, the corresponding data is normal data;

[0026] If not, the photovoltaic data outside the standard range are extracted as questionable data.

[0027] Furthermore, the method of removing abnormal weather data according to meteorological data comprises the following steps:

[0028] Determine the regional wind direction according to meteorological data, and combine several photovoltaic areas in the same wind direction as a regional combination according to the regional wind direction, and record the order in which the multiple photovoltaic areas in the regional combination are arranged in sequence according to the wind direction as the first sorting;

[0029] Screening out suspicious data with high waveform similarity and amplitude difference within a preset threshold between different photovoltaic areas in the same area combination as a comparison group, and sorting the corresponding photovoltaic areas according to the order of recording time of the suspicious data in the comparison group to obtain a second sorting;

[0030] Determine whether the second sorting is consistent with the first sorting of the corresponding region combination,

[0031] If so, the questionable data in the control group will be regarded as abnormal weather data.

[0032] Furthermore, combining a plurality of photovoltaic areas in the same wind direction as a regional combination according to the regional wind direction comprises the following steps:

[0033] According to the wind direction, several photovoltaic areas located in the edge area and facing the wind first are determined as the first areas.

[0034] The wind path is formed according to the windward width perpendicular to the wind direction in each first-place area.

[0035] Calculate the windward area of ​​each photovoltaic area on the windward path.

[0036] Determine whether the windward area reaches more than half of the total area of ​​the corresponding photovoltaic area.

[0037] If so, the PV region is included in the same region combination as the first region.

[0038] Further, determining the inspection priority of the corresponding photovoltaic area according to the abnormal data includes the following steps:

[0039] Determine a first weight value according to the proportion and amplitude of the abnormal data in the corresponding photovoltaic data segment, wherein the greater the proportion or the greater the amplitude of the abnormal data, the higher the first weight value;

[0040] Determine a second weight value according to the pre-stored historical inspection time, wherein the earlier the historical inspection time is, the higher the second weight value is;

[0041] The inspection priority is determined according to the first weight value and the second weight value of the same photovoltaic area.

[0042] Further, determining the first weight value according to the proportion and amplitude of the abnormal data in the corresponding photovoltaic data segment includes the following steps:

[0043] It is determined whether the maximum value exceeds a preset first threshold and whether the proportion exceeds a preset second threshold.

[0044] If any one of them exceeds the limit, the highest inspection priority will be given;

[0045] If neither is exceeded, the first weight value is determined according to a first ratio of the maximum amplitude to the first threshold and a second ratio of the proportion to the second threshold.

[0046] In summary, the present application includes at least one of the following beneficial technical effects:

[0047] 1. Analyze the data anomalies of each photovoltaic area by combining environmental data and power generation data, and then determine the inspection priority based on the data anomalies. With limited manpower and material resources, give priority to inspecting the photovoltaic areas that are more in need, so that faults can be discovered in time;

[0048] 2. Before determining the inspection priority based on data anomalies, identify and exclude abnormal data caused by weather conditions to improve data accuracy and make the final inspection priority more accurate;

[0049] 3. When determining the inspection priority of the PV area, the time of historical inspections is also taken into account, so that each PV area has the opportunity to be inspected in time to update the corresponding equipment parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flowchart of a photovoltaic power station intelligent inspection method in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following is combined with the accompanying drawings Figure 1 , further details of this application are given.

[0052] The present application discloses a method for intelligent inspection of photovoltaic power stations. Figure 1 , including the following steps:

[0053] S100. When a round of inspection is completed, an environmental data set collected during the inspection and a power generation data set corresponding to each pre-divided photovoltaic area are obtained.

[0054] The environmental data set includes meteorological data and time period information corresponding to the meteorological data.

[0055] Meteorological data can be automatically captured from external meteorological platforms or manually entered by staff on duty at the location of the photovoltaic system. Meteorological data includes multiple weather information sorted in chronological order. Weather information includes rainy days, sunny days, cloudy days, overcast days, etc. Weather changes can be learned based on meteorological data.

[0056] The power generation data set is a collection of all data automatically recorded by the photovoltaic power station during the operation of the photovoltaic system.

[0057] The time period information includes a plurality of continuous weather time periods, each weather time period corresponds to a type of weather information, that is, the duration of different weather information can be known according to the time period information.

[0058] The starting time of all data obtained in this step is the start time of the corresponding round of inspection, and the end time is the end time of the corresponding round of inspection. That is, it is confirmed that the data relied on by the new round of inspection are all the data generated during the previous round of inspection to ensure the timeliness of the data.

[0059] S200, integrating time period information into effective working time periods, and integrating each power generation data set into photovoltaic data segments according to the effective working time periods.

[0060] The photovoltaic system does not work at night, on rainy days, snowy days, etc., and the data recorded during non-working hours is useless for subsequent analysis and will result in slow data analysis efficiency.

[0061] Therefore, it is necessary to clean up useless data first: first, screen and integrate the time periods suitable for the operation of the photovoltaic system according to the meteorological data to form an effective working section, and then retrieve the photovoltaic data recorded at the corresponding time from the power generation data set according to the effective working section to integrate it into a photovoltaic data segment.

[0062] The photovoltaic data in the power generation data set corresponds to the timestamp when the data is generated or recorded. The photovoltaic data with the timestamp falling within the effective working segment is extracted and then integrated into photovoltaic data segments in chronological order.

[0063] The data in the photovoltaic data segment still retains the corresponding relationship with time. However, since the effective working segment is composed of multiple working segments, the photovoltaic data segment obtained according to the effective working segment is also composed of multiple data segments, and the time in the same data segment is continuous, it is only necessary to retain the timestamps corresponding to the beginning data and the end data of the same data segment, thereby reducing the number of timestamps required for the photovoltaic data segment.

[0064] S300, determining standard data of each photovoltaic area according to the environmental data set, and determining questionable data in the photovoltaic data segment corresponding to the photovoltaic area according to the standard data.

[0065] Questionable data includes abnormal equipment data and abnormal weather data.

[0066] Standard data refers to the photovoltaic data that should be generated in the corresponding photovoltaic area in theory.

[0067] The best form of determining standard data is that one PV module corresponds to one standard data, but considering that the PV area is a relatively small area formed by pre-division, the same PV area is benchmarked with the same standard data, which can effectively reduce the amount of data calculation. And because there is a certain area span between different PV areas, in order to improve data accuracy, different PV areas need to determine their corresponding standard data.

[0068] Questionable data refers to data in the photovoltaic data segment whose values ​​are different from the standard data.

[0069] S400: Eliminate abnormal weather data in questionable data according to meteorological data to convert the questionable data into abnormal data.

[0070] Equipment abnormal data refers to questionable data caused by factors such as equipment failure. Weather abnormal data refers to questionable data caused by factors such as deteriorating weather. That is, equipment failure and deteriorating weather will cause the photovoltaic system to generate questionable data.

[0071] However, since the weather changes over time, the photovoltaic system is prone to abnormal weather data when the weather is bad, and can regenerate photovoltaic data that meets the standards after the weather recovers, without the need for staff to inspect the photovoltaic system. When abnormal equipment data is generated, the faulty equipment in the photovoltaic system is often difficult to automatically recover over time, so staff are required to inspect the photovoltaic system.

[0072] Furthermore, when human and material resources are relatively tight, abnormal weather data is excluded, and only abnormal equipment data that requires manual maintenance is used as the basis for determining the priority of subsequent inspections.

[0073] S500: Determine the inspection priority of the corresponding photovoltaic area according to the abnormal data.

[0074] The higher the inspection priority, the sooner the corresponding PV area needs to be inspected.

[0075] Generally speaking, the larger the value of the abnormal data and the greater the number of abnormal data, the higher the inspection priority of the corresponding photovoltaic area.

[0076] S600: Determine the inspection area of ​​the current round according to the inspection priority.

[0077] The inspection area for each round is at least one photovoltaic area. When the drone used for inspection has a long endurance, multiple photovoltaic areas can be inspected at one time. When the drone used for inspection has a short endurance, a photovoltaic area can be further partitioned and the inspection of the photovoltaic area can be completed through multiple flights.

[0078] Therefore, when determining the inspection area for the current round, all photovoltaic areas can be sorted in descending order according to the inspection priority, and then the number of photovoltaic areas that can be inspected can be comprehensively determined based on the endurance of the drones used for inspection, the number of drones, the number of operators, etc., and the corresponding number of photovoltaic areas can be selected from the beginning.

[0079] In one embodiment, determining the effective working time period according to the environmental data set includes the following steps:

[0080] S210, dividing the time period information into a plurality of theoretical working time periods and theoretical non-working time periods according to the first preset time and the second preset time;

[0081] S220, respectively judging whether the meteorological data corresponding to each theoretical working time period contains specific weather, the specific weather being weather suitable for the photovoltaic system to work,

[0082] S230: If yes, extract the time period corresponding to the specific weather as the valid period, and sort all the valid periods in chronological order to form a valid working time period.

[0083] The first preset time is the manually set start working time of the photovoltaic system, and the second preset time is the manually set stop working time of the photovoltaic system. The time difference between the first preset time and the second preset time must be controlled within 15 hours, that is, the ratio of the theoretical working time period to the theoretical non-working time period in a day shall not exceed 5:3 at most.

[0084] In addition, when the first preset time and the second preset time are not explicitly set manually, the first preset time may be synchronized to the sunrise time, and the second preset time may be synchronized to the sunset time.

[0085] It is understandable that the theoretical working time period is the daytime period, during which theoretically there is sunlight for the photovoltaic system to generate electricity; while the theoretical non-working time period is the nighttime period, during which the photovoltaic system cannot work.

[0086] Specific weather includes sunny days, cloudy days, and cloudy days with a certain light intensity, wherein a cloudy day with a certain light intensity means that when the weather is cloudy, the light intensity needs to exceed a preset basic value to ensure that the photovoltaic system can work normally.

[0087] By setting the first preset time and the second preset time, the theoretical non-working time period during which the photovoltaic system will definitely not work can be quickly filtered out, and then the theoretical working time period can be further screened more complexly according to the meteorological data, taking into account both the efficiency and accuracy of data processing.

[0088] In one embodiment, determining standard data of each photovoltaic area according to the environmental data set includes the following steps:

[0089] S310, obtaining pre-stored equipment parameters of each photovoltaic area and corresponding storage time.

[0090] S320: Determine whether the storage time is within a preset time.

[0091] S330: If the storage time is within the preset time, standard data is simulated according to corresponding equipment parameters and meteorological data.

[0092] S340: If the storage time is not within the preset time, then according to the meteorological data, the corresponding historical data is matched from the corresponding relationship between the preset meteorological data and the historical data as the standard data.

[0093] Equipment parameters, including the tilt angle of the photovoltaic panels and the flatness of the panel surface, which will affect photovoltaic power generation, can be automatically obtained through image analysis taken during the drone inspection process.

[0094] The preset time is set by the staff. The preset historical data is the photovoltaic data that has been confirmed to be normal during the recent inspections. As the inspection work progresses, the historical data will also be updated regularly.

[0095] When the storage time is still within the preset time, it indicates that the corresponding device parameters are highly timely, and more accurate standard data can be directly estimated based on the device parameters.

[0096] When the stored data is not within the preset time, the accuracy of the standard data obtained by still using device parameter estimation cannot be guaranteed. Therefore, historical data matching is used to obtain more stable standard data.

[0097] In one embodiment, determining the questionable data in the photovoltaic data segment corresponding to the photovoltaic area according to the standard data includes the following steps:

[0098] S350. Determine an adjustment coefficient according to the meteorological data corresponding to each valid segment, and adjust the standard data according to the adjustment coefficient to form a standard interval corresponding to the valid segment.

[0099] S360: Determine whether the photovoltaic data corresponding to each valid segment is within the corresponding standard range.

[0100] S370: If the photovoltaic data corresponding to the valid segment is within the corresponding standard interval, the corresponding data is normal data.

[0101] S380: If the photovoltaic data corresponding to any valid segment is not within the corresponding standard interval, the photovoltaic data outside the standard interval is extracted as questionable data.

[0102] Different meteorological data correspond to different adjustment coefficients. For example, the adjustment coefficient corresponding to cloudy weather is larger than that corresponding to sunny weather. That is, for the same standard data, the range of the standard interval generated by the former is larger than that of the latter.

[0103] Adjusting the standard data according to the adjustment coefficient means generating the upper wave coefficient and the lower wave coefficient according to the adjustment coefficient, and then multiplying the standard data by the upper wave coefficient to get the upper limit of the standard interval, and multiplying the standard data by the lower wave coefficient to get the lower limit of the standard interval. For example, if the adjustment coefficient is 0.1, the upper wave coefficient is 1.1 (1 + adjustment coefficient) and the lower wave coefficient is 0.9 (1-adjustment coefficient).

[0104] In one embodiment, removing abnormal weather data according to meteorological data includes the following steps:

[0105] S410, determining the regional wind direction according to meteorological data, combining a plurality of photovoltaic areas in the same wind direction according to the regional wind direction, and recording the order in which the plurality of photovoltaic areas in the regional combination are arranged in sequence according to the wind direction as a first sort.

[0106] S420, screening out questionable data with high waveform similarity and amplitude difference within a preset threshold between different photovoltaic areas in the same area combination as a comparison group, and sorting the corresponding photovoltaic areas according to the order of recording time of the questionable data in the comparison group to obtain a second sort.

[0107] S430: Determine whether the second sorting is consistent with the first arrangement of the corresponding region combination.

[0108] S440: If the second sorting is consistent with the first sorting of the corresponding area combination, the questionable data in the control group is taken as abnormal weather data.

[0109] The waveform similarity is obtained by comparing the graphics of the questionable data. In order to facilitate the comparison of graphics, the graphics of the questionable data will be scaled so that all graphics to be compared are of the same size. Therefore, the amplitudes of the questionable data also need to be compared to avoid obvious misjudgments.

[0110] When the first sorting and the second sorting are consistent, it can be considered that the clouds have passed through the photovoltaic areas in the corresponding area combination in sequence, resulting in the output power of these photovoltaic areas having nearly consistent changes in time sequence. Therefore, the corresponding questionable data is regarded as weather anomaly data.

[0111] In addition, the weather anomaly data can be further verified: by obtaining the wind force and cloud thickness data corresponding to the regional wind direction to calculate the cloud movement speed, and then determine the cloud passage time based on the cloud movement speed and the distance between each photovoltaic area in the same regional combination, and further verify whether the suspicious data in the control group is weather anomaly data based on the cloud passage time. For example, if the difference between the appearance time of a certain abnormal data in a photovoltaic area and the appearance time in the next adjacent photovoltaic area in the control group is close to the cloud passage time, then it can be determined that the abnormal data is caused by the cloud.

[0112] In one embodiment, a plurality of photovoltaic areas in the same wind direction are combined as a region according to the regional wind direction, including the following steps:

[0113] S411. Determine, according to the wind direction, several photovoltaic areas that are located in the edge area and face the wind first as the primary areas.

[0114] S412, forming a wind path according to the windward width perpendicular to the wind direction in each first-place area.

[0115] S413, calculating the windward area of ​​each photovoltaic area on the windward path.

[0116] S414: Determine whether the wind-passing area reaches more than half of the total area of ​​the corresponding photovoltaic area.

[0117] S415: If yes, the photovoltaic area is included in the same area combination as the first area.

[0118] There are multiple primacy areas corresponding to different wind directions. For example, if the wind direction is from south to north, the photovoltaic area at the southern edge of the entire photovoltaic system can be used as the primacy area.

[0119] A square with the smallest area is constructed outside the first area with the wind direction as one side of the square and the direction perpendicular to the wind direction as the other side of the square. The width of the side of the square perpendicular to the wind direction is the windward width.

[0120] In step S414, the judgment standard of the ratio of the windward area to the total area of ​​the corresponding photovoltaic area is not a fixed half, and can be set by the staff. The larger the judgment standard, the greater the probability that the photovoltaic areas in the regional combination are only affected by the clouds in the same area, and the more accurate the judgment of weather anomaly data.

[0121] In one embodiment, determining the inspection priority of the corresponding photovoltaic area according to the abnormal data includes the following steps:

[0122] S510: Determine a first weight value according to the proportion and amplitude of abnormal data in the corresponding photovoltaic data segment.

[0123] S520: Determine a second weight value according to pre-stored historical inspection time.

[0124] S530: Determine an inspection priority according to the first weight value and the second weight value of the same photovoltaic area.

[0125] The proportion of abnormal data refers to the proportion of the sum of the time of all abnormal data appearing in the photovoltaic data segment to the total time of the corresponding photovoltaic data segment. The amplitude of abnormal data refers to the deviation value of the maximum fluctuation peak in the abnormal data from the closest boundary of the standard range.

[0126] The greater the proportion or amplitude of abnormal data, the higher the first weight value; the earlier the historical inspection time, the higher the second weight value.

[0127] The inspection priority is calculated using the following formula:

[0128] P=K1*R1+K2*R2, where P is the inspection priority, K1 and K2 are both preset coefficients, R1 is the first weight value, and R2 is the second weight value. Generally speaking, K1 is greater than K2 to reflect that the importance of data is higher than the importance of time.

[0129] In one embodiment, determining the first weight value according to the proportion and amplitude of abnormal data in the corresponding photovoltaic data segment includes the following steps:

[0130] S511, respectively determining whether the maximum amplitude exceeds a preset first threshold and whether the proportion exceeds a preset second threshold.

[0131] S512: If any one of them exceeds the limit, the highest inspection priority will be given to them.

[0132] S513: If neither exceeds, determine a first weight value according to a first ratio of the maximum amplitude to the first threshold and a second ratio of the proportion to the second threshold.

[0133] The first threshold and the second threshold are both set manually.

[0134] When the maximum value exceeds the first threshold or the proportion exceeds the second threshold, it means that the data is seriously abnormal and an inspection is urgently needed to determine the cause of the abnormality. Therefore, the influence of the second weight corresponding to the inspection time is no longer considered, and the highest inspection priority is directly given.

[0135] There are many ways to determine the first weight value according to the first ratio and the second ratio. For example, the first weight value can be the one with the largest value between the first ratio and the second ratio; or for another example, the first weight value can be the average value of the first ratio and the second ratio. In this embodiment, the latter determination method is selected.

[0136] In addition, the second weight value is determined by subtracting the historical inspection time from the current time to obtain the time difference, and then using the ratio of the time difference to the preset time period as the second weight value.

[0137] It can be seen that the maximum value of the first weight is 1, while the second weight can exceed 1. Therefore, as long as the photovoltaic data of other photovoltaic areas do not exceed the warning value (that is, the amplitude does not exceed the first threshold or the proportion does not exceed the second threshold), even if a photovoltaic area has no data abnormality, it can eventually get the opportunity for inspection through the continuous increase of the second weight value, and update the corresponding equipment parameters in a more flexible way than regular inspection.

[0138] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A photovoltaic power station intelligent inspection method, characterized in that: The following steps are involved: At the end of each round of inspection, the environmental data set collected during the round of inspection and the power generation data set corresponding to each pre-divided photovoltaic area are obtained, wherein the environmental data set includes meteorological data and time period information corresponding to the meteorological data; The time period information is integrated into the effective working time period, and each power generation data set is integrated into the photovoltaic data segment according to the effective working time period. Specifically, Dividing the time period information into a plurality of theoretical working time periods and theoretical non-working time periods according to the first preset time and the second preset time; It is determined whether the meteorological data corresponding to each theoretical working time period contains specific weather, where the specific weather is suitable for the photovoltaic system to work. If yes, extract the time period corresponding to the specific weather as the valid period, and sort all the valid periods in chronological order to form the valid working time period; Determine the standard data of each photovoltaic area according to the environmental data set, and determine the doubtful data in the photovoltaic data segment of the corresponding photovoltaic area according to the standard data; wherein the doubtful data includes equipment abnormal data and weather abnormal data, specifically, Determine an adjustment coefficient according to the meteorological data corresponding to each valid segment, and adjust the standard data according to the adjustment coefficient to form a standard interval corresponding to the valid segment; Determine whether the photovoltaic data corresponding to each valid segment is within the corresponding standard range. If so, the corresponding data is normal data; If not, the photovoltaic data outside the standard range are extracted as questionable data; Eliminate abnormal weather data in questionable data according to meteorological data to convert the questionable data into abnormal data; Determine the inspection priority of the corresponding photovoltaic area based on abnormal data; The inspection area of ​​the current round is determined based on the inspection priority.

2. A photovoltaic power station intelligent inspection method according to claim 1, characterized in that: Determining standard data for each photovoltaic area based on the environmental data set includes the following steps: Obtain the pre-stored equipment parameters and corresponding storage time of each photovoltaic area; Determine whether the storage time is within the preset time. If so, standard data is simulated based on the corresponding equipment parameters and meteorological data; If not, then the corresponding historical data is matched as standard data from the corresponding relationship between the preset meteorological data and the historical data according to the meteorological data.

3. A photovoltaic power station intelligent inspection method according to claim 1, characterized in that: The method of eliminating abnormal weather data according to meteorological data comprises the following steps: Determine the regional wind direction according to meteorological data, and combine several photovoltaic areas in the same wind direction as a regional combination according to the regional wind direction, and record the order in which the multiple photovoltaic areas in the regional combination are arranged in sequence according to the wind direction as the first sorting; Screening out suspicious data with high waveform similarity and amplitude difference within a preset threshold between different photovoltaic areas in the same area combination as a comparison group, and sorting the corresponding photovoltaic areas according to the order of recording time of the suspicious data in the comparison group to obtain a second sorting; Determine whether the second sorting is consistent with the first sorting of the corresponding region combination, If so, the questionable data in the control group will be regarded as abnormal weather data.

4. A photovoltaic power station intelligent inspection method according to claim 3, characterized in that: The method of combining a plurality of photovoltaic areas in the same wind direction as a regional combination according to the regional wind direction comprises the following steps: According to the wind direction, several photovoltaic areas located in the edge areas and facing the wind first are determined as the first areas. The wind path is formed according to the windward width perpendicular to the wind direction in each first-place area. Calculate the windward area of ​​each photovoltaic area on the windward path. Determine whether the wind-passing area reaches more than half of the total area of ​​the corresponding photovoltaic area. If so, the PV region is included in the same region combination as the first region.

5. A photovoltaic power station intelligent inspection method according to claim 1, characterized in that: Determining the inspection priority of the corresponding photovoltaic area according to the abnormal data includes the following steps: Determine a first weight value according to the proportion and amplitude of the abnormal data in the corresponding photovoltaic data segment, wherein the greater the proportion or the greater the amplitude of the abnormal data, the higher the first weight value; Determine a second weight value according to the pre-stored historical inspection time, wherein the earlier the historical inspection time is, the higher the second weight value is; The inspection priority is determined according to the first weight value and the second weight value of the same photovoltaic area.

6. A photovoltaic power station intelligent inspection method according to claim 5, characterized in that: The method of determining the first weight value according to the proportion and amplitude of the abnormal data in the corresponding photovoltaic data segment includes the following steps: It is determined whether the maximum value exceeds a preset first threshold and whether the proportion exceeds a preset second threshold. If any one of them exceeds the limit, the highest inspection priority will be given; If neither is exceeded, the first weight value is determined according to a first ratio of the maximum amplitude to the first threshold and a second ratio of the proportion to the second threshold.

Images