Mountain photovoltaic single unit power abnormity diagnosis method and device

Abstract

The invention provides a mountain photovoltaic single unit power abnormity diagnosis method and device, and relates to the technical field of mountain photovoltaic power generation, and the method comprises the steps: judging whether a target unit has a power abnormal reduction condition or not according to a current-day period-by-period power sequence and a historical same-period sunny day power sequence of the target unit of a mountain photovoltaic power station; and if yes, executing the subsequent process. And then power-irradiance synchronism is judged, and the fault condition of the target unit is accurately judged based on a synchronism judgment result. In the scene with high synchronism, the fault reason of the target unit is judged by combining the conditions of other units in the mountain photovoltaic power station and weather factors; and in a scene with low synchronism, based on the current-day power curve and the current-day irradiance curve, a DTW algorithm is utilized to judge a fault abnormity reason and quantify a fault degree. The method accurately adapts to the problem of asynchronous power of the mountain photovoltaic units, and can accurately and reliably diagnose the power abnormality of the single photovoltaic unit in the mountain terrain.

Description

Technical Field

[0001] This invention relates to the technical field of mountain photovoltaic power generation, and more specifically, to a method and device for diagnosing power anomalies in a single mountain photovoltaic unit. Background Technology

[0002] Driven by the "dual-carbon" strategic goals, the scale of photovoltaic power generation continues to expand. To alleviate the contradiction of scarce land resources in plain areas, the construction of mountain photovoltaic power stations using unused resources such as barren mountains and slopes has become an important development direction for photovoltaic power generation. However, how to accurately diagnose power anomalies in individual units of mountain photovoltaic power stations has become the core technical bottleneck for the intelligent operation and maintenance of mountain photovoltaic power stations.

[0003] Unlike flat terrain, the undulating terrain of mountainous photovoltaic power stations leads to significant differences in the installation slope, azimuth, and even local microclimate of different photovoltaic units within the same array. This difference directly manifests as substantial variations in the output power amplitude of adjacent units, even under identical illumination conditions. Due to the lack of available healthy reference units with consistent power characteristics for comparison, the core principle of traditional power anomaly diagnosis – "horizontal comparison of power across multiple units" – is essentially ineffective in mountainous environments. Another power loss characteristic analysis method can distinguish anomaly types by analyzing the morphological characteristics of the power curve. However, in mountainous environments, local shading and early equipment failures may present similar power loss curves, resulting in high overlap and low distinguishability. Furthermore, its judgment process heavily relies on the analyst's subjective experience with the curve morphology, making it difficult to establish stable and quantifiable judgment standards, leading to low automation and unreliable performance.

[0004] Performance ratio (PR) monitoring assesses power plant health by calculating the ratio of actual power generation to theoretical power generation. However, this is a macroscopic, a posteriori assessment indicator. Its main drawback is that it cannot pinpoint individual units in real time and accurately, and it cannot effectively distinguish between "normal PR fluctuations due to terrain and weather" and "PR declines caused by equipment malfunctions." Therefore, it offers limited guidance for quickly locating and handling individual unit faults during routine operation and maintenance. Summary of the Invention

[0005] To address the issue of low accuracy in diagnosing power anomalies in mountainous photovoltaic power station units using existing technologies, this invention proposes a method and device for diagnosing power anomalies in single photovoltaic units in mountainous terrain, enabling accurate and reliable power anomaly diagnosis for individual photovoltaic units in mountainous areas.

[0006] To achieve the above-mentioned technical effects, the technical solution of the present invention is as follows: Firstly, this application proposes a method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area, comprising the following steps: S1: Obtain multi-source correlation data for power analysis of each target unit in the mountain photovoltaic power station; the multi-source correlation data for power analysis includes the daily time-period power sequence, the historical same-day sunny day power sequence, and the daily time-period irradiance sequence. S2: Obtain the daily power curve and daily irradiance curve based on the daily time-by-time power sequence and the daily time-by-time irradiance sequence, respectively; S3: Compare the daily power sequence of the target unit with the historical power sequence of the same period on sunny days to determine whether there is an abnormal decrease in power of the target unit. If so, proceed to S4; otherwise, the power fluctuation of the target unit is normal. S4: Based on the daily power curve and daily irradiance curve of the target unit, determine the level of power-irradiance synchronicity. If the synchronicity is high, proceed to S5; if the synchronicity is low, proceed to S6. S5: Calculate the percentage of units in the mountain photovoltaic power station other than the target unit that have abnormal power reduction, and determine the cause of the target unit's failure in a high synchronization scenario based on the percentage of units and the preset weather boundary correlation coefficient. S6: Calculate the DTW distance between the power curve and the irradiance curve of the day, and obtain the fault degree and cause of the target unit under the low synchronicity scenario based on the DTW distance.

[0007] Preferably, the daily time-period power sequence is represented as follows: The daily time-period irradiance sequence is represented as follows: , and The target unit is the kth unit. T An array consisting of power and irradiance at each acquisition time. and These are the kth generating units on that day. Power and irradiance at each sampling time, T Indicates the number of sampling points within the collection period; Based on the daily power sequence, a daily power curve is obtained with the sampling point as the x-axis and power as the y-axis; based on the daily irradiance sequence, a daily irradiance curve is obtained with the sampling point as the x-axis and irradiance as the y-axis.

[0008] Preferably, in S3, the daily power sequence of the target unit is compared with the historical power sequence for the same period under sunny weather conditions to determine whether the target unit has an abnormal power reduction. The process is as follows: S31: Based on the daily time-period power sequence Power sequence of sunny days in the same period of history Calculate the first one respectively k The average daily power of the target unit Average power of sunny days in the same period of history ; S32: Calculate the statistical relative threshold based on the historical sunny-day power series for the same period. ; S33: If If the power output of the target unit is abnormally low, then the power output of the target unit is abnormally low; otherwise, the power fluctuation of the target unit is normal.

[0009] Preferably, in S4, based on the daily power curve and daily irradiance curve of the target unit, the degree of power-irradiance synchronicity is determined. The process is as follows: The synchronicity of the daily power curve and daily irradiance curve of the target unit is calculated using the Pearson correlation coefficient, and the synchronicity judgment threshold is obtained. The expression is:

[0010] in, This represents the average power sequence for each time period of the day. This represents the average value of the daily irradiance sequence for each time period. Compare to preset correlation thresholds With the aforementioned synchronization determination threshold The size, if ≥ If the synchronization is high, then the synchronization is low; otherwise, the synchronization is low.

[0011] Preferably, in S5, the percentage of units in the mountain photovoltaic power station other than the target unit that experience abnormal power reduction is calculated, and based on the percentage of units and a preset weather boundary correlation coefficient, the cause of the target unit's failure in a high-synchronization scenario is determined. The process is as follows: Statistics on the set of other units in the mountain photovoltaic power station besides the target unit. Number of units with abnormally low power And calculate the number of units with abnormally low power. The proportion of units The expression is: ; S This refers to the set of all units in a mountain photovoltaic power station other than the target unit. ... ,in, For the number of other units, and ; Compare the correlation coefficients of preset weather boundaries and the proportion of generating units The size, if > In scenarios with high synchronicity, the cause of the target unit's failure is weather-related; if In scenarios with high synchronicity, the cause of failure in the target unit is generalized equipment aging.

[0012] Preferably, the process of S6 is as follows: Calculate the DTW distance between the power curve and the irradiance curve for the day; The optimal matching path for DTW is extracted based on DTW distance, and the morphological similarity is calculated based on DTW distance. The degree of failure of the target unit in the low synchronization scenario is quantified based on the morphological similarity. Quantization features are obtained based on the optimal matching path of DTW. The causes of failures in target units under low synchronization scenarios are determined based on the quantitative characteristics.

[0013] Preferably, the DTW distance between the daily power curve and the daily irradiance curve is calculated, the optimal DTW matching path is extracted based on the DTW distance, and the morphological similarity is calculated based on the DTW distance. The degree of failure of the target unit in a low-synchronization scenario is then quantified based on the morphological similarity metric. The process is as follows: C1: Normalize the daily power sequence and daily irradiance sequence of the target unit for each time period to obtain the normalized power value. and normalized value of irradiance ; C2: Based on the power normalization value and the normalized value of the irradiance Calculate the element values ​​of the distance matrix And construct the distance matrix The element values ​​of the distance matrix The expression is:

[0014] in, Indicates the order of sampling points. Indicates the number of sampling points; C3: Based on distance matrix Construct the cumulative cost matrix based on the cumulative cost recursive formula. The cumulative cost recursive formula is as follows:

[0015] when or hour, ; C4: Based on the cumulative cost matrix endpoint value The DTW distance between the daily power curve and the daily irradiance curve According to the cumulative cost matrix The optimal matching path and its length L are determined by backtracking the cumulative cost matrix. get; The DTW distance Convert to morphological similarity The expression is:

[0016] Where max(D) is the distance matrix. The maximum value, L The optimal matching path length for DTW; based on morphological similarity As a reference value for the degree of failure of the target unit The closer to 0, the higher the degree of failure. The closer it is to 1, the lower the degree of failure.

[0017] Preferably, based on the DTW optimal matching path, quantization features are obtained, and the process is as follows: D1: Construct a difference sequence based on the DTW optimal matching path. Morphological difference values Satisfying the expression:

[0018] in, , These are the 1st, 2nd, and 3rd digits of the optimal matching path in DTW. The sampling point number of the daily time-period power sequence and daily time-period irradiance sequence corresponding to each matching location; , These are the normalized power value and normalized irradiance value corresponding to the matching point in the DTW optimal matching path, respectively. D2: Calculate the dynamic difference threshold based on the difference sequence, and then calculate the morphological difference values ​​in the difference sequence. Points exceeding the dynamic difference threshold are designated as high difference points, and the intervals occupied by high difference points are designated as high difference intervals. D3: The maximum length of consecutive high-difference intervals in a statistically differential sequence. And calculate the proportion of consecutive high-difference intervals. The expression is: ; D4: Based on the dynamic difference threshold, count the number of isolated high difference points among high difference points. And calculate the proportion of isolated high-difference points. The expression is: The isolated high-difference point is a point that is itself greater than the dynamic difference threshold and whose adjacent points are all less than the dynamic difference threshold. D5: Calculate the coefficient of variation of the differentially expressed sequences. The expression is: ,in, The standard deviation of the differentially expressed series is represented by the standard deviation of the series. This represents the mean of the differentially expressed sequences; D6: The maximum length of a continuous high-difference interval , percentage of consecutive high difference intervals The proportion of isolated high-difference points and coefficient of variation As a quantitative feature.

[0019] Preferably, determining the cause of failure of the target unit in a low synchronization scenario based on the quantification characteristics includes: If the proportion of consecutive high difference intervals Greater than or equal to the preset fixed occlusion threshold and the coefficient of variation If the value is greater than or equal to the preset variation threshold, the cause of the target unit's failure is fixed blockage. If the proportion of isolated high-difference points The maximum length of a continuous high-difference interval that is greater than or equal to a preset random occlusion threshold. If the length is less than the preset threshold, the cause of the target unit's failure is random obstruction; If the cause of the target unit's failure is neither a fixed blockage nor a random blockage, then the cause of the target unit's failure is an asynchronous equipment failure.

[0020] Secondly, this application proposes a power anomaly diagnostic device for a single photovoltaic unit in a mountainous area, used to implement the method, comprising: The data acquisition unit is used to acquire multi-source correlation data of power analysis for each target unit of the mountain photovoltaic power station; the multi-source correlation data of power analysis includes the power sequence of each time period of the day, the power sequence of the same period of the historical sunny day, and the irradiance sequence of each time period of the day; The curve acquisition unit is used to acquire the daily power curve and the daily irradiance curve based on the daily time-by-time power sequence and the daily time-by-time irradiance sequence, respectively. The power anomaly determination unit is used to compare the daily power sequence of the target unit with the historical power sequence of the same period on sunny days to determine whether the target unit has an abnormal power reduction. The synchronization determination unit is used to determine the level of power-irradiance synchronization based on the daily power curve and daily irradiance curve of the target unit. The first fault condition determination unit is used to calculate the proportion of units in the mountain photovoltaic power station other than the target unit that have abnormal power reduction, and to determine the fault cause of the target unit in a high synchronization scenario based on the proportion of units and the preset weather boundary correlation coefficient. The second fault condition determination unit is used to calculate the DTW distance between the power curve and the irradiance curve of the day, and to obtain the fault degree and fault cause of the target unit under the low synchronicity scenario based on the DTW distance.

[0021] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: This invention proposes a method and device for diagnosing power anomalies in single-unit photovoltaic power plants in mountainous areas. The method obtains daily power curves and daily irradiance curves based on daily time-by-time power sequences and daily time-by-time irradiance sequences, respectively. By combining the daily time-by-time power sequences with historical power sequences for the same period under clear weather conditions, it determines whether the target unit experiences an abnormal power reduction. The method confirms the trend synchronization of the daily power curve and the daily irradiance curve, and accurately determines the fault condition based on the synchronization judgment result. In scenarios with high synchronization, a wide-area anomaly diagnosis is performed by considering the conditions of other units in the mountainous photovoltaic power station and relevant weather factors to determine the cause of the target unit's fault. In scenarios with low synchronization, the DTW algorithm is used based on the daily power curve and the daily irradiance curve to determine the cause of the target unit's fault and quantify the degree of fault. This invention is applied to the operation and maintenance and power anomaly diagnosis of mountain photovoltaic power stations. It accurately adapts to the problem of power asynchrony of mountain photovoltaic units, effectively avoids the risk of misjudgment in cross-unit comparison of traditional methods, and improves the accuracy of anomaly identification in complex mountain scenarios compared with existing methods. It can achieve accurate and reliable power anomaly diagnosis of single photovoltaic units in mountainous terrain, and directly uses the existing monitoring data of the power station to achieve diagnosis without the need for additional hardware, which has significant engineering application value. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area, as proposed in this embodiment of the invention. Figure 2 This represents a detailed breakdown of the causes of the target unit failure as proposed in this embodiment of the invention. Figure 3 This diagram illustrates the composition of the power anomaly diagnosis device for a single photovoltaic unit in a mountainous area, as proposed in this embodiment of the invention. Detailed Implementation

[0023] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent. To better illustrate this embodiment, some parts of the accompanying drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions; It is understandable to those skilled in the art that some well-known details may be omitted from the accompanying drawings.

[0024] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments; The positional relationships depicted in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent.

[0025] Example 1 This embodiment provides a method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area. The flowchart of this method can be found here. Figure 1 This includes the following steps: S1: Obtain multi-source correlation data for power analysis of each target unit in the mountain photovoltaic power station; the multi-source correlation data for power analysis includes the daily time-period power sequence, the historical same-day sunny day power sequence, and the daily time-period irradiance sequence. S2: Obtain the daily power curve and daily irradiance curve based on the daily time-by-time power sequence and the daily time-by-time irradiance sequence, respectively; S3: Compare the daily power sequence of the target unit with the historical power sequence of the same period on sunny days to determine whether there is an abnormal decrease in power of the target unit. If so, proceed to S4; otherwise, the power fluctuation of the target unit is normal. S4: Based on the daily power curve and daily irradiance curve of the target unit, determine the level of power-irradiance synchronicity. If the synchronicity is high, proceed to S5; if the synchronicity is low, proceed to S6. S5: Calculate the percentage of units in the mountain photovoltaic power station other than the target unit that have abnormal power reduction, and determine the cause of the target unit's failure in a high synchronization scenario based on the percentage of units and the preset weather boundary correlation coefficient. S6: Calculate the DTW distance between the power curve and the irradiance curve of the day, and obtain the fault degree and cause of the target unit under the low synchronicity scenario based on the DTW distance.

[0026] In this embodiment, firstly, the daily time-period power sequence, daily time-period irradiance sequence, and historical sunny-day power sequence of the target unit of the mountain photovoltaic power station are obtained. Based on the daily time-period power sequence and the daily time-period irradiance sequence, the daily power curve and the daily irradiance curve are obtained respectively. Combining the daily time-period power sequence and the historical sunny-day power sequence, it is determined whether the target unit has an abnormal power reduction. If so, S4 is executed; otherwise, the power fluctuation of the target unit is considered normal, and the process ends. Secondly, the trend synchronization of the daily power curve and the daily irradiance curve is determined, and the fault situation is accurately determined based on the synchronization judgment result. In scenarios with high synchronization, a wide-area anomaly diagnosis is performed by considering the situation of other units in the mountain photovoltaic power station and relevant weather factors to determine the cause of the target unit's fault. In scenarios with low synchronization, the Dynamic Time Warping (DTW) algorithm is used to determine the cause of the target unit's fault based on the daily power curve and the daily irradiance curve, and the degree of fault is quantified. Finally, various anomaly results are compiled, and a list of photovoltaic power station power anomaly types is output.

[0027] This invention is applied to the operation and maintenance and power anomaly diagnosis of photovoltaic power stations in mountainous areas. It accurately adapts to the problem of asynchronous power of photovoltaic units in mountainous areas, effectively avoids the risk of misjudgment in cross-unit comparison of traditional methods, improves the accuracy of anomaly identification in complex mountainous scenarios compared with existing methods, and directly uses the existing monitoring data of the power station to achieve diagnosis without the need for additional hardware, which has significant engineering application value.

[0028] Example 2 In this embodiment, the daily time-period power sequence is represented as follows: The daily time-period irradiance sequence is represented as follows: , and The target unit is the kth unit. T An array consisting of power and irradiance at each acquisition time. and These are the kth generating units on that day. Power and irradiance at each sampling time, T Indicates the number of sampling points within the collection period; Based on the daily power sequence, a daily power curve is obtained with the sampling point as the x-axis and power as the y-axis; based on the daily irradiance sequence, a daily irradiance curve is obtained with the sampling point as the x-axis and irradiance as the y-axis.

[0029] Specifically, irradiance monitoring units are deployed at the target mountain photovoltaic power station, and power data of the output circuit of the target photovoltaic power station is collected simultaneously. This data is then used to monitor the target units of the mountain photovoltaic power station from level 1 to level 2. M Number them.M ≥1, definition k =1,2,…, M ,and k The initial value is 1, from 1 to M Traversal k Obtain the daily time-by-time power sequence of each target unit. With the daily time-by-time irradiance sequence And obtain the historical sunny day power data of the target unit for the same period of the past three consecutive years as the historical sunny day power sequence. T This indicates the number of sampling points within the collection period. T >4, T It is a multiple of 4.

[0030] In S3, the daily time-by-time power sequence of the target unit is compared with the historical power sequence for the same period on sunny days to determine whether there is an abnormal decrease in power of the target unit. The process is as follows: S31: Based on the daily time-period power sequence Power sequence of sunny days in the same period of history Calculate the first one respectively k The average daily power of the target unit Average power of sunny days in the same period of history ; S32: Calculate the statistical relative threshold based on the historical sunny-day power series for the same period. ; S33: If If the power output of the target unit is abnormally low, then the power output of the target unit is abnormally low; otherwise, the power fluctuation of the target unit is normal.

[0031] Specifically, based on the daily time-period power sequence of the target unit. Power sequence of sunny days in the same period of history The statistical relative threshold method is used to quickly determine whether the target unit has an abnormal power reduction.

[0032] In S31, calculate the average power for the day. The expression is: ; Calculate the historical average power output of the k-th generating unit during the same period on sunny days. The expression is ,in, Let t be the power at time t on a sunny day during the same period of year i.

[0033] In S32, based on historical clear-day power data for the same period, this embodiment selects to lower the threshold. If it is 5%, then the statistical relative threshold is... ; In S33, if If the target unit has an abnormal power reduction, then S4 is executed; otherwise, the power fluctuation of the target unit is normal, and the process terminates.

[0034] In S4, based on the target unit's daily power curve and daily irradiance curve, the degree of power-irradiance synchronicity is determined. The process is as follows: The synchronicity of the daily power curve and daily irradiance curve of the target unit is calculated using the Pearson correlation coefficient, and the synchronicity judgment threshold is obtained. The expression is:

[0035] in, This represents the average power sequence for each time period of the day. This represents the average value of the daily irradiance sequence for each time period. Compare to preset correlation thresholds With the aforementioned synchronization determination threshold The size, if ≥ If the synchronization is high, then the synchronization is low; otherwise, the synchronization is low.

[0036] Specifically, the average power sequence for each time period of the day. Average daily irradiance sequence .

[0037] Compare to preset correlation thresholds With the aforementioned synchronization determination threshold The size, when ≥ When, the synchronization is high; when < At that time, the synchronization was low.

[0038] This embodiment details the causes of abnormal power reduction in the target unit. See [link / reference] Figure 2 If the power-irradiance synchronization is high, the causes of the failure include weather effects and general equipment aging; if the power-irradiance synchronization is low, the causes of the failure include fixed shading, random shading, or asynchronous equipment failure.

[0039] In S5, the percentage of units in the mountain photovoltaic power station other than the target unit that experience abnormal power reduction is calculated. Based on this percentage and a preset weather boundary correlation coefficient, the cause of the target unit's failure in a high-synchronization scenario is determined. The process is as follows: Statistics on the set of other units in the mountain photovoltaic power station besides the target unit. Number of units with abnormally low power And calculate the number of units with abnormally low power. The proportion of units The expression is: ; S This refers to the set of all units in a mountain photovoltaic power station other than the target unit. ... ,in, For the number of other units, and ; Compare the correlation coefficients of preset weather boundaries and the proportion of generating units The size, if > In scenarios with high synchronicity, the cause of the target unit's failure is weather-related; if In scenarios with high synchronicity, the cause of failure in the target unit is generalized equipment aging.

[0040] In this embodiment, a weather boundary correlation coefficient is preset. ,like > In scenarios with high synchronicity, the cause of the target unit's failure is weather-related; if If the cause of the target unit's failure in a high synchronization scenario is generalized equipment aging, then after the comparison is completed, the number and type of the abnormal unit are recorded, and the process terminates.

[0041] The process of S6 is as follows: Calculate the DTW distance between the power curve and the irradiance curve for the day; The optimal matching path for DTW is extracted based on DTW distance, and the morphological similarity is calculated based on DTW distance. The degree of failure of the target unit in the low synchronization scenario is quantified based on the morphological similarity. Quantization features are obtained based on the optimal matching path of DTW. The causes of failures in target units under low synchronization scenarios are determined based on the quantitative characteristics.

[0042] In this embodiment, the DTW distance between the daily power curve and the daily irradiance curve is calculated. The optimal DTW matching path is extracted based on the DTW distance. Morphological similarity is calculated based on the DTW distance, and the degree of failure of the target unit in a low-synchronization scenario is quantified according to the morphological similarity. The process is as follows: C1: Normalize the daily power sequence and daily irradiance sequence of the target unit for each time period to obtain the normalized power value. and normalized value of irradiance ; C2: Based on the power normalization value and the normalized value of the irradiance Calculate the element values ​​of the distance matrix And construct the distance matrix The element values ​​of the distance matrix The expression is:

[0043] in, Indicates the order of sampling points. Indicates the number of sampling points; C3: Based on distance matrix Construct the cumulative cost matrix based on the cumulative cost recursive formula. The cumulative cost recursive formula is as follows:

[0044] when or hour, ; C4: Based on the cumulative cost matrix endpoint value The DTW distance between the daily power curve and the daily irradiance curve According to the cumulative cost matrix The optimal matching path and its length L are determined by backtracking the cumulative cost matrix. get; The DTW distance Convert to morphological similarity The expression is:

[0045] Where max(D) is the distance matrix. The maximum value, L The optimal matching path length for DTW; based on morphological similarity As a reference value for the degree of failure of the target unit The closer to 0, the higher the degree of failure. The closer it is to 1, the lower the degree of failure.

[0046] Specifically, in C1, the daily time-by-time power sequence of the target unit is... With the daily time-by-time irradiance sequence Normalization is performed to obtain the normalized power value. and normalized value of irradiance The expression is:

[0047]

[0048] In C3, the boundary conditions are: .

[0049] In C4, the optimal matching path and the length L of the optimal matching path of the DTW are obtained by backtracking the cumulative cost matrix C. The specific steps are as follows: C41: From the end of the cumulative cost matrix Start backtracking, examine three adjacent cells, and select the cell with the lowest cumulative cost as the previous point on the path.

[0050] C42: Repeat C41, each time moving to the adjacent cell with the lowest cumulative cost, until backtracking to the starting point.

[0051] C43: Records the sequence of cell indices traversed during the backtracking process, which is the optimal matching path of DTW; records the number of cells traversed during the backtracking process, which is the length L of the optimal matching path of DTW.

[0052] Morphological similarity It is a percentage ranging from 0 to 1, used to quantify the severity of a fault. The closer the value is to 0, the more severe the fault. The closer the coefficient is to 1, the less severe the fault. This differs from the linear correlation coefficient of the Pearson method. Directly linked to the severity of the fault, output Used as a reference value for the degree of failure of the target unit.

[0053] Based on the DTW optimal matching path, the quantized features are obtained through the following process: D1: Construct a difference sequence based on the DTW optimal matching path. Morphological difference values Satisfying the expression:

[0054] in, , These are the 1st, 2nd, and 3rd digits of the optimal matching path in DTW. The sampling point number of the daily time-period power sequence and daily time-period irradiance sequence corresponding to each matching location; , These are the normalized power value and normalized irradiance value corresponding to the matching point in the DTW optimal matching path, respectively. D2: Calculate the dynamic difference threshold based on the difference sequence, and then calculate the morphological difference values ​​in the difference sequence. Points exceeding the dynamic difference threshold are designated as high difference points, and the intervals occupied by high difference points are designated as high difference intervals. D3: The maximum length of consecutive high-difference intervals in a statistically differential sequence. And calculate the proportion of consecutive high-difference intervals. The expression is: ; D4: Based on the dynamic difference threshold, count the number of isolated high difference points among high difference points. And calculate the proportion of isolated high-difference points. The expression is: The isolated high-difference point is a point that is itself greater than the dynamic difference threshold and whose adjacent points are all less than the dynamic difference threshold. D5: Calculate the coefficient of variation of the differentially expressed sequences. The expression is: ,in, The standard deviation of the differentially expressed series is represented by the standard deviation of the series. This represents the mean of the differentially expressed sequences; D6: The maximum length of a continuous high-difference interval , percentage of consecutive high difference intervals The proportion of isolated high-difference points and coefficient of variation As a quantitative feature.

[0055] In this embodiment, the optimal matching path for DTW is assumed to be:

[0056] in, , These are the 1st, 2nd, and 3rd digits of the optimal matching path in DTW. The sampling point number of the daily time-period power sequence and daily time-period irradiance sequence corresponding to each matching location; In D2, pointwise normalization yields the differential sequences. And calculate the mean difference of the differential sequences. and standard deviation of differences The expressions are as follows:

[0057]

[0058] Using the mean of differences and standard deviation of differences Calculate the dynamic difference threshold The expression is:

[0059] Where 1.5 is the adjustable statistical coefficient; The point is determined to be a high difference point.

[0060] In D4, an isolated high-difference point is a point that is itself greater than the dynamic difference threshold and whose neighboring points are all less than the dynamic difference threshold, i.e., it satisfies... And adjacent points .

[0061] The causes of failures in target units under low synchronicity scenarios are determined based on the quantified characteristics, including: If the proportion of consecutive high difference intervals Greater than or equal to the preset fixed occlusion threshold and the coefficient of variation If the value is greater than or equal to the preset variation threshold, the cause of the target unit's failure is fixed blockage. If the proportion of isolated high-difference points The maximum length of a continuous high-difference interval that is greater than or equal to a preset random occlusion threshold. If the length is less than the preset threshold, the cause of the target unit's failure is random obstruction; If the cause of the target unit's failure is neither a fixed blockage nor a random blockage, then the cause of the target unit's failure is an asynchronous equipment failure.

[0062] In this embodiment, the preset fixed occlusion threshold is 20%, and the variation threshold is 0.8, that is, if , The cause of the target unit's failure is fixed obstruction; The preset random occlusion threshold is 10%, and the length threshold is 5, that is, if And the length of continuous high difference intervals The cause of the target unit's failure is random obstruction; If the above conditions of fixed obstruction and random obstruction are not met, then the cause of the target unit's failure is an asynchronous equipment failure.

[0063] Finally, record the abnormal unit number, the cause of the abnormality and the quantitative judgment basis, output the final diagnosis result, and the diagnosis process terminates.

[0064] Example 3 This embodiment provides a specific example. Table 1 shows the power data and daily irradiance data of the target unit in a mountain photovoltaic system. Table 1

[0065] Table 1 lists the daily power output of the target generating units as a time-period power sequence for that day. Irradiance is listed as the daily time-by-time irradiance sequence. The peak power of sunny days in the same period of history is listed as the power series of sunny days in the same period of history.

[0066] Excluding the 49th sampling point, calculate the historical average power for sunny days during the same period. The expression is:

[0067] Calculate the daily average power of the target unit The expression is:

[0068] If the threshold is lowered to 5% of the historical average power for sunny days in the same period, then the statistical relative threshold is... for:

[0069] Determine if the target unit experiences an abnormal power reduction:

[0070]

[0071] The target unit is experiencing an abnormal power reduction.

[0072] Execute S4 to calculate the synchronicity threshold using the Pearson correlation coefficient. The expression is

[0073] Set correlation threshold ,but This means that the synchronicity between the daily power curve and the daily irradiance curve of the target unit is low.

[0074] Execute S6 to normalize the daily time-by-time power sequence and daily time-by-time irradiance sequence of the target unit, respectively, to obtain the normalized power value. and normalized value of irradiance The expression is:

[0075]

[0076] in, and These are the minimum and maximum values ​​of the sequence, respectively.

[0077] Calculate the element values ​​of the distance matrix And build Distance matrix The element values ​​of the distance matrix The expression is:

[0078] in, , This represents the number of sampling points; The distance matrix D can be found in Table 2.

[0079] Table 2

[0080] in, .

[0081] Based on distance matrix Construct the cumulative cost matrix based on the cumulative cost recursive formula. The cumulative cost recursive formula is as follows:

[0082] when hour

[0083] Cumulative cost matrix See Table 3.

[0084] Table 3

[0085] With cumulative cost matrix endpoint value The DTW distance between the daily power curve and the daily irradiance curve ,Right now

[0086] Backtracking cumulative cost matrix The optimal matching path and its length L are obtained. In this embodiment, the optimal matching path length L is... .

[0087] The DTW distance Convert to morphological similarity The expression is:

[0088] Based on morphological similarity As a reference value for the degree of failure of the target unit A value close to 1 indicates a relatively low level of fault in the target unit.

[0089] Construct a difference sequence based on the DTW optimal matching path, and calculate the difference sequence. Mean difference and standard deviation of differences The expression is:

[0090]

[0091] Based on the mean difference and standard deviation of differences Calculate the dynamic difference threshold The expression is:

[0092] According to the DTW distance inverse calculation analysis, there are a total of 19 high difference points, including 18 continuous high difference points and 1 isolated high difference point. The location of the isolated high difference point is... .

[0093] Calculations show that the differential sequences There is only one Meanwhile, its adjacent points , It meets the criteria for identifying isolated high-difference points, therefore The location of this point .

[0094] The maximum length of consecutive high-difference intervals in a statistically differential series ,get ; Calculate the proportion of consecutive high-difference intervals The expression is:

[0095] The number of isolated high-difference points among high-difference points is counted. ,get ; Calculate the proportion of isolated high-difference points The expression is:

[0096] Calculate the coefficient of variation of the differentially expressed sequences The expression is:

[0097] The specific steps for determining the cause of failure in the target unit under low synchronicity scenarios based on quantitative characteristics are as follows: Step 1: Fixed Occlusion Detection: In this embodiment, a preset fixed occlusion threshold is... If the variation threshold is 0.8, then the proportion of consecutive high-difference intervals is... And coefficient of variation If the value is greater than 0.8, the cause of the target unit's failure is determined to be fixed obstruction.

[0098] Step 2: Random occlusion determination: In this embodiment, the preset random occlusion threshold is: If the length threshold is 5, then the proportion of isolated high-difference points is... And maximum length The cause of the target unit's malfunction was determined to be not random obstruction.

[0099] Step 3: Fault diagnosis of asynchronous equipment: If step one finds that the cause of the target unit's failure is a fixed blockage, then the cause of the target unit's failure is not a non-synchronous equipment failure.

[0100] In conclusion, the cause of the target unit's failure was fixed obstruction.

[0101] Example 4 This embodiment provides a power anomaly diagnosis device for a single photovoltaic unit in a mountainous area. (See also...) Figure 3 The device is used to implement the aforementioned method for diagnosing power anomalies in a mountain photovoltaic single unit, comprising: The data acquisition unit is used to acquire multi-source correlation data of power analysis for each target unit of the mountain photovoltaic power station; the multi-source correlation data of power analysis includes the power sequence of each time period of the day, the power sequence of the same period of the historical sunny day, and the irradiance sequence of each time period of the day; The curve acquisition unit is used to acquire the daily power curve and the daily irradiance curve based on the daily time-by-time power sequence and the daily time-by-time irradiance sequence, respectively. The power anomaly determination unit is used to compare the daily power sequence of the target unit with the historical power sequence of the same period on sunny days to determine whether the target unit has an abnormal power reduction. The synchronization determination unit is used to determine the level of power-irradiance synchronization based on the daily power curve and daily irradiance curve of the target unit. The first fault condition determination unit is used to calculate the proportion of units in the mountain photovoltaic power station other than the target unit that have abnormal power reduction, and to determine the fault cause of the target unit in a high synchronization scenario based on the proportion of units and the preset weather boundary correlation coefficient. The second fault condition determination unit is used to calculate the DTW distance between the power curve and the irradiance curve of the day, and to obtain the fault degree and fault cause of the target unit under the low synchronicity scenario based on the DTW distance.

[0102] The same or similar labels correspond to the same or similar parts; The terms used to describe positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting the invention. Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area, characterized in that, Includes the following steps: S1: Obtain multi-source correlation data for power analysis of each target unit in the mountain photovoltaic power station; the multi-source correlation data for power analysis includes the daily time-period power sequence, the historical same-day sunny day power sequence, and the daily time-period irradiance sequence; S2: Obtain the daily power curve and daily irradiance curve based on the daily time-by-time power sequence and the daily time-by-time irradiance sequence, respectively; S3: Compare the daily power sequence of the target unit with the historical power sequence of the same period on sunny days to determine whether there is an abnormal decrease in power of the target unit. If so, proceed to S4; otherwise, the power fluctuation of the target unit is normal. S4: Based on the daily power curve and daily irradiance curve of the target unit, determine the level of power-irradiance synchronicity. If the synchronicity is high, proceed to S5; if the synchronicity is low, proceed to S6. S5: Calculate the percentage of units in the mountain photovoltaic power station other than the target unit that have abnormal power reduction, and determine the cause of the target unit's failure in a high synchronization scenario based on the percentage of units and the preset weather boundary correlation coefficient. S6: Calculate the DTW distance between the power curve and the irradiance curve of the day, and obtain the fault degree and cause of the target unit under the low synchronicity scenario based on the DTW distance.

2. The method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area according to claim 1, characterized in that, The daily time-period power sequence is represented as follows: The daily time-period irradiance sequence is represented as follows: , and The target unit is the kth unit. T An array consisting of power and irradiance at each acquisition time. and These are the kth generating units on that day. Power and irradiance at each sampling time, T Indicates the number of sampling points within the collection period; Based on the power sequence of each time period of the day, obtain the power curve of the day with the sampling point as the x-axis and the power as the y-axis; Based on the daily irradiance sequence, a daily irradiance curve is obtained with the sampling point as the x-axis and irradiance as the y-axis.

3. The method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area according to claim 2, characterized in that, In S3, the daily time-by-time power sequence of the target unit is compared with the historical power sequence for the same period on sunny days to determine whether there is an abnormal decrease in power of the target unit. The process is as follows: S31: Based on the daily time-period power sequence Power sequence of sunny days in the same period of history Calculate the first one respectively k The average daily power of the target unit Average power of sunny days in the same period of history ; S32: Calculate the statistical relative threshold based on the historical sunny-day power series for the same period. ; S33: If If the power output of the target unit is abnormally low, then the power output of the target unit is abnormally low; otherwise, the power fluctuation of the target unit is normal.

4. The method for diagnosing power anomalies in a mountain photovoltaic single unit according to claim 2, characterized in that, In S4, based on the target unit's daily power curve and daily irradiance curve, the degree of power-irradiance synchronicity is determined. The process is as follows: The synchronicity of the daily power curve and daily irradiance curve of the target unit is calculated using the Pearson correlation coefficient, and the synchronicity judgment threshold is obtained. The expression is: in, This represents the average power sequence for each time period of the day. This represents the average value of the daily irradiance sequence for each time period. Compare to preset correlation thresholds With the aforementioned synchronization determination threshold The size, if ≥ If the synchronization is high, then the synchronization is low; otherwise, the synchronization is low.

5. The method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area according to claim 1, characterized in that, In S5, the percentage of units in the mountain photovoltaic power station other than the target unit that experience abnormal power reduction is calculated. Based on this percentage and a preset weather boundary correlation coefficient, the cause of the target unit's failure in a high-synchronization scenario is determined. The process is as follows: Statistics on the set of other units in the mountain photovoltaic power station besides the target unit. Number of units with abnormally low power And calculate the number of units with abnormally low power. The proportion of units The expression is: ; S This refers to the set of all units in a mountain photovoltaic power station other than the target unit. ... ,in, For the number of other units, and ; Compare the correlation coefficients of preset weather boundaries and the proportion of generating units The size, if > In scenarios with high synchronicity, the cause of the target unit's failure is weather-related; if In scenarios with high synchronicity, the cause of failure in the target unit is generalized equipment aging.

6. The method for diagnosing power anomalies in a single photovoltaic unit in a mountainous area according to claim 1, characterized in that, The process of S6 is as follows: Calculate the DTW distance between the power curve and the irradiance curve for the day; The optimal matching path for DTW is extracted based on DTW distance, and the morphological similarity is calculated based on DTW distance. The degree of failure of the target unit in the low synchronization scenario is quantified based on the morphological similarity. Quantization features are obtained based on the optimal matching path of DTW. The causes of failures in target units under low synchronization scenarios are determined based on the quantitative characteristics.

7. The method for diagnosing power anomalies in a mountain photovoltaic single unit according to claim 6, characterized in that, Based on the morphological similarity metric, the fault degree of the target unit in a low synchronization scenario is quantified. The process is as follows: C1: Normalize the daily power sequence and daily irradiance sequence of the target unit for each time period to obtain the normalized power value. and normalized value of irradiance ; C2: Based on the power normalization value and the normalized value of the irradiance Calculate the element values ​​of the distance matrix And construct the distance matrix The element values ​​of the distance matrix The expression is: in, Indicates the order of sampling points. Indicates the number of sampling points; C3: Based on distance matrix Construct the cumulative cost matrix based on the cumulative cost recursive formula. The cumulative cost recursive formula is as follows: when or hour, ; C4: Based on the cumulative cost matrix endpoint value The DTW distance between the daily power curve and the daily irradiance curve According to the cumulative cost matrix The optimal matching path and its length L are determined by backtracking the cumulative cost matrix. get; The DTW distance Convert to morphological similarity The expression is: Where max(D) is the distance matrix. The maximum value, L The optimal matching path length for DTW; based on morphological similarity As a reference value for the degree of failure of the target unit The closer to 0, the higher the degree of failure. The closer it is to 1, the lower the degree of failure.

8. The method for diagnosing power anomalies in a mountain photovoltaic single unit according to claim 6, characterized in that, Based on the DTW optimal matching path, the quantized features are obtained through the following process: D1: Construct a difference sequence based on the DTW optimal matching path. Morphological difference values Satisfying the expression: in, , These are the 1st, 2nd, and 3rd digits of the optimal matching path in DTW. The sampling point number of the daily time-period power sequence and daily time-period irradiance sequence corresponding to each matching location; , These are the normalized power value and normalized irradiance value corresponding to the matching point in the DTW optimal matching path, respectively. D2: Calculate the dynamic difference threshold based on the difference sequence, and then calculate the morphological difference values ​​in the difference sequence. Points exceeding the dynamic difference threshold are designated as high difference points, and the intervals occupied by high difference points are designated as high difference intervals. D3: The maximum length of consecutive high-difference intervals in a statistically differential sequence. And calculate the proportion of consecutive high-difference intervals. The expression is: ; D4: Based on the dynamic difference threshold, count the number of isolated high difference points among high difference points. And calculate the proportion of isolated high-difference points. The expression is: The isolated high-difference point is a point that is itself greater than the dynamic difference threshold and whose adjacent points are all less than the dynamic difference threshold. D5: Calculate the coefficient of variation of the differentially expressed sequences. The expression is: ,in, The standard deviation of the differentially expressed series is represented by the standard deviation of the series. This represents the mean of the differentially expressed sequences; D6: The maximum length of a continuous high-difference interval , percentage of consecutive high difference intervals The proportion of isolated high-difference points and coefficient of variation As a quantitative feature.

9. A method for diagnosing power anomalies in a mountain photovoltaic single unit according to claim 8, characterized in that, The causes of failures in target units under low synchronicity scenarios are determined based on the quantified characteristics, including: If the proportion of consecutive high difference intervals Greater than or equal to the preset fixed occlusion threshold and the coefficient of variation If the value is greater than or equal to the preset variation threshold, the cause of the target unit's failure is fixed blockage. If the proportion of isolated high-difference points The maximum length of a continuous high-difference interval that is greater than or equal to a preset random occlusion threshold. If the length is less than the preset threshold, the cause of the target unit's failure is random obstruction; If the cause of the target unit's failure is neither a fixed blockage nor a random blockage, then the cause of the target unit's failure is an asynchronous equipment failure.

10. A power anomaly diagnosis device for a mountain photovoltaic single-unit, used to implement the power anomaly diagnosis method for a mountain photovoltaic single-unit as described in any one of claims 1-9, characterized in that, include: The data acquisition unit is used to acquire multi-source correlation data of power analysis for each target unit of the mountain photovoltaic power station; the multi-source correlation data of power analysis includes the power sequence of the current day, the power sequence of the same period in history on sunny days, and the irradiance sequence of the current day. The curve acquisition unit is used to acquire the daily power curve and the daily irradiance curve based on the daily time-by-time power sequence and the daily time-by-time irradiance sequence, respectively. The power anomaly determination unit is used to compare the daily power sequence of the target unit with the historical power sequence of the same period on sunny days to determine whether the target unit has an abnormal power reduction. The synchronization determination unit is used to determine the level of power-irradiance synchronization based on the daily power curve and daily irradiance curve of the target unit. The first fault condition determination unit is used to calculate the proportion of units in the mountain photovoltaic power station other than the target unit that have abnormal power reduction, and to determine the fault cause of the target unit in a high synchronization scenario based on the proportion of units and the preset weather boundary correlation coefficient. The second fault condition determination unit is used to calculate the DTW distance between the power curve and the irradiance curve of the day, and to obtain the fault degree and fault cause of the target unit under the low synchronicity scenario based on the DTW distance.

Images