Program, information processing method, and detection device

Abstract

We provide programs and other tools that can detect abnormalities in solar cell modules with higher accuracy. [Solution] The program acquires time-series current data between the solar cell module and the power conditioner, acquires multiple first frequency conversion data sets with different time zones based on the acquired time-series current data, acquires second frequency conversion data based on a predetermined frequency data set in the multiple first frequency conversion data sets, and causes the computer to perform a process to detect an abnormality in the solar cell module based on the second frequency conversion data.

Description

[Technical Field]

[0001] The present invention relates to a program, an information processing method, and a detection device. [Background Art]

[0002] Means for detecting an abnormality in a solar cell module has been proposed. Patent Document 1 proposes means for detecting an abnormality in a solar cell module by focusing on the magnitude of a harmonic of a specific frequency superimposed on a direct current. [Prior Art Documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Unexamined Patent Publication No. 2023-81332 [Summary of the Invention] [Problem to be Solved by the Invention]

[0004] There is a demand for means for detecting an abnormality in a solar cell module with higher accuracy. An object of the present disclosure is to provide a program and the like that detect an abnormality in a solar cell module with higher accuracy. [Means for Solving the Problem]

[0005] A program according to the present disclosure acquires time-series current data between a solar cell module and a power conditioner, acquires a plurality of first frequency conversion data respectively corresponding to mutually different time zones based on the acquired time-series current data, acquires second frequency conversion data based on a data group of a predetermined frequency in the plurality of first frequency conversion data, and causes a computer to execute processing for detecting an abnormality in the solar cell module based on the second frequency conversion data. [Effect of the Invention]

[0006] According to the present disclosure, it is possible to detect an abnormality in a solar cell module with higher accuracy. [Brief explanation of the drawing]

[0007] [Figure 1] This is a schematic diagram of a method for detecting abnormalities in solar cell modules using a detection device. [Figure 2] This is a block diagram showing an example of the configuration of a detection device. [Figure 3] This is an explanatory diagram illustrating a method for detecting abnormalities in solar cell modules using a detection device. [Figure 4] This graph shows the time-series current data acquired by the measurement unit. [Figure 5] This is an explanatory diagram illustrating how to acquire multiple time-series data. [Figure 6] This is a graph showing the first frequency conversion data. [Figure 7] This is an explanatory diagram showing a method for acquiring signal strength related to the generation of a data set of predetermined frequencies. [Figure 8] This is an explanatory diagram illustrating a method for generating data sets of a predetermined frequency. [Figure 9] This graph shows the signal intensity range at a given frequency for a normal solar cell module. [Figure 10] This graph shows the signal intensity group at a predetermined frequency for a solar cell module where an anomaly occurred. [Figure 11] This graph shows the second frequency conversion data for a normal solar cell module. [Figure 12] This graph shows the second frequency conversion data for a solar cell module that experienced an anomaly. [Figure 13] This is a flowchart illustrating the processing steps taken by the processing unit in detecting anomalies in solar cell modules. [Modes for carrying out the invention]

[0008] Programs, information processing methods, and detection devices according to embodiments of this disclosure will be described below with reference to the drawings. This disclosure is not limited to these examples, but is indicated by the claims and is intended to include all modifications in the sense and scope equivalent to the claims. At least some of the embodiments described below may be combined as desired.

[0009] (First Embodiment) Figure 1 is a schematic diagram of a method for detecting abnormalities in a solar cell module using a detection device. Solar cell module 1 is a power generation module including a solar panel. Power conditioner (Power Conditioning Subsystem: PCS) 2 is an inverter that converts the DC current generated by solar cell module 1 into AC current. Solar cell module 1 and PCS 2 are electrically connected by cable 11, which is a power transmission cable. Power system 21 is a system that includes power plants, substations, transmission lines, or distribution lines. Generally, the term "power system" refers to the entire system that handles everything from power production to consumption, but in this embodiment, power system 21 does not necessarily have to be the entire system and may be a part of the above system. PCS 2 may also be connected to a residential distribution board or household electrical appliances in addition to power system 21.

[0010] PCS2 is connected to the power grid 21 and transmits the converted AC current to the power grid 21. Detection device 3 is a device that detects abnormalities in the solar cell module 1. Detection device 3 is connected to the cable 11 between the solar cell module 1 and PCS2. Detection device 3 detects abnormalities in the solar cell module 1 by measuring the current flowing through the cable 11. The detection device 3 can be connected anywhere between the solar cell module 1 and PCS2, as long as it is a location where the DC current value can be measured. In the following, it is assumed that PCS2 is connected to the power grid 21 carrying 60Hz AC. PCS2 may also be connected to the power grid 21 carrying AC of a frequency other than 60Hz.

[0011] Fig. 2 is a block diagram showing a configuration example of the detection device 3. The detection device 3 comprises a measurement unit 31, a storage unit 32, a processing unit 33, a display unit 34, and an operation unit 35. The detection device 3 may be provided with a reader that can read the recording medium 4 or a communication unit capable of communicating with an external server. Each part of the detection device 3 is connected via a bus.

[0012] The measurement unit 31 is a current sensor that measures the current between the solar cell module 1 and the PCS 2. For example, the measurement unit 31 may be a known Hall-type current sensor including a magnetoelectric conversion element such as a Hall element, but is not limited thereto as long as it is a current sensor capable of measuring direct current. The measurement unit 31 is connected to any position of the cable 11 in a mode capable of measuring the current value of the current flowing through the cable 11. The measurement unit 31 transmits the measured current value to the processing unit 33.

[0013] The storage unit 32 includes memory elements such as RAM (Random Access Memory) or ROM (Read Only Memory), and stores a control program P, data or the like necessary for the processing unit 33 to execute processing. The control program P may be written into the storage unit 32 in the manufacturing stage of the detection device 3, or the detection device 3 may obtain the control program P distributed by an external server through communication via the communication unit and store it in the storage unit 32. The control program P may be obtained by a reader reading the control program 41P, which is readable recorded on a computer-readable recording medium 4 such as a magnetic disk, an optical disk, or a semiconductor memory, from the recording medium 4 and storing it in the storage unit 32. The storage unit 32 may be included in the processing unit 33.

[0014] The processing unit 33 includes an arithmetic processing device such as a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), a GPU (Graphics Processing Unit), an FPGA (Field Programmable Gate Array), a DSP (Digital Signal Processor), or a quantum processor. The processing unit 33 reads the control program P stored in the storage unit 32 and executes processing for detecting an abnormality in the solar cell module 1.

[0015] The display unit 34 is, for example, a display device such as a liquid crystal panel or an organic EL (Electro Luminescence) display. The display unit 34 displays information transmitted from the processing unit 33 so that a user can visually understand the information. The operation unit 35 is, for example, an input device such as a hardware keyboard, a pointing device, or a touch panel. A user of the detection device 3 can input arbitrary information to the detection device 3 using the operation unit 35. The operation unit 35 may be configured integrally with the display unit 34.

[0016] The detection device 3 may be configured such that the measurement unit 31 includes a communication unit, transmits the current value measured by the measurement unit 31 to an information processing device provided outside the detection device 3, and the information processing device executes processing for detecting an abnormality in the solar cell module 1. The information processing device may be a single computer, may be configured with a plurality of computers to perform distributed processing, may be realized by a plurality of virtual machines provided in a single server, or may be realized using a cloud server.

[0017] A method for detecting abnormalities in the solar cell module 1 using the detection device 3 will be described. Figure 3 is an overview diagram of the method for detecting abnormalities in the solar cell module 1 using the detection device 3. The detection device 3 acquires time-series current data between the solar cell module 1 and the PCS2. From the acquired time-series current data, the detection device 3 generates time-series data for multiple different time periods T1, T2, etc. The detection device 3 acquires multiple first frequency-converted data for mutually different time periods by performing frequency conversion on each of the generated time-series data. The detection device 3 acquires a data group of a predetermined frequency from the multiple first frequency-converted data. Based on the data group of the predetermined frequency, the detection device 3 acquires second frequency-converted data. Based on the second frequency-converted data, the detection device 3 detects abnormalities in the solar cell module 1.

[0018] In this embodiment, "frequency conversion" includes decomposing a signal into its frequency components. The frequency conversion can be, for example, a Fast Fourier Transform (FFT), but is not limited to this.

[0019] Figure 4 is a graph showing the time-series current data acquired by the measurement unit 31. In Figure 4, the horizontal axis represents time (milliseconds: ms), and the vertical axis represents the magnitude of the current (amperes: A). The measurement unit 31 of the detection device 3 measures the current value between the solar cell module 1 and the PCS2 for an arbitrary period of time and acquires time-series current data between the solar cell module 1 and the PCS2. The measurement time for the current value may be any time from 0.1 seconds to 3 seconds, for example, but is not limited to this.

[0020] A direct current generated by the solar cell module 1 flows through cable 11. The direct current flowing through cable 11 is superimposed mainly on harmonics originating from PCS2. Therefore, as shown in Figure 4, even when the solar cell module 1 is functioning normally, the direct current flowing between the solar cell module 1 and PCS2 is pulsating.

[0021] The processing unit 33 acquires time-series current data from the measurement unit 31. Based on the acquired time-series current data, the processing unit 33 acquires multiple first frequency conversion data sets for different time periods. In order to acquire multiple first frequency conversion data sets from one time-series current data set, the processing unit 33 generates multiple time-series data sets by extracting the time-series current data for each time period T.

[0022] Figure 5 is an explanatory diagram illustrating how to acquire multiple time-series data. The horizontal axis in Figure 5 is the same as the horizontal axis in Figure 4 and represents time (ms).

[0023] The processing unit 33 acquires time-series data for multiple different time periods by shifting the time period T by predetermined units. Below is an example of how to acquire time-series data for multiple different time periods when the length of the time period T is 400ms. For example, as shown in Figure 5, the time period T is shifted by predetermined units of 1ms, and multiple time periods T are set so that each overlaps with the previous time period T by 399ms. The processing unit 33 acquires time-series data from time-series current data for each set time period T, thereby acquiring time-series data for multiple different time periods.

[0024] In this example, the first time period T1 corresponds to the time period from 0 to 399ms of the time-series current data. The second time period T2 corresponds to the time period from 1 to 400ms of the time-series current data. The first time period T1 and the second time period T2 overlap in the time period from 1 to 399ms. As shown in Figure 3, the processing unit 33 acquires multiple time-series data for each time period T. In this example, if the measurement unit 31 measures the current value for 0.5 seconds, the processing unit 33 can acquire up to 101 time-series data (T1 to T101) from different time periods.

[0025] By shifting the time period T by a predetermined unit and partially overlapping the time period T, the processing unit 33 can obtain more time-series data even if the length of the time period T remains the same. The predetermined unit for shifting the time period T and the length of the time period T are not limited to those described above. The predetermined unit for shifting the time period T may be set to 0, eliminating the overlap of the time period T.

[0026] Figure 6 is a graph showing the first frequency conversion data. In Figure 6, the horizontal axis represents frequency (Hertz: Hz), and the vertical axis represents signal strength (decibels: dB). The processing unit 33 obtains multiple first frequency conversion data by performing frequency conversion on each generated time series data. That is, the processing unit 33 can obtain multiple mutually different first frequency conversion data, up to the number of generated time series data.

[0027] To obtain the graph in Figure 6, the processing unit 33 first acquires time-series current data obtained by the measurement unit 31 when it measures the current of the solar cell module 1 for 2 seconds. From the time-series current data, the processing unit 33 acquires 1501 time-series data points, with a time period T of 500 ms and a predetermined shift unit of 1 ms. The processing unit 33 acquires multiple first frequency conversion data points by performing an FFT once on each time-series data point. The graph in Figure 6 shows one of the acquired first frequency conversion data points.

[0028] 120Hz, 240Hz, 360Hz, and 480Hz are frequencies that correspond to 2 times, 4 times, 6 times, and 8 times the frequency of the power system 21 (60Hz) in this embodiment. In Figure 6, the signal strengths of specific frequencies in the first frequency conversion data, particularly 120Hz, 240Hz, 360Hz, and 480Hz, which are integer multiples of the AC frequency of the power system 21, show higher values ​​than the signal strengths of surrounding frequencies.

[0029] The processing unit 33 acquires second frequency conversion data based on a predetermined frequency data set in a plurality of acquired first frequency conversion data. Therefore, having acquired the first frequency conversion data, the processing unit 33 generates a predetermined frequency data set from the plurality of first frequency conversion data in order to acquire the second frequency data set.

[0030] The method for generating a data set of a predetermined frequency will be described below. Figure 7 is an explanatory diagram showing the method for obtaining signal strength related to the generation of a data set of a predetermined frequency. In Figure 7, the horizontal axis represents frequency (Hz), and the vertical axis represents signal strength (dB). The predetermined frequency in Figure 7 is assumed to be 120 Hz.

[0031] The explanatory diagram in Figure 7 is a schematic diagram of one first frequency conversion data as shown in Figure 6. The processing unit 33 derives the signal strength at a predetermined frequency for each acquired first frequency conversion data. In this example, the signal strength S at the predetermined frequency is the signal strength at 120 Hz.

[0032] Figure 8 is an explanatory diagram illustrating the method for generating a data set of predetermined frequencies. In Figure 8, the horizontal axis represents time (ms), and the vertical axis represents signal strength (dB). The predetermined frequency signal strengths S1, 2, 3... in Figure 8 represent the predetermined frequency signal strength S in the first frequency-converted data obtained by frequency-converting the time-series data of time zones T1, 2, 3..., respectively. The processing unit 33 generates a data set of predetermined frequencies by arranging the predetermined frequency signal strengths S based on the time series of time zone T, as shown in Figure 8. The time intervals of the predetermined frequency signal strengths S may be arranged at equal intervals for predetermined time periods related to the shift of time zone T, but is not limited to this. Hereinafter, the predetermined frequency data set will be referred to as the "signal strength group".

[0033] When acquiring the signal strength group, the predetermined frequency may be any frequency. Preferably, the predetermined frequency is an integer multiple of the frequency of the power system 21. More preferably, the predetermined frequency is 2n times the frequency of the power system 21 (where n is a natural number). Even more preferably, the predetermined frequency is twice the frequency of the power system 21. In this embodiment, since PCS2 is connected to a 60Hz power system 21, the predetermined frequency below will be 120Hz or 240Hz. Only one predetermined frequency needs to be set; it is not necessary to set multiple frequencies.

[0034] Figure 9 is a graph showing the signal strength group at a predetermined frequency for a normal solar cell module 1. In Figure 9, the horizontal axis represents time (ms), and the vertical axis represents signal strength (dB). The graph showing the signal strength group in Figure 9 is based on the signal strength group derived by the processing unit 33, which acquires the signal strength at 120Hz and 240Hz for each first frequency conversion data of a normal solar cell module 1. In Figure 9, the solid line shows the signal strength group at 120Hz, and the dashed line shows the signal strength group at 240Hz.

[0035] Figure 10 is a graph showing the signal strength group at a predetermined frequency for the solar cell module 1 where an anomaly occurred. In Figure 10, the horizontal axis represents time (ms), and the vertical axis represents signal strength (dB). The graph showing the signal strength group in Figure 10 is based on the signal strength group derived by the processing unit 33, which acquires the signal strength at 120Hz and 240Hz for each first frequency conversion data of the solar cell module 1 where the anomaly occurred. In Figure 10, the solid line shows the signal strength group at 120Hz, and the dashed line shows the signal strength group at 240Hz.

[0036] The processing unit 33 acquires second frequency-converted data by performing frequency conversion on a signal intensity group that includes the signal intensity of a predetermined frequency in a plurality of acquired first frequency-converted data.

[0037] Figure 11 is a graph showing the second frequency conversion data of a normal solar cell module 1. In Figure 11, the horizontal axis represents frequency (Hz), and the vertical axis represents signal strength (dB). The graph in Figure 11 is based on the second frequency conversion data obtained by the processing unit 33 performing one FFT on the signal strength group at a predetermined frequency of 120 Hz obtained in Figure 9. As shown in Figure 11, the second frequency conversion data of a normal solar cell module 1 shows higher signal strengths than the surrounding frequencies at frequencies that are integer multiples of the predetermined frequency of 120 Hz, namely 120 Hz, 240 Hz, 360 Hz, and 480 Hz.

[0038] Figure 12 is a graph showing the second frequency conversion data of the solar cell module 1 where the abnormality occurred. In Figure 12, the horizontal axis represents frequency (Hz), and the vertical axis represents signal strength (dB). The graph in Figure 12 is based on the second frequency conversion data obtained by the processing unit 33 performing one FFT on the signal strength group at the predetermined frequency of 120 Hz obtained in Figure 10. As shown in Figure 12, the second frequency conversion data of the solar cell module 1 where the abnormality occurred shows higher signal strength than the surrounding frequencies at 120 Hz, 240 Hz, 360 Hz, and 480 Hz, which are integer multiples of the predetermined frequency of 120 Hz, similar to Figure 11.

[0039] In Figure 12, in addition to integer multiples of the predetermined frequency of 120 Hz, integer multiples of 40 Hz, including 40 Hz and 80 Hz, also show higher signal strength than the surrounding frequencies. As described above, the second frequency conversion data of the malfunctioning solar cell module 1 shows higher signal strength than the surrounding frequencies at specific frequencies that are different from integer multiples of the predetermined frequency and are not seen in the second frequency conversion data of a normal solar cell module 1. In other words, the second frequency conversion data of the malfunctioning solar cell module 1 shows a different waveform pattern from the second frequency conversion data of a normal solar cell module 1.

[0040] The processing unit 33 can detect abnormalities in the solar cell module 1 based on differences in the waveform patterns of the second frequency conversion data. In particular, the processing unit 33 can detect abnormalities in the solar cell module 1 based on the second frequency conversion data within a predetermined range from a predetermined frequency. For example, the processing unit 33 may determine whether or not an abnormality has occurred in the solar cell module 1 by focusing on the signal intensity of frequencies other than integer multiples of the predetermined frequency within a predetermined range of 120Hz, i.e., within the range of 0Hz to 240Hz, which is a predetermined range of the predetermined frequency 120Hz. The predetermined range of the predetermined frequency is not limited to the above.

[0041] The following describes a method for detecting abnormalities in the solar cell module 1 based on second frequency conversion data from the processing unit 33. The user first acquires reference data, which is second frequency conversion data of a normal solar cell module 1, and stores the reference data in the storage unit 32. The processing unit 33 acquires the difference in signal intensity at each frequency between the reference data and the acquired second frequency conversion data. If there is a frequency at which the difference exceeds a predetermined threshold, at a frequency that is not an integer multiple of a predetermined frequency, the processing unit 33 detects an abnormality in the solar cell module 1.

[0042] The method for detecting abnormalities in the solar cell module 1 based on the second frequency conversion data by the processing unit 33 is not limited to the above. Modified examples are described below. (Variation 1) The user sets a threshold for signal strength in advance and stores the threshold in the storage unit 32. If the acquired second frequency conversion data includes frequencies that show a signal strength above the threshold at frequencies that are not integer multiples of a predetermined frequency, the processing unit 33 detects an abnormality in the solar cell module 1. (Modification 2) The user sets a threshold for signal strength in the graph of the second frequency conversion data. The processing unit 33 determines that one planar region in the graph of the second frequency conversion data that shows a signal strength exceeding the threshold is one peak. The user has previously measured the number of peaks in the graph of the reference data and stored it in the storage unit 32. If the number of peaks in the acquired graph of the second frequency conversion data exceeds the number of peaks in the graph of the reference data, the processing unit 33 detects an abnormality in the solar cell module 1. (Variation 3) By applying machine learning-based pattern recognition to the graph showing the second frequency conversion data, the processing unit 33 determines whether the waveform patterns of the reference data and the acquired second frequency conversion data are the same or different. If it determines that the waveform patterns are different, the processing unit 33 detects an abnormality in the solar cell module 1. For example, the machine learning model can be an autoencoder trained on the reference data, a deep learning model (CNN, LSTM, or transformer, etc.) that outputs whether or not there is an abnormality when second frequency conversion data is input, or a support vector machine.

[0043] The processing unit 33 may perform the above processing on second frequency conversion data within a predetermined range from a predetermined frequency, or it may perform the above processing on the entire frequency of the acquired second frequency conversion data without limiting the range. The processing unit 33 may also perform abnormality detection of the solar cell module 1 based on the second frequency conversion data by a method other than the above.

[0044] If an abnormality is detected, the processing unit 33 outputs an output (graph) to the display unit 34 comparing the acquired second frequency conversion data with the second frequency conversion data acquired under normal conditions. The display unit 34, upon receiving the output, displays a graph comparing the second frequency conversion data acquired by the processing unit 33 with the second frequency conversion data acquired under normal conditions. As a comparison graph, the processing unit 33 may output a graph in which the second frequency conversion data acquired under normal conditions and the second frequency conversion data when an abnormality is detected are superimposed on the same plane with frequency on the horizontal axis and signal strength on the vertical axis. Alternatively, the processing unit 33 may output a graph in which the second frequency conversion data acquired under normal conditions and the second frequency conversion data when an abnormality is detected are displayed side by side. By viewing this display, the user of the detection device 3 can recognize that an abnormality has occurred in the solar cell module 1. The user who recognizes that an abnormality has occurred in the solar cell module 1 may repair the solar cell module 1 or perform a more detailed inspection of the abnormality in the solar cell module 1. In addition to displaying the information on the display unit 34, the user of the detection device 3 may be notified of any abnormality in the solar cell module 1 by means of an audible warning or by sending an email to an external computer or smartphone or other terminal.

[0045] Figure 13 is a flowchart illustrating the processing of the processing unit 33 in the detection of anomalies in the solar cell module 1. Before performing anomaly detection of the solar cell module 1, the user of the detection device 3 sets a predetermined frequency.

[0046] The user of the detection device 3 connects the measurement unit 31 to the cable 11. The measurement unit 31 measures the current flowing through the cable 11 for a certain period of time and acquires time-series current data. The measurement unit 31 transmits the acquired time-series current data to the processing unit 33. The processing unit 33 receives the time-series current data (step S101). The processing unit 33 acquires time-series data for multiple different time periods from the received time-series current data (step S102). The processing unit 33 performs frequency conversion based on the received time-series data (step S103). The processing unit 33 acquires multiple first frequency conversion data for different time periods (step S104). The processing unit 33 derives the signal strength at a predetermined frequency for each first frequency conversion data (step S105). The processing unit 33 acquires a signal strength group for the derived signal strength (step S106). The processing unit 33 performs frequency conversion on the acquired signal strength group (step S107). The processing unit 33 acquires the second frequency conversion data (step S108).

[0047] The processing unit 33 detects an abnormality in the solar cell module 1 based on second frequency conversion data within a predetermined range from a predetermined frequency (step S109). If an abnormality is detected (step S109: YES), the processing unit 33 outputs an abnormality indication to the display unit 34 (step S110). The abnormality indication may be, for example, a screen display that combines an output (graph) comparing the second frequency conversion data acquired by the processing unit 33 with the second frequency conversion data acquired in advance when the system is functioning normally, with the text "There is an abnormality in the solar cell module 1." The abnormality indication is not limited to the above.

[0048] If no abnormality is detected (step S109: NO), the processing unit 33 outputs a "no abnormality" message to the display unit 34 (step S111). The "no abnormality" message may be, for example, a screen display that combines an output (graph) comparing the second frequency conversion data acquired by the processing unit 33 with the second frequency conversion data acquired in advance under normal conditions, with the text "There is no abnormality in solar cell module 1." The "no abnormality" message is not limited to the above. If the display unit 34 receives an output from the processing unit 33 indicating an abnormality or no abnormality in step S110 or step S111, it displays the received output.

[0049] Based on the above, it is possible to realize a program that can detect abnormalities in the solar cell module 1 with higher accuracy.

[0050] (Second Embodiment) In the second embodiment, a method for detecting an abnormality in the solar cell module 1 based on a signal intensity group will be described. Content that overlaps with the description in the first embodiment will be omitted. As shown in Figures 9 and 10, the pattern of the signal intensity group differs depending on the state of the solar cell module 1. For example, the signal intensity group of the solar cell module 1 where an abnormality has occurred in Figure 10 exhibits characteristics not seen in the signal intensity group of a normal solar cell module 1 in Figure 9, such as a significant decrease in signal intensity around 900ms at 120Hz and periodic increases and decreases in signal intensity at 240Hz. Therefore, the detection device 3 can detect an abnormality in the solar cell module 1 based on the signal intensity group, which is a data group of predetermined frequencies in a plurality of first frequency conversion data.

[0051] The following describes a method for detecting abnormalities in the solar cell module 1 based on the signal intensity group by the processing unit 33. The processing unit 33 calculates the standard deviation of each signal intensity in the signal intensity group, and if the calculated standard deviation is greater than or equal to a threshold, the processing unit 33 detects an abnormality in the solar cell module 1. The threshold is preferably determined based on the standard deviation of each signal intensity in the signal intensity group of a normal solar cell module 1.

[0052] The method for detecting abnormalities in the solar cell module 1 based on the signal intensity group by the processing unit 33 is not limited to the above. Modified examples are described below. (Variation 1) The processing unit 33 calculates the difference between the maximum and minimum signal strengths in the signal strength group (maximum signal strength - minimum signal strength), and if the calculated difference is greater than or equal to a threshold, it detects an abnormality in the solar cell module 1. The threshold is preferably determined based on the difference between the maximum and minimum signal strengths in the signal strength group of a normal solar cell module 1. (Modification 2) The processing unit 33 calculates the ratio of the maximum signal strength to the minimum signal strength in the signal strength group (maximum signal strength / minimum signal strength), and if the calculated ratio is greater than or equal to a threshold, it detects an abnormality in the solar cell module 1. The threshold is preferably determined based on the ratio of the maximum signal strength to the minimum signal strength in the signal strength group of a normal solar cell module 1. (Variation 3) By applying machine learning-based pattern recognition to a graph showing signal intensity groups, the processing unit 33 determines the similarity or difference between the graph patterns of the signal intensity groups of a normal solar cell module 1 and those of a solar cell module 1 with an abnormality. If it determines that the graph patterns are different, the processing unit 33 detects an abnormality in the solar cell module 1. For example, the machine learning model can be an autoencoder trained on the graph of signal intensity groups of a normal solar cell module 1, a deep learning model (CNN, LSTM, or transformer, etc.) that outputs whether or not there is an abnormality when the signal intensity group of the solar cell module 1 to be inspected is input, or a support vector machine.

[0053] An autoencoder trained on a graph of signal intensity data from a normal solar cell module 1 will output a graph similar to the input graph when a graph of signal intensity data from a normal solar cell module 1 is input. Therefore, when a graph of signal intensity data from a normal solar cell module 1 is input to the autoencoder, the difference between the input and output is small. On the other hand, when a graph of signal intensity data from a malfunctioning solar cell module 1 is input, the difference between the input and output becomes large. If the difference between the input and output is greater than a predetermined threshold, the processing unit 33 can determine that a malfunction has occurred in the solar cell module 1. Even if there is little data on the signal intensity data from the malfunctioning solar cell module 1 and the deep learning model cannot be sufficiently trained, the malfunction in the solar cell module 1 can be detected by the anomaly detection performed by the autoencoder.

[0054] If there is sufficient data on the signal strength of solar cell module 1 when an anomaly occurs, the anomaly in solar cell module 1 can be detected using a deep learning model that has been trained on the signal strength data of both a normal solar cell module 1 and an anomalyed solar cell module 1.

[0055] The processing unit 33 may perform the above processing on signal intensity groups within a predetermined time range, or on signal intensity groups over the entire time range. The processing unit 33 may also perform abnormality detection of the solar cell module 1 based on the signal intensity groups using methods other than those described above.

[0056] The method for detecting abnormalities in the solar cell module 1 will be explained with reference to Figure 13. Steps S101 to S106 are the same as in the first embodiment, so the explanation will be omitted. The processing unit 33 detects an abnormality in the solar cell module 1 based on the signal intensity group acquired in step S106 (step S109). If an abnormality is detected (step S109: YES), the processing unit 33 outputs an indication of abnormality to the display unit 34 (step S110). When an abnormality is detected, the processing unit 33 outputs an output (graph) to the display unit 34 comparing the acquired data group of a predetermined frequency with the data group of a predetermined frequency acquired in advance when the system is functioning normally. The display unit 34, upon receiving the output, displays a graph comparing the data group of a predetermined frequency acquired by the processing unit 33 with the data group of a predetermined frequency acquired in advance when the system is functioning normally. As a comparison graph, the processing unit 33 may output a graph in which the data group of a predetermined frequency acquired in a normal state and the data group of a predetermined frequency when an abnormality is detected are superimposed on the same plane with the horizontal axis being time and the vertical axis being signal intensity. As a comparative graph, the processing unit 33 may output a graph that displays side by side a data set of predetermined frequencies acquired in advance during normal operation and a data set of predetermined frequencies when an abnormality is detected. If no abnormality is detected (step S109: NO), the processing unit 33 outputs "No abnormality" to the display unit 34 (step S111).

[0057] The processing unit 33 may, after performing anomaly detection of the solar cell module 1 based on the signal intensity group, acquire second frequency conversion data based on the signal intensity group and perform anomaly detection of the solar cell module 1 based on the second frequency conversion data. In this case, the processing unit 33 may finally detect an anomaly in the solar cell module 1 if it determines that there is an anomaly in the solar cell module 1 based on either one of the anomaly detection methods, or if it determines that there is an anomaly based on both methods.

[0058] Therefore, abnormalities in the solar cell module 1 can be detected with higher accuracy and more quickly. [Explanation of Symbols]

[0059] 1: Solar cell module, 2: Power conditioner (PCS), 3: Detection device, 4: Recording medium, 21: Power system, 31: Measurement unit, 33: Processing unit, P: Control program

Claims

1. We acquire time-series current data between the solar cell module and the power conditioner. Based on the acquired time-series current data, multiple first frequency conversion data sets with different time zones are acquired. Based on a set of data at a predetermined frequency in multiple first frequency conversion data, second frequency conversion data is acquired. Based on the second frequency conversion data, an abnormality in the solar cell module is detected. A program that instructs a computer to perform a process.

2. By shifting the time period of the acquired time-series current data by a predetermined unit, multiple time-series data for different time periods are generated. By performing frequency conversion on each generated time-series data, multiple first frequency-converted data sets are obtained. The program according to claim 1, which causes a computer to perform a process.

3. By performing frequency conversion on a predetermined frequency data set in multiple acquired first frequency conversion data sets, second frequency conversion data is obtained. The program according to claim 1, which causes a computer to perform a process.

4. Based on the second frequency conversion data within a predetermined range from the predetermined frequency, an abnormality in the solar cell module is detected. The program according to claim 1, which causes a computer to perform a process.

5. The predetermined frequency is an integer multiple of the power system frequency. The program according to claim 1.

6. If an anomaly is detected, the acquired second frequency conversion data will be compared with the second frequency conversion data acquired under normal conditions and output. The program according to claim 1, which causes a computer to perform a process.

7. We acquire time-series current data between the solar cell module and the power conditioner. Based on the acquired time-series current data, multiple first frequency conversion data sets with different time zones are acquired. An abnormality in the solar cell module is detected based on a data set of predetermined frequencies in multiple first frequency conversion data. A program that instructs a computer to perform a process.

8. By shifting the time period of the acquired time-series current data by a predetermined unit, multiple time-series data for different time periods are generated. By performing frequency conversion on each generated time-series data, multiple first frequency-converted data sets are obtained. The program according to claim 7, which causes a computer to perform a process.

9. An abnormality in the solar cell module is detected based on a data set of predetermined frequencies within a predetermined time range. The program according to claim 7, which causes a computer to perform a process.

10. The predetermined frequency is an integer multiple of the power system frequency. The program according to claim 7.

11. When an anomaly is detected, the acquired data set at a predetermined frequency is compared with the data set at a predetermined frequency previously acquired during normal operation, and the results are output. The program according to claim 7, which causes a computer to perform a process.

12. Based on a set of data at a predetermined frequency in multiple first frequency conversion data, second frequency conversion data is acquired. Based on the second frequency conversion data, an abnormality in the solar cell module is detected. The program according to claim 7, which causes a computer to perform a process.

13. We acquire time-series current data between the solar cell module and the power conditioner. Based on the acquired time-series current data, multiple first frequency conversion data sets with different time zones are acquired. Based on a set of data at a predetermined frequency in multiple first frequency conversion data, second frequency conversion data is acquired. Based on the second frequency conversion data, an abnormality in the solar cell module is detected. Information processing methods.

14. A detection device for detecting abnormalities in solar cell modules, The measurement unit, which measures time-series current data between the solar cell module and the power conditioner, acquires the time-series current data it has measured. Based on the acquired time-series current data, multiple first frequency conversion data sets with different time zones are acquired. Based on a set of data at a predetermined frequency in multiple first frequency conversion data, second frequency conversion data is acquired. Based on the second frequency conversion data, an abnormality in the solar cell module is detected. It includes a processing unit that performs the processing. Detection device.

15. We acquire time-series current data between the solar cell module and the power conditioner. Based on the acquired time-series current data, multiple first frequency conversion data sets with different time zones are acquired. An abnormality in the solar cell module is detected based on a data set of predetermined frequencies in multiple first frequency conversion data. Information processing methods.

16. We acquire time-series current data between the solar cell module and the power conditioner. Based on the acquired time-series current data, multiple first frequency conversion data sets with different time zones are acquired. An abnormality in the solar cell module is detected based on a data set of predetermined frequencies in multiple first frequency conversion data. It includes a processing unit that performs the processing. Detection device.

Images