Photovoltaic array monitoring method, device and equipment for high-altitude area and storage medium

Abstract

The invention discloses a photovoltaic array monitoring method, device and equipment in a high altitude area and a storage medium, and relates to the technical field of electrical digital data processing. According to the method, the characteristics of mutual contrast and reference are utilized, so that the reserved signals have no same type and are identified as abnormal characteristics, the misjudgment probability of photovoltaic panel faults is reduced, meanwhile, by randomly defining voltage signal acquisition point positions and acquiring voltage differences in pairs, the partial discharge point positions are acquired through multiple groups of voltage signal acquisition point positions, and the accuracy of partial discharge is improved. Compared with single-time detection, errors can be further reduced through multi-time obtaining of the method, and the positioning precision is improved. Compared with manual regular inspection and meter reading analysis, the photovoltaic panel fault and the fault position can be automatically judged, the fault position can be directly reached through the external aircraft, the inspection cost is reduced, and the fault positioning precision is improved.

Description

Technical Field

[0001] The present application relates to the technical field of electrical digital data processing, and in particular to a photovoltaic array monitoring method, device, equipment and storage medium in high altitude areas. Background Art

[0002] Photovoltaic power station is a facility that uses solar energy to generate electricity. It converts solar energy into electrical energy through special materials such as crystalline silicon panels and electronic components such as inverters, and is connected to the power grid to transmit electricity to the grid. Photovoltaic power stations are green energy projects encouraged by the state, and are clean, environmentally friendly and renewable. Photovoltaic power stations are widely used in various occasions, including providing power for areas without electricity, solar daily electronic products (such as solar chargers, solar street lights, etc.) and grid-connected power generation. With the advancement of technology and policy support, photovoltaic power generation will occupy an important position in the future energy structure.

[0003] The performance monitoring of photovoltaic arrays usually involves a series of manual appearance inspections to determine whether the external mechanical properties of the electrical equipment are complete, and manual meter reading and verification to determine whether the internal electronic performance of the electrical equipment is normal. Finally, a comprehensive judgment is made on the overall performance of the electrical equipment. For example, the appearance inspection is performed by checking the appearance of photovoltaic modules, brackets, cables and other components to confirm that there is no damage, deformation, corrosion, etc.; manual meter reading and verification is performed by using instruments to measure the electrical properties of photovoltaic modules, such as open-circuit voltage, short-circuit current, maximum power point voltage and current, etc., and EL is used to detect whether there are defects such as hidden cracks and black sheets inside the modules; inverter performance testing is performed by checking the operating status of the inverter, including output voltage, current, frequency, power factor and other parameters.

[0004] As can be seen from the above, the above performance monitoring method relies on manual inspection judgment and manual meter reading analysis, which inevitably adds subjective experience errors. In addition, photovoltaic arrays located in high-altitude areas are also subject to extreme conditions such as ice and snow, resulting in inaccurate monitoring results of photovoltaic arrays. Summary of the invention

[0005] The main purpose of the present application is to provide a photovoltaic array monitoring method, device, equipment and storage medium in high altitude areas to solve the problem of inaccurate photovoltaic array monitoring results in high altitude areas in the prior art.

[0006] In order to achieve the above objectives, this application provides the following technical solutions:

[0007] A photovoltaic array monitoring method in a high altitude area, wherein a plurality of photovoltaic arrays are installed in the high altitude area, each photovoltaic array has a plurality of photovoltaic strings connected in series, each photovoltaic string has a plurality of photovoltaic panels, and the photovoltaic array monitoring method comprises:

[0008] Step S1, acquiring a plurality of current signals of each photovoltaic panel based on a preset time period, and extracting a plurality of signal components of each current signal by empirical mode decomposition;

[0009] Step S2, screening out all abnormal signal components from all signal components through a classification algorithm;

[0010] Step S3, respectively obtaining the photovoltaic string corresponding to each abnormal signal component and marking it as an abnormal string;

[0011] Step S4, obtaining the topological relationship of the current abnormal string group, and defining at least two voltage signal acquisition points based on the topological relationship;

[0012] Step S5, combining all voltage signal acquisition points in pairs to acquire at least one voltage difference of the current abnormal string;

[0013] Step S6, locating a theoretical abnormal point of the current abnormal string according to each voltage difference;

[0014] Step S7, obtaining all real geographical locations of the photovoltaic array corresponding to all theoretical abnormal points, and traversing all real geographical locations by the shortest route through an external aircraft, and obtaining at least one visual image based on a real geographical location;

[0015] Step S8, using a visual detection algorithm to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image;

[0016] Step S9, if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is determined to be in a foreign object interference fault state; if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in an internal circuit fault state.

[0017] As a further improvement of the present application, in step S1, a plurality of current signals of each photovoltaic panel are respectively obtained based on a preset time period; and a plurality of signal components of each current signal are respectively extracted by empirical mode decomposition, including:

[0018] Step S11, obtaining all maximum values ​​and all minimum values ​​of the current current signal;

[0019] Step S12, based on the current current signal, all the maximum values ​​are sequentially connected to form an upper envelope, and all the minimum values ​​are sequentially connected to form a lower envelope;

[0020] Step S13, obtaining the average value of the upper envelope and the lower envelope based on the current current signal, and sequentially connecting all the average values ​​to form a mean value line;

[0021] Step S14, subtracting the mean line from the current current signal to obtain a first-order intermediate signal;

[0022] Step S15, repeating steps S11 to S14 with the first-order intermediate signal as the main body, so as to iterate the first-order intermediate signal several times;

[0023] Step S16, respectively obtaining an intermediate signal in which the difference between the number of extreme value points and the number of zero-crossing points after each iteration is 0 or 1, and marking it as a second-order intermediate signal;

[0024] Step S17, obtaining the second-order intermediate signal with a mean line of zero and defining it as a signal component of the current current signal.

[0025] As a further improvement of the present application, step S2, screening out all abnormal signal components from all signal components by a classification algorithm, includes:

[0026] Step S21, integrating all signal components to form a signal set to be classified;

[0027] Step S22, defining a category set according to a preset signal type;

[0028] Step S23, calculating the conditional probability of the signal set to be classified under each preset signal type;

[0029] Step S24, classifying each signal component into the preset signal type with the highest conditional probability;

[0030] Step S25, obtaining the number of signal components of each preset signal type;

[0031] Step S26, defining the preset signal type with the lowest number of signal components as an abnormal signal component type;

[0032] Step S27, acquiring all signal components in the abnormal signal component type and defining them as the abnormal signal components.

[0033] As a further improvement of the present application, step S5, combining all voltage signal acquisition points in pairs to acquire at least one voltage difference of the current abnormal string, includes:

[0034] Step S51, obtaining the total length of the topological relationship;

[0035] Step S52, randomly generating at least two random numbers by a random function in the interval [0,1];

[0036] Step S53, multiplying the current random number by the total length, the obtained random length is the random point corresponding to the current random number on the abnormal cable, and one end of the random length coincides with one end of the topological relationship;

[0037] Step S54, defining each random point as a voltage signal acquisition point;

[0038] Step S55, combining all signal acquisition points in pairs by permutation and combination number C;

[0039] Step S56, obtaining a voltage difference of the current abnormal strings based on a pairwise combination.

[0040] As a further improvement of the present application, step S6, locating a theoretical abnormal point of the current abnormal string according to each voltage difference, includes:

[0041] Step S61, judging whether there is a zero voltage difference based on all voltage differences;

[0042] Step S62, if there is a voltage difference of zero, two voltage signal acquisition points corresponding to the voltage difference of zero are acquired and marked as equidistant propagation points;

[0043] Step S63, connecting two equidistant propagation points along the topological relationship to form an equidistant propagation line;

[0044] Step S64, obtaining the midpoint of the equidistant propagation line as a theoretical abnormal point of the current abnormal string.

[0045] As a further improvement of the present application, step S7, obtaining all real geographical locations of the photovoltaic array corresponding to all theoretical abnormal points, and traversing all real geographical locations by the shortest route through an external aircraft, and obtaining at least one visual image based on a real geographical location, includes:

[0046] Step S71, obtaining a digital elevation model of the high altitude area;

[0047] Step S72, dividing the digital elevation model into cube grids of a preset size;

[0048] Step S73, retaining the topmost cube grid as a candidate grid set;

[0049] Step S74, defining each grid in the candidate grid set as a node, defining the take-off point of the external aircraft as a starting point, defining the landing point of the external aircraft as an end point, defining a real geographical location as a waypoint, and defining all other nodes as passing points;

[0050] Step S75, calculating the minimum number of grids required to traverse all the waypoints from the starting point in sequence and finally reach the end point by using the A_star algorithm;

[0051] Step S76, obtaining grids corresponding to the minimum number of grids, and sequentially connecting them to form the shortest route;

[0052] Step S77, sending the shortest route to an external aircraft, and acquiring at least one visual image of each real geographic location through the external aircraft.

[0053] As a further improvement of the present application, step S8, using a visual detection algorithm to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image, includes:

[0054] Step S81, dividing the current visual image into a plurality of square grids on average;

[0055] Step S82, defining all photovoltaic panels as having the highest confidence;

[0056] Step S83, predicting a number of bounding boxes for all square grids, each bounding box including at least one square grid;

[0057] Step S84, calculating the intersection-over-union ratio of all bounding boxes;

[0058] Step S85, selecting a bounding box whose INR is greater than or equal to a preset INR threshold as a photovoltaic panel detection box;

[0059] Step S86, determining whether all photovoltaic panels based on the current photovoltaic string in the photovoltaic panel detection frame are continuous;

[0060] Step S87: If there is discontinuity among all photovoltaic panels of the current photovoltaic string, the photovoltaic panels at the discontinuity are defined as photovoltaic panels covered by attachments.

[0061] In order to achieve the above objectives, this application also provides the following technical solutions:

[0062] A photovoltaic array monitoring device for high altitude areas, the photovoltaic array monitoring device is applied to the photovoltaic array monitoring method as described above, the photovoltaic array monitoring device comprises:

[0063] A current signal and signal component acquisition module, used to acquire a plurality of current signals of each photovoltaic panel based on a preset time period, and extract a plurality of signal components of each current signal by empirical mode decomposition;

[0064] An abnormal signal component screening module, used for screening out all abnormal signal components from all signal components through a classification algorithm;

[0065] The abnormal string marking module is used to obtain the photovoltaic string corresponding to each abnormal signal component and mark it as an abnormal string;

[0066] A voltage signal acquisition point definition module is used to obtain the topological relationship of the current abnormal string group and define at least two voltage signal acquisition points on the topological relationship;

[0067] An abnormal string voltage difference acquisition module is used to acquire at least one voltage difference of the current abnormal string by combining all voltage signal acquisition points in pairs;

[0068] The module for locating the theoretical abnormal point of the abnormal string is used to locate a theoretical abnormal point of the current abnormal string according to each voltage difference;

[0069] A real geographic location visual image acquisition module is used to acquire all real geographic locations of the photovoltaic array corresponding to all theoretical abnormal points, and traverse all real geographic locations by the shortest route through an external aircraft to obtain at least one visual image based on a real geographic location;

[0070] The module for judging photovoltaic panels covered by attachments is used to judge whether there are photovoltaic panels covered by attachments among all photovoltaic panels in each visual image through a visual detection algorithm;

[0071] The photovoltaic panel fault state determination module is used to determine that if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is in a foreign object interference fault state; if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in an internal circuit fault state.

[0072] In order to achieve the above objectives, this application also provides the following technical solutions:

[0073] An electronic device comprises a processor and a memory coupled to the processor, wherein the memory stores program instructions executable by the processor; when the processor executes the program instructions stored in the memory, the photovoltaic array monitoring method as described above is implemented.

[0074] In order to achieve the above objectives, this application also provides the following technical solutions:

[0075] A storage medium stores program instructions, and when the program instructions are executed by a processor, the photovoltaic array monitoring method as described above can be implemented.

[0076] The present application obtains several current signals of each photovoltaic panel based on a preset time period, and extracts several signal components of each current signal by empirical mode decomposition; screens out all abnormal signal components from all signal components by a classification algorithm; obtains photovoltaic strings corresponding to each abnormal signal component and marks them as abnormal strings; obtains the topological relationship of the current abnormal string, and defines at least two voltage signal acquisition points on the topological relationship; obtains at least one voltage difference of the current abnormal string by combining all voltage signal acquisition points in pairs; locates a theoretical abnormal point of the current abnormal string according to each voltage difference; obtains all real geographical locations corresponding to the photovoltaic array of all theoretical abnormal points, and traverses all real geographical locations by an external aircraft in the shortest route, and obtains at least one visual image based on a real geographical location; determines whether there are photovoltaic panels covered with attachments among all photovoltaic panels in each visual image by a visual detection algorithm; if there are photovoltaic panels covered with attachments, the photovoltaic panels covered with attachments are determined to be in a state of external interference fault, and if there are no photovoltaic panels covered with attachments, the photovoltaic panels located at the theoretical abnormal points are determined to be in an internal circuit fault state. This application uses all photovoltaic panels in the photovoltaic array as a reference to each other, and deletes the same signal features (such as normal current signals; overall grid fluctuations caused by photovoltaic randomness, etc.), so that the remaining signals do not have the same type and are identified as abnormal features (such as external coverings caused by extreme climate at high altitudes, such as ice and snow, which make it difficult for sunlight to reach the photovoltaic panels, and short circuits and disconnections of internal circuits caused by extreme climate at high altitudes, etc.). Due to the characteristics of mutual reference, the probability of misjudgment of photovoltaic panel failures is reduced. At the same time, this application obtains points by randomly defining voltage signal acquisition points and obtaining voltage differences two by two, and obtains points of local discharge through multiple groups of voltage signal acquisition points. Compared with single detection, the multiple acquisitions of this application can further reduce errors and improve positioning accuracy. Compared with manual regular inspections and meter reading analysis, this application can automatically determine photovoltaic panel failures and fault locations, and directly reach the fault location through external aircraft, reducing inspection costs and improving fault location accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A schematic diagram of the steps of an embodiment of a photovoltaic array monitoring method in a high altitude area of ​​the present application;

[0078] Figure 2 This is a functional module diagram of an embodiment of a photovoltaic array monitoring device for high altitude areas of the present application;

[0079] Figure 3 This is a schematic diagram of the structure of an embodiment of the electronic device of the present application;

[0080] Figure 4This is a schematic diagram of the structure of an embodiment of the storage medium of the present application. DETAILED DESCRIPTION

[0081] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0082] The terms "first", "second" and "third" in this application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications in the embodiments of the present application (such as up, down, left, right, front, back...) are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0083] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0084] like Figure 1 As shown, this embodiment provides an embodiment of a photovoltaic array monitoring method in a high altitude area. In this embodiment, a plurality of photovoltaic arrays are installed in the high altitude area, each photovoltaic array has a plurality of photovoltaic strings connected in series, and each photovoltaic string has a plurality of photovoltaic panels.

[0085] Preferably, when photovoltaic modules with the same performance are connected in series, the voltage increases and the current remains unchanged; when photovoltaic modules with the same performance are connected in parallel, the current increases and the voltage remains unchanged. Generally, the connection of the photovoltaic array will first ensure that the voltage of the photovoltaic modules connected in series meets the load operating voltage requirements, and then connect several photovoltaic modules connected in series (also called photovoltaic module strings) in parallel according to the current capacity requirements. In this way, the voltage of the photovoltaic array is the number of times the voltage of a single photovoltaic module in series, and the current is the number of times the current of a single photovoltaic module in parallel.

[0086] Preferably, in a photovoltaic system with multiple battery strings connected in parallel, the string current flowing through any battery string is measured; then, in the battery string, the branch current provided to the cable by each monitored photovoltaic module is measured respectively; finally, by comparing the string current and each branch current, it is determined whether there is an expected difference between them. If there is an expected difference, it is considered that a fault has occurred at the corresponding monitored photovoltaic module; if the currents are consistent, it is considered that the photovoltaic module is operating normally.

[0087] Specifically, the photovoltaic array monitoring method includes the following steps:

[0088] Step S1 : acquiring a plurality of current signals of each photovoltaic panel based on a preset time period, and extracting a plurality of signal components of each current signal by empirical mode decomposition.

[0089] Preferably, in actual applications, the current signal of the photovoltaic array may be unstable or missing to varying degrees due to photovoltaic randomness (for example, extreme climate at high altitudes produces external coverings, such as ice and snow, which make it difficult for sunlight to reach the photovoltaic panels, or short circuits and open circuits in the internal circuits due to extreme climate at high altitudes, etc.), so the collected current signal may be initially pre-processed.

[0090] Specifically, preprocessing can be performed by detrending terms and five-point cubic smoothing.

[0091] Among them, the detrending term is the current signal data collected in the current test. Due to the zero drift of the amplifier caused by temperature changes, the instability of the low-frequency performance outside the sensor frequency range, and the environmental interference around the sensor, it often deviates from the baseline, and even the size of the deviation from the baseline changes with time. The whole process of the deviation from the baseline changing with time is called the trend term of the signal. The trend term directly affects the correctness of the signal and should be removed. The commonly used method to eliminate the trend term is the polynomial least squares method.

[0092] Preferably, a detrend() function can be provided in MATLAB to perform detrending operation, but only the mean and linear trend items can be removed. Therefore, if this function is used for operation, it is recognized that the trend item contained in the sensor is linear. If the trend item is considered to be nonlinear, it is necessary to use a function composed of polyfit() and ployval() to perform the operation, such as Liu_detrend(t,y,m). In actual current signal data processing, a polynomial of degree 1 to 3 is usually used to perform polynomial trend item elimination on the sampled data.

[0093] Among them, the five-point cubic smoothing method can be used for time domain and frequency domain signal smoothing. The main effect of this processing method on time domain data is to reduce the high-frequency random noise mixed in the current signal. And the effect on frequency domain data is to make the spectrum curve smooth, so as to obtain a better fitting effect in modal parameter identification. It should be noted that the five-point cubic smoothing method of frequency domain data will reduce the peak value in the spectrum curve and widen the trapezoid, which may cause the error of identification parameters to increase.

[0094] Preferably, the empirical mode decomposition (EMD) algorithm is an empirical mode decomposition (EMD) based on the concepts of instantaneous frequency and intrinsic mode function (IMF). The empirical mode decomposition (EMD) can decompose complex signals into several IMF components, each of which characterizes the local characteristics of the signal. Signal decomposition is based on the time scale characteristics of the data itself, without the need to pre-set any basis function, so it is adaptive. The advantage of empirical mode decomposition is that it does not use any defined function as a basis, but adaptively generates intrinsic mode functions based on the analyzed signal. It can be used to analyze nonlinear and non-stationary signal sequences, with a high signal-to-noise ratio and good time-frequency focusing. The design intention of this embodiment is to accurately decompose the current signal through the empirical mode decomposition algorithm to distinguish normal current signals from abnormal current signals.

[0095] Step S2, screening out all abnormal signal components from all signal components through a classification algorithm.

[0096] Preferably, the present embodiment may adopt Bayesian classification, which is an irregular classification method. Bayesian classification technology trains the classified sample subsets, learns and summarizes the classification function (the prediction of discrete variables is called classification, and the classification of continuous variables is called regression), and uses the trained classifier to classify the unclassified data. Among the different classification algorithms, the naive Bayesian classification algorithm (Naive Bayes) is a simple Bayesian classification algorithm, and the application effect of the naive Bayesian classification algorithm is better than that of the neural network classification algorithm and the decision tree classification algorithm. In particular, when the amount of data to be classified is very large, the Bayesian classification method has a higher accuracy rate than other classification algorithms. The design intention of the present embodiment to prefer the naive Bayesian classification algorithm rather than the neural network classification algorithm is to achieve high accuracy.

[0097] Preferably, the signal features are mainly divided into three time domain features: short-time energy, zero-crossing rate, and empirical permutation entropy; and six frequency domain features: spectral center of gravity, spectral extension, spectral entropy, spectral flux, spectral roll-off point, and Mel-frequency cepstrum coefficients.

[0098] Preferably, the present embodiment selects only one or two of the above features. If too many features are selected, it may easily lead to a sudden increase in the amount of calculation, which may easily cause the computer to freeze or become unresponsive during actual application.

[0099] Preferably, this embodiment can reduce the difficulty of detection by using one or two of short-time energy, zero-crossing rate, spectrum center of gravity, and spectrum flux.

[0100] Step S3, respectively obtain the photovoltaic string corresponding to each abnormal signal component and mark it as an abnormal string.

[0101] Preferably, the abnormal signal component can select a one-dimensional feature, such as a phase feature, a polarity feature, an amplitude feature, a time-frequency feature, or a wavelet packet feature.

[0102] Preferably, the signal characteristics of the abnormal signal component can be directly obtained through the PRPD spectrum, and the PRPD spectrum is a mature existing technology. The specific extraction of the above-mentioned one-dimensional characteristics is also an existing technology and will not be repeated in this embodiment.

[0103] Step S4, obtaining the topological relationship of the current abnormal string group, and defining at least two voltage signal acquisition points based on the topological relationship.

[0104] Preferably, the complete topological relationship of the photovoltaic array may be obtained first, and then the topological relationship of the abnormal strings may be found.

[0105] Step S5, all voltage signal acquisition points are combined in pairs to acquire at least one voltage difference of the current abnormal string.

[0106] Preferably, the photovoltaic strings are series circuits with characteristics of equal currents and additive voltages.

[0107] Step S6, locating a theoretical abnormal point of the current abnormal string according to each voltage difference.

[0108] Step S7, obtaining all real geographical locations of the photovoltaic arrays corresponding to all theoretical abnormal points, and traversing all real geographical locations by the shortest route through an external aircraft, and obtaining at least one visual image based on a real geographical location.

[0109] Preferably, the shortest route can be realized by Dijkstra algorithm, A* search algorithm, Bellman-Ford algorithm, Floyd-Warshall algorithm, D* algorithm, etc.

[0110] It is worth noting that when choosing the shortest path algorithm, it is necessary to consider factors such as the type of graph (directed graph, undirected graph), whether there are negative weight edges or negative weight loops, and the sparsity of the graph. For example, for dense graphs, the Dijkstra algorithm may be more effective; while for sparse graphs, the Bellman-Ford algorithm may be more appropriate.

[0111] Step S8, using a visual detection algorithm to determine whether there is a photovoltaic panel covered with attachments among all the photovoltaic panels in each visual image.

[0112] Preferably, the visual detection algorithm can be implemented by VJ, HOG, DPMDetector; deep learning Two-stageRCNN, SPPNet, FastRCNN, FasterRCNN; Trick algorithm FPN, CascadeRCNN; deep learning one-stage Yolo, X, SSD, RetinaNet; deep learning Anchor-free CornerNet, CenterNet, FCOS; TransformerDETR, etc.

[0113] Step S9, if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is determined to be in a foreign object interference fault state; if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in an internal circuit fault state.

[0114] Preferably, the attachments in high-altitude areas are usually snow or ice, both of which have high reflectivity to sunlight, so that the solar energy that can be converted by photovoltaic panels is significantly reduced.

[0115] Furthermore, in step S1, a plurality of current signals of each photovoltaic panel are respectively acquired based on a preset time period; and a plurality of signal components of each current signal are respectively extracted by empirical mode decomposition, including:

[0116] Step S11, obtaining all maximum values ​​and all minimum values ​​of the current current signal.

[0117] Step S12: based on the current current signal, all the maximum values ​​are sequentially connected to form an upper envelope, and all the minimum values ​​are sequentially connected to form a lower envelope.

[0118] Step S13, obtaining the average values ​​of the upper envelope and the lower envelope based on the current current signal, and sequentially connecting all the average values ​​to form a mean value line.

[0119] Step S14, subtracting the mean line from the current current signal to obtain a first-order intermediate signal.

[0120] Step S15, repeating steps S11 to S14 with the first-order intermediate signal as the main body, so as to iterate the first-order intermediate signal several times.

[0121] Step S16, respectively obtain the intermediate signal whose difference between the number of extreme value points and the number of zero-crossing points after each iteration is 0 or 1, and mark it as a second-order intermediate signal.

[0122] Preferably, the signal components are intrinsic mode functions (IMF), which are the signal components of each layer obtained after the original signal is decomposed by EMD.

[0123] Step S17, obtaining a second-order intermediate signal with a mean line of zero and defining it as a signal component of the current current signal.

[0124] It can be understood that the steps of empirical mode decomposition (EMD) mainly include the following points:

[0125] ①Extreme point extraction: First, find the local maximum and local minimum of the signal. These local maximum and minimum values ​​represent the peak and valley values ​​of the signal at different time points.

[0126] ② Construct upper and lower envelopes: Use the extracted local maximum and minimum values ​​to construct the upper and lower envelopes of the signal respectively. These two envelopes completely wrap the signal and are tangent to the signal at the extreme points.

[0127] ③ Extract the mean function: Calculate the average value of the upper envelope and the lower envelope to obtain the mean function. This mean function represents the average trend of the signal at the current time scale.

[0128] ④ Iterative decomposition: Subtract the mean function from the original signal to obtain a new signal without the average trend. Then, repeat the above steps for the new signal, that is, extract the extreme points again, construct the envelope, extract the mean function, and subtract the mean function again. This process will be iterated until the intrinsic mode function (IMF) that meets certain conditions is obtained.

[0129] ⑤ Determine the intrinsic mode function (IMF): During the iterative decomposition process, when the new signal meets the IMF conditions, it is considered that an IMF component has been decomposed. The IMF component represents the local characteristic information of the signal at a specific time scale. Then, continue to iteratively decompose the remaining signal until all IMF components are extracted.

[0130] ⑥Reconstruct the signal: Add all the extracted IMF components to the residual (i.e. the remaining signal part) to reconstruct the original signal. This process verifies the perfect reconstruction property of the EMD method, that is, the decomposed signal can reconstruct the original signal without distortion.

[0131] ⑦Through the above steps, empirical mode decomposition can decompose complex non-stationary signals into a finite number of intrinsic mode functions with local characteristics of different time scales, thereby facilitating further analysis and processing of the signal.

[0132] Furthermore, in step S2, all abnormal signal components in all signal components are screened out by a classification algorithm, including:

[0133] Step S21: integrating all signal components to form a signal set to be classified.

[0134] Preferably, the set of signals to be classified can be expressed as X={x 1 ,x 2 ,…,x j ,…,x m}, where x j is the jth signal component, and m is the number of all signal components.

[0135] Step S22: defining a category set according to a preset signal type.

[0136] Preferably, the category set can be expressed as C = {y 1 ,y 2 ,…,y k ,…,y n}, where y k is the kth preset signal type in the category set C, and n is the number of all preset signal types.

[0137] Preferably, the failures of the photovoltaic panels mainly include:

[0138] ① Reduction in power generation or power outage:

[0139] When a photovoltaic panel is damaged, its ability to generate electricity is affected, and there may be a reduction in power generation or a complete power outage.

[0140] ②Appearance damage: Damage on the appearance of photovoltaic panels, such as cracks, fragments, scratches, etc., is a significant feature of photovoltaic panel damage. These damages may cause wire breakage or poor contact, thus affecting the normal operation of the photovoltaic system.

[0141] ③ Deformation or displacement: Long-term heavy pressure or strong wind damage may cause the photovoltaic panel to deform or displace, causing the photovoltaic panel to lose contact with the bracket, thus affecting the power output.

[0142] ④Battery failure:

[0143] Hidden cracks: caused by external forces during welding or handling, or sudden expansion due to high temperature without preheating at low temperature. Hidden cracks will reduce the conversion efficiency of the battery cell, may cause open circuit or short circuit, and further affect the power attenuation of the component.

[0144] Splintering: caused by improper welding operation, incorrect lifting and placement techniques or laminator failure, which will also affect the power attenuation of the component.

[0145] Mixed gears: will reduce the overall power of the components, may cause hot spots or even burn the components.

[0146] Lightning streaks: may be caused by various factors such as hidden cracks in the battery cell, which can easily lead to hot spots and component degradation.

[0147] ⑤ Hot spot effect: Hot spot effect refers to the phenomenon that some cells inside the module are blocked or damaged, resulting in local temperature rise. This usually occurs when the cell is blocked by obstructions (such as snow, ice, leaves, bird droppings, large dust particles, etc.), or when the cell itself has hidden cracks. Hot spots will reduce the output power of photovoltaic modules and may cause aging and damage of the cells.

[0148] ⑥Power attenuation: During long-term operation, the power of photovoltaic modules may gradually attenuate due to factors such as high temperature, light, and humidity. This is a common aging problem of photovoltaic modules.

[0149] ⑦Backplane problem:

[0150] Yellowing of the backplane: Due to material aging caused by ultraviolet radiation and environmental factors, the backplane may turn yellow, causing the molecular structure of the backplane layer to be destroyed, performance to be reduced, reflectivity to be reduced, and affecting the overall output of the component.

[0151] Backplane bulging: Backplane bulging is prone to occur at locations where hot spots exist on the battery cells and where invisible tape is located, mainly due to the vaporization of the material caused by high temperature.

[0152] ⑧ Dirt and shadows:

[0153] Stains or shadows on the panels will affect their power generation efficiency. Therefore, the panels need to be cleaned regularly to avoid stains and shadows.

[0154] In addition, PV panels may also encounter problems with solder ribbons, bus bars and flux, such as cold solder joints, over-soldering and other problems, as well as electrical problems such as loose connections, fuse and circuit breaker failures.

[0155] Step S23, calculating the conditional probability of the signal set to be classified under each preset signal type.

[0156] Preferably, the conditional probability of the signal set to be classified under each preset signal type can be calculated by the following formula:

[0157]

[0158] Among them, P(x|y j ) is the conditional probability of the signal set x to be classified under the jth preset signal type; P(y j ) is the marginal probability of the jth preset signal type; P(x i |y j ) is the conditional probability of the i-th signal component under the j-th preset signal type.

[0159] Step S24, classifying each signal component into the preset signal type with the highest conditional probability.

[0160] Step S25, obtaining the number of signal components of each preset signal type.

[0161] Step S26, defining the preset signal type with the lowest number of signal components as the abnormal signal component type.

[0162] Step S27, acquiring all signal components in the abnormal signal component type and defining them as abnormal signal components.

[0163] Preferably, the naive Bayes classification assumes that the existence of a specific feature in a class is independent of the existence of any other feature, that is, each feature is independent of each other. Therefore, there are some constraints on the actual situation. If there is a correlation between attributes, the classification accuracy will be reduced, but in actual application, the classification effect of naive Bayes is relatively accurate. Naive Bayes solves the probability of each category appearing under the condition that the item appears for a given item to be classified. The largest probability is considered to be the category to which the item to be classified belongs.

[0164] Specifically, the Bayesian classification process is as follows:

[0165] ① Let x = {a1, a2, a3, ..., an} be an item to be classified, and each a is a feature of x.

[0166] ②There is a category set c = {y1, y2, y3,…, ym}.

[0167] ③Calculate P(y1|x), P(y2|x), …, P(ym|x).

[0168] ④If P(yk|x)=max{P(y1|x),P(y2|x),…,P(ym|x)}, then x∈yk.

[0169] ⑤Then calculate the conditional probabilities in step ③ through the following steps:

[0170] ⑥ Find a set of items to be classified with known classification. This set is called the training sample set.

[0171] ⑦ Statistically obtain the conditional probability estimate of each feature attribute in each category. That is:

[0172] P(a 1 |y 1 ),P(a 2 |y 1 ),……,P(a n |y 1 )

[0173] P(a 1 |y 2 ),P(a 2 |y 2 ),……,P(a n |y 2 );

[0174] …

[0175] P(a 1 |y m ),P(a 2 |y m ),……,P(a n |y m );

[0176] ⑧Assuming that each feature attribute is conditionally independent, according to the Bayesian principle:

[0177] P(y i |x)=P(x|y i )P(y i ) / p(x).

[0178] ⑨Since the denominator is a constant for all categories, we only need to maximize the numerator. And because each feature attribute is conditionally independent, then:

[0179] P(x|y i )P(y i )=P(a 1 |y i )P(a 2 |y i )……P(a n |y i )P(y i ).

[0180] It should be noted that the above preferred contents are for explanation of the principle, and the meaning of their symbols is not interchangeable with the meaning of the symbols of other formulas in this embodiment.

[0181] Further, in step S5, all voltage signal acquisition points are combined in pairs to acquire at least one voltage difference of the current abnormal string, including:

[0182] Step S51, obtaining the total length of the topological relationship.

[0183] Step S52: randomly generate at least two random numbers using a random function in the interval [0, 1].

[0184] Preferably, the number of random numbers can be set according to the length of the PV string. For example, it can be ensured that there is one random point every 1 meter or every 0.5 meter on average. For example, a 50m long PV string requires 50 to 100 random numbers, and the number of digits after the decimal point of the random number is generally set to two.

[0185] Step S53, multiplying the current random number by the total length, the obtained random length is the random point corresponding to the current random number on the abnormal cable, and one end of the random length coincides with one end of the topological relationship.

[0186] Preferably, for example, a random number is 0.35, and the same 50m-long photovoltaic string is used, then 50×0.35 is 17.5m, and starting from one end of the photovoltaic string, 17.5m is the random point corresponding to the random number.

[0187] Step S54: define each random point as a voltage signal acquisition point.

[0188] Step S55, all signal acquisition points are combined in pairs through the combination number C of permutations and combinations.

[0189] Preferably, assuming that the total number of random numbers is M, then according to the number of permutations and combinations C, we get

[0190] Step S56, obtaining a voltage difference of the current abnormal strings based on a pairwise combination.

[0191] Further, step S6, locating a theoretical abnormal point of the current abnormal string according to each voltage difference, includes:

[0192] Step S61 , judging whether there is a zero voltage difference based on all the voltage differences.

[0193] Step S62: If there is a voltage difference of zero, two voltage signal acquisition points corresponding to the voltage difference of zero are acquired and marked as equidistant propagation points.

[0194] Step S63, connecting two equidistant propagation points along the topological relationship to form an equidistant propagation line.

[0195] Step S64, obtaining the midpoint of the equidistant propagation line as a theoretical abnormal point of the current abnormal string.

[0196] Further, step S7, obtaining all real geographical locations of the photovoltaic arrays corresponding to all theoretical abnormal points, and traversing all real geographical locations by the shortest route through an external aircraft, and obtaining at least one visual image based on a real geographical location, including:

[0197] Step S71, obtaining a digital elevation model of a high altitude area.

[0198] Step S72, dividing the digital elevation model into cube grids of a preset size.

[0199] Step S73, retaining the top-most cube grid as a candidate grid set.

[0200] Step S74, each grid in the candidate grid set is defined as a node, the take-off point of the external aircraft is defined as the starting point, the landing point of the external aircraft is defined as the end point, a real geographical location is defined as a passing point, and all other nodes are defined as passing points.

[0201] Step S75, using the A_star algorithm to calculate the minimum number of grids required to traverse all the waypoints in sequence from the starting point and finally reach the end point.

[0202] Preferably, the priority is related to the heuristic function of the A-star algorithm:

[0203] F=G+H.

[0204] Among them, G is the movement cost of moving from the starting point to the specified grid, and the path generated along the grid is derived from the known grid information; H is the estimated cost of moving from the specified grid to the end point, and H is derived from the estimation of the unknown grid information; F is the basis for selecting the next node to be traversed.

[0205] Preferably, for a grid-like graph, the following heuristic functions can be used:

[0206] ① If the graph only allows movement in four directions: up, down, left, and right, the Manhattan distance can be used to calculate the number of grids required to move horizontally or vertically from the current grid to the target.

[0207] ② If the graph allows movement in eight directions, diagonal distance can be used. Both horizontal and vertical movement and diagonal movement are legal. In order to improve efficiency, integers 10, 14 ( The coefficient is rounded.

[0208] ③If the graph allows movement in any direction, you can use Euclidean distance, which is the straight-line distance between two points.

[0209] Step S76, obtaining grids corresponding to the minimum number of grids, and connecting them in sequence to form the shortest route.

[0210] Step S77, sending the shortest route to an external aircraft, and obtaining at least one visual image of each real geographic location through the external aircraft.

[0211] Preferably, the calculation process of the A_star algorithm is as follows:

[0212] Step A1: Create a grid set to be checked.

[0213] Step A2, adding the starting point, the end point, and all the waypoints to the grid set to be checked, and setting the priority of the starting point to the highest, the priority of all the waypoints to the middle, and the priority of the end point to the lowest.

[0214] Step A3, determining whether the grid set to be checked is empty, if the grid set to be checked is not empty, executing step A4.

[0215] Step A4: Select the node with the highest priority from the grid set to be checked.

[0216] Step A5, determine whether the node is one of the waypoints. If the node is one of the waypoints, execute step A6.

[0217] Step A6, starting from the current waypoint, gradually track the parent node until reaching the starting point.

[0218] Step A7, respectively obtain the number of grids required for each waypoint to reach the starting point, and define the waypoint with the least number of grids required as the starting point of the next path search.

[0219] Step A8, iterate steps S451 to S457 several times until the starting point, all the waypoints, and all the end points are connected in sequence with a single line to obtain the optimal route.

[0220] Step A9, obtaining all grids covered by the optimal route to obtain the minimum number of grids.

[0221] Preferably, step A3, after determining whether the grid set to be checked is empty, further includes:

[0222] Step A31: If the grid set to be checked is empty, all neighboring nodes of the node are traversed and all neighboring nodes are added to the grid set to be checked.

[0223] Step A32, obtaining the neighboring node with the highest priority among all neighboring nodes as the parent node for the next traversal.

[0224] Step A33, starting from the current parent node, gradually tracing the parent node of the next traversal until reaching one of the waypoints.

[0225] Step A34, defining one of the waypoints reached as the starting point of the next path search.

[0226] Step A35, iterate steps A31 to A34 several times until the starting point, all the waypoints and the end point are connected in sequence with a single line to obtain the optimal route.

[0227] Step A36, obtaining all grids covered by the optimal route to obtain the minimum number of grids.

[0228] Preferably, the grid set to be checked is the open list. In actual application, the A* algorithm implements the shortest path search through the following steps:

[0229] ① Start from the starting point and add the starting point to an open list for storing grids. Currently, there is only one item in the open list, namely the starting point. More grids will be gradually added in the subsequent search process. The grids in the open list are all grids that can be reached in the next step and are used as alternative grids. In the final shortest path, the grids in the open list may be passed along the way or not. That is, the open list is a list of grids to be checked.

[0230] ② Check the grids adjacent to the starting point, and add the walkable or reachable grids to the open list. Set the starting point to the parent node (parentnode or parentsquare) of these grids. When tracing the path, these parent nodes record the last node passed on the shortest path from the starting point to the node.

[0231] ③Remove the starting point from the open list and add it to the close list. Each grid in the close list no longer needs attention.

[0232] ④ Select the grid with the smallest F value from the open list and repeat steps (2) to (3).

[0233] In summary, by repeatedly traversing the open list, selecting the grid with the smallest F value, and generating new available grids, until the grid where the end point is found.

[0234] Preferably, for the cost of each grid, the heuristic function of each grid needs to be calculated separately. The calculation idea of ​​G is similar to the Dijkstra algorithm, which adopts a greedy strategy, that is, "if the shortest path from A to C passes through B, then the section from A to B must be the shortest", find the shortest path from the starting point to each possible point and record it. If 10 and 14 in this embodiment are used as coefficients, the horizontal or vertical movement cost is 10, and the diagonal movement cost is 14.

[0235] Furthermore, step S8, using a visual detection algorithm to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image, includes:

[0236] Step S81, dividing the current visual image into a number of square grids on average.

[0237] Specifically, the size of the original image of the image data may be adjusted to 448×448, and then the adjusted image may be evenly divided into S×S (eg, 7×7) grids, and the size of each grid is 64×64.

[0238] Step S22, predicting a number of bounding boxes for all crops based on all square grids according to a target detection algorithm.

[0239] Preferably, each grid is used to predict the horizontal coordinate, vertical coordinate, width, height, and confidence of each detection frame of N detection frames, that is, each grid needs to predict N×(4+1) values. Each grid needs to predict N (x, y, w, h, confidence); where (x, y) is the offset of the center of the detection frame relative to the grid, (w, h) is the ratio of the detection frame to the above-mentioned resized image, and (confidence) is the confidence of the grid, which takes a value of 1 or 0.

[0240] Step S82, defining all photovoltaic panels with the highest confidence.

[0241] Step S83: predicting a number of bounding boxes for all square grids, each bounding box including at least one square grid.

[0242] Step S84, calculating the intersection-over-union ratio of all bounding boxes.

[0243] Step S85 , selecting a bounding box whose IoU ratio is greater than or equal to a preset IoU ratio threshold as a photovoltaic panel detection box.

[0244] Preferably, the preset intersection-over-union ratio threshold can be set according to the required accuracy, for example, 80%.

[0245] Step S86, determining whether all photovoltaic panels based on the current photovoltaic string in the photovoltaic panel detection frame are continuous.

[0246] Step S87: If there is discontinuity among all photovoltaic panels of the current photovoltaic string, the photovoltaic panels at the discontinuity are defined as photovoltaic panels covered by attachments.

[0247] Preferably, the confidence level can be understood as whether there is a target in the current grid and the accuracy of the detection frame.

[0248] For example: suppose there is an object in a resized image, and the width and height of the resized image are (w a ,h a )but:

[0249] Divide the image into 7×7 (S×S) grids evenly. There is a grid located at the center of the target. The coordinates of the grid are (x i ,y i ), let the coordinates of the center of the target be (x a ,y a ), the above offset (x b ,y b ):

[0250] Preferably, in actual detection, if the predicted detection box and the actual bounding box overlap perfectly, the intersection-and-union ratio is 1. In actual application, the value of the first preset threshold can generally be set to 0.5 to determine whether the predicted bounding box is correct, and the accuracy of the bounding box is positively correlated with the intersection-and-union ratio.

[0251] Preferably, the YOLO algorithm also needs to train the detection frame to improve the accuracy of target detection.

[0252] Next, the training model is trained using a preset pedestrian and vehicle training set, and the weights and biases of the training model are iteratively adjusted a first preset number of times using a back propagation algorithm to reduce the value of the loss function of the training model.

[0253] Preferably, the loss function is as follows:

[0254]

[0255] in, is the indicator function of whether the j detection box of the i-th grid is responsible for the target, and its value is 1 or 0; x i ,y i 、w i 、h i , C i They correspond to the i-th (x, y, w, h, confidence) prediction values ​​respectively.

[0256] It can be understood that the loss function includes the coordinate value deviation of the detection box, the confidence deviation, and the prediction probability deviation (or category deviation).

[0257] in, is the detection frame midpoint loss in the coordinate value deviation, is the loss of detection box width and height in coordinate value deviation, is the confidence deviation, is the deviation of the predicted probability (or class deviation).

[0258] Among them, λ coord is the positioning error penalty, generally λ coord =5;S 2 That is, the S×S grids mentioned above; B is the number of bounding boxes; and is the estimated value of the horizontal and vertical coordinates of the midpoint of the i-th bounding box; and is the estimated value of the width and height of the i-th bounding box; C i is the confidence of the i-th bounding box; is the estimated value of the confidence of the i-th bounding box; noobj is the confidence prediction loss, usually λ noobj =0.5; p i (c) is the category probability of the i-th bounding box; is the estimated value of the category probability of the i-th bounding box; p i (c) with The c in it corresponds to classes.

[0259] It should be noted that since each grid does not necessarily contain a target, if there is no target in the grid, the value of (confidence) will be 0, which will make the gradient span in the subsequent back propagation algorithm too large, so λ is introduced coord To control the loss of the predicted position of the detection box, and introduce λnoobj Controls the penalty for objects not existing within a single grid.

[0260] It should be noted that the meaning of the symbols used in the principle description of the above target detection algorithm is not interchangeable with the meaning of other symbols in the context.

[0261] In this embodiment, a plurality of current signals of each photovoltaic panel are respectively obtained based on a preset time period, and a plurality of signal components of each current signal are respectively extracted by empirical mode decomposition; all abnormal signal components in all signal components are screened out by a classification algorithm; photovoltaic strings corresponding to each abnormal signal component are respectively obtained and marked as an abnormal string; the topological relationship of the current abnormal string is obtained, and at least two voltage signal acquisition points are defined on the topological relationship; at least one voltage difference of the current abnormal string is obtained by combining all voltage signal acquisition points in pairs; a theoretical abnormal point of the current abnormal string is located according to each voltage difference; all real geographical locations of the photovoltaic array corresponding to all theoretical abnormal points are obtained, and all real geographical locations are traversed by an external aircraft in the shortest route, and at least one visual image is obtained based on a real geographical location; and a visual detection algorithm is used to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image; if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is determined to be in a state of external interference fault, and if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in a state of internal circuit fault. This embodiment uses all photovoltaic panels of the photovoltaic array as references to each other, and deletes the same signal features (such as normal current signals; overall power grid fluctuations caused by photovoltaic randomness, etc.), so that the remaining signals do not have the same type and are identified as abnormal features (such as external coverings caused by extreme climate at high altitudes, such as ice and snow, which make it difficult for sunlight to reach the photovoltaic panels, and short circuits and disconnections of internal circuits caused by extreme climate at high altitudes, etc.). Due to the characteristics of mutual reference, the probability of misjudgment of photovoltaic panel failures is reduced. At the same time, this embodiment randomly defines voltage signal acquisition points and obtains voltage differences two by two, and obtains local discharge points through multiple groups of voltage signal acquisition points. Compared with single detection, multiple acquisitions in this embodiment can further reduce errors and improve positioning accuracy. Compared with manual regular inspections and meter reading analysis, this embodiment can automatically determine photovoltaic panel failures and fault locations, and directly reach the fault location through external aircraft, reducing inspection costs and improving fault location accuracy.

[0262] like Figure 2 As shown, this embodiment provides an embodiment of a photovoltaic array monitoring device for high altitude areas. In this embodiment, the photovoltaic array monitoring device is applied to the photovoltaic array monitoring method in the above embodiment.

[0263] Specifically, the photovoltaic array monitoring device includes a current signal and signal component acquisition module 1, an abnormal signal component screening module 2, an abnormal group string marking module 3, a voltage signal acquisition point definition module 4, an abnormal group string voltage difference acquisition module 5, an abnormal group string theoretical abnormal point positioning module 6, a real geographical location visual image acquisition module 7, an attachment covering photovoltaic panel judgment module 8, and a photovoltaic panel fault state judgment module 9, which are electrically connected in sequence.

[0264] Among them, the current signal and signal component acquisition module 1 is used to obtain several current signals of each photovoltaic panel based on a preset time period, and extract several signal components of each current signal by empirical mode decomposition; the abnormal signal component screening module 2 is used to screen out all abnormal signal components in all signal components through a classification algorithm; the abnormal string marking module 3 is used to obtain the photovoltaic string corresponding to each abnormal signal component respectively, and mark it as an abnormal string; the voltage signal acquisition point definition module 4 is used to obtain the topological relationship of the current abnormal string, and define at least two voltage signal acquisition points on the topological relationship; the abnormal string voltage difference acquisition module 5 is used to obtain at least one voltage difference of the current abnormal string by combining all voltage signal acquisition points in pairs; the abnormal string theoretical abnormal point positioning module 6 is used to locate a theoretical abnormal point of the current abnormal string according to each voltage difference; the real geographical location visual image acquisition module 7 is used to obtain all real geographical locations of the photovoltaic arrays corresponding to all theoretical abnormal points, and traverse all real geographical locations by the shortest route through an external aircraft, and obtain at least one visual image based on a real geographical location; the attachment covered photovoltaic panel judgment module 8 is used to determine whether there are photovoltaic panels covered with attachments among all photovoltaic panels in each visual image through a visual detection algorithm; the photovoltaic panel fault state judgment module 9 is used to determine that if there are photovoltaic panels covered with attachments, the photovoltaic panels covered with attachments are in an external interference fault state, and if there are no photovoltaic panels covered with attachments, the photovoltaic panels located at the theoretical abnormal point are judged to be in an internal circuit fault state.

[0265] Furthermore, the current signal and signal component acquisition module 1 specifically includes a first current signal and signal component acquisition submodule, a second current signal and signal component acquisition submodule, a third current signal and signal component acquisition submodule, a fourth current signal and signal component acquisition submodule, a fifth current signal and signal component acquisition submodule, a sixth current signal and signal component acquisition submodule, and a seventh current signal and signal component acquisition submodule, which are electrically connected in sequence; the seventh current signal and signal component acquisition submodule is electrically connected to the abnormal signal component screening module 2.

[0266] Among them, the first current signal and signal component acquisition submodule is used to obtain all the maximum values ​​and all the minimum values ​​of the current current signal; the second current signal and signal component acquisition submodule is used to connect all the maximum values ​​in sequence to form an upper envelope line based on the current current signal, and to connect all the minimum values ​​in sequence to form a lower envelope line; the third current signal and signal component acquisition submodule is used to obtain the average value of the upper envelope line and the lower envelope line based on the current current signal, and to connect all the average values ​​in sequence to form a mean line; the fourth current signal and signal component acquisition submodule is used to subtract the mean line from the current current signal to obtain A first-order intermediate signal; the fifth current signal and signal component acquisition submodule is used to repeatedly execute the first current signal and signal component acquisition submodule to the fourth current signal and signal component acquisition submodule based on the first-order intermediate signal to iterate the first-order intermediate signal several times; the sixth current signal and signal component acquisition submodule is used to respectively obtain the intermediate signal whose difference between the number of extreme points and the number of zero-crossing points after each iteration is 0 or 1, and mark it as a second-order intermediate signal; the seventh current signal and signal component acquisition submodule is used to obtain the second-order intermediate signal with a mean line of zero and define it as the signal component of the current current signal.

[0267] Furthermore, the abnormal signal component screening module 2 specifically includes a first abnormal signal component screening submodule, a second abnormal signal component screening submodule, a third abnormal signal component screening submodule, a fourth abnormal signal component screening submodule, a fifth abnormal signal component screening submodule, a sixth abnormal signal component screening submodule, and a seventh abnormal signal component screening submodule, which are electrically connected in sequence; the first abnormal signal component screening submodule is electrically connected to the seventh current signal and signal component acquisition submodule, and the seventh abnormal signal component screening submodule is electrically connected to the abnormal group string marking module 3.

[0268] Among them, the first abnormal signal component screening submodule is used to integrate all signal components to form a signal set to be classified; the second abnormal signal component screening submodule is used to define a category set according to a preset signal type; the third abnormal signal component screening submodule is used to calculate the conditional probability of the signal set to be classified under each preset signal type; the fourth abnormal signal component screening submodule is used to classify each signal component into the preset signal type with the highest conditional probability; the fifth abnormal signal component screening submodule is used to obtain the number of signal components of each preset signal type; the sixth abnormal signal component screening submodule is used to define the preset signal type with the lowest number of signal components as an abnormal signal component type; the seventh abnormal signal component screening submodule is used to obtain all signal components in the abnormal signal component type and define them as abnormal signal components.

[0269] Furthermore, the abnormal string voltage difference acquisition module 5 specifically includes a first abnormal string voltage difference acquisition submodule, a second abnormal string voltage difference acquisition submodule, a third abnormal string voltage difference acquisition submodule, a fourth abnormal string voltage difference acquisition submodule, a fifth abnormal string voltage difference acquisition submodule, and a sixth abnormal string voltage difference acquisition submodule, which are electrically connected in sequence; the first abnormal string voltage difference acquisition submodule is electrically connected to the voltage signal acquisition point definition module 4, and the sixth abnormal string voltage difference acquisition submodule is electrically connected to the abnormal string theoretical abnormal point positioning module 6.

[0270] Among them, the first abnormal group string voltage difference acquisition submodule is used to obtain the total length of the topological relationship; the second abnormal group string voltage difference acquisition submodule is used to randomly generate at least two random numbers through a random function with an interval of [0,1]; the third abnormal group string voltage difference acquisition submodule is used to multiply the current random number by the total length, and the obtained random length is the random point corresponding to the current random number on the abnormal cable, and one end of the random length coincides with one end of the topological relationship; the fourth abnormal group string voltage difference acquisition submodule is used to define each random point as a voltage signal acquisition point respectively; the fifth abnormal group string voltage difference acquisition submodule is used to combine all signal acquisition points in pairs through the combination number C of permutations and combinations; the sixth abnormal group string voltage difference acquisition submodule is used to obtain a voltage difference of the current abnormal group string based on a pairwise combination.

[0271] Furthermore, the abnormal group string theoretical abnormal point positioning module 6 specifically includes a first abnormal group string theoretical abnormal point positioning submodule, a second abnormal group string theoretical abnormal point positioning submodule, a third abnormal group string theoretical abnormal point positioning submodule, and a fourth abnormal group string theoretical abnormal point positioning submodule, which are electrically connected in sequence; the first abnormal group string theoretical abnormal point positioning submodule is electrically connected to the sixth abnormal group string voltage difference acquisition submodule, and the fourth abnormal group string theoretical abnormal point positioning submodule is electrically connected to the real geographic location visual image acquisition module 7.

[0272] Among them, the first abnormal group string theoretical abnormal point locating submodule is used to determine whether there is a zero voltage difference based on all voltage differences; the second abnormal group string theoretical abnormal point locating submodule is used to obtain two voltage signal acquisition points corresponding to the zero voltage difference if there is a zero voltage difference and mark them as equidistant propagation points; the third abnormal group string theoretical abnormal point locating submodule is used to connect two equidistant propagation points along the topological relationship to form an equidistant propagation line; the fourth abnormal group string theoretical abnormal point locating submodule is used to obtain the midpoint of the equidistant propagation line, which is a theoretical abnormal point of the current abnormal group string.

[0273] Furthermore, the real geographic location visual image acquisition module 7 specifically includes a first real geographic location visual image acquisition submodule, a second real geographic location visual image acquisition submodule, a third real geographic location visual image acquisition submodule, a fourth real geographic location visual image acquisition submodule, a fifth real geographic location visual image acquisition submodule, a sixth real geographic location visual image acquisition submodule, and a seventh real geographic location visual image acquisition submodule, which are electrically connected in sequence; the first real geographic location visual image acquisition submodule is electrically connected to the fourth abnormal group string theoretical abnormal point positioning submodule, and the seventh real geographic location visual image acquisition submodule is electrically connected to the attachment coverage photovoltaic panel judgment module 8.

[0274] Among them, the first real geographic location visual image acquisition submodule is used to obtain the digital elevation model of the high altitude area; the second real geographic location visual image acquisition submodule is used to divide the digital elevation model by a cube grid of a preset size; the third real geographic location visual image acquisition submodule is used to retain the top-level cube grid as a candidate grid set; the fourth real geographic location visual image acquisition submodule is used to define each grid in the candidate grid set as a node, and define the take-off point of the external aircraft as the starting point, the landing point of the external aircraft as the end point, a real geographic location as a waypoint, and all other nodes as passing points; the fifth real geographic location visual image acquisition submodule is used to calculate the minimum number of grids required to traverse all the waypoints in sequence from the starting point and finally reach the end point through the A_star algorithm; the sixth real geographic location visual image acquisition submodule is used to obtain the grids corresponding to the minimum number of grids, and connect them in sequence to form the shortest route; the seventh real geographic location visual image acquisition submodule is used to send the shortest route to the external aircraft, and obtain at least one visual image of each real geographic location through the external aircraft.

[0275] Furthermore, the attachment covering photovoltaic panel judgment module 8 specifically includes a first attachment covering photovoltaic panel judgment submodule, a second attachment covering photovoltaic panel judgment submodule, a third attachment covering photovoltaic panel judgment submodule, a fourth attachment covering photovoltaic panel judgment submodule, a fifth attachment covering photovoltaic panel judgment submodule, a sixth attachment covering photovoltaic panel judgment submodule, and a seventh attachment covering photovoltaic panel judgment submodule, which are electrically connected in sequence; the first attachment covering photovoltaic panel judgment submodule is electrically connected to the seventh real geographic location visual image acquisition submodule, and the seventh attachment covering photovoltaic panel judgment submodule is electrically connected to the photovoltaic panel fault state judgment module 9.

[0276] Among them, the first attachment covering photovoltaic panel judgment submodule is used to divide the current visual image into several square grids on average; the second attachment covering photovoltaic panel judgment submodule is used to define all photovoltaic panels with the highest confidence; the third attachment covering photovoltaic panel judgment submodule is used to predict several bounding boxes for all square grids, each bounding box includes at least one square grid; the fourth attachment covering photovoltaic panel judgment submodule is used to calculate the intersection and union ratio of all bounding boxes; the fifth attachment covering photovoltaic panel judgment submodule is used to select a bounding box with an intersection and union ratio greater than or equal to a preset intersection and union ratio threshold as a photovoltaic panel detection box; the sixth attachment covering photovoltaic panel judgment submodule is used to judge whether all photovoltaic panels based on the current photovoltaic string in the photovoltaic panel detection box are continuous; the seventh attachment covering photovoltaic panel judgment submodule is used to define the photovoltaic panel at the discontinuity as a photovoltaic panel covered by attachments if there is discontinuity among all photovoltaic panels of the current photovoltaic string.

[0277] It should be noted that this embodiment is a functional module embodiment based on the above method embodiment. The optimization, expansion, limitation, example, and principle description of this embodiment can be referred to the above embodiment, and this embodiment will not be repeated.

[0278] In this embodiment, a plurality of current signals of each photovoltaic panel are respectively obtained based on a preset time period, and a plurality of signal components of each current signal are respectively extracted by empirical mode decomposition; all abnormal signal components in all signal components are screened out by a classification algorithm; photovoltaic strings corresponding to each abnormal signal component are respectively obtained and marked as an abnormal string; the topological relationship of the current abnormal string is obtained, and at least two voltage signal acquisition points are defined on the topological relationship; at least one voltage difference of the current abnormal string is obtained by combining all voltage signal acquisition points in pairs; a theoretical abnormal point of the current abnormal string is located according to each voltage difference; all real geographical locations of the photovoltaic array corresponding to all theoretical abnormal points are obtained, and all real geographical locations are traversed by an external aircraft in the shortest route, and at least one visual image is obtained based on a real geographical location; and a visual detection algorithm is used to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image; if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is determined to be in a state of external interference fault, and if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in a state of internal circuit fault. This embodiment uses all photovoltaic panels of the photovoltaic array as references to each other, and deletes the same signal features (such as normal current signals; overall power grid fluctuations caused by photovoltaic randomness, etc.), so that the remaining signals do not have the same type and are identified as abnormal features (such as external coverings caused by extreme climate at high altitudes, such as ice and snow, which make it difficult for sunlight to reach the photovoltaic panels, and short circuits and disconnections of internal circuits caused by extreme climate at high altitudes, etc.). Due to the characteristics of mutual reference, the probability of misjudgment of photovoltaic panel failures is reduced. At the same time, this embodiment randomly defines voltage signal acquisition points and obtains voltage differences two by two, and obtains local discharge points through multiple groups of voltage signal acquisition points. Compared with single detection, multiple acquisitions in this embodiment can further reduce errors and improve positioning accuracy. Compared with manual regular inspections and meter reading analysis, this embodiment can automatically determine photovoltaic panel failures and fault locations, and directly reach the fault location through external aircraft, reducing inspection costs and improving fault location accuracy.

[0279] like Figure 3 As shown, this embodiment provides an embodiment of an electronic device. In this embodiment, the electronic device 10 includes a processor 101 and a memory 102 coupled to the processor 101 .

[0280] The memory 102 stores program instructions for implementing the photovoltaic array monitoring method in high altitude areas of any of the above embodiments.

[0281] The processor 101 is used to execute program instructions stored in the memory 102 to monitor photovoltaic arrays in high altitude areas.

[0282] The processor 101 may also be referred to as a CPU (Central Processing Unit). The processor 101 may be an integrated circuit chip having data processing capabilities. The processor 101 may also be a general-purpose processor, a digital data processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0283] Further, Figure 4 The schematic diagram of the structure of the storage medium of an embodiment of the present application is that the storage medium 11 of the embodiment of the present application stores program instructions 111 that can implement all the above methods, wherein the program instructions 111 can be stored in the above storage medium in the form of a software product, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or terminal devices such as a computer, a server, a mobile phone, and a tablet.

[0284] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0285] In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. The above is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of this application, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present application.

[0286] The specific implementation methods of the present application are described in detail above, but they are only examples, and the present application is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the present application is also within the scope of the present application, and therefore, the equalization, modification, and improvement made without departing from the spirit and principle of the present application should be included in the scope of the present application.

Claims

1. A photovoltaic array monitoring method in a high altitude area, wherein a plurality of photovoltaic arrays are installed in the high altitude area, each photovoltaic array has a plurality of photovoltaic strings connected in series, each photovoltaic string has a plurality of photovoltaic panels, characterized in that: The photovoltaic array monitoring method comprises: Step S1, acquiring a plurality of current signals of each photovoltaic panel based on a preset time period, and extracting a plurality of signal components of each current signal by empirical mode decomposition; Step S2, filtering out all abnormal signal components from all signal components through a classification algorithm; Step S3, respectively obtaining the photovoltaic string corresponding to each abnormal signal component and marking it as an abnormal string; Step S4, obtaining the topological relationship of the current abnormal string group, and defining at least two voltage signal acquisition points based on the topological relationship; Step S5, combining all voltage signal acquisition points in pairs to acquire at least one voltage difference of the current abnormal string; Step S6, locating a theoretical abnormal point of the current abnormal string according to each voltage difference; Step S7, obtaining all real geographical locations of the photovoltaic array corresponding to all theoretical abnormal points, and traversing all real geographical locations by the shortest route through an external aircraft, and obtaining at least one visual image based on a real geographical location; Step S8, using a visual detection algorithm to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image; Step S9, if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is determined to be in a foreign object interference fault state; if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in an internal circuit fault state.

2. The photovoltaic array monitoring method according to claim 1, characterized in that: Step S1, obtaining a plurality of current signals of each photovoltaic panel based on a preset time period; and extracting a plurality of signal components of each current signal by empirical mode decomposition, including: Step S11, obtaining all maximum values ​​and all minimum values ​​of the current current signal; Step S12, based on the current current signal, all the maximum values ​​are sequentially connected to form an upper envelope, and all the minimum values ​​are sequentially connected to form a lower envelope; Step S13, obtaining the average value of the upper envelope and the lower envelope based on the current current signal, and sequentially connecting all the average values ​​to form a mean value line; Step S14, subtracting the mean line from the current current signal to obtain a first-order intermediate signal; Step S15, repeating steps S11 to S14 with the first-order intermediate signal as the main body, so as to iterate the first-order intermediate signal several times; Step S16, respectively obtaining an intermediate signal in which the difference between the number of extreme value points and the number of zero-crossing points after each iteration is 0 or 1, and marking it as a second-order intermediate signal; Step S17, obtaining the second-order intermediate signal with a mean line of zero and defining it as a signal component of the current current signal.

3. The photovoltaic array monitoring method according to claim 1, characterized in that: Step S2, screening out all abnormal signal components from all signal components through a classification algorithm, including: Step S21, integrating all signal components to form a signal set to be classified; Step S22, defining a category set according to a preset signal type; Step S23, calculating the conditional probability of the signal set to be classified under each preset signal type; Step S24, classifying each signal component into the preset signal type with the highest conditional probability; Step S25, obtaining the number of signal components of each preset signal type; Step S26, defining the preset signal type with the lowest number of signal components as an abnormal signal component type; Step S27, acquiring all signal components in the abnormal signal component type and defining them as the abnormal signal components.

4. The photovoltaic array monitoring method according to claim 1, characterized in that: Step S5, combining all voltage signal acquisition points in pairs to acquire at least one voltage difference of the current abnormal string, including: Step S51, obtaining the total length of the topological relationship; Step S52, randomly generating at least two random numbers by a random function in the interval [0,1]; Step S53, multiplying the current random number by the total length, and the obtained random length is the random point corresponding to the current random number on the abnormal cable, and one end of the random length coincides with one end of the topological relationship; Step S54, defining each random point as a voltage signal acquisition point; Step S55, combining all signal acquisition points in pairs by permutation and combination number C; Step S56, obtaining a voltage difference of the current abnormal strings based on a pairwise combination.

5. The photovoltaic array monitoring method according to claim 1, characterized in that: Step S6, locating a theoretical abnormal point of the current abnormal string according to each voltage difference, including: Step S61, judging whether there is a zero voltage difference based on all voltage differences; Step S62, if there is a voltage difference of zero, two voltage signal acquisition points corresponding to the voltage difference of zero are acquired and marked as equidistant propagation points; Step S63, connecting two equidistant propagation points along the topological relationship to form an equidistant propagation line; Step S64, obtaining the midpoint of the equidistant propagation line as a theoretical abnormal point of the current abnormal string.

6. The photovoltaic array monitoring method according to claim 5, characterized in that: Step S7, obtaining all real geographical locations of the photovoltaic array corresponding to all theoretical abnormal points, and traversing all real geographical locations by the shortest route through an external aircraft, and obtaining at least one visual image based on a real geographical location, including: Step S71, obtaining a digital elevation model of the high altitude area; Step S72, dividing the digital elevation model into cube grids of a preset size; Step S73, retaining the topmost cube grid as a candidate grid set; Step S74, defining each grid in the candidate grid set as a node, defining the take-off point of the external aircraft as a starting point, defining the landing point of the external aircraft as an end point, defining a real geographical location as a waypoint, and defining all other nodes as passing points; Step S75, calculating the minimum number of grids required to traverse all the waypoints from the starting point in sequence and finally reach the end point by using the A_star algorithm; Step S76, obtaining grids corresponding to the minimum number of grids, and sequentially connecting them to form the shortest route; Step S77, sending the shortest route to an external aircraft, and acquiring at least one visual image of each real geographic location through the external aircraft.

7. The photovoltaic array monitoring method according to claim 1, characterized in that: Step S8, using a visual detection algorithm to determine whether there is a photovoltaic panel covered with attachments among all photovoltaic panels in each visual image, including: Step S81, dividing the current visual image into a plurality of square grids on average; Step S82, defining all photovoltaic panels as having the highest confidence; Step S83, predicting a number of bounding boxes for all square grids, each bounding box including at least one square grid; Step S84, calculating the intersection-over-union ratio of all bounding boxes; Step S85, selecting a bounding box whose INR is greater than or equal to a preset INR threshold as a photovoltaic panel detection box; Step S86, determining whether all photovoltaic panels based on the current photovoltaic string in the photovoltaic panel detection frame are continuous; Step S87: If there is discontinuity among all photovoltaic panels of the current photovoltaic string, the photovoltaic panels at the discontinuity are defined as photovoltaic panels covered by attachments.

8. A photovoltaic array monitoring device in high altitude areas, the photovoltaic array monitoring device being applied to the photovoltaic array monitoring method according to any one of claims 1 to 7, characterized in that: The photovoltaic array monitoring device comprises: A current signal and signal component acquisition module, used to acquire a plurality of current signals of each photovoltaic panel based on a preset time period, and extract a plurality of signal components of each current signal by empirical mode decomposition; An abnormal signal component screening module, used for screening out all abnormal signal components from all signal components through a classification algorithm; The abnormal string marking module is used to obtain the photovoltaic string corresponding to each abnormal signal component and mark it as an abnormal string; A voltage signal acquisition point definition module is used to obtain the topological relationship of the current abnormal string group and define at least two voltage signal acquisition points on the topological relationship; An abnormal string voltage difference acquisition module is used to acquire at least one voltage difference of the current abnormal string by combining all voltage signal acquisition points in pairs; The module for locating the theoretical abnormal point of the abnormal string is used to locate a theoretical abnormal point of the current abnormal string according to each voltage difference; A real geographic location visual image acquisition module is used to acquire all real geographic locations corresponding to the photovoltaic arrays of all theoretical abnormal points, and traverse all real geographic locations by the shortest route through an external aircraft to obtain at least one visual image based on a real geographic location; The photovoltaic panel covered by attachment judging module is used to judge whether there is a photovoltaic panel covered by attachment among all photovoltaic panels in each visual image through a visual detection algorithm; The photovoltaic panel fault state determination module is used to determine that if there is a photovoltaic panel covered with attachments, the photovoltaic panel covered with attachments is in a foreign object interference fault state; if there is no photovoltaic panel covered with attachments, the photovoltaic panel located at the theoretical abnormal point is determined to be in an internal circuit fault state.

9. An electronic device, characterized in that: It comprises a processor and a memory coupled to the processor, wherein the memory stores program instructions executable by the processor; when the processor executes the program instructions stored in the memory, the photovoltaic array monitoring method as described in any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium stores program instructions, and when the program instructions are executed by the processor, the photovoltaic array monitoring method according to any one of claims 1 to 7 can be implemented.

Images