A photovoltaic off-grid combined toroidal transformer and its detection method

Abstract

The present invention relates to the technical field of electrical equipment, and in particular to an off-grid combined toroidal transformer for photovoltaics and a detection method thereof. The off-grid combined toroidal transformer A and B are combined to replace a single large toroidal transformer. While ensuring the same functions, the outer diameter can be greatly reduced and the chassis length can be increased, which can meet the needs of long strip inverter chassis. The innovative installation method of the inductor is to fix it with the transformer A on the same screw, and through a series of component combinations, it is more stable and reliable than the conventional external suspension and glue fixing methods.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment, in particular to an off-grid combined toroidal transformer for photovoltaic use and a detection method thereof. Background Art

[0002] With the rapid development of photovoltaic power generation technology, off-grid photovoltaic inverters have been increasingly widely used in scenarios such as remote areas and emergency power supplies due to their flexibility and convenience. Most traditional off-grid photovoltaic inverters use single toroidal transformers that are large in volume and heavy in weight, which limits the use of the inverters in some application scenarios with limited space or high portability requirements. In addition, there are also some deficiencies in the winding process of traditional toroidal transformers. For example, the secondary leakage inductance is difficult to control, resulting in a large no-load current of the inverter, affecting the overall efficiency and waveform quality. At the same time, the winding process of traditional toroidal transformers has high requirements for the consistency of the transformers, which increases the production cost and difficulty.

[0003] In order to overcome the above problems, in recent years, some researchers have tried to use combined toroidal transformers to replace traditional single toroidal transformers. By combining multiple small toroidal transformers, the combined toroidal transformer can effectively reduce the outer diameter of the transformer and meet the requirements of the chassis size for some specific application scenarios, such as the common long-strip inverter chassis in the market. However, there is still room for optimization in the inductance installation method and manufacturing process of the existing combined toroidal transformers. In view of this, the present invention proposes a combined toroidal transformer for off-grid inverters for photovoltaic use and a detection method thereof. Summary of the Invention

[0004] The present invention provides an off-grid combined toroidal transformer for photovoltaic use and a detection method thereof to solve the limitation of the space requirement of the long-strip chassis for toroidal transformers in the inverter market.

[0005] The technical solution adopted by the present invention to achieve the above object is as follows:

[0006] The present invention discloses an off-grid combined toroidal transformer for photovoltaic use, including toroidal transformer A, toroidal transformer B, and an inductor;

[0007] The toroidal transformer A is provided with 4 ports, including a first port, a second port, a third port, and a fourth port; the toroidal transformer B is also provided with 4 ports, including a first port, a second port, a third port, and a fourth port;

[0008] The fourth port of the toroidal transformer A is connected to the third port of the toroidal transformer B, the third port of the toroidal transformer A is connected to the inductor, the other end of the inductor is connected to the negative pole of the battery 0VDC, the fourth port of the toroidal transformer B is connected to the positive pole of the battery 48VDC; the first port of the toroidal transformer A is connected to the first port of the toroidal transformer B, the second port of the toroidal transformer A is connected to the second port of the toroidal transformer B;

[0009] Through this connection method, the two windings on the battery side of the off-grid combined toroidal transformer are connected in series and the two windings on the 0-220V side are connected in parallel.

[0010] Further, in a preferred embodiment of the present invention, the toroidal transformer A comprises screws, nuts, an iron core A and a plurality of washers, the inductor, the iron core A and the plurality of washers are all mounted on the screws and fixed by the nuts; wherein the plurality of washers comprise iron flat washers.

[0011] Furthermore, in a preferred embodiment of the present invention, the toroidal transformer A is pre-embedded with hardware fixings to strengthen the fastening force of the iron flat washer.

[0012] The present invention also discloses a detection method for an off-grid combined toroidal transformer for photovoltaic use, which is applied to any of the off-grid combined toroidal transformers for photovoltaic use, and comprises the following steps:

[0013] After the toroidal transformer A and the toroidal transformer B are combined and wound into an off-grid combined toroidal transformer, the off-grid combined toroidal transformer is electrically tested;

[0014] During the electrical test, the test characteristic parameters of the off-grid combined toroidal transformer are obtained at several preset time nodes, and the test characteristic parameters are classified and processed to obtain the test characteristic parameter set of the off-grid combined toroidal transformer within a preset time period;

[0015] Obtaining the working performance requirement information of the off-grid combined toroidal transformer, and determining the standard characteristic parameter set of the off-grid combined toroidal transformer according to the working performance requirement information;

[0016] Compare and analyze the test characteristic parameter set with the standard characteristic parameter set to obtain the performance status of the off-grid combined toroidal transformer;

[0017] If the performance status of the off-grid combined toroidal transformer is abnormal, further analyze the off-grid combined toroidal transformer to obtain suspicious functional components of the off-grid combined toroidal transformer;

[0018] After obtaining the suspicious functional component of the off-grid combined toroidal transformer, the actual electrical parameters of the suspicious functional component during the electrical test are obtained;

[0019] Judge whether the actual electrical parameters of the suspicious functional component during the electrical test are within the preset parameter range; if not, determine the suspicious functional component as a faulty component;

[0020] Repair and replace the faulty components in the off-grid combined toroidal transformer so that the characteristic parameters of the off-grid combined toroidal transformer meet the production requirements.

[0021] Further, in a preferred embodiment of the present invention, the test characteristic parameter set is compared and analyzed with the standard characteristic parameter set to obtain the performance state of the off-grid combined toroidal transformer, specifically:

[0022] Construct a first splitting tree, import the test characteristic parameter set into the first splitting tree, construct a root node in the first splitting tree according to the test characteristic parameter set, and continuously split the test characteristic parameter set in the first splitting tree according to the root node to form several leaf nodes until there is only one test characteristic parameter in each leaf node, then the splitting ends, and a first splitting tree model is obtained;

[0023] Construct a second splitting tree, import the standard characteristic parameter set into the second splitting tree, construct a root node in the second splitting tree according to the standard characteristic parameter set, and continuously split the standard characteristic parameter set in the second splitting tree according to the root node to form several leaf nodes until there is only one standard characteristic parameter in each leaf node, then the splitting ends, and a second splitting tree model is obtained;

[0024] Introduce the iterative closest point matching algorithm, register the first splitting tree model and the second splitting tree model based on the iterative closest point matching algorithm, obtain the registered splitting tree model, and mark the two leaf nodes with the closest distance in the registered splitting tree model as a leaf node pair;

[0025] Calculate the cosine similarity of each leaf node pair in the registered splitting tree model, and perform weighted averaging on the cosine similarities of all leaf node pairs to obtain a similarity score;

[0026] Judge whether the similarity score is greater than the preset score value; if it is greater, mark the performance state of the off-grid combined toroidal transformer as the normal state; if it is not greater, mark the performance state of the off-grid combined toroidal transformer as the abnormal state.

[0027] Further, in a preferred embodiment of the present invention, if the performance state of the off-grid combined toroidal transformer is an abnormal state, further analysis is performed on the off-grid combined toroidal transformer to obtain the suspicious functional components of the off-grid combined toroidal transformer, specifically:

[0028] Obtain the composition of the functional components in the off-grid combined toroidal transformer, and obtain the functional characteristic information of each functional component in the off-grid combined toroidal transformer; and obtain various characteristic parameters of the off-grid combined toroidal transformer;

[0029] Introduce a cross-attention mechanism, and perform a correlation analysis on the functional characteristic information and various characteristic parameters of each functional component in the off-grid combined toroidal transformer through the cross-attention mechanism to obtain the correlation between each functional component and each characteristic parameter;

[0030] If the correlation between a certain functional component and a certain characteristic parameter is greater than a preset correlation threshold, then mark this functional component as the correlation functional component of this characteristic parameter; repeat this judgment process until the correlation between each functional component and each characteristic parameter is judged and analyzed, and the correlation functional components of various characteristic parameters are obtained;

[0031] Construct a knowledge graph, and import the correlation functional components of various characteristic parameters into the knowledge graph;

[0032] If the performance state of the off-grid combined toroidal transformer is an abnormal state, then compare the cosine similarity of each leaf node pair with a preset similarity threshold, and only mark the leaf node pairs whose cosine similarity is not greater than the preset similarity threshold;

[0033] Obtain the characteristic parameters attached to the leaf node pairs whose cosine similarity is not greater than the preset similarity threshold, and define them as the abnormal characteristic parameters of the off-grid combined toroidal transformer;

[0034] Import the abnormal characteristic parameters of the off-grid combined toroidal transformer into the knowledge graph for pairing, pair to obtain the correlation functional components of the abnormal characteristic parameters, and define the paired correlation functional components as the suspicious functional components of the off-grid combined toroidal transformer.

[0035] The present invention solves the technical defects in the background art, and the present invention has the following beneficial effects: Using toroidal transformers A and B combined into an off-grid combined toroidal transformer to replace a single large toroidal transformer can greatly reduce the outer diameter and increase the chassis length while ensuring the same functions, and can meet the requirements of a long-strip inverter chassis. Moreover, the innovative installation method of the inductor, fixing it on the same screw as transformer A and through a series of component combinations, is more stable and reliable than the conventional external suspension and dispensing fixing methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0037] Figure 1 It is a schematic structural diagram of an off-grid combined toroidal transformer;

[0038] Figure 2 It is a schematic connection structure diagram of toroidal transformer A and an inductor;

[0039] Figure 3 It is a schematic structural diagram of toroidal transformer B;

[0040] The description of the reference numerals is as follows: 101, screw; 102, mounting base plate; 103, rubber pad; 104, sponge pad; 105, core A; 106, flat iron washer; 107, spring washer; 108, nut; 109, black iron cover; 202, inductor; 208, hardware fixing piece; 1011, white silicone tube; 304, core B. Detailed implementation manners

[0041] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. The preferred embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0042] The present invention discloses an off-grid combined toroidal transformer for photovoltaic use, including toroidal transformer A, toroidal transformer B, and inductor 202;

[0043] The toroidal transformer A is provided with 4 ports, including a first port, a second port, a third port, and a fourth port; the toroidal transformer B is also provided with 4 ports, including a first port, a second port, a third port, and a fourth port;

[0044] The fourth port of the toroidal transformer A is connected to the third port of the toroidal transformer B, the third port of the toroidal transformer A is connected to the inductor 202, the other end of the inductor 202 is connected to the battery negative electrode 0VDC, the fourth port of the toroidal transformer B is connected to the battery positive electrode 48VDC; the first port of the toroidal transformer A is connected to the first port of the toroidal transformer B, and the second port of the toroidal transformer A is connected to the second port of the toroidal transformer B;

[0045] Through this connection method, the two windings on the battery side of the off-grid combined toroidal transformer are connected in series, and the two windings on the 0-220V side are connected in parallel.

[0046] It should be noted that for this toroidal transformer, toroidal transformer A (220VAC:15VAC) and toroidal transformer B (220VAC:15VAC) are used. The two transformers are combined into an off-grid combined toroidal transformer (inverter system), which has the same function as a large toroidal transformer (capacity = A + B) (220V:30VAC). As Figure 1 shown, the fourth port of transformer A is connected to the third port of transformer B. The third port of transformer A is connected to an inductor, and the other end of the inductor is connected to the negative pole 0VDC of the battery. The fourth port of transformer B is connected to the positive pole 48VDC of the battery. The first port of transformer A is connected to the first port of transformer B, and the second port of transformer A is connected to the second port of transformer B. In this way, the two windings on the battery side are connected in series, and the two windings on the 0-220V side are connected in parallel. Compared with using a large toroidal transformer (capacity = A + B), making two toroidal transformers can greatly reduce the outer diameter, and at the same time, the length occupied by the chassis needs to be increased, which can meet the requirements of the long-strip inverter chassis in the market.

[0047] As Figure 2 shown, the toroidal transformer A includes screw 101, mounting base plate 102, rubber pad 103, sponge pad 104, iron core A105, flat iron washer 106, spring washer 107, nut 108, black iron cover 109, rubber pad 103, rubber pad 103, black iron cover 109, flat iron washer 106, spring washer 107 and nut 108. A hardware fixing part 208 is embedded on the toroidal transformer A to strengthen the fastening force of the flat iron washer.

[0048] It should be noted that the inductor and the iron core A are fixed on the same screw. When the screw passes through the inductor, a white silicone tube 1011 is sleeved, which is used for the isolation between the enameled wire of the inductor and the screw. Compared with the conventional external suspension installation and dispensing fixing methods, this inductor fixing method is more stable.

[0049] As Figure 3 shown, the toroidal transformer B includes screw 101, mounting base plate 102, rubber pad 103, sponge pad 104, iron core B304, flat iron washer 106, spring washer 107, nut 108. A hardware fixing part 208 is embedded on the toroidal transformer B to strengthen the fastening force of the flat iron washer.

[0050] In summary, by using the combined off-grid toroidal transformers A and B to replace a single large toroidal transformer, the outer diameter can be greatly reduced and the chassis length can be increased while ensuring the same functions, meeting the requirements of the long-strip inverter chassis. Moreover, the innovative installation method of the inductor, which fixes it and the iron core A on the same screw and through a series of component combinations, is more stable and reliable than the conventional external suspension and glue-point fixing methods.

[0051] The present invention also discloses a detection method for an off-grid combined toroidal transformer for photovoltaic use, which is applied to any one of the off-grid combined toroidal transformers for photovoltaic use, and includes the following steps:

[0052] After winding the toroidal transformer A and the toroidal transformer B into an off-grid combined toroidal transformer, perform electrical tests on the off-grid combined toroidal transformer;

[0053] During the electrical test, obtain the test characteristic parameters of the off-grid combined toroidal transformer at several preset time nodes, and classify and process the test characteristic parameters to obtain the test characteristic parameter set of the off-grid combined toroidal transformer within the preset time period;

[0054] Obtain the working performance requirement information of the off-grid combined toroidal transformer, and determine the standard characteristic parameter set of the off-grid combined toroidal transformer according to the working performance requirement information;

[0055] Compare and analyze the test characteristic parameter set with the standard characteristic parameter set to obtain the performance state of the off-grid combined toroidal transformer;

[0056] If the performance state of the off-grid combined toroidal transformer is an abnormal state, further analyze the off-grid combined toroidal transformer to obtain the suspected functional components of the off-grid combined toroidal transformer;

[0057] When the suspected functional components of the off-grid combined toroidal transformer are obtained, obtain the actual electrical parameters of the suspected functional components during the electrical test;

[0058] Judge whether the actual electrical parameters of the suspected functional components during the electrical test are within the preset parameter range; if not, determine the suspected functional components as faulty components;

[0059] Repair and replace the faulty components in the off-grid combined toroidal transformer so that the characteristic parameters of the off-grid combined toroidal transformer meet the production requirements.

[0060] Among them, the test characteristic parameters of the off-grid combined toroidal transformer include but are not limited to input voltage, output voltage, voltage ratio, operating frequency, operating efficiency, leakage inductance, and operating temperature.

[0061] The actual electrical parameters of the suspected functional component during the electrical test include but are not limited to operating voltage, operating current, resistance, voltage transfer characteristics, and phase characteristics.

[0062] It should be noted that during the electrical test, test characteristic parameters are obtained at several preset time nodes. This means that measurements are not taken only once, but multiple times at different time points. Different time points may correspond to the performance of the transformer under different operating conditions (such as at the moment of startup, in a stable operating state, when the load changes, etc.). Information on the operating performance requirements of the off-grid combined toroidal transformer is obtained, and based on this information, a standard set of characteristic parameters is determined. The information on the operating performance requirements may come from design specifications, industry standards, or user requirements, etc. The standard set of characteristic parameters is the basis for measuring whether the transformer is qualified. The processed set of test characteristic parameters is compared and analyzed with the standard set of characteristic parameters. Through this comparison, the performance status of the off-grid combined toroidal transformer can be comprehensively evaluated to determine whether it meets the design and usage requirements. If the comparison result shows that the performance status of the off-grid combined toroidal transformer is an abnormal state, further analysis is required to find out the problem. This step reflects the fault diagnosis ability of this process method, not only to judge whether the product is qualified, but also to further locate the problem when it is unqualified. The off-grid combined toroidal transformer is further analyzed to obtain the suspected functional component. This involves the individual analysis of each component of the transformer (such as windings, iron cores, inductors, etc.) or a preliminary judgment of the components that may have problems based on experience and principles. When the suspected functional component is determined, the actual electrical parameters of this suspected functional component during the electrical test are obtained. These actual electrical parameters can more directly reflect the operating state of the suspected component. It is judged whether the actual electrical parameters of the suspected functional component during the electrical test are within the preset parameter range. If not, this suspected functional component is determined as a faulty component. This step is the key step to accurately determine the faulty component. By comparing with the preset parameter range, it can be clearly determined whether the component has truly failed. The component determined to be faulty is repaired and replaced. This is a necessary measure to ensure the normal operation of the off-grid combined toroidal transformer. By replacing the faulty component, the normal function of the transformer can be restored. After repair and replacement, the characteristic parameters of the off-grid combined toroidal transformer meet the production requirements. This ensures the quality of the final product, enabling it to meet the design specifications and usage requirements.

[0063] In summary, by obtaining test characteristic parameters at multiple preset time nodes and classifying and processing them, the performance of the off-grid combined toroidal transformer under different operating conditions can be comprehensively and accurately reflected. Compared with the test at a single time point, this method can capture more performance change information, such as the transient characteristics of the transformer at startup and its stability after long-term operation.

[0064] When an abnormality occurs, the suspicious functional components are determined through step-by-step analysis, and further, the faulty components are determined by comparing the actual electrical parameters with the preset parameter ranges. This hierarchical fault diagnosis method can accurately locate the components with problems, avoiding blind maintenance or replacement of the entire transformer, improving the maintenance efficiency, and reducing the maintenance cost. Finally, it ensures that the characteristic parameters of the off-grid combined toroidal transformer meet the production requirements, which guarantees the product quality. Whether from the perspective of meeting the design specifications to achieve the expected functions or from the perspective of conforming to industry standards to ensure the reliability and safety of the product, this detection method helps to improve the overall quality of the product. And it can determine the standard characteristic parameter set according to the working performance requirement information, making this detection method adaptable to different working requirements. Whether for different power demands, different usage environments, or different user customization requirements, the effective production and quality control of the off-grid combined toroidal transformer can be achieved by adjusting the standard characteristic parameter set.

[0065] Further, in a preferred embodiment of the present invention, the test characteristic parameter set is compared and analyzed with the standard characteristic parameter set to obtain the performance state of the off-grid combined toroidal transformer, specifically as follows:

[0066] Construct a first splitting tree, and import the test characteristic parameter set into the first splitting tree. Build a root node in the first splitting tree according to the test characteristic parameter set, and continuously split the test characteristic parameter set in the first splitting tree according to the root node to form several leaf nodes until there is only one test characteristic parameter in each leaf node, then the splitting ends, and a first splitting tree model is obtained;

[0067] Construct a second splitting tree, and import the standard characteristic parameter set into the second splitting tree. Build a root node in the second splitting tree according to the standard characteristic parameter set, and continuously split the standard characteristic parameter set in the second splitting tree according to the root node to form several leaf nodes until there is only one standard characteristic parameter in each leaf node, then the splitting ends, and a second splitting tree model is obtained;

[0068] Introduce the iterative closest point matching algorithm, register the first splitting tree model and the second splitting tree model based on the iterative closest point matching algorithm to obtain a registered splitting tree model, and mark the two leaf nodes with the closest distance in the registered splitting tree model as a leaf node pair;

[0069] Calculate the cosine similarity of each leaf node pair in the registered splitting tree model, and perform weighted averaging on the cosine similarities of all leaf node pairs to obtain a similarity score;

[0070] Determine whether the similarity score is greater than a preset score value; if it is greater, mark the performance status of the off-grid combined toroidal transformer as the normal state; if it is not greater, mark the performance status of the off-grid combined toroidal transformer as the abnormal state.

[0071] It should be noted that first, a data structure of a first splitting tree is constructed, and the test feature parameter set is imported into this tree structure. This tree structure is a hierarchical data organization form that can effectively manage and analyze complex data sets. The root node is constructed according to the test feature parameter set, and this root node is the starting point of the entire splitting tree. Then, starting from the root node, the test feature parameter set is continuously split according to certain rules. This process is like gradually decomposing a large data set into smaller subsets. Each split will form a new node until there is only one test feature parameter in each leaf node (i.e., the bottommost node), and at this time, the first splitting tree model is constructed. This splitting method helps to deeply analyze the relationship between each individual test feature parameter and the overall performance. Similar to constructing the first splitting tree, a second splitting tree is constructed and the standard feature parameter set of the off-grid combined toroidal transformer is imported into it. The standard feature parameter set is determined according to the working performance requirements of the transformer and is the benchmark for measuring whether the transformer is working properly. Similarly, the root node is constructed according to the standard feature parameter set, and then the standard feature parameter set is continuously split until there is only one standard feature parameter in each leaf node, thereby obtaining the second splitting tree model.

[0072] The iterative closest point matching algorithm is introduced, which is an algorithm for finding the best matching relationship between two data sets. Here, the first split tree model and the second split tree model are used as inputs, and they are registered through this algorithm. The purpose of registration is to align the corresponding nodes in the two tree models as accurately as possible for subsequent comparative analysis. In the registered split tree model, the two leaf nodes with the closest distance are marked as a pair of leaf nodes. For each pair of leaf nodes in the registered split tree model, the cosine similarity between them is calculated. Cosine similarity is an index to measure the similarity between two vectors (here, the characteristic parameters in the leaf nodes can be regarded as vectors), and its value range is between -1 and 1. The closer it is to 1, the more similar the two vectors are. The cosine similarities of all pairs of leaf nodes are weighted and averaged to obtain a comprehensive similarity score. This weighted averaging process takes into account the importance of each pair of leaf nodes in the overall comparison, so that the final similarity score can more comprehensively reflect the similarity between the test characteristic parameter set and the standard characteristic parameter set. Finally, this similarity score is compared with a preset score value. If the similarity score is greater than the preset score value, it indicates that the test characteristic parameter set is similar enough to the standard characteristic parameter set, and the performance state of the off-grid combined toroidal transformer is marked as the normal state; otherwise, if it is not greater than the preset score value, it is marked as the abnormal state.

[0073] By constructing split tree models to process the test characteristic parameter set and the standard characteristic parameter set respectively, then using the iterative closest point matching algorithm for registration, calculating the cosine similarity of the registered leaf node pairs and weighted averaging to obtain the similarity score, so as to judge the performance state of the off-grid combined toroidal transformer. It can comprehensively and meticulously compare the test characteristic parameters and the standard characteristic parameters, fully consider the role of each individual characteristic parameter in the overall performance evaluation, and can more accurately judge the performance state of the transformer compared with simply directly comparing the parameter values. Whether it is the determination of the normal state or the abnormal state is more scientific and reliable.

[0074] Furthermore, in a preferred embodiment of the present invention, if the performance state of the off-grid combined toroidal transformer is the abnormal state, then the off-grid combined toroidal transformer is further analyzed to obtain the suspicious functional components of the off-grid combined toroidal transformer, specifically:

[0075] Obtain the composition of the functional components in the off-grid combined toroidal transformer, and obtain the functional characteristic information of each functional component in the off-grid combined toroidal transformer; and obtain various characteristic parameters of the off-grid combined toroidal transformer;

[0076] Introduce the cross-attention mechanism, and through the cross-attention mechanism, perform a correlation analysis on the functional characteristic information and various characteristic parameters of each functional component in the off-grid combined toroidal transformer to obtain the correlation between each functional component and each characteristic parameter;

[0077] If the correlation between a functional component and a characteristic parameter is greater than a preset correlation threshold, then label the functional component as the correlation functional component of the characteristic parameter; repeat this judgment process until the correlations between all functional components and all characteristic parameters are judged and analyzed, and the correlation functional components of various characteristic parameters are obtained;

[0078] It should be noted that according to the association relationship between the characteristic parameters and each functional component of the off-grid combined toroidal transformer, determine the functional components related to the leaf node pairs with large differences. For example, if there are large differences in the characteristic parameters related to winding winding, then there may be problems with the winding winding component; if there are problems with the characteristic parameters related to the inductor installation method, then the inductor installation component may be the source of the abnormality;

[0079] Construct a knowledge graph and import the correlation functional components of various characteristic parameters into the knowledge graph;

[0080] If the performance state of the off-grid combined toroidal transformer is an abnormal state, then compare the cosine similarity of each leaf node pair with a preset similarity threshold, and only label the leaf node pairs whose cosine similarity is not greater than the preset similarity threshold;

[0081] Obtain the characteristic parameters attached to the leaf node pairs whose cosine similarity is not greater than the preset similarity threshold, and define them as the abnormal characteristic parameters of the off-grid combined toroidal transformer;

[0082] Import the abnormal characteristic parameters of the off-grid combined toroidal transformer into the knowledge graph for pairing, pair to obtain the correlation functional components of the abnormal characteristic parameters, and define the paired correlation functional components as the suspicious functional components of the off-grid combined toroidal transformer.

[0083] It should be noted that when the performance status of the off-grid combined toroidal transformer is abnormal, the composition of its functional components, the functional characteristic information of each functional component, and the various characteristic parameters of the transformer must be obtained first. This step is the basis for subsequent analysis. The composition of the functional components clarifies the scope of the analysis object, the functional characteristic information helps to deeply understand the role and characteristics of each component, and the characteristic parameters are an important basis for judging whether the component is abnormal. The cross-attention mechanism is introduced to perform correlation analysis on the functional characteristic information of each functional component and various characteristic parameters. The cross-attention mechanism can focus on the correlation between different information, and here it can effectively explore the intrinsic connection between the functional components and the characteristic parameters. Through this mechanism, the degree of correlation between each functional component and each characteristic parameter can be quantified. If the correlation between a functional component and a characteristic parameter is greater than the preset correlation threshold, the functional component is marked as the correlation functional component of the characteristic parameter. This process will judge all functional components and characteristic parameters one by one until all the analysis is completed, so as to obtain the correlation functional components corresponding to various characteristic parameters. This step establishes a direct connection between functional components and characteristic parameters, which provides a basis for the subsequent determination of suspicious functional components.

[0084] Build a knowledge graph and import the previously obtained correlation functional components of various feature parameters into it. The knowledge graph is a structured knowledge representation that can clearly present the relationship between different entities (here, feature parameters and functional components) to facilitate subsequent query and reasoning.

[0085] When the performance status is abnormal, the cosine similarity of each leaf node pair calculated previously is compared with the preset similarity threshold, and only the leaf node pairs whose cosine similarity is not greater than the preset similarity threshold are marked. The characteristic parameters attached to these leaf node pairs are defined as the abnormal characteristic parameters of the off-grid combined toroidal transformer. In this step, by comparing the similarity, the characteristic parameters that are greatly different from the normal state are screened out. These abnormal characteristic parameters are likely to be the factors that cause abnormal transformer performance. The abnormal characteristic parameters are imported into the knowledge graph for pairing. Since the knowledge graph has stored the correlation relationship between the characteristic parameters and the functional components, the correlation functional components of the abnormal characteristic parameters can be obtained through this pairing. These correlation functional components are defined as suspicious functional components of the off-grid combined toroidal transformer. This step is the ultimate goal of the entire analysis process. Through a series of previous analysis and preparation work, the suspicious functional components that may cause abnormal transformer performance are finally determined, providing a clear direction for subsequent maintenance and troubleshooting.

[0086] Through a series of complex analysis processes, including obtaining functional component information, analyzing correlations using cross-attention mechanisms, constructing knowledge graphs, and determining abnormal feature parameters based on cosine similarity and finally pairing to obtain suspicious functional components, it is possible to accurately locate the suspicious functional components that may have problems when the off-grid combined toroidal transformer has abnormal performance. Compared with simple fault troubleshooting methods, it fully considers the complex relationship between functional components and feature parameters, utilizes advanced analysis mechanisms (such as cross-attention mechanisms) and knowledge representation forms (such as knowledge graphs), improves the accuracy and efficiency of fault troubleshooting, provides strong support for quickly repairing off-grid combined toroidal transformers, and thus effectively improves production efficiency.

[0087] Among them, the test feature parameters are classified to obtain the test feature parameter set of the off-grid combined toroidal transformer within a preset time period. Specifically:

[0088] Construct a three-dimensional grid coordinate system and divide several sub-grids in the three-dimensional grid coordinate system;

[0089] The collected test feature parameters are discretized, and the discretized single test feature parameters are respectively mapped to a separate sub-grid;

[0090] Feature extraction processing is performed on the test feature parameters in each sub-grid to obtain the data feature information of the test feature parameters in each sub-grid;

[0091] Calculate the mutual information value between the data feature information of the test feature parameters in each sub-grid; and merge the two sub-grids with the largest mutual information value to obtain several grid groups;

[0092] Calculate the parameter mean of the test feature parameters in each grid group, and use the calculated parameter mean as the grid center of the corresponding grid group;

[0093] Calculate the Euclidean distance between each test feature parameter in each grid group and the corresponding grid center, and perform weighted average processing on the Euclidean distance between each test feature parameter in each grid group and the corresponding grid center to obtain the compactness of each grid group;

[0094] If the compactness of each grid group is greater than the preset compactness, stop the merge iteration operation;

[0095] If there is at least one grid group whose compactness is not greater than the preset compactness, then regard each grid group as a sub-grid again and perform merge processing until the compactness of each grid group is greater than the preset compactness, and then stop the merge iteration operation;

[0096] After receiving the instruction to stop the merging iteration operation, each grid group is segmented in the three-dimensional grid coordinate system to obtain the test characteristic parameter set of the off-grid combined toroidal transformer within a preset time period.

[0097] It should be noted that a three-dimensional grid coordinate system is constructed, which provides a spatial framework for the subsequent processing of test characteristic parameters. Several sub-grids are segmented in this three-dimensional space, and these sub-grids will serve as the basic units for discretely processing test characteristic parameters. The collected test characteristic parameters are discretely processed, which means converting the continuous test characteristic parameters into discrete values or states according to certain rules. Then, the discrete single test characteristic parameters are respectively mapped to a separate sub-grid, thus realizing the preliminary positioning of the test characteristic parameters in the three-dimensional grid space. Feature extraction processing is performed on the test characteristic parameters in each sub-grid, aiming to obtain the data characteristic information of these test characteristic parameters (such as mean, median, mode, standard deviation, etc.). These data characteristic information can more concisely and effectively represent the characteristics of the test characteristic parameters within the sub-grid, for example, they may be some statistical characteristics or specific characteristics related to the performance of the transformer.

[0098] Calculate the mutual information value between the data characteristic information of the test characteristic parameters in each sub-grid. Mutual information is a measure of the degree of mutual dependence between two random variables and is used here to measure the degree of association between the data characteristic information in different sub-grids. Through the mutual information value, the sub-grids with the strongest degree of association can be found. The two sub-grids with the largest mutual information value are merged, and by continuously repeating this process, several grid groups are obtained. This merging process is a clustering operation based on the degree of association of data characteristic information, clustering the sub-grids where the data characteristic information with strong correlation is located together to form grid groups. For each grid group, calculate the parameter mean of the test characteristic parameters therein and use this mean as the grid center of the corresponding grid group. The grid center can be regarded as a representative position of this grid group in the test characteristic parameter space, which reflects a certain average state of the test characteristic parameters within this grid group.

[0099] Calculate the Euclidean distance between each test feature parameter in each grid group and the corresponding grid center. The Euclidean distance is a common method for measuring the spatial distance between two points. Then, perform a weighted average process on these distances to obtain the compactness of each grid group. The compactness reflects the degree of aggregation of the test feature parameters within the grid group relative to the grid center. If the compactness is high, it indicates that the test feature parameters within the grid group are relatively concentrated near the grid center. Determine whether the compactness of each grid group is greater than the preset compactness. If so, stop the merging iteration operation, which means that the current grid group division already meets certain compactness requirements; if there is at least one grid group whose compactness is not greater than the preset compactness, regard each grid group as a sub-grid again and perform the merging process again until the compactness of all grid groups meets the requirements. This iterative process is to continuously optimize the grid group division so that the test feature parameters within each grid group have high compactness. After stopping the merging iteration operation, cut out each grid group in the three-dimensional grid coordinate system, and these grid groups constitute the test feature parameter set of the off-grid combined toroidal transformer within the preset time period. This test feature parameter set is obtained through a series of complex processing and optimization, and it integrates the original test feature parameters in a more orderly and representative way.

[0100] By constructing a three-dimensional grid coordinate system, performing discrete mapping, feature extraction, sub-grid merging based on mutual information, calculating the grid group compactness and performing iterative merging operations on the test feature parameters, and finally obtaining the test feature parameter set, it can effectively classify the test feature parameters, explore the internal relationships between the test feature parameters, and make the finally obtained test feature parameter set have good structural and representativeness. Compared with the simple method of directly processing the test feature parameters, it can better reflect the complex performance characteristics of the off-grid combined toroidal transformer within the preset time period, and provides a more reliable data basis for subsequent comparative analysis with the standard feature parameter set and accurate judgment of the transformer performance status.

[0101] The above is inspired by the ideal embodiments of the present invention, and the description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent should be subject to the appended claims.

Claims

1. A detection method for off-grid combined toroidal transformer for photovoltaic use, characterized in that: The off-grid combined toroidal transformer comprises a toroidal transformer A, a toroidal transformer B and an inductor; The toroidal transformer A is provided with 4 ports, including a first port, a second port, a third port and a fourth port; the toroidal transformer B is also provided with 4 ports, including a first port, a second port, a third port and a fourth port; The fourth port of the toroidal transformer A is connected to the third port of the toroidal transformer B, the third port of the toroidal transformer A is connected to the inductor, the other end of the inductor is connected to the negative pole of the battery 0VDC, the fourth port of the toroidal transformer B is connected to the positive pole of the battery 48VDC; the first port of the toroidal transformer A is connected to the first port of the toroidal transformer B, the second port of the toroidal transformer A is connected to the second port of the toroidal transformer B; Through this connection method, the two windings on the battery side of the off-grid combined toroidal transformer are connected in series and the two windings on the 0-220V side are connected in parallel; The detection method comprises the following steps: After the toroidal transformer A and the toroidal transformer B are combined and wound into an off-grid combined toroidal transformer, an electrical test is performed on the off-grid combined toroidal transformer; During the electrical test, the test characteristic parameters of the off-grid combined toroidal transformer are obtained at several preset time nodes, and the test characteristic parameters are classified to obtain a set of test characteristic parameters of the off-grid combined toroidal transformer within a preset time period; Acquiring working performance requirement information of the off-grid combined toroidal transformer, and determining a standard characteristic parameter set of the off-grid combined toroidal transformer according to the working performance requirement information; Compare and analyze the test characteristic parameter set with the standard characteristic parameter set to obtain the performance status of the off-grid combined toroidal transformer; If the performance state of the off-grid combined toroidal transformer is abnormal, further analyzing the off-grid combined toroidal transformer to obtain suspicious functional components of the off-grid combined toroidal transformer; After obtaining the suspicious functional components of the off-grid combined toroidal transformer, obtaining actual electrical parameters of the suspicious functional components during the electrical testing process; Determine whether the actual electrical parameters of the suspicious functional component during the electrical test process are within a preset parameter range; if not, determine the suspicious functional component as a faulty component; The faulty parts in the off-grid combined toroidal transformer are repaired and replaced so that the characteristic parameters of the off-grid combined toroidal transformer meet the production requirements.

2. The detection method of a photovoltaic off-grid combined toroidal transformer according to claim 1, characterized in that: The toroidal transformer A comprises screws, nuts, an iron core A and a plurality of gaskets. The inductor, the iron core A and the plurality of gaskets are all mounted on the screws and fixed by the nuts. The plurality of gaskets comprise iron flat gaskets.

3. The detection method of a photovoltaic off-grid combined toroidal transformer according to claim 2, characterized in that: The toroidal transformer A is pre-embedded with hardware fixings to strengthen the fastening force of the iron flat washer.

4. The detection method of a photovoltaic off-grid combined toroidal transformer according to claim 1, characterized in that: The test characteristic parameter set is compared and analyzed with the standard characteristic parameter set to obtain the performance status of the off-grid combined toroidal transformer, specifically: Constructing a first splitting tree, and importing the test feature parameter set into the first splitting tree, constructing a root node in the first splitting tree according to the test feature parameter set, and continuously splitting the test feature parameter set in the first splitting tree according to the root node to form a plurality of leaf nodes, until only one test feature parameter exists in each leaf node, the splitting ends, and a first splitting tree model is obtained; Constructing a second splitting tree, and importing the standard feature parameter set into the second splitting tree, constructing a root node in the second splitting tree according to the standard feature parameter set, and continuously splitting the standard feature parameter set in the second splitting tree according to the root node to form a plurality of leaf nodes, until only one standard feature parameter exists in each leaf node, the splitting ends, and a second splitting tree model is obtained; An iterative closest point matching algorithm is introduced, and the first split tree model and the second split tree model are registered based on the iterative closest point matching algorithm to obtain a registered split tree model, and two leaf nodes closest to each other in the registered split tree model are marked as a leaf node pair; Calculate the cosine similarity of each leaf node pair in the split tree model after registration, and perform weighted average of the cosine similarities of all leaf node pairs to obtain a similarity score; Determine whether the similarity score is greater than a preset score value; If it is greater than, the performance status of the off-grid combined toroidal transformer is marked as normal; If it is not greater than, the performance status of the off-grid combined toroidal transformer is marked as an abnormal state.

5. The detection method of a photovoltaic off-grid combined toroidal transformer according to claim 4, characterized in that: If the performance status of the off-grid combined toroidal transformer is abnormal, further analysis is performed on the off-grid combined toroidal transformer to obtain suspicious functional components of the off-grid combined toroidal transformer, specifically: Obtaining the composition of functional components in the off-grid combined toroidal transformer, and obtaining functional characteristic information of each functional component in the off-grid combined toroidal transformer; and obtaining various characteristic parameters of the off-grid combined toroidal transformer; The cross-attention mechanism is introduced to analyze the correlation between the functional characteristic information of each functional component in the off-grid combined toroidal transformer and various characteristic parameters, and obtain the correlation between each functional component and each characteristic parameter. If the correlation between a certain functional component and a certain characteristic parameter is greater than a preset correlation threshold, the functional component is marked as the correlation functional component of the characteristic parameter; repeat this judgment process until the correlation between each functional component and each characteristic parameter is judged and analyzed, and the correlation functional components of various characteristic parameters are obtained; Constructing a knowledge graph, and importing correlation functional components of various feature parameters into the knowledge graph; If the performance state of the off-grid combined toroidal transformer is abnormal, the cosine similarity of each leaf node pair is compared with a preset similarity threshold, and only the leaf node pairs whose cosine similarity is not greater than the preset similarity threshold are marked; Acquire characteristic parameters attached to leaf node pairs whose cosine similarity is not greater than a preset similarity threshold, and define them as abnormal characteristic parameters of the off-grid combined toroidal transformer; The abnormal characteristic parameters of the off-grid combined toroidal transformer are imported into the knowledge graph for pairing, and the correlation functional components of the abnormal characteristic parameters are obtained by pairing, and the correlation functional components obtained by pairing are defined as suspicious functional components of the off-grid combined toroidal transformer.

Images