High-voltage cable intelligent grounding box and high-voltage cable cross interconnection misconnection defect detection method
Through the detection method of intelligent grounding box and cross-interconnect error defects of high-voltage cable, the fault type is identified by equivalent circuit model and artificial neural network, the problem of difficult detection of cross-interconnect error defects of high-voltage cable sheath is solved, achieving efficient and accurate Defect diagnosis and repair measures.
Patent Information
- Application Number
- CN202410788125.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-19
- Publication Date
- 2025-02-18
AI Technical Summary
The prior art is difficult to effectively detect and diagnose defects in cross-interconnection of high-voltage cable sheath, resulting in high defect detection costs and limiting the verification and application of detection methods.
The high-voltage cable intelligent grounding box and cross-connection fault defect detection method are used to establish an equivalent circuit model, extract the fault feature quantity and criterion, build the error connection method classification table, and use artificial neural network to identify the fault type.
Laboratory equivalent simulation detection of cross-connection grounding defects of high-voltage cable sheath is realized, which can identify errors and provide targeted maintenance measures, reducing inspection costs and complexity.
Smart Images
Figure CN119471479A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of high-voltage cable monitoring, and in particular to a high-voltage cable intelligent grounding box and a method for detecting cross-connection and misconnection defects of high-voltage cables. Background Art
[0002] In order to limit the induced voltage on the sheath, high-voltage cables over 1.2 kilometers are generally grounded directly at both ends and cross-connected in the middle. The cross-connection and misconnection of the high-voltage cable sheath is a typical defect. The test of the cross-connection and misconnection of the high-voltage cable sheath is very expensive, which limits the verification and application of the defect detection method. It is urgent to propose a laboratory equivalent simulation method for the cross-connection and misconnection of the high-voltage cable sheath. Summary of the invention
[0003] The purpose of the embodiments of the present application is to provide a high-voltage cable intelligent grounding box and a method for detecting cross-connection errors of high-voltage cables, which converts the misconnection information into a data form that is easy to process and analyze, thereby facilitating the statistics, analysis and comparison of the misconnection situations. The misconnection mode can be identified through on-site detection of characteristic quantities, providing a basis for formulating targeted maintenance measures.
[0004] To achieve the above objectives, this application provides the following technical solutions:
[0005] The present application provides a method for detecting cross-connection faults of high-voltage cables, including the following specific steps:
[0006] Establish an equivalent circuit model to simulate different high-voltage cable sheath cross-connection wiring error faults;
[0007] Extract fault characteristics and fault judgment criteria, construct a classification table of misconnection modes, and encode the misconnection modes;
[0008] The intelligent grounding box of the high-voltage cable collects the sheath current phase information and the collection time of the first and last sheath currents. It uses the sheath current amplitude, the ratio of the sheath currents of the three loops and the sheath current phasor difference of the same loop as characteristic quantities to query the corresponding code and identify the fault type.
[0009] The equivalent circuit model adopts a three-phase nine-segment cross-interconnection. According to the three-phase nine-segment cross-interconnection, simulation models of different cross-interconnection misconnection modes are established to simulate different high-voltage cable sheath cross-interconnection wiring error faults.
[0010] The extraction of fault characteristic quantities and fault criteria is specifically to set the cable parameters such as core current, three-section lengths, phase-to-phase distance, and ambient temperature, consider the influence of cable wiring and unbalanced cross-connection parts, set the influence coefficient, and output the extracted fault characteristic quantities and fault criteria as sheath current amplitude and phase.
[0011] The construction of the misconnection mode classification table and encoding of the misconnection modes are specifically as follows:
[0012] According to the type of cross-connection misconnection, it is divided into four categories, namely normal situation, two typical cross-connection misconnection methods, double-circuit cross-connection misconnection, and three-circuit cross-connection misconnection. Under normal circumstances, both ends are connected with left-falling-left-falling-right or both ends are connected with right-falling-right-left, and the code is set to 0; the two typical connection methods are one end is connected with left-falling-left-falling-right and the other end is connected with right-falling-right-left, and the codes are set to A1 and A2; there are 9 types of double-circuit cross-connection misconnection, coded as B1 to B9; there are 23 types of three-circuit cross-connection misconnection, coded as C1 to C23.
[0013] The fault type identification adopts an artificial neural network method. The artificial neural network includes an input layer, a hidden layer and an output layer. The input layer represents input data, including on-site cable parameters and cable core current; the hidden layer is used for calculation and processing. There can be one or more hidden layers. Each neuron in the hidden layer will calculate the input it receives from the previous layer and generate an output; the output layer outputs the result of the identified fault type.
[0014] A high-voltage cable intelligent grounding box is used to implement the high-voltage cable cross-connection and misconnection defect detection method as described above, characterized in that it includes a detection host, a sheath current detection component, a cable current sensor, and a high-frequency sensor. A grounding terminal is arranged in the grounding box, and the single-core cable is grounded through the grounding terminal. The cross-connected sheath current detection component between the grounding terminals includes a sheath current detection sensor, and the sheath current detection sensor is sleeved on the metal sheath of the single-core cable. A waveform extraction module is arranged in the detection host to provide a phase difference feature value recognized by CNN, and the high-frequency sensor and the sheath current detection component are electrically connected to the input end of the detection host.
[0015] Compared with the prior art, the beneficial effects of this application are:
[0016] Starting from the actual needs of cross-connection fault diagnosis of high-voltage cable sheath grounding system, the present invention aims at the theories and bottlenecks existing in existing research, combines the existing theoretical basis and detection technology, and faces the actual working conditions and distributed online monitoring scenarios. It proposes to use the phase difference of the sheath current at the head and tail of the high-voltage cable as a characteristic quantity, classify and encode 36 cross-connection connection methods, and use ANN to identify different cross-connection misconnection types. At the same time, in terms of on-site detection, a low-cost, 100kHz, 0.2-level sheath grounding current online monitoring sensor and sampling hardware unit and a high-precision wide-band high-voltage cable intelligent grounding box are used to realize the collection of sheath current phase information and the synchronization of the head and tail sheath current collection time, ultimately realizing the innovation of high-voltage cable on-site detection technology and intelligent diagnosis of sheath cross-connection grounding defects. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0018] Figure 1 A flow chart of the method of this application;
[0019] Figure 2 A schematic diagram of a cross-connected high-voltage cable for this application;
[0020] Figure 3 The equivalent circuit diagram of the normal and typical misconnection conditions of this application;
[0021] Figure 4 A cross-connection model diagram of the high-voltage cable sheath of this application;
[0022] Figure 5 This is a typical cross-connect wiring error simulation result diagram of this application;
[0023] Figure 6 This is a simulation diagram of the cross-connection wiring error of the 2nd and 3rd loops of this application;
[0024] Figure 7 This is a three-circuit defect simulation diagram of this application;
[0025] Figure 8 This is a diagram of the high-voltage cable intelligent grounding box for this application. DETAILED DESCRIPTION
[0026] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.
[0027] The terms "comprises," "comprising," or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.
[0028] The terms "first", "second", etc. are only used to distinguish one entity or operation from another entity or operation, and should not be understood as indicating or implying relative importance, nor should they be understood as requiring or implying any such actual relationship or order between these entities or operations.
[0029] like Figure 1 As shown, a method for detecting cross-connection faults of high-voltage cables comprises the following specific steps:
[0030] Establish an equivalent circuit model to simulate different high-voltage cable sheath cross-connection wiring error faults;
[0031] Extract fault characteristics and fault judgment criteria, build a classification table of misconnection modes, and encode the misconnection modes;
[0032] The intelligent grounding box of the high-voltage cable collects the sheath current phase information and the collection time of the first and last sheath currents. It uses the sheath current amplitude, the ratio of the sheath currents of the three loops and the sheath current phasor difference of the same loop as characteristic quantities to query the corresponding code and identify the fault type.
[0033] The equivalent circuit model adopts a three-phase nine-segment cross-interconnection. According to the three-phase nine-segment cross-interconnection, simulation models of different cross-interconnection misconnection modes are established to simulate different high-voltage cable sheath cross-interconnection wiring error faults.
[0034] The extraction of fault characteristic quantities and fault criteria is specifically to set the cable parameters such as core current, three-section lengths, phase-to-phase distance, and ambient temperature, consider the influence of cable wiring and unbalanced cross-connection parts, set the influence coefficient, and output the extracted fault characteristic quantities and fault criteria as sheath current amplitude and phase.
[0035] The construction of the misconnection mode classification table and encoding of the misconnection modes are specifically as follows:
[0036] According to the type of cross-connection misconnection, it is divided into four categories, namely normal situation, two typical cross-connection misconnection methods, double-circuit cross-connection misconnection, and three-circuit cross-connection misconnection. Under normal circumstances, both ends are connected with left-falling-left-falling-right or both ends are connected with right-falling-right-left, and the code is set to 0; the two typical connection methods are one end is connected with left-falling-left-falling-right and the other end is connected with right-falling-right-left, and the codes are set to A1 and A2; there are 9 types of double-circuit cross-connection misconnection, coded as B1 to B9; there are 23 types of three-circuit cross-connection misconnection, coded as C1 to C23.
[0037] The method for identifying the fault type uses an artificial neural network, which includes an input layer, a hidden layer, and an output layer. The input layer represents the input data, including on-site cable parameters and cable core current. The hidden layer is used for calculation and processing, and there can be one or more of them. Each neuron in the hidden layer will calculate the input received from the previous layer and generate an output. The output result of the output layer is the identified fault type.
[0038] Embodiment:
[0039] 1) Equivalent circuits for normal conditions and typical connection errors
[0040] To suppress the power frequency induced voltage, high-voltage cables over 1.2 kilometers generally adopt the cross-connected grounding method. A long cable line consists of multiple cross-connected segments as shown in Figure 2 Currently, high-voltage cables often adopt a three-phase nine-segment cross-connection, including two cross-connected grounding boxes and two grounding boxes. The 9 small segments of the high-voltage cable in the figure are represented by A1, A2, A3, B1, B2, B3, C1, C2, C3. The cable realizes metal transposition connection in the cross-connected grounding boxes CB1 and CB2. a1, a2, a3, a4, a5, a6 are current transformers installed on the direct grounding line of the high-voltage cable.
[0041] Due to short construction and maintenance periods and insufficient on-site personnel management, cross-connection methods are often incorrect. The equivalent circuits of the cross-connected high-voltage cable sheath grounding system under normal and typical connection errors are as shown in Figure 3 (a)(b)(c).
[0042] 2) Simulation models for different cross-connection errors are as shown in Figure 4 ,
[0043] The correct cross-connection method is that both cross-connected grounding boxes are connected as "撇撇捺" (a specific connection pattern description) or "捺捺撇", and a typical cross-connection wiring error is that one side is connected as "撇撇捺" and the other side is connected as "捺捺撇". The simulation diagrams of the three loops are as shown in Figure 5 .
[0044] The faults finally caused by cross-connection wiring errors are divided into two categories according to the affected loops. One category is three-loop defects, and the other is double-loop defects. The characteristics of double-loop defects are more obvious, as shown in Figure 6 .
[0045] It can be seen from the pointer diagram that when the front or back section of the cross-connection is connected incorrectly, the loop phase difference is much larger than the normal value, and when both the front and back sections of the cross-connection are connected incorrectly, the loop phase difference is larger than the normal value. In addition, the phase difference amplitudes of the first and last ends of the wrongly connected loops 2 and 3 are basically the same, and the phase difference is 180°. There are three types of wrong connection of loops 2 and 3, and the phase difference pointer diagrams of each type have obvious differences and are easy to distinguish.
[0046] The same is true for the wrong connection of circuits 1 and 2 and circuits 1 and 3. However, the current phase difference characteristics of the first and last sheath layers of the three-circuit defect are not obvious, such as Figure 7 As shown, three loop sheath current amplitudes need to be used for encoding.
[0047] The coding is divided into four categories. The connection mode in two normal situations is coded as 0, the cross-connection connection mode in two typical situations is coded as A1 and A2, there are 9 types of two-circuit defects, coded as B1, B2 to B9, and there are 23 types of three-circuit defects, coded as C1, C2 to C23.
[0048] 3) Artificial neural network (ANN) method for high voltage cable sheath grounding fault pattern recognition,
[0049] The working principle of artificial neural network (ANN) is to imitate the working mode of human brain neurons. Simply put, ANN is a network structure composed of a large number of artificial neurons (also called "nodes" or "units") connected in a certain hierarchical structure. A basic ANN model contains three levels, namely input layer, hidden layer and output layer. The input layer represents input data, such as field cable parameters, cable core current, etc.; the hidden layer is used for calculation and processing, and there can be one or more layers. Each neuron in the hidden layer will calculate the input it receives from the previous layer and generate output; the output layer represents the final output result.
[0050] In view of the influencing factors such as cable core current, length of three small cable segments, ground impedance and core voltage, it is necessary to obtain data on on-site cable parameter errors, cable core current, length of three small cable segments, ground impedance and core voltage. This may include historical records and real-time data; for cable parameter errors, the ill-conditioned tensor equation is preliminarily solved based on the minimum entropy principle and Tikhonov regularization principle; for measurement errors, the advantage of online monitoring that can obtain multiple sets of equations is combined with the advantage of high accuracy of deep neural network parameter estimation, and a deep neural network is established to solve the ill-conditioned tensor equation. The objective function is to minimize the Frobenius norm of the difference between the tensor equation parameter matrices obtained by the two solution methods, and an intelligent solution method for ill-conditioned tensor equations suitable for online monitoring is adopted.
[0051] After that, data preprocessing is performed. Data preprocessing usually includes steps such as removing outliers, filling missing values, and data standardization to clean and organize the data to ensure that it is suitable for subsequent analysis. In the feature engineering stage, it is necessary to determine the key factors that affect the performance of cables on site. This can be done through correlation analysis, principal component analysis (PCA), and other methods. For the features related to the target task, we should retain them for model training. Finally, the effect of the model needs to be evaluated through representative evaluation criteria (such as precision, recall rate, F1 score, etc.). If the effect is not good, it may be necessary to return to the feature engineering stage, reselect or create features, or adjust the structure and parameters of the model. If the evaluation results of the model meet expectations, the model can be applied to the actual scenario of cable fault detection.
[0052] 4) Data collection of sheath current phasor difference of high voltage cable sheath grounding fault
[0053] A low-cost, 100kHz, 0.2-level sheath ground current online monitoring sensor and sampling hardware unit are used, including FPGA sampling, synchronous trigger signal, etc. The synchronous trigger signal source can be sent to all sensor devices to tell them when to start measuring or sampling. At the same time, this synchronous trigger signal can also be used to tell the data processing or recording device when to start receiving and processing data.
[0054] Use high-precision broadband high-voltage cable intelligent grounding box such as Figure 8 As shown, a high-precision cable detection main component is arranged in the grounding box, which includes a detection host 1, a sheath current detection component 2, a cable current sensor 3, and a high-frequency sensor 4. A plurality of grounding terminals 5 are arranged in the grounding box, and the single-core cable is grounded through the grounding terminal 5, and the grounding terminals are cross-connected. The sheath current detection component includes a plurality of sheath current detection sensors, and the sheath current detection sensors are sleeved on the metal sheath of the single-core cable. A waveform extraction module is arranged in the detection host 1 to provide a phase difference feature quantity for CNN recognition. The high-frequency sensor 4 and the sheath current detection component 2 are electrically connected to the input end of the detection host 1.
[0055] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A method for detecting cross-connection faults of high-voltage cables, characterized in that: The specific steps include: Establish an equivalent circuit model to simulate different high-voltage cable sheath cross-connection wiring error faults; Extract fault characteristics and fault judgment criteria, build a classification table of misconnection modes, and encode the misconnection modes; The intelligent grounding box of the high-voltage cable collects the sheath current phase information and the collection time of the first and last sheath currents. It uses the sheath current amplitude, the ratio of the sheath currents of the three loops and the sheath current phasor difference of the same loop as characteristic quantities to query the corresponding code and identify the fault type.
2. A method for detecting cross-connection errors of high-voltage cables according to claim 1, characterized in that: The equivalent circuit model adopts a three-phase nine-segment cross-interconnection. According to the three-phase nine-segment cross-interconnection, simulation models of different cross-interconnection misconnection modes are established to simulate different high-voltage cable sheath cross-interconnection wiring error faults.
3. A method for detecting cross-connection errors of high-voltage cables according to claim 1, characterized in that: The extraction of fault characteristic quantities and fault criteria is specifically to set the cable parameters such as core current, three-section lengths, phase-to-phase distance, and ambient temperature, consider the influence of cable wiring and unbalanced cross-connection parts, set the influence coefficient, and output the extracted fault characteristic quantities and fault criteria as sheath current amplitude and phase.
4. A method for detecting cross-connection faults of high-voltage cables according to claim 1, characterized in that: The construction of the misconnection mode classification table and encoding of the misconnection modes are specifically as follows: According to the type of cross-connection misconnection, it is divided into four categories, namely normal situation, two typical cross-connection misconnection methods, double-circuit cross-connection misconnection, and three-circuit cross-connection misconnection. Under normal circumstances, both ends are connected with left-falling-left-falling-right or both ends are connected with right-falling-right-left, and the code is set to 0; the two typical connection methods are one end is connected with left-falling-left-falling-right and the other end is connected with right-falling-right-left, and the codes are set to A1 and A2; there are 9 types of double-circuit cross-connection misconnection, coded as B1 to B9; there are 23 types of three-circuit cross-connection misconnection, coded as C1 to C23.
5. A method for detecting cross-connection faults of high-voltage cables according to claim 1, characterized in that: The fault type identification adopts an artificial neural network method. The artificial neural network includes an input layer, a hidden layer and an output layer. The input layer represents input data, including on-site cable parameters and cable core current; the hidden layer is used for calculation and processing. There can be one or more hidden layers. Each neuron in the hidden layer will calculate the input it receives from the previous layer and generate an output; the output layer outputs the result of the identified fault type.
6. A high-voltage cable intelligent grounding box, used to implement the high-voltage cable cross-connection fault detection method as described in any one of claims 1 to 5, characterized in that: It includes a detection host, a sheath current detection component, a cable current sensor, and a high-frequency sensor. A grounding terminal is arranged in the grounding box, and the single-core cable is grounded through the grounding terminal. The grounding terminals are cross-connected. The sheath current detection component includes a sheath current detection sensor, and the sheath current detection sensor is sleeved on the metal sheath of the single-core cable. A waveform extraction module is arranged in the detection host to provide phase difference feature quantities for CNN recognition, and the high-frequency sensor and the sheath current detection component are electrically connected to the input end of the detection host.