Combiner box fault monitoring method
By setting multiple monitoring indicators and fault impact models, we can quickly judge the bus box fault and evaluate the degree of impact, and solve the problem of difficulty in detecting bus box faults in the existing technology, and improve the timeliness of fault monitoring and operation and maintenance efficiency.
Patent Information
- Application Number
- CN202410974744.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2024-11-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, it is difficult to detect the fault of the bus box in a timely manner, resulting in low operation and maintenance efficiency and high cost, and failures will cause other equipment failures or downtime.
By setting multiple monitoring indicators, obtain real-time monitoring data and generate monitoring evaluation value, filter fault characteristic data, establish a fault impact model, send alarm signals based on the fault impact coefficient, formulate operation and maintenance strategies, and improve fault monitoring timeliness and operation and maintenance efficiency. .
It realizes rapid judgment of bus box failures, screens fault characteristic data, evaluates the degree of impact on other equipment, sends alarm signals in a timely manner, and formulates operation and maintenance strategies, which greatly improves the timeliness of fault monitoring and operation and maintenance efficiency.
Smart Images

Figure CN119051590A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of fault monitoring, and in particular to a combiner box fault monitoring method. Background Art
[0002] The combiner box is a wiring device that ensures the orderly connection and convergence of photovoltaic modules in the photovoltaic power generation system. This device can ensure that the photovoltaic system is easy to cut off the circuit during maintenance and inspection, and reduce the scope of power outage when the photovoltaic system fails. Although the function of the combiner box is relatively simple, the failure rate in the photovoltaic system is not low. Fuse blowing, fire, communication failure and other problems are also common problems in photovoltaic power stations.
[0003] In the prior art, combiner box fault monitoring mainly relies on monitoring system alarms and manual on-site inspections. However, the number of combiner boxes is large and their distribution range is wide, which often makes it difficult to detect faults in time after they occur. Combiner box failures can also cause other equipment failures or shutdowns, greatly reducing operation and maintenance efficiency and increasing operation and maintenance costs. Summary of the invention
[0004] In order to solve the above technical problems, the present application provides a method for monitoring a junction box fault, which sets multiple monitoring indicators, obtains real-time monitoring data of each monitoring indicator, and generates a monitoring evaluation value of the corresponding monitoring indicator, thereby obtaining a total monitoring evaluation value of the junction box. According to the total monitoring evaluation value, it is possible to quickly determine whether a junction box fault occurs. If so, the fault characteristic data is screened and the fault influence coefficient is obtained. According to the fault influence coefficient, the degree of impact on other equipment is obtained, and an alarm signal is sent in time to formulate an operation and maintenance strategy, which greatly improves the timeliness of fault monitoring and the operation and maintenance efficiency.
[0005] In some embodiments of the present application, a combiner box fault monitoring method is provided, comprising:
[0006] Set multiple monitoring indicators according to the historical fault data of the combiner box, obtain real-time monitoring data of each monitoring indicator, and generate a real-time operation curve of the corresponding monitoring indicator according to the real-time monitoring data;
[0007] Generate monitoring evaluation values of corresponding monitoring indicators according to real-time operation curves, obtain total monitoring evaluation values of combiner boxes according to monitoring evaluation values, and obtain fault risk coefficients according to total monitoring evaluation values. If the fault risk coefficient is greater than a preset risk coefficient threshold, filter out fault feature data;
[0008] A fault impact model is established, fault feature data is input into the fault impact model, a fault impact coefficient is obtained, and an alarm signal is sent according to the fault impact coefficient.
[0009] In some embodiments of the present application, multiple monitoring indicators are set according to the historical fault data of the combiner box, including:
[0010] Preprocessing the historical fault data, wherein the preprocessing includes unifying time units, removing outliers, filling missing values, and data cleaning;
[0011] Determine the historical fault level of the combiner box according to the preprocessed historical fault data, obtain historical fault data of different historical fault levels, and set monitoring indicators according to the historical fault levels, wherein the monitoring indicators include primary monitoring indicators, secondary monitoring indicators and tertiary monitoring indicators;
[0012] According to the historical fault data of different historical fault levels, the monitoring equipment and monitoring points of the corresponding monitoring indicators are set, and the real-time monitoring data of the corresponding monitoring indicators are obtained according to the monitoring equipment and monitoring points.
[0013] In some embodiments of the present application, generating a real-time operation curve of a corresponding monitoring indicator according to real-time monitoring data includes:
[0014] Draw a first-level real-time operation curve for the real-time monitoring data of the first-level monitoring indicators collected at the monitoring points, where the abscissa of the first-level real-time operation curve is time and the ordinate is the real-time monitoring data of the first-level monitoring indicators;
[0015] Draw a secondary real-time operation curve for the real-time monitoring data of the secondary monitoring indicators collected at the monitoring points, where the horizontal axis of the secondary real-time operation curve is time and the vertical axis is the real-time monitoring data of the secondary monitoring indicators;
[0016] A three-level real-time operation curve is drawn based on the real-time monitoring data of the three-level monitoring indicators collected at the monitoring points. The abscissa of the three-level real-time operation curve is time, and the ordinate is the real-time monitoring data of the three-level monitoring indicators.
[0017] In some embodiments of the present application, generating a monitoring evaluation value of a corresponding monitoring indicator according to a real-time operation curve includes:
[0018] Screen the fluctuation characteristic quantities and fluctuation time nodes of the real-time operation curve of the same monitoring indicator at the same monitoring point, and map the fluctuation time nodes to the fluctuation characteristic quantities one by one;
[0019] Comprehensively analyze the fluctuation characteristic quantity of the same monitoring indicator at the same monitoring point at the same fluctuation time node to obtain a comprehensive fluctuation characteristic quantity, compare the comprehensive fluctuation characteristic quantity with a preset fluctuation characteristic threshold, and perform amplitude analysis on the comprehensive fluctuation characteristic quantity according to the comparison result to obtain a fluctuation evaluation value at the corresponding fluctuation time node, wherein the fluctuation evaluation value includes a first fluctuation evaluation value P1, a second fluctuation evaluation value P2 and a third fluctuation evaluation value P3;
[0020] According to the fluctuation time node and fluctuation evaluation value of the same monitoring point, a one-to-one mapping is performed, and the comprehensive fluctuation evaluation value of the corresponding monitoring point is calculated;
[0021]
[0022] Wherein, mi is the comprehensive fluctuation evaluation value of the ith monitoring point, k is the number of fluctuation time nodes of the ith monitoring point, ts is the sth fluctuation time node, Pjs is the jth fluctuation evaluation value of the sth fluctuation time node, where s = 1, 2, ... k, j = 1, 2, 3;
[0023] Generate a monitoring evaluation value of the corresponding monitoring indicator based on the comprehensive fluctuation evaluation values of multiple monitoring points of the same monitoring indicator and the weight coefficients of the corresponding monitoring points;
[0024]
[0025] Among them, Hr is the monitoring evaluation value of the r-level monitoring indicator, where r = 1, 2, 3, nr is the number of monitoring points of the r-level monitoring indicator, mir is the comprehensive fluctuation evaluation value of the i-th monitoring point in the r-level monitoring indicator, eir is the weight coefficient of the i-th monitoring point in the r-level monitoring indicator, i = 1, 2, ...n;
[0026] The total monitoring evaluation value of the combiner box is generated according to the weight coefficient of the monitoring indicator and the monitoring evaluation value.
[0027] In some embodiments of the present application, the failure risk coefficient is obtained according to the total monitoring evaluation value, including:
[0028] Presetting a first preset evaluation value interval, a second preset evaluation value interval, a third preset evaluation value interval and a fourth preset evaluation value interval;
[0029] When the total monitoring evaluation value is within the first preset evaluation value interval, selecting the fourth preset fault risk coefficient as the current fault risk coefficient;
[0030] When the total monitoring evaluation value is within the second preset evaluation value interval, the third preset fault risk coefficient is selected as the current fault risk coefficient;
[0031] When the total monitoring evaluation value is within the third preset evaluation value interval, the second preset fault risk coefficient is selected as the current fault risk coefficient;
[0032] When the total monitoring evaluation value is in the fourth preset evaluation value interval, the first preset fault risk coefficient is selected as the current fault risk coefficient.
[0033] In some embodiments of the present application, if the fault risk coefficient is greater than a preset risk coefficient threshold, the fault characteristic data is screened out, including:
[0034] If the fault risk coefficient is greater than the preset risk coefficient threshold, the fluctuation evaluation values of different fluctuation time nodes at multiple monitoring points in each monitoring indicator of the combiner box are analyzed, and the fluctuation time nodes and monitoring points whose fluctuation evaluation values are greater than the preset fluctuation evaluation value threshold are screened, and the corresponding fluctuation time nodes and monitoring points are set as feature nodes and feature points;
[0035] Obtain the real-time monitoring data of the adjacent time nodes at the same monitoring point of the current feature node. If the real-time monitoring data at the adjacent time nodes are in the standard monitoring data interval, the current feature node is removed, and the real-time monitoring data at the remaining feature nodes are set as fault feature data;
[0036] According to the degree of deviation between the real-time monitoring data at the adjacent time nodes of the remaining characteristic nodes and the standard detection data interval, the first coefficient of the fault characteristic coefficient is set;
[0037] The second coefficient of the fault feature data is set according to the weight coefficient of the feature point corresponding to the remaining feature nodes, the fault feature data is sorted according to the first coefficient and the second coefficient, and the fault feature data is input into the fault impact model in the order of arrangement to obtain the fault impact coefficient.
[0038] In some embodiments of the present application, establishing a fault impact model includes:
[0039] Obtain historical operation data of associated devices corresponding to historical fault data of the combiner box;
[0040] The historical operation data of the associated equipment is classified and judged according to the classification rules of the equipment operation status, and the historical operation data affected by the historical fault data under the same mapping node is determined, and the current corresponding historical fault data is used as the input data set, and the historical operation data of the same mapping node of the current historical fault data is used as the first output data set;
[0041] Comparing the historical operation data in the first output data set with the corresponding preset standard operation data, obtaining the state coefficient of the historical operation data in the first output data set according to the comparison result, and using the state coefficient of the historical operation data in the first output data set as the second output data set;
[0042] The input data set, the first output data set and the second output data set are used to generate a training set and a test set, and a neural network model is trained according to the training set to generate an initial fault impact model;
[0043] The credibility of the initial fault impact model is obtained according to the test set. If the credibility is greater than the preset credibility threshold, the initial fault impact model is set as the fault impact model. If the credibility is less than the preset credibility threshold, the training set is iteratively trained to regenerate the initial fault impact model.
[0044] In some embodiments of the present application, the fault feature data is input into the fault impact model to obtain the fault impact coefficient, including:
[0045] Input the fault characteristic data into the fault impact model to obtain the real-time operation data and real-time status coefficient of the current associated equipment;
[0046] Generate a fault influence coefficient of current fault feature data according to the influence degree coefficient of the associated equipment and the real-time status coefficient;
[0047]
[0048] Wherein, d is the fault impact coefficient, q is the number of associated devices, yw is the real-time status coefficient of the w-th associated device, and hw is the influence degree coefficient of the w-th associated device.
[0049] In some embodiments of the present application, before sending an alarm signal according to the fault influence coefficient, the method further includes:
[0050] Comprehensively analyze the impact coefficient and real-time status coefficient of associated equipment to obtain the fault urgency of each associated equipment;
[0051] Sort the faults according to their urgency to obtain a fault sequence. According to the fault feature data corresponding to each fault urgency in the fault operation and maintenance sequence, filter out the fault type and fault point corresponding to the fault feature data. According to the fault type and fault point, filter out the corresponding historical operation and maintenance strategy.
[0052] Map the historical operation and maintenance strategies to the fault urgency level one by one to obtain the operation and maintenance sequence;
[0053] Send an alarm signal based on the fault impact coefficient, and perform operation and maintenance management according to the corresponding operation and maintenance sequence.
[0054] In some embodiments of the present application, sending an alarm signal according to the fault influence coefficient includes:
[0055] Presetting a first preset fault influence coefficient threshold and a second preset fault influence coefficient threshold;
[0056] When the fault impact coefficient is less than the first preset fault impact coefficient threshold, a first-level alarm signal is sent;
[0057] When the fault influence coefficient is within the first preset fault influence coefficient threshold and the second preset fault influence coefficient threshold, a secondary alarm signal is sent;
[0058] When the fault influence coefficient is greater than the second preset fault influence coefficient threshold, a third-level alarm signal is sent.
[0059] Compared with the prior art, the combiner box fault monitoring method of the embodiment of the present application has the following beneficial effects:
[0060] By setting multiple monitoring indicators, obtaining real-time monitoring data for each monitoring indicator and generating monitoring evaluation values for the corresponding monitoring indicators, the total monitoring evaluation value of the combiner box can be obtained. According to the total monitoring evaluation value, it is possible to quickly determine whether the combiner box has a fault. If so, the fault characteristic data is screened and the fault impact coefficient is obtained. According to the fault impact coefficient, the degree of impact on other equipment is obtained, and alarm signals are sent in time to formulate operation and maintenance strategies, which greatly improves the timeliness of fault monitoring and the efficiency of operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS
[0061] Figure 1 It is a flow chart of a combiner box fault monitoring method in a preferred embodiment of the present application. DETAILED DESCRIPTION
[0062] The specific implementation methods of the present application are further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present application but are not intended to limit the scope of the present application.
[0063] In the description of the present application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.
[0064] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.
[0065] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.
[0066] like Figure 1 As shown, a combiner box fault monitoring method according to a preferred embodiment of the present application includes:
[0067] Set multiple monitoring indicators according to the historical fault data of the combiner box, obtain real-time monitoring data of each monitoring indicator, and generate a real-time operation curve of the corresponding monitoring indicator according to the real-time monitoring data;
[0068] Generate monitoring evaluation values of corresponding monitoring indicators according to real-time operation curves, obtain total monitoring evaluation values of combiner boxes according to monitoring evaluation values, and obtain fault risk coefficients according to total monitoring evaluation values. If the fault risk coefficient is greater than a preset risk coefficient threshold, filter out fault feature data;
[0069] A fault impact model is established, fault feature data is input into the fault impact model, a fault impact coefficient is obtained, and an alarm signal is sent according to the fault impact coefficient.
[0070] In some embodiments of the present application, multiple monitoring indicators are set according to the historical fault data of the combiner box, including:
[0071] Preprocessing the historical fault data, wherein the preprocessing includes unifying time units, removing outliers, filling missing values, and data cleaning;
[0072] Determine the historical fault level of the combiner box according to the preprocessed historical fault data, obtain historical fault data of different historical fault levels, and set monitoring indicators according to the historical fault levels, wherein the monitoring indicators include primary monitoring indicators, secondary monitoring indicators and tertiary monitoring indicators;
[0073] According to the historical fault data of different historical fault levels, the monitoring equipment and monitoring points of the corresponding monitoring indicators are set, and the real-time monitoring data of the corresponding monitoring indicators are obtained according to the monitoring equipment and monitoring points.
[0074] In some embodiments of the present application, generating a real-time operation curve of a corresponding monitoring indicator according to real-time monitoring data includes:
[0075] Draw a first-level real-time operation curve for the real-time monitoring data of the first-level monitoring indicators collected at the monitoring points, where the abscissa of the first-level real-time operation curve is time and the ordinate is the real-time monitoring data of the first-level monitoring indicators;
[0076] Draw a secondary real-time operation curve for the real-time monitoring data of the secondary monitoring indicators collected at the monitoring points, where the horizontal axis of the secondary real-time operation curve is time and the vertical axis is the real-time monitoring data of the secondary monitoring indicators;
[0077] A three-level real-time operation curve is drawn based on the real-time monitoring data of the three-level monitoring indicators collected at the monitoring points. The abscissa of the three-level real-time operation curve is time, and the ordinate is the real-time monitoring data of the three-level monitoring indicators.
[0078] In some embodiments of the present application, generating a monitoring evaluation value of a corresponding monitoring indicator according to a real-time operation curve includes:
[0079] Screen the fluctuation characteristic quantities and fluctuation time nodes of the real-time operation curve of the same monitoring indicator at the same monitoring point, and map the fluctuation time nodes to the fluctuation characteristic quantities one by one;
[0080] Comprehensively analyze the fluctuation characteristic quantity of the same monitoring indicator at the same monitoring point at the same fluctuation time node to obtain a comprehensive fluctuation characteristic quantity, compare the comprehensive fluctuation characteristic quantity with a preset fluctuation characteristic threshold, and perform amplitude analysis on the comprehensive fluctuation characteristic quantity according to the comparison result to obtain a fluctuation evaluation value at the corresponding fluctuation time node, wherein the fluctuation evaluation value includes a first fluctuation evaluation value P1, a second fluctuation evaluation value P2 and a third fluctuation evaluation value P3;
[0081] According to the fluctuation time node and fluctuation evaluation value of the same monitoring point, a one-to-one mapping is performed, and the comprehensive fluctuation evaluation value of the corresponding monitoring point is calculated;
[0082]
[0083] Wherein, mi is the comprehensive fluctuation evaluation value of the ith monitoring point, k is the number of fluctuation time nodes of the ith monitoring point, ts is the sth fluctuation time node, Pjs is the jth fluctuation evaluation value of the sth fluctuation time node, where s = 1, 2, ... k, j = 1, 2, 3;
[0084] Generate a monitoring evaluation value of the corresponding monitoring indicator based on the comprehensive fluctuation evaluation values of multiple monitoring points of the same monitoring indicator and the weight coefficients of the corresponding monitoring points;
[0085]
[0086] Among them, Hr is the monitoring evaluation value of the r-level monitoring indicator, where r = 1, 2, 3, nr is the number of monitoring points of the r-level monitoring indicator, mir is the comprehensive fluctuation evaluation value of the i-th monitoring point in the r-level monitoring indicator, eir is the weight coefficient of the i-th monitoring point in the r-level monitoring indicator, i = 1, 2, ...n;
[0087] The total monitoring evaluation value of the combiner box is generated according to the weight coefficient of the monitoring indicator and the monitoring evaluation value.
[0088] In this embodiment, the preset fluctuation characteristic threshold is set based on the normal fluctuation boundary data in the monitoring data of the monitoring point corresponding to the junction box. When the comprehensive fluctuation characteristic quantity at the same fluctuation time node of the same monitoring point is smaller than the preset fluctuation characteristic threshold, the fluctuation evaluation value is larger, that is, the real-time monitoring data fluctuates less and is less likely to fail. When it is greater than the preset fluctuation characteristic threshold, the fluctuation evaluation value is smaller, that is, it is more likely to fail.
[0089] In this embodiment, the monitoring evaluation value of the corresponding monitoring indicator is calculated according to the comprehensive fluctuation evaluation value at multiple monitoring points of each monitoring indicator, so as to improve the accuracy of the monitoring evaluation value of the corresponding monitoring indicator of the junction box, thereby ensuring accurate judgment of the fault state of the junction box.
[0090] In some embodiments of the present application, the failure risk coefficient is obtained according to the total monitoring evaluation value, including:
[0091] Presetting a first preset evaluation value interval, a second preset evaluation value interval, a third preset evaluation value interval and a fourth preset evaluation value interval;
[0092] When the total monitoring evaluation value is within the first preset evaluation value interval, selecting the fourth preset fault risk coefficient as the current fault risk coefficient;
[0093] When the total monitoring evaluation value is within the second preset evaluation value interval, the third preset fault risk coefficient is selected as the current fault risk coefficient;
[0094] When the total monitoring evaluation value is within the third preset evaluation value interval, the second preset fault risk coefficient is selected as the current fault risk coefficient;
[0095] When the total monitoring evaluation value is in the fourth preset evaluation value interval, the first preset fault risk coefficient is selected as the current fault risk coefficient.
[0096] In some embodiments of the present application, if the fault risk coefficient is greater than a preset risk coefficient threshold, the fault characteristic data is screened out, including:
[0097] If the fault risk coefficient is greater than the preset risk coefficient threshold, the fluctuation evaluation values of different fluctuation time nodes at multiple monitoring points in each monitoring indicator of the combiner box are analyzed, and the fluctuation time nodes and monitoring points whose fluctuation evaluation values are greater than the preset fluctuation evaluation value threshold are screened, and the corresponding fluctuation time nodes and monitoring points are set as feature nodes and feature points;
[0098] Obtain the real-time monitoring data of the adjacent time nodes at the same monitoring point of the current feature node. If the real-time monitoring data at the adjacent time nodes are in the standard monitoring data interval, the current feature node is removed, and the real-time monitoring data at the remaining feature nodes are set as fault feature data;
[0099] According to the degree of deviation between the real-time monitoring data at the adjacent time nodes of the remaining characteristic nodes and the standard detection data interval, the first coefficient of the fault characteristic coefficient is set;
[0100] The second coefficient of the fault feature data is set according to the weight coefficient of the feature point corresponding to the remaining feature nodes, the fault feature data is sorted according to the first coefficient and the second coefficient, and the fault feature data is input into the fault impact model in the order of arrangement to obtain the fault impact coefficient.
[0101] In this embodiment, characteristic nodes are screened out according to the fluctuation evaluation values at the fluctuation time nodes of the monitoring points, and real-time monitoring data after the characteristic nodes are obtained to determine whether the real-time monitoring data after the characteristic nodes are in the standard monitoring data interval. If so, it means that the data after the current characteristic node is in a normal state, and the corresponding characteristic nodes are eliminated. If not, the real-time monitoring data at the characteristic nodes are set as fault characteristic data.
[0102] In this embodiment, characteristic points and characteristic nodes are screened out according to the fluctuation evaluation value of the fluctuation time node at each monitoring point, and it is judged again whether the characteristic node has a fault to ensure the accuracy of the junction box fault judgment. If so, the fault characteristic data is obtained and the arrangement order of the fault characteristic data is set, so as to quickly obtain the fault influence coefficient, send an alarm signal, and reduce the fault loss.
[0103] In some embodiments of the present application, establishing a fault impact model includes:
[0104] Obtain historical operation data of associated devices corresponding to historical fault data of the combiner box;
[0105] The historical operation data of the associated equipment is classified and judged according to the classification rules of the equipment operation status, and the historical operation data affected by the historical fault data under the same mapping node is determined, and the current corresponding historical fault data is used as the input data set, and the historical operation data of the same mapping node of the current historical fault data is used as the first output data set;
[0106] Comparing the historical operation data in the first output data set with the corresponding preset standard operation data, obtaining the state coefficient of the historical operation data in the first output data set according to the comparison result, and using the state coefficient of the historical operation data in the first output data set as the second output data set;
[0107] The input data set, the first output data set and the second output data set are used to generate a training set and a test set, and a neural network model is trained according to the training set to generate an initial fault impact model;
[0108] The credibility of the initial fault impact model is obtained according to the test set. If the credibility is greater than the preset credibility threshold, the initial fault impact model is set as the fault impact model. If the credibility is less than the preset credibility threshold, the training set is iteratively trained to regenerate the initial fault impact model.
[0109] In some embodiments of the present application, the fault feature data is input into the fault impact model to obtain the fault impact coefficient, including:
[0110] Input the fault characteristic data into the fault impact model to obtain the real-time operation data and real-time status coefficient of the current associated equipment;
[0111] Generate a fault influence coefficient of current fault feature data according to the influence degree coefficient of the associated equipment and the real-time status coefficient;
[0112]
[0113] Wherein, d is the fault impact coefficient, q is the number of associated devices, yw is the real-time status coefficient of the w-th associated device, and hw is the influence degree coefficient of the w-th associated device.
[0114] In this embodiment, the associated equipment is the equipment that is affected when the combiner box fails. The real-time status coefficient of the associated equipment is obtained according to the impact of the combiner box failure on the associated equipment. The impact degree coefficient of the associated equipment refers to the importance of the associated equipment in the operation process. According to the fault impact coefficient, an alarm signal is sent in time and a reasonable operation and maintenance strategy is formulated to reduce the loss and cost caused by the combiner box failure.
[0115] In some embodiments of the present application, before sending an alarm signal according to the fault influence coefficient, the method further includes:
[0116] Comprehensively analyze the impact coefficient and real-time status coefficient of associated equipment to obtain the fault urgency of each associated equipment;
[0117] Sort the faults according to their urgency to obtain a fault sequence. According to the fault feature data corresponding to each fault urgency in the fault operation and maintenance sequence, filter out the fault type and fault point corresponding to the fault feature data. According to the fault type and fault point, filter out the corresponding historical operation and maintenance strategy.
[0118] Map the historical operation and maintenance strategies to the fault urgency level one by one to obtain the operation and maintenance sequence;
[0119] Send an alarm signal based on the fault impact coefficient, and perform operation and maintenance management according to the corresponding operation and maintenance sequence.
[0120] In some embodiments of the present application, sending an alarm signal according to the fault influence coefficient includes:
[0121] Presetting a first preset fault influence coefficient threshold and a second preset fault influence coefficient threshold;
[0122] When the fault impact coefficient is less than the first preset fault impact coefficient threshold, a first-level alarm signal is sent;
[0123] When the fault influence coefficient is within the first preset fault influence coefficient threshold and the second preset fault influence coefficient threshold, a secondary alarm signal is sent;
[0124] When the fault influence coefficient is greater than the second preset fault influence coefficient threshold, a third-level alarm signal is sent.
[0125] In this embodiment, when the fault impact coefficient is less than the first preset fault impact coefficient threshold, it means that the combiner box has a fault but has no impact on the associated equipment. The first-level alarm signal means that only the combiner box is faulty, and an operation and maintenance strategy is formulated based on the fault characteristic data of the combiner box. When the fault impact coefficient is between the first preset fault impact coefficient threshold and the second preset fault impact coefficient threshold, it means that the combiner box has a fault and has a small impact on the associated equipment. A second-level alarm signal is sent and operation and maintenance are performed in accordance with the corresponding operation and maintenance sequence. When the fault impact coefficient is greater than the second preset fault impact coefficient, it means that the combiner box has a fault and has a large impact on the associated equipment. A third-level alarm signal is sent and operation and maintenance are performed in accordance with the corresponding operation and maintenance sequence.
[0126] In summary, the present invention discloses a method for monitoring a junction box fault, comprising: setting a plurality of monitoring indicators according to historical fault data of the junction box, and obtaining real-time monitoring data of each monitoring indicator, and generating a real-time operation curve of the corresponding monitoring indicator according to the real-time monitoring data; generating a monitoring evaluation value of the corresponding monitoring indicator according to the real-time operation curve, obtaining a total monitoring evaluation value of the junction box according to the monitoring evaluation value, and obtaining a fault risk coefficient according to the total monitoring evaluation value, and if the fault risk coefficient is greater than a preset risk coefficient threshold, screening out fault feature data; establishing a fault impact model, inputting the fault feature data into the fault impact model, obtaining a fault impact coefficient, and sending an alarm signal according to the fault impact coefficient. By setting a plurality of monitoring indicators, obtaining real-time monitoring data of each monitoring indicator and generating a monitoring evaluation value of the corresponding monitoring indicator, the total monitoring evaluation value of the junction box is obtained, and whether a fault occurs in the junction box can be quickly determined according to the total monitoring evaluation value. If so, screening the fault feature data and obtaining the fault impact coefficient, obtaining the degree of influence on other equipment according to the fault impact coefficient, sending an alarm signal in time and formulating an operation and maintenance strategy, thereby greatly improving the timeliness and operation and maintenance efficiency of fault monitoring.
[0127] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and substitutions can be made without departing from the technical principles of the present application. These improvements and substitutions should also be regarded as the scope of protection of the present application.
Claims
1. A combiner box fault monitoring method, characterized in that: include: Set multiple monitoring indicators according to the historical fault data of the combiner box, obtain real-time monitoring data of each monitoring indicator, and generate a real-time operation curve of the corresponding monitoring indicator according to the real-time monitoring data; Generate monitoring evaluation values of corresponding monitoring indicators according to real-time operation curves, obtain total monitoring evaluation values of combiner boxes according to monitoring evaluation values, and obtain fault risk coefficients according to total monitoring evaluation values. If the fault risk coefficient is greater than a preset risk coefficient threshold, filter out fault feature data; A fault impact model is established, fault feature data is input into the fault impact model, a fault impact coefficient is obtained, and an alarm signal is sent according to the fault impact coefficient.
2. The combiner box fault monitoring method according to claim 1, characterized in that: According to the historical fault data of the combiner box, multiple monitoring indicators are set, including: Preprocessing the historical fault data, wherein the preprocessing includes unifying time units, removing outliers, filling missing values, and data cleaning; Determine the historical fault level of the combiner box according to the preprocessed historical fault data, obtain historical fault data of different historical fault levels, and set monitoring indicators according to the historical fault levels, wherein the monitoring indicators include primary monitoring indicators, secondary monitoring indicators and tertiary monitoring indicators; According to the historical fault data of different historical fault levels, the monitoring equipment and monitoring points of the corresponding monitoring indicators are set, and the real-time monitoring data of the corresponding monitoring indicators are obtained according to the monitoring equipment and monitoring points.
3. The combiner box fault monitoring method according to claim 2, characterized in that: Generate real-time operation curves of corresponding monitoring indicators based on real-time monitoring data, including: Draw a first-level real-time operation curve for the real-time monitoring data of the first-level monitoring indicators collected at the monitoring points, where the abscissa of the first-level real-time operation curve is time and the ordinate is the real-time monitoring data of the first-level monitoring indicators; Draw a secondary real-time operation curve for the real-time monitoring data of the secondary monitoring indicators collected at the monitoring points, where the horizontal axis of the secondary real-time operation curve is time and the vertical axis is the real-time monitoring data of the secondary monitoring indicators; A three-level real-time operation curve is drawn based on the real-time monitoring data of the three-level monitoring indicators collected at the monitoring points. The abscissa of the three-level real-time operation curve is time, and the ordinate is the real-time monitoring data of the three-level monitoring indicators.
4. The combiner box fault monitoring method according to claim 3, characterized in that: Generate monitoring evaluation values of corresponding monitoring indicators based on real-time operation curves, including: Screen the fluctuation characteristic quantities and fluctuation time nodes of the real-time operation curve of the same monitoring indicator at the same monitoring point, and map the fluctuation time nodes to the fluctuation characteristic quantities one by one; Comprehensively analyze the fluctuation characteristic quantity of the same monitoring indicator at the same monitoring point at the same fluctuation time node to obtain a comprehensive fluctuation characteristic quantity, compare the comprehensive fluctuation characteristic quantity with a preset fluctuation characteristic threshold, and perform amplitude analysis on the comprehensive fluctuation characteristic quantity according to the comparison result to obtain a fluctuation evaluation value at the corresponding fluctuation time node, wherein the fluctuation evaluation value includes a first fluctuation evaluation value P1, a second fluctuation evaluation value P2 and a third fluctuation evaluation value P3; According to the fluctuation time node and fluctuation evaluation value of the same monitoring point, a one-to-one mapping is performed, and the comprehensive fluctuation evaluation value of the corresponding monitoring point is calculated; Wherein, mi is the comprehensive fluctuation evaluation value of the ith monitoring point, k is the number of fluctuation time nodes of the ith monitoring point, ts is the sth fluctuation time node, Pjs is the jth fluctuation evaluation value of the sth fluctuation time node, where s = 1, 2, ... k, j = 1, 2, 3; Generate a monitoring evaluation value of the corresponding monitoring indicator based on the comprehensive fluctuation evaluation values of multiple monitoring points of the same monitoring indicator and the weight coefficients of the corresponding monitoring points; Among them, Hr is the monitoring evaluation value of the r-level monitoring indicator, where r = 1, 2, 3, nr is the number of monitoring points of the r-level monitoring indicator, mir is the comprehensive fluctuation evaluation value of the i-th monitoring point in the r-level monitoring indicator, eir is the weight coefficient of the i-th monitoring point in the r-level monitoring indicator, i = 1, 2, ...n; The total monitoring evaluation value of the combiner box is generated according to the weight coefficient of the monitoring indicator and the monitoring evaluation value.
5. The combiner box fault monitoring method according to claim 4, characterized in that: The failure risk factor is obtained based on the total monitoring evaluation value, including: Presetting a first preset evaluation value interval, a second preset evaluation value interval, a third preset evaluation value interval and a fourth preset evaluation value interval; When the total monitoring evaluation value is within the first preset evaluation value interval, selecting the fourth preset fault risk coefficient as the current fault risk coefficient; When the total monitoring evaluation value is within the second preset evaluation value interval, the third preset fault risk coefficient is selected as the current fault risk coefficient; When the total monitoring evaluation value is within the third preset evaluation value interval, the second preset fault risk coefficient is selected as the current fault risk coefficient; When the total monitoring evaluation value is in the fourth preset evaluation value interval, the first preset fault risk coefficient is selected as the current fault risk coefficient.
6. The combiner box fault monitoring method according to claim 5, characterized in that: If the fault risk factor is greater than the preset risk factor threshold, the fault feature data is screened out, including: If the fault risk coefficient is greater than the preset risk coefficient threshold, the fluctuation evaluation values of different fluctuation time nodes at multiple monitoring points in each monitoring indicator of the combiner box are analyzed, and the fluctuation time nodes and monitoring points whose fluctuation evaluation values are greater than the preset fluctuation evaluation value threshold are screened, and the corresponding fluctuation time nodes and monitoring points are set as feature nodes and feature points; Obtain the real-time monitoring data of the adjacent time nodes at the same monitoring point of the current feature node. If the real-time monitoring data at the adjacent time nodes are in the standard monitoring data interval, the current feature node is removed, and the real-time monitoring data at the remaining feature nodes are set as fault feature data; According to the degree of deviation between the real-time monitoring data at the adjacent time nodes of the remaining characteristic nodes and the standard detection data interval, the first coefficient of the fault characteristic coefficient is set; The second coefficient of the fault feature data is set according to the weight coefficient of the feature point corresponding to the remaining feature nodes, the fault feature data is sorted according to the first coefficient and the second coefficient, and the fault feature data is input into the fault impact model in the order of arrangement to obtain the fault impact coefficient.
7. The combiner box fault monitoring method according to claim 6, characterized in that: Build a failure impact model, including: Obtain historical operation data of associated devices corresponding to historical fault data of the combiner box; The historical operation data of the associated equipment is classified and judged according to the classification rules of the equipment operation status, and the historical operation data affected by the historical fault data under the same mapping node is determined, and the current corresponding historical fault data is used as the input data set, and the historical operation data of the same mapping node of the current historical fault data is used as the first output data set; Comparing the historical operating data in the first output data set with the corresponding preset standard operating data, obtaining the state coefficient of the historical operating data in the first output data set according to the comparison result, and using the state coefficient of the historical operating data in the first output data set as the second output data set; The input data set, the first output data set and the second output data set are used to generate a training set and a test set, and a neural network model is trained according to the training set to generate an initial fault impact model; The credibility of the initial fault impact model is obtained according to the test set. If the credibility is greater than the preset credibility threshold, the initial fault impact model is set as the fault impact model. If the credibility is less than the preset credibility threshold, the training set is iteratively trained to regenerate the initial fault impact model.
8. The combiner box fault monitoring method according to claim 7, characterized in that: Input the fault characteristic data into the fault impact model to obtain the fault impact coefficient, including: Input the fault characteristic data into the fault impact model to obtain the real-time operation data and real-time status coefficient of the current associated equipment; Generate a fault influence coefficient of current fault feature data according to the influence degree coefficient of the associated equipment and the real-time status coefficient; Wherein, d is the fault impact coefficient, q is the number of associated devices, yw is the real-time status coefficient of the w-th associated device, and hw is the influence degree coefficient of the w-th associated device.
9. The combiner box fault monitoring method according to claim 2, characterized in that: Before sending an alarm signal according to the fault influence coefficient, it also includes: Comprehensively analyze the impact coefficient and real-time status coefficient of associated equipment to obtain the fault urgency of each associated equipment; Sort the faults according to their urgency to obtain a fault sequence. According to the fault feature data corresponding to each fault urgency in the fault operation and maintenance sequence, filter out the fault type and fault location corresponding to the fault feature data. According to the fault type and fault location, filter out the corresponding historical operation and maintenance strategy. Map the historical operation and maintenance strategies to the fault urgency level one by one to obtain the operation and maintenance sequence; Send an alarm signal based on the fault impact coefficient, and perform operation and maintenance management according to the corresponding operation and maintenance sequence.
10. The combiner box fault monitoring method according to claim 9, characterized in that: Send alarm signals according to the fault impact coefficient, including: Presetting a first preset fault influence coefficient threshold and a second preset fault influence coefficient threshold; When the fault impact coefficient is less than the first preset fault impact coefficient threshold, a first-level alarm signal is sent; When the fault influence coefficient is within the first preset fault influence coefficient threshold and the second preset fault influence coefficient threshold, a secondary alarm signal is sent; When the fault influence coefficient is greater than the second preset fault influence coefficient threshold, a third-level alarm signal is sent.
Citation Information
Cited By
Digital processing equipment visual management system and method
CN119417456A
Helical gear speed reducer fault diagnosis method under intelligent monitoring
CN119827145A
Fault diagnosis method for helical gear reducer under intelligent monitoring
CN119827145B
Multi-parameter online monitoring and fault early warning system based on intelligent busbar
CN120891307A
