A method for isolating combiner box fault

Abstract

The present invention relates to the technical field of combiner box maintenance, and discloses a combiner box fault isolation method, including: constructing a combiner box monitoring network based on the acquired combiner box layout information of a photovoltaic power station; constructing a voltage monitoring model library based on the acquired historical operation data of the combiner boxes of the photovoltaic power station; setting monitoring time nodes, monitoring the combiner box monitoring network based on the voltage monitoring model library, and generating fault monitoring results of the combiner box monitoring network; generating a combiner box maintenance strategy based on the fault monitoring results of the combiner box monitoring network. The present invention realizes real-time monitoring and fault diagnosis of combiner boxes by constructing a combiner box monitoring network and a voltage monitoring model library, thereby improving maintenance efficiency and accuracy; at the same time, the present invention also takes into account the special environmental factors of water photovoltaic power stations, such as the influence of water level changes, water flow, etc. on the combiner boxes.

Description

Technical Field

[0001] The present invention relates to the technical field of combiner box maintenance, and in particular to a combiner box fault isolation method. Background Art

[0002] The combiner box is one of the key equipment in a photovoltaic power station. Its main function is to collect the DC power generated by multiple photovoltaic strings and safely transmit the power to the subsequent DC distribution cabinet or inverter through protection devices such as circuit breakers and lightning arresters. The performance of the combiner box directly affects the power generation efficiency and safety of the photovoltaic power station.

[0003] At present, there are many small ponds in large-scale ground photovoltaics. The junction box is installed on the pond. Due to the large number of ponds, it takes a long time for the boat to find the faulty junction box from the pond. In particular, there are many DC junction boxes under the box inverter and they are scattered. It is possible that half a day has passed by the time the faulty junction box is found. The power generation time of a box inverter is more than half of the day, which greatly reduces the power generation efficiency. At the same time, if the fault is not eliminated, it may develop into an accident under sufficient sunlight, which increases the risk. Secondly, it is difficult to achieve real-time monitoring by manual inspection, and it is difficult to find some sudden faults in time. In addition, the accuracy of manual inspection is also affected by the experience and skill level of the inspection personnel. Summary of the invention

[0004] The purpose of the present invention is to obtain the combiner box layout information of the photovoltaic power station, build a combiner box monitoring network, and monitor the combiner box monitoring network through the voltage monitoring model library to generate the location of the faulty combiner, effectively monitor and maintain the combiner box, and improve the operating efficiency and safety of the water photovoltaic power station.

[0005] In order to achieve the above object, the present invention provides a combiner box fault isolation method, comprising:

[0006] A combiner box monitoring network is constructed based on the acquired combiner box layout information of the photovoltaic power station;

[0007] A voltage monitoring model library is constructed based on the historical operation data of the combiner box of the photovoltaic power station;

[0008] Set monitoring time nodes, monitor the combiner box monitoring network based on the voltage monitoring model library, and generate fault monitoring results for the combiner box monitoring network;

[0009] Generate a combiner box maintenance instruction based on the fault monitoring result of the combiner box monitoring network.

[0010] In some embodiments of the present invention, the construction of a combiner box monitoring network includes:

[0011] Get the connection relationship between the box-type inverter and the combiner box in the ground photovoltaic power station, set the box-type inverter as the first-level network node and the combiner box as the second-level network node;

[0012] Based on the transmission line, the primary network node and the secondary network node are connected to generate a combiner box monitoring subnetwork, each of which includes a primary network node and multiple secondary network nodes;

[0013] Based on the number M of secondary network nodes in the combiner box monitoring subnetwork and the regional coverage area S of the combiner box monitoring subnetwork, a level evaluation model of the combiner box monitoring subnetwork is constructed to generate a multi-level combiner box monitoring subnetwork.

[0014] Set up a gateway node to connect all primary network nodes to generate a combiner box monitoring network.

[0015] In some embodiments of the present invention, the step of constructing a grade evaluation model for a combiner box monitoring subnetwork includes:

[0016] Connect all the secondary network nodes of the combiner box monitoring sub-network, and generate multiple triangular areas based on the connection between the current secondary network node and the adjacent secondary network node and the connection between the secondary network node and the primary network node;

[0017] Calculate the area St of each triangular area and sum them up to generate the regional coverage area S of the current combiner box monitoring subnetwork;

[0018] Based on the regional coverage area S of the current combiner box monitoring subnetwork, generating an evaluation value Ps of the regional coverage area of ​​the current combiner box monitoring subnetwork;

[0019] Count the number M of secondary network nodes of the current combiner box monitoring sub-network and generate a network node number evaluation value Pm;

[0020] Among them, the value range of the regional coverage area evaluation value Ps and the network node quantity evaluation value Pm is the same;

[0021] Constructing a grade evaluation model for the combiner box monitoring subnetwork and generating a grade evaluation value P for the combiner box monitoring subnetwork;

[0022] P = k 1 *Pm+k 2 *Ps;

[0023] Among them, k 1 is the weight of the evaluation value of the number of network nodes, k 2 is the weight of the regional coverage area evaluation value Ps;

[0024] The level evaluation value threshold is set based on the level evaluation value P to obtain a multi-level combiner box monitoring subnetwork.

[0025] In some embodiments of the present invention, the step of constructing a voltage monitoring model library includes:

[0026] Grouping the primary network nodes based on the number M of corresponding secondary network nodes;

[0027] Obtain the historical voltage data of each group of primary network nodes, and generate a voltage evaluation value v1 of the peak value of the corresponding voltage data, an evaluation value f1 of the occurrence frequency of the peak value, and an evaluation value t1 of the duration of the peak value;

[0028] And generate a state evaluation value E of the first-level network node;

[0029] E=α*v1+β*f1+γ*t1;

[0030] Among them, α is the weight of the voltage evaluation value v1, β is the weight of the peak occurrence frequency evaluation value f1, and γ is the weight of the peak duration evaluation value t1; the values ​​of α, β and γ vary with the grouping of the first-level network nodes;

[0031] Based on the status evaluation value E of the first-level network node, a voltage monitoring model of the current first-level network node group is constructed;

[0032] The voltage monitoring models of all groups are obtained to generate a voltage monitoring model library.

[0033] In some embodiments of the present invention, the monitoring of the combiner box monitoring network based on the voltage monitoring model library and generating the fault monitoring result of the combiner box monitoring network include:

[0034] Set monitoring time nodes based on the historical operation data of the PV power station, and set the interval of monitoring time nodes based on the level of the combiner box monitoring subnetwork;

[0035] Selecting a corresponding voltage monitoring model from a voltage monitoring model library based on the acquired primary network node grouping of the current combiner box monitoring subnetwork;

[0036] Based on the voltage monitoring model of the current primary network node grouping, determine whether there is an abnormality in the current combiner box monitoring sub-network;

[0037] If there is an abnormality in the current combiner box monitoring subnetwork, the secondary network nodes of the current combiner box monitoring subnetwork are detected respectively to generate the location of the faulty combiner.

[0038] In some embodiments of the present invention, the step of determining whether the current combiner box monitoring sub-network is abnormal includes:

[0039] Obtain voltage data of the current first-level network node, and generate a state evaluation value E of the current first-level network node through a corresponding voltage monitoring model;

[0040] The state evaluation value E and the preset state evaluation value (th 1 , th 2 ) for comparison;

[0041] Among them, 1 is the first preset state evaluation value, th 2 is a second preset state evaluation value;

[0042] If E <th 1 or E>th 2 , it is determined that the current combiner box monitoring subnetwork is abnormal, and the current status evaluation value E and voltage data related information are recorded;

[0043] If 1 <E≤th 2 , it is determined that the current combiner box monitoring subnetwork is normal.

[0044] In some embodiments of the present invention, the generating of the position of the faulty current combiner includes:

[0045] Generate an environmental impact model of secondary network nodes based on historical data of photovoltaic power plants;

[0046] Obtain the current photovoltaic power station environmental prediction data, and generate the predicted power curve L of the current secondary network node based on the environmental impact model;

[0047] Obtain the actual power curve LT of the current secondary network node through actual monitoring;

[0048] Compare the predicted power curve L with the actual power curve LT to generate a power deviation value b of the secondary network node;

[0049] By comparing the power deviation value b with the preset deviation value, it is determined whether the current secondary network node is faulty, and the location of the faulty combiner in the combiner box monitoring sub-network is determined.

[0050] In some embodiments of the present invention, the generating of the predicted power curve L of the current secondary network node includes:

[0051] The environmental prediction data of photovoltaic power stations include: light intensity data and ambient temperature data;

[0052] Based on the environmental prediction data of the photovoltaic power station and the conversion efficiency data of the photovoltaic panels, the power prediction value of the secondary network node at the current time node is calculated through the environmental impact model to generate the predicted power curve L, L = [L 1 ,L 2 …L i …L n ];

[0053] Among them, L irepresents the power value at the i-th moment; n represents the total number of moments at the current time node.

[0054] In some embodiments of the present invention, the generating of the power deviation value b of the secondary network node includes:

[0055] Get the actual power curve LT of the current secondary network node, LT = [LT 1 ,LT 2 …LT i …LT n ];

[0056] Based on the predicted power curve L and the actual power curve LT, the power deviation value b of the secondary network node is generated:

[0057]

[0058] In some embodiments of the present invention, the generating of the combiner box maintenance instruction includes:

[0059] A DC circuit breaker is set on the transmission cable between the box inverter and the combiner box. When the position of the faulty combiner of the combiner box monitoring subnetwork is determined, the DC circuit breaker is controlled to disconnect the connection between the box inverter and the faulty combiner box, and a maintenance instruction for the current combiner box is generated.

[0060] The present invention provides a combiner box fault isolation method, which has the following beneficial effects compared with the prior art:

[0061] By constructing a combiner box monitoring network, the structure of the combiner box monitoring network is clear, which is convenient for subsequent management and data collection; it helps to adopt different monitoring and maintenance strategies for sub-networks of different levels, and improve the rationality of resource allocation and monitoring efficiency.

[0062] By constructing a grade evaluation model for the combiner box monitoring subnetwork, the structure and scale factors of the combiner box monitoring subnetwork are comprehensively considered, and its grade is determined more accurately, thereby more effectively allocating resources and monitoring faults.

[0063] By constructing the voltage evaluation value and then generating the state evaluation value, this multi-dimensional evaluation method can more comprehensively reflect the voltage characteristics of the primary network nodes;

[0064] By constructing a voltage monitoring model through different groups and forming a model library, a diverse model basis is provided for the subsequent accurate voltage monitoring of the combiner box monitoring sub-networks in different groups, thereby improving the accuracy and adaptability of voltage monitoring.

[0065] By setting the monitoring time node interval according to the level of the combiner box monitoring subnetwork, monitoring resources can be reasonably allocated according to the importance or failure risk of different levels of networks, thereby improving monitoring efficiency.

[0066] By constructing an environmental impact model for the secondary network nodes, the impact of environmental factors on the combiner box is taken into account, making fault troubleshooting more comprehensive, effectively locating the faulty combiner, and improving the accuracy and efficiency of fault troubleshooting.

[0067] By installing a DC circuit breaker between the integrated inverter and the combiner box, the scope of the fault can be prevented from expanding, thus ensuring the safe and stable operation of the entire system. This provides a better and more flexible power generation mode for centralized inverter power generation, and also provides a faster way to handle faults, thus avoiding the risk of small faults developing into major accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 The present invention provides a flow chart of a method for isolating a combiner box fault. DETAILED DESCRIPTION

[0069] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0070] In the description of the present invention, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0071] 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 the present invention, unless otherwise specified, "plurality" means two or more.

[0072] In the description of the present invention, 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 the present invention can be understood according to specific circumstances.

[0073] Embodiment 1:

[0074] A combiner box fault isolation method is preferably provided in an embodiment of the present invention, such as Figure 1 As shown, including:

[0075] A combiner box monitoring network is constructed based on the acquired combiner box layout information of the photovoltaic power station;

[0076] A voltage monitoring model library is constructed based on the historical operation data of the combiner box of the photovoltaic power station;

[0077] Set monitoring time nodes, monitor the combiner box monitoring network based on the voltage monitoring model library, and generate fault monitoring results for the combiner box monitoring network;

[0078] Generate a combiner box maintenance instruction based on the fault monitoring result of the combiner box monitoring network.

[0079] Embodiment 2:

[0080] The construction of the combiner box monitoring network includes:

[0081] Get the connection relationship between the box-type inverter and the combiner box in the ground photovoltaic power station, set the box-type inverter as the first-level network node and the combiner box as the second-level network node;

[0082] Based on the transmission line, the primary network node and the secondary network node are connected to generate a combiner box monitoring subnetwork, each of which includes a primary network node and multiple secondary network nodes;

[0083] Based on the number M of secondary network nodes in the combiner box monitoring subnetwork and the regional coverage area S of the combiner box monitoring subnetwork, a level evaluation model of the combiner box monitoring subnetwork is constructed to generate a multi-level combiner box monitoring subnetwork.

[0084] Set up a gateway node to connect all primary network nodes to generate a combiner box monitoring network.

[0085] In this embodiment, for the ground photovoltaic power station, the connection relationship between the box inverter and the combiner box is carefully obtained; the box inverter is clearly set as a primary network node, and the combiner box is set as a secondary network node; this setting is based on their different functions and positions in the power transmission and conversion process of the photovoltaic power station.

[0086] The inverter boxes and combiner boxes are numbered or marked to facilitate accurate construction of network relationships later.

[0087] According to the actual physical connection situation, use the transmission line to connect the primary network node (box inverter integrated machine) and the secondary network node (combiner box). During the connection process, ensure that the connection direction of the transmission line is correct and record the length, specifications and other relevant parameters of each transmission line.

[0088] For each connection relationship, a corresponding connection record is created, including the numbers of the two connected nodes, the identification of the transmission line and other information, so as to generate a combiner box monitoring subnetwork. Each combiner box monitoring subnetwork contains one primary network node and multiple secondary network nodes according to the definition.

[0089] Acquisition of basic data for building a grade evaluation model

[0090] For each combiner box monitoring sub-network, the number of secondary network nodes M is counted. This number reflects the size of the sub-network.

[0091] Calculate the regional coverage area S of the combiner box monitoring sub-network. This may require using a suitable geometric method (such as polygon area calculation method) to calculate based on the geographical coordinates (if any) of each node in the sub-network.

[0092] Embodiment 3:

[0093] The construction of the grade evaluation model of the combiner box monitoring sub-network includes:

[0094] Connect all the secondary network nodes of the combiner box monitoring sub-network, and generate multiple triangular areas based on the connection between the current secondary network node and the adjacent secondary network node and the connection between the secondary network node and the primary network node;

[0095] Calculate the area St of each triangular area and sum them up to generate the regional coverage area S of the current combiner box monitoring subnetwork;

[0096] Based on the regional coverage area S of the current combiner box monitoring subnetwork, generating an evaluation value Ps of the regional coverage area of ​​the current combiner box monitoring subnetwork;

[0097] Count the number M of secondary network nodes of the current combiner box monitoring sub-network and generate a network node number evaluation value Pm;

[0098] Among them, the value range of the regional coverage area evaluation value Ps and the network node quantity evaluation value Pm is the same;

[0099] Constructing a grade evaluation model for the combiner box monitoring subnetwork and generating a grade evaluation value P for the combiner box monitoring subnetwork;

[0100] P = k 1 *Pm+k 2 *Ps;

[0101] Among them, k 1 is the weight of the evaluation value of the number of network nodes, k 2 is the weight of the regional coverage area evaluation value Ps;

[0102] The level evaluation value threshold is set based on the level evaluation value P to obtain a multi-level combiner box monitoring subnetwork.

[0103] In this embodiment, for all secondary network nodes in the combiner box monitoring subnetwork, operations are performed according to their connection relationships. When connecting the current secondary network node with the adjacent secondary network node and the secondary network node with the primary network node, multiple triangular areas are formed. When constructing these triangular areas, the accuracy of the connection must be ensured to truly reflect the topological structure of the network.

[0104] For example, a secondary network node is taken as a vertex, and its adjacent secondary network node and the primary network node connected thereto are the other two vertices, forming a triangular area.

[0105] Collect the regional coverage area S data of all combiner box monitoring sub-networks and generate the regional coverage area evaluation value Ps

[0106] Then, the calculation method of the regional coverage area evaluation value Ps is determined according to the distribution of these data. For example, the value of S can be mapped to a specific value range (such as [0,1]) through a normalization method. If S is the minimum value in all combiner box monitoring subnetworks, Ps = 0; if S is the maximum value, Ps = 1; for intermediate values, the value of Ps is calculated according to a linear scale.

[0107] Count the number of secondary network nodes M of the current combiner box monitoring sub-network.

[0108] Similarly, normalization or other appropriate methods are used to convert the value of M into the network node quantity evaluation value Pm, so that its value range is the same as the regional coverage area evaluation value Ps (such as [0,1]). For example, if M is the minimum value of the number of secondary network nodes in all combiner box monitoring subnetworks, Pm = 0; if M is the maximum value, Pm = 1; intermediate values ​​are calculated proportionally.

[0109] Embodiment 4:

[0110] The construction of the voltage monitoring model library includes:

[0111] Grouping the primary network nodes based on the number M of corresponding secondary network nodes;

[0112] Obtain the historical voltage data of each group of primary network nodes, and generate a voltage evaluation value v1 of the peak value of the corresponding voltage data, an evaluation value f1 of the occurrence frequency of the peak value, and an evaluation value t1 of the duration of the peak value;

[0113] And generate a state evaluation value E of the first-level network node;

[0114] E=α*v1+β*f1+γ*t1;

[0115] Among them, α is the weight of the voltage evaluation value v1, β is the weight of the peak occurrence frequency evaluation value f1, and γ is the weight of the peak duration evaluation value t1; the values ​​of α, β and γ vary with the grouping of the first-level network nodes;

[0116] Based on the status evaluation value E of the first-level network node, a voltage monitoring model of the current first-level network node group is constructed;

[0117] The voltage monitoring models of all groups are obtained to generate a voltage monitoring model library.

[0118] In this embodiment, the information of each group is recorded in detail, including the number of the primary network nodes in the group, the connection mode, the corresponding secondary network node quantity range, etc., to facilitate subsequent data processing and model construction.

[0119] Obtain historical voltage data of each group of primary network nodes from the database of the photovoltaic power station. Ensure the integrity of the data, covering data in different seasons, different weather conditions and different operating periods.

[0120] Voltage evaluation value v1 generation

[0121] Analyze each group of historical voltage data to determine the peak value of the voltage data. This can be done through data statistics methods, such as finding the maximum value within a period of time as the peak value.

[0122] The voltage evaluation value v1 is generated by comparing the peak value with the normal operating voltage range. For example, if the peak value is within the normal operating voltage range, v1 can be set to 1; if the peak value is slightly higher than the normal range, v1 may be set to 0.8; if the peak value is much higher than the normal range, v1 may be set to 0.5 or lower. The specific value is determined according to the actual voltage deviation.

[0123] Based on the calculated state evaluation value E of the first-level network node, a voltage monitoring model of the current first-level network node group is constructed.

[0124] The input of the model can include real-time voltage data, voltage fluctuation range, etc., and the output is the status evaluation result (such as normal or abnormal) of the current first-level network node group. Different thresholds and judgment logic can be set within the model according to the calculation method of E and the characteristics of the group. According to the above steps, the voltage monitoring model is constructed for all groups.

[0125] Ensure that each model can accurately reflect the electrical characteristics and operating status of the first-level network nodes of the corresponding group.

[0126] The voltage monitoring models of all groups are integrated together to generate a voltage monitoring model library. The model library is numbered and indexed to facilitate the subsequent rapid call of the corresponding voltage monitoring model according to different primary network node groups.

[0127] Embodiment 5:

[0128] The monitoring of the combiner box monitoring network based on the voltage monitoring model library and generating the fault monitoring result of the combiner box monitoring network include:

[0129] Set monitoring time nodes based on the historical operation data of the PV power station, and set the interval of monitoring time nodes based on the level of the combiner box monitoring subnetwork;

[0130] Selecting a corresponding voltage monitoring model from a voltage monitoring model library based on the acquired primary network node grouping of the current combiner box monitoring subnetwork;

[0131] Based on the voltage monitoring model of the current primary network node grouping, determine whether there is an abnormality in the current combiner box monitoring sub-network;

[0132] If there is an abnormality in the current combiner box monitoring subnetwork, the secondary network nodes of the current combiner box monitoring subnetwork are detected respectively to generate the location of the faulty combiner.

[0133] In this embodiment, the historical operation data of the photovoltaic power station is deeply analyzed, including the operation data of the combiner box monitoring network in different seasons and weather conditions. For example, the power output, voltage fluctuation and other conditions in the historical data are checked.

[0134] Determine monitoring time nodes based on the frequency and pattern of faults in historical data. If historical data shows that faults are more likely to occur in the early morning and evening, more intensive monitoring time nodes can be set during these periods.

[0135] The combiner box monitoring subnetwork is divided into levels, for example, high, medium and low levels according to factors such as the number of photovoltaic panels connected to it and the power transmission capacity.

[0136] For high-level combiner box monitoring subnetworks, due to their high importance in the entire PV power station and the wide range of fault impact, a shorter monitoring time node interval is set, such as monitoring every 10 minutes. For medium-level combiner box monitoring subnetworks, monitoring can be set every 15 minutes, while low-level combiner box monitoring subnetworks can be monitored every 20 minutes.

[0137] Embodiment 6:

[0138] The process of judging whether the current combiner box monitoring sub-network is abnormal includes:

[0139] Obtain voltage data of the current first-level network node, and generate a state evaluation value E of the current first-level network node through a corresponding voltage monitoring model;

[0140] Compare the state evaluation value E with the preset state evaluation values (th 1 , th 2 );

[0141] Among them, th 1 is the first preset state evaluation value, and th 2 is the second preset state evaluation value;

[0142] If E < th 1 or E > th 2 , it is determined that there is an abnormality in the current sub-network for monitoring the busbar trunking unit, and the current state evaluation value E and the relevant information of the voltage data are recorded;

[0143] If th 1 < E ≤ th 2 , it is determined that the current sub-network for monitoring the busbar trunking unit is normal.

[0144] In this embodiment, the calculated state evaluation value E is strictly compared with the preset state evaluation values th1 and th2. During the comparison process, the consistency of the data types should be ensured. For example, if E is a normalized value, th1 and th2 should also be values on the same normalization scale.

[0145] Abnormality determination and information recording:

[0146] If E < th1, it indicates that the state evaluation value of the current first-level network node is lower than the normal lower limit, and there may be abnormal situations such as undervoltage. If E > th2, there may be abnormal situations such as overvoltage. In both cases, it is determined that there is an abnormality in the current sub-network for monitoring the busbar trunking unit. At the same time, the current state evaluation value E and the information related to the voltage data, such as the acquisition time and the corresponding first-level network node number, are recorded in detail. These recorded information will be helpful for subsequent fault analysis and maintenance work.

[0147] If th1 < E ≤ th2, it means that the state evaluation value of the current first-level network node is within the normal range, and thus it is determined that the current sub-network for monitoring the busbar trunking unit is normal. In this case, the current state evaluation value E and the information related to the voltage data can also be selectively recorded for long-term statistical analysis of the network operation status.

[0148] Embodiment 7:

[0149] When generating the location of the faulty current collector, it includes:

[0150] Generate an environmental impact model for the secondary network nodes based on the historical data of the photovoltaic power station;

[0151] Obtain the current environmental prediction data of the photovoltaic power station, and generate a predicted power curve L for the current secondary network nodes based on the environmental impact model;

[0152] Obtain the actual power curve LT of the current secondary network node through actual monitoring;

[0153] Compare the predicted power curve L with the actual power curve LT to generate a power deviation value b of the secondary network node;

[0154] By comparing the power deviation value b with the preset deviation value, it is determined whether the current secondary network node is faulty, and the location of the faulty combiner in the combiner box monitoring sub-network is determined.

[0155] In this embodiment, the current photovoltaic power station environment prediction data is obtained, including light intensity, ambient temperature, wind speed, humidity, etc. These data are preprocessed, such as data format conversion, unit unification, etc. For example, the light intensity data is uniformly converted into the standard unit of watts per square meter.

[0156] Conduct a quality assessment of environmental forecast data. If the data is missing or incomplete, try to use interpolation methods (such as linear interpolation or spline interpolation) to supplement it. At the same time, make a preliminary judgment on the accuracy of the data. For example, based on local weather forecast experience, judge whether the predicted values ​​of light intensity and temperature are reasonable.

[0157] The processed current photovoltaic power station environmental prediction data is substituted into the environmental impact model to calculate the power prediction values ​​of the current secondary network nodes at different times.

[0158] The actual power is monitored by the power monitoring equipment installed at the secondary network node. Ensure that the accuracy of the monitoring equipment meets the requirements, for example, the power measurement accuracy is within ±1%.

[0159] The actual power data is collected at fixed time intervals (eg, data is collected every 5 minutes) to form an original actual power data sequence.

[0160] Clean the collected actual power data. Remove abnormal data points caused by monitoring equipment failure or interference. Data smoothing techniques such as moving average (for example, taking the average of three adjacent data points to replace the middle data point) can be used to process small fluctuations in the data, thereby obtaining a relatively smooth actual power curve LT.

[0161] Embodiment 8:

[0162] The generating of the predicted power curve L of the current secondary network node includes:

[0163] The environmental prediction data of photovoltaic power stations include: light intensity data and ambient temperature data;

[0164] Based on the environmental prediction data of the photovoltaic power station and the conversion efficiency data of the photovoltaic panels, the power prediction value of the secondary network node at the current time node is calculated through the environmental impact model to generate the predicted power curve L, L = [L 1 ,L 2 …L i …L n ];

[0165] Among them, L i represents the power value at the i-th moment; n represents the total number of moments at the current time node.

[0166] In this embodiment, the light intensity data and ambient temperature data in the photovoltaic power station environment prediction data are obtained. For the light intensity data, ensure that its measurement unit is unified (for example, watts / square meter), and the time resolution of the data meets the requirements (such as one data point every 15 minutes).

[0167] Check the rationality of the light intensity data. Set reasonable upper and lower limits based on local meteorological characteristics and the historical light intensity data range. If the data exceeds this range, further verify the data source or use a data correction algorithm, for example, compare and correct the data with the surrounding meteorological stations.

[0168] Similarly, the ambient temperature data is processed to ensure the accuracy of the temperature data (such as accurate to 0.1 degrees Celsius) and verify its rationality. Considering the operating environment of the photovoltaic power station, abnormal high or low temperature data points are excluded, and verification can be carried out by comparing with historical temperature data of the same period.

[0169] Determine the conversion efficiency data of the photovoltaic panel. First, obtain the nominal conversion efficiency value based on the photovoltaic panel model used in the photovoltaic power station. Then, consider the actual influencing factors such as aging, dirt, and shading of the photovoltaic panel to correct the conversion efficiency. For example, through regular photovoltaic panel performance test data, establish a conversion efficiency attenuation model, and calculate the actual conversion efficiency based on the service life and maintenance of the photovoltaic panel.

[0170] For the i-th moment, the power prediction value L is calculated based on the environmental impact model. i Calculation.

[0171] In the environmental impact model, the theoretical power generation of the photovoltaic panel is first calculated based on the light intensity data. Assuming that the light intensity is I (watts / square meter), the area of ​​the photovoltaic panel is A (square meter), and the ideal conversion efficiency of the photovoltaic panel is η 0 (without considering actual influencing factors), the theoretical power generation power P 0 =I*A*η 0 .

[0172] Then, the effect of ambient temperature on conversion efficiency is considered. Based on the actual conversion efficiency η of the photovoltaic panel (taking into account factors such as aging, dirt, and shielding) and the ambient temperature T (degrees Celsius), a function η(T) related to conversion efficiency and temperature is established (for example, a linear function or a quadratic function obtained by fitting experimental data), and the theoretical power generation is corrected to obtain the corrected power generation P i '=P 0 *η(T).

[0173] Finally, consider other loss factors in the secondary network nodes, such as line loss, junction box loss, etc. Assume that the line loss coefficient is k 3 (0 <k 3 <1), the combiner box loss coefficient is k 4 (0 <k 4 <1), then the power prediction value at the i-th moment Li=Pi'*k 3 *k 4 .

[0174] According to the above method, the power prediction value L at each time i (1≤i≤n) is calculated in turn. i , thereby generating a predicted power curve L = [L 1 ,L 2 …L i …L n ]. During the calculation process, ensure that each L i The calculation process is accurate and the intermediate parameters in each calculation step (such as theoretical power generation P 0 , Corrected power generation P i ', etc.) for use in subsequent analysis and optimization.

[0175] Embodiment 9:

[0176] The generating of the power deviation value b of the secondary network node includes:

[0177] Get the actual power curve LT of the current secondary network node, LT = [LT 1 ,LT 2 …LT i …LT n ];

[0178] Based on the predicted power curve L and the actual power curve LT, the power deviation value b of the secondary network node is generated:

[0179]

[0180] In this embodiment, the actual power curve LT of the current secondary network node is accurately obtained from the monitoring device. In the process of obtaining data, the data is checked for integrity to ensure that the data sequence [LT 1 ,LT 2 …LT i …LT n ] There are no missing values. If there are missing values, they can be supplemented by interpolation. For example, for a power curve with a linear trend, linear interpolation can be used; for a power curve with periodic characteristics, interpolation can be performed based on historical data for the same period.

[0181] For each data point LT in the actual power curve LT obtained i Perform rationality check. Set reasonable upper and lower limits based on the rated power range of the power station and the historical power fluctuation range. If a data point exceeds the range, it is considered an abnormal data point. For abnormal data points, data smoothing can be used, such as using the sliding average method (for example, taking the average of 3-5 adjacent data points to replace the abnormal data point) for correction.

[0182] Based on the predicted power curve L = [L 1 ,L 2 …L i …L n Before calculating the power deviation value, ensure that the predicted power curve is generated based on accurate PV power station environment prediction data (including light intensity data and ambient temperature data, etc.). i Conduct accuracy assessments and set prediction accuracy requirements for each data point based on historical prediction accuracy data. For example, prediction accuracy within ±5% is considered acceptable. If the prediction accuracy of a data point exceeds this range, the environmental impact model can be re-evaluated or the prediction algorithm can be adjusted.

[0183] Embodiment 10:

[0184] The generation of the combiner box maintenance strategy includes:

[0185] A DC circuit breaker is set on the transmission cable between the box inverter and the combiner box. When the position of the faulty combiner of the combiner box monitoring subnetwork is determined, the DC circuit breaker is controlled to disconnect the connection between the box inverter and the faulty combiner box, and a maintenance instruction for the current combiner box is generated.

[0186] In this embodiment, a DC circuit breaker of a suitable type is selected according to the rated voltage, rated current, and maximum short-circuit current of the transmission line between the box-type inverter and the combiner box. Ensure that the rated voltage of the DC circuit breaker is not lower than the voltage of the transmission line, the rated current can meet the current requirements under normal operation and overload conditions, and the maximum short-circuit current tolerance is sufficient to cope with possible short-circuit faults.

[0187] After determining the location of the faulty combiner in the combiner box monitoring subnetwork, the control center first sends a control command to the DC circuit breaker. During the command transmission process, encryption and verification mechanisms are used to ensure the accuracy and security of command transmission. For example, the CRC (cyclic redundancy check) algorithm is used to verify the command.

[0188] After receiving the control command, the DC circuit breaker quickly disconnects the connection between the inverter and the faulty combiner box within 100 milliseconds (the response time is determined according to the specific circuit breaker performance). At the same time, the DC circuit breaker feeds back its own operating status (such as successful disconnection) to the control center so that the control center can confirm the validity of the operation.

[0189] The generated maintenance instructions should contain detailed information about the faulty combiner. For example, the maintenance instructions should clearly state the number of the faulty combiner, its location (accurate to a specific area or coordinates, if any), and the type of fault (such as short circuit fault, overcurrent fault, etc., determined based on previous fault monitoring results).

[0190] The maintenance instructions should also include operational suggestions for maintenance personnel. For example, if it is a short circuit fault, it is recommended that maintenance personnel first check whether the fuse inside the combiner is blown; if it is an overcurrent fault, it is recommended that maintenance personnel check whether there are overload devices in the input and output circuits of the combiner, etc.

[0191] In addition, the maintenance instructions should be accompanied by relevant operating data of the current combiner box monitoring subnetwork (such as the voltage and current values ​​when the fault occurs) so that maintenance personnel can fully understand the operating conditions when the fault occurs, which helps to quickly locate and repair the fault.

[0192] Finally, it should be noted that it is obvious that those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technology, the present invention is also intended to include these modifications and variations.

[0193] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention.

Claims

1. A method for isolating a combiner box fault, characterized in that: include: A combiner box monitoring network is constructed based on the acquired combiner box layout information of the photovoltaic power station; A voltage monitoring model library is constructed based on the historical operation data of the combiner box of the photovoltaic power station; Monitor the combiner box monitoring network through the voltage monitoring model library and generate fault monitoring results of the combiner box monitoring network; Generate a combiner box maintenance strategy based on the fault monitoring results of the combiner box monitoring network.

2. The combiner box fault isolation method according to claim 1, characterized in that: The construction of the combiner box monitoring network includes: Get the connection relationship between the box-type inverter and the combiner box in the ground photovoltaic power station, set the box-type inverter as the first-level network node and the combiner box as the second-level network node; Based on the transmission line, the primary network node and the secondary network node are connected to generate a combiner box monitoring subnetwork, each of which includes a primary network node and multiple secondary network nodes; Based on the number M of secondary network nodes in the combiner box monitoring subnetwork and the regional coverage area S of the combiner box monitoring subnetwork, a level evaluation model of the combiner box monitoring subnetwork is constructed to generate a multi-level combiner box monitoring subnetwork. Set up a gateway node to connect all primary network nodes to generate a combiner box monitoring network.

3. The combiner box fault isolation method according to claim 2, characterized in that: The construction of the grade evaluation model of the combiner box monitoring sub-network includes: Connect all the secondary network nodes of the combiner box monitoring sub-network, and generate multiple triangular areas based on the connection between the current secondary network node and the adjacent secondary network node and the connection between the secondary network node and the primary network node; Calculate the area St of each triangular area and sum them up to generate the regional coverage area S of the current combiner box monitoring subnetwork; Based on the regional coverage area S of the current combiner box monitoring subnetwork, generating an evaluation value Ps of the regional coverage area of ​​the current combiner box monitoring subnetwork; Count the number M of secondary network nodes of the current combiner box monitoring sub-network and generate a network node number evaluation value Pm; Among them, the value range of the regional coverage area evaluation value Ps and the network node quantity evaluation value Pm is the same; Constructing a grade evaluation model for the combiner box monitoring subnetwork and generating a grade evaluation value P for the combiner box monitoring subnetwork; P = k1*Pm+k2*Ps; Among them, k1 is the weight of the evaluation value of the number of network nodes, and k2 is the weight of the evaluation value Ps of the regional coverage area; The level evaluation value threshold is set based on the level evaluation value P to obtain a multi-level combiner box monitoring subnetwork.

4. The combiner box fault isolation method according to claim 3, characterized in that: The construction of the voltage monitoring model library includes: Grouping the primary network nodes based on the number M of corresponding secondary network nodes; Obtain the historical voltage data of each group of primary network nodes, and generate a voltage evaluation value v1 of the peak value of the corresponding voltage data, an evaluation value f1 of the occurrence frequency of the peak value, and an evaluation value t1 of the duration of the peak value; And generate a state evaluation value E of the first-level network node; E=α*v1+β*f1+γ*t1; Among them, α is the weight of the voltage evaluation value v1, β is the weight of the peak occurrence frequency evaluation value f1, and γ is the weight of the peak duration evaluation value t1; the values ​​of α, β and γ vary with the grouping of the first-level network nodes; Based on the status evaluation value E of the first-level network node, a voltage monitoring model of the current first-level network node group is constructed; Obtain the voltage monitoring models of all groups to generate a voltage monitoring model library.

5. The combiner box fault isolation method according to claim 4, characterized in that: When generating the fault monitoring result of the busbar box monitoring network, it includes: Set the monitoring time nodes based on the historical operation data of the photovoltaic power station, and set the interval of the monitoring time nodes based on the level of the busbar box monitoring subnetworks; Select the corresponding voltage monitoring model from the voltage monitoring model library based on the first-level network node grouping of the currently obtained busbar box monitoring subnetworks; Based on the voltage monitoring model of the current first-level network node grouping, determine whether there is an abnormality in the currently monitored busbar box subnetwork; If there is an abnormality in the currently monitored busbar box subnetwork, detect the second-level network nodes of the currently monitored busbar box subnetwork respectively to generate the location of the faulty busbar collector.

6. The combiner box fault isolation method according to claim 5, characterized in that: When determining whether there is an abnormality in the currently monitored busbar box subnetwork, it includes: Obtain the voltage data of the current first-level network node, and generate the state evaluation value E of the current first-level network node through the corresponding voltage monitoring model; Compare the state evaluation value E with the preset state evaluation values (th1, th2); Among them, th1 is the first preset state evaluation value, and th2 is the second preset state evaluation value; If E < th1 or E > th2, it is determined that there is an abnormality in the currently monitored busbar box subnetwork, and record the current state evaluation value E and the relevant information of the voltage data; If th1 < E ≤ th2, it is determined that the currently monitored busbar box subnetwork is normal.

7. The combiner box fault isolation method according to claim 6, characterized in that: When generating the location of the faulty busbar collector, it includes: Generate an environmental impact model for the second-level network nodes based on the historical data of the photovoltaic power station; Obtain the current environmental prediction data of the photovoltaic power station, and generate the predicted power curve L of the current second-level network node based on the environmental impact model; Obtain the actual power curve LT of the current second-level network node through actual monitoring; Compare the predicted power curve L with the actual power curve LT to generate the power deviation value b of the second-level network node; By comparing the power deviation value b with the preset deviation value, determine whether the current second-level network node has a fault, and determine the location of the faulty busbar collector in the busbar box monitoring subnetwork.

8. The combiner box fault isolation method according to claim 7, characterized in that: When generating the predicted power curve L of the current second-level network node, it includes: The environmental prediction data of the photovoltaic power station includes: light intensity data and environmental temperature data; Based on the environmental prediction data of the photovoltaic power station and the conversion efficiency data of the photovoltaic panels, the power prediction value of the secondary network node at the current time node is calculated through the environmental impact model to generate the predicted power curve L, L = [L1, L2…L i …L n ]; Among them, L i represents the power value at the i-th moment; n represents the total number of moments at the current time node.

9. The combiner box fault isolation method according to claim 7, characterized in that: When generating the power deviation value b of the second-level network node, it includes: Get the actual power curve LT of the current secondary network node, LT = [LT1, LT2…LT i …LT n ]; Generate the power deviation value b of the second-level network node based on the predicted power curve L and the actual power curve LT:

10. The combiner box fault isolation method according to claim 9, characterized in that: When generating the busbar box maintenance strategy, it includes: Set a DC circuit breaker on the transmission cable between the box inverter integrated machine and the busbar box. When determining the location of the faulty busbar collector in the busbar box monitoring subnetwork, control the DC circuit breaker to disconnect the connection between the box inverter integrated machine and the faulty busbar box, and generate a maintenance instruction for the current busbar box.

Images