A method and system for verifying and correcting distribution network line loss data
By establishing a data verification training tree and repairing abnormal data, the problem of poor data quality in line loss calculation in distribution network is solved, efficient line loss data verification and correction is achieved, and calculation accuracy and data synchronization are improved.
Patent Information
- Application Number
- CN201811240705.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2018-10-23
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2038-10-23
AI Technical Summary
The components and lines involved in the calculation of power distribution network line loss are complex and the data quality is poor, which leads to the inability to accurately calculate the power loss and the inability to accurately estimate the high-loss areas and lines, and the line loss management work is in a blind spot.
By establishing a data verification training tree, obtaining attribute data of the line loss data to be measured, traversing the data verification training tree, calculating the average height, filtering and repairing abnormal data, ensuring that the data quality meets the calculation accuracy of the actual project.
It improves the efficiency of line loss data verification, solves the problem of poor data quality during line loss calculation in distribution network, realizes the synchronization of data upload, and meets the calculation accuracy of actual projects.
Smart Images

Figure CN109767349B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of distribution network line loss data verification and correction, and in particular to a distribution network line loss data verification and correction method and system. Background Art
[0002] The distribution network is at the end of the power system and is directly connected to the user. It has distinct characteristics such as wide geographical distribution, large grid scale, many types of equipment, diverse network connections, and changeable operating modes. The power loss of the distribution network, especially the management line loss, has always been high. In particular, the line loss problem of public lines and low-voltage users in the substation area is more significant, resulting in a considerable proportion of the distribution network line loss in the entire grid line loss (about 70%). The main reasons for the high line loss of the distribution network are as follows: (1) Unreasonable planning and construction: the power supply radius of the distribution line is too large, the conductor cross-section is small, the transformer capacity is not matched, and the equipment selection is unscientific; the on-site reactive power source configuration is insufficient; the three-phase load distribution of the substation low-voltage line is uneven. (2) Unlean operation mode: the distribution network control strategy has not yet considered the line loss factor; the maintenance plan arrangement is not matched with the operation mode; the reactive loss is large or the capacitor commissioning rate is low. The load rate is high and the equipment is not in the economic operation zone; the reactive power flow adjustment is not in place and the voltage level is low. (3) The stock of high-loss equipment is large: the mismatch between the distribution transformer capacity and the dynamic changes in the load results in excessive transformer losses; the uneven quality of the main equipment in the distribution network leads to excessive line losses; the installation process is rough, the equipment is old, and the joints are oxidized. (4) There are illegal electricity use phenomena: illegal electricity use by power customers occurs from time to time, and the means and methods are quite different, which affects the lean management of line losses. Therefore, the distribution characteristics and change patterns of line losses are very complex, and in-depth theoretical research on the mechanism is needed. According to the differences in the causes of line losses, through targeted construction and transformation, operation and maintenance, equipment investment and optimized operation methods, the energy saving and loss reduction level of the distribution network can be improved.
[0003] However, the distribution network has many nodes, many branches, and many components, and most components do not have the conditions for measuring and recording operating parameters. Therefore, it is difficult to accurately calculate the power loss of the distribution network. The main problem is that the components and lines involved in the distribution network line loss calculation are relatively complex, and the data quality is poor, which cannot meet the calculation accuracy of the actual project. The data upload is not synchronized, and the results have synchronization problems. Therefore, only line loss estimates can be obtained, and it is impossible to accurately estimate high-loss areas and lines, and line loss management work falls into a blind spot. Summary of the Invention
[0004] In order to solve the problem in the prior art that it is impossible to accurately obtain distribution network line loss data, the present invention provides a distribution network line loss data verification and correction method and system.
[0005] The technical solution provided by the present invention is: a method for correcting line loss data of a distribution network, comprising:
[0006] Obtaining attribute data of the line loss data to be measured;
[0007] Based on the attribute data of the line loss data to be measured, a pre-established data verification training tree is sequentially traversed to obtain an average height of the line loss data to be measured;
[0008] Compare the average height of the line loss data to be tested with the preset height threshold, filter out abnormal attribute data and repair it.
[0009] Preferably, the establishment of the data verification training tree includes:
[0010] Selecting a set amount of line loss data from a pre-acquired line loss data set as sample data;
[0011] Sort the sample data in order of size and establish root nodes, where each root node corresponds to one sample data;
[0012] Circularly selecting attributes of the sample data in sequence, and randomly generating a cutting point for each attribute;
[0013] Taking the cutting point as a reference, a hyperplane is taken, and the hyperplane is divided into a left child node and a right child node according to the hyperplane to complete the construction of the data verification training tree.
[0014] Preferably, the establishment of the data verification training tree further includes:
[0015] For any piece of sample data in the training sample data, obtaining an average height of the sample data in each tree based on the height of the sample data in each tree;
[0016] Based on the average height, a preset height threshold is obtained by setting an upper and lower interval.
[0017] Preferably, the step of sequentially traversing a pre-established data verification training tree based on the attribute data of the line loss data to be measured to obtain an average height of the line loss data to be measured includes:
[0018] Traversing the attribute data of the line loss data to be measured sequentially through the data verification training tree, obtaining the height at which the current line loss data falls within the data verification training tree, and recording the current height;
[0019] Repeat the above steps, traverse the attribute data of the line loss data to be measured through all the data verification training trees in sequence, and calculate the average height of the line loss data falling on the data verification training trees.
[0020] Preferably, the comparing the height of the line loss data to be measured with a preset height range, screening out abnormal attribute data and repairing it, includes:
[0021] If the measured line loss data is outside the preset height threshold, the line loss data is abnormal; otherwise, the line loss data is normal;
[0022] When the line loss data is abnormal, repair the abnormal test data.
[0023] Preferably, the repairing of abnormal test data includes:
[0024] According to the cycle sequence of building the data verification training tree, obtain the data values of the previous and next cycles of the current line loss data;
[0025] Repair abnormal test data based on the data values of the previous and next cycles of the current line loss data.
[0026] Preferably, the abnormal test data is repaired based on the data values of the previous and next cycles of the current line loss data, as shown in the following formula:
[0027]
[0028] Where S i is the corrected line loss data; S i-1 is the data value of the previous cycle of the current line loss data; S i+1 It is the data value of the next cycle of the current line loss data.
[0029] Preferably, the line loss data includes: outlet power data, line transformer ledger data, user power data, and line topology data.
[0030] Preferably, the attribute data includes: line length, line model, transformer model, user power, line topology, power supply, and metered power.
[0031] A distribution network line loss data correction system, comprising:
[0032] Attribute data acquisition module: obtains the attribute data of the line loss data to be tested;
[0033] Average height calculation module: based on the attribute data of the line loss data to be measured, traverses the pre-established data verification training tree in sequence to obtain the average height of the line loss data to be measured;
[0034] Screening and repair module: compares the average height of the line loss data to be tested with the preset height threshold, screens out abnormal attribute data and repairs it.
[0035] Preferably, the average height calculation module includes: a training tree establishment submodule;
[0036] The training tree establishment submodule includes:
[0037] Sample data selection unit: selects a set amount of line loss data from the pre-acquired line loss data set as sample data;
[0038] Sorting unit: sorting the sample data in order of size and establishing root nodes, each of which corresponds to one sample data;
[0039] Cutting point generating unit: cyclically selecting attributes of the sample data in sequence, and randomly generating a cutting point for each attribute;
[0040] Construction unit: taking the cutting point as a reference, taking a hyperplane, dividing it into a left child node and a right child node according to the hyperplane, and completing the construction of the data verification training tree.
[0041] Preferably, the average height calculation module further includes:
[0042] An average value obtaining unit: for any piece of sample data in the training sample data, obtaining the average value of the height of the sample data in each tree based on the height of the sample data in each tree;
[0043] Threshold setting unit: Based on the average height value, a preset height threshold is obtained by setting upper and lower intervals.
[0044] Compared with the prior art, the present invention has the following beneficial effects:
[0045] The present invention provides a method for correcting distribution network line loss data, comprising: obtaining attribute data of the line loss data to be measured; traversing a pre-established data verification training tree in sequence based on the attribute data of the line loss data to be measured to obtain the average height of the line loss data to be measured; and comparing the average height of the line loss data to be measured with a preset height threshold, screening out abnormal attribute data, and repairing it. The present invention provides a method for verifying and correcting distribution network line loss data, which improves the efficiency of line loss data verification and effectively solves the problem of complex components and lines and poor data quality when calculating distribution network line loss. The verification and correction results meet the calculation accuracy of actual projects, allowing data upload to be synchronized, and solving the problem of synchronicity of distribution network line loss data. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 This is the flow chart of the method for verifying and correcting line loss data of the distribution network of the present invention;
[0047] Figure 2 This is a flow chart of the distribution network line loss data verification and correction steps of the present invention. DETAILED DESCRIPTION
[0048] In order to better understand the present invention, the present invention is further described below with reference to the accompanying drawings and examples.
[0049] Example 1:
[0050] A method for verifying and correcting the line loss data of a distribution network provided by the present invention is as Figure 1 shown:
[0051] S1: Obtain the attribute data of the line loss data to be measured:
[0052] The calculation of the line loss of the distribution network includes the calculation of the theoretical line loss and the statistical line loss. According to the Figure 2 flow chart of the steps for verifying and correcting the line loss data of the distribution network shown, the attributes of the input quantity of the line loss data in the figure include the line length L i , the line model LT i , the transformer model Tt i , the user power W i , the line topology Topo i ; the statistical line loss is the power supply electricity EP i , the meter electricity mEP i . Among them, the high-incidence problems of data quality are that the accuracy rate of the line topology is relatively low, the loss is serious, the synchronization rate of the user power data and the power supply electricity data is relatively low, and the error rate of the transformer and line models is relatively high. Set the size of the selected data sample set to n, and the elements in the sample set are S(L i ,LT i ,Tt i ,W i ,opo i ,mEP i ,EP i ), a total of 7 attributes;
[0053] Establish a data verification training tree T, select m samples (m < n) from the set S as the samples for constructing the tree Three, and put them into the root node of the data verification training tree;
[0054] For the data verification training tree T, sequentially select each attribute (line length L i , the line model LT i , the transformer model Tt i , the user power W i , the line topology Topo i , the power supply electricity EP i , the meter electricity mEP i ) from the m samples. Randomly generate a cut point p in the current attribute, and the cut point is generated between the maximum and minimum values of the dimension in the current node data, and put the p point into the data verification training tree T;
[0055] Taking point p as the reference, take its hyperplane, cyclically read the data of the 7 attributes of the sample set S, divide it into 2 subspaces according to the hyperplane, put the data less than p in the left child node of the current node, and put the data greater than or equal to p in the right child node of the current node;
[0056] The above steps are repeated for all k attributes in the m subset of the sample set S to complete the construction of the data verification training tree T. The entire training tree is constructed.
[0057] S2: based on the attribute data of the line loss data to be measured, traverse the pre-established data verification training tree in sequence to obtain the average height of the line loss data to be measured;
[0058] For any data S(L i ,LT i ,Tt i ,W i ,opo i ,mEP i ,EP i ), let it traverse each iTree, and then calculate S i Finally, it falls on the layer of each tree (the height of x in the tree). We can get S i The average height of each tree Hs i .
[0059] S3: Compare the average height of the line loss data to be tested with the preset height threshold, filter out abnormal attribute data and repair it:
[0060] A threshold value th is set, and an upper and lower interval is set based on the average height of the attribute data. Test data with a height lower than the preset threshold value is considered abnormal;
[0061] The numerical width of the upper and lower intervals is determined by the actual working conditions of the bureau;
[0062] For the outlier S i , can be smoothed and repaired by the normal data of the previous and next cycles. The specific method is to assume that the sample set S is collected in periodic order, and the abnormal value can be corrected as
[0063] Example 2
[0064] Based on the same inventive concept, the present invention also provides a distribution network line loss data correction system, comprising:
[0065] A distribution network line loss data correction system, comprising:
[0066] Attribute data acquisition module: obtains the attribute data of the line loss data to be tested;
[0067] Average height calculation module: based on the attribute data of the line loss data to be measured, traverses the pre-established data verification training tree in sequence to obtain the average height of the line loss data to be measured;
[0068] Screening and repair module: compares the average height of the line loss data to be tested with the preset height threshold, screens out abnormal attribute data and repairs it.
[0069] The average height calculation module includes: a training tree establishment submodule;
[0070] The training tree establishment submodule includes:
[0071] Sample data selection unit: selects a set amount of line loss data from the pre-acquired line loss data set as sample data;
[0072] Sorting unit: sorting the sample data in order of size and establishing root nodes, each of which corresponds to one sample data;
[0073] Cutting point generating unit: cyclically selecting attributes of the sample data in sequence, and randomly generating a cutting point for each attribute;
[0074] Construction unit: taking the cutting point as a reference, taking a hyperplane, dividing it into a left child node and a right child node according to the hyperplane, and completing the construction of the data verification training tree.
[0075] The average height calculation module further includes:
[0076] An average value obtaining unit: for any piece of sample data in the training sample data, obtaining the average value of the height of the sample data in each tree based on the height of the sample data in each tree;
[0077] Threshold setting unit: Based on the average height value, a preset height threshold is obtained by setting upper and lower intervals.
[0078] The average height calculation module includes:
[0079] Height recording unit: traverses the attribute data of the line loss data to be measured through the data verification training tree in sequence, obtains the height at which the current line loss data falls on the data verification training tree, and records the current height;
[0080] Average calculation unit: repeat the above steps, traverse the attribute data of the line loss data to be measured through all the data verification training trees in sequence, and calculate the average height of the line loss data falling on the data verification training trees,
[0081] The screening and repair module includes:
[0082] Judgment unit: if the measured line loss data is outside the preset height threshold, the line loss data is abnormal; otherwise, the line loss data is normal;
[0083] Repair unit: When the line loss data is abnormal, repair the abnormal test data.
[0084] The repair unit comprises:
[0085] Cycle acquisition subunit: obtains the data values of the previous and next cycles of the current line loss data according to the cycle sequence of building the data verification training tree;
[0086] Repair subunit: repairs abnormal test data based on the data values of the previous and next cycles of the current line loss data.
[0087] In the repair subunit, the abnormal test data is repaired as shown in the following formula:
[0088]
[0089] Where S i is the corrected line loss data; S i-1 is the data value of the previous cycle of the current line loss data; S i+1 It is the data value of the next cycle of the current line loss data.
[0090] The line loss data in the attribute data acquisition module includes: outlet power data, line transformer ledger data, user power data, and line topology data.
[0091] The attribute data in the attribute data acquisition module include: line length, line model, transformer model, user power, line topology, power supply, and metered power.
[0092] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0093] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.
[0094] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0095] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.
[0096] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.
[0097] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.
Claims
1. A method for correcting line loss data in a distribution network. It is characterized in that include: Obtaining attribute data of the line loss data to be tested; Based on the attribute data of the line loss data to be measured, a pre-established data verification training tree is traversed in sequence to obtain an average height of the line loss data to be measured; Compare the average height of the line loss data to be tested with the preset height threshold, filter out abnormal attribute data and repair them; The establishment of the data verification training tree includes: Selecting a set amount of line loss data from the total set of line loss data acquired in advance as sample data; Sort the sample data in order of size and establish root nodes, each of which corresponds to a sample data; Circularly selecting the attributes of the sample data in turn, and randomly generating a cutting point for each attribute; Taking the cutting point as a reference, taking a hyperplane, dividing it into a left child node and a right child node according to the hyperplane, and completing the construction of the data verification training tree; The establishment of the data verification training tree also includes: For any piece of sample data in the training sample data, based on the height of the sample data in each tree, obtain the average height of the sample data in each tree; Based on the average height, a preset height threshold is obtained by setting an upper and lower interval; The method of sequentially traversing a pre-established data verification training tree based on the attribute data of the line loss data to be measured to obtain an average height of the line loss data to be measured includes: The attribute data of the line loss data to be measured are sequentially traversed through the data verification training tree to obtain the height at which the current line loss data falls on the data verification training tree, and the current height is recorded; Repeat the above steps, traverse all the data verification training trees in sequence with the attribute data of the line loss data to be measured, and calculate the average height of the line loss data falling in the data verification training trees.
2. A method for correcting line loss data of a distribution network as claimed in claim 1, It is characterized in that The step of comparing the height of the line loss data to be measured with a preset height range, screening out abnormal attribute data and repairing the data includes: If the line loss data to be measured is outside the preset height threshold, the line loss data is abnormal; otherwise, the line loss data is normal; When the line loss data is abnormal, repair the abnormal test data.
3. A method for correcting line loss data of a distribution network as claimed in claim 2, It is characterized in that The repairing of abnormal test data includes: According to the periodic sequence of building the data verification training tree, obtain the data values of the previous and next periods of the current line loss data; Repair abnormal test data based on the data values of the previous and next cycles of the current line loss data.
4. A method for correcting line loss data of a distribution network as claimed in claim 3, It is characterized in that The abnormal test data is repaired based on the data values of the previous and next cycles of the current line loss data, as shown in the following formula: In the formula, S i is the corrected line loss data; S i-1 is the data value of the previous cycle of the current line loss data; S i+1 It is the data value of the next cycle of the current line loss data.
5. A method for correcting line loss data of a distribution network as claimed in claim 1, It is characterized in that The line loss data includes: export power data, line transformer ledger data, user power data, and line topology data.
6. A method for correcting line loss data of a distribution network as claimed in claim 1, It is characterized in that The attribute data include: line length, line model, transformer model, user power, line topology, power supply, and metered power.
7. A distribution network line loss data correction system, It is characterized in that include: Attribute data acquisition module: acquires attribute data of the line loss data to be tested; Average height calculation module: based on the attribute data of the line loss data to be measured, traverse the pre-established data verification training tree in sequence to obtain the average height of the line loss data to be measured; Screening and repair module: compares the average height of the line loss data to be tested with the preset height threshold, screens out abnormal attribute data and repairs them; The average height calculation module includes: a training tree establishment submodule; The training tree establishment submodule includes: Sample data selection unit: selects a set number of line loss data from the total set of line loss data acquired in advance as sample data; Sorting unit: sorting the sample data in order of size, establishing root nodes, each of which corresponds to a sample data; A cutting point generating unit: cyclically selecting attributes of the sample data in sequence, and randomly generating a cutting point for each attribute; Construction unit: taking the cutting point as a reference, taking a hyperplane, dividing it into a left child node and a right child node according to the hyperplane, and completing the construction of the data verification training tree; The average height calculation module also includes: An average value obtaining unit: for any sample data in the training sample data, obtaining the average value of the height of the sample data in each tree based on the height of the sample data in each tree; Threshold setting unit: based on the height average value, by setting the upper and lower intervals, to obtain a preset height threshold; The average height calculation module also includes Height recording unit: traverses the data verification training tree in sequence with the attribute data of the line loss data to be measured, obtains the height at which the current line loss data falls on the data verification training tree, and records the current height; Average calculation unit: repeat the above steps, traverse the attribute data of the line loss data to be measured through all the data verification training trees in sequence, and calculate the average height of the line loss data falling in the data verification training trees.
Citation Information
Patent Citations
Power network element loss abnormity determination method
CN104332987A
Method for lean distribution network emergency maintenance and management system based on big data technology
CN105701596A
Method and system for calculating theoretical line loss of power grid
CN105826918A
Line court anomaly analysis method based on big data
CN107527114A
Region network distribution loss reduction method
CN107909208A
