A main body loss assessment method and system for multi-body transmission channel network loss sharing
By calculating the target abnormality degree of sensor data in a multi-subject power grid and cleaning the data, combined with the current tracking algorithm, the network loss calculation error problem caused by abnormal data is solved, and the accuracy and fairness of network loss sharing are achieved.
Patent Information
- Application Number
- CN202411052647.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-02
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-02
AI Technical Summary
In multi-subject power grids, the network loss calculation error caused by abnormal sensor data affects the fairness of network loss sharing, and existing trend analysis algorithms are difficult to accurately evaluate the network loss of each subject.
By obtaining neighborhood data of sensor data sequence, calculating the target anomaly, cleaning the data, and evaluating network loss using a trend tracking algorithm to ensure the accuracy and fairness of network loss sharing.
Through the combination of data cleaning and trend tracking algorithms, the accuracy of network loss assessment is improved and the fairness and accuracy of network loss sharing is ensured.
Smart Images

Figure CN118585754B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power systems, and in particular to a method and system for evaluating the main body loss of multi-main body transmission channel network loss sharing. Background Art
[0002] The current multi-subject transmission channel refers to a power transmission network involving multiple subjects, also referred to as a multi-subject power grid, and the subjects refer to the participants in the power market, including at least power generation companies, independent power producers, transmission companies and distributors. Since the multi-subject power grid will generate losses in the process of transmitting electric energy, the reasonable and fair sharing of network losses is of great significance both in terms of improving the economic efficiency of the entire power system and ensuring the fairness of power grid users.
[0003] Since each subject has different degrees of use and methods of the power grid or is responsible for different nodes in the multi-subject power grid, the network loss that each subject needs to bear is also different in the process of network loss sharing. The node in the multi-subject power grid usually refers to a specific location in the multi-subject power grid, such as a substation, a connection point or a load center in the multi-subject power grid. In order to achieve fairness in network loss sharing, a power flow analysis algorithm is usually used to directly analyze and calculate the sensor data of each node in the multi-subject power grid that is monitored and collected. Based on the results of the analysis and calculation, the network loss of each node can be obtained, and then the network loss of each node in the node set in the multi-subject power grid that each subject is responsible for or controlled by is calculated. The network loss of each node can be obtained by calculating the network loss of each node to obtain the network loss that each node in the multi-agent power grid needs to bear. However, in the process of power transmission in the multi-agent power grid, transformer failure, line failure or extreme weather may occur, and the occurrence of transformer failure, line failure or extreme weather will cause abnormal sensor data in the sensor data collected at each node, and the existence of abnormal sensor data will cause errors or inaccuracies in the network loss of the node calculated and evaluated based on the power flow analysis algorithm, thereby causing errors or inaccuracies in the network loss that each node in the multi-agent power grid needs to bear, thereby affecting the fairness of network loss sharing. Therefore, how to accurately obtain the network loss that each node in the multi-agent power grid needs to bear is an urgent problem to be solved. Summary of the invention
[0004] In order to solve the above problems, the present invention provides a main body loss assessment method and system for multi-body transmission channel network loss allocation, and the technical scheme adopted is as follows:
[0005] In a first aspect, an embodiment of the present invention provides a method for evaluating the loss of a multi-subject transmission channel network loss, comprising the following steps:
[0006] In a multi-agent power grid, obtain the sensor data sequence of each node;
[0007] According to the sensor data sequence, a neighborhood data sequence of each data in the sensor data sequence is obtained;
[0008] Acquire associated nodes corresponding to the respective nodes, and obtain target abnormality levels of respective data in the sensor data sequences of the respective nodes according to neighborhood data sequences of respective data in the sensor data sequences of the respective nodes and neighborhood data sequences of respective data in the sensor data sequences of the associated nodes;
[0009] According to the target abnormality degree of each data, data cleaning is performed on the abnormal data in the sensor data sequence of each node to obtain the target sensor data sequence of each node;
[0010] According to the target sensor data sequence of each node and the power flow tracking algorithm, a network loss assessment value of each node is obtained;
[0011] In the multi-agent power grid, a node set corresponding to each subject is obtained, and the cumulative sum of network loss assessment values of all nodes in the node set corresponding to each subject is used as the network loss that the corresponding subject needs to bear.
[0012] In a second aspect, the present invention provides a subject loss assessment system for multi-subject transmission channel network loss sharing, comprising a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the above-mentioned subject loss assessment method for multi-subject transmission channel network loss sharing.
[0013] Beneficial effects: The present invention obtains the target abnormality degree of each data in the sensor data sequence of each node according to the neighborhood data sequence of each data in the sensor data sequence of the associated node and the neighborhood data sequence of each data in the sensor data sequence of each node; then, according to the target abnormality degree of each data, the abnormal data in the sensor data sequence of each node is cleaned to obtain the target sensor data sequence of each node, and the network loss assessment value of each node is obtained according to the target sensor data sequence of each node and the power flow tracking algorithm; finally, the node set corresponding to each subject in the multi-subject power grid is obtained, and the cumulative sum of the network loss assessment values of all nodes in the node set corresponding to each subject is used as the network loss that the corresponding subject needs to bear. The present invention obtains the target sensor data sequence of each node by performing data cleaning on the abnormal data in the sensor data sequence of each node, which can make the network loss that each subject needs to bear obtained subsequently more accurate, and thus can also ensure the fairness of network loss allocation. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0015] Figure 1 The present invention is a flow chart of a method for evaluating the main body loss of multi-body transmission channel network loss sharing. DETAILED DESCRIPTION
[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the embodiments of the present invention.
[0017] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0018] This embodiment provides a method for evaluating the loss of a main body for allocating network losses of a multi-body transmission channel, which is described in detail as follows:
[0019] like Figure 1 As shown, the main body loss assessment method for multi-body transmission channel network loss allocation includes the following steps:
[0020] Step S001: acquiring a sensor data sequence of each node in a multi-agent power grid.
[0021] The main purpose of this embodiment is to accurately obtain or evaluate the network loss that each subject in the multi-subject power grid needs to bear, and by accurately evaluating the network loss that each subject in the multi-subject power grid needs to bear, the fairness of network loss sharing can be guaranteed; in addition, network loss refers to the energy loss caused by factors such as resistance in the process of transmitting electricity or electric energy from the power generation point to the consumption point, and these energy losses mainly occur in the transmission and distribution links, including line loss, transformer loss, etc.
[0022] For the convenience of analysis and description, this embodiment will be described below by taking any multi-subject power grid as an example, that is, the multi-subject power grids that appear later in this embodiment all refer to the same multi-subject power grid; the multi-subject power grid in this embodiment refers to a power transmission network involving multiple subjects, and the subjects in this embodiment include but are not limited to power generation companies, independent power producers, transmission companies and distributors, etc., and each subject may control or be responsible for multiple nodes, and each subject may also own power generation facilities, transmission facilities or consumer loads.
[0023] In addition, the multi-subject power grid in this embodiment also includes multiple nodes. Each node in the multi-subject power grid refers to a specific location in the multi-subject power grid or a convergence point or distribution point of power flow in the multi-subject power grid. In the multi-subject power grid, not only are the network losses of different nodes different, but the locations and numbers of nodes controlled by each subject are also different, and the network loss of each node is determined by factors such as current load, line length and resistance.
[0024] Since the time-series voltage data and time-series current data corresponding to each node are needed to evaluate the network loss corresponding to each node in the multi-agent power grid based on the power flow tracking algorithm, that is, the sensor data in this embodiment refers to voltage data and current data. Therefore, this embodiment will collect and obtain the sensor data sequence corresponding to each node in the multi-agent power grid, that is, obtain the current data sequence and voltage sensor data corresponding to each node in the multi-agent power grid. The specific acquisition process is:
[0025] For any node a in a multi-agent power grid: all voltage data and all current data corresponding to node a are collected during the monitoring time period, and two sensor data of node a can be collected at a monitoring moment in the monitoring time period, which are the voltage data and current data of node a respectively; then a sequence constructed by collecting all voltage data corresponding to node a during the monitoring time period is recorded as the voltage data sequence corresponding to node a during the monitoring time period, a sequence constructed by collecting all current data corresponding to node a during the monitoring time period is recorded as the current data sequence corresponding to node a during the monitoring time period, and the current data sequence corresponding to node a and the voltage data sequence corresponding to node a are both recorded as the sensor data sequence corresponding to node a during the monitoring time period, and the data in the sensor data sequence are arranged in chronological order of collection time.
[0026] Therefore, based on the above, it can be known that at a monitoring moment, the current data and voltage data corresponding to node a can be collected, and then two sensor data sequences corresponding to node a can be obtained during the monitoring time period, namely, the voltage data sequence corresponding to node a during the monitoring time period and the current data sequence corresponding to node a during the monitoring time period.
[0027] In addition, this embodiment obtains the sensor data sequence corresponding to each node through a data acquisition and monitoring control system (SCADA), that is, a sensor, such as a current sensor and a voltage sensor, is arranged at each node position, and then the SCADA system collects the current or voltage data corresponding to each node by connecting to the sensor at each node position, and the data collected by the SCADA system will be stored in a central database.
[0028] Therefore, according to the above method of obtaining the two sensor data sequences corresponding to node a, the two sensor data sequences corresponding to each node in the monitoring time period can be obtained, and the lengths of the sensor data sequences corresponding to different nodes are required to be the same, and the collection time corresponding to the sensor data at the same position in different sensor data sequences is the same, that is, at any monitoring time in the monitoring time period, the sensor data corresponding to each node needs to be collected.
[0029] In addition, in specific applications, the implementer needs to set the monitoring time period and the time interval between adjacent monitoring moments according to actual conditions. For example, in this embodiment, the monitoring time period can be set to the multi-subject power grid from the start of operation to the current operation moment, and then the time intervals between adjacent monitoring moments can be set to be consistent, for example, all set to 1 second.
[0030] Therefore, this embodiment can obtain sensor data sequences of two different data types corresponding to each node in the multi-agent power grid through the above process, namely, a voltage data sequence and a current data sequence.
[0031] Step S002, obtaining a neighborhood data sequence of each data in the sensor data sequence according to the sensor data sequence; obtaining an associated node corresponding to each node, and obtaining a target abnormality degree of each data in the sensor data sequence of each node according to the neighborhood data sequence of each data in the sensor data sequence of each node and the neighborhood data sequence of each data in the sensor data sequence of the associated node.
[0032] Since this embodiment will subsequently determine the network loss of each subject in the multi-subject power grid by obtaining the network loss of each node, and the network loss of each subject refers to the network loss that each subject needs to bear, and the network loss of each node is usually evaluated by a power flow analysis algorithm. However, in the process of power transmission in the multi-subject power grid, transformer failure, line failure or extreme weather may occur, and the occurrence of transformer failure, line failure or extreme weather will cause abnormal data in the sensor data sequence of each node monitored and collected, and the existence of abnormal data will cause errors or inaccuracies in the network loss of the node calculated and evaluated based on the power flow analysis algorithm, which will lead to the inability to accurately obtain the network loss that each subject needs to bear. In order to accurately obtain the network loss that each subject needs to bear, this embodiment also ensures Regarding the fairness of network loss sharing, this embodiment will analyze the possibility that each data in the sensor data sequence of each node is abnormal data to eliminate the influence of abnormal data in the sensor data sequence on the subsequent evaluation of the network loss corresponding to each node as much as possible. That is, this embodiment will mainly determine the target sensor data sequence of each node by obtaining the target abnormality degree of each data in the sensor data sequence. The target abnormality degree of the data can reflect the possibility that the corresponding data is abnormal data. Then, based on the power flow tracking algorithm, the target sensor data sequence of each node is subjected to power flow analysis calculation, and the network loss assessment value of each node is obtained based on the result of the power flow analysis calculation. Then, based on the obtained network loss assessment value of each node, the purpose of accurately evaluating the network loss of each subject in the multi-subject power grid is finally achieved.
[0033] Therefore, based on the above analysis, it can be known that in order to achieve the purpose of accurately obtaining the network loss of each subject, this embodiment needs to first analyze the possibility that each data in the sensor data sequence of each node is abnormal data, that is, it is necessary to first obtain the target abnormality degree of each data in the sensor data sequence of each node. Since the target abnormality degree of data in the sensor data sequences of different data types in this embodiment is obtained in the same way, and for the convenience of analysis, this embodiment will take the method of obtaining the target abnormality degree of each data in the voltage data sequence corresponding to each node as an example for analysis, that is, in step S002, the sensor data sequence or the data in the sequence that appear later refers to voltage data, then in this embodiment, the specific process of obtaining the target abnormality degree of each data in the sensor data sequence of each node is:
[0034] Based on the above analysis, it can be known that transformer failure, line failure or the occurrence of extreme weather is the root cause of the abnormal data in the sensor data sequence of each node monitored and collected, and the abnormal data has a mutation feature in a local range or the local area where the abnormal data is located has a large fluctuation. Therefore, when obtaining the target abnormality degree of each data, this embodiment needs to first obtain the neighborhood data sequence of each data in the sensor data sequence of each node, and the specific acquisition process of the neighborhood data sequence of each data in this embodiment is:
[0035] For the bth data in the sensor data sequence A of any node a in a multi-agent power grid: in the sensor data sequence A, the bth data and a sequence constructed by a preset number of consecutive data adjacent to the bth data and located on the left of the bth data are recorded as the neighborhood data sequence of the bth data, that is, the neighborhood data sequence of the bth data refers to the data sequence formed by the bRth data to the bth data in the sensor data sequence A, and at this time, the neighborhood data sequence of the bth data includes the bRth data and the bth data, where R refers to the value of the preset number.
[0036] In addition, in this embodiment, there may be a situation where the number of data located on the left side of the b-th data is less than the preset number. In this embodiment, when the number of data located on the left side of the b-th data is less than the preset number, this embodiment records the sequence constructed by all the data located on the left side of the b-th data as the neighborhood data sequence of the b-th data.
[0037] Therefore, this embodiment can obtain the neighborhood data sequence of each data in the sensor data sequence through the above method; and in specific applications, the implementer needs to set the value of the preset number according to the actual situation, for example, in this embodiment, the value of the preset number is set to 49 or 50.
[0038] In addition, in the process of transmitting electricity in a multi-agent power grid, the load equipment corresponding to the node may be suddenly loaded or unloaded, and the sudden loading or unloading of the load equipment corresponding to the node will also cause mutation data in the sensor data sequence of the node or large fluctuations in the sequence, and may also cause mutation data or large fluctuations in the sensor data sequence of the node's associated nodes. The mutation data caused by sudden loading or unloading is not abnormal data, but normal data; therefore, if it is based only on the sensor data sequence corresponding to the node itself, normal data will appear in the acquired abnormal data, that is, the mutation data or fluctuating data caused by loading or unloading will also be identified as abnormal data. Therefore, in order to avoid the fluctuation data or mutation data caused by the loading or unloading of high-load equipment in the power grid from being identified as abnormal data, when acquiring the abnormal data in the sensor data sequence of each node or when acquiring the target abnormality degree of each data in the sensor data sequence, this embodiment will analyze the degree of association between the node and its corresponding associated node to obtain the first abnormal index value corresponding to each data in the sensor data sequence of each node, and then determine the target abnormality degree based on the first abnormal index value; based on the above analysis, it can be seen that the associated node of the node is crucial to the acquisition of the target abnormality degree, so this embodiment needs to obtain the associated node corresponding to each node, and the specific acquisition process is:
[0039] For node a in a multi-agent power grid: according to the power transmission direction of the multi-agent power grid, obtain all connected nodes of node a, and record them as associated nodes corresponding to node a; for example, the power transmission direction is from node a to node a0, and the power flowing through node a will also flow through node a0 later, and node a is directly connected to node a0, then it means that node a0 is the associated node of node a.
[0040] Therefore, all associated nodes corresponding to all nodes can be obtained through the above process; and after obtaining the associated nodes corresponding to the nodes, this embodiment will then obtain the target abnormality degree of each data in the sensor data sequence of each node according to the neighborhood data sequence of each data in the sensor data sequence of each node and the neighborhood data sequence of each data in the sensor data sequence of the associated node. The specific process is:
[0041] In this embodiment, firstly, the first abnormality index value corresponding to each data in the sensor data sequence of each node is obtained according to the neighborhood data sequence of each data in the sensor data sequence of each node and the neighborhood data sequence of each data in the sensor data sequence of the associated node. The first abnormality index value is a key indicator for the subsequent acquisition of the target abnormality degree corresponding to the corresponding data. Therefore, this embodiment needs to specifically analyze and describe the acquisition process of the first abnormality index value. In addition, for the sake of ease of understanding, this embodiment will take the acquisition process of the first abnormality index value corresponding to the b-th data in the sensor data sequence A of the node a as an example for description, that is, the acquisition process of the first abnormality index value corresponding to the b-th data in the sensor data sequence A is:
[0042] First, the set constructed by the b-th data in the sensor data sequence of each associated node corresponding to node a is recorded as the associated data set F0 corresponding to the b-th data in the sensor data sequence A, and the number of data in the associated data set F0 is consistent with the number of associated nodes corresponding to node a, that is, the associated data set F0 is composed of the b-th data in the sensor data sequences of all associated nodes corresponding to node a.
[0043] Then, according to the neighborhood data sequence of the b-th data in the sensor data sequence A and the neighborhood data sequence of the associated data in the set F0, the contribution characterization value corresponding to each associated data in the set F0 is obtained, and the contribution characterization value is a key indicator for subsequently obtaining the first abnormal index value; in addition, for ease of understanding, this embodiment will take the acquisition process of the contribution characterization value corresponding to the m-th associated data in the associated data set F0 as an example for analysis, so the acquisition process of the contribution characterization value corresponding to the m-th associated data is:
[0044] First, all target feature values between the bth data and the mth associated data in the sensor data sequence A are obtained, and the target feature values can reflect the degree of association between the bth data and the mth associated data in the sensor data sequence A, and the strength of the degree of association is the basis for subsequently determining the contribution representation value corresponding to the mth associated data; in addition, the number of target feature values between the bth data and the mth associated data in the sensor data sequence A is the same as the number of data in the neighborhood data sequence A0 or the neighborhood data sequence Am, and the neighborhood data sequence A0 is the neighborhood data sequence of the bth data in the sensor data sequence A, and the neighborhood data sequence Am is the neighborhood data sequence of the mth associated data.
[0045] In this embodiment, the specific process of obtaining all target feature values between the b-th data and the m-th associated data in the sensor data sequence A is as follows:
[0046] First, according to the cth data in the neighborhood data sequence of each associated data in the set F0, the neighborhood data sequence A0, the cth data in the neighborhood data sequence A0, and the cth data in the neighborhood data sequence Am, the cth target feature value between the bth data in the sensor data sequence A and the mth associated data is obtained; and the specific acquisition process of the cth target feature value between the bth data in the sensor data sequence A and the mth associated data in this embodiment is:
[0047] First, the mean of the neighborhood data sequence A0 is obtained, and then the normalized value of the absolute value of the difference between the mean of the neighborhood data sequence A0 and the c-th data in the neighborhood data sequence A0 is calculated, and recorded as the first difference value corresponding to the c-th data in the neighborhood data sequence A0; then the difference between the preset first constant and the first difference value is obtained, and recorded as the c-th first eigenvalue between the b-th data and the m-th associated data in the sensor data sequence A.
[0048] Then, according to the cth data in the neighborhood data sequence of each associated data in the set F0 and the cth sensor data in the neighborhood data sequence A0, the cth second eigenvalue between the bth data in the sensor data sequence A and the mth associated data is obtained; and the specific acquisition process of the cth second eigenvalue is: firstly, the first distance set corresponding to the set F0 is obtained, and then the minimum first distance in the first distance set is selected, and the value obtained by negative correlation mapping the minimum first distance in the first distance set is recorded as the mapping value, and then the difference between the preset first constant and the mapping value is obtained, and recorded as the cth second eigenvalue between the bth data in the sensor data sequence A and the mth associated data, and the fth first distance in the first distance set is the Euclidean distance between the cth data in the neighborhood data sequence A0 and the cth data in the neighborhood data sequence Af, and the neighborhood data sequence Af is the neighborhood data sequence of the fth associated data in the set F0.
[0049] Then, the Euclidean distance between the cth data in the neighborhood data sequence A0 and the cth data in the neighborhood data sequence Am is calculated and recorded as the cth third eigenvalue between the bth data in the sensor data sequence A and the mth associated data.
[0050] Then, the product of the cth first eigenvalue, the cth second eigenvalue, and the cth third eigenvalue between the bth data and the mth associated data in the sensor data sequence A is obtained, and recorded as the cth target eigenvalue between the bth data and the mth associated data in the sensor data sequence A.
[0051] In a specific application, the implementer needs to set the preset first constant according to the actual situation. For example, in this embodiment, the preset first constant is set to constant 1.
[0052] Therefore, this embodiment can obtain all target feature values between the bth data and the mth associated data in the sensor data sequence A according to the above process of obtaining the cth target feature value; after obtaining all target feature values between the bth data and the mth associated data in the sensor data sequence A, the cumulative sum of all target feature values between the bth data and the mth associated data in the sensor data sequence A is obtained and recorded as a comprehensive value, and then the value obtained after negative correlation mapping of the comprehensive value is recorded as the correlation degree representation value between the bth data and the mth associated data in the sensor data sequence A, and the correlation degree representation value between the bth data and the mth associated data in the sensor data sequence A is recorded as the contribution representation value corresponding to the mth associated data in the relational set F0.
[0053] In addition, in this embodiment, the specific expression for obtaining the contribution representation value corresponding to the mth associated data in the set F0 is:
[0054]
[0055] in, is the contribution representation value corresponding to the mth associated data in the set F0, and is also the representation value of the degree of association between the bth data and the mth associated data in the sensor data sequence A; softmax() is a normalization function, is the cth first eigenvalue between the bth data and the mth associated data in the sensor data sequence A, is the cth second eigenvalue between the bth data and the mth associated data in the sensor data sequence A, is the cth third eigenvalue between the bth data and the mth associated data in the sensor data sequence A, and L is the total number of data in the neighborhood data sequence A0 or the neighborhood data sequence Am.
[0056] in addition, , , , where Norm() is the normalization function, g0 is the absolute value of the difference between the mean of the neighborhood data sequence A0 and the cth data in the neighborhood data sequence A0, and exp() is an exponential function with a constant e as the base. is the minimum first distance in the first distance set corresponding to the associated data set F0, It is the Euclidean distance between the cth data in the neighborhood data sequence A0 and the cth data in the neighborhood data sequence Am.
[0057] And when , as well as The smaller the value of The larger the value of The larger the value of , the larger the correlation degree representation value between the b-th data and the m-th data in the sensor data sequence A. The larger the correlation degree representation value, the stronger the correlation between the b-th data and the m-th associated data in the sensor data sequence A. The stronger the correlation, the larger the contribution representation value corresponding to the m-th data.
[0058] Therefore, this embodiment can obtain the contribution representation value corresponding to each associated data in the set F0 according to the above method for obtaining the contribution representation value corresponding to the mth associated data; after obtaining the contribution representation value corresponding to each associated data in the set F0, this embodiment will then obtain the association feature mean corresponding to the bth data in the sensor data sequence A, and the specific acquisition process is:
[0059] The characteristic data value corresponding to each associated data in the set F0 is obtained, and then the mean of the characteristic data values corresponding to all the associated data in the set F0 is calculated and recorded as the associated characteristic mean corresponding to the b-th data in the sensor data sequence A; and the process of obtaining the characteristic data value is: for any associated data G in the set F0, the product of the associated data G and its corresponding contribution representation value is recorded as the characteristic data value corresponding to the associated data G.
[0060] Next, the Euclidean distance between the bth data in the sensor data sequence A and its corresponding associated feature mean is obtained, and the Euclidean distance between the bth data in the sensor data sequence A and its corresponding associated feature mean is negatively correlated and mapped, and the value obtained by the negative correlation mapping is recorded as the feature mapping value, and then the difference between the constant 1 and the feature mapping value is calculated, and recorded as the first abnormality index value corresponding to the bth data in the sensor data sequence A; and the first abnormality index value obtained above is the normalized value of the Euclidean distance between the bth data in the sensor data sequence A and its corresponding associated feature mean; in addition, at this time, the method of negatively correlating the Euclidean distance between the bth data in the sensor data sequence A and its corresponding associated feature mean is exp(-W), wherein W is the Euclidean distance between the bth data in the sensor data sequence A and its corresponding associated feature mean, and the first abnormality index value can reflect the possibility that the bth data is abnormal data.
[0061] Therefore, this embodiment can obtain the first abnormality index value corresponding to each data in the sensor data sequence of each node through the process of obtaining the first abnormality index value corresponding to the bth data in the sensor data sequence A of the node a.
[0062] After obtaining the first abnormality index value, this embodiment uses the COF abnormality detection algorithm to obtain the second abnormality index value corresponding to each data. The specific process is as follows:
[0063] For node a, according to sensor data sequence A, the COF anomaly detection algorithm is used to perform anomaly detection on each data in the sensor data sequence A, and the detection anomaly value corresponding to each data in the sensor data sequence A is obtained, and recorded as the second anomaly index value corresponding to the corresponding data.
[0064] Therefore, the second abnormality index value corresponding to each data in the sensor data sequence of each node can be obtained through the above process; and the process of obtaining the detection abnormality value using the COF abnormality detection algorithm is a well-known technology, so it will not be described in detail in this embodiment.
[0065] Then, the normalized value of the product of the second abnormal index value corresponding to each data in the sensor data sequence of each node and its corresponding first abnormal index value is calculated and recorded as the target abnormality degree of the corresponding data. The greater the target abnormality degree of the data, the greater the possibility that the corresponding data is abnormal data. Conversely, the smaller the target abnormality degree of the data, the smaller the possibility that the corresponding data is abnormal data. In addition, the greater the second abnormal index value corresponding to the data and its corresponding first abnormal index value, the greater the target abnormality degree of the corresponding data.
[0066] Step S003, according to the target abnormality degree of each data, the abnormal data in the sensor data sequence of each node is cleaned to obtain the target sensor data sequence of each node; according to the target sensor data sequence of each node and the power flow tracking algorithm, the network loss assessment value of each node is obtained; in the multi-agent power grid, the node set corresponding to each subject is obtained, and the cumulative sum of the network loss assessment values of all nodes in the node set corresponding to each subject is used as the network loss that the corresponding subject needs to bear.
[0067] Next, this embodiment will perform data cleaning on the abnormal data in the sensor data sequence based on the target abnormal degree of each data, obtain the target sensor data sequence of each node, and then determine the network loss assessment value of each node based on the target sensor data sequence, and determine the network loss that each subject in the multi-subject power grid needs to bear based on the network loss assessment value of each node, specifically:
[0068] Since what is obtained in step S002 is the target abnormality level of each voltage data in the voltage data sequence, and this embodiment is based on the method of obtaining the target abnormality level of each data in the voltage data sequence in step S002, it is possible to obtain the target abnormality level of each current data in the current data sequence of each node, that is, this embodiment is able to obtain the target abnormality level of each data in each sensor data sequence.
[0069] Then, based on the obtained target abnormality degree, the abnormal data in the sensor data sequence is cleaned to obtain the target sensor data sequence of each node, specifically: for the bth data in the sensor data sequence A of the node a, it is determined whether the target abnormality degree of the bth data is less than the preset judgment threshold value, if so, the bth data is determined to be normal data, and the bth data is used as the target sensor data corresponding to the bth sensor data, otherwise, the bth data is determined to be abnormal data, and in the sensor data sequence A, the data closest to the bth sensor data and less than the preset judgment threshold value is obtained, and the obtained data closest to the bth data and less than the preset judgment threshold value is used as the target sensor data corresponding to the bth data; that is, this embodiment replaces the abnormal data with normal data, thereby avoiding the problem of inaccurate assessment caused by abnormal data during subsequent network loss assessment.
[0070] In addition, there may be two data that are closest to the b-th data and less than the preset judgment threshold, and they are located on the left and right of the b-th data respectively. When this happens, this embodiment selects the data located on the left side of the b-th data as the target sensor data corresponding to the b-th data.
[0071] In specific applications, the implementer needs to set the preset determination threshold according to the actual situation. For example, in this embodiment, the preset determination threshold is set to 0.7. In addition, the above process of replacing the abnormal data in the sequence is the process of cleaning the abnormal data in the sequence.
[0072] Therefore, through the above process, the target sensor data corresponding to each data in each sensor data sequence of node a can be obtained, and then the sequence constructed by the target sensor data corresponding to each data in any sensor data sequence corresponding to node a is recorded as the target sensor data sequence of the corresponding sensor data sequence. Therefore, this embodiment can obtain the target sensor data sequence of each sensor data sequence of each node, and the target sensor data sequences of each sensor data sequence of each node are recorded as the target sensor data sequence of the corresponding node, that is, each node corresponds to two target sensor data sequences.
[0073] Then, the power flow tracking algorithm is used to perform power flow analysis and calculation on the target sensor data sequence of each node to obtain the network loss assessment value of each node; then, in the multi-agent power grid, the node set corresponding to each subject is obtained, and the node set corresponding to each subject is composed of all nodes controlled or responsible by the corresponding subject; then, the cumulative sum of the network loss assessment values of all nodes in the node set corresponding to each subject is obtained, and used as the network loss that the corresponding subject needs to bear or as the network loss of the corresponding subject.
[0074] Furthermore, since the process of using a power flow tracking algorithm to perform power flow analysis and calculation on the target sensor data sequence of each node to obtain the network loss assessment value of each node is a well-known technology, it will not be described in detail in this embodiment.
[0075] A subject loss assessment system for multi-subject transmission channel network loss allocation in this embodiment includes a memory and a processor, and the processor executes a computer program stored in the memory to implement the above-mentioned subject loss assessment method for multi-subject transmission channel network loss allocation.
[0076] In summary, this embodiment obtains the target abnormality degree of each data in the sensor data sequence of each node according to the neighborhood data sequence of each data in the sensor data sequence of the associated node and the neighborhood data sequence of each data in the sensor data sequence of each node; then, according to the target abnormality degree corresponding to each data, the abnormal data in the sensor data sequence of each node is cleaned to obtain the target sensor data sequence of each node, and the network loss assessment value of each node is obtained according to the target sensor data sequence of each node and the power flow tracking algorithm; finally, the node set corresponding to each subject in the multi-subject power grid is obtained, and the cumulative sum of the network loss assessment values of all nodes in the node set corresponding to each subject is used as the network loss that the corresponding subject needs to bear. This embodiment obtains the target sensor data sequence of each node by cleaning the abnormal data in the sensor data sequence of each node, which can make the network loss that each subject needs to bear obtained later more accurate, and thus can also ensure the fairness of network loss allocation.
[0077] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.
Claims
1. A method for assessing the loss of a subject for sharing the network loss of multiple subject delivery channels, characterized in that: The method comprises the following steps: In a multi-agent power grid, obtain the sensor data sequence of each node; According to the sensor data sequence, a neighborhood data sequence of each data in the sensor data sequence is obtained; Acquire associated nodes corresponding to the respective nodes, and obtain target abnormality levels of respective data in the sensor data sequences of the respective nodes according to neighborhood data sequences of respective data in the sensor data sequences of the respective nodes and neighborhood data sequences of respective data in the sensor data sequences of the associated nodes; According to the target abnormality degree of each data, data cleaning is performed on the abnormal data in the sensor data sequence of each node to obtain the target sensor data sequence of each node; According to the target sensor data sequence of each node and the power flow tracking algorithm, a network loss assessment value of each node is obtained; In the multi-agent power grid, a node set corresponding to each subject is obtained, and the cumulative sum of network loss assessment values of all nodes in the node set corresponding to each subject is used as the network loss that the corresponding subject needs to bear.
2. A method for assessing the loss of a subject for sharing the network loss of a multi-subject transmission channel according to claim 1, characterized in that: The method for obtaining a neighborhood data sequence of each data in the sensor data sequence comprises: For the bth data in the sensor data sequence A of any node a: In the sensor data sequence A, a sequence constructed by the b-th data and a preset number of consecutive data adjacent to the b-th data and located on the left side of the b-th data is recorded as a neighborhood data sequence of the b-th data.
3. A method for assessing the loss of a subject for sharing the network loss of a multi-subject transmission channel according to claim 1, characterized in that: The method for obtaining the associated nodes corresponding to each of the nodes includes: For any node a: According to the power transmission direction of the multi-agent power grid, all connected nodes of the node a are obtained and recorded as associated nodes corresponding to the node a.
4. A method for assessing the loss of a subject for sharing the network loss of a multi-subject transmission channel according to claim 1, characterized in that: The method for obtaining the target abnormality degree of each data in the sensor data sequence of each node includes: Obtaining, according to the neighborhood data sequence of each data in the sensor data sequence of each node and the neighborhood data sequence of each data in the sensor data sequence of the associated node, a first abnormality index value corresponding to each data in the sensor data sequence of each node; Using the COF anomaly detection algorithm to perform anomaly detection on each data in the sensor data sequence of each node, obtain a detection anomaly value corresponding to each data in the sensor data sequence of each node, and record it as a second anomaly index value; The normalized value of the product of the second abnormal index value corresponding to each data and the first abnormal index value corresponding to the data is recorded as the target abnormality degree of the corresponding data.
5. A method for assessing the loss of a subject for sharing the network loss of a multi-subject transmission channel according to claim 4, characterized in that: The method for obtaining the first abnormality index value corresponding to each data in the sensor data sequence of each node includes: For the bth data in the sensor data sequence A of the node a: The set constructed by the b-th data in the sensor data sequence of each associated node corresponding to the node a is recorded as the associated data set F0 corresponding to the b-th data in the sensor data sequence A; Obtaining contribution representation values corresponding to respective associated data in the associated data set F0 according to a neighborhood data sequence of the b-th data in the sensor data sequence A and a neighborhood data sequence of associated data in the associated data set F0; For any associated data G in the associated data set F0, the product of the associated data G and its corresponding contribution representation value is recorded as the feature data value corresponding to the associated data G; the mean of the feature data values corresponding to all associated data in the associated data set F0 is recorded as the associated feature mean corresponding to the b-th data in the sensor data sequence A; The normalized value of the Euclidean distance between the b-th data in the sensor data sequence A and the corresponding associated feature mean is recorded as the first abnormality indicator value corresponding to the b-th data in the sensor data sequence A.
6. A method for assessing the loss of a subject for sharing the network loss of a multi-subject transmission channel according to claim 5, characterized in that: The method for obtaining the contribution representation value corresponding to each associated data in the associated data set F0 includes: For the mth associated data in the associated data set F0: According to the cth data in the neighborhood data sequence of each associated data in the associated data set F0, the neighborhood data sequence A0 of the bth data in the sensor data sequence A, the cth data in the neighborhood data sequence A0, and the cth data in the neighborhood data sequence Am of the mth associated data, the cth target feature value between the bth data in the sensor data sequence A and the mth associated data is obtained; the cumulative sum of all target feature values between the bth data in the sensor data sequence A and the mth associated data is recorded as a comprehensive value, the value obtained after negative correlation mapping of the comprehensive value is recorded as a correlation degree representation value between the bth data in the sensor data sequence A and the mth associated data, and the correlation degree representation value is recorded as a contribution representation value corresponding to the mth associated data.
7. A method for assessing the loss of a subject for network loss sharing of a multi-subject transmission channel as claimed in claim 6, characterized in that: The method for obtaining the cth target feature value between the bth data in the sensor data sequence A and the mth associated data comprises: Record the normalized value of the absolute value of the difference between the mean of the neighborhood data sequence A0 and the c-th data in the neighborhood data sequence A0 as the first difference value corresponding to the c-th data in the neighborhood data sequence A0, and record the difference between the preset first constant and the first difference value as the c-th first eigenvalue between the b-th data in the sensor data sequence A and the m-th associated data; According to the cth data in the neighborhood data sequence of the associated data and the cth data in the neighborhood data sequence A0, obtaining the cth second eigenvalue between the bth data in the sensor data sequence A and the mth associated data; The Euclidean distance between the cth data in the neighborhood data sequence A0 and the cth data in the neighborhood data sequence Am is recorded as the cth third eigenvalue between the bth data in the sensor data sequence A and the mth associated data; The product of the cth first eigenvalue, the cth second eigenvalue and the cth third eigenvalue is recorded as the cth target eigenvalue between the bth data in the sensor data sequence A and the mth associated data.
8. A method for assessing the loss of a subject for sharing the network loss of multiple subject delivery channels as claimed in claim 7, characterized in that: The method for obtaining the cth second characteristic value between the bth data in the sensor data sequence A and the mth associated data includes: A first distance set corresponding to the associated data set F0 is obtained, and a value obtained by negatively correlating the minimum first distance in the first distance set is recorded as a mapping value. A difference between a preset first constant and the mapping value is recorded as the cth second eigenvalue between the bth data in the sensor data sequence A and the mth associated data. The fth first distance in the first distance set is the Euclidean distance between the cth data in the neighborhood data sequence A0 and the cth data in the neighborhood data sequence Af. The neighborhood data sequence Af is the neighborhood data sequence of the fth associated data in the associated data set F0.
9. A method for assessing the loss of a subject for sharing the network loss of a multi-subject transmission channel according to claim 1, characterized in that: The method of performing data cleaning on abnormal data in the sensor data sequence of each node according to the target abnormal degree of each data to obtain the target sensor data sequence of each node includes: For the bth data in the sensor data sequence A of any node a, determine whether the target abnormality degree of the bth data is less than a preset determination threshold value. If so, determine that the bth data is normal data, and use the bth data as the target sensor data corresponding to the bth data. Otherwise, determine that the bth data is abnormal data, and use the data closest to the bth data and less than the preset determination threshold value in the sensor data sequence A as the target sensor data corresponding to the bth data. A sequence constructed by the target sensor data corresponding to each data in the sensor data sequence A is recorded as the target sensor data sequence of the node a.
10. A main body loss assessment system for multi-main body transmission channel network loss sharing, comprising a memory and a processor, characterized in that: The processor executes the computer program stored in the memory to implement a subject loss assessment method for multi-subject transmission channel network loss sharing as described in any one of claims 1-9.
