A Method and System for Identifying the Switch Status of Distribution Network Lines Based on Dynamic Combination Weighting and Evidence Reasoning
By using dynamic combination weighting and evidence reasoning methods, the problems of unreasonable weight settings and uncertainty in the identification of switch status in distribution network lines are solved. This enables accurate identification of switch status and traceability of degradation, dynamic updating of weights, retention of uncertain information, identification of status level jumps, and location of degradation sources.
Patent Information
- Application Number
- CN202611312214.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-27
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies struggle to balance the inherent engineering importance of indicators with real-time equipment degradation information in the identification of distribution network line switch status. This leads to unreasonable weighting, uncertainties caused by conflicting evidence and incomplete information, and an inability to accurately identify the status and degradation sources of multi-source heterogeneous data.
A dynamic combination weighting and evidence reasoning method is adopted. A state index comparison and judgment matrix is constructed through the analytic hierarchy process. By combining information entropy and degradation degree variable weights, dynamic combination weight calculation of state indexes is performed. Evidence reasoning is used to fuse state levels and identify the state of the switch body, secondary part and external environment.
It enables accurate identification and degradation tracing of distribution network line switch status, dynamically updates weighted response equipment degradation, retains uncertain information, avoids information loss, and can identify status level jumps and confidence changes to pinpoint the source of degradation.
Smart Images

Figure CN123116009A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of intelligent monitoring of distribution network lines and artificial intelligence technology, and in particular to a method and system for identifying the switch status of distribution network lines based on dynamic combination weighting and evidence reasoning. Background Technology
[0002] The distribution network, located at the end of the power system, directly undertakes the tasks of power distribution and user supply. Distribution network line switches are used for line segmentation, load transfer, fault isolation, and power restoration; their operating status directly affects the fault range, outage duration, and power supply reliability of the distribution network. With the continuous advancement of distribution automation, equipment such as pole-mounted switches and ring main unit switches are no longer merely primary devices for opening and closing operations, but are gradually forming integrated primary and secondary equipment comprised of the switch body, operating mechanism, measurement and control unit, communication unit, backup power supply, and the field environment. The status information of distribution network line switches comes from online monitoring, live-line testing, routine tests, inspections, defect and fault records, and environmental monitoring, representing typical multi-source heterogeneous data.
[0003] Traditional operation and maintenance are divided into reactive maintenance and periodic maintenance. Reactive maintenance often leads to unplanned power outages; periodic maintenance is carried out according to a fixed cycle, which can result in over-maintenance of equipment in good condition and untimely maintenance of equipment that is deteriorating. Condition-based maintenance determines the timing of maintenance based on the actual operating status of the equipment, provided that a comprehensive identification of the equipment's condition is completed.
[0004] Existing technologies can utilize multiple status indicators of distribution network equipment for health assessment, and fuzzy evaluation, hierarchical analysis, and entropy weighting methods are also used to process multi-source status information. However, there are still many technical problems in identifying the status of distribution network line switches at multiple operating times. For example, the weights are difficult to balance the inherent engineering importance of the indicators with real-time equipment degradation information; when determining the level, there are uncertainties caused by lost level boundaries, conflicting evidence, and incomplete information; the output results at the whole-machine level cannot distinguish the status of the switch body, secondary components, and external environment, and cannot locate the dominant indicators that cause status degradation. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a method and system for identifying the status of distribution network line switches based on dynamic combination weighting and evidence reasoning. This invention can completely preserve the uncertainty information in the status identification process and can realize the status identification and degradation tracing of distribution network line switches by functional group.
[0006] In a first aspect, the present invention provides a method for identifying the status of distribution network line switches based on dynamic combination weighting and evidence reasoning, including: The system acquires the number of status indicators of the distribution network line switch to be identified at multiple times, and preprocesses the status indicator data to obtain valid status indicator data. The effective state index data is divided into multiple state index groups, and the state indexes within each state index group are processed to be directionally consistent and dimensionless, resulting in dimensionless samples of state indexes. Based on the evaluation batch samples at the current target evaluation time, the information entropy and the degradation degree of each state index are calculated using dimensionless samples. A state index comparison and judgment matrix is constructed based on the analytic hierarchy process. After consistency verification, the matrix is normalized to obtain the constant weights of the importance of the inherent engineering importance of the state index. The deterioration degree variable weight and importance constant weight of the same state indicator are multiplied one by one, and then normalized within the corresponding state indicator group to obtain the dynamic combination weight of the state indicators at each target evaluation time. According to the pre-configured evaluation rules, the observed values of the state indicators are converted into confidence vectors of different state levels, and the confidence vectors of all state indicators in the same state indicator group are combined to form a state confidence matrix. For each state indicator group, the basic probability quality allocation is completed by combining the dynamic combination weights within the state indicator group with the state confidence matrix, and the conflict quality between evidences is calculated. The evidence reasoning recursively synthesizes the evidence to obtain the comprehensive quality after the fusion of all state indicators within the state indicator group. Based on the integrated quality calculation, the comprehensive reliability and comprehensive ignorance of each state indicator group are calculated, and the state level of each state indicator group is determined according to the principle of maximum reliability. The state levels of the state indicator groups at adjacent evaluation times are compared, and candidate state indicators are selected to determine the dominant indicator.
[0007] Secondly, the present invention provides a distribution network line switch status identification system based on dynamic combination weighting and evidence reasoning, comprising: The data acquisition module is used to acquire the number of status indicators of the distribution network line switch to be identified at multiple times, and to preprocess the number of status indicators to obtain valid status indicator data. The preprocessing module is used to divide the valid state index data into multiple state index groups, and to perform directional consistency and dimensionless processing on the state indexes within each state index group to obtain dimensionless samples of state indexes. The degradation degree variable weight calculation module is used to calculate the information entropy and degradation degree variable weight of each state index based on the evaluation batch samples at the current target evaluation time using dimensionless samples. The importance constant weight calculation module is used to construct a state index comparison and judgment matrix based on the analytic hierarchy process, and after consistency verification, it is normalized to obtain the importance constant weight that represents the inherent engineering importance of the state index. The dynamic combination weight calculation module is used to multiply the degradation degree variable weight and the importance constant weight of the same state indicator item by item, and then normalize them within the corresponding state indicator group to obtain the dynamic combination weight of the state indicators at each target evaluation time. The state reliability calculation module is used to convert the observed values of state indicators into reliability vectors of different state levels according to the pre-configured evaluation rules. The reliability vectors of all state indicators in the same state indicator group are combined to form the state reliability matrix. The comprehensive quality calculation module is used to complete the basic probability quality allocation for each state indicator group by combining the dynamic combination weights within the state indicator group with the state confidence matrix, calculate the conflict quality between evidence, perform recursive synthesis of evidence reasoning, and obtain the comprehensive quality after the fusion of all state indicators within the state indicator group.
[0008] The state determination module is used to calculate the comprehensive reliability and comprehensive ignorance of each state indicator group based on the integrated quality after fusion, determine the state level of each state indicator group according to the principle of maximum reliability, compare the state levels of the state indicator groups at adjacent evaluation times, and screen candidate state indicators to determine the dominant indicator.
[0009] Thirdly, the present invention provides an electronic device, comprising: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by at least one processor. When the computer program is executed by at least one processor, it implements the distribution network line switch status identification method based on dynamic combination weighting and evidence reasoning.
[0010] Fourthly, the present invention provides a computer storage medium storing a computer program, which, when executed by a processor, implements the method for identifying the status of distribution network line switches based on dynamic combination weighting and evidence reasoning.
[0011] The beneficial effects of this invention are as follows: 1. This invention proposes a dynamic combination weight, which combines the variable weight of the degradation degree representing the real-time degradation information of the indicator with the constant weight of the importance representing the inherent importance of the project. The weight is dynamically updated with the equipment operation data. It not only retains the engineering significance of the indicators determined by domain experts, but also amplifies the role of the currently deteriorating indicators in the evaluation, and overcomes the shortcomings of fixed weights that cannot respond to equipment degradation and the lack of physical meaning in simple objective weighting.
[0012] 2. This invention uses evidence reasoning to fuse the reliability of the four-level state, and outputs the comprehensive reliability and comprehensive ignorance of each state level. It retains the uncertainty caused by the fuzzy boundary of the indicators, the conflict of evidence, and the incompleteness of information, and avoids the information loss caused by the compression of multi-dimensional information into a single health score in traditional technology.
[0013] 3. This invention divides the indicators into three groups: switch body, secondary part, and non-operational performance. By grouping and fusing evidence, the state of the primary body of the equipment, the secondary control system, and the historical factors of the external environment can be obtained separately, so that local anomalies will not be masked by the overall results of the whole machine.
[0014] 4. The present invention introduces the comparison of adjacent time points, which not only outputs the current state of the device, but also identifies level jumps or internal changes in reliability. It combines weight and reliability to screen the dominant indicators and locate the source of degradation. Attached Figure Description
[0015] Figure 1 This is a flowchart illustrating the method of an embodiment of the present invention; Figure 2 This is a schematic diagram of the system framework according to an embodiment of the present invention; In the diagram, 10 is the data acquisition module; 20 is the preprocessing module; 30 is the degradation degree variable weight calculation module; 40 is the importance constant weight calculation module; 50 is the dynamic combination weight calculation module; 60 is the state reliability calculation module; 70 is the comprehensive quality calculation module; and 80 is the state determination module. Detailed Implementation
[0016] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings: like Figure 1 As shown, this embodiment provides a method for identifying the status of distribution network line switches based on dynamic combination weighting and evidence reasoning, including: S1: Obtain the number of status indicators of the distribution network line switch to be identified at multiple times, and preprocess the number of status indicators to obtain valid status indicator data; S2: Divide the effective state index data into multiple state index groups, and perform directional consistency and dimensionless processing on the state indexes within each state index group to obtain dimensionless samples of state indexes. S3: Based on the evaluation batch samples at the current target evaluation time, use dimensionless samples to calculate the information entropy and the degradation degree of each state index with variable weights; S4: Construct a state index comparison and judgment matrix based on the analytic hierarchy process, and normalize it after consistency verification to obtain the constant weight of importance representing the inherent engineering importance of the state index. S5: Multiply the degradation weight and importance weight of the same state indicator by one item, and then normalize them in the corresponding state indicator group to obtain the dynamic combination weight of the state indicators at each target evaluation time. S6: According to the pre-configured evaluation rules, the observed values of the state indicators are converted into confidence vectors of different state levels. The confidence vectors of all state indicators in the same state indicator group are combined to form a state confidence matrix. S7: For each state indicator group, the basic probability quality allocation is completed by combining the dynamic combination weights within the state indicator group with the state confidence matrix, and the conflict quality between evidences is calculated. The evidence reasoning recursively synthesizes the evidence to obtain the comprehensive quality after the fusion of all state indicators within the state indicator group. S8: Calculate the overall reliability and overall ignorance of each state indicator group based on the integrated quality after fusion, and determine the state level of each state indicator group according to the principle of maximum reliability; compare the state levels of the state indicator groups at adjacent evaluation times, and screen candidate state indicators to determine the dominant indicator.
[0017] In this embodiment, step S1 involves acquiring multi-time status indicator data, importance parameters, and status evaluation configuration parameters of the distribution network line switch to be identified, and preprocessing the status indicator data to obtain effective status indicator data, including: S11: Obtain the status indicator data of the distribution network line switch to be identified. The status indicator data includes the device unique identifier ID, status indicator number, data collection time timestamp, and measured value of the status indicator.
[0018] S12: Obtain importance parameters and state evaluation configuration parameters. The importance parameters include the analytic hierarchy process (AHP) judgment matrix for each state indicator, the consistency check threshold, and the random consistency index. The state evaluation configuration parameters include the indicator configuration table, the single-indicator reliability evaluation rule base, and the state level definition.
[0019] S13: Based on the device's unique identifier ID, retain the status indicator data of the current switch to be identified and remove the status indicator data of other switches; convert the time of the status indicator data of the current switch to be identified to the same standard time; S14: Perform integrity and range checks on the status index data of the current switch to be identified during calibration time to obtain valid status index data.
[0020] In this embodiment, in step S2, the valid state index data is divided into multiple state index groups, and the state indices within each state index group are subjected to direction consistency and dimensionless processing to obtain dimensionless samples of state indices, including: S21: Divide the effective status indicator data into three status indicator groups: switch body indicator group, secondary part indicator group, and non-operational efficiency indicator group, among which: The switch body index group G1 includes vacuum degree I01, relative contact wear I02, contact temperature I03, and vibration similarity I04.
[0021] The secondary indicator group G2 includes battery health performance I05, terminal online rate I06, control box temperature I07, control box humidity I08, and self-test abnormality frequency I09.
[0022] The non-operational performance indicator group G3 includes air temperature I10, atmospheric humidity I11, atmospheric pollution level I12, cumulative number of actions I13, and average number of annual failures I14.
[0023] S22: For each target evaluation time t, construct the time series data structure of the indicators. : ; In the formula, Number the status indicator groups; For the first Total number of indicators for the group; For the first Group 1 The indicator data at time t.
[0024] S23: For each state indicator group and each target evaluation time, extract the most recent value within the sliding time window. Each valid sample constitutes an evaluation batch, that is: ; In the formula, For the g-th state index group, the th The original observation sequence of a state index over k time points within a sliding window; For the first The original observations of a state index at k time points within a sliding window.
[0025] S24: Read the indicator direction type of each status indicator in the group from the indicator configuration library. The indicator direction type includes smaller values for better status, larger values for better status, the middle range for best status, and level type.
[0026] S25: Convert the state indices of "larger values indicate better state," "optimal in the middle range," and "level-based" into a form where smaller values indicate better state; the unified dimensionless representation of the "smaller values indicate better state" form is: ; In the formula, , These represent the maximum and minimum values of the batch samples used to evaluate the status indicators. For the first Item status index number A dimensionless sample, When all sample values of the condition index within the evaluation batch are equal, the dimensionless results are all set to 0.
[0027] In this embodiment, step S3, based on the evaluation batch samples at the current target evaluation time, uses dimensionless samples to calculate the information entropy and the degradation degree weighting of each state index, including: S31: Regarding the first Item status index number The proportion of information calculated for each dimensionless sample is as follows: ; In the formula, For the first Item status index number The proportion of information in a dimensionless sample.
[0028] S32: Calculation of the first based on information proportion The information entropy of the status index is used to calculate the degradation degree weighting, i.e.: ; ; In the formula, For the first Information entropy of the item's state index; For the first The degradation degree of the status index is variable-weighted; This represents the total number of valid state indicators participating in the calculation within the current state indicator group.
[0029] S33: Output the degradation degree weighted vector of the current state index group at the current target evaluation time, i.e.: ; In the formula, For the current target evaluation time t, the current state index group The degradation degree variable weight vector.
[0030] In this embodiment, step S4 involves constructing a state index comparison and judgment matrix based on the analytic hierarchy process (AHP), and then normalizing it after consistency verification to obtain the constant weights representing the inherent engineering importance of the state indexes; including:
[0031] S41: Obtain all state indicators for each state indicator group and the importance comparison scale for any two state indicators, and construct a state indicator comparison judgment matrix A based on the state indicator importance comparison scale, that is: ; In the formula, the importance scale of the diagonal elements is compared. This indicates that any two state indicators are equally important; and ; This represents the total number of state indicators in the corresponding state indicator group.
[0032] S42: Find the largest eigenvalue and the corresponding eigenvector of the state index judgment matrix; Judgment matrix of state indicators Solve the characteristic equation To obtain the largest eigenvalue And take the eigenvector corresponding to the largest eigenvalue. ; in, Represents eigenvalues; Represents the eigenvector and eigencomponents. Reflecting on the group's first The initial values for the relative importance of the status indicators.
[0033] S43: Based on the maximum eigenvalue Calculate the consistency index CI, and based on the consistency index CI and the random consistency index RI, calculate the consistency ratio CR. Then, determine whether the consistency verification is successful based on the consistency ratio CR. ; ; In this embodiment, the random consistency index RI is obtained from the AHP standard table based on the total number of state indices in the corresponding state index group; When RI=0.89; When RI = 1.12.
[0034] When the consistency ratio If consistency is not achieved, a prompt will be made to reconfigure the status indicator judgment matrix. ; The consistency check was deemed successful.
[0035] S44: Extract the eigenvector corresponding to the largest eigenvalue. Normalization yields the constant weight vector of the importance of the state indicators in the current state indicator group, i.e.: ; In the formula, For the current status indicator group status indicators Importance constant weight vector; S45: Set the current status indicator group All status indicators The current state index group is obtained by concatenating the importance constant weight vectors. The importance constant weight vector, i.e.: .
[0036] In this embodiment, step S5: Multiply the degradation degree variable weight and importance constant weight of the same state index item by item, and then normalize them within the corresponding state index group to obtain the dynamic combination weight of the state index at each target evaluation time, including: ; In the formula, For the current status indicator group The The dynamic combination weight of the status indicators at the target evaluation time t.
[0037] In this embodiment, in step S6, according to the pre-configured evaluation rules, the observed values of the state indicators are converted into confidence vectors for different state levels. The confidence vectors of all state indicators in the same state indicator group are combined to form a state confidence matrix, including: S61: Based on the pre-configured evaluation rules, the observed values of the status indicators are converted into confidence vectors for four status levels: normal, attentive, abnormal, and severe. ; In the formula, A set of state levels; These represent four status levels: normal, attention, abnormal, and serious.
[0038] S62: For the first The first of the status indicator group The confidence vectors of each indicator at evaluation time t are assembled to obtain the single-indicator confidence vector, i.e.: ; In the formula, For the first The first of the status indicator group The status information of each indicator at evaluation time t; For the first The first of the status indicator group The confidence vector of an indicator at the nth state level at evaluation time t.
[0039] S63: Combine the confidence vectors of all state indicators in the same state indicator group to form a state confidence matrix, i.e.: ; In the formula, For the first The state reliability matrix of the state index group at evaluation time t; For the first The number of status indicators in a status indicator group.
[0040] S64: Calculate the undetermined information margin corresponding to each row of the state confidence matrix. The margin generated when the sum of the confidence vectors is less than 1 is directly retained.
[0041] In this embodiment, in step S7, for each state indicator group, the basic probability quality allocation is completed by combining the dynamic combination weights within the state indicator group with the state confidence matrix, and the conflict quality between evidences is calculated. Then, recursive synthesis of evidence reasoning is performed to obtain the comprehensive quality after fusing all state indicators within the state indicator group, including: S71: For each state indicator group, combining the dynamic combination weights and the confidence vector of the single state indicator, calculate the basic probability quality assigned to a given state level, the global ignorance quality due to incomplete information, and the residual support quality generated by the relative weights, i.e.: The basic probability mass assigned to a given state level is represented as: ; In the formula, Indicates the first The first of the status indicator group Item status indicators are assigned to determine the status level The basic probability mass.
[0042] The quality of global ignorance caused by incomplete information is represented as: ; In the formula, For the first The first of the status indicator group The quality of global ignorance arises from incomplete information in single-state indicators.
[0043] The residual support quality generated by the relative weights is expressed as: ; In the formula, For the first The first of the status indicator group The residual support quality of a status index is generated by the relative weights of the status indices.
[0044] In this embodiment, the basic probabilistic quality for determining the state level, the global ignorance quality due to incomplete information, and the residual support quality generated by relative weights satisfy the mass conservation law: .
[0045] S72: Starting from the second state indicator in the state indicator group, iterate through the loop, calculate the conflict quality between the integrated evidence and the current new indicator evidence by combining the basic probability quality; determine whether the evidence is conflicting, and calculate the conflict normalization coefficient of incompletely conflicting evidence; that is: ; In the formula, For the first The first of the status indicator group The quality of conflict in the status indicators; Indicates the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. Overall quality; Indicates the first The first of the status indicator group The status index is assigned to a specific status level at evaluation time t. The basic probability mass; When conflict quality Greater than or equal to 1 minus the preset value tolerance : If the evidence is determined to be completely conflicting, the calculation for this group will be terminated and an alarm will be output.
[0046] If the evidence does not produce a complete conflict, then calculate the conflict normalization coefficient, i.e.: ; In the formula, This is the conflict normalization coefficient.
[0047] S73: Using the conflict normalization coefficient, recursively update and determine the overall quality of the hierarchy, the overall quality of global ignorance, and the overall quality of residual support; that is: The updated overall quality of the determined state level is represented as follows: ; In the formula, To indicate the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. Overall quality; For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The overall quality of global ignorance; For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The remaining support for overall quality.
[0048] In this embodiment, the updated global ignorance synthesis quality is represented as: ; In the formula, For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The overall quality of global ignorance.
[0049] In this embodiment, the updated remaining support overall quality is represented as: ; In the formula, For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The remaining support for overall quality.
[0050] S74: Iterate until all state indicators in this state indicator group have completed evidence synthesis, and obtain the overall quality of the determined state level after the fusion of all state indicators. Overall ignorance of quality Remaining support overall quality .
[0051] S75: Overall quality for determining state levels Overall ignorance of quality Remaining support overall quality Perform a mass conservation check, that is: .
[0052] In this embodiment, in step S8, the comprehensive reliability and comprehensive ignorance of each state indicator group are calculated based on the fused comprehensive quality, and the state level of each state indicator group is determined according to the maximum reliability principle; the state levels of the state indicator groups at adjacent evaluation times are compared, and candidate state indicators are screened to determine the dominant indicator, including: S81: Determining the overall quality of the state level after the fusion of all state indicators. Overall ignorance of quality Remaining support overall quality Calculate the overall reliability and overall ignorance of each state index group, i.e.: In this embodiment, the comprehensive reliability is calculated based on the comprehensive quality of the determined state level after all state indicators have been fused and the comprehensive quality of the remaining support, and is expressed as: ; In the formula, For the first The state indicator group evaluates the state level at time t. Overall reliability; To determine the overall quality of the status level; To support the overall quality of the remaining support.
[0053] In this embodiment, the comprehensive ignorance degree is calculated based on the comprehensive global ignorance quality and the comprehensive residual support quality after all state indicators have been fused, and is expressed as: ; In the formula, For the first The overall ignorance of the state indicator group at evaluation time t; The overall quality is based on a lack of global awareness.
[0054] In this embodiment, the comprehensive confidence and comprehensive ignorance satisfy: .
[0055] S82: The state level of this state indicator group is determined using the maximum reliability principle; that is: Compare the overall reliability of the four status levels (normal, attention, abnormal, and severe), and take the status level corresponding to the highest overall reliability as the status level of this status index group at the current moment. ; If the combined confidence level of multiple state levels is the same, select the state level with the more severe degradation and record the uncertainty log.
[0056] S83: Obtain historical data from the previous evaluation time t-1, including state level, comprehensive reliability of each state indicator, and dynamic combination weight; S84: If the state level of the previous evaluation time t-1 is different from that of the current evaluation time, then... If the state level is the same at the previous evaluation time t-1 and the current evaluation time, but the overall reliability change of the state level exceeds the preset threshold, then the level remains unchanged - internal reliability change; otherwise, the level is not significantly changed. S85: Using non-zero confidence support for the current judgment state level as the screening criterion, a set of candidate indicators is obtained; including: If the corresponding state index at the current evaluation time corresponds to the state level The overall reliability is greater than the preset numerical tolerance, that is... If so, the corresponding status indicator is added to the candidate indicator set; S86: Select the indicator with the largest dynamic combination weight at the current evaluation time from the candidate indicator set as the dominant indicator; if there is a change in the state level, prioritize the candidate indicator whose comprehensive reliability shifts to the current state level and whose dynamic combination weight increases.
[0057] like Figure 2 As shown, embodiments of this application provide a distribution network line switch status identification system based on dynamic combination weighting and evidence reasoning, including a data acquisition module 10, a preprocessing module 20, a deterioration degree variable weight calculation module 30, an importance degree constant weight calculation module 40, a dynamic combination weight calculation module 50, a status reliability calculation module 60, a comprehensive quality calculation module 70, and a status determination module 80.
[0058] In this embodiment, the data acquisition module 10 is configured to acquire the number of status indicators of the distribution network line switch to be identified at multiple times, and preprocess the number of status indicators to obtain valid status indicator data.
[0059] In this embodiment, the data acquisition module 10 acquires the status indicator data of the distribution network line switch to be identified. The status indicator data includes the device unique identifier ID, status indicator number, acquisition time timestamp, and measured value of the status indicator.
[0060] In this embodiment, the data acquisition module 10 acquires importance parameters and state evaluation configuration parameters. The importance parameters include the analytic hierarchy process (AHP) judgment matrix for each state indicator, the consistency check threshold, and the random consistency index. The state evaluation configuration parameters include an indicator configuration table, a single-indicator reliability evaluation rule base, and state level definitions.
[0061] In this embodiment, the data acquisition module 10 retains the status indicator data of the current switch to be identified based on the device's unique identifier ID, and removes the status indicator data of other switching devices; it also converts the time of the status indicator data of the current switch to be identified to the same standard time. In this embodiment, the data acquisition module 10 performs integrity verification and numerical range verification on the status index data of the current switch to be identified during calibration time to obtain valid status index data.
[0062] In this embodiment, the preprocessing module 20 is configured to divide the valid state index data into multiple state index groups, and perform directional consistency and dimensionless processing on the state indexes within each state index group to obtain dimensionless samples of the state indexes.
[0063] In this embodiment, the preprocessing module 20 divides the effective status index data into three status index groups: the switch body index group, the secondary part index group, and the non-operational performance index group, wherein: The switch body index group G1 includes vacuum degree I01, relative contact wear I02, contact temperature I03, and vibration similarity I04.
[0064] The secondary indicator group G2 includes battery health performance I05, terminal online rate I06, control box temperature I07, control box humidity I08, and self-test abnormality frequency I09.
[0065] The non-operational performance indicator group G3 includes air temperature I10, atmospheric humidity I11, atmospheric pollution level I12, cumulative number of actions I13, and average number of annual failures I14.
[0066] In this embodiment, the preprocessing module 20 constructs an index time series data structure for each target evaluation time t. : ; In the formula, Number the status indicator groups; For the first Total number of indicators for the group; For the first Group 1 The indicator data at time t.
[0067] In this embodiment, the preprocessing module 20 extracts the most recent value within the sliding time window for each state indicator group and each target evaluation time. Each valid sample constitutes an evaluation batch, that is: ; In the formula, For the g-th state index group, the th The original observation sequence of a state index over k time points within a sliding window; For the first The original observations of a state index at k time points within a sliding window.
[0068] In this embodiment, the preprocessing module 20 reads the indicator direction type of each status indicator in the group from the indicator configuration library. The indicator direction type includes the following: the smaller the value, the better the status; the larger the value, the better the status; the middle range is the best; and the level type.
[0069] In this embodiment, the preprocessing module 20 converts the state indices of "larger values indicate better state," "intermediate range indicates best state," and "level-based state indices" into a form where "smaller values indicate better state." The unified dimensionless representation of the "smaller values indicate better state" form is as follows: ; In the formula, , These represent the maximum and minimum values of the batch samples used to evaluate the status indicators. For the first Item status index number A dimensionless sample, When all sample values of the condition index within the evaluation batch are equal, the dimensionless results are all set to 0.
[0070] In this embodiment, the degradation degree variable weight calculation module 30 is configured to calculate the information entropy and degradation degree variable weight of each state index based on the evaluation batch samples at the current target evaluation time using dimensionless samples.
[0071] In this embodiment, the degradation degree variable weight calculation module 30 calculates the degradation degree of the first... Item status index number The proportion of information calculated for each dimensionless sample is as follows: ; In the formula, For the first Item status index number The proportion of information in a dimensionless sample.
[0072] In this embodiment, the degradation degree variable weight calculation module 30 calculates the first degradation degree based on the information proportion. The information entropy of the status index is used to calculate the degradation degree weighting, i.e.: ; ; In the formula, For the first Information entropy of the item's state index; For the first The degradation degree of the status index is variable-weighted; This represents the total number of valid state indicators participating in the calculation within the current state indicator group.
[0073] In this embodiment, the degradation degree variable weight calculation module 30 outputs the degradation degree variable weight vector of the current state index group at the current target evaluation time, that is: ; In the formula, For the current target evaluation time t, the current state index group The degradation degree variable weight vector.
[0074] In this embodiment, the importance constant weight calculation module 40 is configured to construct a state index comparison and judgment matrix based on the analytic hierarchy process, and then normalize it after consistency verification to obtain the importance constant weight that represents the inherent engineering importance of the state index.
[0075] In this embodiment, the step involves obtaining all state indicators for each state indicator group and the importance comparison scale for any two state indicators, and constructing a state indicator comparison judgment matrix A based on the state indicator importance comparison scale, i.e.: ; In the formula, the importance scale of the diagonal elements is compared. This indicates that any two state indicators are equally important; and ; This represents the total number of state indicators in the corresponding state indicator group.
[0076] In this embodiment, the solution involves finding the maximum eigenvalue and the corresponding eigenvector of the state index judgment matrix; Judgment matrix of state indicators Solve the characteristic equation To obtain the largest eigenvalue And take the eigenvector corresponding to the largest eigenvalue. ; in, Represents eigenvalues; Represents the eigenvector and eigencomponents. Reflecting on the group's first The initial values for the relative importance of the status indicators.
[0077] In this embodiment, the step based on the maximum eigenvalue... Calculate the consistency index CI, and based on the consistency index CI and the random consistency index RI, calculate the consistency ratio CR. Then, determine whether the consistency verification is successful based on the consistency ratio CR. ; ; In this embodiment, the random consistency index RI is obtained from the AHP standard table based on the total number of state indices in the corresponding state index group; When RI=0.89; When RI = 1.12.
[0078] When the consistency ratio If consistency is not achieved, a prompt will be made to reconfigure the status indicator judgment matrix. ; The consistency check was deemed successful.
[0079] In this embodiment, the eigenvector corresponding to the largest eigenvalue is... Normalization yields the constant weight vector of the importance of the state indicators in the current state indicator group, i.e.: ; In the formula, For the current status indicator group status indicators Importance constant weight vector; In this embodiment, the current status indicator group is... All status indicators The current state index group is obtained by concatenating the importance constant weight vectors. The importance constant weight vector, i.e.: .
[0080] In this embodiment, the dynamic combination weight calculation module 50 is configured to multiply the degradation degree variable weight and the importance constant weight of the same state indicator item by item, and then normalize them within the corresponding state indicator group to obtain the dynamic combination weight of the state indicators at each target evaluation time, that is: ; In the formula, For the current status indicator group The The dynamic combination weight of the status indicators at the target evaluation time t.
[0081] In this embodiment, the state reliability calculation module 60 is configured to convert state index observations into reliability vectors of different state levels according to pre-configured evaluation rules, and the reliability vectors of all state indices in the same state index group are combined to form a state reliability matrix.
[0082] In this embodiment, the state reliability calculation module 60 converts the observed state indicators into reliability vectors for four state levels—normal, attentive, abnormal, and severe—according to pre-configured evaluation rules, i.e.: ; In the formula, A set of state levels; These represent four status levels: normal, attention, abnormal, and serious.
[0083] In this embodiment, the state confidence calculation module 60 calculates the state confidence score for the first... The first of the status indicator group The confidence vectors of each indicator at evaluation time t are assembled to obtain the single-indicator confidence vector, i.e.: ; In the formula, For the first The first of the status indicator group The status information of each indicator at evaluation time t; For the first The first of the status indicator group The confidence vector of an indicator at the nth state level at evaluation time t.
[0084] In this embodiment, the state reliability calculation module 60 combines the reliability vectors of all state indicators in the same state indicator group to form a state reliability matrix, that is: ; In the formula, For the first The state reliability matrix of the state index group at evaluation time t; For the first The number of status indicators in a status indicator group.
[0085] In this embodiment, the state reliability calculation module 60 calculates the undetermined information margin corresponding to each row of the state reliability matrix and directly retains the margin generated when the sum of the reliability vectors is less than 1.
[0086] In this embodiment, the comprehensive quality calculation module 70 is configured to perform basic probability quality allocation for each state indicator group by combining the dynamic combination weights within the state indicator group with the state confidence matrix, calculate the conflict quality between evidence, perform recursive synthesis of evidence reasoning, and obtain the comprehensive quality after the fusion of all state indicators within the state indicator group.
[0087] In this embodiment, the comprehensive quality calculation module 70, for each group of state indicators, combines the dynamic combination weights and the confidence vector of a single state indicator to calculate the basic probability quality assigned to a given state level, the global ignorance quality due to incomplete information, and the residual support quality generated by the relative weights, namely: The basic probability mass assigned to a given state level is represented as: ; In the formula, Indicates the first The first of the status indicator group Item status indicators are assigned to determine the status level The basic probability mass.
[0088] The quality of global ignorance caused by incomplete information is represented as: ; In the formula, For the first The first of the status indicator group The quality of global ignorance arises from incomplete information in single-state indicators.
[0089] The residual support quality generated by the relative weights is expressed as: ; In the formula, For the first The first of the status indicator group The residual support quality of a status index is generated by the relative weights of the status indices.
[0090] In this embodiment, the basic probabilistic quality for determining the state level, the global ignorance quality due to incomplete information, and the residual support quality generated by relative weights satisfy the mass conservation law: .
[0091] In this embodiment, the comprehensive quality calculation module 70 iterates from the second state indicator in the state indicator group, calculates the conflict quality between the integrated evidence and the current new indicator evidence by combining the basic probability quality calculation; and determines whether the evidence is conflicting, and calculates the conflict normalization coefficient of the incompletely conflicting evidence; that is: ; In the formula, For the first The first of the status indicator group The quality of conflict in the status indicators; Indicates the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. Overall quality; Indicates the first The first of the status indicator group The status index is assigned to a specific status level at evaluation time t. The basic probability mass; When conflict quality Greater than or equal to 1 minus the preset value tolerance : If the evidence is determined to be completely conflicting, the calculation for this group will be terminated and an alarm will be output.
[0092] If the evidence does not produce a complete conflict, then calculate the conflict normalization coefficient, i.e.: ; In the formula, This is the conflict normalization coefficient.
[0093] In this embodiment, the comprehensive quality calculation module 70 uses conflict normalization coefficients to recursively update and determine the overall quality of the grade, the overall quality of global ignorance, and the overall quality of remaining support; that is: The updated overall quality of the determined state level is represented as follows: ; In the formula, To indicate the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. Overall quality; For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The overall quality of global ignorance; For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The remaining support for overall quality.
[0094] In this embodiment, the updated global ignorance synthesis quality is represented as: ; In the formula, For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The overall quality of global ignorance.
[0095] In this embodiment, the updated remaining support overall quality is represented as: ; In the formula, For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The remaining support for overall quality.
[0096] In this embodiment, the comprehensive quality calculation module 70 iterates until all state indicators of the current state indicator group have completed evidence synthesis, thus obtaining the comprehensive quality of the determined state level after the fusion of all state indicators. Overall ignorance of quality Remaining support overall quality .
[0097] Comprehensive quality for determining state level Overall ignorance of quality Remaining support overall quality Perform a mass conservation check, that is: .
[0098] In this embodiment, the state determination module 80 is configured to calculate the comprehensive reliability and comprehensive ignorance of each state indicator group based on the integrated quality after fusion, determine the state level of each state indicator group according to the principle of maximum reliability, compare the state levels of the state indicator groups at adjacent evaluation times, and screen candidate state indicators to determine the dominant indicator.
[0099] In this embodiment, the state determination module 80 determines the overall quality of the state level based on the fusion of all state indicators. Overall ignorance of quality Remaining support overall quality Calculate the overall reliability and overall ignorance of each state index group, i.e.: In this embodiment, the comprehensive reliability is calculated based on the comprehensive quality of the determined state level after all state indicators have been fused and the comprehensive quality of the remaining support, and is expressed as: ; In the formula, For the first The state indicator group evaluates the state level at time t. Overall reliability; To determine the overall quality of the status level; To support the overall quality of the remaining support.
[0100] In this embodiment, the comprehensive ignorance degree is calculated based on the comprehensive global ignorance quality and the comprehensive residual support quality after all state indicators have been fused, and is expressed as: ; In the formula, For the first The overall ignorance of the state indicator group at evaluation time t; The overall quality is based on a lack of global awareness.
[0101] In this embodiment, the comprehensive confidence and comprehensive ignorance satisfy: .
[0102] In this embodiment, the state determination module 80 uses the maximum confidence principle to determine the state level of this state index group; that is: Compare the overall reliability of the four status levels (normal, attention, abnormal, and severe), and take the status level corresponding to the highest overall reliability as the status level of this status index group at the current moment. ; If the combined confidence level of multiple state levels is the same, select the state level with the more severe degradation and record the uncertainty log.
[0103] In this embodiment, the state determination module 80 acquires historical data from the previous evaluation time t-1, including the state level, the comprehensive reliability of each state indicator, and the dynamic combination weight. In this embodiment, if the state determination module 80 determines that the state level of the previous evaluation time t-1 is different from the current evaluation time, If the state level is the same at the previous evaluation time t-1 and the current evaluation time, but the overall reliability change of the state level exceeds the preset threshold, then the level remains unchanged - internal reliability change; otherwise, the level is not significantly changed.
[0104] In this embodiment, the state determination module 80 uses non-zero confidence support for the current determined state level as a screening condition to obtain a set of candidate indicators, including: If the corresponding state index at the current evaluation time corresponds to the state level The overall reliability is greater than the preset numerical tolerance, that is... If so, the corresponding status indicator is added to the candidate indicator set; In this embodiment, the state determination module 80 selects the indicator with the largest dynamic combination weight at the current evaluation time from the candidate indicator set as the dominant indicator; if there is a change in state level, the candidate indicator whose comprehensive reliability shifts to the current state level and whose dynamic combination weight increases is selected first.
[0105] A third embodiment of the present invention provides an electronic device, comprising: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by at least one processor. When the computer program is executed by at least one processor, it implements the distribution network line switch status identification method based on dynamic combination weighting and evidence reasoning.
[0106] The fourth embodiment of the present invention provides a computer storage medium storing a computer program, which, when executed by a processor, implements the distribution network line switch status identification method based on dynamic combination weighting and evidence reasoning.
[0107] The embodiments and descriptions above are merely illustrative of the principles and preferred embodiments of the present invention. Various changes and modifications may be made to the present invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed.
Claims
1. A method for identifying the state of a switch on a distribution line based on dynamic combination weighting and evidence reasoning, characterized in that, include: The number of status indicators of the distribution network line switch to be identified at multiple times is obtained, and the number of status indicators is preprocessed to obtain effective status indicator data. The effective state index data is divided into multiple state index groups, and the state indexes within each state index group are subjected to direction consistency and dimensionless processing to obtain dimensionless samples of state indexes. Based on the evaluation batch samples at the current target evaluation time, the information entropy and the degradation degree of each state index are calculated using dimensionless samples. A state index comparison and judgment matrix is constructed based on the analytic hierarchy process. After consistency verification, the matrix is normalized to obtain the constant weights of the importance of the inherent engineering importance of the state index. The deterioration degree variable weight and importance constant weight of the same state indicator are multiplied one by one, and then normalized within the corresponding state indicator group to obtain the dynamic combination weight of the state indicators at each target evaluation time. According to the pre-configured evaluation rules, the observed values of the state indicators are converted into confidence vectors of different state levels, and the confidence vectors of all state indicators in the same state indicator group are combined to form a state confidence matrix. For each state indicator group, the basic probability quality allocation is completed by combining the dynamic combination weights within the state indicator group with the state confidence matrix, and the conflict quality between evidences is calculated. The evidence reasoning recursively synthesizes the evidence to obtain the comprehensive quality after the fusion of all state indicators within the state indicator group. Based on the integrated quality calculation, the comprehensive reliability and comprehensive ignorance of each state indicator group are calculated, and the state level of each state indicator group is determined according to the principle of maximum reliability. The state levels of the state indicator groups at adjacent evaluation times are compared, and candidate state indicators are selected to determine the dominant indicator.
2. The method according to claim 1, characterized in that: The effective status index data is divided into three status index groups: switch body index group, secondary part index group, and non-operational efficiency index group. The switch body index group G1 includes vacuum degree I01, relative contact wear I02, contact temperature I03, and vibration similarity I04; The secondary component index group G2 includes battery health performance I05, terminal online rate I06, control box temperature I07, control box humidity I08, and self-test abnormality frequency I09. The non-operational performance indicator group G3 includes air temperature I10, atmospheric humidity I11, atmospheric pollution level I12, cumulative number of actions I13, and average number of annual failures I14.
3. The method according to claim 1, characterized in that: Based on the evaluation batch samples at the current target evaluation time, the information entropy and degradation degree weighting of each state index are calculated using dimensionless samples, including: The proportion of information calculated from dimensionless samples based on state indicators; Information entropy of state indicators is calculated based on information proportion, and degradation degree is calculated based on information entropy; By concatenating the degradation weights of all state indicators, we obtain the degradation weight vector of the current state indicator group at the current target evaluation time.
4. The method according to claim 1, characterized in that: A state index comparison and judgment matrix is constructed based on the analytic hierarchy process (AHP). After consistency verification and normalization, the importance constant weights representing the inherent engineering importance of the state indexes are obtained, including: Obtain all state indicators for each state indicator group and the importance comparison scale for any two state indicators, and construct a state indicator comparison judgment matrix based on the state indicator importance comparison scale. Find the largest eigenvalue and the corresponding eigenvector of the state index judgment matrix; The consistency index is calculated based on the largest eigenvalue, and the consistency ratio is calculated based on the consistency index and the random consistency index. The consistency ratio is then used to determine whether the consistency verification is successful. Normalize the eigenvector corresponding to the largest eigenvalue to obtain the constant weight vector of the importance of the state indicators in the current state indicator group; The importance constant weight vector of the current state indicator group is obtained by concatenating the importance constant weight vectors of all state indicators in the current state indicator group.
5. The method according to claim 1, characterized in that: According to the pre-configured evaluation rules, the observed values of state indicators are converted into reliability vectors for different state levels. The reliability vectors of all state indicators in the same state indicator group are combined to form a state reliability matrix, including: Based on the pre-configured evaluation rules, the observed values of the status indicators are converted into confidence vectors for four status levels: normal, attentive, abnormal, and severe. The confidence vector of a single indicator is obtained by assembling the confidence vectors of the state indicators of the state indicator group at the evaluation time. The confidence vectors of all state indicators in the same state indicator group are combined to form a state confidence matrix; Calculate the undetermined information margin corresponding to each row of the state confidence matrix. The margin generated when the sum of the confidence vectors is less than 1 is directly retained.
6. The method according to claim 1, characterized in that: For each state indicator group, a basic probability quality allocation is completed by combining the dynamic combination weights within the state indicator group with the state confidence matrix. The quality of conflict between pieces of evidence is then calculated, and recursive synthesis of the evidence reasoning is performed to obtain the comprehensive quality after fusing all state indicators within the state indicator group, including: For each state indicator group, the basic probability quality assigned to a given state level, the global ignorance quality due to incomplete information, and the residual support quality generated by relative weights are calculated by combining the dynamic combination weights and the confidence vector of a single state indicator. Starting from the second state indicator in the state indicator group, the iterative loop is used to calculate the conflict quality between the integrated evidence and the current new indicator evidence, combined with the basic probability quality; and it is determined whether the evidence is conflicting, and the conflict normalization coefficient of the incompletely conflicting evidence is calculated. Using the conflict normalization coefficient, the overall quality of the grade, the overall quality of global ignorance, and the overall quality of residual support are recursively updated and determined. The iteration loop continues until all state indicators in this state indicator group have completed evidence synthesis, resulting in the overall quality of the determined state level, the overall quality of global ignorance, and the overall quality of remaining support after all state indicators have been fused.
7. The method according to claim 6, characterized in that: The updated overall quality representation of the determined state level is as follows: ; In the formula, To indicate the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. Overall quality; Here is the conflict normalization coefficient; Indicates the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. Overall quality; Indicates the first The first of the status indicator group The status index is assigned to a specific status level at evaluation time t. The basic probability mass; For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The overall quality of global ignorance; For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The remaining support overall quality; For the first The first of the status indicator group The quality of global ignorance arising from incomplete information in single-state indicators; For the first The first of the status indicator group The residual support quality of the status index is generated by the relative weights of the status index; The updated global ignorance comprehensive quality representation is as follows: ; In the formula, For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The overall quality of global ignorance; The updated remaining support overall quality is expressed as follows: ; In the formula, For the first The state indicator group before evaluation time t After the status indicators are integrated, the status level is determined. The remaining support for overall quality.
8. The method according to claim 7, characterized in that: Based on the integrated quality calculation of each state indicator group, the comprehensive reliability and comprehensive ignorance are calculated, and the state level of each state indicator group is determined according to the principle of maximum reliability. Compare the state levels of state indicator groups at adjacent evaluation times, and screen candidate state indicators to determine the dominant indicator, including: Based on the comprehensive quality of the determined state level, the comprehensive quality of global ignorance, and the comprehensive quality of remaining support after all state indicators have been fused, the comprehensive reliability and comprehensive ignorance of each state indicator group are calculated. The state level of this state indicator group is determined using the maximum reliability principle; Obtain historical data from the previous evaluation time t-1, including state level, overall reliability of each state indicator, and dynamic combination weight; If the state level of the previous evaluation time is different from that of the current evaluation time, it is marked as a level jump change; if the state level of the previous evaluation time is the same as that of the current evaluation time, but there is a change in the overall reliability of the state level that exceeds the preset threshold, it is marked as no change in level - internal reliability change; otherwise, it is marked as no significant change. The candidate indicator set is obtained by using the non-zero confidence level of the current judgment status as the screening condition; The indicator with the largest dynamic combination weight at the current evaluation time is selected from the candidate indicator set as the dominant indicator; if there is a change in the state level, the candidate indicator whose comprehensive reliability shifts to the current state level and whose dynamic combination weight increases is given priority.
9. A distribution network line switch status identification system based on dynamic combination weighting and evidence reasoning, characterized in that, include: The data acquisition module is used to acquire the number of status indicators of the distribution network line switch to be identified at multiple times, and to preprocess the number of status indicators to obtain valid status indicator data. The preprocessing module is used to divide the valid state index data into multiple state index groups, and to perform directional consistency and dimensionless processing on the state indexes within each state index group to obtain dimensionless samples of state indexes. The degradation degree variable weight calculation module is used to calculate the information entropy and degradation degree variable weight of each state index based on the evaluation batch samples at the current target evaluation time using dimensionless samples. The importance constant weight calculation module is used to construct a state index comparison and judgment matrix based on the analytic hierarchy process, and after consistency verification, it is normalized to obtain the importance constant weight that represents the inherent engineering importance of the state index. The dynamic combination weight calculation module is used to multiply the degradation degree variable weight and the importance constant weight of the same state indicator item by item, and then normalize them within the corresponding state indicator group to obtain the dynamic combination weight of the state indicators at each target evaluation time. The state reliability calculation module is used to convert the observed values of state indicators into reliability vectors of different state levels according to the pre-configured evaluation rules. The reliability vectors of all state indicators in the same state indicator group are combined to form the state reliability matrix. The comprehensive quality calculation module is used to complete the basic probability quality allocation for each state indicator group by combining the dynamic combination weights within the state indicator group with the state confidence matrix, calculate the conflict quality between evidence, perform recursive synthesis of evidence reasoning, and obtain the comprehensive quality after the fusion of all state indicators within the state indicator group. The state determination module is used to calculate the comprehensive reliability and comprehensive ignorance of each state indicator group based on the integrated quality after fusion, determine the state level of each state indicator group according to the principle of maximum reliability, compare the state levels of the state indicator groups at adjacent evaluation times, and screen candidate state indicators to determine the dominant indicator.
10. An electronic device, comprising: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by at least one processor, characterized in that, when the computer program is executed by at least one processor, it implements the distribution network line switch status identification method based on dynamic combination weighting and evidence reasoning as described in any one of claims 1-8.
