A cable line state assessment and maintenance decision method and system

Abstract

This invention discloses a method and system for cable line condition assessment and maintenance decision-making, relating to the field of power system technology. The method includes: acquiring cable condition data and determining missing data information; the missing data information includes the location and type of the missing data; based on the missing data information, employing a corresponding data completion strategy to complete the cable condition data, obtaining completed values ​​and corresponding uncertainty indicators; inputting the completed values ​​and corresponding uncertainty indicators into a pre-constructed fault prediction model to obtain the cable's fault probability and confidence interval within a specified future time window; and determining the optimal maintenance strategy based on the fault probability. This invention enables accurate assessment and scientific decision-making regarding cable line conditions even with incomplete information.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method and system for cable line condition assessment and maintenance decision-making. Background Technology

[0002] As a critical infrastructure for power transmission, cable lines have a complex structure, mainly composed of conductors, insulation layers, and sheaths. Since 2008, the State Grid Corporation of China has used condition assessment as the basis for condition-based maintenance, and related methods have been gradually applied to condition assessment of equipment such as power transformers and circuit breakers. However, in the case of cable lines, due to problems such as monitoring equipment failures, data transmission delays, difficulty in real-time acquisition of some parameters, and missing historical data, the monitoring data naturally possesses the characteristics of being "incomplete and unbalanced."

[0003] Existing cable condition assessment methods typically rely on comprehensive data support, but in actual operation, data gaps often arise due to reasons such as tests not being conducted or monitoring interruptions. To address these situations, current technologies often employ simple methods such as "sample removal" or "fixed value imputation," which frequently lead to distorted assessment results and fail to accurately reflect the actual condition of the cables. Furthermore, current condition assessment results are mostly qualitative descriptions, lacking quantitative analysis of fault probability and risk, thus failing to provide accurate basis for subsequent maintenance decisions.

[0004] In terms of data completion, existing methods are relatively simplistic. For example, commonly used methods such as mean imputation do not fully consider the characteristics of the cable equipment itself, such as the coupling relationship between insulation aging and multiple state quantities, leading to large completion errors. Furthermore, there is currently a lack of a closed-loop mechanism from data completion to condition assessment and maintenance decision-making. Errors generated during the completion process are propagated step by step to the decision-making stage, potentially leading to unnecessary increases in maintenance costs or missed fault risks. Therefore, how to achieve accurate assessment and scientific decision-making of cable conditions under incomplete information has become a pressing technical problem to be solved in the current cable operation and maintenance field. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method and system for cable line condition assessment and maintenance decision-making, which can achieve accurate assessment and scientific decision-making on cable line condition under conditions of incomplete information.

[0006] To achieve the above objectives, the present invention provides a method for cable line condition assessment and maintenance decision-making, comprising: Acquire cable status data and determine any missing data information in the cable status data; the missing data information includes the location and type of missing data. Based on the missing data information, a corresponding data completion strategy is used to complete the cable status data to obtain the completed value and the corresponding uncertainty index. The completed value and the corresponding uncertainty index are input into the pre-built fault prediction model to obtain the fault probability and confidence interval of the cable within a specified time window in the future. Based on the failure probability, the optimal maintenance strategy is determined.

[0007] Optionally, the missing type includes at least one of random missing, fixed missing, and complete missing.

[0008] Optionally, the adoption of a corresponding data completion strategy includes: For the aforementioned random missing values, the median of the same state level is used for filling. For the fixed missing values, Bayesian inference is used to obtain the imputation values ​​and standard deviations; For the complete absence, an industry standard value is used as the completion value, and a priori coefficient is introduced to characterize the credibility of the industry standard value.

[0009] Optionally, the fault prediction model is specifically an improved Cox proportional hazards model, and the risk function of the improved Cox proportional hazards model is: in, Represents the risk function, The benchmark risk rate; For the first A state vector of each influencing factor; For the first The regression coefficients of the influencing factors; The total number of influencing factors; Weights for missing labels; A marker variable to indicate whether data is missing; Here is the uncertainty coefficient; To fix the missing data and complete the standard deviation.

[0010] Optionally, based on the failure probability, an optimal maintenance strategy is determined, including: Based on the aforementioned failure probability and the preset failure consequence cost, calculate the risk value before maintenance; Generate or invoke multiple maintenance plans, and estimate the implementation cost and post-implementation risk value of each maintenance plan; Calculate the net benefit of each maintenance plan based on the risk value before maintenance, the implementation cost, and the risk value after maintenance. The optimal maintenance strategy is to select the maintenance plan with the greatest net benefit.

[0011] Optionally, the formula for calculating the risk value before maintenance is: in, The risk value before maintenance. Specify a time window for the future Internal failure probability, For direct failure costs, For indirect failure costs, Let be the average load power of the i-th type of user. Let be the unit power outage loss coefficient for user type i. Total number of user categories This indicates the expected duration of the power outage.

[0012] Optionally, the formula for calculating the net benefit is: in, For the first One maintenance plan Net benefits The risk value before maintenance. Maintenance plan Risk values ​​after implementation; Maintenance plan The effective period, Maintenance plan The implementation cost.

[0013] This invention also provides a cable line condition assessment and maintenance decision-making system, comprising: A data acquisition and missing data identification unit is used to acquire cable status data and determine the missing data information of the cable status data; the missing data information includes the missing location and the missing type. The data completion unit is used to complete the cable status data based on the missing data information and using the corresponding data completion strategy to obtain the completed value and the corresponding uncertainty index. The fault probability prediction unit is used to input the completed value and the corresponding uncertainty index into the pre-built fault prediction model to obtain the fault probability and confidence interval of the cable in a specified time window in the future. The maintenance decision output unit is used to determine the optimal maintenance strategy based on the fault probability.

[0014] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: The cable line condition assessment and maintenance decision-making method provided by this invention acquires cable condition data and identifies missing data information, accurately identifying incomplete and unbalanced data characteristics. This lays the foundation for subsequent targeted processing and solves the problem of distorted assessment results caused by "sample removal" or "fixed value imputation" in existing technologies when data is missing (such as when tests are not conducted or monitoring is interrupted). Based on the missing data information, a corresponding data completion strategy is used to complete the data, outputting the completed value and corresponding uncertainty indicators. Differentiated completion methods are used for different types of missing data, such as random missing data, fixed missing data, and complete missing data. Considering the characteristics of cable equipment (such as the coupling relationship between insulation aging and multiple state quantities), this method solves the problems of existing data completion methods being singular and having large completion errors, reducing error sources. The completed value and uncertainty indicators are input into a pre-constructed fault prediction model to obtain the cable's fault probability and confidence interval within a specified future time window. This transforms traditional qualitative condition levels into quantitative fault probability and risk values ​​with confidence intervals, providing a precise basis for subsequent maintenance decisions and overcoming the shortcomings of existing technologies where condition assessment results are mostly qualitative descriptions and lack quantitative analysis of fault occurrence probability and risk. The optimal maintenance strategy is determined based on the failure probability, and a closed-loop mechanism is constructed from data completion to condition assessment and maintenance decision-making. The model output is corrected by using uncertainty indicators to avoid the completion error being passed on to the decision-making stage. This solves the technical problem of wasted operation and maintenance costs or missed failure risk due to the lack of a closed-loop mechanism in the existing technology, and improves the scientificity and practicality of cable condition assessment and operation and maintenance decision-making under incomplete information. Attached Figure Description

[0015] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same parts in the exemplary embodiments of the invention.

[0016] Figure 1 This is a schematic diagram of the method flow for cable line condition assessment and maintenance decision-making in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the cable line status data processing and adaptive evaluation decision-making process according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the module structure of the cable line condition assessment and maintenance decision-making system according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 and Figure 2 , Figure 1 This is a flowchart illustrating the methodology for cable line condition assessment and maintenance decision-making. Figure 2 This is a schematic diagram of the cable line status data processing and adaptive evaluation decision-making process.

[0019] Methods for cable line condition assessment and maintenance decision-making include: S101: Obtain cable status data and determine any missing data information for the cable status data.

[0020] The missing data information includes the location of the missing data and the type of missing data; the type of missing data includes at least one of random missing data, fixed missing data, and complete missing data.

[0021] In applications, the first step is to acquire cable condition data during the operation of the cable line, such as insulation resistance, partial discharge, grounding current, and service life, among other multi-dimensional parameters. These parameters form the basis for assessing cable health. After data acquisition, due to temporary equipment failures, data transmission delays, difficulties in real-time acquisition of some parameters, or missing historical data, the obtained cable condition data often inherently possesses characteristics of being "incomplete and unbalanced." Therefore, this raw data needs to be preprocessed, specifically including noise filtering and outlier removal to retain valid data.

[0022] Next, the preprocessed cable status data undergoes missing data detection and identification. This involves not only marking the locations of missing data but, more importantly, identifying and determining the type of missing data—that is, fine-grained classification of missing data information. Based on the cause and pattern of the missing data, it can be mainly divided into three types: random missing data, such as single-point data loss due to temporary sensor malfunctions; fixed missing data, such as long-term systematic data loss due to the lack of specific monitoring devices installed on a batch of cables; and complete missing data, such as overall data gaps caused by the absence of any historical data for newly commissioned cables. This accurate identification and determination of missing data types (random missing data, fixed missing data, and complete missing data) serves as the basis for subsequent differentiated processing strategies.

[0023] S102: Based on the missing data information, the corresponding data completion strategy is used to complete the cable status data, and the completed values ​​and corresponding uncertainty indicators are obtained.

[0024] Traditional methods, when faced with missing data, often employ single approaches such as sample removal or fixed-value imputation, without considering the underlying causes of the missing information. This leads to significant imputation errors and distorted evaluation results. To address this issue, this invention proposes a multi-strategy adaptive data imputation mechanism. This mechanism dynamically selects and executes the optimal imputation algorithm based on the specific missing type determined earlier, thereby achieving high-precision repair of incomplete information.

[0025] Specifically, the output of this step consists of two parts: first, the complete dataset of cable state variables after completion, i.e., the completed values; and second, the uncertainty index generated with each completed value. This uncertainty index is a quantitative representation of the error or confidence level that may be introduced by the completion operation, and can be expressed as the standard deviation of the statistical distribution, the width of the confidence interval, or a confidence coefficient. By introducing the uncertainty index, the subsequent fault prediction model can recognize and quantify the uncertainty introduced by data completion, rather than using the completed values ​​as completely accurate true values, thus fundamentally preventing the hidden transmission of completion errors.

[0026] For example, the corresponding data completion strategy adopted above includes: For random missing values, the median of the same state level is used for imputation; For fixed missing values, Bayesian inference is used to obtain the imputation values ​​and standard deviations; For complete missing values, industry standard values ​​are used as the filler values, and a priori coefficient is introduced to characterize the reliability of the industry standard values.

[0027] In applications, for random missing data, the completion strategy focuses on using the historical data distribution characteristics of similar devices under the same health condition to fill in the missing data. For example, the median of data at the same health level is used, and a missing marker is generated simultaneously to record that the data at that position has been completed. 补全 =median(X 同等级样本 ), M=1 (missing label); where M is the missing label variable (0 = complete, 1 = missing), used for subsequent model correction; X 同等级样本 For data of the same state level, X 补全 This is for completing data that is randomly missing.

[0028] For fixed missing values, Bayesian inference can be used for completion. This involves constructing a prior distribution based on historical data of cables of the same type, and then combining this with the correlation of measured state variables (such as insulation resistance) to output a posterior distribution. The posterior mean and standard deviation were extracted as the completion result and uncertainty index. Among them, The measured state variable data, This is missing state data.

[0029] For cases where information is completely missing, industry standards, technical specifications, or expert experience are introduced as prior knowledge to fill the gaps. A dynamically decaying confidence coefficient (i.e., a priori coefficient) is assigned to this prior value. This coefficient gradually decreases as measured data accumulates after the equipment is put into operation, allowing the evaluation conclusion to smoothly transition from relying on prior knowledge to relying on actual data. For example, virtual state quantities (such as the upper limit of the normal level) are set based on the "Cable Condition Evaluation Guidelines," and a priori coefficient is labeled. α decreases from 1 to 0 as measured data accumulates.

[0030] By employing the refined and differentiated completion strategies based on missing data types, this invention effectively overcomes the limitations of traditional single completion methods. It not only provides estimates of missing data but, more importantly, offers quantitative measures of the reliability of these estimates (uncertainty indicators), providing a data foundation for subsequent more scientific and robust state assessments and risk predictions.

[0031] S103: Input the completed value and the corresponding uncertainty index into the pre-built fault prediction model to obtain the fault probability and confidence interval of the cable within a specified time window in the future.

[0032] After obtaining the complete dataset with added uncertainty indicators, it is input into a pre-built and trained fault prediction model. The core task of this model is to transform the multi-dimensional state information of the cable into a quantitative prediction of the probability of a fault occurring within a future period. Unlike existing technologies that typically input only the completed data as deterministic values, this fault prediction model is specifically designed or adapted to simultaneously receive completed values ​​representing the current state of the cable, as well as corresponding uncertainty indicators representing the reliability of these completed values. This dual-channel input mechanism ensures that the model clearly understands the reliability of the data itself.

[0033] For example, the failure prediction model is specifically an improved Cox proportional hazards model.

[0034] Specifically, an uncertainty index for the completed data is introduced to modify the Cox proportional hazards model, outputting a quantified probability of failure. The mathematical expression for the Cox proportional hazards model is: ,in, The benchmark risk rate (all independent variables) All risk values ​​are 0, and are only related to time t. This is the state vector of the relevant influencing factors. These are regression coefficients (which need to be fitted with data to reflect the strength and direction of the independent variable's influence on the risk rate). The model input consists of the completed state variables X=[X1,X2,……Xk], missing marker variables M, uncertainty index σ, and operating years t; the risk function is a supplement based on the Cox function.

[0035] The risk function of the improved Cox proportional hazards model is: in, Represents the risk function, The benchmark risk rate; For the first A state vector of each influencing factor; For the first The regression coefficients of the influencing factors; The total number of influencing factors; Weights for missing labels (quantifying the amplifying effect of missing labels on risk). A marker variable to characterize whether data is missing (0 = data complete, 1 = data randomly missing); This represents the uncertainty coefficient (the larger the standard deviation, the higher the risk). To fix the imputation standard deviation (an uncertainty index) for missing data, reflecting the reliability of the Bayesian imputation values ​​(the larger the standard deviation, the greater the imputation error), a modified Cox function is used. To obtain the probability of failure at time T in the future. and confidence interval .

[0036] S104: Determine the optimal maintenance strategy based on the failure probability.

[0037] Based on the aforementioned probability of cable failure within a specified future time window, the final step is to formulate and output maintenance strategies. This aims to transform abstract failure risk predictions into concrete, executable, and cost-effective operation and maintenance action plans, thereby forming a complete closed loop from condition assessment to maintenance decision-making, and thoroughly resolving the problem of disconnect between assessment and operation and maintenance in traditional methods.

[0038] For example, determining the optimal maintenance strategy based on the failure probability includes: Based on the failure probability and the preset failure consequence cost, calculate the risk value before maintenance; Generate or invoke multiple maintenance plans, and estimate the implementation cost and post-implementation risk of each maintenance plan; Calculate the net benefits of each maintenance plan based on the risk value before maintenance, the implementation cost, and the risk value after maintenance. The optimal maintenance strategy is to select the maintenance plan with the greatest net benefit.

[0039] In application, the predicted failure probability is first transformed into a risk value that can directly guide decision-making. This transformation is achieved by combining a pre-defined failure consequence cost matrix. Failure consequence costs not only include direct economic losses such as equipment damage repair, spare parts replacement, and labor costs, but also quantify indirect costs such as production interruptions and social impacts caused by power outages, calculated for example using the average load of various user groups and the unit power outage loss coefficient. By multiplying the failure probability by the failure consequence cost, a quantitative risk value that comprehensively reflects the likelihood and severity of a failure is obtained, thus directly linking technical condition assessment with economic benefit analysis.

[0040] Subsequently, based on this risk quantification result, the process moves to the generation and comparison of maintenance plans. According to the current condition level of the cable, multiple targeted maintenance plans are constructed or retrieved from a predefined measure library, resulting in a maintenance plan set. These solutions can cover various dimensions and levels of operational and maintenance measures, ranging from localized repairs (such as insulation repair) and component replacements (such as joint replacements) to enhanced condition monitoring. For each maintenance solution, it is necessary to estimate the total cost required for its implementation and assess the potential improvement in cable condition after implementation. The latter can be achieved by updating the Cox proportional hazards model parameters to re-predict the failure probability and then calculating the risk value after implementation.

[0041] The current condition level of a cable is a classification result reflecting the cable's health status based on a comprehensive assessment of multiple state quantities (such as insulation resistance, partial discharge, and grounding current). This assessment is primarily based on national and industry standards (such as DL / T 5210.1-2018 Power Construction Quality Acceptance Code and Cable Condition Evaluation Guidelines), combined with cable equipment characteristics (such as insulation aging patterns and the impact of the operating environment) for a comprehensive scoring mapping. Specifically, by scoring each state quantity and calculating a weighted total score, the cable condition can be divided into four levels: normal, warning, abnormal, and critical. A normal state indicates that all state quantities are stable and within standard limits, allowing for safe operation. A warning state indicates that the trend of state quantity changes is close to or partially exceeds the standard limits, requiring enhanced monitoring. An abnormal state indicates that important state quantities are close to or slightly exceed the limits, requiring monitoring and timely maintenance. A critical state indicates that important state quantities are severely exceeded, requiring immediate maintenance. In application, the current status level result can be derived from the assessment conclusions of on-site personnel on the status of each component in engineering practice. It not only provides a basis for filling in the same status level in the data completion strategy, but also serves as the basic input for generating or calling targeted maintenance plans, thereby realizing a closed-loop connection from status assessment to maintenance decision-making.

[0042] Ultimately, the optimal maintenance strategy is determined through a systematic cost-benefit quantitative analysis. Specifically, the net benefit of each maintenance plan is calculated, balancing the risk reduction gains from implementation with the implementation costs. Risk reduction gains represent the reduction in risk value before and after maintenance, taking into account the effective period of the measures. By comparing the net benefits of all candidate plans, the plan with the highest net benefit or the best overall evaluation is selected as the final maintenance strategy. This mechanism ensures that decisions are based not only on technical necessity but also on economic rationality, thereby achieving optimal allocation of maintenance resources, avoiding cost waste due to over-maintenance or missed fault risk assessment due to under-maintenance, and significantly improving the refinement and scientific level of cable line operation and maintenance management.

[0043] The formula for calculating the risk value before maintenance is as follows: In the formula, The risk value before maintenance. Specify a time window for the future Internal failure probability, Direct failure costs include equipment repair costs, spare parts replacement costs, and labor maintenance costs. For indirect failure costs; Let be the average load power (kW) of the i-th type of user. Let be the unit power outage loss coefficient (yuan / kWh) for the i-th type of user. Total number of user categories The estimated duration of the power outage is in hours (h).

[0044] The formula for calculating net benefit is as follows: In the formula, For the first One maintenance plan Net benefits =1, 2, ... , This represents the total number of maintenance plans. The risk value before maintenance. Maintenance plan Post-implementation risk value (calculated by updating Cox model parameters); Maintenance plan The effective period (in years). Maintenance plan The implementation costs, such as equipment, labor, and power outage losses.

[0045] The core objective of updating the Cox model parameters is to adapt the model to scenarios with incomplete cable information and accurately quantify the impact of state variables and data loss uncertainties on fault risk. Specifically, the Cox model parameters mainly include state variable-related parameters, missing uncertainty parameters, and baseline risk parameters. State variable-related parameters β Vectors are used to quantify the strength and direction of the impact of various core state variables (such as partial discharge and insulation resistance) on fault risk; their values ​​can be positive or negative, for example... β 绝缘电阻 A positive value indicates a higher insulation resistance and a higher risk of failure. Missing uncertainty parameters include... γ (Missing label weights) and uncertainty coefficient δ (Complete the standard deviation correction factor), where γ Used to amplify the risks posed by randomly missing data. δ This is used to quantify the risk impact of fixed missing data completion errors. Both are positive, reflecting the engineering logic that "the more unreliable the data, the more cautious the risk needs to be." Benchmark risk parameters. It reflects the natural aging trend of a cable group, is related only to operating time, and does not depend on the specific cable's condition parameters; for example... (5 years) = 0.002 / month, which means that after the cable has been running for 5 years under the baseline condition, the instantaneous probability of a fault occurring in the current month is 0.2%.

[0046] In the decision-making process, to predict the effects of a maintenance plan, it is necessary to simulate the expected changes in the cable condition after the plan is implemented. Specifically, based on the technical content of the maintenance plan (such as partial repair, component replacement, etc.), the degree of improvement it will bring to various cable condition parameters (such as insulation resistance, partial discharge, etc.) is estimated, thereby generating a set of simulated "future condition parameter data" and its corresponding uncertainty indicators. Inputting this set of simulated data into the Cox model allows for the model fitting process to derive the model parameters (including updated condition parameter coefficients) corresponding to the future condition. β Vector, missing uncertain parameters γ and δ (etc.). Using this Cox model updated based on simulated data, the probability of new cable failures within a specified future time window can be calculated, and then substituted into the risk value formula to obtain the estimated risk value after the maintenance plan is implemented. This process realizes quantitative simulation and forward-looking comparison of the effects of each candidate plan during the decision-making stage.

[0047] Corresponding to the aforementioned application function implementation method embodiments, the present invention also provides a cable line condition assessment and maintenance decision system and corresponding embodiments.

[0048] Please see Figure 3 , Figure 3This is a schematic diagram of the module structure of a cable line condition assessment and maintenance decision-making system.

[0049] A cable line condition assessment and maintenance decision-making system, including: The data acquisition and missing data identification unit 31 is used to acquire cable status data and determine the missing data information of the cable status data; the missing data information includes the missing location and the missing type. The data completion unit 32 is used to complete the cable status data based on the missing data information and adopt the corresponding data completion strategy to obtain the completed value and the corresponding uncertainty index. The fault probability prediction unit 33 is used to input the completed value and the corresponding uncertainty index into the pre-built fault prediction model to obtain the fault probability and confidence interval of the cable in a specified time window in the future. The maintenance decision output unit 34 is used to determine the optimal maintenance strategy based on the failure probability.

[0050] In one embodiment, the maintenance decision output unit 34 is specifically used for determining the optimal maintenance strategy based on the failure probability, in particular for: Based on the failure probability and the preset failure consequence cost, calculate the risk value before maintenance; Generate or invoke multiple maintenance plans, and estimate the implementation cost and post-implementation risk of each maintenance plan; Calculate the net benefits of each maintenance plan based on the risk value before maintenance, the implementation cost, and the risk value after maintenance. The optimal maintenance strategy is to select the maintenance plan with the greatest net benefit.

[0051] Regarding the system in the above embodiments, the specific manner in which each unit module performs operations has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0052] Please see Figure 4 The electronic device 4000 includes a memory 4010 and a processor 4020.

[0053] The processor 4020 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0054] Memory 4010 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by processor 4020 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 4010 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, memory 4010 may include removable storage devices that are readable and / or writable, such as laser discs (CDs), read-only digital multifunction optical discs (e.g., DVD-ROMs, dual-layer DVD-ROMs), read-only Blu-ray discs, ultra-high density optical discs, flash memory cards (e.g., SD cards, mini SD cards, Micro-SD cards, etc.), magnetic floppy disks, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0055] The memory 4010 stores executable code, which, when processed by the processor 4020, can cause the processor 4020 to execute part or all of the methods described above.

[0056] Furthermore, the method according to the present invention can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the above-described method of the present invention.

[0057] Alternatively, the present invention may also be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) thereon, which, when executed by a processor of an electronic device (or server, etc.), causes the processor to perform part or all of the steps of the methods described above according to the present application.

[0058] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for cable line condition assessment and maintenance decision-making, characterized in that, include: Acquire cable status data and determine any missing data in the cable status data; The missing data information includes the missing location and the missing type; Based on the missing data information, a corresponding data completion strategy is used to complete the cable status data to obtain the completed value and the corresponding uncertainty index. The completed value and the corresponding uncertainty index are input into the pre-built fault prediction model to obtain the fault probability and confidence interval of the cable within a specified time window in the future. Based on the failure probability, the optimal maintenance strategy is determined.

2. The cable line condition assessment and maintenance decision-making method according to claim 1, characterized in that, The missing types include at least one of random missing, fixed missing, and complete missing.

3. The cable line condition assessment and maintenance decision-making method according to claim 2, characterized in that, The corresponding data completion strategy adopted includes: For the aforementioned random missing values, the median of the same state level is used for filling. For the fixed missing values, Bayesian inference is used to obtain the imputation values ​​and standard deviations; For the complete absence, an industry standard value is used as the completion value, and a priori coefficient is introduced to characterize the credibility of the industry standard value.

4. The cable line condition assessment and maintenance decision-making method according to claim 1, characterized in that, The fault prediction model is specifically an improved Cox proportional hazards model, and the risk function of the improved Cox proportional hazards model is: in, Represents the risk function, The benchmark risk rate; For the first A state vector of each influencing factor; For the first The regression coefficients of the influencing factors; The total number of influencing factors; Weights for missing labels; A marker variable to indicate whether data is missing; Here is the uncertainty coefficient; To fix the missing data and complete the standard deviation.

5. The cable line condition assessment and maintenance decision-making method according to claim 1, characterized in that, Based on the aforementioned failure probabilities, the optimal maintenance strategy is determined, including: Based on the aforementioned failure probability and the preset failure consequence cost, calculate the risk value before maintenance; Generate or invoke multiple maintenance plans, and estimate the implementation cost and post-implementation risk value of each maintenance plan; Calculate the net benefit of each maintenance plan based on the risk value before maintenance, the implementation cost, and the risk value after maintenance. The optimal maintenance strategy is to select the maintenance plan with the greatest net benefit.

6. The cable line condition assessment and maintenance decision-making method according to claim 5, characterized in that, The formula for calculating the risk value before maintenance is as follows: in, The risk value before maintenance. Specify a time window for the future Internal failure probability, For direct failure costs, For indirect failure costs, Let be the average load power of the i-th type of user. Let be the unit power outage loss coefficient for user type i. Total number of user categories This indicates the expected duration of the power outage.

7. The cable line condition assessment and maintenance decision-making method according to claim 5, characterized in that, The formula for calculating the net benefit is as follows: in, For the first Individual maintenance plan Net benefits The risk value before maintenance. Maintenance plan Risk values ​​after implementation; Maintenance plan The effective period, Maintenance plan The implementation cost.

8. A cable line condition assessment and maintenance decision-making system, characterized in that, include: A data acquisition and missing data identification unit is used to acquire cable status data and determine the missing data information of the cable status data. The missing data information includes the missing location and the missing type; The data completion unit is used to complete the cable status data based on the missing data information and using the corresponding data completion strategy to obtain the completed value and the corresponding uncertainty index. The fault probability prediction unit is used to input the completed value and the corresponding uncertainty index into the pre-built fault prediction model to obtain the fault probability and confidence interval of the cable in a specified time window in the future. The maintenance decision output unit is used to determine the optimal maintenance strategy based on the fault probability.

Images