Urban sewage pipe network system operation and maintenance effect evaluation system and method

Abstract

The invention discloses an urban sewage pipe network system operation and maintenance effect evaluation system and method, and belongs to the technical field of sewage pipe network operation and maintenance management. According to the method, through four core models of data quality scoring, improved AHP-entropy weight method combination weighting, pipe network toughness approximate quantification and public security risk Bayesian updating, robust evaluation of data incomplete and error scenes is realized; and in combination with a core-maneuvering-wide area three-layer three-dimensional perception network and a four-step progressive data governance process, outputting a system health index with a confidence interval, designing a double-layer performance payment function associated with data quality, and generating a data short board diagnosis report. According to the method, the problem that efficiency quantification is difficult in the landing of an effect-based payment mechanism is solved, an enhanced closed loop of evaluation, data improvement and accurate operation and maintenance is constructed, a scientific quantification basis is provided for performance management of the sewage pipe network, and the method is suitable for operation and maintenance evaluation of the urban sewage pipe network under various data conditions.

Description

Technical Field

[0001] This invention belongs to the field of sewage pipe network operation and maintenance management technology, and particularly relates to a system and method for evaluating the operation and maintenance effect of urban sewage pipe network systems. Background Technology

[0002] Currently, the "integrated plant and network" operation and the "pay-per-performance" mechanism directly link government payments with the system performance output of the operation service provider, which can effectively improve the overall efficiency of the drainage system.

[0003] However, the implementation of the "pay-for-performance" mechanism urgently requires a scientific, quantitative, and fair evaluation system for operation and maintenance effectiveness. Most urban sewage pipe networks in my country face severe data challenges: First, basic data suffers from "archive loss," with widespread issues such as missing design drawings, inaccurate records, and unclear pipe material properties due to the large span of pipe network construction years, resulting in ambiguous system topology. Second, operational data suffers from "perception blind spots," with low coverage density of online monitoring equipment, high investment and maintenance costs, and frequent data transmission interruptions and distortions, making it difficult to obtain systematic real-time operational status. Third, data quality is shrouded in uncertainty, with monitoring data exhibiting significant errors due to untimely calibration, improper installation, or external interference, and lacking effective verification and quality control mechanisms.

[0004] Existing pipeline network assessment research largely focuses on specific technical fields, such as intelligent structural defect identification algorithms based on CCTV, overflow simulation analysis based on hydraulic models like SWMM, and economic assessment based on the asset lifecycle theory. These studies all implicitly rely on the key assumption of "obtaining relatively complete and accurate basic data and continuously reliable monitoring data," making them ill-suited to the reality of my country's "data desert" and "data fog." Directly applying such idealized "refined models" often leads to distorted evaluation results and frequent disputes, ultimately causing the "pay-for-performance" mechanism to stagnate or become merely a formality. Therefore, there is an urgent need to construct an evaluation method that can tolerate data defects and objectively reflect the true effectiveness of the system amidst uncertainty, providing technical support for the implementation of the "pay-for-performance" mechanism. Summary of the Invention

[0005] To address the problems mentioned in the background section, this invention proposes an evaluation system and method for the operation and maintenance effectiveness of urban sewage pipe network systems. This system and method are adaptable to data uncertainty and possess fault tolerance and dynamic adjustment capabilities. It can still output reliable evaluation results even under conditions of missing or poor-quality data, supporting performance-based payment and management decisions.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] This invention provides a method for evaluating the operation and maintenance effectiveness of urban sewage pipe network systems, comprising the following steps:

[0008] S1. Construct a three-dimensional evaluation framework of "collection and transportation efficiency - pipeline network resilience - public safety", with the weight ratio of the three dimensions set at 5:3:2;

[0009] S2. Configure a three-level quantitative path of "standard-alternative path-safety net" for each final evaluation indicator;

[0010] S3. Construct a combined weighting model that couples data quality;

[0011] S4. Approximate quantification of pipeline network resilience based on key node monitoring:

[0012] S5. Use a Bayesian update model to dynamically assess public safety risks;

[0013] S6. Construct a three-layer three-dimensional perception network of "core-mobile-wide area" to collect multi-source data, and process it through a four-step progressive data governance process;

[0014] S7. Calculate the scores of the three-dimensional evaluation dimensions, and obtain the system health index HSI = μ ± δ by weighted summation, with a confidence level of 1 - α, where α is the significance level. Design a two-level performance-based payment function. ,in For the actual amount paid, The base service fee, The performance coefficient is based on the HSI central value μ. When μ≥85, f(μ)=1.2; when 70≤μ<85, f(μ)=1.0; when 60≤μ<70, f(μ)=0.8; when μ<60, f(μ)=0.5. g(δ) is the data quality coefficient (when δ≤3, g(δ)=1.0; when 3<δ≤5, g(δ)=0.9; when δ>5, g(δ)=0.8).

[0015] S8. Generate a data value and investment priority diagnosis report, identify areas with data shortcomings, indicators with shortcomings, and investment priorities for data improvement, drive the supplementation of data acquisition equipment and the improvement of pipeline archives, and formulate precise operation and maintenance plans based on evaluation results to form an enhanced closed loop of "evaluation-data improvement-precise operation and maintenance-effective payment".

[0016] Preferably, in step S1,

[0017] Collection and transportation efficiency dimension: Evaluate the core function of the pipeline network of "collecting all waste and transporting smoothly", which is directly related to the pollutant reduction target, and select COD collection load rate, sewage collection rate and pipeline unobstructed rate;

[0018] Pipeline resilience dimension: Evaluate the robustness of the system in maintaining function and recovering quickly under disturbances such as rainfall and faults, focusing on the system's dynamic shock resistance;

[0019] Public safety dimension: Quantify the effectiveness of operation and maintenance in controlling public safety risks such as road collapse, sewage overflow, and missing manhole covers, covering core scenarios of public concern.

[0020] Preferably, in step S2,

[0021] Standard approach: Calculated based on periodic, comprehensive professional inspection data; pipeline structure integrity rate based on comprehensive CCTV inspection results; sewage collection rate based on full-basin flow monitoring data.

[0022] Alternative Path A: When full coverage is not possible, based on the statistical inference results of "key testing in high-risk areas + sampling testing in medium- and low-risk areas", the sampling error range and confidence level are marked.

[0023] Alternative Path B: When no test data is available, a multiple regression model based on "pipe age - material - historical maintenance records - ground anomaly reports - regional environmental characteristics" is used for estimation, and the model parameters are calibrated using historical valid data;

[0024] Fallback Path: When none of the above paths can be achieved, a qualitative description is used to record the potential impact on the evaluation of higher-level indicators. The impact level is divided into three levels: high, medium, and low.

[0025] Preferably, in step S3,

[0026] Data quality quantification: based on four sub-dimensions—completeness, continuity, timeliness, and cross-validation error—according to the formula. Calculate the data quality score q ij ∈[0,1], where The data integrity score is the percentage of valid data within the statistical period. The score is for timeliness, corresponding to the time interval between data collection and application. The data frequency score indicates whether the data collection density meets the standard. Cross-validation error refers to the degree of deviation from multi-source corroborating data. For each sub-dimension weight, and =0.3、 =0.3、 =0.2、 =0.2;

[0027] Improved entropy weighting method: After standardizing the data matrix R=(rij)m×n, data quality weights are introduced to calculate the feature proportions. Then calculate the improved entropy value. Coefficient of difference and objective weight ;

[0028] Combination and dynamic adjustment: Incorporating AHP subjective weights The weighting was determined through expert scoring and consistency testing, and was consistent with objective weighting. , to obtain the combined weights ,in This is the weighting coefficient, ranging from 0.3 to 0.5; if the q of a certain indicator is greater than or equal to 0.5 for two consecutive evaluation periods... ij If the mean is lower than the threshold θ=0.6, the combined weight will be reduced by 20%-50% in the next period until the data quality score rises back above the threshold.

[0029] Preferably, in step S4,

[0030] Key nodes are selected, including the main inlet pipe of the sewage treatment plant, the forebay of the booster pump station, the inspection well at the lowest point of the terrain, the inspection well at historically flood-prone points, the intersection of main pipes, and the watershed boundary nodes. The number of nodes is configured according to the standard of no less than 2 per 5km² of the pipeline coverage area.

[0031] Monitoring and surrogate variable definition: Install high-precision level gauges at key nodes, with a sampling frequency of no less than 15 minutes / time, to continuously monitor the filling degree. Set the filling degree safety threshold to 85% of the pipe diameter, and define a surrogate variable for system functional health. ,in The node topology importance weights are determined through pipeline topology analysis and expert evaluation, and their sum is 1; I( The function is an indicator function that takes the value 1 if the condition is met, and 0 otherwise. =0.85 is the safety threshold coefficient. For the pipe diameter corresponding to the node, The weighted proportion of key nodes whose fill degree did not exceed the threshold during the disturbance period;

[0032] Resilience index calculation: recording disturbance events This is the start time of the disturbance. During the period when functionality recovers to more than 80% of its pre-disturbance level. The change curve, according to the formula Calculate the toughness index, where The proxy reliability coefficient (0.7≤ρ≤1.0) is determined by the proportion of traffic coverage of the key node to the entire network and the results of historical simulation verification.

[0033] Preferably, in step S5,

[0034] Prior probabilities are generated based on pipeline inspection data X, including structural defect level and corrosion degree; soil type data S, including soil bearing capacity and moisture content; and traffic load data T, including average daily traffic flow and proportion of heavy-load vehicles. Prior risk probabilities P(Risk|X,S,T) are generated through a logistic regression model, which is trained and optimized using historical risk event data.

[0035] Construct a likelihood function for multi-source evidence: collect municipal hotline complaints E1, including abnormal road sounds, depressions, and odors; visual records from inspectors E2, including ground subsidence around pipelines and loose manhole covers; leakage reports from other infrastructure along the same road section E3, including leakage data from water supply pipes and gas pipes; and InSAR remote sensing ground subsidence data E4, millimeter-level subsidence trend Ev. Evaluate the likelihood ratio of each piece of evidence under the condition that risk exists and risk does not exist.

[0036] Update posterior probability: When a new evidence set Enew is obtained, follow Bayes' theorem. Update risk probability, where =1- .

[0037] Preferably, in step S6, the three-layer sensing network consists of: a core verification layer deploying highly reliable online instruments at the wastewater treatment plant inlet, booster pump stations, and key watershed boundaries, with data transmission stability ≥95%; a mobile inspection layer equipped with vehicle-mounted mobile monitoring units and handheld detection devices to perform quarterly planned inspections and 24-hour responsive inspections; and a wide-area sensing layer accessing meteorological radar rainfall data, pump station operating power consumption data, key area video surveillance water accumulation identification results, and social crowdsourcing information.

[0038] Four-step data governance: rule-based automatic cleaning, filtering outlier data based on physical constraints and industry standards; multi-source spatiotemporal alignment and verification, aligning data under a unified spatiotemporal benchmark and identifying contradictory data according to the reliability level of the data source; machine learning imputation based on spatiotemporal kriging interpolation or LSTM sequence prediction, outputting the imputation value and the uncertainty range of the 95% confidence interval; generating a data lineage report, recording the original data source, acquisition equipment information, all processing steps, other data sources integrated, and the final uncertainty estimate.

[0039] Preferably, the multi-source evidence in step S5 also includes construction activity records around the pipeline (E5), groundwater level monitoring data (E6), and frequency of extreme weather events (E7). The construction activity records must include the construction type, construction depth, and horizontal distance from the pipeline. The groundwater level monitoring data must be updated monthly. Extreme weather is defined as "daily rainfall ≥ 50 mm". The likelihood ratio (LR) is set as follows: municipal hotline complaints LR = 2.5-3.0, inspection records LR = 3.0-4.0, related facility leakage LR = 2.0-2.5, InSAR settlement data LR = 4.0-5.0, construction activities LR = 3.5-4.5, groundwater level LR = 2.0-3.0, and extreme weather LR = 1.5-2.0. The specific values ​​are calibrated based on historical data of the region.

[0040] Preferably, the criteria for determining the regularized automatic cleaning in step S6 also include: the deviation between the daily average upstream flow rate and the daily average downstream flow rate of the pump station is ≤10%; the daily average downstream liquid level is not higher than the daily average upstream liquid level; the correlation coefficient between the pipeline liquid level and flow rate during rainfall is ≥0.7; the time synchronization accuracy of multi-source spatiotemporal alignment is ≤1 minute, and the spatial matching accuracy is ≤10 meters; the data lineage report includes the data acquisition timestamp, equipment number, operator, processing algorithm version, and uncertainty calculation process, supports tracing the entire life cycle of each evaluation data point, and the report format is compatible with the import requirements of mainstream data management platforms.

[0041] On the other hand, the present invention also provides an evaluation system for the operation and maintenance effect of urban sewage pipe network system, including the following modules:

[0042] The three-dimensional evaluation framework module is used to construct an evaluation system from the following three dimensions: collection and transportation efficiency, pipeline network resilience, and public safety.

[0043] The fault tolerance index quantification module designs a three-level quantification path of "standard-substitution-backup" for each evaluation index, and supports switching evaluation methods under different data conditions;

[0044] The data quality perception and empowerment module includes a data quality scoring unit that assesses data quality from four dimensions: completeness, continuity, timeliness, and consistency.

[0045] An improved entropy weighting unit based on data quality weighting is used to dynamically adjust indicator weights, combine weight calculation units, and integrate subjective and objective weights.

[0046] The resilience approximation quantification module, based on monitoring data from key nodes, approximates the system's disturbance resistance and recovery capabilities through proxy variables and resilience integral algorithms.

[0047] The multi-source evidence fusion risk assessment module uses a Bayesian inference model to integrate multi-source information such as inspection records, citizen complaints, and remote sensing data, and dynamically updates the probability of public safety risks.

[0048] The economical monitoring network construction module adopts a three-layer perception architecture of "core-mobile-wide area" to achieve low-cost and high-coverage data acquisition;

[0049] The data governance and transparent traceability module enables automatic data cleaning, multi-source alignment, intelligent interpolation, and full-process traceability.

[0050] The performance-based payment linkage module designs a two-layer payment function based on the confidence interval of the evaluation results, linking data quality with payment.

[0051] Compared with the prior art, the present invention has the following characteristics:

[0052] 1. A "Resilience Evaluation for Data Uncertainty" paradigm is proposed, which internalizes data uncertainty from the external challenge of evaluation into the core parameters of the model. The evaluation system itself has the ability to tolerate data defects, the ability to integrate multi-source information, and the ability to infer uncertainty. It is perfectly adapted to the current data status of my country's sewage pipe network and solves the core technical bottleneck of "pay-for-performance" implementation.

[0053] 2. The constructed "three-dimensional-fault-tolerant-dynamic" evaluation framework and four core model groups have achieved full-process technological innovation from data processing and indicator weighting to performance evaluation and risk assessment. The evaluation results are objective, robust and have confidence intervals, providing credible and traceable quantitative basis for performance-based payment and reducing evaluation disputes.

[0054] 3. By adopting a three-dimensional sensing network of "core-mobile-wide area" and a four-step data governance process, we can improve data quality and availability while controlling monitoring and operation and maintenance costs. Through data quality linkage mechanism and data shortcoming diagnosis, we can drive the continuous accumulation of data assets and form a virtuous cycle of "data improvement-accurate evaluation-operation and maintenance optimization".

[0055] 4. The method is highly standardized and easy to operate, providing a standardized toolkit that can be used out of the box. Evaluation can be started on urban sewage pipe networks under various data conditions without waiting for complete data, promoting the rapid implementation of the "pay-for-performance" mechanism and having broad practical value and prospects for promotion. Attached Figure Description

[0056] Figure 1 System overall architecture diagram

[0057] Figure 2 Evaluation Flowchart

[0058] Figure 3Data quality scoring model calculation flowchart

[0059] Figure 4 Flowchart of the Improved Entropy Weight Method Weighting Model

[0060] Figure 5 Flowchart of the approximate quantitative model for pipeline resilience

[0061] Figure 6 Bayesian risk update model calculation flowchart

[0062] Figure 7 System collaboration workflow diagram. Detailed Implementation

[0063] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0064] like Figure 1 As shown in the figure, this embodiment provides an evaluation system for the operation and maintenance effect of an urban sewage pipe network system, including the following modules:

[0065] The three-dimensional evaluation framework module is used to construct an evaluation system from the following three dimensions: collection and transportation efficiency, pipeline network resilience, and public safety.

[0066] The fault tolerance index quantification module designs a three-level quantification path of "standard-substitution-backup" for each evaluation index, and supports switching evaluation methods under different data conditions;

[0067] The data quality perception and empowerment module includes a data quality scoring unit that assesses data quality from four dimensions: completeness, continuity, timeliness, and consistency.

[0068] An improved entropy weighting unit based on data quality weighting is used to dynamically adjust indicator weights, combine weight calculation units, and integrate subjective and objective weights.

[0069] The resilience approximation quantification module, based on monitoring data from key nodes, approximates the system's disturbance resistance and recovery capabilities through proxy variables and resilience integral algorithms.

[0070] The multi-source evidence fusion risk assessment module uses a Bayesian inference model to integrate multi-source information such as inspection records, citizen complaints, and remote sensing data, and dynamically updates the probability of public safety risks.

[0071] The economical monitoring network construction module adopts a three-layer perception architecture of "core-mobile-wide area" to achieve low-cost and high-coverage data acquisition;

[0072] The data governance and transparent traceability module enables automatic data cleaning, multi-source alignment, intelligent interpolation, and full-process traceability.

[0073] The performance-based payment linkage module designs a two-layer payment function based on the confidence interval of the evaluation results, linking data quality with payment.

[0074] This embodiment also provides a performance-based evaluation method for the operation and maintenance of sewage pipe networks, characterized by the following steps:

[0075] S1. Construct a three-dimensional evaluation framework of "collection and transportation efficiency - pipeline resilience - public safety", with a preset weight ratio of 5:3:2 for the three dimensions, such as... Figure 2 As shown, where:

[0076] Collection and transportation efficiency dimension: Evaluate the core function of the pipeline network of "collecting all waste and transporting smoothly", which is directly related to the pollutant reduction target, and select key indicators such as COD collection load rate, sewage collection rate, and pipeline unobstructed rate;

[0077] Pipeline network resilience dimension: Evaluates the robustness of the system in maintaining function and recovering quickly under disturbances such as rainfall and faults, focusing on the system's dynamic shock resistance;

[0078] Public safety dimension: Quantify the effectiveness of operation and maintenance in controlling public safety risks such as road collapse, sewage overflow, and missing manhole covers, covering core scenarios of public concern;

[0079] S2. Configure a three-level quantitative path of "standard-substitute-backup" for each final evaluation indicator:

[0080] Standard approach: Calculated based on periodic, comprehensive professional testing data, such as pipeline structure integrity rate based on comprehensive CCTV testing results, and sewage collection rate based on full-basin flow monitoring data;

[0081] Alternative Path A: When full coverage is not possible, based on the statistical inference results of "key testing in high-risk areas + sampling testing in medium- and low-risk areas", the sampling error range and confidence level are marked.

[0082] Alternative Path B: When no test data is available, a multiple regression model based on "pipe age - material - historical maintenance records - ground anomaly reports - regional environmental characteristics" is used for estimation, and the model parameters are calibrated using historical valid data;

[0083] fallback path: When none of the above paths can be achieved, a qualitative description is used and the potential impact on the evaluation of higher-level indicators is recorded (high, medium, and low levels).

[0084] S3. Construct a combined weighting model that couples data quality, such as... Figure 3 As shown, it specifically includes:

[0085] Data quality quantification: based on four sub-dimensions—completeness, continuity, timeliness, and cross-validation error—according to the formula. Calculate the data quality score q ij ∈[0,1], where The score is for data integrity (the percentage of valid data within the statistical period). The timeliness score is the score corresponding to the time interval between data collection and application. The data frequency score (the degree to which the data collection density meets the standard). Cross-validation error (the degree of deviation from multi-source corroboration data). For each sub-dimension weight, and =0.3、 =0.3、 =0.2、 =0.2;

[0086] like Figure 4 As shown, the improved entropy weight method calculates: the standardized data matrix R = (r ij After m×n, data quality weights are introduced to calculate the feature proportions. Then calculate the improved entropy value. Coefficient of difference and objective weighting;

[0087] Combination and dynamic adjustment: Incorporating AHP subjective weights (Determined through expert scoring and consistency testing) and objective weighting , to obtain the combined weights ,in This is the weighting coefficient (range 0.3-0.5); if the q of a certain indicator is greater than or equal to 0.5 for two consecutive evaluation periods... ij If the mean is lower than the threshold θ=0.6, the combined weight will be reduced by 20%-50% in the next period until the data quality score rises back above the threshold.

[0088] S4. Achieve approximate quantification of pipeline network resilience based on key node monitoring, such as Figure 5 As shown:

[0089] Select key nodes: including the main inlet pipe of the sewage treatment plant, the forebay of the booster pump station, the inspection well at the lowest point of the terrain, the inspection well at the historically flood-prone point, the intersection of the main pipes, the watershed boundary node, and other core nodes that "affect the whole body if one part is affected". The number of nodes is configured according to the standard of no less than 2 per 5km² of pipeline coverage area.

[0090] Monitoring and surrogate variable definition: Install high-precision level gauges at key nodes, sampling at a frequency of no less than 15 minutes per measurement, to continuously monitor the filling degree (level / diameter). Set the filling degree safety threshold to 85% of the pipe diameter, and define a surrogate variable for system functional health. ,in The node topology importance weights (determined through pipeline topology analysis and expert evaluation, with a sum of 1), I( () is an indicator function (it takes the value 1 if the condition is met, and 0 otherwise). =0.85 is the safety threshold coefficient, where is the pipe diameter corresponding to the node. The weighted proportion of key nodes whose fill degree did not exceed the threshold during the disturbance period;

[0091] Resilience index calculation: recording disturbance events This is the start time of the disturbance. During the period when functionality recovers to more than 80% of its pre-disturbance level. The change curve, according to the formula Calculate the toughness index, where The proxy reliability coefficient (0.7≤ρ≤1.0) is determined by the proportion of traffic coverage of the key node to the entire network and the results of historical simulation verification.

[0092] S5, such as Figure 6 , 7 As shown, a Bayesian update model is used to dynamically assess public safety risks:

[0093] Generate prior probabilities: Based on pipeline inspection data (X, including structural defect level and corrosion degree), soil type data (S, including soil bearing capacity and moisture content), and traffic load data (T, including average daily traffic flow and proportion of heavy-load vehicles), a prior risk probability P(Risk|X,S,T) is generated through a logistic regression model. The model is trained and optimized using historical risk event data.

[0094] Construct a multi-source evidence likelihood function: Collect multiple sources of evidence, including complaints from municipal hotlines (E1, such as abnormal road noises, depressions, and odors), visual records from inspectors (E2, such as ground subsidence around pipelines and loose manhole covers), reports of leaks in other infrastructure along the same road section (E3, such as data on leaks in water supply pipes and gas pipes), and InSAR remote sensing data on ground subsidence (E4, millimeter-level subsidence trends). Ev is then used to assess the likelihood ratio of each piece of evidence under conditions of presence (Risk) and absence (¬Risk). ;

[0095] Update posterior probability: Obtain new evidence set E new At that time, according to Bayes' theorem Update risk probability, where =1- ;

[0096] S6. Construct a three-layered, multi-source sensing network ("core-mobile-wide area") to collect multi-source data, which is then processed through a four-step progressive data governance process:

[0097] The three-layer sensing network consists of: a core verification layer deploying highly reliable online instruments (flow rate, liquid level, COD concentration) at wastewater treatment plant inlets, booster pump stations, and key watershed boundaries, with data transmission stability ≥95%; a mobile inspection layer equipped with vehicle-mounted mobile monitoring units and handheld testing devices (portable flow meters, rapid water quality analyzers) to perform quarterly planned inspections and 24-hour responsive inspections; and a wide-area sensing layer integrating meteorological radar rainfall data, pump station operating power consumption data, water accumulation identification results from video surveillance in key areas, and social crowdsourcing information (reported through citizen apps and the 12345 hotline).

[0098] Four-step data governance: Rule-based automatic cleaning, filtering outlier data (such as negative flow rate or COD concentration exceeding the reasonable range of 50-1000 mg / L) based on physical constraints and industry standards; multi-source spatiotemporal alignment and verification, aligning data under a unified spatiotemporal reference (UTC time + plane coordinate system), and identifying contradictory data according to the reliability level of the data source (core layer > mobile layer > wide area layer); machine learning interpolation based on spatiotemporal kriging interpolation or LSTM sequence prediction, outputting the interpolated value and the uncertainty range of the 95% confidence interval; generating a "Data Lineage Report" to record the original data source, acquisition equipment information, all processing steps, other data sources integrated, and the final uncertainty estimate.

[0099] S7. Calculate the scores of the three-dimensional evaluation dimensions, and obtain the system health index HSI (HSI = μ ± δ, confidence level 1 - α, where α is the significance level) by weighted summation. Design a two-level performance-based payment function. ,in For the actual amount paid, The base service fee, The performance coefficient is based on the HSI center value μ (f(μ)=1.2 when μ≥85, f(μ)=1.0 when 70≤μ<85, f(μ)=0.8 when 60≤μ<70, f(μ)=0.5 when μ<60), and the data quality coefficient is g(δ)=1.0 when δ≤3, g(δ)=0.9 when 3<δ≤5, g(δ)=0.8 when δ>5).

[0100] S8. Generate the "Data Value and Investment Priority Diagnosis Report" to identify data deficiency areas (by pipeline network zoning), deficiency indicators (by three-dimensional evaluation dimensions), and data improvement investment priorities (high, medium, and low levels). This will drive the addition of data acquisition equipment and the improvement of pipeline network archives. Based on the evaluation results, a precise operation and maintenance plan will be formulated (such as targeted dredging, mixed connection point renovation, and defective pipeline repair), forming an enhanced closed loop of "evaluation - data improvement - precise operation and maintenance - effective payment".

[0101] In step S3, the data quality score q ijThe corresponding percentage scoring criteria are as follows: 85 points or above is "Excellent", 60-85 points is "Pass", and below 60 points is "Unreliable". The percentage score Q = 0.3 × Completeness Score + 0.3 × Continuity Score + 0.2 × Reasonableness Score + 0.2 × Consistency Score. The continuity score is based on "single missing data ≤ 4 hours and daily cumulative missing data ≤ 6 hours" as the full score standard, with 5 points deducted for each hour exceeding this standard. The reasonableness score is based on "values ​​within the physically feasible range and without abrupt anomalies" as the full score standard, with 3 points deducted for each outlier found. The consistency score is based on "no contradictions with related data" as the full score standard, with 10 points deducted for each contradiction found.

[0102] In step S4, the selection of key nodes must meet the following conditions: core pipe sections covering ≥80% of the total pipeline flow, including inspection wells within 500 meters of all historical risk event occurrence points, and key locations such as the lowest point and main pipeline intersection nodes must be selected; the fullness monitoring data must be transmitted to the data platform in real time, and the data loss rate must not exceed 1 hour for a single instance and not exceed 3 hours cumulatively per day, otherwise the mobile inspection layer will be triggered for on-site verification and data supplementation.

[0103] The multi-source evidence in step S5 also includes construction activity records around the pipeline (E5), groundwater level monitoring data (E6), and frequency of extreme weather events (E7). Construction activity records must include information such as construction type, construction depth, and horizontal distance from the pipeline. Groundwater level monitoring data must be updated monthly. Extreme weather is defined as "daily rainfall ≥ 50mm". The likelihood ratio (LR) is set as follows: E1 (municipal hotline complaints) LR = 2.5-3.0, E2 (inspection records) LR = 3.0-4.0, E3 (related facility leakage) LR = 2.0-2.5, E4 (InSAR settlement data) LR = 4.0-5.0, E5 (construction activities) LR = 3.5-4.5, E6 (groundwater level) LR = 2.0-3.0, and E7 (extreme weather) LR = 1.5-2.0. The specific values ​​are calibrated based on historical data of the region.

[0104] The criteria for determining the regularized automatic cleaning in step S6 include: the deviation between the daily average upstream flow rate and the daily average downstream flow rate of the pump station is ≤10%; the daily average downstream liquid level is not higher than the daily average upstream liquid level; the correlation coefficient between the pipeline liquid level and flow rate during rainfall is ≥0.7; the time synchronization accuracy of multi-source spatiotemporal alignment is ≤1 minute, and the spatial matching accuracy is ≤10 meters; the "Data Lineage Report" must include key information such as data collection timestamp, equipment number, operator, processing algorithm version, and uncertainty calculation process, support the tracing of the entire life cycle of each evaluation data point, and the report format is compatible with the import requirements of mainstream data management platforms.

[0105] Taking an old urban area as an example, the implementation steps of this system are explained as follows:

[0106] 1. Deploy a three-tiered monitoring network consisting of a core layer, a mobile layer, and a wide-area layer;

[0107] 2. Collect data from multiple sources and conduct data quality scoring and governance;

[0108] 3. Implement a quantitative fault tolerance index and an approximate resilience assessment model;

[0109] 4. Integrate multi-source evidence to update public safety risks;

[0110] 5. Calculate the System Health Index (HSI) and its confidence interval;

[0111] 6. Calculate performance fees based on the payment function and output a data diagnostic report;

[0112] 7. Optimize monitoring layout and operation and maintenance measures based on diagnostic results to form closed-loop management.

[0113] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for evaluating the operation and maintenance effectiveness of an urban sewage pipe network system, characterized in that, Includes the following steps: S1. Construct a three-dimensional evaluation framework of "collection and transportation efficiency - pipeline network resilience - public safety", with the weight ratio of the three dimensions set at 5:3:2; S2. Configure a three-level quantitative path of "standard-alternative path-safety net" for each final evaluation indicator; S3. Construct a combined weighting model that couples data quality; S4. Approximate quantification of pipeline network resilience based on key node monitoring: S5. Use a Bayesian update model to dynamically assess public safety risks; S6. Construct a three-layer three-dimensional perception network of "core-mobile-wide area" to collect multi-source data, and process it through a four-step progressive data governance process; S7. Calculate the scores of the three-dimensional evaluation dimensions, and obtain the system health index HSI = μ ± δ by weighted summation, with a confidence level of 1 - α, where α is the significance level. Design a two-level performance-based payment function. ,in For the actual amount paid, The base service fee, The performance coefficient is based on the HSI central value μ. When μ≥85, f(μ)=1.2; when 70≤μ<85, f(μ)=1.0; when 60≤μ<70, f(μ)=0.8; when μ<60, f(μ)=0.

5. g(δ) is the data quality coefficient (when δ≤3, g(δ)=1.0; when 3<δ≤5, g(δ)=0.9; when δ>5, g(δ)=0.8). S8. Generate a data value and investment priority diagnosis report to identify data deficiency areas, deficiency indicators, and data improvement investment priorities. Drive the addition of data acquisition equipment and the improvement of pipeline archives. Combine the evaluation results to formulate a precise operation and maintenance plan, forming an enhanced closed loop of "evaluation-data improvement-precise operation and maintenance-effective payment".

2. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, In step S1 Collection and transportation efficiency dimension: Evaluate the core function of the pipeline network of "collecting all waste and transporting smoothly", which is directly related to the pollutant reduction target, and select COD collection load rate, sewage collection rate and pipeline unobstructed rate; Pipeline resilience dimension: Evaluate the robustness of the system in maintaining function and recovering quickly under disturbances such as rainfall and faults, focusing on the system's dynamic shock resistance; Public safety dimension: Quantify the effectiveness of operation and maintenance in controlling public safety risks such as road collapse, sewage overflow, and missing manhole covers, covering core scenarios of public concern.

3. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, In step S2 Standard approach: Calculated based on periodic, comprehensive professional inspection data; pipeline structure integrity rate based on comprehensive CCTV inspection results; sewage collection rate based on full-basin flow monitoring data. Alternative Path A: When full coverage is not possible, based on the statistical inference results of "key detection in high-risk areas + sampling detection in medium- and low-risk areas", the sampling error range and confidence level are marked; Alternative Path B: When no test data is available, estimate the parameters using a multiple regression model constructed based on "pipe age - material - historical maintenance records - ground anomaly reports - regional environmental characteristics". The model parameters are calibrated using historical valid data. Fallback Path: When none of the above paths can be achieved, a qualitative description is used to record the potential impact on the evaluation of higher-level indicators. The impact level is divided into three levels: high, medium, and low.

4. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, In step S3 Data quality quantification: based on four sub-dimensions—completeness, continuity, timeliness, and cross-validation error—according to the formula. Calculate the data quality score q ij ∈[0,1], where The data integrity score is the percentage of valid data within the statistical period. The score is for timeliness, corresponding to the time interval between data collection and application. The data frequency score indicates whether the data collection density meets the standard. Cross-validation error refers to the degree of deviation from multi-source corroborating data. For each sub-dimension weight, and =0.3、 =0.3、 =0.2、 =0.2; Improved entropy weighting method: After standardizing the data matrix R=(rij)m×n, data quality weights are introduced to calculate the feature proportions. Then calculate the improved entropy value. Coefficient of difference and objective weight ; Combination and dynamic adjustment: Incorporating AHP subjective weights The weighting was determined through expert scoring and consistency testing, and was consistent with objective weighting. , to obtain the combined weights ,in This is the weighting coefficient, ranging from 0.3 to 0.5; if the q of a certain indicator is greater than or equal to 0.5 for two consecutive evaluation periods... ij If the mean is lower than the threshold θ=0.6, the combined weight will be reduced by 20%-50% in the next period until the data quality score rises back above the threshold.

5. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, In step S4 Key nodes are selected, including the main inlet pipe of the sewage treatment plant, the forebay of the booster pump station, the inspection well at the lowest point of the terrain, the inspection well at historically flood-prone points, the intersection of main pipes, and the watershed boundary nodes. The number of nodes is configured according to the standard of no less than 2 per 5km² of the pipeline coverage area. Monitoring and surrogate variable definition: Install high-precision level gauges at key nodes, with a sampling frequency of no less than 15 minutes / time, to continuously monitor the filling degree. Set the filling degree safety threshold to 85% of the pipe diameter, and define a surrogate variable for system functional health. ,in The node topology importance weights are determined through pipeline topology analysis and expert evaluation, and their sum is 1. I( The function is an indicator function that takes the value 1 if the condition is met, and 0 otherwise. =0.85 is the safety threshold coefficient. For the pipe diameter corresponding to the node, The weighted proportion of key nodes whose fill degree did not exceed the threshold during the disturbance period; Resilience index calculation: recording disturbance events This is the start time of the disturbance. During the period when functionality recovers to more than 80% of its pre-disturbance level. The change curve, according to the formula Calculate the toughness index, where The proxy reliability coefficient (0.7≤ρ≤1.0) is determined by the proportion of traffic coverage of the key node to the entire network and the results of historical simulation verification.

6. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, In step S5 Prior probabilities are generated based on pipeline inspection data X, including structural defect level and corrosion degree; soil type data S, including soil bearing capacity and moisture content; and traffic load data T, including average daily traffic flow and proportion of heavy-load vehicles. Prior risk probabilities P(Risk|X,S,T) are generated through a logistic regression model, which is trained and optimized using historical risk event data. Construct a multi-source evidence likelihood function: collect municipal hotline complaints E1, including abnormal road noises, dents, and odors; Inspectors visually record E2, including ground subsidence around pipelines and loose manhole covers; E3, reports of leaks in other infrastructure along the same road section, including data on leaks in water and gas pipes; E4, InSAR remote sensing data on ground subsidence, millimeter-level subsidence trends; and assess the likelihood ratio of each piece of evidence in the presence and absence of risk. ; Update posterior probability: When a new evidence set Enew is obtained, follow Bayes' theorem. Update risk probability, where =1- .

7. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, In step S6, the three-layer sensing network consists of a core verification layer that deploys highly reliable online instruments at the wastewater treatment plant inlet, booster pump stations, and key watershed boundaries, with data transmission stability ≥95%. The mobile inspection layer is equipped with vehicle-mounted mobile monitoring units and handheld detection equipment to carry out quarterly planned inspections and 24-hour responsive inspections; the wide-area perception layer accesses meteorological radar rainfall data, pump station operation power consumption data, water accumulation identification results from video surveillance in key areas, and social crowdsourcing information; Four-step data governance: rule-based automatic cleaning, filtering outlier data based on physical constraints and industry standards; multi-source spatiotemporal alignment and verification, aligning data under a unified spatiotemporal benchmark and identifying contradictory data according to the reliability level of the data source; machine learning imputation based on spatiotemporal kriging interpolation or LSTM sequence prediction, outputting the imputation value and the uncertainty range of the 95% confidence interval; generating a data lineage report, recording the original data source, acquisition equipment information, all processing steps, other data sources integrated, and the final uncertainty estimate.

8. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, The multi-source evidence in step S5 also includes construction activity records around the pipeline (E5), groundwater level monitoring data (E6), and frequency of extreme weather events (E7). The construction activity records must include the construction type, construction depth, and horizontal distance from the pipeline. The groundwater level monitoring data must be updated monthly. Extreme weather is defined as "daily rainfall ≥ 50mm". The likelihood ratio (LR) is calculated as follows: municipal hotline complaints LR = 2.5-3.0, inspection records LR = 3.0-4.0, related facility leakage LR = 2.0-2.5, InSAR settlement data LR = 4.0-5.0, construction activities LR = 3.5-4.5, groundwater level LR = 2.0-3.0, and extreme weather LR = 1.5-2.

0. The specific values ​​are calibrated based on historical data of the region.

9. The method for evaluating the operation and maintenance effect of an urban sewage pipe network system according to claim 1, characterized in that, The criteria for determining the regularized automatic cleaning in step S6 also include: the deviation between the daily average upstream flow rate and the daily average downstream flow rate of the pump station is ≤10%; the daily average downstream liquid level is not higher than the daily average upstream liquid level; the correlation coefficient between the pipeline liquid level and flow rate during rainfall is ≥0.7; the time synchronization accuracy of multi-source spatiotemporal alignment is ≤1 minute, and the spatial matching accuracy is ≤10 meters; the data lineage report includes the data acquisition timestamp, equipment number, operator, processing algorithm version, and uncertainty calculation process, supports tracing the entire life cycle of each evaluation data point, and the report format is compatible with the import requirements of mainstream data management platforms.

10. A system for evaluating the operation and maintenance effectiveness of an urban sewage pipe network system, characterized in that, Includes the following modules: The three-dimensional evaluation framework module is used to construct an evaluation system from the following three dimensions: collection and transportation efficiency, pipeline network resilience, and public safety. The fault tolerance index quantification module designs a three-level quantification path of "standard-substitution-backup" for each evaluation index, and supports switching evaluation methods under different data conditions; The data quality perception and empowerment module includes a data quality scoring unit that assesses data quality from four dimensions: completeness, continuity, timeliness, and consistency. An improved entropy weighting unit based on data quality weighting is used to dynamically adjust indicator weights, combine weight calculation units, and integrate subjective and objective weights; The resilience approximation quantification module, based on monitoring data from key nodes, approximates the system's disturbance resistance and recovery capabilities through proxy variables and resilience integral algorithms. The multi-source evidence fusion risk assessment module uses a Bayesian inference model to integrate multi-source information such as inspection records, citizen complaints, and remote sensing data, and dynamically updates the probability of public safety risks. The economical monitoring network construction module adopts a three-layer perception architecture of "core-mobile-wide area" to achieve low-cost and high-coverage data acquisition; The data governance and transparent traceability module enables automatic data cleaning, multi-source alignment, intelligent interpolation, and full-process traceability. The performance-based payment linkage module designs a two-layer payment function based on the confidence interval of the evaluation results, linking data quality with payment.

Images