An Automated Scheduling Optimization Method for Engineering Project Inspection Plans

Abstract

This application discloses an automatic scheduling optimization method for engineering project inspection plans, relating to the field of scheduling optimization. The method includes: when the risk accumulation density exceeds a first preset threshold, splitting the original global inner wall inspection task into an emergency local inspection task covering the corresponding continuous pipe segment interval and a regular inspection task covering the remaining pipe segment, and generating split task data; when the risk accumulation density in multiple consecutive outer wall inspection cycles is lower than a second preset threshold, updating the trigger priority of the global inner wall inspection task and postponing its execution time, generating postponed task data; acquiring inspection cost data, including detection tool call cost, task waiting cost, and risk retention cost; and outputting a dynamic scheduling scheme including execution coordinate range, execution time, task priority, and detection tool type based on the split task data, postponed task data, and inspection cost data. This application has the effect of improving scheduling optimization efficiency.

Description

Technical Field

[0001] This application relates to the field of scheduling optimization technology, and in particular to an automatic scheduling optimization method for inspection plans of engineering projects. Background Technology

[0002] In the field of inspection scheduling for large-scale engineering projects such as long-distance pipelines, traditional planning methods mainly rely on fixed industry standards and periodic time nodes, treating high-cost internal wall inspection operations and low-cost external wall maintenance inspections as isolated and independent task flows. This scheduling logic exhibits obvious lag and rigidity, resulting in the generation of inspection plans based solely on historical experience in a single dimension, failing to effectively identify and utilize the physical correlations between different inspection tasks.

[0003] In related technologies, existing scheduling methods neglect the double-wall effect during the damage evolution process of the inner and outer walls of pipelines, making it impossible to translate failure clues of the outer wall anti-corrosion layer into decision-making basis for adjusting the timing of inner wall inspections. Under conventional scheduling models, the expensive task of inspecting the entire inner wall lacks dynamic resilience; even if only a few pipe sections have a high corrosion risk, the scheduling logic still forcibly executes a global task covering the entire line, resulting in significant cost losses and wasted equipment resources. Simultaneously, because existing scheduling methods lack a quantitative assessment of the decay of risk evidence over time, they cannot scientifically prioritize or postpone tasks based on the continuous and stable performance of environmental conditions. This current state of affairs, lacking dynamic control mechanisms at the task granularity level, leads to significant spatial overlap in engineering scheduling and a lack of sufficient flexibility in time allocation, making it difficult to cope with the complex and ever-changing spatial environment and risk characteristics of long-distance transportation projects, thus reducing scheduling optimization efficiency and warranting improvement. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this application provides an automatic scheduling optimization method for engineering project inspection plans.

[0005] This application provides an automatic scheduling optimization method for engineering project inspection plans, including the following steps: The system obtains the electric field gradient of the outer wall anti-corrosion layer, the damage level of the outer wall anti-corrosion layer, the historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift time since the previous inner wall inspection task for each pipe section in the pipeline project to be inspected. The system then performs time-series alignment according to the spatial coordinates of the pipe sections to generate the health fingerprint vector of each pipe section. Evidence weight decay factor is generated based on drift duration, and the evidence weight decay factor is used to correct the electric field gradient of the outer wall anti-corrosion layer and the damage level of the outer wall anti-corrosion layer to obtain the corrected outer wall state data. The corrected external wall condition data is coupled with the historical corrosion rate of the internal wall to generate the coupling risk value of the internal wall characteristics of each pipe section. Based on the spatial continuity of adjacent pipe segments, the risk values ​​of inner wall feature coupling are aggregated in intervals to obtain the risk accumulation density of continuous pipe segment intervals. When the risk accumulation density exceeds the first preset threshold, the original global inner wall inspection task is split into an emergency local inspection task covering the corresponding continuous pipe section and a regular inspection task covering the remaining pipe section, and the split task data is generated. When the cumulative risk density is lower than the second preset threshold in multiple consecutive external wall inspection cycles, the trigger priority of the global internal wall inspection task is updated and its execution time is postponed, and the postponed task data is generated. Obtain inspection cost data, which includes the cost of calling up detection tools, the cost of waiting for tasks, and the cost of retaining risks. Based on the split task data, the postponed task data, and the inspection cost data, a dynamic scheduling scheme is output, which includes the execution coordinate range, execution time, task priority, and detection tool type.

[0006] In summary, this application includes at least one of the following beneficial technical effects: 1. This application provides an automatic scheduling optimization method for engineering project inspection plans. By aligning the spatial coordinates of the electric field gradient of the outer wall anti-corrosion layer, the damage level of the outer wall anti-corrosion layer, the historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift time since the previous inner wall inspection task for each pipe section, and encapsulating them into a health fingerprint vector reflecting the real-time status of the pipe section, an evidence weight decay factor is introduced to dynamically correct the outer wall status. This allows the coupled calculation process to adaptively adjust according to the freshness and validity of the data, thus effectively avoiding the assessment distortion problem caused by ignoring the decay of data over time in traditional pipeline risk assessment. By deeply coupling the corrected outer wall status data with the historical corrosion law of the inner wall, the generated inner wall feature coupled risk value can accurately capture the inducing effect of external environmental degradation on the acceleration of internal corrosion, greatly improving the sensitivity and objectivity of risk judgment. This transforms the characterization of pipeline health status from a single-dimensional static record to a high-precision simulation of multi-field coupled dynamic evolution, laying a precise data foundation for subsequent differentiated scheduling and big data management. 2. By leveraging the spatial continuity of risk values ​​coupled with inner wall features for interval aggregation, and constructing a dual-threshold intelligent task splitting and trigger priority adjustment mechanism based on risk accumulation density, the original global inspection tasks are reconstructed into emergency local tasks or routine tasks with delayed execution according to the degree of risk concentration. The dynamic scheduling scheme is output by comprehensively considering the cost of calling detection tools, task waiting costs, and risk retention costs. This not only enables the rapid splitting of emergency tasks in high-risk connected areas to eliminate safety hazards and prevent accidents, but also effectively avoids unnecessary over-maintenance during stable periods when the risk level is consistently below the safety threshold through priority decay and time delay. Thus, while ensuring the inherent safety of the entire pipeline, the optimal allocation of inspection resources in the spatiotemporal dimension is achieved, significantly reducing the comprehensive operation and maintenance costs of long-distance pipelines and improving the scientific level of decision-making in pipeline network big data management and digital management. Attached Figure Description

[0007] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a flowchart illustrating the method for automatically scheduling and optimizing the inspection plan of an engineering project according to an embodiment of this application.

[0009] Figure 2 This is a schematic diagram illustrating the implementation logic of the health fingerprint vector in an embodiment of this application.

[0010] Figure 3 This is a schematic diagram illustrating the implementation logic of risk accumulation density in an embodiment of this application.

[0011] Figure 4 This is a schematic diagram illustrating the implementation logic of the dynamic scheduling scheme in this application embodiment. Detailed Implementation

[0012] The following description, in conjunction with the implementation of the present invention, is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention, and all such modifications and additions should fall within the protection scope of the present invention.

[0013] Example This application discloses an automatic scheduling optimization method for engineering project inspection plans.

[0014] Reference Figure 1-4 An automatic scheduling optimization method for engineering project inspection plans includes the following steps: The system obtains the electric field gradient of the outer wall anti-corrosion layer, the damage level of the outer wall anti-corrosion layer, the historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift time since the previous inner wall inspection task for each pipe section in the pipeline project to be inspected. The system then performs time-series alignment according to the spatial coordinates of the pipe sections to generate the health fingerprint vector of each pipe section. Among them, for the external wall status parameters, the electric field gradient of the external wall anti-corrosion layer, which characterizes the surface potential distribution, is obtained by retrieving the field data collected by the external wall inspection equipment such as buried pipeline current mapping or DC voltage gradient. Based on the measured voltage drop ratio and the defect evaluation standard, the system logic determines and extracts the corresponding external wall anti-corrosion layer damage level. For the historical parameters of the inner wall, by accessing the database of the pipeline integrity management platform, the historical report of the most recent inner wall inspection operation (such as magnetic flux leakage detection or ultrasonic detection) of the pipe section is retrieved. The previous inner wall inspection results containing defect depth, location and wall thickness reduction are read from the report. Combined with the detection data of the pipe section in an earlier time period, the historical corrosion rate of the inner wall of the pipe section is calculated by dividing the wall thickness loss increment by the time interval between the two inspections. Finally, for the parameters reflecting the timeliness of the data, the system time of the current external wall inspection task is subtracted from the completion time of the previous internal wall inspection task recorded in the database to obtain the time span of each pipe segment since the last internal inspection, that is, the drift time since the last internal wall inspection task. Through the above detection acquisition, historical database reading and time sequence logic calculation, the full collection of multiple physical quantity status data of the inner and outer walls of the pipe segment is realized.

[0015] In the temporal alignment based on pipe segment spatial coordinates, temporal alignment specifically means that since environmental data such as the electric field gradient of the outer wall anti-corrosion layer and internal detection data such as the historical corrosion rate of the inner wall come from different inspection cycles and technical methods, their acquisition time points and sampling frequencies differ significantly. Therefore, it is necessary to use the fixed physical spatial coordinates of the pipe segment as a unified anchor point to map multi-source heterogeneous data belonging to the same physical pipe segment but carrying different timestamps onto the same reference time axis for synchronous sorting. Specifically, during data integration, the system will clarify the absolute time relationship between the acquisition time of the current outer wall status data and the previous on-site detection time of the inner wall data, giving multi-dimensional data features a unified and standardized time label. The above operation not only eliminates the temporal dimension confusion caused by asynchronous acquisition and provides a strict time scale for accurately calculating drift duration, but also ensures that in the subsequent generation of health fingerprint vectors and the coupling calculation of inner and outer wall features, the actual superposition effect of the time decay of the inner wall historical data and the current outer wall deterioration state can be accurately measured within a logically consistent time frame.

[0016] Furthermore, by performing temporal alignment according to the spatial coordinates of the pipe segments, a health fingerprint vector for each pipe segment is generated, including: Read the coordinates and acquisition time of abnormal points in the external wall inspection record, and convert the coordinates of abnormal points into the corresponding pipe section number; Read the inspection mileage, historical corrosion rate of the inner wall and the previous inner wall inspection results from the internal wall inspection history, and convert the inspection mileage into the corresponding pipe section number; Using the pipe segment number as the associated field, the electric field gradient of the outer wall anti-corrosion layer, the damage level of the outer wall anti-corrosion layer, the historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift duration under the same pipe segment number are merged. Based on the time difference between the data collection time and the previous inner wall inspection time, the merged data is time-stamped to form a health fingerprint vector for each pipe section.

[0017] In a specific embodiment, the implementation logic for time-series alignment based on pipe segment spatial coordinates and generation of health fingerprint vectors is as follows: First, the external wall inspection record is retrieved through the data interface. The external wall inspection record specifically includes feature information such as the coordinates of abnormal points, the time of abnormal point detection, the detection voltage difference, and the corresponding anti-corrosion layer peeling strength. During this process, the latitude and longitude data of the abnormal point coordinates are converted into the cumulative mileage value along the pipeline using a preset engineering geographic information mapping table. Based on the preset length segmentation rules of the pipeline, the cumulative mileage value is placed into the corresponding discretization interval, thereby realizing the accurate conversion of the abnormal point coordinates into the corresponding pipe segment number.

[0018] Simultaneously, the system reads the internal wall inspection history database, extracting data including the inspection mileage reported by the internal detector, the calculated historical corrosion rate of the internal wall for each pipe segment, and the percentage of residual wall thickness from the previous internal wall inspection results. Using the same mileage conversion logic, the internal wall inspection mileage is mapped to the standard pipe segment numbering system, ensuring spatial consistency between internal and external wall data. Based on spatial alignment, using the pipe segment number as the unique primary key, the database aggregation operator merges the external wall anti-corrosion layer electric field gradient, external wall anti-corrosion layer damage level, historical internal wall corrosion rate, previous internal wall inspection results, and automatically calculated drift duration across tables under the same pipe segment number dimension. The drift duration refers to the time difference between the current external wall inspection time and the time of the most recent internal wall inspection completion for that pipe segment.

[0019] Finally, according to the time step between the collection time and the previous inner wall inspection time, a timestamp index is attached to each group of aggregated data, and the above multi-dimensional feature parameters are encapsulated into a standardized numerical sequence according to the preset vector dimension order, thereby forming a health fingerprint vector of each pipe segment under specific spatiotemporal coordinates, providing a structured data foundation for subsequent risk coupling calculation.

[0020] In a specific implementation, a pre-defined engineering geographic information mapping table serves as the spatiotemporal index benchmark in the pipeline digital twin system. It records the nonlinear mapping relationship between the latitude and longitude coordinates, altitude, and cumulative mileage of the pipeline's physical centerline. This table is obtained by retrieving as-built drawings, survey records, and raw GIS (Geographic Information System) data from the pipeline project. During the initialization phase, various feature points along the pipeline (such as elbows, valve chambers, tees, and pile locations) are used as control anchor points. A coordinate-mileage transformation model covering the entire pipeline is established using piecewise linear interpolation or curve fitting algorithms. In this embodiment, when the coordinates of an anomaly point with GPS timestamps and location information are received from external inspection equipment, the mapping table is retrieved in real time. The discrete geographic coordinate points are projected onto the pipeline centerline, and the precise linear cumulative mileage of that point along the entire pipeline is calculated, thereby eliminating errors between geographic distance and physical mileage caused by terrain undulations or the meandering of the pipeline.

[0021] The preset length segmentation rule is a logical discretization standard defined for the operation and maintenance management of long-distance pipelines, used to divide continuous pipelines, some hundreds of kilometers long, into statistical units. This rule is set based on operation and maintenance specifications, valve chamber spacing, or specific physical protection units. In terms of acquisition method, it is preset by operation and maintenance personnel in the system configuration interface according to the required detection accuracy; for example, setting each 50-meter or 100-meter section as a standard evaluation segment. In this embodiment, this rule serves as a data bucket allocation mechanism, performing modular arithmetic or interval determination on the linear cumulative mileage obtained through the mapping table, mapping continuous mileage values ​​to a unique, serialized segment number space. In this way, it is possible to aggregate and classify point data from different batches, sources, and with varying collection densities of external wall inspections and continuous segment data from internal wall inspections according to the same physical intervals.

[0022] By adopting the above technical solution, the dimensionality reduction from external geographic space to pipeline linear space is effectively solved by the preset engineering geographic information mapping table, while the preset length segmentation rule solves the attribution determination from linear continuous space to logical discrete unit. The synergistic application of the two ensures that when the system constructs the health fingerprint vector, it can logically anchor the external wall anomaly points with inherent deviations in positioning accuracy and the internal wall corrosion data to the same physical pipe segment entity, thereby providing a reliable coordinate basis for the subsequent splitting operator to accurately locate and segment specific high-risk areas.

[0023] Evidence weight decay factor is generated based on drift duration, and the evidence weight decay factor is used to correct the electric field gradient of the outer wall anti-corrosion layer and the damage level of the outer wall anti-corrosion layer to obtain the corrected outer wall state data. Furthermore, an evidence weight attenuation factor is generated based on the drift duration, and this factor is used to correct the electric field gradient of the outer wall anti-corrosion layer and the damage level of the outer wall anti-corrosion layer, including: The drift duration of each pipe segment is extracted based on the health fingerprint vector of each pipe segment, and the evidence weight attenuation factor is determined based on the position of the drift duration in the preset time segment. The corrected electric field gradient data is obtained based on the electric field gradient of the outer wall anti-corrosion layer and the evidence weight attenuation factor, and the corrected damage level data is obtained based on the damage level of the outer wall anti-corrosion layer and the evidence weight attenuation factor. The corrected electric field gradient data and the corrected damage level data are combined into corrected outer wall condition data.

[0024] In a specific embodiment, the detailed logic for generating an evidence weight attenuation factor based on the drift duration and correcting the data is as follows: First, by parsing the health fingerprint vectors generated for each pipe segment, the drift duration reflecting the timeliness characteristics of the pipe segment is extracted, i.e., the time span from the completion of the last inner wall inspection task to the current outer wall inspection data collection. Subsequently, the extracted drift duration is incorporated into a preset time segmentation strategy. This preset time segmentation is based on multiple time intervals pre-defined according to the pipeline corrosion failure evolution cycle. For example, a year is divided into an initial stable period, a mid-term fluctuating period, and a late-term risk period. The specific interval into which the drift duration falls is determined, and the evidence weight attenuation factor is determined according to the nonlinear attenuation function corresponding to the interval. The evidence weight attenuation factor measures the immediate effectiveness of the outer wall inspection clues for predicting inner wall risks. Generally, the longer the drift duration, the lower the reference value of the old inner wall data, thus causing a corresponding shift in the weight of outer wall evidence in the coupled judgment.

[0025] After obtaining the evidence weight attenuation factor, the data correction stage begins: based on the electric field gradient data of the outer wall anti-corrosion layer, its value is multiplied by the evidence weight attenuation factor or gain compensation is performed to eliminate the data deviation caused by the detection time being too far from the original reference time, and the corrected electric field gradient data is obtained; simultaneously, based on the grading index of the damage level of the outer wall anti-corrosion layer, the magnitude of the damage level is adjusted using the evidence weight attenuation factor, and the qualitative damage level is transformed into corrected damage level data that can reflect the real-time risk intensity.

[0026] Finally, using a data encapsulation protocol, the corrected electric field gradient data and the corrected damage level data are logically aggregated according to a standard state description format, merging them into unified corrected outer wall state data. This process ensures that the originally acquired outer wall physical features can be adaptively calibrated over time, enabling the subsequently generated inner wall feature coupling risk values ​​to more accurately reflect the structural health status under the double-wall effect and avoiding scheduling misjudgments caused by data timeliness issues.

[0027] In a specific implementation, the nonlinear decay function corresponding to the interval is obtained by performing correlation analysis on historical operational big data of the pipeline project. Specifically, multiple sets of data on external anti-corrosion layer damage from historical archives are retrieved, and spatiotemporal matching records of actual corrosion points detected during subsequent internal pipe cleaning are used to statistically analyze the prediction accuracy of external wall anomalies on internal wall corrosion under different drift durations. Through regression analysis modeling, the slope and curvature of the prediction accuracy decreasing over time are determined, thereby fitting and generating a nonlinear decay curve.

[0028] In the application of this application embodiment, the specific execution logic of the function is as follows: After the system extracts the drift duration of a certain pipe segment through the health fingerprint vector, it first determines which preset time segment the duration falls within. If the drift duration is in the initial stable phase (such as a very short time after inspection), the nonlinear decay function outputs a high weighting coefficient, and at this time, the outer wall status data is regarded as an immediate supplement to the inner wall status. As the drift duration enters the mid-term fluctuation phase, the function shows an exponential downward trend. The function calculates an decay ratio between zero and one to weaken those signals that may have lost their reference value due to environmental self-repair or data staleness.

[0029] Specifically, in order to accurately adapt to the preset time segmentation strategy and the corresponding nonlinear decay function of the interval, the system adopts the following formula for calculating the segmented evidence weight decay factor: ; in, Represented as the evidence weight attenuation factor, its physical meaning lies in measuring the immediate effectiveness of external wall inspection clues in predicting internal wall risks. It is used to adjust and correct the magnitude of the electric field gradient and damage level of the external wall anti-corrosion layer. The value range is (0,1]. When the drift duration is 0, the value of the evidence weight decay factor is 1, which means that the internal inspection data is completely reliable. As the drift duration increases, the evidence weight decay factor approaches 0, which means that the reference value of the old data disappears.

[0030] The drift duration refers to the time span from the completion of the last inner wall inspection task to the current outer wall inspection data collection. The value range is [0,20], and the unit is years. In long-distance pipeline management, the interval between two large-scale inner inspections is usually 5 to 10 years. 20 years is considered an extremely high-risk old data range.

[0031] , These are time segmentation thresholds, which are pre-set boundary values ​​based on the pipeline corrosion failure evolution cycle. They are used to precisely divide the drift time span into multiple specific intervals: an initial stable segment, a mid-term fluctuating segment, and a late-stage risk segment. The value range is [1,5] years, which represents that the data is still in the initial stable phase with high reliability. The value range is usually in the range of (5, 12] years, representing the data entering a medium-term fluctuation period, exceeding This is then classified as a later-stage risk segment.

[0032] , , This is expressed as the interval attenuation sensitivity coefficient, which is the nonlinear attenuation rate control parameter corresponding to different time periods. And it is set... < < This indicates that as the drift time enters the later fluctuation or risk phase, the reference value of the old inner wall data decreases rapidly. Therefore, the weight of the outer wall evidence in the final coupling judgment must shift accordingly. The value range is [0.01, 0.08], ensuring that the initial data decays slowly; The value range is [0.08, 0.20], which increases the uncertainty weight in the intermediate stage; The value range is [0.20, 0.50], which ensures that the weight of old data drops rapidly in the later stages, forcing the system to rely more on the real-time collected outer wall features.

[0033] , It is represented as the interval smoothing gain coefficient, with a value range of [0.8, 1.2]. It can be obtained by fitting historical data and is used as a compensation term in the nonlinear decay function. It aims to ensure that the value of the decay factor can maintain physical and logical continuity or make reasonable step adjustments when the drift time crosses different time segment boundaries (such as from the initial segment to the middle segment). e is a natural constant.

[0034] By adopting the above technical solution, the problem of data spatiotemporal misalignment in long-distance pipeline inspection is effectively solved by using a nonlinear decay function. This allows the scheduling system to assign completely different decision priorities when faced with external wall clues collected six months ago and external wall clues collected one week ago, ensuring that the scheduling plan is always generated based on the most timely and valuable evidence chain.

[0035] The corrected external wall condition data is coupled with the historical corrosion rate of the internal wall to generate the coupling risk value of the internal wall characteristics of each pipe section. Furthermore, the corrected external wall condition data is coupled with the historical corrosion rate of the internal wall to generate the coupling risk value of the internal wall characteristics for each pipe segment, including: Extract the corrected electric field gradient data and the corrected damage level data from the corrected outer wall condition data; The electrical induced increment, reflecting the degree of loss of the pipe section protection current, is determined based on the corrected electric field gradient data, and the physical penetration coefficient, reflecting the degree of entry of the external medium, is determined based on the corrected damage level data. The external environmental corrosion stress index of each pipe section is obtained by weighted synthesis of electrical induced increment and physical penetration coefficient. The external environmental corrosion stress index is used to map and correct the historical corrosion rate of the inner wall to obtain the dynamic corrosion acceleration factor. The dynamic corrosion acceleration factor and drift time are cumulatively calculated to generate the inner wall characteristic coupling risk value for each pipe section.

[0036] In a specific embodiment, the detailed steps for performing the coupled calculation of the corrected outer wall condition data and the historical corrosion rate of the inner wall to generate a risk value are as follows: First, through the data parsing engine, the corrected electric field gradient data and the corrected damage level data are accurately extracted from the encapsulated corrected outer wall condition data according to the field index.

[0037] Next, based on the corrected electric field gradient data, the preset electrochemical loss mapping function is retrieved, and the electrical induction increment reflecting the leakage intensity of the cathodic protection current due to defects in the outer wall anti-corrosion layer is calculated. Simultaneously, based on the corrected damage level data and combined with the medium permeability characteristics of the environment in which the pipe section is located, the physical penetration coefficient reflecting the ease with which corrosive media such as moisture and oxygen penetrate the anti-corrosion layer and enter the surface of the pipe is determined.

[0038] Subsequently, the electrical induced increment and the physical penetration coefficient are fused to obtain the external environmental corrosion stress index for each pipe section, which comprehensively reflects the erosion pressure exerted on the pipe body by external environmental failure. Based on this, using mapping correction logic, this index is applied as a regulating variable to the read historical corrosion rate of the inner wall. By correcting the deviation of historical data under the current environmental stress, a dynamic corrosion acceleration factor reflecting the current double-wall effect is obtained.

[0039] In the specific implementation process, the electrical induced increment and the physical penetration coefficient are combined to obtain the external environmental corrosion stress index for each pipe section: ; in, It is expressed as the external environmental corrosion stress index. This index comprehensively quantifies the potential for comprehensive corrosion damage to the pipeline body caused by the current external environment of the pipe section. The higher the value, the stronger the catalytic and stressing effect of the comprehensive external deterioration environment caused by the failure of the anti-corrosion layer on the continued thinning of the pipe wall. Expressed as an electrically induced increment, it characterizes the accelerating effect of external potential anomalies (such as stray current interference, cathodic protection current loss, etc.) on electrochemical corrosion reactions. This increment represents the additional corrosion energy induced by purely electrical environmental anomalies.

[0040] Expressed as the physical penetration coefficient, it characterizes the ease and extent to which corrosive media (such as groundwater, soil corrosive ions, dissolved oxygen, etc.) in the surrounding environment penetrate the failed anti-corrosion layer and reach the exposed metal surface of the pipe wall. The higher the damage level, the more thorough the physical channel for the medium to reach the metal, and the larger the coefficient value.

[0041] and Dynamically allocated weights are pre-configured based on the macroscopic environmental characteristics of the pipeline segment (satisfying) If the pipeline is laid in an area with strong stray currents, such as near subway lines or high-voltage power lines, the system will automatically adjust the height based on big data management logic. If the pipe section is located in a highly corrosive soil environment such as high water content or high salinity, the direct erosion by physical media will be more prominent, thus requiring a higher [pressure level]. The introduction of these two weights enables the system to accurately quantify the differences in pipeline failures caused by electrical and physical factors under different geological or interference environments.

[0042] Finally, the dynamic corrosion acceleration factor is multiplied by the corresponding drift time. The logic behind this is to estimate the potential damage increment on the inner wall within this timeframe by accumulating the current acceleration rate and the time span since the last inspection, thereby generating the inner wall characteristic coupling risk value for each pipe segment. This risk value not only includes the historical corrosion patterns of the inner wall itself but also couples with the dynamic risk factors caused by the failure of the outer wall protection, providing physically meaningful decision data for subsequent task splitting and scheduling optimization.

[0043] In this embodiment, the preset electrochemical loss mapping function is a mathematical model describing the quantitative conversion relationship between the change in the potential gradient of the outer wall of the pipe section and the cathodic protection current leakage intensity. When the anti-corrosion layer is damaged, the current originally used to protect the pipe body will be lost to the ground through the damage point, causing the electric field gradient to be distorted. The preset electrochemical loss mapping function is expressed as a formula simplified from the electrochemical polarization curve. It is obtained by performing polarization current simulation tests on pipe sections with different degrees of damage in the experimental section, and fitting it with the design parameters of the cathodic protection system adopted in the project (such as output current and soil resistivity benchmark value), and finally determining the proportional coefficient between the potential gradient magnitude and the current density loss. In the application of this embodiment, the corrected electric field gradient data is substituted into the function as an input variable, and then the corresponding current compensation gap is mapped according to the current gradient intensity, thereby calculating and generating the electrical induced increment, thereby quantifying the degree of loss of electrical protection capability.

[0044] In determining the physical infiltration coefficient, a multi-dimensional data cross-judgment logic was adopted, focusing on combining the medium permeability characteristics of the environment in which the pipe section is located. First, by retrieving the environmental layer in the engineering geographic information system (GIS), environmental medium data corresponding to each pipe section was obtained, including soil type (such as clay, sand, rock), groundwater level, soil moisture content, and environmental pH, etc., and the above characteristics were converted into a quantitative environmental permeability level according to the preset environmental permeability evaluation matrix.

[0045] The specific determination logic is as follows: A two-dimensional relational logic table is established, with the corrected damage level data as the horizontal axis and the environmental permeability level as the vertical axis. In this logic table, even if two pipe sections have the same damage level, if one is in a highly permeable sandy soil environment with abundant groundwater (high environmental permeability level), the system will assign it a higher weighted score based on the matrix intersection point; conversely, if it is in a dry, low-permeability clay environment, its score will be relatively lower. In this way, the static damage description is combined with the dynamic environmental medium inducing depth to calculate the physical penetration coefficient. This coefficient not only reflects the degree of damage in the geometric dimensions of the anti-corrosion layer, but also characterizes the physical flux of external corrosive media reaching the pipe surface through the damage point under the current environmental pressure, thus providing accurate physical parameter support for the subsequent calculation of the external environmental corrosion stress index.

[0046] In practical implementation, the functional relationship between mapping correction and dynamic corrosion acceleration factor is achieved by using a nonlinear amplification mapping correction function based on the natural index. The external environmental corrosion stress index is used as the excited variable to correct the historical corrosion rate of the inner wall. The specific formula is as follows: ; in, It is expressed as a dynamic corrosion accelerator factor, which characterizes the real-time corrosion rate that the inner wall of the pipe section is actually experiencing under the current environmental stress of environmental degradation, after being accelerated.

[0047] It represents the historical corrosion rate of the inner wall, which is the inherent corrosion evolution trend of the pipe section under the background condition.

[0048] It is expressed as the external environmental corrosion stress index. The larger the external environmental corrosion stress index value, the stronger the internal and external pinch effect generated by the external environment.

[0049] It is represented as the environmental stress sensitivity mapping coefficient, used to characterize the sensitivity of a specific pipe material or a specific transport medium to changes in external environmental stress, ensuring that the mapping process conforms to the objective laws of material dynamics of pipe electrochemical corrosion, and e is a natural constant.

[0050] After deriving the dynamic corrosion acceleration factor, and combining the timeliness of the data with the initial defect state, an inner wall characteristic coupling risk value, characterizing the current true degree of danger of the pipe section, is generated through time calculation of integral accumulation. The specific formula is as follows: ; in, This is represented as the inner wall feature coupling risk value, a quantified dimensionless risk score. This value will be directly used to compare with a preset threshold, thereby triggering task splitting or delayed scheduling.

[0051] This represents the initial defect baseline from the previous internal inspection, and is the maximum defect depth or wall thickness reduction of this pipe section that was found during the previous internal wall inspection, retrieved from the database. Represented as a dynamic corrosion acceleration factor; This is expressed as drift duration. This product term represents the incremental corrosion depth caused by external wall deterioration during this drift period without on-site internal inspection.

[0052] The risk normalization adjustment coefficient is used to convert the physical depth accumulation value into a standardized risk rating score in the range of 0-1, so that the big data management system can perform global sorting and task decision-making.

[0053] Based on the spatial continuity of adjacent pipe segments, the risk values ​​of inner wall feature coupling are aggregated in intervals to obtain the risk accumulation density of continuous pipe segment intervals. Furthermore, based on the spatial continuity of adjacent pipe segments, the risk values ​​of inner wall feature coupling are aggregated across intervals to obtain the cumulative risk density of continuous pipe segment intervals, including: Read the characteristic coupling risk value of the inner wall of each pipe segment according to the order of the pipe segment numbers; Pipe segments whose inner wall feature coupling risk values ​​reach a preset suspicious threshold are marked as suspicious pipe segments; Adjacent suspicious pipe segments are continuously merged to generate suspicious pipe segment intervals; The total risk value of the inner wall feature coupling and the length of the interval are calculated in each suspicious pipe segment interval. The total risk value of the inner wall feature coupling is then normalized according to the interval length to obtain the risk accumulation density of the suspicious pipe segment interval.

[0054] In a specific embodiment, the detailed logic for performing interval aggregation of inner wall feature coupling risk values ​​and calculation of risk accumulation density is as follows: First, through a database query command, the inner wall feature coupling risk values ​​corresponding to each pipe segment are read sequentially from the starting station to the ending station according to the arrangement order of the pipe segment numbers, ensuring that the topological relationship of the data in the geographic spatial dimension is maintained.

[0055] Subsequently, the read inner wall feature coupling risk value is compared with the preset suspicious threshold in real time. The preset suspicious threshold is a key limit set according to the minimum failure probability benchmark defined in the pipeline safety evaluation standard. Any pipe section whose inner wall feature coupling risk value reaches or exceeds the threshold is assigned a monitoring label at the logic layer and uniformly marked as a suspicious pipe section.

[0056] Next, the aforementioned discrete marked suspicious pipe segments are processed. By searching the continuity of the pipe segment numbers, it is determined whether adjacent suspicious pipe segments are geographically connected. If the numbers of two or more adjacent suspicious pipe segments are continuously distributed, a continuity merging operation is performed on them, thereby converging scattered risk points into suspicious pipe segment intervals with spatial practical significance, and automatically recording the start and end mileage of the interval.

[0057] After the interval is generated, the internal wall characteristic coupling risk values ​​of all included pipe segments within the interval are accumulated to obtain the total internal wall characteristic coupling risk value of the region. At the same time, the physical span between the starting point and the ending point within the interval is calculated, i.e., the interval length.

[0058] Finally, the total risk values ​​of the acquired inner wall features are divided according to the interval length to obtain the degree of risk concentration per unit length, i.e., the risk accumulation density. The physical meaning of the risk accumulation density is to eliminate the interference of interval length on risk assessment, enabling the system to identify those local high-risk areas with highly concentrated risk values, providing key quantitative basis for subsequent judgment on whether to trigger the task split logic.

[0059] In practical implementation, the specific formula for calculating the cumulative risk density is as follows: ; in, The interval length is... Represented as risk accumulation density, its physical meaning lies in eliminating the interference of different physical spans of merged intervals on risk assessment. It can intuitively quantify the risk density of local high-risk areas and is the core quantitative basis for subsequent judgment on whether to split off an emergency inspection task from that area. The value range is [value range missing]. This value is a real number that is always greater than zero. Its specific upper limit depends on the maximum risk definition value of a single pipe segment and the smallest granularity of the pipe segment division. The larger the value, the more concentrated and critical the safety hazards in that local area are.

[0060] Expressed as the total risk value of inner wall feature coupling, it refers to all included risk values ​​within the continuous suspicious pipe segment interval generated by spatial clustering. The sum of the coupling risk values ​​of the inner wall characteristics of each monitored pipe section, with a value range of [value range missing]. Due to the accumulated single-point risk value All have reached or exceeded the preset suspicious threshold, therefore the total value must be greater than or equal to ( ). (Preset suspicious threshold).

[0061] This is expressed as the characteristic coupling risk value of the inner wall of a single pipe segment, with a value range of [value missing]. The preset suspicious threshold is the lower limit of the benchmark, representing the lowest failure probability benchmark, which can be obtained by fitting historical data. This is the normalized score for the maximum physical limit risk.

[0062] When the risk accumulation density exceeds the first preset threshold, the original global inner wall inspection task is split into an emergency local inspection task covering the corresponding continuous pipe section and a regular inspection task covering the remaining pipe section, and the split task data is generated. Furthermore, the original global inner wall inspection task is split into emergency local inspection tasks covering corresponding continuous pipe sections and regular inspection tasks covering the remaining pipe sections, and the split task data is generated, including: Extract continuous pipe segment intervals where the risk accumulation density exceeds a first preset threshold, and read the start and end coordinates of the continuous pipe segment intervals; The remaining pipe segment range is obtained by spatially subtracting the pipe segment coverage area from the continuous pipe segment interval in the original global inner wall inspection task. Emergency local inspection task data is generated based on the starting coordinates, ending coordinates, risk accumulation density, and current outer wall inspection cycle of the continuous pipe section. Based on the remaining pipe section range, the execution time of the original global internal wall inspection task, and the task priority of the original global internal wall inspection task, generate regular inspection task data; The data from emergency local inspection missions and routine inspection missions are merged to form split mission data.

[0063] In a specific embodiment, the detailed logic for splitting the original global inner wall inspection task and generating split task data is as follows: First, continuously connected pipe segments with a risk accumulation density exceeding a first preset threshold are extracted in real time from all aggregated pipe segment segments. The first preset threshold is a critical indicator in the engineering safety benchmark that requires immediate detection and intervention. For each identified high-risk segment, the starting and ending coordinates of the continuously connected pipe segment segment are accurately read through the spatial indexing engine, thereby locking the detection target area at the physical construction level.

[0064] Subsequently, the pre-set coverage range of the pipe sections covering the entire line in the original global inner wall inspection task is spatially deducted from the continuous pipe section interval. The logic is to extract the remaining pipe section range where the risk level is still under control through the interval elimination operation.

[0065] Next, based on the starting and ending coordinates of the continuous pipe segment interval, the quantified risk accumulation density, and the real-time timestamp of the current outer wall inspection cycle, the urgency weight of local inspections is calculated. Based on this, emergency local inspection task data containing specific inspection process requirements is generated, ensuring that high-risk sections can be prioritized for execution, bypassing the original cycle limitations. Simultaneously, based on the calculated remaining pipe segment range, the execution time and priority of the original global inner wall inspection tasks are inherited to generate regular inspection task data, maintaining the basic monitoring frequency of the large-scale pipeline network.

[0066] Finally, by using a data structure encapsulation protocol, the newly generated emergency local inspection task data and regular inspection task data are logically merged to form split task data containing multi-level task attributes. This process realizes a granular transformation from unified scheduling across the entire line to segmented differentiated scheduling, ensuring the dynamic allocation of inspection resources to local high-risk areas.

[0067] When the cumulative risk density is lower than the second preset threshold in multiple consecutive external wall inspection cycles, the trigger priority of the global internal wall inspection task is updated and its execution time is postponed, and the postponed task data is generated. Furthermore, update the trigger priority of the global inner wall inspection task and postpone its execution time, generating the postponed task data, including: Read the cumulative risk density corresponding to each cycle in the order of the external wall inspection cycle; Within multiple consecutive external wall inspection cycles, the cumulative risk density of each cycle is compared with a second preset threshold. When the cumulative risk density in each period is lower than the second preset threshold, a risk stability marker is generated. The trigger priority of the global inner wall inspection task is reduced based on the risk stability marker, and the execution time of the global inner wall inspection task is updated according to the preset time step. The updated global inner wall inspection task's execution coordinate range, execution time, task priority, and detection tool type are encapsulated to generate the subsequent task data.

[0068] In a specific embodiment, the detailed logic for executing the update of the global inner wall inspection task trigger priority and delaying its execution time is as follows: First, the risk accumulation density corresponding to each cycle is read sequentially according to the time sequence of the outer wall inspection cycle, in order to obtain continuous evidence of the safety status of the pipe section through long-term historical trend analysis.

[0069] Next, the system enters the stable state determination phase. Within a preset sliding time window, monitoring results from multiple consecutive external wall inspection cycles are extracted. The cumulative risk density of each cycle is compared one by one with a second preset threshold. The second preset threshold is typically set as a safety warning line significantly lower than the first preset threshold, indicating that the pipe section is in a stable state with extremely low risk. The system logic determines that the current corrosion development of the project has stagnated or is extremely slow only when the cumulative risk density of each cycle within the window is lower than the second preset threshold. Subsequently, a signal generation mechanism is triggered to generate a risk stability marker.

[0070] Subsequently, the system enters the dynamic correction phase of task scheduling parameters: Based on the generated risk stability markers, the system automatically retrieves the current list of tasks to be executed, locates the corresponding global inner wall inspection task, and calls the priority correction operator to reduce the trigger priority of the task according to a preset weight decay ratio, thus shifting its competitive position in the scheduling queue. Simultaneously, according to a preset time step—that is, the allowable safe extension time span calculated based on industry experience—an additive offset operation is performed on the execution time of the task, thereby updating the global inner wall inspection task.

[0071] Finally, the updated execution coordinate range, updated execution time, adjusted task priority, and compatible detection tool types of the global internal wall inspection tasks are structurally integrated to generate the postponed task data. This logic enables the scheduling scheme to adjust negative feedback to environmental feedback, effectively extending the cycle of high-cost inspection operations by scientifically identifying low-risk signals, and achieving optimal allocation of operation and maintenance costs.

[0072] In the specific implementation process, the determination method of the first and second preset thresholds is based on statistical algorithms calibrated using historical pipeline failure samples and the physical limits of the pipeline segment itself. First, the design and operation parameters of the target pipeline segment, the historical high-risk defect sample library, and the severity of failure consequences of the geographical environment along the pipeline are retrieved to deduce the risk extreme value boundary when the pipeline segment faces a critical state of leakage or rupture. A preset safety margin coefficient is then added downwards to calculate the first preset threshold, which represents the system safety red line and is used as the split judgment benchmark for forcibly triggering emergency local operations. At the same time, by fitting the probability density distribution of the risk values ​​of large-scale daily inspections of the pipeline segment and similar healthy pipeline segments during the long-term normal service stable period, the confidence upper limit of its background risk fluctuation or the benchmark characteristic value in a stable state is extracted. Based on this, the second preset threshold, which represents the system safety green line and is used to verify the feasibility of task postponement and resource release, is calibrated. Through this dual-threshold calculation mechanism based on historical objective data and extreme state deduction, a scientific and rigorous quantitative judgment boundary is provided for the dynamic scheduling scheme.

[0073] In this embodiment, the priority correction operator is the core logical unit for reorganizing the task sequence within the scheduling system. Essentially, it's a numerical processing mechanism based on multi-objective decision weights. The priority correction operator retrieves the original level parameters of the tasks to be executed from the task library and combines them with real-time generated risk status markers to downgrade the urgency of tasks in the global scheduling queue. The preset weight decay ratio is the core coefficient when the priority correction operator performs the correction, representing the quantification degree to which task priority is postponed under the premise of obtaining stable risk evidence.

[0074] The preset weight decay ratio is determined by outlier analysis and utility modeling of historical reliability maintenance data for pipeline projects. Specifically, safety feedback records of delayed internal inspection tasks under stable risk conditions over the past few years are retrieved, and the correlation between delay duration and structural reliability degradation is analyzed to calculate a proportional constant that ensures a safety margin while maximizing the saving of operation and maintenance resources. This proportional constant is set between zero and one and is stored in the background configuration parameter library during the system initialization phase.

[0075] In this embodiment, the application logic of the priority correction operator is as follows: When a risk stabilization marker is detected, the priority correction operator is automatically activated and directed to the global inner wall inspection task corresponding to the current pipe segment. The priority correction operator first extracts the current trigger priority of the global inner wall inspection task, and then retrieves the corresponding preset weight decay ratio from the parameter library. During the calculation process, the priority correction operator uses the current trigger priority as the base and performs a product operation with its weight decay ratio, and the result is used as the new priority weight of the task. In this way, the global inspection task that originally had a high execution priority due to its approaching time period will have its ranking level in the task pool decrease accordingly due to the continuous low level of environmental risk. This dynamic adjustment logic ensures that the system can identify those tasks that are due but have low risk, and give way to other urgently split local high-risk inspection tasks, thereby achieving the optimal redistribution of limited inspection resources across the entire line.

[0076] Obtain inspection cost data, which includes the cost of calling up detection tools, the cost of waiting for tasks, and the cost of retaining risks. Based on the split task data, the postponed task data, and the inspection cost data, a dynamic scheduling scheme is output, which includes the execution coordinate range, execution time, task priority, and detection tool type.

[0077] Furthermore, based on the split task data, the postponed task data, and the inspection cost data, a dynamic scheduling scheme is output, including: Extract emergency local inspection task data and regular inspection task data from the split task data, and read the execution coordinate range, execution time, task priority and detection tool type from the emergency local inspection task data and regular inspection task data; Read the execution coordinate range, execution time, task priority, and detection tool type of the global inner wall inspection task from the postponed task data; Read the inspection cost data to determine the cost of calling up inspection tools, the cost of waiting for tasks, and the cost of retaining risks. The data of emergency local inspection tasks, routine inspection tasks, and postponed tasks are sorted according to their execution time, and the order of tasks within the same execution period is adjusted according to task priority. By combining the cost of calling the detection tool, the cost of waiting for the task, and the cost of retaining the risk, the comprehensive cost corresponding to the adjusted task order is calculated. Select the order of tasks whose comprehensive cost value meets the preset cost conditions, and generate a dynamic scheduling scheme that includes the execution coordinate range, execution time, task priority and detection tool type.

[0078] In a specific embodiment, the dynamic scheduling scheme is implemented as follows: First, from the generated split task data, emergency local inspection task data and regular inspection task data are separated, and the global inner wall inspection tasks in the delayed task data are read simultaneously. From these, the execution coordinate range, execution time, task priority, and required detection tool type (such as magnetic leakage detectors or ultrasonic detectors of different precision) corresponding to each task to be executed are accurately parsed.

[0079] Subsequently, the inspection cost data was retrieved to obtain the costs of calling up inspection tools due to equipment transportation and personnel allocation, the costs of resource idleness and task waiting caused by work queuing, and the potential risk retention costs caused by high-risk pipelines not being handled in a timely manner. Among them, the risk retention costs increased non-linearly with time delay.

[0080] Based on this, all categories of tasks to be executed are first initially linearly sorted according to their execution time. For overlapping tasks that fall within the same execution time period, the internal order is adjusted according to the priority of the tasks to ensure that emergency local inspection tasks receive the highest execution priority when resources are limited.

[0081] Next, each set of arranged task sequences is substituted into the multi-objective cost function as input, and dynamically weighted summation is performed by combining the corresponding detection tool call cost, task waiting cost, and risk retention cost to calculate the comprehensive cost value corresponding to the adjusted task order. Its core logic is to find the optimal balance between eliminating high risks as soon as possible and minimizing the frequency of equipment deployment through quantitative evaluation.

[0082] Finally, the task sequence that satisfies the preset cost conditions (such as lowest total cost or compliance with budget constraints) is selected, and the execution coordinate range, execution time, task priority, and detection tool type under this optimal sequence are structurally encapsulated to generate a dynamic scheduling plan with on-site guidance. This process realizes the transformation from isolated task allocation to dynamic optimization and allocation of global resources, ensuring that the inspection plan has extremely high response efficiency and economic rationality when dealing with sudden risks.

[0083] In this embodiment, the multi-objective cost function is a comprehensive evaluation model used to quantitatively assess the merits of different scheduling sequences. Its core function is to transform multiple, even conflicting, evaluation indicators into a single, comparable comprehensive cost value. The multi-objective cost function mathematically aggregates logically independent and dimensionally different detection tool call costs, task waiting costs, and risk retention costs through weighted summation. In terms of acquisition, this multi-objective cost function is established by combining industry experience models with historical engineering financial data. First, the baseline weight of the detection tool call cost is determined based on equipment leasing, logistics transportation, and labor cost databases. Second, the loss coefficient of task waiting costs is derived based on resource scheduling efficiency data analysis. Through correlation analysis of past accident losses and risk levels, an estimation model for risk retention costs is established; that is, the greater the cumulative risk density of a pipeline segment and the longer it remains in an undetected state, the more exponentially the risk retention cost increases. These parameters are weighted using the analytic hierarchy process (AHP) during the system configuration phase and ultimately encapsulated into callable mathematical function expressions.

[0084] In the application logic of this application embodiment, the specific execution process of the function is as follows: Each time the system generates a possible task arrangement order (i.e., a candidate scheduling scheme), it passes it as an input variable to the multi-objective cost function. The multi-objective cost function first calculates the overall tool call cost based on the frequency of task switching and tool type requirements in the sequence; then, based on the gaps between tasks on the time axis, it accumulates and calculates the task waiting cost caused by resource occupation; subsequently, it performs a time-series integral calculation, multiplying the waiting time of each task before execution in the schedule by its corresponding risk accumulation density to obtain the overall risk retention cost of the entire sequence.

[0085] Ultimately, the multi-objective cost function weights and sums these three costs, outputting the comprehensive cost value corresponding to the given sequence. Through iterative calculation and comparison of multiple candidate sequences, it logically favors sequences that allow high-risk tasks to execute as early as possible (to reduce risk retention costs) while also having compact task switching logic (to reduce invocation and waiting costs). In this way, the multi-objective cost function achieves optimal selection of inspection tasks under complex constraints, ensuring that the final dynamic scheduling scheme possesses optimal operational economics while guaranteeing absolute pipeline safety.

[0086] In the specific implementation process, the calculation formula for the comprehensive cost is as follows: ; in, This is the risk retention cost. This is represented as the comprehensive cost value, used to quantitatively evaluate the overall execution cost of the current task prioritization scheme. The system uses a traversal optimization algorithm to find the scheme with the minimum value, achieving an optimal balance between quickly eliminating high risks and minimizing equipment relocation frequency. The value range is... Furthermore, the smaller the value, the higher the overall benefit of the task scheduling plan.

[0087] It represents the total number of permutation tasks, used to indicate the total number of all tasks to be executed that participate in the current spatiotemporal sequence permutation logic, and its value range is positive integer.

[0088] This represents the cost of calling up detection tools, which is the economic and resource consumption cost incurred due to equipment transportation, personnel allocation, and switching between different pipe sections after analyzing the type of detection tools required for the task (such as magnetic leakage or ultrasonic).

[0089] This is represented as task waiting cost, which is the cost of queuing, idle resources, and management redundancy caused by multiple tasks falling into the same execution period and overlapping.

[0090] This represents the task priority / risk assessment weight, a weighted indicator precisely parsed from the task data. For urgent local inspection tasks, this value is extremely high, ensuring that they receive execution priority when resources are limited. This is expressed as a time delay, meaning the time difference between the actual scheduled execution time of a task and the original optimal preset time window for that task.

[0091] Represented as a nonlinear growth sensitivity coefficient, its value range is... Represented as dynamic weighting coefficients, with values ​​ranging from [value range missing]. And usually satisfy When system resources are extremely scarce, the speed will be increased. and When the system is extremely sensitive to security, the settings will be increased. The aforementioned nonlinear growth sensitivity coefficient and dynamic weighting coefficient can both be obtained by fitting historical data.

[0092] Furthermore, after outputting the dynamic scheduling plan, a scheduling feedback update step is also included, which includes: Obtain the actual internal wall corrosion results generated by the executed internal wall inspection task, and associate the actual internal wall corrosion results with the corresponding pipe section according to the execution coordinate range; Update the previous internal wall inspection results and historical internal wall corrosion rate of the corresponding pipe section using the actual internal wall corrosion results; The drift duration of the corresponding pipe section is recalculated based on the updated results of the previous internal wall inspection. The updated historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift duration are written into the health fingerprint vector, and the next round of inner wall feature coupling risk value calculation is triggered. The task priority and execution time of unexecuted tasks are adjusted based on the risk value of the inner wall feature coupling in the next round.

[0093] In a specific embodiment, the detailed logic of the scheduling feedback update step is as follows: First, after the inspection operation guided by the dynamic scheduling scheme is completed, the actual internal wall corrosion results generated by the internal wall inspection task are obtained from the original report uploaded by the detection equipment, including the actual wall thickness reduction and defect geometric features. The actual internal wall corrosion results are then accurately associated with the corresponding pipe segment in the physical topology according to the execution coordinate range.

[0094] Subsequently, the old data of the corresponding pipe section is replaced or corrected using the actual internal wall corrosion results, thereby updating the previous internal wall inspection results. The historical corrosion rate of the internal wall is recalculated and updated in combination with the corrosion increment detected this time and the time span, ensuring that the historical baseline data and the physical measured values ​​are highly synchronized.

[0095] Based on this, the drift duration of the corresponding pipe segment is recalculated according to the updated results of the previous inner wall inspection (i.e., using the latest inspection completion time as the new baseline). The drift duration will be reset or shortened upon completion of the inspection task. Next, the updated historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift duration are written as core fields into the health fingerprint vector of the pipe segment, immediately triggering the next round of inner wall feature coupling risk value calculation. In this closed-loop logic, the system uses the latest measured data as initial values ​​to calibrate the risk evolution curve originally generated based on speculation, thereby generating a more accurate next round of inner wall feature coupling risk value. Finally, based on the next round of inner wall feature coupling risk value, the task priority and execution time of unexecuted tasks in the task pool are adjusted in real time: if the measured corrosion is more severe than expected, the priority of the relevant tasks is increased and the execution time is advanced; otherwise, it is appropriately slowed down. This feedback mechanism ensures that the scheduling system has self-learning and adaptive capabilities, achieving a deep closed-loop update of inspection data and decision logic.

[0096] Furthermore, before generating the health fingerprint vector for each pipe segment, a data consistency processing step is included, which includes: Read the electric field gradient and damage level of the outer wall anti-corrosion layer of the same pipe section within the same outer wall inspection cycle; Calculate the difference between multiple electric field gradients of the outer wall anti-corrosion layer according to the adjacent relationship of the collection time, and mark the data whose difference exceeds the preset jump threshold as data to be verified. By using data from adjacent acquisition times of the same pipe segment to replace and correct the data to be verified, the corrected electric field gradient of the outer wall anti-corrosion layer is obtained. The corrected electric field gradient of the outer wall anti-corrosion layer and the corresponding damage level of the outer wall anti-corrosion layer are written into the health fingerprint vector.

[0097] In a specific embodiment, the logic for performing data consistency processing steps to ensure the quality of underlying data is as follows: First, the distributed task scheduler accurately reads multiple external wall anti-corrosion layer electric field gradients and external wall anti-corrosion layer damage levels of the same pipe segment within the same external wall inspection cycle from the underlying inspection database according to the pipe segment number. This process ensures that multiple measurement sampling points for the same physical object are completely collected.

[0098] Subsequently, the data are sorted in chronological order according to the adjacent relationship of the acquisition time, and the difference between the electric field gradients of multiple outer wall anti-corrosion layers is calculated in turn. The physical logic of this step is that the pipeline corrosion and electrical state should have physical continuity within a small time step. If a certain measuring point is affected by the instantaneous grounding resistance fluctuation of the sensor or environmental electromagnetic interference, causing the gradient reading to fluctuate violently and the difference to exceed the preset jump threshold, the system will logically determine that the reading does not have true representativeness and mark it as data to be verified.

[0099] Next, data from adjacent acquisition times of the same pipe section (such as the mean of adjacent valid measurement points before and after the data to be verified or the value estimated by linear interpolation) are used to replace and correct the data to be verified. This is achieved by eliminating isolated outlier noise through logical extrapolation of nearby spatiotemporal evidence, thereby obtaining the corrected electric field gradient of the outer wall anti-corrosion layer that truly reflects the distribution law of the surface electric potential.

[0100] Finally, the corrected electric field gradient of the outer wall anti-corrosion layer is time-aligned with the corresponding damage level of the outer wall anti-corrosion layer and written together into the health fingerprint vector. This preprocessing process eliminates the interference of random sensor errors on risk assessment, ensuring that all subsequent decisions regarding risk accumulation and task splitting are based on logically rigorous and consistent fundamental data.

[0101] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention, they should all fall within the protection scope of the present invention.

[0102] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0103] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A method for automatically scheduling and optimizing inspection plans for engineering projects, characterized in that, Includes the following steps: The system obtains the electric field gradient of the outer wall anti-corrosion layer, the damage level of the outer wall anti-corrosion layer, the historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift time since the previous inner wall inspection task for each pipe section in the pipeline project to be inspected. The system then performs time-series alignment according to the spatial coordinates of the pipe sections to generate the health fingerprint vector of each pipe section. Evidence weight decay factor is generated based on drift duration, and the evidence weight decay factor is used to correct the electric field gradient of the outer wall anti-corrosion layer and the damage level of the outer wall anti-corrosion layer to obtain the corrected outer wall state data. The corrected external wall condition data is coupled with the historical corrosion rate of the internal wall to generate the coupling risk value of the internal wall characteristics of each pipe section. Based on the spatial continuity of adjacent pipe segments, the risk values ​​of inner wall feature coupling are aggregated in intervals to obtain the risk accumulation density of continuous pipe segment intervals. When the risk accumulation density exceeds the first preset threshold, the original global inner wall inspection task is split into an emergency local inspection task covering the corresponding continuous pipe section and a regular inspection task covering the remaining pipe section, and the split task data is generated. When the cumulative risk density is lower than the second preset threshold in multiple consecutive external wall inspection cycles, the trigger priority of the global internal wall inspection task is updated and its execution time is postponed, and the postponed task data is generated. Obtain inspection cost data, which includes the cost of calling up detection tools, the cost of waiting for tasks, and the cost of retaining risks. Based on the split task data, the postponed task data, and the inspection cost data, a dynamic scheduling scheme is output, which includes the execution coordinate range, execution time, task priority, and detection tool type.

2. The automatic scheduling optimization method for engineering project inspection plans according to claim 1, characterized in that, And according to the spatial coordinates of the pipe segments, time-series alignment is performed to generate health fingerprint vectors for each pipe segment, including: Read the coordinates and acquisition time of abnormal points in the external wall inspection record, and convert the coordinates of abnormal points into the corresponding pipe section number; Read the inspection mileage, historical corrosion rate of the inner wall and the previous inner wall inspection results from the internal wall inspection history, and convert the inspection mileage into the corresponding pipe section number; Using the pipe segment number as the associated field, the electric field gradient of the outer wall anti-corrosion layer, the damage level of the outer wall anti-corrosion layer, the historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift duration under the same pipe segment number are merged. Based on the time difference between the data collection time and the previous inner wall inspection time, the merged data is time-stamped to form a health fingerprint vector for each pipe section.

3. The automatic scheduling optimization method for engineering project inspection plans according to claim 2, characterized in that, An evidence weight attenuation factor is generated based on the drift duration, and this factor is used to correct the electric field gradient and damage level of the outer wall anti-corrosion layer, including: The drift duration of each pipe segment is extracted based on the health fingerprint vector of each pipe segment, and the evidence weight attenuation factor is determined based on the position of the drift duration in the preset time segment. The corrected electric field gradient data is obtained based on the electric field gradient of the outer wall anti-corrosion layer and the evidence weight attenuation factor, and the corrected damage level data is obtained based on the damage level of the outer wall anti-corrosion layer and the evidence weight attenuation factor. The corrected electric field gradient data and the corrected damage level data are combined into corrected outer wall condition data.

4. The automatic scheduling optimization method for engineering project inspection plans according to claim 3, characterized in that, The corrected external wall condition data is coupled with the historical corrosion rate of the internal wall to generate the coupling risk value of the internal wall characteristics for each pipe segment, including: Extract the corrected electric field gradient data and the corrected damage level data from the corrected outer wall condition data; The electrical induced increment, reflecting the degree of loss of the pipe section protection current, is determined based on the corrected electric field gradient data, and the physical penetration coefficient, reflecting the degree of entry of the external medium, is determined based on the corrected damage level data. The external environmental corrosion stress index of each pipe section is obtained by weighted synthesis of electrical induced increment and physical penetration coefficient. The external environmental corrosion stress index is used to map and correct the historical corrosion rate of the inner wall to obtain the dynamic corrosion acceleration factor. The dynamic corrosion acceleration factor and drift time are cumulatively calculated to generate the inner wall characteristic coupling risk value for each pipe section.

5. The automatic scheduling optimization method for engineering project inspection plans according to claim 2, characterized in that, Based on the spatial continuity of adjacent pipe segments, the risk values ​​of inner wall feature coupling are aggregated in intervals to obtain the risk accumulation density of continuous pipe segment intervals, including: Read the characteristic coupling risk value of the inner wall of each pipe segment according to the order of the pipe segment numbers; Pipe segments whose inner wall feature coupling risk values ​​reach a preset suspicious threshold are marked as suspicious pipe segments; Adjacent suspicious pipe segments are continuously merged to generate suspicious pipe segment intervals; The total risk value of the inner wall feature coupling and the length of the interval are calculated in each suspicious pipe segment interval. The total risk value of the inner wall feature coupling is then normalized according to the interval length to obtain the risk accumulation density of the suspicious pipe segment interval.

6. The automatic scheduling optimization method for engineering project inspection plans according to claim 5, characterized in that, The original global inner wall inspection task was split into an emergency local inspection task covering the corresponding continuous pipe section and a regular inspection task covering the remaining pipe section, and the split task data was generated, including: Extract continuous pipe segment intervals where the risk accumulation density exceeds a first preset threshold, and read the start and end coordinates of the continuous pipe segment intervals; The remaining pipe segment range is obtained by spatially subtracting the pipe segment coverage area from the continuous pipe segment interval in the original global inner wall inspection task. Emergency local inspection task data is generated based on the starting coordinates, ending coordinates, risk accumulation density, and current outer wall inspection cycle of the continuous pipe section. Based on the remaining pipe section range, the execution time of the original global internal wall inspection task, and the task priority of the original global internal wall inspection task, generate regular inspection task data; The data from emergency local inspection missions and routine inspection missions are merged to form split mission data.

7. The automatic scheduling optimization method for engineering project inspection plans according to claim 6, characterized in that, Update the trigger priority of the global inner wall inspection task and postpone its execution time, generating the postponed task data, including: Read the cumulative risk density corresponding to each cycle in the order of the external wall inspection cycle; Within multiple consecutive external wall inspection cycles, the cumulative risk density of each cycle is compared with a second preset threshold. When the cumulative risk density in each period is lower than the second preset threshold, a risk stability marker is generated. The trigger priority of the global inner wall inspection task is reduced based on the risk stability marker, and the execution time of the global inner wall inspection task is updated according to the preset time step. The updated global inner wall inspection task's execution coordinate range, execution time, task priority, and detection tool type are encapsulated to generate the subsequent task data.

8. The automatic scheduling optimization method for engineering project inspection plans according to claim 7, characterized in that, Based on the split task data, the postponed task data, and the inspection cost data, a dynamic scheduling scheme is output, including: Extract emergency local inspection task data and regular inspection task data from the split task data, and read the execution coordinate range, execution time, task priority and detection tool type from the emergency local inspection task data and regular inspection task data; Read the execution coordinate range, execution time, task priority, and detection tool type of the global inner wall inspection task from the postponed task data; Read the inspection cost data to determine the cost of calling up inspection tools, the cost of waiting for tasks, and the cost of retaining risks. The data of emergency local inspection tasks, routine inspection tasks, and postponed tasks are sorted according to their execution time, and the order of tasks within the same execution period is adjusted according to task priority. By combining the cost of calling the detection tool, the cost of waiting for the task, and the cost of retaining the risk, the comprehensive cost corresponding to the adjusted task order is calculated. Select the order of tasks whose comprehensive cost value meets the preset cost conditions, and generate a dynamic scheduling scheme that includes the execution coordinate range, execution time, task priority and detection tool type.

9. The automatic scheduling optimization method for engineering project inspection plans according to claim 8, characterized in that, After outputting the dynamic scheduling plan, a scheduling feedback update step is also included, which includes: Obtain the actual internal wall corrosion results generated by the executed internal wall inspection task, and associate the actual internal wall corrosion results with the corresponding pipe section according to the execution coordinate range; Update the previous internal wall inspection results and historical internal wall corrosion rate of the corresponding pipe section using the actual internal wall corrosion results; The drift duration of the corresponding pipe section is recalculated based on the updated results of the previous internal wall inspection. The updated historical corrosion rate of the inner wall, the previous inner wall inspection results, and the drift duration are written into the health fingerprint vector, and the next round of inner wall feature coupling risk value calculation is triggered. The task priority and execution time of unexecuted tasks are adjusted based on the risk value of the inner wall feature coupling in the next round.

10. The automatic scheduling optimization method for engineering project inspection plans according to claim 1, characterized in that, Before generating the health fingerprint vectors for each pipe segment, a data consistency processing step is included, which includes: Read the electric field gradient and damage level of the outer wall anti-corrosion layer of the same pipe section within the same outer wall inspection cycle; Calculate the difference between multiple electric field gradients of the outer wall anti-corrosion layer according to the adjacent relationship of the collection time, and mark the data whose difference exceeds the preset jump threshold as data to be verified. By using data from adjacent acquisition times of the same pipe segment to replace and correct the data to be verified, the corrected electric field gradient of the outer wall anti-corrosion layer is obtained. The corrected electric field gradient of the outer wall anti-corrosion layer and the corresponding damage level of the outer wall anti-corrosion layer are written into the health fingerprint vector.

Images