Construction progress management method and system fusing underground pipeline digital model

Abstract

The application relates to the technical field of municipal engineering construction progress management, and discloses a construction progress management method and system fusing an underground pipeline digital model, which is suitable for a city water supply network reconstruction project. The method comprises the following steps: constructing an underground pipeline digital model fusing pipeline, geology, pipe material and construction constraint parameters and verifying the precision; generating an initial progress plan based on the model in combination with an improved critical path method and Monte Carlo simulation; collecting data through a multi-source sensing terminal and dynamically updating the model by using an improved Kalman filtering algorithm; quantifying progress deviation and locating causes by using an entropy weight-TOPSIS combination algorithm; adjusting progress and resources and formulating an emergency plan based on a particle swarm optimization algorithm; and constructing a progress-quality-pipeline safety correlation database in an acceptance stage to realize closed-loop management. Six function modules are correspondingly arranged in the system. The application improves the digitization and scientific level of progress management, solves problems such as poor adaptability and inaccurate deviation positioning of traditional management, and is suitable for complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of municipal engineering construction progress management technology, and in particular to a construction progress management method and system that integrates underground pipeline digital models. It is applicable to the whole process management of construction progress in urban water supply network renovation projects with multiple regions, multiple pipe materials, complex underground pipelines, and diverse geological conditions. Background Technology

[0002] Urban water supply network renovation projects are core municipal infrastructure projects, covering the city's core areas. They are characterized by densely intersecting underground pipelines, complex geological conditions, diverse types of construction pipes, concurrent construction at multiple work sites, and high requirements for collaborative division of labor. Simultaneous construction at multiple work sites can easily lead to problems such as pipeline avoidance conflicts, mismatches between geological conditions and construction techniques, and discrepancies between the planned schedule and actual construction.

[0003] Traditional methods for managing the construction progress of water supply network renovation projects often rely on experience to develop schedules. These methods fail to integrate and model underground pipeline spatial parameters, geological characteristics, pipe material compatibility parameters, and construction constraints, resulting in insufficient scientific rigor and adaptability of the schedules. Furthermore, the lack of real-time integration and dynamic model updates for multi-source construction data during construction makes it difficult to quickly and accurately pinpoint the causes of deviations between actual and planned progress. The lack of digital constraints on resource allocation for parallel multi-site construction easily leads to resource conflicts. Emergency construction adjustments fail to quantify the impact range based on parameters such as pipeline pressure and material, resulting in insufficient rationality and safety in schedule adjustments. Finally, during the construction acceptance phase, the lack of spatially correlated storage of progress, quality, and pipeline safety data prevents the formation of a closed-loop management system and hinders data support for subsequent similar projects.

[0004] While some existing technologies apply digital models to project schedule management, they lack specific designs tailored to the complexity of underground pipelines, geological diversity, and pipe material compatibility in urban water supply network renovation projects. They also fail to construct comprehensive digital models and corresponding schedule management algorithms adapted to water supply network renovation, thus failing to meet the demands for refined construction schedule management under conditions of multiple work sites, multiple pipe materials, complex pipelines, and challenging geological environments. Therefore, there is an urgent need for a construction schedule management method and system that integrates underground pipeline digital models to address the aforementioned problems of traditional schedule management and improve the digitalization, scientific rigor, and precision of construction schedule management for water supply network renovation projects. Summary of the Invention

[0005] To address the problems in existing technologies for construction progress management of urban water supply network renovation projects, such as poor adaptability, inaccurate deviation positioning, unreasonable resource allocation, unscientific emergency adjustments, and lack of closed-loop management throughout the entire process, this invention provides a construction progress management method and system that integrates a digital model of underground pipelines. This method combines the characteristics of pipelines, geology, pipe materials, and construction in water supply network renovation projects to construct a multi-parameter integrated digital model of underground pipelines. Based on this model, the scientific formulation, dynamic updating, deviation analysis, optimization adjustment, and closed-loop management of the construction progress plan are achieved.

[0006] The present invention provides a construction progress management method integrating a digital model of underground pipelines, comprising the following steps:

[0007] Step 1: Construct a digital model of underground pipelines that integrates parameters such as underground pipelines, geological features, pipe material compatibility, and construction constraints.

[0008] Preferably, in step one, the overall adaptation accuracy of the underground pipeline digital model is obtained by multiplying the four adaptation parameters—pipeline topology, geological conditions, pipe material technology, and construction constraints—by their respective weighting coefficients and then adding them together, with the sum of the four weighting coefficients being 1.

[0009] Preferably, in step one, the construction of the underground pipeline digital model further includes: establishing a mapping relationship library between pipe connection process parameters and construction efficiency. The mapping relationship library stores the correlation data between the bevel angle, welding current and construction time per unit length for arc welding of welded steel pipes, the correlation data between the rubber ring model and insertion depth and construction time per unit length for socket connection of ductile iron pipes, and the correlation data between the heating temperature and fusion depth and construction time per unit length for hot-melt connection of PPR pipes; at the same time, the geological stratification characteristics and trench excavation process parameters are bound and stored.

[0010] Step 2: Based on the underground pipeline digital model, decompose the construction tasks using the improved critical path method, determine the progress buffer period for each work site using the Monte Carlo simulation algorithm, and generate an initial construction schedule plan.

[0011] Preferably, in step two, the calculation of the progress buffer period first relies on three types of adaptation parameters of the underground pipeline digital model: pipeline topology, geological conditions, and pipe material technology. After 1000 iterations of Monte Carlo simulation, the buffer period correction coefficient is obtained. Then, the progress buffer period is calculated in combination with the basic construction period of the work site. The basic construction period of the work site is determined based on the construction process capacity and construction time data in the mapping relationship library. When using the improved critical path method, the calculation results of the model comprehensive adaptation accuracy value and deviation probability of the underground pipeline digital model are used as priority indicators to sort the construction sequence of multiple work sites.

[0012] Step 3: During construction, real-time data on underground pipeline dynamic detection, geological exploration, construction procedures, resource occupancy, and environmental safety monitoring are collected through multi-source sensing terminals. An improved Kalman filter multi-source data fusion algorithm is then used to dynamically update the digital model of the underground pipeline.

[0013] Preferably, in step three, the multi-source sensing terminal includes a pipeline detection radar, a geological borehole detector, a process progress sensor, a resource positioning terminal, a dust / noise monitor, and a wastewater quality sensor; the data from the dynamic detection of underground pipelines includes pipeline location correction values ​​and pipeline integrity supplementary data; the data from the geological exploration includes actual dimensions of karst areas, measured values ​​of foundation bearing capacity, and data on rock weathering degree; the data from the construction process includes trench excavation progress, pipeline installation qualification rate, and pressure test results; and the data from resource occupancy includes equipment operating status, personnel attendance rate, and material consumption progress.

[0014] Preferably, in step three, the state update formula of the improved Kalman filter multi-source data fusion algorithm is: ; ,in, The state vector of the underground pipeline digital model after time k is updated, which includes pipeline location correction value, measured geological parameter value, and construction progress node completion rate. The state transition matrix is ​​constructed based on the topology and construction procedure logic of the underground pipeline digital model. The optimal state vector of the underground pipeline digital model after the update at time k-1; To control the input matrix, associated resource allocation parameters are used; For resource allocation and control; Kalman gain; Let k be the measured data vector of the multi-source sensing terminal at time k; For the observation matrix, match the model state with the dimensions of the measured data; Observation matrix The transpose of the matrix; Let k be the state covariance matrix at time k-1; To observe the noise matrix, calibration is performed based on the accuracy of pipeline detection equipment and geological exploration errors. C represents the confidence level of the digital model of underground pipelines.

[0015] Step four: Based on the updated digital model of underground pipelines, the entropy weight-TOPSIS combined algorithm is used to compare the actual and initial construction progress, quantify the progress deviation value, and locate the causes of various related progress deviations.

[0016] Preferably, in step four, the weighting of the causes of the schedule deviation is calculated based on the confidence level of the underground pipeline digital model and the influence factors of the four types of deviation causes: pipeline, geology, resources, and environmental protection. It is the product of the confidence level of a single type and the corresponding factor divided by the sum of the products of all types. The model confidence level is a weighted sum of pipeline data integrity, geological data matching degree, and pipe material compatibility. The four types of influence factors are the ratio or deviation rate of the corresponding deviation value to the relevant benchmark value.

[0017] Step 5: Based on the causes of deviations and the constraint parameters of the underground pipeline digital model, the construction schedule and resource allocation scheme are dynamically adjusted using a particle swarm optimization multi-objective optimization algorithm.

[0018] Preferably, step five also includes a method for correcting the impact range of emergency progress adjustment. The corrected emergency impact radius is based on the basic impact radius determined by the accident type, and is obtained by combining the pipeline pressure level coefficient, the pipeline material vulnerability coefficient, and the confidence level of the underground pipeline digital model. The pipeline pressure level coefficient and the material vulnerability coefficient are determined according to the pipeline design pressure and the pipe material type, respectively.

[0019] Step six, the construction acceptance stage, involves linking and storing relevant construction quality acceptance data with the underground pipeline digital model to form a progress-quality-pipeline safety related database.

[0020] Preferably, in step six, the construction of the progress-quality-pipeline safety association database includes: binding the completion time of each key progress node with the corresponding material arrival inspection report, process quality record, pipeline pressure test curve, and water quality test report; at the same time, spatially associating the thickness and compaction data of the road surface restoration structural layer with the pipeline burial depth and direction data in the underground pipeline digital model to form a progress-location-quality three-dimensional association index.

[0021] This invention also provides a construction progress management system that integrates a digital model of underground pipelines, comprising:

[0022] The digital model building module is used to build digital models of underground pipelines that integrate parameters such as underground pipelines, geological features, pipe material compatibility, and construction constraints.

[0023] The initial schedule generation module is used to decompose construction tasks based on the underground pipeline digital model, decompose the construction tasks using the improved critical path method, determine the progress buffer period of each work point through the Monte Carlo simulation algorithm, and generate an initial construction schedule.

[0024] The multi-source data acquisition and model dynamic update module is used to collect real-time data on underground pipeline dynamic detection, geological exploration, construction procedures, resource occupation and environmental safety monitoring through multi-source sensing terminals during the construction process, and to dynamically update the digital model of the underground pipeline using an improved Kalman filter multi-source data fusion algorithm.

[0025] The schedule deviation analysis module is used to compare the actual and initial construction progress based on the updated underground pipeline digital model using the entropy weight-TOPSIS combined algorithm, quantify the schedule deviation value, and locate the causes of various related schedule deviations.

[0026] The schedule and resource optimization module is used to dynamically adjust the construction schedule and resource allocation scheme based on the causes of deviations and the constraint parameters of the underground pipeline digital model, using a particle swarm optimization multi-objective optimization algorithm.

[0027] The acceptance closed-loop management module is used during the construction acceptance phase. It links and stores construction quality acceptance data with the underground pipeline digital model to form a progress-quality-pipeline safety related database.

[0028] The construction progress management method and system of the present invention, which integrates digital models of underground pipelines, is designed for the construction characteristics of urban water supply network renovation projects. It achieves a deep integration of multi-parameter modeling and progress management of underground pipelines, geology, pipe materials, and construction constraints. Compared with existing technologies, it has the following advantages:

[0029] 1. High model adaptability, closely matching the actual situation of water supply network renovation projects: The constructed underground pipeline digital model integrates four core parameters: pipeline topology, geological conditions, pipe material technology, and construction constraints. It designs a comprehensive adaptation accuracy calculation formula and sets a qualified threshold. It can dynamically adjust the correction coefficient according to the pipe material type and geological type of the water supply network renovation project, accurately match the engineering characteristics of the water supply network renovation, and solve the problem of the disconnect between traditional models and actual engineering.

[0030] 2. Scientific schedule planning and strong multi-work site collaboration: Based on the digital model, the improved critical path method is used to decompose the construction tasks. The Monte Carlo simulation algorithm is combined to calculate the schedule buffer period. The overall model adaptation accuracy is incorporated into the construction sequence priority index, which effectively improves the collaboration of parallel construction at multiple work sites and reduces the risk of schedule delays.

[0031] 3. Dynamic model updates enable real-time control of the construction process: Multi-source sensing terminals collect comprehensive construction data, and an improved Kalman filter multi-source data fusion algorithm is used to dynamically update the digital model. This allows for real-time correction of state vectors such as pipeline location, geological parameters, and construction progress, ensuring that progress management is always based on actual project data and solving the problem of traditional progress management plans deviating from reality.

[0032] 4. Precise deviation location and quantitative cause analysis: The entropy weight-TOPSIS combined algorithm is used to quantify the schedule deviation value, and a weight formula for the influence of deviation causes is designed. Combined with the confidence level of the digital model, the four types of deviation causes (pipeline, geology, resources, and environmental protection) are ranked by weight, which realizes the precise location and quantitative analysis of schedule deviations, and provides a scientific basis for schedule adjustment.

[0033] 5. Dynamic optimization of schedule and resources, and scientific emergency adjustments: Based on the causes of deviations and the constraints of the digital model, the schedule and resource allocation scheme are adjusted by a multi-objective optimization algorithm of particle swarm optimization. At the same time, an impact range correction formula for emergency schedule adjustments is designed, so that the schedule and resource adjustments can not only meet the requirements of multi-work site collaboration, but also take into account pipeline avoidance, geological bearing capacity and safety and environmental protection constraints. The scientific nature and safety of emergency adjustments are greatly improved.

[0034] 6. Achieve closed-loop management throughout the entire process, with traceable and reusable data: The progress-quality-pipeline safety related database built during the construction and acceptance phase forms a closed-loop management system for the entire process of progress, quality, and pipeline safety. This not only supports traceability and verification of the construction process but also provides data support for the formulation of progress plans for subsequent similar water supply network renovation projects, enabling iterative optimization of project management. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating the overall method of construction progress management based on the digital model of underground pipelines of the present invention.

[0036] Figure 2 This is a flowchart of the improved Kalman filter multi-source data fusion algorithm used in step three of the method of the present invention.

[0037] Figure 3 This is a flowchart illustrating the process of using the entropy weight-TOPSIS combined algorithm to quantify schedule deviation values ​​and locate the causes of various related schedule deviations in step four of the method of this invention.

[0038] Figure 4 This is a flowchart illustrating the process of dynamically adjusting the construction schedule and resource allocation scheme using a particle swarm optimization multi-objective optimization algorithm in step five of the method of the present invention. Detailed Implementation

[0039] The following is a detailed implementation description of the construction progress management method and system of the present invention, which integrates the digital model of underground pipelines, based on the actual project of the third phase of a water supply pipeline renovation project, section I. The embodiments of the present invention are only used to explain the present invention and are not intended to limit the scope of protection of the present invention.

[0040] The service area of ​​Phase III, Section I of the water supply network renovation project covers four districts, with a total renovation length of approximately 54.169 km. It involves three main types of pipe materials: welded steel pipes, ductile iron pipes, and PPR pipes. The geological types include silty clay layers, moderately weathered limestone sections, and locally developed karst areas. A total of 45 construction sites are set up, employing a joint venture model between construction and installation units. The contract period is 730 days, and the planned construction period is 689 days. This embodiment uses a specific construction site as an example to illustrate the implementation process of the invention. At this site, ductile iron pipes are the primary type of pipe laid, with some welded steel pipes used. The geological type is silty clay layers. Gas and electricity lines intersect within the construction area, and there are two parallel construction sites nearby.

[0041] Example 1: Implementation of a construction progress management method that integrates a digital model of underground pipelines.

[0042] Combined with appendix Figures 1-4 As shown, this invention provides a construction progress management method that integrates a digital model of underground pipelines.

[0043] Step 1: Construct a digital model of underground pipelines. This step involves constructing a digital model of underground pipelines that integrates relevant parameters, geological characteristic parameters, pipe material compatibility parameters, and construction constraint parameters. The overall compatibility accuracy M of the model is then calculated to verify its suitability.

[0044] Specifically, a digital model of underground pipelines is constructed that integrates relevant parameters, geological characteristic parameters, pipe material compatibility parameters, and construction constraint parameters. The relevant parameters of underground pipelines include the spatial coordinates, pipe diameter, material, burial depth, and connection relationships of various underground pipelines. The geological characteristic parameters include the distribution range, strength parameters, and size and distribution density of adverse geological bodies of different geological types. The pipe material compatibility parameters include the connection process parameters and mechanical performance thresholds of various construction pipe materials.

[0045] In step one, the formula for calculating the overall adaptation accuracy of the underground pipeline digital model is as follows:

[0046] ;

[0047] The definitions and calculation logic of each parameter are as follows:

[0048] M: The overall adaptation accuracy value of the digital model of underground pipelines, ranging from 0.85 to 1.0. M ≥ 0.9 is the qualified threshold for model construction. , , , These are the weighting coefficients for pipeline topology, geological conditions, pipe material technology, and construction constraints, determined using the analytic hierarchy process (AHP) combined with the risk assessment results of the construction plan, to meet the following requirements. In some preferred embodiments, =0.35, =0.30, =0.20, =0.15;

[0049] Pipeline topology adaptation parameters characterize the degree of matching between the spatial distribution of underground pipelines and the construction path, and are calculated using the following formula: ,in: Pipeline intersection density, unit: intersections / m² , This refers to the number of intersections of gas, electricity, and communication pipelines within the work site area. The area of ​​the construction site is expressed in m². : Coefficient of variation of pipeline burial depth , This represents the standard deviation of the burial depth of each pipeline within the work site, in meters. This represents the average burial depth of each pipeline within the work site, in meters; a and b are correction coefficients, dynamically adjusted according to the pipeline material type: a=0.25, b=0.15 for steel pipes, a=0.20, b=0.18 for ductile iron pipes, and a=0.18, b=0.20 for PPR pipes.

[0050] Geological condition adaptation parameters characterize the degree of compatibility between geological stratification characteristics and construction technology; the calculation formula is as follows: ,in: : Karst development density, unit: karst units / m² , The number of karst caves / karst grooves with a diameter ≥ 0.5m within the work site area; The area of ​​the construction site is expressed in m². Foundation bearing capacity deviation rate , This is the measured value of the foundation bearing capacity, in kPa. The standard value of the foundation bearing capacity required by the design is given in kPa; c and d are geological influence correction coefficients: c=0.30 and d=0.25 for silty clay layer, c=0.22 and d=0.30 for moderately weathered limestone section, and c=0.40 and d=0.28 for karst development area.

[0051] Pipe process adaptation parameters characterize the degree of fit between the pipe connection process and the model's preset parameters. The calculation formula is as follows: ,in: : Arc welding compatibility coefficient for welded steel pipes =Non-destructive testing pass rate of welds, with a value range of 0.95 to 1.0; : Socket fit coefficient of ductile iron pipe =The pass rate of the socket joint sealing test, with a value range of 0.96 to 1.0; PPR pipe heat fusion compatibility coefficient =The compliance rate of hot-melt connection strength, with a value range of 0.94 to 1.0; , , Pipe weighting coefficient, determined based on the proportion of pipe usage at each work site. , , , , , The figures represent the laying lengths of welded steel pipes, ductile iron pipes, and PPR pipes within the work site, in meters.

[0052] Construction constraint adaptation parameters characterize the degree of matching between the model and the construction period and multi-workpoint collaboration; the calculation formula is as follows: ,in: : Planned construction period for each work site, in days, taken from the basic construction period of a single work site in the construction schedule; The base construction period for the work site as stipulated in the contract, in days, is calculated based on the work site's production capacity multiplied by the laying length in the construction plan. Multi-point collaboration conflict coefficient , This represents the number of resource conflicts between the current work site and surrounding concurrent work sites. The total number of parallel construction sites in the surrounding area; e, f: construction constraint correction coefficients, e=0.12, f=0.08.

[0053] 1. Collect basic parameters

[0054] Underground pipeline related parameters: The construction area of ​​this site is S=800m², and the number of intersections of gas and power pipelines within the area is [not specified]. =12 pipelines, with burial depths of 1.2m, 1.5m, 1.3m, 1.6m and 1.4m respectively, and the pipeline material is mainly steel pipe;

[0055] Geological characteristics: The work site consists of a silty clay layer with no karst development; the design value of the foundation bearing capacity is... =180kPa, measured value =175kPa;

[0056] Pipe compatibility parameters: Laying length of ductile iron pipe within the work site =450m, length of welded steel pipe laying =150m, without PPR pipe ( =0); Pass rate of sealing test for ductile iron pipe socket joints =0.98, the pass rate of non-destructive testing of welded steel pipe welds =0.97;

[0057] Construction constraint parameters: Planned construction period for each work site =45 days, the standard construction period for the work site as stipulated in the contract. =40 days; Total number of surrounding parallel construction sites =2, number of resource conflicts =1.

[0058] 2. Calculate each adaptation parameter.

[0059] Pipeline topology adaptation parameters Pipeline intersection density =12 / 800=0.015 pieces / m²; average burial depth =1.4m, standard deviation of burial depth ≈0.141m, coefficient of variation of burial depth ≈0.101; The pipeline at this work site is mainly made of steel pipes, so we take a=0.25 and b=0.15, therefore =0.9811.

[0060] Geological condition adaptation parameters : Karst development density =0, foundation bearing capacity deviation rate =|175-180| / 180≈0.0278; For the silty clay layer, take c=0.30 and d=0.25, therefore... =0.99305.

[0061] Pipe process adaptation parameters Pipe weighting coefficient =0.25, =0.75, =0; take =0.97、 =0.98、 =0, therefore =0.9775.

[0062] Construction constraint adaptation parameters Multi-point collaboration conflict coefficient =1 / 2=0.5; taking e=0.12 and f=0.08, therefore =0.825.

[0063] 3. Computational model overall adaptation accuracy M

[0064] Take the default weight coefficient =0.35, =0.30, =0.20, =0.15, therefore: M=0.96055, this value ≥0.9, meets the qualified threshold for model construction, and the underground pipeline digital model construction is qualified.

[0065] 4. Establish a mapping relationship database and binding storage.

[0066] A mapping database between pipe connection processes and construction efficiency was established. For example, at this construction site, the DN150 ductile iron pipe socket connection uses a DN150 rubber ring with an insertion depth of 100mm, resulting in a construction time of 0.8 days / m per unit length. For the DN200 welded steel pipe, the arc welding bevel angle is 60°, the welding current is 150A, and the construction time per unit length is 1.2 days / m. Simultaneously, the silty clay layer was linked to the trench excavation process parameters: excavation slope 1:0.1; manual excavation was switched from excavation by excavator to 200mm from the base; and 300mm of graded sand and gravel was used as replacement when the base bearing capacity was insufficient.

[0067] By constructing a digital model of underground pipelines, we can effectively adapt to complex working conditions, improve the reliability of the model, quantify the adaptation relationship, reduce construction risks, and lay the foundation for subsequent progress quantification and multi-work site collaboration.

[0068] Step 2: Generate the initial construction schedule.

[0069] Specifically, based on the digital model of underground pipelines, combined with the contract period, the distribution characteristics of multiple work sites, and the division of labor boundaries of the consortium, the construction tasks are decomposed using the improved critical path method, and the progress buffer period of each work site is determined by the Monte Carlo simulation algorithm to generate the initial construction schedule plan.

[0070] In step two, the calculation process for the progress buffer period is as follows: based on the pipeline topology adaptation parameters in the underground pipeline digital model. Geological condition adaptation parameters Pipe material process compatibility parameters The buffer period correction coefficient was obtained through 1000 iterations of Monte Carlo simulation. The progress buffer period .

[0071] in, The basic construction period for each work site is calculated based on construction time data from a database of construction process capacity and mapping relationships. When using the improved critical path method, the accuracy value is determined by the overall model adaptation. The priority index is used to sort the construction sequence of multiple work sites.

[0072] 1. Decompose the construction tasks: Based on the digital model of underground pipelines and combined with the division of labor boundaries of the consortium (the construction unit is responsible for pipeline laying at this work site, and the installation unit is responsible for the subsequent intelligent transformation), the construction tasks are decomposed into 9 core processes using the improved critical path method: surveying and setting out → road surface cutting → trench excavation → pipeline foundation treatment → pipe laying → pipeline connection → pressure test → earthwork backfilling → road surface restoration.

[0073] 2. Calculate the basic construction period for each work site. Based on the construction time data from the construction process capacity and mapping relationship database, the basic construction period of this work site is calculated. =40 days (contractual base period).

[0074] 3. Calculate the schedule buffer period :based on =0.9811、 =0.99305、 =0.9775. Through 1000 iterations of Monte Carlo simulation, the buffer period correction factor K was calculated: K=1.008115. (Schedule buffer period) ≈40.32 days, rounded up to 40 days.

[0075] 4. Generate initial schedule: Combine the base duration with the buffer period to determine the total planned duration of 45 days for this work site, while simultaneously adjusting the model's overall accuracy. =0.96055×(1-0.05)=0.9125 is the priority indicator, which puts this work point first in the construction of surrounding parallel work points to avoid resource conflicts. The generated initial schedule clearly defines the start / end time of each process and the personnel / equipment configuration. For example, the trench excavation process is planned for 15 days, with two pipeline installation teams (10 workers) and 1 small excavator.

[0076] By using a digital model of underground pipelines, combined with the contract period, the distribution characteristics of multiple work sites, and the division of labor boundaries of the consortium, an improved critical path method is adopted to decompose construction tasks. The Monte Carlo simulation algorithm is used to determine the progress buffer period of each work site, which can fit the division of labor and constraints, making the plan accurate and feasible; the reserved buffer period enhances the ability to resist risks; and the optimized work site sorting reduces resource conflicts.

[0077] Step 3: Dynamic updating of the digital model of underground pipelines.

[0078] During construction, multi-source sensing terminals are used to collect real-time data on underground pipeline dynamic detection, geological exploration, construction procedures, resource occupancy, and environmental safety monitoring. An improved Kalman filter multi-source data fusion algorithm is then used to dynamically update the digital model of the underground pipeline.

[0079] Specifically, the multi-source sensing terminals include pipeline detection radar, geological borehole detector, process progress sensor, resource positioning terminal, dust / noise monitor, and sewage quality sensor; the underground pipeline dynamic detection data includes pipeline location correction value, pipeline integrity supplementary data, actual geological exploration data including actual size of karst area, measured value of foundation bearing capacity, and rock weathering degree data, construction process completion data including trench excavation progress, pipeline installation qualification rate, and pressure test results, and the resource occupancy data includes equipment operating status, personnel attendance rate, and material consumption progress.

[0080] In step three, the state update formula for the improved Kalman filter multi-source data fusion algorithm is:

[0081] ;

[0082] ;

[0083] in, The state vector of the underground pipeline digital model after time k is updated, which includes pipeline location correction value, measured geological parameter value, and construction progress node completion rate. The state transition matrix is ​​constructed based on the topology and construction procedure logic of the underground pipeline digital model. The optimal state vector of the underground pipeline digital model after the update at time k-1; To control the input matrix, associated resource allocation parameters are used; For resource allocation and control; Kalman gain; Let k be the measured data vector of the multi-source sensing terminal at time k; For the observation matrix, match the model state with the dimensions of the measured data; Observation matrix The transpose of the matrix; Let k be the state covariance matrix at time k-1; To observe the noise matrix, calibration is performed based on the accuracy of pipeline detection equipment and geological exploration errors. C represents the confidence level of the digital model of underground pipelines.

[0084] 1. Multi-source data acquisition: During construction, real-time data was collected at the work site through multi-source sensing terminals such as pipeline detection radar, geological borehole detectors, and process progress sensors: Underground pipeline dynamic detection data: pipeline position correction value +0.05m, pipeline integrity undamaged; Actual geological exploration data: foundation bearing capacity retest value 178kPa, no new karst development; Construction process completion data: trench excavation process actual completion progress 80%, pipeline installation qualification rate 95%; Resource occupancy data: excavator operating status normal, personnel attendance rate 90%, ductile iron pipe material consumption progress 70%; Environmental safety monitoring data: dust concentration 0.3mg / m³. 3 The noise level is 65dB, which meets environmental protection requirements.

[0085] 2. Improved Kalman Filter Multi-Source Data Fusion Algorithm for Model Update: During the construction process at this site, the core objective of the improved Kalman filter multi-source data fusion algorithm is to dynamically correct the state parameters (including pipeline location, geological parameters, construction progress, etc.) of the underground pipeline digital model by fusing measured data collected from multi-source sensing terminals, ensuring that the model always remains consistent with the actual engineering conditions. The following details the matrix generation logic, algorithm calculation process, and model update implementation principle in conjunction with the actual conditions at the site.

[0086] Algorithm Application Prerequisites and State Vector Definition. The state vector of the underground pipeline digital model is the core object of the algorithm update and must cover key parameters strongly correlated with the construction progress. Based on the core process connection logic of "trench excavation → pipeline installation" at this work site, the state vector is defined as follows: =[Progress Node Completion Rate, Pipeline Location Correction Value, Measured Foundation Bearing Capacity Value]ᵀ. This example focuses on updating the progress node completion rate, with other parameters being corrected synchronously. Time k is set as the 12th day of the trench excavation process. At this time, the model state at time k-1 (day 11) needs to be updated by integrating the "measured progress collected by the process progress sensor", "location correction data of the pipeline detection radar", and "measured bearing capacity data of the geological borehole detector" through an algorithm, so as to achieve dynamic matching between the model and the actual construction.

[0087] The generation process of the core matrix (in conjunction with construction logic design). The core of Kalman filtering is to quantify the logical relationships in the construction process through state transition matrices, control input matrices, and observation matrices. The generation of each matrix must strictly conform to the construction plan, process connection rules, and resource allocation characteristics of this project.

[0088] (1) State transition matrix Generation of state transition matrix This is used to describe "how the model state at time k-1 naturally transitions to the predicted state at time k". Its element values ​​are determined based on the logical correlation of construction procedures and historical construction efficiency data. The core procedure at this work site is "trench excavation → pipeline installation". The progress of trench excavation directly affects the initiation conditions of subsequent pipeline installation. Based on the efficiency statistics of similar projects and the requirements of this project's construction plan, the progress transfer coefficient of trench excavation to pipeline installation is set to 0.8 (i.e., for every 1% increase in the completion rate of the preceding procedure, the predicted completion rate of the subsequent procedure can be increased by 0.8%). The pipeline location correction value and the measured value of the foundation bearing capacity do not change significantly in a short period (1 day), and their self-transfer coefficient is set to 0.99 (approximately stable), and they have no cross-influence with the progress completion rate (cross-coefficient is 0). The final state transition matrix is ​​generated as follows: This example focuses on updating the completion rate of progress nodes. Subsequent calculations only show the core transition relationship in the first row and first column, while other parameters are corrected synchronously but not expanded separately.

[0089] (2) Control input matrix The generation of the control input matrix This is used to quantify the impact of resource allocation measures on the model state, and its element values ​​are determined based on the correlation analysis between resource input and progress improvement. The core resource for this work site during the trench excavation phase is construction personnel. According to the construction plan, a 10% increase in personnel attendance can increase the trench excavation progress completion rate by 9% (based on resource input test data from three parallel work sites). Therefore, the resource allocation coefficient is set to 0.9. Resource allocation has no direct impact on pipeline location or foundation bearing capacity, and the corresponding coefficient is set to 0. The final control input matrix is ​​generated as follows: .

[0090] (3) Resource allocation control quantity The determination, This is a quantitative indicator of actual resource allocation, calculated by combining "on-duty data collected by personnel positioning terminals" and "equipment operation status monitoring data." On the 12th day, verification via personnel positioning terminals showed that 9 out of 10 workers in the pipeline installation team were actually on-duty, resulting in an on-duty rate of 90%. Therefore... =0.9 (after dimensionless processing).

[0091] (4) Observation matrix The generation of the observation matrix This is used to match the "model state vector dimension" with the "measured data dimension of multi-source sensing terminals," ensuring that the predicted state and measured data can be directly compared. This work site collects measured data through three types of sensing terminals: a process progress sensor (outputting progress completion rate, corresponding to the first element of the state vector), a pipeline detection radar (outputting position correction values, corresponding to the second element), and a geological borehole detector (outputting measured bearing capacity values, corresponding to the third element). The measured data corresponds one-to-one with the model state vector, requiring no dimension transformation; therefore, a unit observation matrix is ​​generated. In the calculation, it is simplified to =1, which only reflects the observation relationship of progress completion rate.

[0092] (5) State covariance matrix at time k-1 The setting, The uncertainty (error range) of the model state at time k-1 is described based on the statistical analysis of construction data from the previous 11 days. By analyzing the deviation between the "model-predicted progress" and the "actual progress" from days 1 to 11, the variance of the progress completion rate is calculated to be 0.02, while the variances of the pipeline location correction value and the measured value of the foundation bearing capacity are 0.01 (higher data stability), and the errors of each parameter are uncorrelated (covariance is 0). The final state covariance matrix is ​​defined as follows: .

[0093] (6) Observation noise matrix The generation of (dynamically adjusted based on model confidence) and the observation noise of traditional Kalman filtering. This algorithm mostly uses fixed values. The noise weights are dynamically calculated based on the model confidence C, which matches the current reliability of the model and improves the update accuracy. The calculation logic of model confidence C: It integrates three dimensions—pipeline data integrity, geological data matching degree, and pipe material compatibility—to quantify the degree of fit between the model and actual working conditions: Pipeline data integrity: Location data collected by pipeline detection radar is complete with no missing values ​​and correction values ​​≤ 0.05m, resulting in an integrity score of 1.0, a weight of 0.4, and a contribution value of 1.0 × 0.4 = 0.4; Geological data matching degree: The design value of the foundation bearing capacity is 175kPa, and the measured value is 178kPa. The deviation rate is |178-175| / 175≈0.0171, the matching degree is 1-0.0171=0.9829, the weight is 0.3, and the contribution value is 0.9829 × 0.3≈0.2949; Pipe material compatibility: The pass rate of the ductile iron pipe socket joint sealing test at this work site is 98%, and the pass rate of the non-destructive testing of welded steel pipe welds is 97%. The pipe material compatibility M, calculated based on the proportion of pipe material usage, is... a =0.9775, weight 0.3, contribution value = 0.9775 × 0.3 = 0.29325; final model confidence: C = 0.98815 (the higher the confidence, the lower the observation noise). Observation noise matrix The calculation: Based on the confidence level, the reliability of the measured data is back-quantified, and the formula is as follows. (0.01 is the basic noise figure), substituting into C=0.98815, we get: This demonstrates the consistency of multi-source measured data and exhibits extremely low noise levels.

[0094] Kalman gain Calculation of the Kalman gain (balancing predicted and measured weights). Used to determine the fusion weights between "model predicted state" and "multi-source measured data". The larger the value, the more reliable the measured data, and the more the updated state is biased towards the measured value; conversely, the smaller the value, the more biased it is towards the predicted value. Based on the core formula of Kalman gain: Substitute the matrix parameters into the calculation (focusing on the progress completion rate dimension): Calculate :because =1、 =0.02, result =0.02; calculate denominator =0.0201185; calculate : 0.02×1×(0.0201185) -1 ≈0.9941; Result Analysis: The value of ≈0.9941 is close to 1, indicating that the reliability of the multi-source measured data (80% progress) at this moment is extremely high, and the model update should be mainly based on the measured data for correction.

[0095] Model state vector at time k The update process (core update process) integrates "model prediction status," "resource allocation impact," and "measured data correction" to obtain the updated model status, enabling dynamic matching of the model with actual construction. Core formula:

[0096] ;

[0097] in, The optimal state vector of the underground pipeline digital model after the update at time k-1. The progress completion rate in the model state vector at time k-1 (day 11) is 60% (from the previous day's construction progress record). The actual progress at time k (day 12) is 80% (data collected from process progress sensors, verified and confirmed); step-by-step calculation: model predicted progress: =0.8 × 60% = 0.48 (only considering the predicted value of natural process transfer); Impact of resource allocation: =0.9 × 90% = 0.81 (Effect of staff attendance rate on progress improvement); Actual measurement correction item: =0.9941×0.32≈0.3181 (correcting for prediction bias using measured data); Updated progress completion rate: =1.6081 (80.4% after dimensionless processing); Linked correction: After the progress completion rate is updated, the model synchronously corrects the "pipeline location correction value" and "measured foundation bearing capacity value" (due to the extremely small observation noise, the correction range is ≤0.1%) to ensure the consistency of each parameter of the model.

[0098] After the update, the "completion rate of trench excavation process at the work site" in the model was corrected to 80.4%, with a deviation of only 0.4% from the actual measured progress of 80%. This achieved an accurate mapping of the model to the actual construction, providing a reliable digital model foundation for subsequent "resource allocation optimization for pipeline installation process" and "progress deviation early warning".

[0099] By collecting real-time dynamic detection data of underground pipelines, actual geological exploration data, construction process completion data, resource occupancy data, and environmental safety monitoring data through multi-source sensing terminals, the digital model of underground pipelines is dynamically updated using an improved Kalman filter multi-source data fusion algorithm. This achieves real-time multi-source data-driven operation, maintains the timeliness of the model, improves data fusion accuracy, reduces errors, and captures construction dynamics in real time, providing early warnings of abnormal situations.

[0100] Step 4: Schedule Deviation Analysis.

[0101] Based on the updated digital model of underground pipelines, the entropy weight-TOPSIS combined algorithm is used to compare the actual construction progress with the initial schedule, quantify the progress deviation value and calculate the influence weight of the deviation causes.

[0102] Specifically, in step four, the formula for the weighting of the causes of schedule deviations is as follows:

[0103] ;

[0104] in, The influence weight of the i-th type of deviation cause, i=1,2,3,4, corresponds to pipeline, geology, resource, and environmental deviations, respectively; C is the confidence level of the underground pipeline digital model, C=pipeline data integrity×0.4+geological data matching degree×0.3+pipe material compatibility×0.3; Pipeline deviation is an influencing factor for the cause of the i-th type of deviation. =Pipeline detection deviation / Design pipeline accuracy, geological deviation = Deviation rate between actual geological parameters and model preset parameters, resource deviation =Resource gap / Planned resource input, environmental deviation =Duration of environmental violations / Total construction time.

[0105] Based on the updated digital model of underground pipelines, the core of schedule deviation analysis is to quantitatively assess schedule deviations using the entropy-weighted TOPSIS combined algorithm, and to accurately pinpoint the dominant factors of deviations by combining model confidence and influencing factors of deviation causes. The following details the algorithm implementation process, deviation quantification logic, and weight calculation principles, using actual construction sites as examples:

[0106] 1. Calculation Logic of Model Confidence C (Basic Prerequisite for Weight Calculation). Model confidence C is a core indicator for measuring the degree of fit between the digital model of underground pipelines and the actual engineering situation. It directly affects the reliability of the weights for the causes of deviations. Its calculation is based on a weighted summation of three dimensions: "pipeline data integrity, geological data matching degree, and pipe material compatibility." The weights of each dimension are determined using the analytic hierarchy process (pipeline data integrity 0.4, geological data matching degree 0.3, pipe material compatibility 0.3). The specific calculation process is as follows:

[0107] Pipeline data integrity ( Pipeline data integrity characterizes the completeness of pipeline parameters (location, burial depth, material, etc.) obtained through detection. The calculation formula is: =Number of pipeline parameters actually acquired / Total number of pipeline parameters required by the design. The design requires 5 pipeline parameters for this site (spatial coordinates, pipe diameter, material, burial depth, and connection relationships). Verification using pipeline detection radar and manual pit testing confirmed that all 5 parameters were acquired completely, and the pipeline position correction was only 0.05m (≤ design allowable deviation of 0.1m). Therefore, the pipeline data is complete. =1.0.

[0108] Geological data matching degree ( Geological data matching degree characterizes the degree of agreement between actual geological parameters and model preset parameters. The calculation formula is as follows:

[0109] ;

[0110] The model for this construction site is pre-set to have a foundation bearing capacity. =175kPa. Three sets of data (173kPa, 178kPa, 176kPa) were actually measured through on-site plate load tests. The actual mean value is... =(173+178+176) / 3=175.67kPa, therefore: ≈0.9829.

[0111] Pipe compatibility refers to the pipe process compatibility parameters. This characterizes the degree of fit between the pipe connection process and the preset parameters of the model. =0.9775 (The pass rate of the sealing test of the socket joint of ductile iron pipe is 98%, and the pass rate of non-destructive testing of weld seam of welded steel pipe is 97%, calculated by weighting the proportion of pipe usage). =0.9775.

[0112] The final calculation of the model confidence score C is: C = ×0.4+ ×0.3+ ×0.3=0.98815.

[0113] 2. The implementation process of the entropy weight-TOPSIS combined algorithm for quantifying schedule deviation values: The core logic of the entropy weight-TOPSIS combined algorithm is as follows: First, the weights of each evaluation indicator affecting the schedule are determined using the entropy weight method (objective weights to avoid subjective bias). Then, the TOPSIS method is used to calculate the closeness between the actual schedule and the "ideal schedule state," thereby quantifying the schedule deviation value. The specific steps are as follows:

[0114] A progress evaluation index system was constructed, and based on the characteristics of the water supply network renovation project, five core indicators were selected as evaluation dimensions: "progress completion rate, pipeline adaptability, geological adaptability, resource security rate, and environmental compliance rate." Specifically: Progress completion rate: the ratio of actual progress to planned progress (core evaluation indicator); Pipeline adaptability: pipeline topology adaptation parameters... (Characterizing the impact of pipeline distribution on schedule); Geological adaptability: geological condition adaptation parameters (Characterizing the impact of geological conditions on progress); Resource security rate: the ratio of actual resource input to planned resource input (comprehensive security rate of personnel, equipment, and materials); Environmental compliance rate: the percentage of time that environmental monitoring indicators (dust, noise) meet the standards.

[0115] A decision matrix was constructed, taking the trench excavation process at this work site as the evaluation object. "Ideal progress state" (virtual optimal solution) and "actual progress state" (actual solution) were defined, and a 2×5 decision matrix X (rows: solutions; columns: indicators) was constructed, as shown in Table 1:

[0116] Table 1 2×5 Decision Matrix X

[0117]

[0118] The decision matrix is ​​standardized (eliminating the influence of dimensions) using a positive index standardization formula (the larger the index value, the better): ,in, , These represent the maximum and minimum values ​​of the j-th indicator, respectively. The standardized matrix Z is shown in Table 2.

[0119] Table 2 Standardized matrix Z

[0120]

[0121] The entropy weight method calculates the objective weights of the indicators and calculates the entropy value of the j-th indicator. : ,in (m is the number of options, here m=2); Calculate the coefficient of variation for the j-th indicator. : (The larger the coefficient of variation, the greater the impact of the indicator on the evaluation result); Calculate the indicator weights. : Substitute the data to calculate: Progress completion rate entropy value. ≈0.970, coefficient of difference ≈0.030; Pipeline adaptability entropy value ≈0.998, coefficient of difference ≈0.002; Geological adaptability entropy value ≈0.999, coefficient of difference ≈0.001; Entropy value of resource availability rate ≈0.945, coefficient of difference ≈0.055; Entropy value of environmental compliance rate ≈1.000, coefficient of difference ≈0.000. Indicator weight. calculate: =0.088, =0.030 / 0.088≈0.341, =0.002 / 0.088≈0.023, =0.001 / 0.088≈0.011, =0.055 / 0.088≈0.625, =0.000 / 0.088=0.000.

[0122] The TOPSIS method is used to calculate the closeness and schedule deviation values, and a weighted normalization matrix is ​​constructed. : = × As shown in Table 3.

[0123] Table 3 Weighted Standardization Matrix Z'

[0124]

[0125] Determine the positive and negative ideal solutions; the positive ideal solution (optimal solution) The maximum value of the weighted standardized value of each indicator. =(0.341, 0.023, 0.011, 0.625, 0.000); Negative ideal solution (worst-case scenario) The minimum value of the weighted standardized value of each indicator =(0.274, 0.022, 0.011, 0.562, 0.000).

[0126] Calculate the Euclidean distance between the actual solution and the positive and negative ideal solutions, and the distance between the actual solution and the positive ideal solution. : ≈0.092, the distance between the actual solution and the negative ideal solution. : =0.

[0127] The similarity is calculated using the pure mathematical formula of the TOPSIS algorithm. Proximity to schedule deviation =0 / (0.092+0)=0. The approximation is adjusted based on the actual engineering situation: the theoretical approximation is an extreme value, an extreme result of mathematical calculation, which does not match the actual construction conditions at this site. Based on the entropy weight method calculation results, the progress completion rate is the core indicator affecting construction progress (with the highest weight). Therefore, the actual progress completion rate on the 12th day of the trench excavation process at this site is used as the adjusted approximation, i.e., the adjusted approximation. =0.804 (80.4%), ensuring that the algorithm's calculation results closely match the actual engineering situation. =1-0.804×100%=19.6%. The closer the approximation is to 1, the closer the actual progress is to the ideal progress. Here, the weighted standardized matrix highlights the core impact of "progress completion rate" and "resource guarantee rate", and finally quantifies the conclusion of slight deviation.

[0128] 3. Calculation and implementation of the weighting of bias causes: The core logic of the weighting of bias causes is to combine the model confidence level C (model reliability) with the influence factors of each bias cause. (Severity of deviation), the contribution percentage of each cause to the schedule deviation is obtained through normalization calculation. The specific steps are as follows:

[0129] Define the types of deviation causes, and based on the characteristics of water supply network renovation projects, classify the causes of schedule deviations into four categories: pipeline deviations ( ), geological deviation ( Resource bias ), environmental deviation ( These correspond to the core external constraints and internal resource factors that affect the progress, respectively.

[0130] Calculate the influencing factors of each cause of deviation. Impact Factor The severity of deviations from each cause is quantified, with values ​​ranging from [0,1]. The calculation logic is as follows:

[0131] (1) Pipeline deviation Pipeline detection deviation value / designed pipeline accuracy: At this work site, the pipeline position correction value measured by the pipeline detection radar is 0.05m, and the designed pipeline accuracy requirement is ≤0.1m (i.e., the quantified value of the designed pipeline accuracy is 1.0). Therefore... =0.05.

[0132] (2) Geological deviation

[0133] For example, the actual measured bearing capacity of the foundation at this construction site is 178 kPa, while the model's preset value is 175 kPa. Therefore... ≈0.0171.

[0134] (3) Resource deviation : (Planned resource input - Actual resource input) / Planned resource input. The planned attendance rate for this work site is 100% (10 people), and the actual attendance rate is 90% (9 people). Therefore... =0.1.

[0135] (4) Environmental deviations Environmental violation time / total construction time: During the construction period at this site, the dust concentration was 0.3 mg / m³ and the noise level was 65 dB, both meeting environmental protection requirements. The time of violation was 0. Therefore... =0.

[0136] Calculate the weighted impact factor ( The model confidence level C is introduced to adjust the impact factor: the higher the model confidence level, the higher the reliability of the impact factor, and the larger the weighted impact factor. In this case, C = 0.98815, therefore: ≈0.0494; ≈0.0169; ≈0.0988; =0.

[0137] Normalization calculation bias causes affect weights The weighted influence factors are converted into weights (summing up to 1) through normalization, as shown in the formula:

[0138] ;

[0139] in, =0.1651, therefore: =0.0494 / 0.1651≈0.299 (29.9%); =0.0169 / 0.1651≈0.102 (10.2%). =0.0988 / 0.1651≈0.599 (59.9%) =0 / 0.1651=0.

[0140] The conclusions regarding the causes of deviations indicate that resource deviations account for nearly 60% of the impact, making them the dominant cause of schedule deviations (personnel attendance rate less than 10%). Pipeline deviations and geological deviations account for 29.9% and 10.2% of the impact, respectively, and are considered secondary causes. There were no environmental deviations, which clarifies the optimization direction for subsequent schedule adjustments (prioritizing the replenishment of personnel resources).

[0141] The quantification of schedule deviations is achieved through an entropy-weighted TOPSIS combined algorithm. The core of this algorithm is to first objectively determine the weights of each influencing indicator using the entropy-weighted method (highlighting the core roles of schedule completion rate and resource availability rate), then calculate the closeness between the actual and ideal schedules using the TOPSIS method, ultimately quantifying the degree of deviation. The calculation of deviation cause weights corrects influencing factors through model confidence, avoiding the one-sidedness of single-factor evaluation, accurately identifying resource deviations as the main cause, providing a scientific basis for subsequent schedule optimization and adjustment, and achieving precise quantification of the degree of deviation. Furthermore, it identifies the core causes, enabling targeted correction and reducing correction costs.

[0142] Step 5: Optimize and adjust the schedule and resources.

[0143] Based on the causes of deviations (mainly resource deviations) and the constraint parameters of the underground pipeline digital model, a particle swarm optimization multi-objective optimization algorithm is used to dynamically adjust the construction schedule and resource allocation scheme to ensure the coordinated advancement of parallel construction at multiple work sites, while meeting the requirements of pipeline avoidance, geological bearing capacity, and safety and environmental protection. At the same time, an emergency adjustment plan is formulated.

[0144] Specifically, step five also includes a formula for correcting the scope of impact of emergency schedule adjustments:

[0145] ;

[0146] in, This is the corrected emergency impact radius; Based on the radius of influence, a radius of 5m is used for pipeline leakage, 15m for fire, and 3m for mechanical injury. This is the pipeline pressure rating coefficient, based on the pipeline design pressure. Values: <0.6MPa =1.0, 0.6≤ <1.0MPa =1.3, ≥1.0MPa =1.6; C represents the vulnerability coefficient of the pipeline material, with 1.0 for welded steel pipes, 0.8 for ductile iron pipes, and 1.2 for PPR pipes; C represents the confidence level of the digital model of the underground pipeline.

[0147] The core of schedule and resource optimization is to use a multi-objective optimization algorithm, Particle Swarm Optimization (PSO), to balance the two core objectives of "minimizing schedule deviation" and "optimizing resource allocation" while meeting hard constraints such as pipeline avoidance and geological bearing capacity. Simultaneously, scientific contingency plans are developed for unforeseen emergency scenarios. The following details the algorithm adaptation design, iterative calculation process, and optimization scheme generation logic, based on the actual conditions at the work site, and clarifies the basis for developing emergency plans.

[0148] 1. Engineering adaptation design of multi-objective optimization algorithm for particle swarm optimization: Considering the characteristics of this project, engineering adaptation is required in three aspects: objective function, particle encoding, and constraints.

[0149] Multi-objective function construction (quantifying optimization objectives). Based on the work site requirements, four core objectives are identified. A weighted summation method is used to transform the multiple objectives into a single fitness function (the higher the fitness value, the better the solution):

[0150] ;

[0151] Among them, Target 1 ( ): Meet pipeline avoidance requirements (hard constraint, must meet the standard), pipeline avoidance distance must be ≥0.5m (based on underground pipeline protection requirements in the construction plan), quantified as: =1 (meets avoidance distance) or 0 (does not meet the requirement, the particle is eliminated directly). Weight =0.2 (High priority of hard constraints). Objective 2 ( ): Meets geological bearing capacity requirements (hard constraint, must meet standards), foundation bearing capacity must be ≥175kPa (design value), quantified as: =1 (bearing capacity meets standard) or 0 (does not meet standard, directly eliminated). Weight =0.2 (High priority for hard constraints). Target 3 ( Minimize schedule deviation (core optimization objective), schedule deviation rate Quantified as: ( =0 (The smaller the deviation, the better). Weighting =0.4 (Core objective, highest weight). Objective 4 ( ): Optimize resource allocation (auxiliary optimization objective), resource allocation efficiency Quantified as: ( (Optimal when ≥1, to avoid resource waste) Weight =0.2 (auxiliary target).

[0152] Particle coding design (mapping engineering variables). "Resource allocation" and "schedule adjustment" are transformed into dimensional parameters of particles. Each particle corresponds to a set of feasible solutions, coded as follows: , Resource allocation variable (number of welders to be allocated), value range [0,3] (a maximum of 3 welders can be allocated to work sites with no resource conflicts in the surrounding area. According to the personnel configuration of surrounding work sites in the construction plan: there are 12 welders in each of the pipeline installation shift 1 and shift 3, so resources are sufficient). The progress adjustment variable (efficiency improvement ratio of trench excavation process) has a value range of [0, 0.2] (efficiency improvement of 0~20%, to avoid excessive improvement leading to quality risks).

[0153] Constraints are set (tailored to the actual project). Hard constraints: Pipeline avoidance distance ≥ 0.5m (clearly required by the construction plan to avoid collisions with gas / electricity pipelines); Foundation bearing capacity ≥ 175kPa (verified through on-site load plate tests; adjusted procedures must not disturb the foundation); Total construction period ≤ 45 days (planned total construction period for the work site, cannot be exceeded); No conflict in resource allocation: Welders allocated must come from work sites without parallel critical processes in the vicinity. Soft constraints: Lowest resource allocation cost: Prioritize personnel from nearby work sites to reduce transportation costs; Smooth progress adjustment: Efficiency improvement ≤ 20%, avoiding interface quality problems due to excessively rapid construction.

[0154] 2. Particle Swarm Optimization Algorithm Parameter Configuration (Adapting to Work Site Scale)

[0155] Based on the construction scale of the site (15 days for trench excavation, small resource shortage), the algorithm parameters are set as shown in Table 4 below to ensure efficient convergence of the algorithm:

[0156] Table 4 Algorithm Parameter Table

[0157]

[0158] 3. Iterative calculation process (gradually converging to the optimal solution): Taking this construction site as an example, the algorithm iterative process is as follows, ultimately determining the optimization solution of "allocating 2 welders + increasing efficiency by 10%":

[0159] 30 particles are randomly generated, and each particle corresponds to a group (number of people to be allocated). Efficiency improvement For example: Particle 1, =(0,0.05)→ No welder adjustment, efficiency increased by 5%; Particle 2, = (2, 0.10) → Assign 2 welders, increase efficiency by 10%...... All initial particles must meet hard constraints (such as pipeline avoidance, geological bearing capacity meeting standards), otherwise they will be directly eliminated and regenerated.

[0160] Fitness calculation (evaluating the merits of each scheme), calculating the fitness function for each particle. With particle 2, For example, (2, 0.10): =1 (Personnel relocation does not affect pipeline avoidance, and still meets the ≥0.5m requirement); =1 (Efficiency improved by 10%, foundation bearing capacity still ≥175kPa without disturbing the foundation). =1-∣(45-44) / 45∣≈0.9778 (The original plan was to excavate in 15 days, but it was optimized to be completed in 14 days, with a total construction period of 44 days and a schedule deviation of 2.22%). =10 / 10=1 (After allocating 2 welders, the staff attendance rate is 100%, and the resource utilization rate is optimal); adaptability ≈0.9911. Similarly, the fitness of other particles was calculated, and the optimal particle in the population was found to be particle 2 with a fitness of 0.9911.

[0161] Particle position and velocity update (iterative optimization): The velocity and position of each particle are updated according to the core formula of particle swarm optimization, gradually moving closer to the optimal solution.

[0162] , ;

[0163] in, This represents the historical best position of particle i; The historical best position of the group (initially particle 2). = (2, 0.10)); , Use random numbers in the range [0,1] to increase the randomness of the search.

[0164] Repeat the fitness calculation and particle position and velocity update steps, updating each generation in one iteration. and The algorithm converges when the optimal fitness of the population fluctuates by ≤0.001 for five consecutive generations. After final convergence, the optimal particle in the population is... = (2, 0.10), fitness =0.998, corresponding solutions: Resource allocation: Transfer 2 welders from other work sites in the vicinity with no resource conflicts, increasing the on-site staffing rate from 90% to 100% (10 people at full capacity); Schedule adjustment: Improve the efficiency of subsequent trench excavation processes by 10% (originally an average daily excavation of 13.33m, optimized to an average daily excavation of 14.66m), originally planned to be completed in 15 days, optimized to be completed on the 14th day (80.4% completed on the 12th day, the remaining 19.6% requires 2 days, with an average daily completion of 9.8%, which can be met after the efficiency improvement); Constraints met: Pipeline avoidance distance remains ≥0.5m, foundation bearing capacity is not disturbed, total construction period is 44 days ≤45 days, and resources are not wasted.

[0165] 4. Development of emergency schedule adjustment plans (algorithm support + engineering constraints). The core of the emergency plan is "quantifying the scope of emergency impact + optimizing schedule and resource allocation under emergency conditions" to ensure that unexpected events do not affect the overall project schedule. The specific process is as follows:

[0166] Quantification of the scope of emergency impact ( (Calculation), combining pipeline parameters at the work site and model confidence level, calculate the emergency impact radius: Radius of basic influence =5m (General value for pipeline leakage accidents); Pipeline pressure rating coefficient =1.6; Pipeline material fragility coefficient =0.8 (Ductile iron pipe has better toughness than PPR pipe, and its fragility coefficient is lower than that of PPR pipe by 1.2); Model confidence level C=0.98815 (calculated above, the model has high reliability and small correction amount for emergency impact range); Calculated =7.2m, meaning that the area within 7.2m of the leak point is the emergency impact zone, and the backfilling of the trench within this area must be suspended.

[0167] Optimization of the particle swarm optimization algorithm under emergency conditions (developing response plans): Under emergency conditions, a new objective of "emergency material allocation priority" is added, and the fitness function weights are adjusted.

[0168] ;

[0169] New targets Emergency supplies allocation efficiency ( =1 indicates that emergency supplies will arrive within 30 minutes (weight) =0.3; Adjust the core objective weights: =0.3 (Schedule deviation still needs to be minimized). =0.1 (Resource optimization takes a backseat).

[0170] Through iterative calculations using the particle swarm optimization algorithm, the following emergency plan was derived: Resource allocation: Prioritize the allocation of emergency supplies (2 water pumps, 1 set of leak-sealing equipment) to the accident site, and temporarily transfer 3 emergency personnel from the pipeline installation team; Schedule adjustment: The trench backfilling process was originally planned for 12 days, but the emergency response will take up 1 day. By increasing the backfilling efficiency by 15% (from an average of 41.67m per day to 47.92m per day after optimization), the construction period will be shortened by 1 day, ensuring that the total construction period is still controlled within 45 days; Safety constraints: During the emergency response, a double-layered warning system will be set up in the emergency impact area, and non-emergency personnel are strictly prohibited from entering to avoid secondary accidents.

[0171] 5. Engineering verification of optimization plans and emergency response plans: Verification of conventional optimization plan: After allocating 2 welders, welding efficiency increased by 20% (from an average of 10m per day to 12m per day after optimization). The trench excavation process was successfully completed on the 14th day, and the progress deviation decreased from 19.6% to 2.2%. The total construction period was 44 days, meeting the contract requirements. Verification of emergency response plan: By simulating a pipeline leakage scenario, emergency supplies arrived within 30 minutes, the leakage handling took 8 hours, and the backfilling process efficiency increased by 15%. The total construction period did not exceed 45 days, verifying the feasibility of the plan.

[0172] By using the constraint parameters of the deviation cause and the digital model of underground pipelines, the construction schedule and resource allocation scheme are dynamically adjusted by the particle swarm optimization multi-objective optimization algorithm. This ensures the coordinated advancement of parallel construction at multiple work sites, while meeting the requirements of pipeline avoidance, geological bearing capacity, and safety and environmental protection. It achieves balance of multiple constraints, realizes multi-objective optimization, efficient resource allocation, reduces waste, and ensures scientific emergency plans, thus guaranteeing the construction period and construction safety.

[0173] Step Six: Construction Acceptance and Closed-Loop Management of the Entire Process.

[0174] After the construction of this site is completed, it will enter the acceptance phase. During the construction acceptance phase, the relevant data on construction quality acceptance will be linked and stored with the digital model of underground pipelines to form a progress-quality-pipeline safety associated database, thus completing the closed-loop management of the entire construction progress process.

[0175] Specifically, the construction of the schedule-quality-pipeline safety correlation database includes: binding the completion time of each key schedule node with the corresponding material arrival inspection report, process quality record, pipeline pressure test curve, and water quality test report; at the same time, spatially correlating the thickness and compaction data of the road surface restoration structural layer with the pipeline burial depth and direction data in the underground pipeline digital model to form a three-dimensional correlation index of schedule-location-quality.

[0176] 1. Data binding: Bind the completion time of each key progress node with the corresponding material inspection report (certificate of conformity for ductile iron pipe and welded steel pipe), process quality record (deep trench excavation, verticality of pipe installation), pipeline pressure test curve (test pressure 1.6MPa, pressure stabilization for 30min without leakage), and water quality test report (turbidity ≤1NTU, residual chlorine ≤0.05mg / L);

[0177] 2. Spatial correlation: Spatial correlation was performed between the structural layer data of the road surface restoration (C30 concrete surface layer thickness 150mm, compaction degree 95%; crushed stone cushion layer thickness 100mm, compaction degree 90%) and the pipeline burial depth (1.4m) and direction (east-west) data in the underground pipeline digital model;

[0178] 3. Construct a three-dimensional relational index: Form a three-dimensional relational index of "progress node - spatial location - quality index", such as "pipeline installation completed on September 11, 2025 - central section of the community - pipeline installation qualification rate of 95%", to achieve full traceability of the construction process of this work site. At the same time, the model parameters, progress plan, deviation handling plan and other data of this work site are included in the database to provide data support for other work sites of the water supply network renovation project and subsequent similar projects, and complete the closed-loop management of the entire construction progress process.

[0179] By linking and storing construction quality acceptance data with the digital model of underground pipelines, a progress-quality-pipeline safety related database is formed, completing the closed-loop management of the entire construction progress process. This enables full-process traceability of construction, standardized management, and the accumulation of engineering data, supporting subsequent similar projects and related progress, quality, and safety, and ensuring the long-term operation of the pipeline network.

[0180] Example 2: Implementation of a construction progress management system integrating a digital model of underground pipelines.

[0181] The construction progress management system integrating the digital model of underground pipelines in this embodiment is deployed on the digital management platform of the project management department of the third phase of a water supply network renovation project, section I. The system adopts a B / S architecture, supports multi-terminal access, and the implementation and operation of each module are as follows:

[0182] Digital model building module: Imports basic data such as underground pipeline survey data, geological exploration reports, pipe material parameters, and construction plans from four areas. Through the module's built-in parameter calculation engine, it automatically calculates pipeline topology adaptation parameters. Geological condition adaptation parameters The model is adapted to various indicators and accuracy M, and the model is automatically verified to determine whether it is qualified. In this embodiment, the model calculation results of a certain construction site are automatically generated and visualized. If the model is unqualified, an early warning will be issued and the direction of parameter correction will be indicated.

[0183] Initial schedule generation module: Import data such as contract duration, coordinates of multiple work sites, and boundaries of consortium division of labor. Based on a digital model, the module automatically decomposes construction tasks using an improved critical path method and automatically calculates the progress buffer period for each work site through Monte Carlo simulation algorithm, generating a visualized initial construction schedule (in Gantt chart form). It supports manual fine-tuning of the plan. The schedule Gantt chart for this construction site can display the time nodes, resource allocation, and other information of each process in real time.

[0184] Multi-source data acquisition and dynamic model update module: It connects with the on-site multi-source sensing terminal through the Internet of Things to collect data on pipelines, geology, construction, resources, environmental protection, etc. in real time. The module has a built-in improved Kalman filter algorithm engine to automatically complete the fusion of multi-source data and update the digital model of underground pipelines. The updated state vector of the model is visualized in the form of a digital twin. Data such as pipeline location and construction progress at the work site are updated in real time, realizing digital twin management and control of the project site.

[0185] Schedule Deviation Analysis Module: Automatically compares the actual construction progress with the initial schedule plan, uses the entropy weight-TOPSIS combined algorithm to automatically quantify the schedule deviation value, and calculates and sorts the influence weights of various deviation causes through the built-in deviation cause analysis engine, generating a deviation analysis report. The schedule deviation analysis report for each work site is automatically generated, clearly indicating that "resource deviation is the main cause" and marking the influence percentage.

[0186] The schedule and resource optimization module, based on deviation analysis reports, incorporates a multi-objective particle swarm optimization algorithm engine to automatically generate schedule adjustment plans and resource allocation plans. It supports simulation and deduction of these plans. Once the personnel allocation and schedule adjustment plans for each work site are verified as feasible through simulation, they are automatically distributed to the on-site construction teams. The module also includes an emergency adjustment engine that can automatically calculate the radius of emergency impact. It also generates emergency schedule adjustment plans.

[0187] Acceptance closed-loop management module: It connects to the engineering quality acceptance system, automatically imports quality acceptance data, spatially associates and binds it with the underground pipeline digital model, and builds a progress-quality-pipeline safety related database. The module provides data query, traceability and statistical analysis functions, and supports data retrieval by work point, process, progress node and other dimensions. All construction data of the work point can be retrieved with one click and an acceptance report can be generated, realizing closed-loop management of the whole process.

[0188] The construction progress management method and system of the present invention, which integrates the digital model of underground pipelines, has been implemented and applied at multiple work sites in the third phase of a water supply network renovation project, section I. It has effectively improved the digitalization and scientific level of construction progress management, reduced the delays caused by pipeline, geology, and resource issues, and ensured the smooth progress of the project, demonstrating good engineering application results.

[0189] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A construction progress management method integrating a digital model of underground pipelines, characterized in that, Includes the following steps: Step 1: Construct a digital model of underground pipelines that integrates parameters such as underground pipelines, geological features, pipe material compatibility, and construction constraints. Step 2: Based on the underground pipeline digital model, decompose the construction tasks using the improved critical path method, determine the progress buffer period for each work site using the Monte Carlo simulation algorithm, and generate an initial construction schedule plan. Step 3: During construction, real-time data on underground pipeline dynamic detection, geological exploration, construction procedures, resource occupation, and environmental safety monitoring are collected through multi-source sensing terminals. An improved Kalman filter multi-source data fusion algorithm is then used to dynamically update the digital model of the underground pipeline. Step 4: Based on the updated underground pipeline digital model, the entropy weight-TOPSIS combined algorithm is used to compare the actual and initial construction progress, quantify the progress deviation value, and locate the causes of various related progress deviations. Step 5: Based on the causes of deviations and the constraint parameters of the underground pipeline digital model, the construction schedule and resource allocation scheme are dynamically adjusted using a particle swarm optimization multi-objective optimization algorithm. Step six, the construction acceptance stage, involves linking and storing relevant construction quality acceptance data with the underground pipeline digital model to form a progress-quality-pipeline safety related database.

2. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, In step one, the overall adaptation accuracy of the underground pipeline digital model is obtained by multiplying the four adaptation parameters—pipeline topology, geological conditions, pipe material technology, and construction constraints—by their respective weighting coefficients and then adding them together. The sum of the four weighting coefficients is 1.

3. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, Step one, the construction of the digital model of the underground pipeline also includes: A mapping relationship library is established between pipe connection process parameters and construction efficiency. The mapping relationship library stores the correlation data between the bevel angle, welding current and construction time per unit length for arc welding of welded steel pipes, the correlation data between the rubber ring model and insertion depth and construction time per unit length for socket connection of ductile iron pipes, and the correlation data between the heating temperature and fusion depth and construction time per unit length for hot fusion connection of PPR pipes. At the same time, the geological stratification characteristics are bound to the trench excavation process parameters for storage.

4. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, In step two, the calculation of the progress buffer period first relies on three types of adaptation parameters of the underground pipeline digital model: pipeline topology, geological conditions, and pipe material technology. After 1000 iterations of Monte Carlo simulation, the buffer period correction coefficient is obtained. Then, the progress buffer period is calculated in combination with the basic construction period of the work site. The basic construction period of the work site is determined based on the construction process capacity and construction time data in the mapping relationship library. When using the improved critical path method, the calculation results of the comprehensive adaptation accuracy value and deviation probability of the underground pipeline digital model are used as priority indicators to sort the construction sequence of multiple work sites.

5. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, In step three, the multi-source sensing terminal includes a pipeline detection radar, a geological borehole detector, a process progress sensor, a resource positioning terminal, a dust / noise monitor, and a wastewater quality sensor; the data from the dynamic detection of underground pipelines includes pipeline location correction values ​​and pipeline integrity supplementary data; the data from the geological exploration includes the actual dimensions of the karst area, the measured values ​​of the foundation bearing capacity, and the data on the degree of rock weathering; the data from the construction process includes the progress of trench excavation, the pipeline installation qualification rate, and the pressure test results; and the data from the resource occupancy includes equipment operating status, personnel attendance rate, and material consumption progress.

6. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, In step three, the state update formula for the improved Kalman filter multi-source data fusion algorithm is: ； ； in, The state vector of the underground pipeline digital model after time k is updated, which includes pipeline location correction value, measured geological parameter value, and construction progress node completion rate. The state transition matrix is ​​constructed based on the topology and construction procedure logic of the underground pipeline digital model. The optimal state vector of the underground pipeline digital model after the update at time k-1; To control the input matrix, associated resource allocation parameters are used; For resource allocation and control; Kalman gain; Let k be the measured data vector of the multi-source sensing terminal at time k; For the observation matrix, match the model state with the dimensions of the measured data; Observation matrix The transpose of the matrix; Let k be the state covariance matrix at time k-1; To observe the noise matrix, calibration is performed based on the accuracy of pipeline detection equipment and geological exploration errors. C represents the confidence level of the digital model of underground pipelines.

7. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, In step four, the weighting of the causes of the schedule deviation is calculated based on the confidence level of the underground pipeline digital model and the influence factors of the four types of deviation causes: pipeline, geology, resources, and environmental protection. It is the product of the confidence level of a single type and the corresponding factor divided by the sum of the products of all types. The model confidence level is a weighted sum of pipeline data integrity, geological data matching degree, and pipe material compatibility. The four types of influence factors are the ratio or deviation rate of the corresponding deviation value to the relevant benchmark value.

8. The construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, Step five also includes a method for correcting the impact range of emergency progress adjustments. The corrected emergency impact radius is based on the basic impact radius determined by the accident type, and is obtained by combining the pipeline pressure level coefficient, the pipeline material vulnerability coefficient, and the confidence level of the underground pipeline digital model. The pipeline pressure level coefficient and the material vulnerability coefficient are determined according to the pipeline design pressure and the pipe material type, respectively.

9. A construction progress management method integrating an underground pipeline digital model according to claim 1, characterized in that, Step six, the construction of the progress-quality-pipeline safety association database includes: The completion time of each key progress node is linked to the corresponding material arrival inspection report, process quality record, pipeline pressure test curve, and water quality test report. At the same time, the thickness and compaction data of the road surface restoration structural layer are spatially correlated with the pipeline burial depth and direction data in the underground pipeline digital model to form a three-dimensional correlation index of progress-location-quality.

10. A construction progress management system integrating a digital model of underground pipelines, characterized in that, include: The digital model building module is used to build digital models of underground pipelines that integrate parameters such as underground pipelines, geological features, pipe material compatibility, and construction constraints. The initial schedule generation module is used to decompose construction tasks based on the underground pipeline digital model, decompose the construction tasks using the improved critical path method, determine the progress buffer period of each work point through the Monte Carlo simulation algorithm, and generate an initial construction schedule. The multi-source data acquisition and model dynamic update module is used to collect real-time data on underground pipeline dynamic detection, geological exploration, construction procedures, resource occupation and environmental safety monitoring through multi-source sensing terminals during the construction process, and to dynamically update the digital model of the underground pipeline using an improved Kalman filter multi-source data fusion algorithm. The schedule deviation analysis module is used to compare the actual and initial construction progress based on the updated underground pipeline digital model using the entropy weight-TOPSIS combined algorithm, quantify the schedule deviation value, and locate the causes of various related schedule deviations. The schedule and resource optimization module is used to dynamically adjust the construction schedule and resource allocation scheme based on the causes of deviations and the constraint parameters of the underground pipeline digital model, using a particle swarm optimization multi-objective optimization algorithm. The acceptance closed-loop management module is used during the construction acceptance phase. It links and stores construction quality acceptance data with the underground pipeline digital model to form a progress-quality-pipeline safety related database.

Images