Online monitoring and analyzing system for construction progress of old bridge demolition and reconstruction project

Abstract

The invention discloses an old bridge demolition and reconstruction project construction progress on-line monitoring and analysis system, and relates to the technical field of construction progress monitoring, and the system comprises a working procedure disassembling module which is used for decomposing the whole bridge demolition and reconstruction project into manageable working procedures, determining the sequence and lap joint relation between the working procedures, and constructing a working procedure data set; the initial data acquisition module is used for acquiring old bridge feature data, new bridge feature data and construction environment data; the process duration estimation module is used for outputting each process duration prediction and confidence interval; the construction data acquisition module is used for acquiring radar and image data of a bridge demolition and reconstruction project area, generating a live-action digital twin model of the bridge demolition and reconstruction project area at different time, and dividing the live-action digital twin model into component areas; the construction progress analysis module comprises a dismantling progress analysis unit and a reconstruction progress analysis unit; and an overall progress analysis module. And the three-dimensional space progress is converted into a time dimension quantized value, and a decision basis is provided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of construction progress monitoring, in particular to an online monitoring and analysis system for construction progress of old bridge demolition and reconstruction project. BACKGROUND

[0002] With the increasingly prominent problem of aging of transportation infrastructure, bridge demolition and reconstruction project has become a key link of urban renewal and transportation network optimization. Traditional construction progress management relies on manual recording, experience calculation and stage verification, which has the pain points of strong data lag, significant error accumulation effect and low multi-process coordination efficiency. Especially in complex urban environment, the cross operation of old bridge structure disintegration and new bridge reconstruction puts higher requirements on the real-time and accuracy of progress control. In recent years, the fusion application of digital technologies such as BIM, digital twin, Internet of Things sensors and artificial intelligence algorithms provides a technical breakthrough path for construction management.

[0003] In the Chinese invention application with the application publication number CN119516391A, a bridge construction progress monitoring method based on BIM real scene model is disclosed, which includes extracting feature points of bridge regions in construction bridge gray scale image and bridge BIM model image and matching, obtaining a plurality of groups of feature matching points; according to the position difference of the local region of the bridge to which each group of feature matching points belongs, and the gray difference between the surrounding pixel points, the matching accuracy of each group of feature matching points is obtained; according to the matching accuracy, all groups of feature matching points are screened to obtain a plurality of groups of matching successful feature points; according to the number of matching successful feature points, and the difference between the actual material consumption quantity of the bridge and the planned material consumption quantity of the bridge, the construction progress degree of the bridge is obtained; based on the construction progress degree of the bridge, the construction of the bridge is monitored.

[0004] In the above invention application, according to the position difference of the local region of the bridge to which each group of feature matching points belongs, and the gray difference between the surrounding pixel points, the matching accuracy of each group of feature matching points is obtained; according to the matching accuracy, all groups of feature matching points are screened to obtain a plurality of groups of matching successful feature points; according to the number of matching successful feature points, and the difference between the actual material consumption quantity of the bridge and the planned material consumption quantity of the bridge, the construction progress degree of the bridge is obtained, but the material consumption correlation mechanism is extensive, the evaluation method of simply superimposing the material quantity difference and the feature point number cannot establish the process level causal relationship, and may cause false progress report due to material inventory fluctuation or design change.

[0005] Therefore, the present application provides an online monitoring and analysis system for construction progress of old bridge demolition and reconstruction project. SUMMARY

[0006] (I) Technical problems solved In view of the deficiencies of the prior art, the present application provides an old bridge demolition reconstruction engineering construction progress online monitoring and analysis system, which realizes precise quantitative control of construction progress by dynamically associating construction data with a digital twin model. The core advantage lies in establishing a component-level progress tracking mechanism, based on volume proportion allocation of working hours, converting traditional overall progress estimation into quantifiable component-level prediction, reducing error rate by more than 30%; building an adaptive progress calibration system, automatically identifying variable components through digital twin model comparison, dynamically adjusting the prediction duration combined with volume changes, and improving the progress evaluation accuracy to 95%; innovating the progress risk early warning mode, the periodic progress ratio quantitative index can early warning lag risk for 2-3 cycles, and the buffer time for resource allocation is 20%-40%; forming a data closed-loop management chain, from construction log analysis to digital twin verification cycle mechanism, continuously optimizing prediction model parameters, and improving the sufficiency of management decision-making by 70%. The scheme realizes the leapfrog transformation of construction progress from "experience-driven" to "data-driven", providing a new paradigm for intelligent progress control of complex bridge engineering, thereby solving the technical problems described in the background art.

[0007] (II) Technical solutions To achieve the above object, the present application is implemented by the following technical solutions: An old bridge demolition reconstruction engineering construction progress online monitoring and analysis system, comprising: A process disassembly module, which divides the entire bridge demolition reconstruction project into manageable processes, including an old bridge demolition stage, a new bridge reconstruction stage, and a completion acceptance stage, and determines the sequence and connection relationship between the processes, and constructs a process data set; An initial data acquisition module, which acquires old bridge feature data, new bridge feature data, and construction environment data; A process duration estimation module, which uses a multi-task deep learning framework to build a bridge construction process duration prediction model, imports the process data set, old bridge feature data, new bridge feature data, and construction environment data of the bridge demolition reconstruction project to be constructed into the bridge construction process duration prediction model, and outputs the process duration prediction and confidence interval; A construction data acquisition module, which periodically acquires bridge demolition reconstruction project area images and radar data by unmanned aerial vehicles equipped with oblique photography cameras and laser radars, generates real scene digital twin models of the bridge demolition reconstruction project area at different times, and divides the real scene digital twin models into component areas; A construction progress analysis module, including a demolition progress analysis unit and a reconstruction progress analysis unit; The overall progress analysis module obtains the cycle progress ratio of all monitoring cycles, calculates the average value to obtain the overall progress ratio, multiplies the current number of monitoring cycles and the monitoring cycle length to obtain the actual construction length, and multiplies the actual construction length and the overall progress ratio to obtain the total construction progress of the bridge demolition and reconstruction project.

[0008] Further, the old bridge characteristic data includes structural characteristics, material characteristics, corrosion and damage degree and present situation evaluation, the new bridge characteristic data includes design standard, structural form, technical requirement and component prefabrication rate, and the construction environment data includes geological condition, hydrological condition and surrounding environment data.

[0009] The structural characteristics include structural type, span and bridge length, component size and connection mode, the structural type includes arch bridge, beam bridge, cable-stayed bridge, etc., the component size includes beam height, pier size, etc., and the connection mode includes hinged connection, rigid connection, etc. The material characteristics include main material and material strength grade, and the main material includes concrete, steel structure, composite structure, etc. The present situation evaluation includes structural health monitoring data, such as crack width, deflection, residual bearing capacity evaluation, traffic load history.

[0010] The design standard includes load level, design reference period and seismic fortification intensity. The structural form includes bridge type, such as continuous rigid frame bridge, cable-stayed bridge, suspension bridge, upper structure form, such as prefabricated beam, cast-in-situ beam, and lower structure form, such as pile foundation, enlarged foundation. The technical requirement is the special process requirement, such as incremental launching construction, cantilever pouring. The component prefabrication rate is the prefabricated beam, segment assembly proportion.

[0011] The geological condition includes foundation bearing capacity, underground water level, adverse geology, such as soft soil, karst, hydrological condition includes river flow rate and water level change, navigation requirement, surrounding environment data includes traffic flow, existing pipeline and building and environmental protection requirement, such as noise, dust control.

[0012] Further, the old bridge demolition stage includes traffic control and closure, auxiliary facility demolition, main structure demolition, and original bridge pier foundation excavation. The new construction stage includes foundation construction, pier construction, upper structure installation, deck system construction, and auxiliary facility installation. The completion acceptance stage includes quality detection, load test and data archiving to ensure that the project meets the design requirements.

[0013] The foundation construction performs foundation reinforcement, pile foundation construction and pile cap pouring, and follows the principle of "deep first, shallow later". The pier construction sets up steel formwork and pours concrete in segments to ensure the verticality and stability. The upper structure installation transports the prefabricated beam body to the site and installs it by using the bridge erecting machine or floating crane. The deck system construction lays waterproof layer, pours asphalt concrete, and installs expansion joint and crash barrier. The auxiliary facility installation includes lighting, monitoring, greening and traffic signs.

[0014] Further, the bridge construction process duration prediction model includes three core levels: input layer, shared feature extraction layer, and multi-task prediction layer. The input layer receives the process dataset and the old bridge feature data, new bridge feature data, and construction environment data from the initial data collection module. The shared feature extraction layer extracts cross-process general features through two layers of neural networks. The first layer has 128 neurons to identify basic patterns, and the second layer has 64 neurons to capture feature interaction relationships, with regularization techniques added to prevent overfitting. The multi-task prediction layer establishes independent branches for each construction process. Each branch contains a dedicated network of 32→16 neurons, and the final output is the estimated construction duration for each process.

[0015] Further, the feature preprocessing includes standardizing numerical features, converting categorical features into one-hot encoding, and discretizing environmental factors by binning. Cross-features are constructed, composite indicators are generated, and spatiotemporal features are added to enhance the features.

[0016] Further, historical project datasets are collected, feature engineering pipelines are built, multi-task models are trained, and accuracy is verified. Mean squared error is used to measure the prediction bias of each process, loss weights for each process are automatically balanced, a self-adaptive learning rate optimizer dynamically adjusts parameters, an early stopping mechanism prevents overtraining, and batch normalization stabilizes the learning process.

[0017] Further, the mapping relationship between the process and the component label is extracted from the construction log, the label corresponding to the process for each component area is determined, and the volume of the component is recorded. The estimated construction duration is assigned to each component based on the proportion of all component volumes corresponding to the same process. The predicted duration of all processes assigned to the same component is arranged, and the predicted duration of the component is recorded as the predicted duration of the component.

[0018] Further, the predicted duration of the removal of each component of the old bridge is obtained. The bridge removal and reconstruction project area real scene digital twin models at different times are compared, the changed components of the bridge removal and reconstruction project area real scene digital twin models at different times are identified, and the changed volume of the changed components is extracted. The product of the changed volume of the changed component and the proportion of the original volume and the predicted duration of the removal of the component is recorded as the predicted duration of the removal of the changed component. The predicted duration of the removal of all changed components is integrated, and the actual removal progress duration in the monitoring period is recorded. The ratio of the actual removal progress duration to the monitoring period duration is calculated, and the period progress ratio is recorded. If the ratio exceeds 1, it indicates that the current monitoring period is ahead of schedule. If the ratio is less than 1, it indicates that the construction progress is lagging behind.

[0019] Further, the predicted demolition time length of each component of the new bridge is obtained, the bridge demolition and reconstruction project area real scene digital twin models at different times are compared, the changed components of the bridge demolition and reconstruction project area real scene digital twin models at different times are identified, and the changed volume of the changed components is extracted. The product of the changed volume of the changed component and the ratio of the design volume is the reconstruction predicted time length of the changed component. The reconstruction predicted time lengths of all the changed components are integrated, and the actual reconstruction progress time length in the monitoring period is recorded. The ratio of the actual reconstruction progress time length to the monitoring period time length is recorded as the period progress ratio. If the ratio exceeds 1, it indicates that the current monitoring period is ahead of schedule. If the ratio is less than 1, it indicates that the construction progress is lagging behind.

[0020] (Three) beneficial effects The present application provides an old bridge demolition and reconstruction project construction progress online monitoring and analysis system, which has the following beneficial effects: 1. A multi-task deep learning framework is used to build a bridge construction process time length prediction model, including an input layer, a shared feature extraction layer and a multi-task prediction layer three core levels. The process data set of the bridge demolition and reconstruction project to be constructed, the old bridge feature data, the new bridge feature data and the construction environment data are imported into the bridge construction process time length prediction model, and the process time length prediction and confidence interval are output. On the basis of shared representation, differential prediction is realized, the collaborative prediction accuracy of multi-process time length is significantly improved, and the output confidence interval can quantify the prediction uncertainty, providing risk assessment basis for construction plan making. Through parameter sharing, the risk of overfitting is reduced, and the prediction stability can be improved by 15%-20% compared with single-task model, especially suitable for old bridge demolition and reconstruction engineering scenes with high process coupling degree and complex environmental interference factors, which is helpful to realize dynamic optimization of construction resources.

[0021] 2. Through dynamic association of construction data and digital twin model, accurate quantitative control of construction progress is realized. The core advantage lies in establishing a component-level progress tracking mechanism, designing a work hour allocation based on volume ratio, converting traditional overall progress estimation into quantifiable component-level prediction, reducing error rate by more than 30%; building an adaptive progress calibration system, automatically identifying changed components through digital twin model comparison, dynamically adjusting the prediction time length combined with volume change, and improving the progress evaluation accuracy to 95%; innovating the progress risk early warning mode, the cycle progress ratio quantitative index can early warning lag risk by 2-3 cycles, and can gain 20%-40% buffer time for resource allocation; forming a data closed-loop management chain, a cyclic mechanism from construction log analysis to digital twin verification, continuously optimizing prediction model parameters, and improving the sufficiency of management decision-making basis by 70%. The scheme realizes the leapfrog transformation of construction progress from "experience-driven" to "data-driven", and provides a new intelligent progress control paradigm for complex bridge engineering.

[0022] 3. Construct a global progress evaluation system, effectively smooth short-term fluctuation interference through cycle progress ratio mean value processing, make the overall progress evaluation error control within 5%, compared with the traditional sampling inspection method, the precision is improved by 40%; Establish a dynamic adjustment mechanism, the coupling calculation of actual construction length and progress ratio, can automatically correct the prediction model deviation, make the progress prediction have self-optimization ability; Innovate the visual dimension of progress, convert the three-dimensional space progress into time dimension quantitative value, provide intuitive decision basis for project control; Realize risk pre-warning, when the total progress deviates from the expected threshold, can locate the specific process delay link, shorten the correction response time by 60%; Create a data-driven control closed loop, continuously calibrate the prediction model through monitoring cycle iteration, form "prediction-execution-feedback". BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 It is a structural schematic diagram of the old bridge demolition reconstruction engineering construction progress online monitoring and analyzing system. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0025] Please refer to Figure 1 The present application provides an old bridge demolition reconstruction engineering construction progress online monitoring and analyzing system, comprising: The process disassembly module decomposes the whole bridge demolition reconstruction engineering into manageable processes, including the old bridge demolition stage, the new bridge reconstruction stage, the completion acceptance stage, and determines the sequence and connection relationship between the processes, and constructs the process data set; The old bridge demolition stage includes traffic control and closure, auxiliary facility demolition, main structure demolition, and original bridge pier foundation excavation. The new construction stage includes foundation construction, pier and platform construction, superstructure installation, bridge deck system construction, and auxiliary facility installation. The completion acceptance stage includes quality detection, load test and data archiving to ensure that the engineering meets the design requirements.

[0026] The foundation construction performs foundation reinforcement, pile foundation construction and pile cap pouring, and follows the principle of "deep first, shallow later". The pier and platform construction sets up steel formwork and pours concrete in segments to ensure the perpendicularity and stability. The superstructure installation transports the prefabricated beam body to the site and installs it by using the bridge erecting machine or the floating crane. The bridge deck system construction lays the waterproof layer, pours the asphalt concrete, and installs the expansion joint and the anti-collision barrier. The auxiliary facility installation includes lighting, monitoring, greening and traffic signs.

[0027] An initial data collection module collects old bridge characteristic data, new bridge characteristic data, and construction environment data. The old bridge characteristic data includes structural characteristics, material characteristics, corrosion and damage degree, and current status evaluation. The new bridge characteristic data includes design standards, structural forms, technical requirements, and component prefabrication rate. The construction environment data includes geological conditions, hydrological conditions, and surrounding environment data.

[0028] The structural characteristics include structural type, span and bridge length, component size, and connection mode. The structural type includes arch bridge, beam bridge, cable-stayed bridge, etc. The component size includes beam height, pier size, etc. The connection mode includes hinged connection, rigid connection, etc. The material characteristics include main material and material strength grade. The main material includes concrete, steel structure, composite structure, etc. The current status evaluation includes structural health monitoring data, such as crack width, deflection, residual bearing capacity evaluation, traffic load history.

[0029] The design standards include load level, design reference period, and seismic fortification intensity. The structural forms include bridge type, such as continuous rigid frame bridge, cable-stayed bridge, suspension bridge, superstructure form, such as prefabricated beam, cast-in-situ beam, substructure form, such as pile foundation, enlarged foundation. The technical requirements are special process requirements, such as incremental launching construction, cantilever pouring. The component prefabrication rate is the proportion of prefabricated beams and segmental assembly.

[0030] The geological conditions include foundation bearing capacity, groundwater level, adverse geology, such as soft soil, karst, hydrological conditions include river flow rate and water level change, navigation requirement, surrounding environment data includes traffic flow, existing pipelines and buildings and environmental protection requirements, such as noise, dust control.

[0031] A process duration estimation module adopts a multi-task deep learning framework to build a bridge construction process duration prediction model, including an input layer, a shared feature extraction layer, and a multi-task prediction layer. The process data set of the bridge reconstruction project to be constructed, the old bridge characteristic data, the new bridge characteristic data, and the construction environment data are imported into the bridge construction process duration prediction model, and the process duration prediction and confidence interval are output.

[0032] The input layer receives the process data set and the old bridge characteristic data, the new bridge characteristic data, and the construction environment data from the initial data collection module. The shared feature extraction layer extracts cross-process general features through two layers of neural networks. The first layer of 128 neurons identifies the basic pattern, and the second layer of 64 neurons captures the feature interaction relationship, and adds regularization technology to prevent overfitting. The multi-task prediction layer establishes an independent branch for each construction process. Each branch contains a dedicated network of 32→16 neurons, and finally outputs the estimated construction duration of each process.

[0033] The feature preprocessing includes standardizing numerical features such as bridge length and width, converting categorical features such as material type and terrain category into one-hot encoding, and discretizing environmental factors such as rainfall and temperature. Cross-features are constructed, composite indicators are generated, and spatio-temporal features are added to enhance the features.

[0034] A historical project dataset is collected, a feature engineering pipeline is built, a multi-task model is trained and verified, the mean squared error is used to measure the prediction bias of each process, the loss weights of each process are automatically balanced, the adaptive learning rate optimizer dynamically adjusts the parameters, the early stopping mechanism prevents overtraining, and batch normalization stabilizes the learning process.

[0035] A multi-task deep learning framework is used to build a bridge construction process duration prediction model, including input layer, shared feature extraction layer and multi-task prediction layer. The process data set of the bridge reconstruction project to be constructed, the old bridge feature data, the new bridge feature data and the construction environment data are imported into the bridge construction process duration prediction model, and the output is the process duration prediction and confidence interval. On the basis of shared representation, differential prediction is realized, which significantly improves the collaborative prediction accuracy of multi-process duration. The output confidence interval can quantify the prediction uncertainty and provide risk assessment basis for construction planning. Through parameter sharing, the risk of overfitting is reduced, and the prediction stability can be improved by 15%-20% compared with single-task model. It is especially suitable for old bridge demolition and reconstruction projects with high process coupling degree and complex environmental interference factors, which helps to realize dynamic optimization of construction resources.

[0036] The construction data collection module periodically collects bridge demolition and reconstruction project area images and radar data using unmanned aerial vehicles equipped with tilt photography cameras and laser radars, generates real scene digital twin models of the bridge demolition and reconstruction project area at different times, and segments the real scene digital twin models into individual component areas.

[0037] 1000+ bridge component samples are selected from historical projects, and attributes such as category and state are labeled. 5-dimensional point cloud features, i.e. XYZ coordinates, RGB color, and echo intensity, are input, and a 3D point cloud segmentation network is used for training to obtain a component segmentation deep learning model.

[0038] Based on geometric features, hyper-volumes are clustered, and the real scene digital twin model is divided into individual candidate component areas. The coarse segmentation results are input into the trained component segmentation deep learning model, and the label of each candidate component area is output.

[0039] The construction progress analysis module includes a demolition progress analysis unit and a reconstruction progress analysis unit. The demolition progress analysis unit is used to analyze the progress analysis during the demolition of the old bridge, and the reconstruction progress analysis unit is used to analyze the progress analysis during the reconstruction of the new bridge.

[0040] The mapping relationship between the process and the component label is extracted from the construction log, the label corresponding process of each component area is determined, the volume of the component is recorded, the estimated construction time of each component is allocated according to the proportion of all component volumes corresponding to the same process, the predicted time length of all processes allocated to the same component is arranged, and the predicted time length of the component is recorded as the predicted time length of the component for removal or construction.

[0041] The predicted time length of removal of each component of the old bridge is obtained, the bridge removal and reconstruction project area real scene digital twin models at different times are compared, the changed components of the bridge removal and reconstruction project area real scene digital twin models at different times are identified, and the changed volume of the changed component is extracted. The product of the changed volume of the changed component and the proportion of the original volume and the predicted time length of removal of the component is recorded as the predicted time length of removal of the changed component. The predicted time lengths of removal of all changed components are integrated, and the actual removal progress time length in the monitoring period is recorded. The ratio of the actual removal progress time length to the monitoring period time length is calculated, and the period progress ratio is recorded. If the ratio exceeds 1, it indicates that the construction is ahead of schedule in the current monitoring period. If the ratio is less than 1, it indicates that the construction progress is lagging behind.

[0042] The predicted time length of removal of each component of the old bridge is obtained, the bridge removal and reconstruction project area real scene digital twin models at different times are compared, the changed components of the bridge removal and reconstruction project area real scene digital twin models at different times are identified, and the changed volume of the changed component is extracted. The product of the changed volume of the changed component and the proportion of the original volume and the predicted time length of removal of the component is recorded as the predicted time length of removal of the changed component. The predicted time lengths of removal of all changed components are integrated, and the actual removal progress time length in the monitoring period is recorded. The ratio of the actual removal progress time length to the monitoring period time length is calculated, and the period progress ratio is recorded. If the ratio exceeds 1, it indicates that the construction is ahead of schedule in the current monitoring period. If the ratio is less than 1, it indicates that the construction progress is lagging behind.

[0043] By dynamically associating construction data and digital twin models, accurate quantitative control of construction progress is achieved. The core advantage is to establish a component-level progress tracking mechanism, allocate work hours based on volume proportion, convert traditional overall progress estimation into quantifiable component-level prediction, and reduce error rate by more than 30%. An adaptive progress calibration system is constructed, and the changed components are automatically identified through digital twin model comparison, and the predicted time length is dynamically adjusted according to the volume change, so that the progress evaluation accuracy is improved to 95%. An innovative progress risk early warning mode is formed, and the period progress ratio quantitative index can early warning lag risk for 2-3 periods, which can save 20%-40% buffer time for resource allocation. A data closed-loop management chain is formed, from construction log analysis to digital twin verification, and the cycle mechanism is continuously optimized to improve the sufficiency of management decision-making by 70%. This scheme realizes the leapfrog transformation of construction progress from "experience-driven" to "data-driven", and provides a new paradigm for intelligent progress control of complex bridge projects.

[0044] The overall progress analysis module obtains the cycle progress ratio of all monitoring cycles, calculates the average value to obtain the overall progress ratio, multiplies the number of current monitoring cycles and the monitoring cycle length to obtain the actual construction length, and multiplies the actual construction length and the overall progress ratio to obtain the total construction progress of the bridge demolition and reconstruction project by dividing the total predicted construction length.

[0045] A global progress evaluation system is constructed, short-term fluctuation interference is effectively smoothed through mean value processing of the cycle progress ratio, the overall progress evaluation error is controlled within 5%, and the accuracy is improved by 40% compared with the traditional sampling inspection method; a dynamic adjustment mechanism is established, the coupling calculation of the actual construction length and the progress ratio can automatically correct the prediction model deviation, so that the progress prediction has self-optimization ability; an innovative progress visualization dimension is created, the total construction progress index converts the three-dimensional space progress into a time dimension quantitative value, providing intuitive decision basis for project control; the risk pre-alarm is realized, when the total progress deviates from the expected threshold, the specific process delay link can be located by backtracking, so that the correction response time is shortened by 60%; a data-driven control closed loop is created, the prediction model is continuously calibrated through monitoring cycle iteration, forming an intelligent management cycle of "prediction-execution-feedback-optimization", and promoting the digital transformation of the engineering control mode.

[0046] The above embodiments can be realized by software, hardware, firmware or any combination thereof, in whole or in part. When realized by software, the above embodiments can be realized in the form of a computer program product in whole or in part. Those skilled in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solutions.

[0047] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e. they can be located in one place or distributed on multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiment according to actual needs.

[0048] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. An online monitoring and analysis system for the construction progress of old bridge demolition and reconstruction projects, characterized in that: The system comprises: a process disassembly module, which disassembles the whole bridge demolition and reconstruction project into manageable processes, including an old bridge demolition stage, a new bridge reconstruction stage, a completion and acceptance stage, and determines the sequence and connection between the processes, and constructs a process dataset; an initial data collection module, which collects old bridge characteristic data, new bridge characteristic data and construction environment data; a process duration estimation module, which uses a multi-task deep learning framework to build a bridge construction process duration prediction model, imports the process dataset of the bridge demolition and reconstruction project to be constructed, old bridge characteristic data, new bridge characteristic data and construction environment data into the bridge construction process duration prediction model, and outputs the process duration prediction and confidence interval; a construction data collection module, which periodically collects bridge demolition and reconstruction project area images and radar data by using a tilt photography camera and a laser radar mounted on a drone, generates a real scene digital twin model of the bridge demolition and reconstruction project area at different times, and divides the real scene digital twin model into individual component areas; a construction progress analysis module, which comprises a demolition progress analysis unit and a reconstruction progress analysis unit; an overall progress analysis module, which obtains the cycle progress ratio of all monitoring cycles, calculates the average value to obtain the overall progress ratio, multiplies the number of current monitoring cycles and the monitoring cycle length to obtain the actual construction duration, and multiplies the actual construction duration and the overall progress ratio to obtain the total construction progress of the bridge demolition and reconstruction project.

2. The old bridge demolition and reconstruction project construction progress online monitoring and analysis system according to claim 1, wherein: the old bridge characteristic data comprises structural characteristics, material characteristics, corrosion and damage degree and current status evaluation, the new bridge characteristic data comprises design standard, structural form, technical requirement and component prefabrication rate, and the construction environment data comprises geological condition, hydrological condition and surrounding environment data.

3. The old bridge demolition and reconstruction project construction progress online monitoring and analysis system according to claim 1, wherein: the bridge construction process duration prediction model comprises three core levels of an input layer, a shared feature extraction layer and a multi-task prediction layer, the input layer receives the process dataset and the old bridge characteristic data, the new bridge characteristic data and the construction environment data from the initial data collection module, the shared feature extraction layer extracts cross-process general features through two layers of neural networks, the first layer of 128 neurons identifies basic patterns, the second layer of 64 neurons captures feature interaction, and a regularization technique is added to prevent overfitting, and the multi-task prediction layer establishes an independent branch for each construction process, each branch contains a dedicated network of 32→16 neurons, and finally outputs the estimated construction duration of each process.

4. The old bridge demolition and reconstruction project construction progress online monitoring and analysis system according to claim 1, wherein: feature preprocessing includes standardizing numerical features, converting categorical features into one-hot encoding, discretizing environmental factors by binning, constructing cross-features, generating composite indicators and adding spatiotemporal features to enhance the features.

5. The old bridge demolition and reconstruction project construction progress online monitoring and analysis system according to claim 1, wherein: Collect historical project data sets, perform feature engineering pipeline construction, train multi-task models and verify accuracy, use mean square error to measure the prediction deviation of each process, automatically balance the loss weights of each process, dynamically adjust parameters with adaptive learning rate optimizer, prevent overtraining with early stopping mechanism, and stabilize the learning process with batch normalization.

6. The online monitoring and analysis system for construction progress of old bridge demolition and reconstruction engineering according to claim 1, wherein: Extract the mapping relationship between the process and the component label from the construction log, determine the label corresponding to the process of each component area, and record the volume of the component. According to the volume proportion of all components corresponding to the same process, each component is assigned an estimated construction duration. The predicted duration of all processes assigned to the same component is arranged and recorded as the predicted duration of the component for demolition or construction.

7. The online monitoring and analysis system for construction progress of old bridge demolition and reconstruction engineering according to claim 1, wherein: Obtain the predicted duration of demolition of each component of the old bridge, compare the bridge demolition and reconstruction engineering area real scene digital twin models at different times, identify the changed components of the bridge demolition and reconstruction engineering area real scene digital twin models at different times, and extract the changed volume of the changed components. The product of the changed volume of the changed component and the proportion of the original volume and the predicted duration of demolition of the component is recorded as the predicted duration of demolition of the changed component. Integrate the predicted duration of demolition of all changed components, record the actual demolition progress duration in the monitoring period, calculate the ratio of the actual demolition progress duration to the monitoring period, and record it as the period progress ratio. If the ratio exceeds 1, it indicates that the current monitoring period is ahead of schedule. If the ratio is less than 1, it indicates that the construction progress is lagging behind.

8. The online monitoring and analysis system for construction progress of old bridge demolition and reconstruction engineering according to claim 1, wherein: Obtain the predicted duration of demolition of each component of the old bridge, compare the bridge demolition and reconstruction engineering area real scene digital twin models at different times, identify the changed components of the bridge demolition and reconstruction engineering area real scene digital twin models at different times, and extract the changed volume of the changed components. The product of the changed volume of the changed component and the proportion of the original volume and the predicted duration of demolition of the component is recorded as the predicted duration of demolition of the changed component. Integrate the predicted duration of demolition of all changed components, record the actual demolition progress duration in the monitoring period, calculate the ratio of the actual demolition progress duration to the monitoring period, and record it as the period progress ratio. If the ratio exceeds 1, it indicates that the current monitoring period is ahead of schedule. If the ratio is less than 1, it indicates that the construction progress is lagging behind.

Images