BIM + AI-based steel bridge full-life-cycle design method and system and medium

Abstract

The invention relates to a BIM + AI-based steel bridge full-life-cycle design method and system and a medium, and the method comprises the following steps: constructing a steel bridge BIM parameterized component library, the component library comprises standardized parameter components of beam segments, nodes, supports and connecting pieces, and each component has multi-dimensional data of geometric information, material attributes, mechanical parameters and cost information; establishing a data standard based on an IFC extension standard, and newly adding design parameters, material attributes, a construction process, a monitoring sensor and a steel bridge specific attribute set of maintenance records on the basis of an IFC framework; and deploying an AI engine, wherein the AI engine comprises a structure type selection recommendation model based on the historical project database. Through AI-driven intelligent design assistance, the structure type selection recommendation model can quickly give technical feasibility scores, economic scores and comprehensive recommendation indexes based on not less than 10,000 groups of historical project data, and a traditional trial and error process depending on experience is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of building information modeling and artificial intelligence technology, and in particular to a BIM+AI-based steel bridge life cycle design method, system and medium. BACKGROUND

[0002] As an important part of modern transportation infrastructure, steel bridges have the advantages of light weight, strong spanning capacity, short construction period, and recyclability, and are widely used in long-span bridge construction. Traditional steel bridge design methods mainly rely on the experience and manual calculation of designers, and the design process involves multiple links such as structure selection, section optimization, node design, and construction scheme development. Each link is relatively independent, information transmission relies on drawings and documents, and there are problems such as long design cycle, serious information island, and difficult optimization of design scheme. At the same time, the traditional design method lacks overall consideration of the whole life cycle of the steel bridge, and the design stage is disconnected from the construction and maintenance stages, which may lead to problems such as construction difficulties and high maintenance costs, and cannot achieve the performance optimization and cost minimization of the whole life cycle of the steel bridge.

[0003] With the development of building information modeling (BIM) technology, its application in bridge engineering is increasing. BIM technology establishes a three-dimensional digital model containing geometric information, physical information, and functional information, realizes integrated management and visual expression of design information, and to some extent solves the information island problem in traditional design methods. However, the existing BIM technology in steel bridge design mainly stays at the level of three-dimensional modeling and collision checking, lacks intelligent design assistance functions, and designers still need to rely on experience for scheme decision-making. In addition, the existing BIM model is mostly a static model, which is difficult to reflect the actual state changes in the construction process and the performance degradation in the maintenance stage, and cannot form a data closed loop of design, construction, and maintenance, limiting the application effect of BIM technology in the whole life cycle management of steel bridges.

[0004] The rapid development of artificial intelligence (AI) technology provides a new way to solve the above problems, but the application of AI technology in steel bridge design is still in the exploratory stage, and there is a lack of systematic integrated solutions. Existing researches mainly focus on single-point applications, such as using machine learning for structure damage identification and using optimization algorithms for section design, and have not formed a complete technology system covering the whole life cycle of steel bridges. At the same time, the training of AI models relies on a large amount of high-quality data, while steel bridge engineering data is scattered in the design, construction, and maintenance stages, lacking a unified data standard and collection mechanism, which makes it difficult for AI technology to realize its potential. Therefore, it is urgent to develop a steel bridge whole life cycle design method that deeply integrates BIM and AI, realizes intelligent design, precise construction, and smart maintenance, and improves the overall quality and efficiency of steel bridge engineering. SUMMARY

[0005] In view of the low design efficiency of steel bridges, the fragmentation of life cycle data, and the lack of intelligent design assistance in the prior art, the present application provides a steel bridge life cycle design method, system and medium based on BIM+AI, which realizes the intelligentization and data closed-loop management of the whole process of steel bridge design, construction and operation and maintenance through the deep integration of BIM technology and AI technology.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0007] In a first aspect, the present application provides a steel bridge life cycle design method based on BIM+AI, comprising the following steps: constructing a steel bridge BIM parameterized component library, the component library containing standardized parameter components of beam segments, nodes, supports and connecting pieces, each component having multidimensional data of geometric information, material properties, mechanical parameters and cost information; establishing a data standard based on IFC extension standard, adding design parameters, material properties, construction technology, monitoring sensors and maintenance records of steel bridge specific attribute set on the basis of IFC framework; deploying an AI engine, the AI engine including a structure selection recommendation model based on a historical project database, an intelligent optimization model of cross section and node using a deep neural network combined with a topology optimization algorithm, and a construction feasibility prediction model based on a graph neural network; collecting steel performance data, welding quality data and construction monitoring data through an IoT sensor network in the construction stage, and automatically updating the BIM model to form a completed BIM model; in the operation and maintenance stage, real-time data stream of a structure health monitoring system is accessed, the residual life of the structure is predicted through a time series prediction algorithm and a machine learning model, and a maintenance decision is generated; the performance degradation data, damage evolution law and maintenance effect evaluation in the operation and maintenance stage are fed back to the AI training database, and the AI model is optimized through transfer learning and incremental learning.

[0008] Further, the structure selection recommendation model is constructed based on a database containing not less than 10,000 groups of historical bridge project data, covering bridge type selection cases under different geographical environments, load conditions and economic indicators, the model uses a reinforcement learning algorithm, inputs bridge site environment parameters, load conditions and economic indicators, and outputs the optimal bridge type scheme, its technical feasibility score, economic score and comprehensive recommendation index.

[0009] Further, the deep neural network of the intelligent optimization model of cross section and node includes an input layer, multiple hidden layers and an output layer, inputs parameters such as span, load and material properties, and outputs optimized cross section size, stiffener arrangement scheme and node construction details, the topology optimization algorithm finds the optimal material distribution under the constraints of strength, stiffness and stability, realizes lightweight design of the structure, and optimizes the response time to be controlled within 30 seconds for cross section optimization and 2 minutes for node optimization.

[0010] Further, the construction feasibility prediction model uses a graph neural network to model the construction process, taking hoisting equipment, components, and temporary supports as nodes in the graph and their mutual relationships as edges, learns the complex dependency relationships in the construction process, predicts the rationality of hoisting sequence, the feasibility of welding process, and the safety of temporary support scheme, and outputs feasibility scores and risk levels. The F1-score of the model prediction is above 0.85.

[0011] Further, the IoT sensor network includes strain gauges, displacement meters, temperature sensors, and inclinometers to collect real-time data on actual material performance, welding quality detection results, and stress and strain deformation at key positions, and upload the data to the cloud platform through wireless transmission. The system automatically analyzes the data and updates the corresponding attributes of the BIM model to realize digital recording of the construction process.

[0012] Further, the structural health monitoring system includes vibration data collected by acceleration sensors, deformation data collected by displacement sensors, and corrosion rate data collected by corrosion monitoring sensors. An AI model performs real-time analysis based on the monitoring data, uses a time series prediction algorithm to predict performance degradation trends, evaluates the structural health status, and generates personalized recommendations including detection, maintenance, and reinforcement schemes.

[0013] Further, a BIM collaborative working environment based on a cloud platform is established to support real-time collaboration among professionals in structural, mechanical and electrical, and construction specialties on the same platform using a distributed architecture. The system automatically manages versions, records the time, person, and content of each modification, and ensures traceability in the design process.

[0014] In a second aspect, the present application provides a steel bridge life cycle design system based on BIM+AI, including: a data layer for building a unified data platform and realizing unified storage and management of heterogeneous data sources such as BIM model data, AI model parameters, real-time monitoring data, and design specification library through data lake technology; an engine layer including an AI computing engine supporting TensorFlow and PyTorch deep learning frameworks, a BIM engine realizing deep integration with mainstream BIM software through RevitAPI and NavisworksAPI, and a rule engine with built-in national and industry design specifications for automatic compliance verification; an application layer using SaaS design to provide AI-driven intelligent design assistant modules, conflict warning modules for real-time detection of multi-specialty design conflicts, life prediction modules based on monitoring data, and maintenance recommendation modules for generating personalized schemes; the system uses a microservice architecture, with each functional module being independently deployed and extended, service calling through an API gateway, and supporting flexible deployment in private clouds, public clouds, or hybrid cloud environments.

[0015] In a third aspect, the present application provides a computer readable medium, wherein a computer program is stored on the computer readable medium, and the computer program is executed by a processor to implement the BIM+AI-based steel bridge life cycle design method.

[0016] The present application realizes intelligent management of the whole process of the steel bridge from design, construction to operation and maintenance by constructing a BIM digital twin base, integrating an AI intelligent engine and establishing a full life cycle data closed loop, significantly improves design efficiency and quality, reduces construction and operation and maintenance costs, and provides a complete technical solution for the digitalization and intelligent transformation of steel bridge projects.

[0017] The present application has the following advantages:

[0018] 1. Greatly improve design efficiency and quality. Through AI-driven intelligent design assistance, the structure selection recommendation model based on no less than 10,000 historical project data can quickly give technical feasibility score, economic score and comprehensive recommendation index, avoiding the traditional trial-and-error process relying on experience; the cross-section and node intelligent optimization model has fast response speed, the cross-section optimization is completed within 30 seconds, and the node optimization is completed within 2 minutes, the design cycle is shortened by 42%. At the same time, the optimized design scheme is more reasonable, the steel consumption in the implementation example is reduced by 15.3%, the node fatigue life is increased by 35%, and the optimal balance between structure performance and economy is achieved.

[0019] 2. Realize full life cycle data closed loop management. The present application establishes a complete data link from design to construction and then to operation and maintenance, the construction phase collects material performance, welding quality, construction monitoring and other data in real time through the IoT sensor network, automatically updates the BIM model, and controls the geometric deviation between the completion model and the actual structure within ±10mm; the performance degradation data and damage evolution law in the operation and maintenance phase are fed back to the AI training database, and the model performance is continuously optimized through transfer learning and incremental learning. This data closed loop mechanism enables the experience of each project to be converted into system wisdom, continuously improving the design quality of subsequent projects.

[0020] 3. Significantly improve construction and operation and maintenance effect. The F1-score of the construction feasibility prediction model reaches more than 0.85, which can accurately predict the feasibility of hoisting sequence, welding process and temporary support scheme, effectively avoid construction risks, and the construction rework rate is reduced to 0.8% in the implementation example. In the operation and maintenance phase, the system performs real-time analysis based on structure health monitoring data, accurately predicts the remaining life of the structure and generates personalized maintenance suggestions, and the operation and maintenance cost is reduced by 28%. At the same time, the system can discover potential problems in advance, such as discovering the stiffness degradation of the cross diaphragm and giving reinforcement suggestions in the implementation example, ensuring the long-term safe operation of the bridge.

[0021] 4. Promote multi-disciplinary collaboration and standardization. The cloud platform-based distributed BIM collaborative environment supports real-time collaborative design of multi-disciplinary such as structure, mechanical and electrical, construction, automatic version management ensures the traceability of the design process; unified IFC extension standard and parametric component library promotes the standardization and standardization of steel bridge design, improves the reusability and interoperability of design results, and lays a solid foundation for the digital transformation of steel bridge engineering. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the present application or prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only a part of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0023] Fig. 1 The BIM+AI-based steel bridge life cycle design method flowchart of the present application;

[0024] Fig. 2 The three core model architecture of the AI engine of the present application;

[0025] Fig. 3 The system architecture and data flow diagram of the present application;

[0026] Fig. 4 The full life cycle data collection and feedback mechanism diagram of the present application. DETAILED DESCRIPTION

[0027] The present application will be described in detail below in conjunction with the drawings and specific embodiments. It should be noted that in order to make the embodiments more detailed, the following embodiments are the best, preferred embodiments, and other alternative ways can also be used by those skilled in the art to implement them; and the drawings are only used to describe the embodiments more specifically, and are not intended to specifically limit the present application.

[0028] It should be noted that in the specification, "one embodiment", "embodiment", "exemplary embodiment", "some embodiments" and the like indicate that the described embodiments can include specific features, structures or characteristics, but not necessarily every embodiment includes the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of those skilled in the related art to implement such a feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0029] In general, terminology can be understood at least in part from the context of use. For example, the term "one or more" as used herein, depending at least in part upon context, can be used to describe any feature, structure, or characteristic in a singular sense or can be used to describe combinations of features, structures or characteristics in a plural sense. Additionally, the term "based on" can be understood as not necessarily being confined to factors that are exclusively related to the factor that is based on, but can instead be broadly understood as being based on at least in part on a factor that can or can not have any direct relation to the factor that is based on.

[0030] Referring to Figs. 1 to 4 as shown

[0031] The application provides a "BIM+AI" based steel bridge life cycle design method, system and computer readable medium, which will be further described in detail below in combination with specific embodiments.

[0032] I. Construction and implementation of steel bridge life cycle BIM digital twin base

[0033] The application first constructs a unified steel bridge BIM parameterized component library, which contains standardized parameter components such as beam segments, nodes, supports, connectors, etc. Each component has complete geometric information, material properties, mechanical parameters, cost information and other multi-dimensional data. The component library adopts a parameterized design idea, supports quickly generating component instances of different specifications by adjusting key parameters. In terms of data standard definition, the application adopts IFC extension standard, adds a set of properties specific to steel bridges on the basis of the original IFC framework, including design parameters such as design load, safety factor, fatigue life; material properties such as steel grade, yield strength, elastic modulus; construction technology such as welding method, hoisting sequence, temporary support scheme; monitoring sensors such as strain gauge position, accelerometer number, data acquisition frequency; maintenance records such as detection date, damage type, repair measures and other life cycle information. The BIM collaborative working environment based on the cloud platform adopts a distributed architecture, supports real-time collaborative design of multi-professional personnel such as structure professionals, mechanical and electrical professionals, and construction professionals on the same platform, and the system automatically performs version management, records the time, modifier, and modification content of each modification, and ensures the traceability of the design process.

[0034] II. Realization of AI engine integration and intelligent design assistance

[0035] In terms of AI engine integration, the application deploys three types of core AI models. The structure selection recommendation model is based on a historical project database, which contains no less than 10,000 sets of historical bridge project data, covering different geographical environments, load conditions, and economic indicators for bridge type selection cases. The model uses reinforcement learning algorithm, by inputting bridge site environment parameters: such as topography, hydrological conditions, seismic intensity; load conditions: such as design lane number, design load grade, crowd load; economic indicators: such as total investment limit, per square meter cost requirement, etc., automatically recommends the optimal bridge type scheme, such as continuous steel box girder, steel truss, cable-stayed bridge, etc., the recommended results include the technical feasibility score, economic score and comprehensive recommendation index of the scheme. The cross-section and node intelligent optimization model uses deep neural network combined with topology optimization algorithm, the DNN network includes input layer, multiple hidden layers and output layer, the input parameters include span, load, material properties, etc., the output is the optimized cross-section size, stiffener arrangement scheme, node construction details, etc. The topology optimization algorithm finds the optimal distribution of materials under the constraints of strength, stiffness, stability, etc., to achieve lightweight design of the structure. The optimization process response time is controlled within 30 seconds for cross-section optimization and less than 2 minutes for node optimization. The construction feasibility prediction model is based on construction simulation data, using graph neural network to model the construction process, taking hoisting equipment, components, temporary supports, etc. as nodes in the graph, and mutual relationships as edges, learning the complex dependency relationships in the construction process through GNN, predicting the rationality of hoisting sequence, the feasibility of welding process, the safety of temporary support scheme, outputting the feasibility score and risk level, the F1-score of the model prediction reaches more than 0.85.

[0036] III. Establishment of full life cycle data closed loop mechanism

[0037] In the construction phase, the system realizes automatic data collection and dynamic BIM model updating through Internet of Things technology. The IoT sensor network includes strain gauges, displacement meters, temperature sensors, inclinometers, etc., which collect real-time material performance data such as actual yield strength and elastic modulus of steel, welding quality data such as ultrasonic testing results and radiographic testing results of welds, and construction monitoring data such as stress, strain, and deformation of key parts. These data are uploaded to the cloud platform through wireless transmission, and the system automatically analyzes the data and updates the corresponding attributes in the BIM model to form a "completed BIM" model reflecting the actual construction state. In the operation and maintenance phase, the system accesses real-time data streams from the structural health monitoring system, including vibration data collected by acceleration sensors, deformation data collected by displacement sensors, and corrosion rate data collected by corrosion monitoring sensors. The AI model performs real-time analysis based on these monitoring data, uses time series prediction algorithms and machine learning models to predict the remaining life of the structure, assess the current health status, and generate maintenance decision recommendations. The system establishes a complete "operation-design" feedback mechanism, which inputs performance degradation data, damage evolution rules, and maintenance effectiveness evaluation information collected during the operation and maintenance phase into the AI training database, continuously optimizes the prediction accuracy of the AI model and the quality of design recommendations through transfer learning and incremental learning techniques, and makes the design parameters and safety redundancy settings of future projects more reasonable.

[0038] IV. Technical implementation of system architecture

[0039] The "BIM+AI" collaborative platform system adopts a layered architecture design. The data layer builds a unified data platform, integrating and managing heterogeneous data sources such as BIM model data, AI model parameters, real-time monitoring data, and design specification library, and using data lake technology to realize the unified storage and management of structured, semi-structured, and unstructured data. The engine layer includes three core engines: the AI computing engine supports TensorFlow and PyTorch deep learning frameworks, providing model training, inference, and updating functions; the BIM engine realizes deep integration with mainstream BIM software through RevitAPI and NavisworksAPI, supporting model creation, editing, querying, and analysis operations; the rule engine has built-in national and industry-related design specifications, enabling automatic compliance checking of design schemes. The application layer adopts SaaS design, providing intelligent design assistant modules (including AI-driven scheme recommendation, parameter optimization, performance prediction, etc.), conflict warning modules (real-time detection of multi-specialty design conflicts and provision of solutions), life prediction modules (prediction of remaining service life based on monitoring data), and maintenance recommendation modules (generation of personalized detection, maintenance, and reinforcement schemes). The entire system adopts a microservices architecture, with each functional module deployed and extended independently, and service calls made through an API gateway, supporting flexible deployment in private clouds, public clouds, or hybrid clouds.

[0040] Example 1: Design application of a certain highway river-crossing steel box girder bridge

[0041] A certain highway needs to build a river-crossing bridge, with a main span of 200 meters, a design load of highway-I level, and an earthquake fortification intensity of 7 degrees. The project uses the "BIM+AI" system of the invention for whole life cycle design. In the early stage of the project, the design team inputs the basic information of the bridge site into the system, including terrain data (river width 195 meters, land height difference 3.5 meters), hydrological data (100-year flood level 25.6 meters, normal water level 18.2 meters), geological data (overburden thickness 8-12 meters, bedrock is moderately weathered granite), traffic demand (six-lane in both directions, design speed 100 km / h), etc. After calculation and analysis by the AI structure selection recommendation model, three recommended schemes are given: continuous steel box girder bridge (technical feasibility score 92, economic score 88, comprehensive recommendation index 90), steel truss bridge (technical feasibility score 88, economic score 85, comprehensive recommendation index 86), and cable-stayed bridge (technical feasibility score 95, economic score 78, comprehensive recommendation index 85). According to the comprehensive evaluation, the system recommends the continuous steel box girder scheme.

[0042] In the detailed design stage, the cross-section optimization AI model intelligently optimizes the main girder cross-section. The initial design adopts a single-box three-chamber cross-section, with a top plate thickness of 20 mm, a bottom plate thickness of 16 mm, and a web thickness of 14 mm. After AI optimization, the system gives the optimization scheme: the top plate adopts variable thickness design, 16 mm in the middle of the span and 24 mm at the support point; the bottom plate is optimized as an orthotropic plate with a longitudinal rib spacing of 300 mm and a transverse rib spacing of 2500 mm; the web is locally thickened to 16 mm in areas with large shear. After optimization, the steel consumption is reduced by 15.3%, while meeting the design requirements of strength, stiffness, stability, and fatigue, etc. In the node design, the AI model performs topology optimization on the connection node of the diaphragm and the main girder, and uses circular arc transition to reduce stress concentration. The fatigue life of the optimized node is increased by 35%.

[0043] In the construction stage, the project adopts a construction scheme of segmental precast and on-site assembly. The IoT sensor network includes 120 strain gauges, 48 displacement meters, and 36 inclinometers, distributed at key cross-sections and node positions. During the hoisting process of the beam segment, the system collects real-time hoisting stress data, and when it finds that the hoisting stress of a certain beam segment exceeds 85% of the design value, the AI model immediately issues a warning and suggests adjusting the hoisting point position. During construction, the system automatically updates the BIM model daily, recording information such as actual installation position, welding quality test results, high-strength bolt pre-tension, etc. When the project is completed, the geometric deviation between the completion BIM model generated by the system and the actual structure is controlled within ±10 mm.

[0044] In the operation and maintenance stage, the bridge is installed with a structural health monitoring system including 32 acceleration sensors, 24 strain sensors and 8 displacement sensors, and the data acquisition frequency is once every 5 minutes. After two years of operation, the system found that the No. 3 cross diaphragm of the main span showed signs of stiffness degradation by analyzing the change trend of vibration frequency. The AI life prediction model evaluated that the remaining safe service life of this position at the current degradation rate is 18 years, which is lower than the requirement of 100 years of design service life. The system automatically generates a maintenance suggestion: add stiffening ribs at this cross diaphragm position, which is expected to increase the remaining life to more than 85 years. At the same time, this finding is fed back to the AI design model, and in subsequent similar projects, the system automatically strengthens the design of the cross diaphragm to avoid the recurrence of similar problems. Through the application of the present application, the design cycle of the project is shortened by 42%, the steel consumption is saved by 15.3%, the construction rework rate is reduced to 0.8%, and the operation and maintenance cost is reduced by 28%, fully verifying the significant advantages of "BIM+AI" technology in the whole life cycle management of steel bridges.

[0045] The present application encompasses any substitutions, modifications, equivalent methods and solutions made to the essence and scope of the present application. In order for the public to have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can be fully understood without the description of these details to those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits, etc. are not described in detail.

[0046] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which should be considered as the protection scope of the present application.

Claims

1. A BIM+AI-based full life-cycle design method for steel bridges, characterized in that, Includes the following steps: Construct a BIM parametric component library for steel bridges. The component library contains standardized parametric components for beam segments, nodes, supports, and connectors. Each component has multi-dimensional data including geometric information, material properties, mechanical parameters, and cost information. Establish a data standard based on the IFC extended standard, and add a set of steel bridge-specific attributes such as design parameters, material properties, construction technology, monitoring sensors, and maintenance records on the basis of the IFC framework; Deploy an AI engine, which includes a structure selection recommendation model based on a historical project database, an intelligent optimization model for cross sections and nodes using deep neural networks combined with topology optimization algorithms, and a construction feasibility prediction model based on graph neural networks. During the construction phase, data on steel performance, welding quality, and construction monitoring are collected through an IoT sensor network, and the BIM model is automatically updated to form an as-built BIM model. During the operation and maintenance phase, the real-time data stream of the structural health monitoring system is accessed, and the remaining lifespan of the structure is predicted and maintenance decisions are generated through time series prediction algorithms and machine learning models. The performance degradation data, damage evolution patterns, and maintenance effectiveness evaluations during the operation and maintenance phase are fed back into the AI ​​training database, and the AI ​​model is optimized through transfer learning and incremental learning.

2. The BIM+AI-based full life-cycle design method for steel bridges according to claim 1, characterized in that, The historical project database of the structural selection recommendation model contains no less than 10,000 sets of historical bridge project data. The model adopts a reinforcement learning algorithm, takes bridge site environmental parameters, load conditions, and economic indicators as input, and outputs the technical feasibility score, economic score, and comprehensive recommendation index of the bridge type scheme.

3. The BIM+AI-based full life-cycle design method for steel bridges according to claim 1, characterized in that, The deep neural network of the intelligent optimization model for cross sections and nodes includes an input layer, multiple hidden layers, and an output layer. The input parameters include span, load, and material properties. The outputs are the optimized cross section dimensions, stiffening rib arrangement scheme, and node construction details. The cross section optimization response time is less than 30 seconds, and the node optimization response time is less than 2 minutes.

4. The BIM+AI-based full life-cycle design method for steel bridges according to claim 1, characterized in that, The construction feasibility prediction model uses hoisting equipment, components, and temporary supports as nodes in a graph, and their interrelationships as edges. It learns the dependencies in the construction process through a graph neural network to predict the rationality of the hoisting sequence, the feasibility of the welding process, and the safety of the temporary support scheme. The model's predicted F1-score reaches over 0.

85.

5. The BIM+AI-based full life-cycle design method for steel bridges according to claim 1, characterized in that, The IoT sensor network includes strain gauges, displacement gauges, temperature sensors, and inclinometers. The collected data is wirelessly transmitted and uploaded to the cloud platform. The system automatically parses the data and updates the corresponding attributes in the BIM model.

6. The BIM+AI-based full life-cycle design method for steel bridges according to claim 1, characterized in that, The structural health monitoring system includes an acceleration sensor, a displacement sensor, and a corrosion monitoring sensor, which collect vibration data, deformation data, and corrosion rate data, respectively.

7. The BIM+AI-based full life-cycle design method for steel bridges according to claim 1, characterized in that, The cloud-based BIM collaborative work environment adopts a distributed architecture, supports real-time collaborative design by multiple professionals, and the system automatically manages versions, recording modification time, modifier, and modification content.

8. A BIM+AI-based full life-cycle design system for steel bridges, characterized in that, include: In the data layer, a unified data platform is built, and data lake technology is used to achieve unified storage and management of BIM model data, AI model parameters, real-time monitoring data, and design specification library; The engine layer includes an AI computing engine that supports TensorFlow and PyTorch deep learning frameworks, a BIM engine that integrates BIM software through Revit API and Navisworks API, and a rule engine with built-in design specifications. The application layer adopts a SaaS design, including an intelligent design assistant module, a conflict early warning module, a lifespan prediction module, and a maintenance suggestion module; The system adopts a microservice architecture, with each functional module deployed independently and services invoked through an API gateway.

9. A computer-readable medium, characterized in that, The computer-readable medium stores a computer program that, when executed by a processor, implements the BIM+AI-based full life-cycle design method for steel bridges as described in any one of claims 1 to 7.

Images