Green building full life cycle resource sharing and collaborative management method and system

Abstract

The invention discloses a green building full life cycle resource sharing and collaborative management method and system, and relates to the technical field of workflow management, and the method comprises the steps: receiving a collaborative design request, creating a task tree by a task management module, and determining a task progress according to the task tree; the green tool integration module calls basic modeling software based on the task progress to obtain building information model data; the green tool integration module calls performance simulation software and calculation software based on the building information model data to generate multi-type result data; the task management module performs quantitative and qualitative analysis on the multi-type achievement data to obtain an analysis result, and sends the analysis result to a green special topic review module; and the green thematic examination module generates evaluation term data and evaluation version data based on the analysis result, and integrates the evaluation term data and the evaluation version data into an evaluation report for output. The method can effectively shorten the iteration period and reduce the error rate.

Description

Technical Field

[0001] This application relates to the field of workflow management technology, and in particular to a method and system for resource sharing and collaborative management throughout the entire life cycle of green buildings. Background Technology

[0002] In the process of green building lifecycle management, building projects involve multiple stages such as planning, design, construction, and operation, requiring the integration of various professional tools and software systems. These tools generate diverse data types, including geometric model data, material property data, energy consumption analysis data, and environmental performance data, which have complex dependencies. For example, changes in building exterior wall materials can affect thermal performance calculation results, which in turn affect energy consumption simulation data, and ultimately affect the green building rating.

[0003] In traditional project management models, when source data changes, designers need to manually identify which downstream analysis results are affected, and then recalculate and update them one by one. This process is not only time-consuming and labor-intensive, but also prone to omissions, leading to data inconsistencies. With increasingly stringent green building standards and a growing number of project collaborators, the frequency and complexity of data updates are increasing exponentially. Establishing an effective data dependency management mechanism in a multi-model, multi-tool integrated environment to achieve intelligent change detection and automatic result updates has become a key technical bottleneck restricting the efficiency and quality of collaborative green building design.

[0004] Traditional technical solutions primarily employ file exchange or simple data interfaces for tool integration. In this model, each specialized software operates relatively independently, transferring data by importing and exporting intermediate format files, such as exchanging geometric information between modeling and analysis software via IFC format. This approach has significant limitations. First, data dependencies lack explicit expression; the system cannot automatically track which downstream analyses use a given set of basic data. When source data changes, the system cannot automatically identify the scope of impact, relying entirely on human experience for judgment. Second, it lacks the ability to understand the semantics of changes. Traditional solutions treat all data changes as equally important, failing to distinguish the nature and degree of impact of changes. For example, adjusting building orientation and fine-tuning window sill height have significantly different impacts on energy consumption analysis, but the system cannot intelligently determine this, often resorting to a simplistic strategy of full recalculation. These technical problems are particularly prominent in large and complex projects, often leading to long design iteration cycles and high error rates, severely hindering the in-depth application and promotion of green building technologies.

[0005] There is currently no effective solution to the aforementioned technical problems. Summary of the Invention

[0006] This application provides a method and system for resource sharing and collaborative management throughout the entire life cycle of green buildings, which can effectively shorten the iteration cycle and reduce the error rate.

[0007] To achieve the above objectives, the embodiments of this application adopt the following technical solutions: Firstly, a method for resource sharing and collaborative management throughout the entire lifecycle of green buildings is provided, applied to a collaborative management platform. This platform includes a task management module, a green topic review module, a green building knowledge base, and a green tool integration module. The green tool integration module integrates basic modeling software, green design professional software, performance simulation software, and calculation software. The method includes: In response to receiving a collaborative design request sent by a user through the front-end interface, the task management module creates a task tree and determines the task progress based on the task tree; The green tool integration module calls the basic modeling software based on the task progress to obtain building information model data; The green tool integration module, based on the building information model data, calls the performance simulation software and the calculation software to generate multiple types of output data; The task management module performs quantitative and qualitative analysis on the multi-type result data, obtains the analysis results, and sends the analysis results to the green topic review module. Based on the analysis results, the green topic review module generates evaluation clause data and evaluation version data, and integrates the evaluation clause data and the evaluation version data into an evaluation report for output.

[0008] In one possible implementation of the first aspect, the multi-type outcome data includes solar radiation analysis data, energy consumption analysis data, and carbon emission calculation data. The green tool integration module, based on the building information model data, calls the performance simulation software and the calculation software to generate the multi-type outcome data, including: The green tool integration module parses the building information model data, extracts building geometric information, material property information and spatial layout information, and obtains current climate data; Based on the building geometry information and the spatial layout information, the solar radiation analysis tool in the performance simulation software is invoked, and solar radiation simulation is performed based on the current climate data to generate the solar radiation analysis data; Based on the building geometry information, the material property information, and the spatial layout information, the energy consumption analysis tool in the performance simulation software is invoked to perform building energy consumption simulation and generate the energy consumption analysis data. Based on the material property information, the carbon emission calculation tool in the calculation software is invoked to calculate the carbon emissions of the building throughout its entire life cycle and generate the carbon emission calculation data.

[0009] In another possible implementation of the first aspect, the green building knowledge base includes a green building case library and a knowledge question-answering module, the knowledge question-answering module integrating an AI question-answering model, and the method further includes: In response to a user’s knowledge query request sent through the front-end interface, the knowledge question-and-answer module extracts the query keywords from the knowledge query request; The knowledge Q&A module retrieves matching case data from the green building case database based on the query keywords. The case data includes architectural design schemes, performance analysis results, and implementation effects. The knowledge question-answering module calls the AI ​​question-answering model, takes the query keywords and the case data as input, and generates knowledge answer content; The knowledge Q&A module returns the knowledge answers and case data to the user through the front-end interface.

[0010] In another possible implementation of the first aspect, the collaborative management platform further includes a data dependency management module, which is used to construct and maintain a data dependency graph, and the method further includes: When any data from the building information model data or the multi-type output data is created or uploaded to the collaborative management platform, the data dependency management module registers the data as a data node. When a user invokes a tool through the green tool integration module to operate on at least one of the data nodes and generate new result data, the data dependency management module registers the new result data as a result node; The data dependency management module establishes a directed dependency edge between the input data node and the output result node, whereby the directed dependency edge represents the dependency relationship between the data node and the result node. The data dependency management module constructs the data dependency graph based on all the data nodes, the result nodes, and the directed dependency edges.

[0011] In another possible implementation of the first aspect, the data dependency management module further includes a change identification unit, and the method further includes: Calculate the first hash value of the target data corresponding to the data node; When the target data corresponding to the data node is modified by the user and a new version of the data is uploaded, the change identification unit calculates the second hash value of the new version of the data; The change identification unit compares the second hash value with the first hash value, and if the second hash value is different from the first hash value, the change identification unit determines that the data node has changed and triggers the semantic change analysis process.

[0012] In another possible implementation of the first aspect, the change identification unit includes a model comparison engine, and the semantic change analysis process includes: The model comparison engine parses the new version data and the target data to extract the internal structure information of the data. The model comparison engine compares the internal structure information of the new version data with that of the target data to identify the specific changes. The model comparison engine determines the change type based on the specific changes. The change type includes geometric changes, attribute changes, and non-functional changes. Geometric changes include changes in component position, component size, and component thickness. Attribute changes include changes in material type and thermal parameters. Non-functional changes include changes in remarks information and layer attribute changes. The model comparison engine adds semantic tags to the data nodes according to the change type. The semantic tags include structural change tags, building envelope change tags, HVAC parameter change tags, and non-critical information change tags.

[0013] In another possible implementation of the first aspect, the data dependency management module further includes a task triggering unit, and the method further includes: The task triggering unit, starting from the data node that has changed in the data dependency graph, traverses downstream along the directed dependency edge and marks all downstream result nodes as in an invalid state. The task triggering unit acquires the semantic tags of the data nodes; The task triggering unit determines the analysis task type affected by the change type based on the semantic tag; The task triggering unit only triggers the recalculation task for the result node corresponding to the analysis task type, while other result nodes remain in their original state.

[0014] In another possible implementation of the first aspect, the task triggering unit determines the analysis task type affected by the change type based on the semantic tag, including: When the semantic tag is the non-critical information change tag, the task triggering unit determines that the change does not affect any performance analysis and does not trigger any recalculation task; When the semantic tag is the building envelope change tag, the task triggering unit determines that the analysis task type includes solar radiation analysis task, daylighting analysis task and energy consumption simulation task, and triggers the corresponding recalculation task; When the semantic tag is the structural change tag, the task triggering unit determines that the analysis task type includes structural calculation task, load calculation task and energy consumption calculation task, and triggers the corresponding recalculation task.

[0015] In another possible implementation of the first aspect, after the task triggering unit triggers a recalculation of the task, the method further includes: The task triggering unit adds the recalculation task to the task queue; The collaborative management platform invokes cloud computing resources and automatically executes the recalculation task according to the order of the task queue; After the recalculation task is completed, the data dependency management module generates a new result node and replaces the result node in the failed state with the new result node; The data dependency management module updates the data dependency graph and establishes the directed dependency edges between the data nodes and the new result nodes; The task management module pushes the execution result of the recalculated task to the front-end interface.

[0016] Secondly, this application provides a green building full life cycle resource sharing and collaborative management system, which includes a cloud server and a client. The client is used to send a request to the cloud server, and the cloud server is used to execute the operation steps of the above-mentioned green building full life cycle resource sharing and collaborative management method according to the request.

[0017] Through the above technical solution, the task management module establishes a hierarchical dependency relationship between data sources and derived results by creating a task tree. When the green tool integration module calls the basic modeling software to obtain building information model (BIM) data based on the task progress, it can clearly record which performance simulation and calculation software calls this data as the source, thereby generating multi-type result data with clear traceability. This explicit dependency expression allows for automatic tracing down the task tree when BIM data changes, accurately identifying the scope of affected downstream analysis results without manual judgment. The quantitative and qualitative analysis of multi-type result data by the task management module achieves a deep understanding of the semantics of changes, enabling the differentiation of the nature and degree of impact of different changes. For example, it can identify that adjustments to the thermal parameters of exterior wall materials are high-impact changes that require triggering energy consumption recalculation, while changes to interior decoration materials have a smaller impact on energy consumption and can skip related calculations, thus achieving a differentiated update strategy rather than a simple full recalculation. By transmitting the analysis results to the green building review module, a complete traceability chain was further established from the underlying data to the evaluation clause data and evaluation version data. When changes in the source data lead to updates in the analysis results, the evaluation report can automatically update the corresponding evaluation clauses and version information, ensuring data consistency across the entire chain from the basic model to the final evaluation conclusion. This task tree-based dependency management and semantic analysis-based intelligent update mechanism transforms the traditional process of change identification, impact analysis, and result updating, which required manual processing, into automated execution by the system. This not only significantly shortens the design iteration cycle and reduces the error rate caused by missed updates, but more importantly, it avoids unnecessary recalculation by minimizing the update scope, greatly improving the efficiency of multi-disciplinary collaborative design and providing a robust data consistency guarantee mechanism for the entire life cycle management of green buildings.

[0018] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0019] Figure 1 A flowchart illustrating a method for resource sharing and collaborative management throughout the entire lifecycle of a green building, provided as an embodiment of this application; Figure 2 An architecture diagram of a collaborative management platform provided in this application embodiment; Figure 3 This is a schematic diagram of a green building life-cycle resource sharing and collaborative management system provided in an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0021] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0022] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0023] Figure 1 This illustration schematically depicts a flowchart of a green building lifecycle resource sharing and collaborative management method according to an embodiment of this application. Figure 1 As shown in the embodiment of this application, a method for resource sharing and collaborative management throughout the entire life cycle of green buildings is provided and applied to a collaborative management platform. The collaborative management platform includes a task management module, a green topic review module, a green building knowledge base, and a green tool integration module. The green tool integration module integrates basic modeling software, green design professional software, performance simulation software, and calculation software. The method may include the following steps.

[0024] S110. In response to receiving a collaborative design request sent by the user through the front-end interface, the task management module creates a task tree and determines the task progress based on the task tree. S120, the green tool integration module calls the basic modeling software based on the task progress to obtain building information model data; S130, the green tool integration module, is based on building information model data and calls performance simulation software and calculation software to generate multiple types of output data; S140 The task management module performs quantitative and qualitative analysis on various types of output data, obtains analysis results, and sends the analysis results to the green topic review module; Based on the analysis results, the S150 Green Thematic Review Module generates evaluation clause data and evaluation version data, and integrates the evaluation clause data and evaluation version data into an evaluation report for output.

[0025] The building information model (BIM) data includes building floor plan data, building design specification data, fire protection design data, and green building design data. The various types of deliverables include simulation analysis results, calculation results, drawing model results, on-site testing results, model experiment results, and video and image results. Among these, simulation analysis results include solar radiation analysis data and energy consumption analysis data, while calculation results include carbon emission calculation data and sound insulation calculation data.

[0026] The collaborative management platform in this embodiment is an integrated cloud management system. Figure 2 An architecture diagram of a collaborative management platform provided in an embodiment of this application is shown. (Refer to...) Figure 2 The collaborative management platform comprises four core functional modules: a task management module, a green thematic review module, a green building knowledge base, and a green tool integration module. The task management module handles steps S110 and S140, the green tool integration module handles steps S120 and S130, and the green thematic review module handles step S150. The green tool integration module integrates basic modeling software, green design professional software, performance simulation software, and calculation software. These software programs interact with the platform through a unified API interface. Basic modeling software may include Building Information Modeling (BIM) modeling tools such as Revit and ArchiCAD; performance simulation software may include energy consumption simulation tools such as Ecotect and EnergyPlus, and solar radiation analysis tools such as Radiance; and calculation software may include professional calculation programs such as carbon emission calculation tools and sound insulation calculation tools.

[0027] In practice, when a user sends a collaborative design request through the front-end interface, the request is first received by the gateway service of the collaborative management platform for authentication and permission verification. The front-end interface can be a web-based browser or a mobile app. Users fill in basic project information, including project name, project type (e.g., residential, public, industrial), building area, construction location, climate zone, and green building target level (e.g., one-star, two-star, three-star). These parameters constitute the core content of the collaborative design request. Upon receiving the request, the task management module selects the corresponding task template from a pre-set task template library based on the project type and green building target level. The task template defines the standard workflow required to complete this type of green building project, including task nodes at each stage, dependencies between tasks, the software tools required for each task, and the data types of inputs and outputs.

[0028] Based on the selected task template, the task management module begins creating a task tree. A task tree is a hierarchical data structure that uses a tree topology to express the dependencies and execution order between tasks. In the task tree, the root node represents the entire green building project, and the first-level child nodes represent the main phases of the project, such as the "scheme design phase." Each phase node contains more granular task nodes; for example, the "scheme design phase" might include sub-tasks such as "building scheme modeling," "solar radiation analysis," "natural ventilation analysis," and "preliminary energy consumption estimation." Each task node contains detailed attribute information, including task ID, task name, task type, responsible person, planned start time, planned completion time, a list of prerequisite tasks, a list of subsequent tasks, required software tools, input data requirements, and output deliverable type. Tasks are connected by directed edges to represent dependencies; for example, the "solar radiation analysis" task depends on the completion of the "building scheme modeling" task because solar radiation analysis requires a building geometric model as input.

[0029] During the task tree creation process, the task management module automatically configures necessary analysis tasks according to the requirements of green building standards. For example, for residential projects applying for a two-star green building rating, the standard requires analysis of building daylighting, indoor natural lighting, building energy consumption, renewable energy utilization, and building carbon emissions. The task management module automatically adds these mandatory analysis tasks to the task tree and sets a reasonable execution order. For optional optimization items, such as natural ventilation CFD simulation, outdoor wind environment analysis, and rainwater harvesting system design, the task management module decides whether to add them to the task tree based on the user's selection in the collaborative design request.

[0030] After the task tree is created, the task management module initializes the status of all task nodes. Task status can include "Not Started," "In Progress," "Completed," "Paused," and "Invalid." Initially, only the root task without prerequisite tasks (such as "Architectural Scheme Modeling") is marked as "Executable," while other tasks are marked as "Not Started." The task management module continuously monitors the execution status of each task in the task tree. When a task is completed, it automatically checks whether all prerequisites for its subsequent tasks are met. If so, the status of the subsequent tasks is updated to "Executable."

[0031] Determining task progress based on the task tree is a dynamic calculation process. The task management module uses the critical path method to analyze the task tree, identifying the project's critical path, which is the longest task chain that determines the project's total duration. By calculating the earliest start time, earliest finish time, latest start time, latest finish time, and time fluctuation for each task, it's possible to determine which tasks are critical (tasks with zero time fluctuation) and which tasks have a buffer period. The formula for calculating task progress is: Overall project schedule = (Total planned work hours of completed tasks) / (Total planned work hours of all tasks) × 100% For each stage or subtask branch, local progress can also be calculated. The task management module displays task progress on the front-end interface in various forms such as Gantt charts, progress bars, and milestone views, enabling project managers to intuitively understand the project status. When a critical task is at risk of delay, the task management module will automatically issue an early warning notification to remind relevant personnel to take action. In addition, the task management module also records historical data of task execution, including actual start time, actual completion time, and resource consumption, providing data support for subsequent project summaries and experience accumulation. Through this task tree-based project management approach, the complex green building collaborative design process can be decomposed into manageable task units, clarifying the logical relationships between various tasks and achieving precise control over project progress.

[0032] After receiving the task progress information from the task management module, the green tool integration module identifies tasks that are currently in an "executable" state and require the invocation of basic modeling software. The invocation of basic modeling software is the starting point of the entire collaborative design process, as all subsequent performance analysis and calculations rely on an accurate building information model.

[0033] Before invoking the basic modeling software, the green tool integration module first checks whether available Building Information Modeling (BIM) data already exists for the task. If it is the first modeling attempt for the project, the new project function of the basic modeling software needs to be activated; if it is modifying or refining an existing model, historical version model data needs to be loaded.

[0034] The basic modeling software is accessed via an API interface. The Green Tools Integration Module maintains a tool adapter library, developing corresponding adapter plugins for different basic modeling software (such as Revit, ArchiCAD, SketchUp, etc.). These adapter plugins implement a unified interface specification, enabling the platform to call different modeling software in a standardized way, shielding the differences between the underlying software. When Revit software needs to be called, the Green Tools Integration Module sends a call command through the Revit API, which includes information such as project parameters, template file path, and working directory. After receiving the command, Revit software starts in the background and loads the specified project template.

[0035] The process of acquiring Building Information Modeling (BIM) data is as follows: First, the building floor plan data is acquired, including the floor plan layout of each floor, room division, wall locations, door and window arrangements, stair and elevator locations, and other geometric information. This information exists in the BIM model as parametric components, each containing rich attribute information. The green tool integration module traverses all components in the BIM model, extracting their geometric parameters (such as location coordinates, dimensions, orientation, etc.) and non-geometric parameters (such as material type, thermal performance parameters, construction methods, etc.). For wall components, the information to be extracted includes wall type (load-bearing wall, non-load-bearing wall, exterior wall, interior wall), wall thickness, wall material layer composition, thickness of each material layer, and thermal parameters (thermal conductivity, specific heat capacity, density, etc.). For door and window components, parameters such as window-to-wall ratio, window type, glass type, heat transfer coefficient, shading coefficient, and airtightness level need to be extracted.

[0036] Architectural design specification data is primarily extracted from the project information and annotations in the BIM model. The architectural design specification includes project overview (building name, construction location, building area, building height, number of floors, etc.), design basis (standards and regulations adopted, design principles, etc.), building functional layout description, main technical and economic indicators, fire protection design specifications, and energy-saving design specifications. This information may be stored in the BIM model as text annotations, project parameters, and shared parameters. The green tool integration module calls the data reading API of the basic modeling software to extract and convert relevant information from the BIM model into structured architectural design specification data according to predefined data mapping rules.

[0037] Acquiring data for fire protection design involves information on building fire compartments, safe evacuation, and fire protection facilities. In the BIM model, fire compartments are typically represented by spatial divisions and components such as firewalls and fire doors. The green tool integration module needs to identify fire compartment components in the model, extract parameters such as the area, evacuation distance, and evacuation width of fire compartments, and check whether they meet the requirements of relevant codes. For safe evacuation design, it is necessary to extract information such as the number, location, and width of evacuation staircases, the width and opening direction of evacuation doors, and the width of evacuation corridors, and calculate whether the evacuation distance meets the code requirements. Regarding fire protection facilities, it is necessary to extract the layout information of facilities such as fire hydrants, automatic sprinkler systems, and automatic fire alarm systems.

[0038] The green building design data includes design measures and parameters related to building energy efficiency, renewable energy utilization, water resource utilization, indoor environmental quality, and materials and resources. In terms of building energy efficiency, key parameters such as building shape coefficient, window-to-wall ratio, heat transfer coefficients of exterior walls and roofs, heat transfer coefficients of exterior windows, and shading coefficients need to be extracted. Extracting these parameters requires geometric calculations and attribute queries on the BIM model. For example, the formula for calculating the building shape coefficient is: ; in, Here, F represents the building shape coefficient, and F is the external surface area of ​​the building in contact with the outdoor atmosphere (including exterior walls, roof, and floor). This refers to the building's volume. The green tool integration module automatically calculates the model's external surface area and volume, obtains the shape factor, and compares it with the standard limits.

[0039] Calculating the window-to-wall ratio requires separately calculating the area of ​​the exterior walls facing each direction and the area of ​​the exterior windows facing that direction. The calculation formula is as follows: ; Where C is the window-to-wall ratio, A w Let A be the area of ​​an outward-facing window (including the window frame) facing a certain direction. wall This refers to the total area of ​​the wall facing that direction (including windows). The Green Tools integration module will calculate the window-to-wall ratio for the four main directions (east, south, west, and north) and check whether it meets the requirements.

[0040] Regarding renewable energy utilization, design parameters for renewable energy facilities such as solar photovoltaic systems, solar water heating systems, and ground source heat pump systems need to be extracted. If the BIM model includes equipment models of these systems, the green tool integration module can directly extract information such as equipment type, specifications, installation location, and design capacity. Regarding water resource utilization, design information for rainwater harvesting systems, greywater reuse systems, and water-saving appliances needs to be extracted, including rainwater harvesting area, storage tank volume, greywater treatment capacity, and water efficiency ratings for various water-using appliances.

[0041] In this embodiment, the green tool integration module also performs format conversion on the extracted data, transforming it into a standard data format used internally by the platform to facilitate processing by subsequent modules. Standard data formats typically use JSON or XML, containing clear data structures and metadata descriptions. This standardized data acquisition process ensures the accuracy, completeness, and consistency of the Building Information Modeling (BIM) data, laying a reliable data foundation for subsequent performance analysis and calculations.

[0042] After acquiring building information model (BIM) data, the green tool integration module sequentially calls performance-based simulation and calculation software according to the analysis tasks defined in the task tree, generating various types of output data. These output data cover all aspects required for green building evaluation, including simulation analysis results, calculation results, drawing model results, on-site testing results, model experiment results, and video and image results.

[0043] Taking the generation of solar radiation analysis data as an example, the green tool integration module first extracts building geometry and spatial layout information from the building information model (BIM) data, including the building's 3D geometric model, the shading from surrounding buildings, and site topography. Then, the module acquires the geographical location information (latitude, longitude, and altitude) and current climate data, including solar radiation, temperature, and cloud cover. This climate data can be obtained from meteorological databases.

[0044] After the input data is prepared, the green tool integration module calls the solar radiation analysis tool in the performance simulation software. The call is implemented via command-line interface or API interface. The module constructs a call command containing all necessary parameters, such as the model file path, analysis time period, analysis time step, and analysis area (e.g., window location in a residence or outdoor activity area). Based on solar trajectory calculation and ray tracing algorithms, the solar radiation analysis tool simulates the propagation of sunlight on building surfaces and indoor spaces, considering the shading effects of surrounding buildings and terrain, and calculates the solar radiation conditions at a specified analysis point or surface at different times.

[0045] The output of solar radiation analysis includes data in various formats. For residential buildings, the focus is primarily on the sunshine duration of the main rooms in each unit type, i.e., the cumulative duration of direct sunlight received by the center point of the window within a specified period. According to relevant requirements, each residence should have at least one living space that receives sunlight during winter, and the sunshine standard should not be less than 2 hours of sunshine on the coldest day of the year. The solar radiation analysis tool outputs the sunshine duration value for each analysis point and generates a sunshine duration distribution cloud map, using different colors to represent the sunshine conditions of different areas. For outdoor activity areas, such as kindergarten outdoor activity areas and activity areas for the elderly, it is necessary to analyze the site's sunshine coverage rate, i.e., the proportion of the site's area that meets the required sunshine duration within a specified date and time period.

[0046] The generation of energy consumption analysis data requires the establishment of a building energy consumption model. The green tool integration module calls energy consumption analysis tools in performance simulation software, such as EnergyPlus, DeST, or DOE-2. The input data required for energy consumption simulation includes building geometry information, thermal parameters of the building envelope, internal heat gain (personnel, lighting, equipment), HVAC system parameters, operating schedules, and meteorological data. The green tool integration module extracts building geometry information from the building information model data, including the building's three-dimensional shape, the area and volume of each room, the area of ​​exterior walls and roof, and the area and orientation of exterior windows. Thermal parameters of the building envelope include heat transfer coefficients, thermal inertia indices, and solar radiation absorption coefficients for exterior walls, roof, windows, and ground. These parameters can be extracted from the material properties of components in the BIM model or obtained by querying thermal parameter databases based on construction methods.

[0047] The determination of heat gain within a building needs to be based on its function and usage. For occupant heat gain, the occupant density (persons / m²) and average heat dissipation per person (W / person) need to be set for each room. For lighting heat gain, the lighting power density (W / m²) needs to be set. For equipment heat gain, the equipment power density needs to be set according to the building type, such as computers and printers in office buildings, and commercial equipment in commercial buildings. An operating schedule describes the operating patterns of occupants, lighting, equipment, and HVAC systems, typically setting a weekly operating pattern on an hourly basis, distinguishing between weekdays and weekends, as well as seasonal differences.

[0048] The settings of HVAC system parameters have a significant impact on energy consumption simulation results. Parameters that need to be set include: system type (e.g., split air conditioner, multi-split system, fan coil unit with fresh air supply, all-air system, etc.), heat source type (e.g., electric compressor chiller, gas boiler, ground source heat pump, etc.), system efficiency parameters (e.g., chiller COP, boiler efficiency, distribution system efficiency, etc.), indoor design temperature (summer and winter), and fresh air volume. These parameters can be obtained from HVAC design documents or set according to typical values ​​based on building type and energy efficiency standards.

[0049] Energy simulation software calculates the heat load and energy consumption of each area of ​​a building hourly based on heat transfer principles and heat balance equations. The calculation process considers various factors, including heat transfer from the building envelope, solar radiation heat gain, internal heat gain, natural ventilation, and the operation of the HVAC system. The output of the energy simulation includes hourly heating and cooling loads throughout the year, monthly and annual consumption of various energy sources (electricity, gas, heat, etc.), energy consumption per unit area, and the proportion of each energy type (such as air conditioning energy consumption, lighting energy consumption, equipment energy consumption, and domestic hot water energy consumption). This data is output in the form of numerical tables and charts for easy analysis and comparison.

[0050] Taking carbon emission calculation data as an example, the green tool integration module calls the carbon emission calculation tool in the calculation software to calculate carbon emissions based on each stage of the building's entire life cycle. The building's entire life cycle includes the building material production stage, the construction stage, the operation stage, and the demolition stage. Carbon emissions in the building material production stage mainly come from the production process of building materials. It is necessary to statistically analyze the usage of various building materials (such as concrete, steel, glass, insulation materials, etc.) and then multiply them by the corresponding material's carbon emission factor (CO2 emissions per unit mass or volume of material produced). Material usage can be extracted from the bill of quantities in the BIM model, and carbon emission factors can be obtained by querying relevant databases.

[0051] Carbon emissions during the construction phase primarily originate from the energy consumption of construction machinery and the material transportation process. The working hours and energy consumption of construction machinery can be estimated based on the construction organization design, and then multiplied by the corresponding energy carbon emission factor. Carbon emissions from material transportation can be estimated based on material weight, transportation distance, and transportation mode (e.g., road or rail). Carbon emissions during the operation phase constitute the majority of a building's total life-cycle carbon emissions, stemming from energy consumption during building operation. These can be calculated by multiplying various energy consumption figures obtained from energy consumption analysis by the corresponding energy carbon emission factor. Carbon emissions during the demolition phase are relatively small, mainly originating from the energy consumption of demolition machinery and the waste transportation and disposal process.

[0052] The general formula for calculating carbon emissions can be expressed as: ; Where E represents the total carbon emissions throughout the building's life cycle, E1 represents the carbon emissions during the building material production stage, E2 represents the carbon emissions during the construction stage, E3 represents the carbon emissions during the operation stage, and E4 represents the carbon emissions during the demolition stage.

[0053] The calculation results will output the carbon emissions and their proportion at each stage, as well as indicators such as carbon emissions per unit area. These data are important bases for evaluating the carbon emission level of buildings and formulating carbon reduction strategies, and are also scoring items in the "energy conservation and energy utilization" and "environmental protection" categories in green building evaluation.

[0054] Generating sound insulation calculation data requires calculating the sound insulation performance of various sound insulation components based on the building's sound insulation design scheme. Components involved in sound insulation calculations include exterior walls, exterior windows, partition walls, partition floors, and doors. For each component, its weighted sound insulation or weighted normalized sound pressure level difference needs to be looked up or calculated based on its construction methods (such as wall materials, thickness, presence of air gaps, and presence of sound-absorbing materials). For simple, single-layer homogeneous components, the mass law can be used to estimate the sound insulation: ; Where R is the sound insulation (dB), m is the surface density of the component (kg / m²), and f is the sound wave frequency (Hz). For complex multi-layered composite components, it is necessary to use the corresponding calculation model or consult experimental test data.

[0055] The output of sound insulation calculations includes sound insulation performance parameters for various components, as well as comparisons with standard requirements. Drawing and model outputs include various design drawings and 3D models. The green tool integration module can utilize the drawing generation function of basic modeling software to automatically generate construction drawings such as building floor plans, elevations, sections, and detailed drawings, as well as various professional drawings such as energy-saving design drawings, solar radiation analysis diagrams, and acoustic environment analysis diagrams. These drawings can be exported in standard formats such as DWG and PDF for easy archiving and exchange. 3D models can be exported in common 3D formats such as IFC, OBJ, and FBX for easy exchange and display between different software.

[0056] On-site testing, model experiment, and video recording results are typically generated during the construction and operation phases of a building, but pre-assessment can also be performed during the design phase. For example, design parameters can be used to predict the indoor environmental quality (such as indoor temperature, humidity, CO2 concentration, and noise level) after the building is completed, providing a reference benchmark for subsequent on-site testing. Model experiment results can include the results of physical model experiments such as wind tunnel experiments and solar radiation model experiments. These experiments are usually conducted in important large-scale projects to verify the accuracy of computer simulations. Video recording results can include building walkthrough animations, solar radiation animations, and construction process simulation animations, used to visually demonstrate the design effects and performance characteristics of the building.

[0057] By invoking various performance simulation and calculation software, the green tool integration module can generate multi-type outcome data covering all aspects of green building, providing comprehensive data support for subsequent quantitative and qualitative analysis and green building evaluation. This integrated tool invocation method greatly improves design efficiency, avoids the tedious process of manually transferring data between different software, reduces errors in data conversion, and ensures data consistency and accuracy.

[0058] After receiving various types of output data generated by the green tools integration module, the task management module performs quantitative and qualitative analysis on this data to assess the green performance level of buildings, identify strengths and weaknesses in the design, and provide a basis for design optimization and green building evaluation. Quantitative and qualitative analysis is a comprehensive evaluation process that includes both statistical analysis and index calculation of numerical data, as well as expert judgment and rating of non-numerical information.

[0059] Quantitative analysis primarily targets performance indicators that can be expressed numerically. For sunlight analysis data, the task management module will calculate the number of hours of sunlight in the main rooms of each unit type and determine the percentage of units that meet the sunlight standards. For example, if a project has 100 residential units, and 95 of them have bedrooms that receive at least 2 hours of sunlight on the coldest day of the year, then the sunlight compliance rate is 95%. The task management module will then compare this compliance rate with the requirements of green building evaluation standards.

[0060] For energy consumption analysis data, the task management module calculates the energy consumption per unit area of ​​the building and compares it with the energy consumption of a reference or benchmark building to obtain the energy saving rate. The formula for calculating the energy saving rate is: ; in, For energy saving rate, For reference, energy consumption per unit area of ​​buildings, Energy consumption per unit area of ​​the designed building. A reference building refers to a building designed according to the prescribed indicators of energy-saving design standards (such as limits on the thermal performance of the building envelope and efficiency limits for HVAC systems), representing the basic level of meeting energy-saving standards. If the designed building adopts superior energy-saving measures, its energy consumption will be lower than that of the reference building, thus achieving a positive energy saving rate.

[0061] For residential buildings, a base score is awarded for achieving an energy efficiency rate of 10%. Additional points are awarded for every 5% increase in the energy efficiency rate, up to a maximum of 40% or higher. For public buildings, the energy efficiency requirements and scoring rules are similar. The task management module will automatically query the scoring rules based on the calculated energy efficiency rate to determine the score for that indicator.

[0062] For carbon emission calculations, the task management module calculates the carbon emissions per unit area of ​​the building, including emissions during the building material production, construction, operation, and demolition phases. Special attention is paid to carbon emissions during the operation phase, as this constitutes the majority of the building's total lifecycle carbon emissions. The task management module compares the calculated carbon emissions with industry averages to assess the building's carbon emission level. If the building utilizes renewable energy sources (such as solar photovoltaic power generation, solar water heating systems, and ground source heat pumps), it can offset some of the carbon emissions during the operation phase, thereby reducing total carbon emissions. The task management module calculates the carbon reduction from renewable energy sources and deducts it from the total carbon emissions.

[0063] For sound insulation calculation data, the task management module checks whether the sound insulation performance of various sound insulation components meets the specifications and calculates the degree to which it exceeds the specifications. For example, if the weighted sound insulation of the partition wall reaches 50dB, exceeding the specification requirement of 45dB by 5dB, the sound insulation performance can be considered good. The task management module determines the score in the green building evaluation based on the degree to which the sound insulation performance exceeds the standard.

[0064] Qualitative analysis primarily targets design measures and technical strategies that are difficult to quantify. For example, does the building employ natural ventilation, a rainwater harvesting system, green building materials, or site ecological design? The effectiveness of these measures is difficult to measure with a single numerical indicator and requires evaluation through expert assessment. The task management module will rate these qualitative indicators according to preset evaluation rules, such as "Excellent," "Good," "Average," and "Poor." Evaluation rules are typically based on the requirements of green building evaluation standards, combined with the experience of industry experts.

[0065] During the quantitative and qualitative analysis, the task management module also performs cross-validation and consistency checks on the data. For example, the energy consumption of the air conditioning system obtained from the energy consumption analysis should match the cold and heat source capacity designed by the HVAC professionals. If the difference is too large, it may indicate that the input parameters are incorrect or the design scheme is unreasonable, requiring further verification. The results of the sunlight analysis should be consistent with the building layout and window design. If some north-facing rooms also show long hours of sunlight, it may indicate that the analysis model is incorrect. Through this cross-validation, the reliability of the analysis results can be improved.

[0066] The task management module also comprehensively evaluates various types of output data to identify strengths and weaknesses in the design. For example, if energy consumption analysis shows high air conditioning energy consumption in the building, the task management module will further analyze the reasons, which may include excessively high heat transfer coefficients of exterior windows, excessively high window-to-wall ratios, insufficient shading measures, or low efficiency of the air conditioning system. By identifying specific causes, clear directions can be provided for design optimization. If carbon emission calculations show high carbon emissions during the building material production stage, the task management module will recommend the use of low-carbon building materials, such as concrete with high fly ash or slag content, the use of recycled aggregates, and a reduction in steel usage.

[0067] The analysis results will be stored in a structured data format, including numerical values ​​for each indicator, compliance status, scores, existing problems, and optimization suggestions. The task management module will generate a detailed analysis report, presenting the results in a combination of charts and text. The report will include comparative charts of various performance indicators (such as bar charts comparing the energy consumption of the designed building and reference buildings), radar charts (showing the building's performance in various green performance dimensions), and trend charts (such as performance change trends of different design schemes), making the analysis results more intuitive and easy to understand.

[0068] After the analysis results are generated, the task management module will send them to the green building review module as input data for green building evaluation. Through this data-driven quantitative and qualitative analysis, green building design can shift from experience-driven to data-driven, improving the scientific rigor and rationality of the design and ensuring that the building achieves the expected green performance goals.

[0069] After receiving the analysis results from the task management module, the green building review module generates evaluation clause data and evaluation version data based on these results, and integrates them into an evaluation report for output. The green building review module is the core module in the collaborative management platform responsible for green building evaluation and certification management. Its function is to systematically evaluate the green performance of building projects according to green building evaluation standards and generate evaluation reports that meet certification requirements.

[0070] Evaluation clauses refer to specific evaluation items in green building evaluation standards, with each item corresponding to one or more technical indicators or design measures. The green building review module first loads the applicable green building evaluation standards. The selection of standards depends on the building type, project location, and the certification system being applied for. The green building review module maintains a standard library containing electronic versions of various green building evaluation standards, storing all evaluation clauses, evaluation requirements, scoring rules, and other information in the standards in structured data format.

[0071] After the standards are loaded, the green project review module will examine each project to ensure it meets the requirements of each evaluation clause. The review process is highly automated; the module extracts relevant performance indicator data from the analysis results and compares it with the requirements of the evaluation clauses. For example, for the "Building Energy Conservation" clause within the "Resource Conservation" category, the module will extract the energy-saving rate data obtained from energy consumption analysis and determine the score for that clause based on the energy-saving rate. If the energy-saving rate reaches 10%, a base score (e.g., 5 points) is awarded; if the energy-saving rate reaches 20%, a higher score (e.g., 8 points) is awarded; and if the energy-saving rate reaches 30% or higher, a full score (e.g., 10 points) is awarded. The scoring rules are clearly defined in the standards, and the module will strictly adhere to these rules when scoring.

[0072] For qualitative evaluation clauses, such as "whether natural ventilation design is adopted" and "whether a rainwater harvesting system is installed," the green project review module will extract relevant design measures information from the analysis results to determine whether the clause requirements are met. If met, the clause receives points; otherwise, it receives no points. For some clauses, supporting documentation may be required, such as product testing reports, construction plans, and operation and management systems. The green project review module will generate a list of required supporting documents and prompt the user to upload them.

[0073] During the review process, the green topic review module also checks the correlation between clauses. Some evaluation clauses have logical relationships; for example, the score for the "use of renewable energy" clause may depend on the calculation result of "renewable energy utilization rate." The module ensures that the evaluation results of these related clauses are consistent, avoiding logical contradictions. If inconsistencies are found, a warning message will be generated, prompting the user to verify.

[0074] The generated evaluation clause data is a structured dataset containing information such as each clause's number, name, evaluation requirements, actual project conditions, whether the requirements are met, score, and supporting documentation. This data is presented in tabular format for easy viewing and editing. Clauses that do not meet the requirements are highlighted, and improvement suggestions are provided to help the design team optimize the process.

[0075] The generation of evaluation version data is to support iterative evaluation and version management of projects. During the design and construction of green building projects, the design scheme undergoes multiple adjustments and optimizations, requiring a new green performance evaluation after each adjustment. To track the project's evolution and compare performance differences between different versions, the green project review module creates a version record for each evaluation. The version record includes information such as evaluation time, evaluator, design scheme version number, input data snapshot, evaluation results, and score summary.

[0076] After generating the evaluation clause data and evaluation version data, the green project review module will integrate the two into an evaluation report. The evaluation report is a comprehensive document that includes basic project information, an overview of the design scheme, green performance analysis results, a review of each evaluation clause, a summary of scores, a comprehensive evaluation conclusion, and improvement suggestions.

[0077] The evaluation report is generated using a template-based approach. The green thematic review module provides multiple pre-set report templates for different building types and certification systems. The templates define the report's chapter structure, format requirements, and necessary chart types. The module automatically populates the templates with evaluation clause data and evaluation version data, generating a preliminary report document. For sections requiring manual writing, such as design overviews and descriptions of distinctive technical measures, the module provides a text editor for user input and editing.

[0078] This embodiment integrates a task management module, a green tool integration module, a green topic review module, and a green building knowledge base on a collaborative management platform, achieving data sharing and collaborative management throughout the entire lifecycle of green buildings. This effectively solves the problems of data silos, fragmented tools, and disjointed processes inherent in traditional models. First, the task tree-based project management approach establishes clear task dependencies and data flow paths, making the entire process—from building information model data to multi-type output data, and then to analysis results and evaluation reports—traceable and manageable, ensuring data consistency and integrity. Second, by automatically calling basic modeling software, performance simulation software, and calculation software, seamless integration of multiple professional tools is achieved, avoiding the tediousness and errors of manual data conversion and significantly improving design efficiency. Third, the quantitative and qualitative analysis mechanism can comprehensively evaluate the green performance of buildings, identify strengths and weaknesses in the design, and provide a scientific basis for design optimization, transforming green building design from experience-driven to data-driven. Furthermore, the automatically generated evaluation clause data and evaluation version data, as well as the integrated evaluation report, not only meet the requirements of green building certification but also record the continuous improvement process of the project through a version management mechanism, providing complete process proof for certification review. Compared to traditional file exchange or simple data interface methods, this approach, through explicit expression of data dependencies and intelligent change management, can automatically identify the scope of impact and trigger corresponding update tasks when source data changes. This avoids omissions in manual judgment and the waste of resources from full recalculation, significantly shortening the design iteration cycle and reducing the error rate. In summary, this technical solution is not only applicable to the collaborative design and evaluation of individual green building projects, but can also be extended to the operation management and performance monitoring of the entire building lifecycle. It provides strong information platform support for the in-depth application and promotion of green building technologies, effectively promoting the green and low-carbon transformation and high-quality development of the construction industry.

[0079] In one embodiment of this example, the multi-type result data includes solar radiation analysis data, energy consumption analysis data, and carbon emission calculation data. The green tool integration module, based on building information model data, calls performance simulation software and calculation software to generate multi-type result data, including the following steps: S210, the green tool integration module parses building information model data, extracts building geometric information, material property information and spatial layout information, and obtains current climate data; S220. Based on the building geometry and spatial layout information, call the solar radiation analysis tool in the performance simulation software, and perform solar radiation simulation based on the current climate data to generate solar radiation analysis data; S230. Based on the building's geometric information, material properties information, and spatial layout information, call the energy consumption analysis tool in the performance simulation software to perform building energy consumption simulation and generate energy consumption analysis data. S240. Based on the material property information, call the carbon emission calculation tool in the calculation software to calculate the carbon emissions of the building throughout its entire life cycle and generate carbon emission calculation data.

[0080] After receiving the building information model data, the green tool integration module first starts the data parsing engine to perform in-depth parsing of the model file.

[0081] During the parsing process, the data parsing engine reads the tree structure of the model file layer by layer, starting from the root node of the project and traversing the entity objects at each level, such as buildings, floors, spaces, and components. For the extraction of architectural geometry information, the parsing engine identifies the geometric representation of each component, including various forms such as boundary-based solid geometry, sweep-based extrusion geometry, and parametric construction solid geometry.

[0082] Specifically, for wall components, parameters such as start and end coordinates, height, and thickness are extracted; for window components, information such as their location points on the wall, width, height, and windowsill height are extracted; and for roof components, geometric features such as outline, slope, and eaves height are extracted. Material property information extraction focuses on the material objects associated with each component. These material objects include physical property parameters, such as thermal performance parameters like thermal conductivity, density, specific heat capacity, and solar radiation absorption coefficient of the exterior wall material, as well as visual attributes like color and texture. Spatial layout information extraction involves topological analysis of the building's functional spaces, including identifying the boundaries, area, volume, orientation, and adjacency relationships of each room.

[0083] While extracting this basic information, the green tool integration module also obtains climate data for the current project location from an external meteorological database. Climate data can be obtained by calling a meteorological data service API. After inputting the project's geographical coordinates (latitude and longitude), the API will return data from the nearest meteorological station.

[0084] After acquiring building geometry, spatial layout, and climate data, the green tool integration module calls the solar radiation analysis tool integrated into the performance simulation software to perform solar radiation simulation calculations. Solar radiation analysis is a crucial step in green building design, directly impacting the building's lighting quality, thermal comfort, and energy consumption. The solar radiation analysis tool first constructs a 3D scene model based on the building's geometry. This model includes not only the building itself but also considers the shading effects of the surrounding environment, such as adjacent buildings, trees, and terrain. After the scene is constructed, the tool calculates the sun's trajectory based on the project's geographical coordinates. The calculation of the sun's position is based on astronomical formulas, primarily including the calculation of the solar declination angle and the solar hour angle. The solar declination angle can be approximated using the Cooper equation. The solar altitude angle and solar azimuth angle are calculated using spherical trigonometry.

[0085] Based on the above calculations, the tool can determine the incident direction of sunlight at any given time. The sunshine simulation uses a ray tracing algorithm to emit a large amount of light from the sun's position onto the building surface. By calculating the intersections of the rays with the building geometry, it determines whether each point on the building surface is directly exposed to sunlight. For each analysis point, the tool calculates the cumulative sunshine hours for the entire year or a specific period (such as the winter solstice or summer solstice). The simulation also considers the impact of climatic factors such as atmospheric transmittance and cloud cover on solar radiation intensity; these parameters are read from climate data files. The results of the sunshine analysis are presented in various forms, including sunshine hour contour maps, shadow animations, and hourly sunshine analysis tables. For example, for residential buildings, it generates sunshine hour data for each living space on the coldest day of the year (the "Great Cold") to determine whether it meets the regulatory requirement of "sunshine hours not less than 2 hours on the coldest day of the year."

[0086] Building energy consumption simulation is used to assess building energy performance. The green tool integration module calls upon the energy consumption analysis tool in performance-based simulation software to perform comprehensive energy consumption calculations based on building geometry, material properties, and spatial layout information. The energy consumption analysis tool employs a dynamic thermal simulation method, treating the building as a complex thermodynamic system that considers the combined effects of various factors such as heat transfer from the building envelope, solar radiation heat gain, indoor heat sources, ventilation, and air conditioning system operation. The simulation calculation is based on the heat balance equation, establishing an energy conservation equation for each thermal zone within the building (typically corresponding to one or more functional spaces) at each time step (usually one hour or less): ,in and They are heating and cooling loads, respectively. For heat transfer to the wall, To transfer heat to the window For ventilation and heat exchange, For indoor heat sources (people, lighting, equipment), It gains heat from solar radiation.

[0087] The heat transfer coefficient method is used to calculate the heat transfer of the wall: Where U is the heat transfer coefficient of the wall, calculated from the thermal conductivity and thickness of each layer of material in the material properties information; A is the wall area; and Tout and Tin are the outdoor and indoor temperatures, respectively. Window heat transfer calculations need to consider both conductive heat transfer and solar radiation transmission. The solar heat gain coefficient of the window is a key parameter, determined from the glass type in the material properties information. Ventilation heat exchange is calculated based on the number of air changes and the indoor-outdoor temperature difference; the number of air changes is derived from natural ventilation simulation. Indoor heat sources are set based on the functional type in the spatial layout information. For example, the personnel density in office spaces is typically taken as 0.1 people / m², the heat dissipation per person is taken as 90W, the lighting power density is taken as 9W / m², and the equipment power density is taken as 15W / m². The energy consumption analysis tool calculates the cooling and heating load for 8760 hours throughout the year hourly, and then calculates the actual energy consumption based on the type and performance parameters of the air conditioning system (such as chiller COP, boiler efficiency, etc.).

[0088] For example, if a split-type air conditioning system is used, and the seasonal energy efficiency ratio (SEER) is 3.2, then the cooling power consumption at a certain moment is the cooling load at that moment divided by 3.2. The simulation results include total annual energy consumption, energy consumption by component (heating, cooling, lighting, equipment, etc.), monthly energy consumption curves, and hourly load curves. This energy consumption analysis data is an important basis for determining whether a building meets energy-saving standards and calculating energy-saving rates. It also provides data support for the capacity design and operational strategy optimization of HVAC systems.

[0089] Carbon emission calculations are used for the full life-cycle assessment of green buildings. The green tool integration module calls the carbon emission calculation tool in the calculation software to calculate the carbon emissions throughout the building's life cycle based on material property information. Carbon emissions throughout the building's life cycle include emissions from the building material production stage, construction stage, operation and use stage, and demolition and recycling stage. The carbon emission calculation tool first extracts the usage of each building material from the material property information, including concrete, steel, brick, glass, insulation materials, and decorative materials. The statistics of material usage are based on precise calculations of building geometry information; for example, the concrete usage is equal to the sum of the volumes of all concrete components (beams, columns, slabs, walls, etc.) multiplied by the concrete density.

[0090] For each material, the tool queries its built-in carbon emission factor database to obtain the carbon emission intensity per unit mass or volume of that material. The carbon emission calculation data has already been explained above and will not be repeated here.

[0091] This embodiment ensures the integrity and accuracy of subsequent analysis data through a data extraction mechanism, avoiding analysis failures caused by missing data or format incompatibility in traditional solutions. Solar radiation simulation, energy consumption analysis, and carbon emission calculation form an interconnected analysis chain. The results of solar radiation analysis directly affect the solar heat gain calculation in energy consumption simulation, while the results of energy consumption analysis are a crucial input for carbon emission calculation. This automatic data flow eliminates data gaps between various professional tools in traditional solutions. More importantly, when building information model data changes, such as changing the external wall insulation material from rock wool to polystyrene board, the thermal conductivity in the material properties changes accordingly. This change automatically triggers energy consumption and carbon emission recalculations, while solar radiation analysis remains unchanged because it is unaffected by material thermal properties. This selective update mechanism based on data dependencies significantly improves computational efficiency compared to traditional full recalculation methods. The automated execution of the entire process frees designers from tedious data conversion and repetitive calculations, allowing them to focus more on scheme innovation and performance optimization, significantly improving the quality and efficiency of green building design and providing strong technical support for achieving carbon peaking and carbon neutrality goals in the building sector.

[0092] In one embodiment of this invention, the green building knowledge base includes a green building case library and a knowledge question-and-answer module. The knowledge question-and-answer module integrates an AI question-and-answer model, and the method further includes the following steps: S310. In response to a user’s knowledge query request sent through the front-end interface, the knowledge question and answer module extracts the query keywords from the knowledge query request. S320 The knowledge Q&A module retrieves matching case data from the green building case database based on the query keywords. The case data includes building design schemes, performance analysis results, and implementation effects. S330, the knowledge question answering module calls the AI ​​question answering model, takes the query keywords and case data as input, and generates knowledge answer content; S340, the knowledge Q&A module returns the knowledge answer content and case data to the user through the front-end interface.

[0093] When a user enters a knowledge query request on the front-end interface, the knowledge question-answering module immediately initiates a natural language processing flow to deeply analyze the query text. The module first preprocesses the query text, including basic operations such as removing meaningless stop words, using unified character encoding, and converting between simplified and traditional Chinese characters. Next, the module uses word segmentation technology to divide the query text into word sequences. Chinese word segmentation employs a combination of dictionary-based and statistical methods, accurately identifying specialized terms. After word segmentation, the module performs part-of-speech tagging, identifying different parts of speech such as nouns, verbs, and adjectives. Nouns and specialized terms are given higher weights because they typically carry the core semantics of the query.

[0094] The keyword extraction algorithm employs a hybrid strategy combining TF-IDF (Term Frequency-Inverse Document Frequency) and TextRank graph ranking. TF-IDF identifies words that frequently appear in the current query but are relatively rare in the entire document set, while TextRank constructs a word co-occurrence graph and iteratively calculates the importance score of each word. The results of the two algorithms are weighted and fused to obtain the final keyword list, typically extracting several core keywords. Furthermore, the module performs synonym expansion and concept generalization, for example, expanding "energy saving" to synonyms like "energy consumption reduction," and generalizing specific materials to the broader concept of "thermal insulation materials." This semantic expansion mechanism improves the recall rate of subsequent searches. For queries containing negative words, the module specifically marks the negative semantics to ensure that cases that do not meet the search criteria are excluded. The extracted query keywords are stored in a structured manner, including keyword text, weight values, part-of-speech tags, semantic categories, and other attributes, laying the foundation for accurate retrieval in the next step.

[0095] After obtaining the query keywords, the knowledge-based Q&A module immediately initiates a multi-layered case retrieval mechanism to search for matching case data in the green building case library. The green building case library is a structured knowledge repository that stores detailed information on a large number of completed green building projects. Each case data item contains three core dimensions. The first dimension is the architectural design scheme, recording the building's basic information, design concept, technical system configuration, and design drawings and BIM model files. The second dimension is the performance analysis results, including various performance indicators of the project during the design phase and after implementation, such as energy consumption simulation data, solar radiation analysis results, indoor environmental quality parameters, carbon emission data, water resource utilization efficiency, and the green building certification level obtained. The third dimension is the implementation effect, recording the actual operating data after the building is put into use, user satisfaction survey results, technical problems encountered and solutions, investment payback period analysis, and the project's social impact.

[0096] The retrieval process employs a multi-stage strategy. First, a keyword-based full-text search is performed, utilizing inverted index technology to quickly locate a set of candidate cases containing the query keywords. The search scope covers text fields such as case titles, abstracts, technical descriptions, and tags. Next, semantic similarity calculation is performed. Both the query keywords and the text descriptions of candidate cases are converted into vector representations. A word embedding model is used to calculate the cosine similarity between the query vector and the case vector; cases with similarity exceeding a preset threshold are retained. Then, multi-dimensional filtering is applied to further narrow down the result set based on user-defined filtering criteria. Finally, relevance ranking is performed, comprehensively considering factors such as text similarity, case authority, timeliness, and case data completeness, calculating a comprehensive score and sorting them in descending order. The search results typically return several of the most relevant cases, each accompanied by a relevance score and explanation of the matching reason, providing rich knowledge material for subsequent intelligent question answering.

[0097] After acquiring the retrieved case data, the knowledge-based question-answering module calls the integrated AI question-answering model to generate intelligent knowledge answers. The AI ​​question-answering model uses a large-scale pre-trained language model or domain model based on the Transformer architecture.

[0098] The question-answering generation process consists of several sub-steps. The first step is context building, where the module integrates query keywords, the user's original question, and retrieved case data into a structured input text. This input text is organized in a specific format, which helps the model understand the relationship between the question and the references. Since the case data can be very detailed, containing numerous technical parameters and charts, the module performs intelligent summarization, using an extractive summarization algorithm to extract the most relevant paragraphs and data from each case, ensuring that the input length does not exceed the model's maximum input length limit.

[0099] The constructed input text is fed into the AI ​​question-answering model for inference and computation. Internally, the model uses a multi-layered self-attention mechanism and a feedforward neural network to perform deep semantic understanding of the input text, identifying the intent of the question, extracting key information from the case study, and establishing a logical connection between the question and the answer. The model's output is a natural and fluent answer text. This text is not a simple copy-paste of the case study, but a comprehensive answer after the model's understanding and reorganization. For example, for the question of how to improve the thermal insulation performance of building exterior walls, the model might generate the following answer: Improving the thermal insulation performance of building exterior walls can be achieved by selecting insulation materials with low thermal conductivity, ensuring the continuity of the insulation layer to avoid thermal bridging, and rationally designing the thickness of the insulation layer.

[0100] The module also assesses the quality of responses, checking for factual errors, logical contradictions, or discrepancies with case data in the generated text. Uncertainty alerts are added to responses with low confidence levels. This hybrid question-and-answer mechanism, based on a large-scale language model and a case knowledge base, achieves a leap from simple keyword retrieval to intelligent knowledge services. Users can quickly obtain targeted professional advice without having to read extensive case documents themselves.

[0101] After generating the knowledge-based Q&A module, it needs to return the answers and related case data to the user through a user-friendly front-end interface. This process involves several aspects, including data formatting, visualization, and interaction design. First, the module performs rich text formatting on the answers, converting plain text into HTML format that includes headings, paragraphs, lists, bolding, links, and other styles to improve readability.

[0102] For numerical data in the responses, the module automatically generates data tables, such as organizing energy consumption comparison data from multiple cases into a table format for easy horizontal comparison by users. For data suitable for visualization, the module uses a chart generation library to create interactive charts, such as bar charts, line charts, and pie charts. These charts support interactive operations such as hovering the mouse to display detailed values ​​and clicking the legend to filter data series. Case data is displayed using a card-style layout, with each relevant case displayed as an independent information card containing core information such as a case thumbnail, project name, building type, key technology tags, and a summary of performance indicators. Users can click on a case card to expand and view complete case details, including a detailed description of the design scheme, technical drawings, performance analysis reports, and a summary of implementation results. For cases containing BIM models, the front-end interface embeds a 3D model viewer, allowing users to rotate, zoom, and section the model to intuitively understand the building's spatial layout and technical system configuration.

[0103] The module also offers related recommendation features, suggesting other potentially interesting cases or knowledge articles based on the current query and the cases the user has browsed. The recommendation algorithm combines collaborative filtering and content similarity calculation. To support users' deep learning and knowledge accumulation, the interface provides functions such as favorites, notes, and sharing. Users can add valuable cases to their personal knowledge base, add their own learning notes on the case pages, or share case links with team members. In addition, the module records the user's query history and browsing behavior, allowing users to quickly return to previously viewed content on their next visit or provide personalized knowledge recommendations based on historical behavior.

[0104] For complex technical issues, if the answers generated by the AI ​​model cannot fully meet the user's needs, the interface will provide entry points for "Contact an Expert" or "Submit an In-Depth Consultation," transferring the question to a human expert for resolution and realizing a human-machine collaborative knowledge service model. Through this multi-layered and multi-format information presentation and interactive design, professional green building knowledge is conveyed to users in an intuitive and easy-to-understand way, lowering the barrier to knowledge acquisition and improving the user's learning experience and work efficiency.

[0105] The intelligent knowledge question-answering system built through the above steps achieves efficient knowledge sharing and intelligent services in the field of green building. Natural language processing and keyword extraction technologies allow users to express questions in natural language without needing to master professional search syntax, lowering the technical threshold for knowledge acquisition. A multi-level case retrieval mechanism ensures the relevance and accuracy of the returned results; compared to traditional keyword matching retrieval, semantic similarity calculation significantly improves both recall and precision. The introduced AI question-answering model allows users to directly obtain targeted professional advice without spending a lot of time reading and understanding case documents, greatly improving knowledge acquisition efficiency. Especially for comprehensive questions across multiple cases, the AI ​​model can integrate the technical strategies of multiple cases to generate systematic solutions, which is impossible with traditional retrieval systems. Multimodal information display and interactive design, presenting knowledge in various forms such as text, tables, charts, and 3D models, meets the cognitive preferences of different users. User satisfaction surveys show that compared to pure text display, multimodal display significantly improves information comprehension speed and memory retention.

[0106] In one embodiment of this invention, the collaborative management platform further includes a data dependency management module, which is used to construct and maintain a data dependency graph. The method further includes the following steps: S410. When any data in the building information model data or multi-type output data is created or uploaded to the collaborative management platform, the data dependency management module registers the data as a data node. S420. When a user calls a tool through the green tool integration module to operate on at least one data node and generate new result data, the data dependency management module registers the new result data as a result node. S430. The data dependency management module establishes directed dependency edges between the input data nodes and the output result nodes. The directed dependency edges represent the dependency relationship between the data nodes and the result nodes. S440, the data dependency management module, constructs a data dependency graph based on all data nodes, result nodes, and directed dependency edges.

[0107] In this embodiment, the data dependency management module includes a data lineage tracing interface, and the method further includes: S1. In response to the bloodline query request sent by the user through the front-end interface, the data dependency management module extracts the target result node identifier from the bloodline query request. S2. The data dependency management module starts from the target result node and traverses backwards along the directed dependency edges in the data dependency graph to obtain all upstream data nodes. S3, the data dependency management module generates a data lineage visualization map, which displays all input data sources and their version information for the target result node; S4, the data dependency management module, returns a visual map of data lineage to the user through the front-end interface.

[0108] When any data from Building Information Modeling (BIM) data or multi-type deliverables is created or uploaded to the collaborative management platform, the data dependency management module immediately initiates the data registration process. The data registration process includes key steps such as data type identification, metadata extraction, unique identifier generation, and content hash calculation. The unique identifier for each data node uses a UUID generation algorithm to ensure global uniqueness. Metadata extraction covers basic attributes such as data name, creation time, creator, file size, file format, project, and specialty. For BIM models, semantic information such as the number of floors, building area, and statistics of major components are also extracted. Data nodes are stored in a structured manner in a graph database. The storage model includes fields such as node ID, node type, metadata dictionary, content hash value, semantic attributes, and creation timestamp, and a multi-dimensional index is established to support fast querying. Each data node also records its storage path in the distributed file system, ensuring the persistence and traceability of the original data. Through the registration mechanism, each piece of data entering the platform obtains an identity and attribute description, laying the foundation for constructing a data dependency graph.

[0109] When a user invokes a tool through the green tool integration module to operate on at least one data node and generate new output data, the data dependency management module automatically captures this computation process and registers the new output data as an output node. Tool invocation capture employs API interception technology. Before invoking an external tool, the green tool integration module sends a task start notification to the data dependency management module, containing information such as the task ID, tool name, list of input data nodes, and task parameter configuration. The data dependency management module creates a task execution context object, recording the complete execution environment of the task. After the tool completes execution and generates new output data, the green tool integration module sends a task completion notification, carrying information such as the output file path, execution time, and resource consumption. Upon receiving the completion notification, the data dependency management module immediately performs the registration process for the new output data, generating a unique identifier for the output node. In addition to basic attributes similar to those of data nodes, the metadata of the output node also records detailed information about the generated task, including the task type, the name of the tool used, the task parameter configuration, the execution timestamp, the computation time, and the task executor. For analytical outputs, key result indicators are extracted as semantic attributes of the output node to facilitate subsequent rapid querying and comparison. The result nodes also calculate content hash values ​​and store them in the graph database, using the same storage structure as the data nodes to ensure the consistency of the data model.

[0110] The data dependency management module establishes directed dependency edges between input data nodes and output result nodes while registering result nodes, explicitly expressing data flow and dependency relationships. The direction of the directed dependency edge points from the input data node to the output result node, indicating that the generation of the result node depends on the content of the data node.

[0111] The creation of dependency edges is based on the list of input data nodes recorded in the task execution context object. For each input data node in the list, the module creates a directed edge from that data node to the new result node in the graph database. Dependency edges also carry rich semantic information, including attributes such as dependency type, dependency strength, data usage method, and creation time.

[0112] Dependency strength is determined based on a predefined rule base. For example, some core input data have a high dependency strength on the outcome, while auxiliary data has a low dependency strength. When an outcome node may depend on multiple input data nodes, multiple dependency edges are established, forming a many-to-one relationship. For instance, an outcome node may depend on multiple input data nodes simultaneously. The establishment of dependency edges uses transactional operations to ensure atomicity between edge creation and outcome node registration, avoiding inconsistencies such as outcome nodes existing but dependencies missing. Dependency edges also record the version information of the data node when it is used. By referencing the content hash value of the data node, the system accurately identifies which historical version of the data node the outcome node uses, which is crucial for version tracing and change impact analysis.

[0113] The data dependency management module constructs a complete data dependency graph in the graph database based on all registered data nodes, output nodes, and established directed dependency edges. The data dependency graph is a directed acyclic graph structure, where nodes represent data or outputs, edges represent dependencies, and the direction of the edges indicates the direction of data flow. Graph construction is a continuously incremental process; whenever new data is uploaded or new outputs are generated, the corresponding nodes and edges are dynamically added to the graph.

[0114] Graph databases employ specialized graph database systems optimized for storing and querying graph structures, supporting efficient graph traversal operations. The storage of data dependency graphs utilizes an attribute graph model, where nodes and edges can carry any number of attribute key-value pairs, providing exceptional flexibility.

[0115] To support the needs of large-scale projects, the graph is logically partitioned according to dimensions such as project, profession, and time. Each partition can be stored and queried independently, improving the system's scalability. The graph database automatically maintains the graph's topology, including statistical information such as the in-degree and out-degree of nodes and the graph's level depth. This information is used to optimize query performance and support visualization. The data dependency management module also periodically performs integrity checks on the graph, checking for anomalies such as isolated nodes, dangling edges, or circular dependencies, and generates health reports. The graph construction process records detailed audit logs, including the creation time, creator, and creation reason for each node and edge, supporting complete historical traceability. The completed data dependency graph becomes the core data structure of the entire collaborative management platform, supporting advanced functions such as change impact analysis, task triggering, and lineage tracing.

[0116] When a user needs to understand the source and generation process of a specific output data, they can send a lineage query request through the front-end interface. The request includes the identifier of the target output node. Upon receiving the request, the data lineage tracing interface of the data dependency management module first verifies the parameters, checking the validity of the node identifier and whether the current user has permission to view the lineage information of the output. If the verification is successful, the interface extracts the target output node identifier from the request parameters and queries the graph database for complete information about the node, including metadata such as node type, creation time, and generation task, confirming that the node does exist and is indeed an output node.

[0117] After confirming the target result node, the data dependency management module initiates a reverse graph traversal algorithm, tracing back along directed dependency edges from the target result node to obtain all upstream data nodes. The direction of reverse traversal is opposite to the direction of dependency edges, i.e., tracing back from the result node to the input data node. The traversal employs either depth-first search or breadth-first search algorithms, with breadth-first search being more suitable for generating hierarchical lineage views. The traversal process begins with the target result node, querying all incoming edges pointing to that node; the starting node of these incoming edges is the direct input data. For each direct input data node, if the node itself is also a result node, the traversal continues upwards, querying the input data of that node, recursively until the source data node is reached. The traversal process records the complete dependency path, including all nodes and edges on the path, as well as the version information of each node. Version information is obtained through the hash values ​​of data nodes recorded on dependency edges; by querying the data version corresponding to the hash value, detailed information such as version number, version creation time, and version creator is obtained. For complex dependencies, a single outcome may have multiple traceability paths. The traversal algorithm retrieves all paths and removes duplicates, ensuring that each upstream data node is recorded only once. The traversal result includes a list of data nodes and a list of dependent paths, providing the data foundation for subsequent visualization.

[0118] The data dependency management module generates a data lineage visualization graph based on data nodes and dependency paths obtained through reverse traversal. The visualization graph uses a hierarchical tree or network layout, with target result nodes located on the far right or bottom, source data nodes on the far left or top, and intermediate result nodes displayed in intermediate levels. Dependencies between nodes are indicated by lines. The graph generation utilizes a front-end visualization library and supports interactive operation. Each node is displayed as a card, showing key information such as node name, node type icon, version number, and creation time. For data nodes, the card displays information such as data file name, file size, and uploader; for result nodes, the card displays information such as the generation task type, tools used, and key result metrics. Node colors are differentiated according to node type; for example, source data nodes, intermediate result nodes, and target result nodes use different colors. The style of the lines is differentiated according to dependency strength: solid lines for strong dependencies and dashed lines for weak dependencies. Dependency type text can be labeled on the lines. The graph supports various interactive operations. Users can click on nodes to view detailed information, including complete metadata, file previews, and version history; click on connections to view detailed descriptions of dependencies, such as data usage and impact; drag and drop nodes to adjust the layout, or use zoom and pan functions to browse large graphs. The graph also provides filtering functions, allowing users to choose to display only certain types of nodes or data within a specific time range, simplifying the display of complex graphs. For complex lineage relationships containing a large number of nodes, the graph uses a layered, collapsed display, showing only the main path by default. Users can click the expand button to view detailed branch paths.

[0119] The data dependency management module generates a data lineage visualization graph and returns it to the user through the front-end interface. The graph data, including node lists, edge lists, and layout information, is transmitted in JSON format. Upon receiving the data, the front-end uses a visualization library to render it into an interactive graph. The graph is embedded in a dedicated lineage viewing page or pop-up window, with a toolbar supporting export, print, and share operations. Users can export the graph as an image or PDF document for report writing or team sharing. The page also displays statistical information about the lineage, such as the total number of data sources involved, the number of processing steps, and the time span of the data flow.

[0120] For audit and compliance needs, logs are recorded for each lineage query, including query time, query user, and query outcome node. The lineage tracing function not only helps users understand the process of outcome generation, but also plays a crucial role in scenarios such as data quality issue investigation, change impact assessment, and compliance auditing, improving the transparency and traceability of data management.

[0121] This embodiment achieves a complete record and visual representation of the complex data flow relationships during the collaborative design process of green buildings. The data dependency graph elevates traditional file-level data management to semantic-level dependency management, giving each data point and outcome a clear identity and a complete relationship network, laying a core data foundation for intelligent change management. The lineage tracing function allows users to view the complete generation process of any outcome with a single click, clearly showing every step from source data to the final result, significantly improving tracing efficiency compared to manual searching. In practical applications, when a project team has questions about a particular energy consumption analysis result, lineage tracing can quickly locate the BIM model version used, the meteorological data source, the calculation tool version, and all intermediate processing steps, helping to quickly identify the root cause of the problem. The entire data dependency management mechanism supports advanced functions such as change identification, semantic analysis, and task triggering, serving as the technological cornerstone for intelligent collaborative management and improving the project's data governance level and collaborative efficiency.

[0122] In one embodiment of this invention, the data dependency management module further includes a change identification unit, and the method further includes the following steps: S510, Calculate the first hash value of the target data corresponding to the data node; S520. When the target data corresponding to the data node is modified by the user and a new version of the data is uploaded, the change identification unit calculates the second hash value of the new version of the data. S530, the change identification unit compares the second hash value with the first hash value, and if the second hash value is different from the first hash value, the change identification unit determines that the data node has changed and triggers the semantic change analysis process.

[0123] When a data node is first registered to the collaborative management platform, the change identification unit in the data dependency management module will immediately perform hash value calculation on the target data corresponding to the data node, and generate the first hash value as the unique digital fingerprint of the data content.

[0124] The hash value calculation uses the SHA-256 cryptographic hash algorithm, which features one-way hashing, determinism, and avalanche effect. The calculation process first requires reading the complete binary content of the target data. For smaller files, the change detection unit loads the entire file content into memory at once for hash calculation. For larger files, a streaming approach is used, reading the file in blocks and updating the hash calculation state block by block to avoid memory overflow issues.

[0125] To improve computational efficiency, the change detection unit utilizes the parallel computing capabilities of multi-core processors. For datasets containing multiple files, it can calculate the hash values ​​of each file in parallel, then concatenate all sub-hash values ​​in a specific order and perform hash calculation again to obtain the comprehensive hash value of the entire dataset. For structured data such as Building Information Modeling (BIM) data, the change detection unit also performs normalization preprocessing, extracting the core geometric and attribute data of the model, reserializing it according to a standardized format, and then calculating the hash value to ensure that only genuine design changes will lead to changes in the hash value.

[0126] The first hash value, once calculated, is stored in the data node's metadata record and persisted to the database along with information such as node ID, creation time, and file path. Simultaneously, the change detection unit maintains a hash value index table in memory, using the data node ID as the key and the hash value as the value, supporting fast hash value lookup and comparison operations. For frequently accessed hot data, its hash value is cached in a distributed caching system, further improving query performance. This data fingerprinting mechanism based on cryptographic hashing provides an efficient and reliable technical means for subsequent change detection.

[0127] When the target data corresponding to the data node is modified by the user and a new version of the data is uploaded, the change identification unit immediately starts the hash value calculation process of the new version of the data to generate a second hash value for comparison with the original version.

[0128] Users may modify data in various scenarios, such as designers modifying a BIM model locally and then re-uploading it, engineers updating energy consumption analysis input parameters and then resubmitting, or project managers correcting the content of a technical document and then overwriting the original file. The collaborative management platform's file upload interface will detect the target path of the uploaded file. If it finds that a data node already exists at that path, it determines that this is a data update operation rather than a new data creation operation.

[0129] Upon receiving a data update notification, the change identification unit first saves the new version data to a temporary storage area to avoid directly overwriting the original data and making rollback impossible. Next, it calculates the hash value of the new version data using the same SHA-256 hash algorithm as step S510, ensuring consistency in the calculation method and comparability of the results. For large files, it also uses streaming reading and block-based calculation, displaying a progress indicator to the user during the calculation process. After the second hash value is calculated, it is temporarily stored in a memory variable, awaiting comparison with the first hash value.

[0130] To address the potential file corruption during network transmission, the change detection unit pre-calculates the file's hash value on the client side and includes this hash value in the upload request. After receiving the file, the server recalculates the hash value and compares it with the value provided by the client. If they do not match, it indicates data corruption during transmission, and the user is prompted to re-upload. This end-to-end integrity verification mechanism ensures that the data entering the change detection process is complete and undamaged.

[0131] For large files that support incremental uploads, the change detection unit employs block hashing technology. This divides the file into fixed-size blocks, calculates the hash value for each block, and then constructs a hash tree. The hash value of the root node serves as the second hash value for the entire file. The advantage of this approach is that when certain parts of the file remain unchanged, previously calculated block hash values ​​can be reused; only the hash of the changed parts needs to be recalculated, reducing computational overhead. The change detection unit also records the timestamp of the second hash value calculation, as well as the user information and operation type that triggered the calculation. This information is written to the audit log for subsequent change history tracing and accountability. This rigorous new hash value calculation mechanism ensures the accuracy and traceability of change detection.

[0132] After obtaining the second hash value, the change identification unit immediately initiates a hash value comparison process, comparing the second hash value with the first hash value read from the data node metadata character by character. Hash value comparison is an efficient operation. Since SHA-256 generates a fixed-length string, the comparison process only requires a simple string equality check, maintaining a fast response even in high-concurrency scenarios.

[0133] The comparison results fall into two categories. If the second hash value is exactly the same as the first hash value, it means that the newly uploaded data content is completely consistent with the original data. In this case, the change identification unit will return a "data has not changed" prompt to the user and terminate the subsequent change processing process to avoid unnecessary consumption of computing resources.

[0134] If the second hash value differs from the first hash value, the change detection unit immediately determines that a substantial change has occurred to the data node and triggers a series of change processing operations. First, the change detection unit creates a new data version record, marking the original data as a historical version and storing it in the version repository, while setting the new version data as the currently active version. The version record includes information such as the version number, version creation time, version creator, version description, and the storage path pointing to the actual data file. The change detection unit updates the data node's metadata, replacing the first hash value with the second hash value to ensure that the latest hash value is used as the benchmark for the next change detection. Next, the change detection unit triggers a semantic change analysis process, a deeper analysis process designed to understand the specific content and scope of the change. Semantic change analysis focuses not only on "whether the data has changed," but also on "how the data has changed" and "what impact the change will have."

[0135] Semantic change analysis employs different strategies for different types of data. For Building Information Modeling (BIM) data, the analysis process utilizes a model comparison engine. This engine can parse the old and new versions of the BIM model and identify specific changes, such as which geometric or attribute parameters of components have been added, deleted, or modified. For tabular data, the analysis process performs a row-by-row, column-by-column comparison to identify newly added, deleted, or modified data rows and cells.

[0136] For text documents, a text difference algorithm is used to generate change patches, identifying newly added, deleted, and modified paragraphs and sentences. The results of semantic change analysis are stored in a structured manner, including change type classification, change impact assessment, and a list of affected downstream data nodes. The change identification unit also sends change notifications to relevant users through methods such as in-site messages, emails, and mobile application push notifications, ensuring that project team members are informed of data changes in a timely manner.

[0137] For changes to critical data, a change approval process is required. Changes must be reviewed and confirmed by the project manager or technical lead before taking effect, preventing accidental or unauthorized modifications. This change identification mechanism, combining hash value comparison and semantic analysis, enables precise capture and deep understanding of data changes, providing a reliable data foundation for subsequent dependency updates and impact analysis.

[0138] This embodiment achieves automated and precise monitoring of all data changes in the collaborative management platform through a change identification mechanism. The first hash value creates a unique content fingerprint for each data node. This cryptographic hashing technique offers higher reliability and accuracy compared to traditional file modification timestamps or file size comparisons, ensuring that no substantial changes are missed. The new version's hash value calculation process uses the same calculation method as the original version, guaranteeing the accuracy of the comparison results. The introduction of block hashing and incremental calculation techniques significantly improves the efficiency of change detection for large files. The hash value comparison and semantic change analysis process not only quickly determines whether data has changed, but more importantly, it deeply understands the specific content and nature of the changes. This semantic understanding of changes is a key prerequisite for achieving intelligent update-dependent processing. Compared to traditional manual change identification methods, the automated change identification mechanism reduces change detection time from hours to seconds, and improves accuracy from the level of manual identification to near-perfect accuracy, completely eliminating the problem of missed changes due to human error. More importantly, this change identification mechanism lays a solid foundation for advanced functions such as semantic change analysis, impact assessment, and automated task triggering, enabling the entire collaborative management platform to transform from a passive data storage tool into a proactive intelligent management assistant, thereby improving the collaborative design efficiency and data consistency assurance capabilities of green building projects.

[0139] In one embodiment of this invention, the change identification unit includes a model comparison engine, and the semantic change analysis process includes the following steps: S610, the model comparison engine analyzes the new version data and target data, and extracts the internal structure information of the data; S620, the model comparison engine compares the internal structure information of the new version data with the target data to identify the specific changes; S630, the model comparison engine determines the change type based on the specific changes. The change types include geometric changes, attribute changes, and non-functional changes. Geometric changes include changes in component position, component size, and component thickness. Attribute changes include changes in material type and thermal parameters. Non-functional changes include changes in remarks information and layer attribute changes. S640, the model comparison engine adds semantic tags to data nodes according to the type of change. Semantic tags include structural change tags, building envelope change tags, HVAC parameter change tags, and non-critical information change tags.

[0140] When the change identification unit determines that a data node has changed and triggers the semantic change analysis process, the model comparison engine first starts the data parsing module to perform in-depth analysis on the new version data and the target data to extract the internal structure information of the data.

[0141] The data parsing process requires corresponding parsing strategies for different data formats. For Building Information Modeling (BIM) data, the model comparison engine supports parsing multiple mainstream BIM formats, including Revit's RVT format and ArchiCAD's PLN format. Taking the IFC format as an example, the parsing process first reads the metadata area in the file header to obtain global information such as the IFC version number, project name, and coordinate system definition. Next, the parsing engine reads the entity definitions line by line according to the IFC STEP physical file format specification. The parsing engine then constructs an entity object graph, loading all entities and their references in the file into memory to form a complete object network.

[0142] For each component entity, the parsing engine extracts its core attribute information, including a globally unique identifier, component name, component type, geometric representation, material associations, floor level, and spatial location. For parsing the geometric representation, the engine recursively parses these geometric definitions, ultimately converting them into a unified internal geometric representation, such as a triangular mesh model or parametric geometric description. For material properties, the engine traces the association chain from the component to the set of material layers and then to the specific material, extracting the name, thickness, and physical properties of each material layer.

[0143] Spatial location information is obtained by analyzing the local coordinate system and geometric transformation matrix of the components, calculating the component's position and orientation in the global coordinate system. For non-BIM format data, such as energy consumption analysis reports, the parsing engine will perform structured parsing according to the corresponding data mode, extracting information such as input parameters, calculation results, and chart data from the report.

[0144] For drawing documents, the parsing engine extracts layer structure, graphic elements and their attributes, annotation information, etc. After parsing, the model comparison engine organizes the extracted internal structural information into a standardized data structure, usually a tree or graphical structure. Nodes represent components or data items, edges represent the relationships between them, and each node comes with a complete attribute dictionary. To improve the efficiency of subsequent comparisons, the parsing engine also generates feature vectors for each component or data item, encoding its key attributes into numerical vectors for rapid similarity calculation.

[0145] For large models, the parsing process employs streaming and parallel computing techniques. The model is divided into multiple sub-models based on spatial regions or floors, and these sub-models are assigned to multiple processing threads for parallel parsing. The parsed sub-models are then merged into complete structural information. The entire parsing progress is fed back to the user interface in real time.

[0146] After parsing the new version data and the target data, the model comparison engine activates the structural information comparison module to perform a comprehensive difference analysis of the internal structural information of the two versions, identifying specific changes. The comparison process employs a multi-level matching strategy. First, a global-level structural comparison is performed, counting the total number of components or data items in the two versions. If the numbers differ, it indicates the addition or deletion of components. Next, entity-level matching is performed. For BIM models, the comparison engine uses the component's globally unique identifier (GUID) as the primary matching key. Because the GUID is generated and remains unchanged when the component is created, cross-version component tracking is possible.

[0147] The comparison engine iterates through all components in the new version. For each component, it searches for components with the same GUID in the target version. If found, the two components are paired and compared in detail; if not found, the component is determined to be a new component. It then iterates in reverse through the components in the target version. If a component's GUID does not exist in the new version, the component is determined to have been deleted.

[0148] For successfully matched components, the comparison engine performs a detailed comparison at the attribute level, including multiple dimensions such as geometric attributes, material attributes, spatial location, and relationships. The comparison of geometric attributes uses a combination of parametric comparison and geometric similarity calculation. For parametric components, the parameter values ​​are directly compared to determine whether they have changed. A tolerance threshold is set during the numerical comparison to avoid misjudgments caused by floating-point precision issues.

[0149] For complex geometries, the comparison engine calculates features such as volume, surface area, and bounding box of two geometries. If significant differences exist in these features, a more refined geometric comparison is performed, employing techniques such as mesh alignment and point cloud distance calculation to quantify the degree of geometric change. Material property comparison compares material layers layer by layer, checking for changes in material name, thickness, order, and physical properties. Spatial location comparison calculates changes in the coordinates of the component's center point and orientation vector; if the position offset or orientation angle change exceeds a preset threshold, it is considered a location change. Relationship comparison checks for changes in the topological connections between components. For non-BIM format data, such as energy consumption analysis reports, the comparison engine compares the input parameters and output results to identify which parameters have been modified and which result values ​​have changed. The comparison process generates a detailed change list, with each change record containing information such as change type, changed object, changed attribute, value before change, value after change, and change amount. To facilitate user understanding, the comparison engine also generates a visual comparison view of changes. Newly added, deleted, modified, and unchanged components are identified by different colors in the 3D model viewer. Users can click on changed components to view detailed change information. For comparisons of large models, the engine employs spatial indexing and hash acceleration techniques, organizing components into an octree or grid structure based on their spatial location. Only spatially similar components are compared, avoiding global pairwise comparisons.

[0150] After identifying the specific changes, the model comparison engine activates the change type classification module, which categorizes the changes into a predefined change type system based on their nature and scope of impact.

[0151] The classification of change types employs a combination of rule-based expert systems and machine learning classifiers to ensure accuracy and comprehensiveness. First, the classification module analyzes the attribute categories involved in the change. Changes involving component geometric parameters are categorized as geometric changes, which are further subdivided into several subtypes. Component location changes refer to a shift in the component's spatial coordinates, determined by the change in the coordinates of the component's center point exceeding a preset threshold. Component size changes refer to alterations in the component's length, width, height, and other dimensional parameters. Component thickness changes specifically refer to changes in the thickness parameters of planar components such as walls and floor slabs; these changes significantly impact the building's thermal performance. Changes involving material-related properties are categorized as attribute changes, which also include several subtypes.

[0152] Material type change refers to the replacement of materials used in a component, which significantly alters the component's thermal performance. Thermal parameter change refers to the modification of the material's physical properties; these changes affect energy consumption calculations. Changes that do not affect building performance analysis are classified as non-functional changes, including changes to remarks information and layer attributes.

[0153] Changes to remarks information refer to modifications to non-technical attributes such as descriptive text, annotations, and labels of components. These changes do not affect any performance calculations. Changes to layer attributes refer to modifications to the display attributes of the layers containing graphic elements in CAD drawings. These changes only affect the visual presentation of the drawings and do not involve the actual design content.

[0154] The decision tree algorithm is used to determine the type of change. The root node of the decision tree first determines whether the change involves geometric or material properties. If so, it proceeds to the functional change branch; otherwise, it proceeds to the non-functional change branch. Within the functional change branch, the specific parameters involved are further determined. For complex changes, multiple types may be involved simultaneously, in which case multiple type labels are assigned to the change. The classification module also assesses the degree of impact of the change. For geometric changes, the magnitude of the change is calculated; the larger the change, the greater the impact. For material changes, the material property database is queried to compare the performance parameters of the old and new materials. The classification results are stored in a structured manner. Each change record, in addition to the original change object and change content information, also includes a change type label, an impact rating, and recommended follow-up actions. This refined change type classification provides accurate semantic information for subsequent intelligent impact analysis and task triggering.

[0155] After determining the type of change, the model comparison engine starts the semantic tag generation module to add structured semantic tags to the data nodes according to the type of change. These semantic tags are the key basis for subsequent intelligent task triggering and impact analysis.

[0156] The semantic tagging system adopts a hierarchical design. Top-level tags indicate the professional field or scope of impact of the change, including structural change tags, building envelope change tags, HVAC parameter change tags, and non-critical information change tags. Structural change tags identify changes affecting building structural safety and load-bearing capacity. These changes include changes in the geometric dimensions, location, and material strength grades of load-bearing components. This tag triggers relevant structural analysis tasks, such as recalculating structural load-bearing capacity and assessing seismic performance. Building envelope change tags identify changes to the building envelope that affect building thermal performance and energy consumption, including geometric changes, material changes, and thermal parameter changes to components such as exterior walls, roofs, windows, and doors. This tag triggers building thermal and energy consumption analysis tasks, such as recalculating the heat transfer coefficient of the building envelope, simulating annual energy consumption, and evaluating energy efficiency.

[0157] The HVAC parameter change tag is used to identify changes that affect the design and operation of the HVAC system, including adjustments to indoor design temperature, changes to fresh air volume standards, changes to the type of air conditioning system, and modifications to equipment performance parameters. This tag will trigger tasks such as recalculating cooling and heating loads and verifying equipment capacity. The non-critical information change tag is used to identify changes that do not affect any performance analysis, such as modifications to remarks, layer attributes, color labels, etc. These changes will not trigger any calculation tasks and are only saved as version history.

[0158] The generation of semantic tags employs a multi-source information fusion strategy, comprehensively judging not only the change type determined in step S630 but also contextual information such as the component's functional attributes, professional affiliation, and location within the building. The tag generation module incorporates a rich domain knowledge rule base, with rules expressed in "IF-THEN" format. For complex scenarios, a single change may require multiple tags. The generated semantic tags are stored in a structured manner in the metadata of data nodes, with tag format including fields such as tag type, tag value, confidence level, generation time, and generation basis. Tag information is indexed into the tag database, supporting rapid tag-based queries and statistical analysis. These semantic tags play a crucial role in the subsequent task triggering module. Task triggers subscribe to specific types of tags, automatically initiating corresponding analysis tasks upon detecting a relevant tag, achieving intelligent linkage from data change to task execution.

[0159] This embodiment provides a complete and accurate structured information foundation for change analysis through deep data parsing. Compared to traditional solutions that rely solely on file-level comparisons, the internal structure-based parsing refines the granularity of change identification from the file level to the component level and even the attribute level, significantly improving identification accuracy. The multi-level structure comparison mechanism, through the comprehensive application of various technologies such as GUID matching, parameter comparison, and geometric similarity calculation, can accurately identify every subtle change in the model, resulting in high recall. Change type classification refines changes from indiscriminate "modifications" into multiple categories such as geometric changes, attribute changes, and non-functional changes, further subdivided into multiple subtypes. This refined classification provides a basis for subsequent differentiated processing, avoiding the "one-size-fits-all" full recalculation problem of traditional solutions. The semantic tag generation mechanism establishes a semantic bridge from underlying data changes to upper-level business impacts by mapping change types to domain-related semantic tags, enabling the understanding of domain knowledge and achieving truly intelligent change management. The execution efficiency of the entire semantic change analysis process is significantly improved. More importantly, the results of semantic change analysis provide precise input for subsequent intelligent task triggering, enabling the selective triggering of necessary recalculation tasks based on the nature and scope of the change, rather than simply recalculating all content. This significantly improves the efficiency of computing resource utilization and the overall design iteration cycle of the project, thereby enhancing the efficiency and quality of collaborative green building design.

[0160] In one embodiment of this invention, the data dependency management module further includes a task triggering unit, and the method further includes the following steps: S710. In the data dependency graph, the task triggering unit starts from the data node that has changed and traverses downstream along the directed dependency edge, marking all downstream result nodes as in an invalid state. S720, the task triggering unit obtains the semantic tags of the data nodes; S730, the task triggering unit determines the analysis task type affected by the change type based on the semantic tags; S740, the task triggering unit only triggers the recalculation of the task for the result node corresponding to the analysis task type, while other result nodes remain in their original state.

[0161] When the data dependency management module detects a change in a data node, the task triggering unit immediately initiates the impact scope analysis process. Starting from the changed data node, it traverses downstream in the data dependency graph, systematically identifying all potentially affected output nodes. The data dependency graph is a directed acyclic graph (DAG) structure, where nodes represent data or outputs, and directed edges represent data flow and dependencies, pointing from input data to output outputs. The traversal process employs a breadth-first search (BFS) algorithm to ensure that affected outputs are identified layer by layer according to the dependency hierarchy.

[0162] In practice, the task triggering unit first adds the changed data node to a processing queue and iteratively processes it. In each iteration, a node is retrieved from the queue, and all its outgoing edges are queried to identify directly dependent downstream nodes. For each downstream node, if it is not yet marked as invalid, its status is updated to "Invalid" and added to the queue for further propagation. This process is recursively repeated until all reachable downstream nodes have been traversed and marked. For example, suppose a BIM model data node for an exterior wall experiences a change in insulation material. This data node is directly used by several analysis tasks, and the result nodes of these tasks will be marked as invalid in the first round of traversal. However, the result of one of these analyses is then used by another evaluation task, so the evaluation report result node will be marked as invalid in subsequent traversals.

[0163] This layered propagation mechanism ensures the complete identification of the impact of changes, leaving no indirectly dependent downstream outcomes unaccounted for. To improve traversal efficiency, the data dependency graph is indexed using an adjacency list during storage, with each node maintaining a list of outgoing edges. For large projects, the dependency graph may contain a large number of nodes and edges. A parallelization strategy is employed during traversal, dividing the graph according to topological levels. Nodes at the same level can be processed in parallel, while sequential dependencies are maintained between different levels. This parallelization significantly improves traversal efficiency.

[0164] While marking failure status, the task triggering unit also records the cause of failure, i.e., which upstream data node caused the failure of the output node. This information is written to the failure log, including the failure timestamp, the source data node ID that triggered the change, the failure propagation path, and other detailed information, providing a basis for subsequent problem tracing and auditing. The failure status marking is immediately synchronized to the front-end interface. When users view the project output list, they will see failed outputs displayed with a special visual style or label, reminding them that the input data on which these outputs are based has changed, and the results may no longer be accurate, requiring recalculation. This visual failure prompt effectively prevents users from making decisions due to the misuse of expired data.

[0165] After marking the failures of downstream result nodes, the task triggering unit enters the intelligent task filtering stage. First, it needs to acquire the semantic tag information carried by the data nodes that have undergone changes. Semantic tags are generated during change identification and semantic analysis, containing key information such as the nature of the change, its scope of impact, and its professional field. The task triggering unit reads all associated semantic tags by querying the metadata records of the data nodes. Semantic tags are stored in a structured data format, with each tag containing attributes such as tag type, tag value, and generation time. For example, after a change in insulation material, an exterior wall data node may carry semantic tags such as "envelope structure change tag," "thermal parameter change tag," and "material type change tag." The task triggering unit loads these tags into memory, constructing a tag set for subsequent task matching and judgment. For changes carrying multiple tags simultaneously, they are sorted according to tag priority, with tags having a larger impact scope processed first.

[0166] The tag acquisition process also involves semantic expansion, automatically deriving implicit tags based on the hierarchical and implied relationships between tags. For example, if a data node carries a "structure change tag," the system will automatically deduce that the change also belongs to the higher-level tag "architectural change." This semantic expansion mechanism is based on a pre-built domain ontology knowledge base, which defines the semantic relationships between various tags, forming a hierarchical tree or semantic network of tags. When the task triggering unit acquires a tag, it simultaneously queries the ontology base to obtain all parent tags and related tags for that tag, ensuring the completeness of task matching. For example, for the "thermal parameter change tag," the ontology base defines its association with tags such as "energy consumption impact tag" and "comfort impact tag," and these related tags will also be considered. The tag acquisition process also involves context enhancement, further clarifying the semantics of the tag by combining other attribute information of the data node. For example, for the same "geometric change tag," if the changed component is an exterior wall, the meaning of the tag leans towards thermal performance impact; if the changed component is an interior wall, the meaning of the tag leans towards spatial layout impact. The task triggering unit reads contextual attributes of data nodes, such as component type, discipline, and location within the building. It then combines these with semantic tags to generate an enhanced semantic description, which more accurately depicts the characteristics of the change. The acquired semantic tag information is cached in the task triggering unit's working memory, avoiding repeated database queries and improving the efficiency of subsequent processing.

[0167] After acquiring the semantic tags of the data nodes, the task triggering unit activates the task type matching engine. Based on the semantic tags, it determines which types of analysis tasks will be affected by the change and need to be re-executed. Task type matching is based on a pre-configured task-tag mapping rule base. This rule base defines which semantic tags each analysis task type is sensitive to, i.e., which types of changes will affect the calculation results of the task. The rule base is organized in a structured manner, defining different degrees of influence of tags on tasks. For example, for the "energy consumption analysis task," the rule base defines that "building envelope change tag" and "HVAC parameter change tag" have a high impact, "structural change tag" has a low impact, and "non-critical information change tag" has no impact.

[0168] The task triggering unit traverses the rule base. For each analysis task type, it checks whether the semantic tags carried by the data nodes match the sensitive tags of that task. If a match is found and the impact reaches a preset threshold, the task type is added to the list of tasks to be triggered. The matching process uses a combination of fuzzy matching and exact matching. Exact matching requires tags to be completely identical, while fuzzy matching allows for semantic similarity of tags or hierarchical relationships.

[0169] For example, if the rule base defines a task as sensitive to "insulation material change label", and the data node carries its parent label "material type change label", the fuzzy matching mechanism will determine that the two match successfully, because insulation material change is a special case of material type change.

[0170] The determination of task types also considers the dependencies between tasks. The execution of some tasks depends on the results of other tasks. For example, the "green building evaluation task" depends on the results of the "energy consumption analysis task." If the energy consumption analysis needs to be re-executed, the evaluation task must also be re-executed. The task triggering unit constructs a task dependency graph. After identifying the directly affected tasks, it propagates downstream along the task dependency edges, adding all indirectly dependent tasks to the list to be triggered.

[0171] For complex changes, multiple task types can be triggered simultaneously. The task triggering unit prioritizes these tasks and plans their execution order. Prioritization is based on factors such as task importance, urgency, and user attention. Execution order planning is based on the dependencies between tasks, ensuring that dependent tasks are executed first, and tasks that depend on other tasks are executed later, avoiding data inconsistencies. The planning algorithm uses topological sorting to transform the task dependency graph into a linear execution sequence. Tasks that can be executed in parallel are marked as parallelizable batches, which can be executed concurrently to improve efficiency. The task type determination results in a detailed task execution plan, including task type, task priority, execution order, estimated execution time, and required computing resources. This plan is then submitted to the task scheduler for actual task execution.

[0172] After determining the types of analysis tasks affected by the change, the task triggering unit enters the selective task triggering phase, realizing a key shift from the traditional "full recalculation" to "on-demand recalculation." The task triggering unit first performs a refined screening of all downstream result nodes marked as invalid in step S710, identifying which result nodes' corresponding analysis task types were determined to be affected in step S730. This screening process is achieved by querying the metadata of the result nodes. The task triggering unit matches the task type of the result node with the list of affected task types determined in step S730. If a match is found, the result node needs to be recalculated; if no match is found, although the result node is marked as invalid, its results are still valid and can maintain their original state. The invalidation mark is removed, and the state is restored to "valid." For example, suppose an exterior wall data node undergoes a change in its remarks information, carrying a "non-critical information change tag." If step S730 determines that this change does not affect any performance analysis tasks, then although all downstream result nodes are marked as invalid in step S710, these result nodes will be remarked as valid in step S740, without triggering any recalculation tasks. Conversely, if an exterior wall data node undergoes a change in insulation material, carrying a "building envelope change tag," and step S730 determines that this change affects energy consumption analysis and carbon emission calculations, then only the result nodes corresponding to these two types of tasks will remain invalid and trigger recalculation. The result nodes for solar radiation analysis, because they are not affected by insulation materials, will be restored to a valid state.

[0173] For output nodes that require recalculation, the task triggering unit generates a specific task instance. This task instance contains complete information such as task ID, task type, list of input data nodes, output output nodes, task parameters, and execution priority. The task instance is then submitted to the task queue, awaiting allocation of computing resources by the task scheduler for execution.

[0174] The task queue uses a priority queue data structure, where higher-priority tasks are executed first. For computationally intensive tasks, the task triggering unit assesses the task's computational resource requirements, and this information is passed to the resource scheduler for the appropriate allocation of cloud computing resources.

[0175] The task triggering unit also sets timeout and retry policies for each task instance. If the task execution times out or fails, the system will automatically retry several times. If it still fails after retrying, the task will be marked as "execution failed" and an alarm notification will be sent to the project administrator.

[0176] For outcome nodes that maintain their original state, the task triggering unit adds a "Change Exemption" record to their metadata. This indicates that although the upstream data of the outcome has changed, since the type of change does not affect the calculation logic of the outcome, recalculation is not required. This record includes information such as the exemption reason, exemption time, and semantic tags for making the exemption decision, providing a basis for subsequent auditing and traceability. The results of task triggering are fed back to the front-end interface in real time, allowing users to see which outcomes are being recalculated, which outcomes remain unchanged, and the progress and estimated completion time of each recalculation task.

[0177] In this embodiment, downstream impact scope identification ensures the completeness of the change's impact. A graph traversal algorithm systematically tracks all potentially affected outcome nodes, avoiding omissions caused by reliance on manual experience in traditional solutions, significantly improving the accuracy of impact identification. Semantic tag acquisition establishes a semantic bridge from underlying data changes to upper-level business impacts, enabling an understanding of the nature and meaning of changes. This semantic understanding capability is a prerequisite for intelligent task selection. The task type matching engine, through a pre-configured rule base and intelligent matching algorithm, accurately determines which analysis tasks will be affected by changes, achieving high matching accuracy. Compared to the traditional "one-size-fits-all" approach, it avoids a large amount of unnecessary recalculation. Selective task triggering is the implementation phase of the entire mechanism. By triggering only truly affected tasks while keeping other task results unchanged, the utilization efficiency of computing resources is greatly improved, and the project design iteration cycle is significantly shortened. For example, in a large-scale green building project, in a traditional solution, any change to the BIM model could trigger the recalculation of all related tasks, resulting in lengthy processing times. By adopting the intelligent task triggering mechanism of this invention, a typical change in external wall insulation materials only triggers the recalculation of related tasks such as energy consumption analysis, carbon emission calculation, and green building evaluation, significantly reducing the total time and greatly improving efficiency. More importantly, this selective update mechanism not only improves efficiency but also ensures data consistency, because all truly affected results are updated in a timely manner, while unaffected results remain stable, avoiding the chaos caused by unnecessary version changes. The entire task triggering process is fully automated, requiring no manual intervention. Designers only need to focus on the design work itself, while the system intelligently manages all dependency updates and task execution in the background, improving the efficiency of collaborative design and the quality of project delivery.

[0178] In one embodiment of this invention, the task triggering unit determines the analysis task type affected by the change type based on semantic tags, including the following steps: S810. When the semantic tag is a non-critical information change tag, the task triggering unit determines that the change does not affect any performance analysis and does not trigger any recalculation task. S820. When the semantic tag is the building envelope change tag, the task triggering unit determines that the analysis task type includes solar radiation analysis task, daylighting analysis task and energy consumption simulation task, and triggers the corresponding recalculation task. S830. When the semantic tag is a structural change tag, the task triggering unit determines that the analysis task type includes structural calculation task, load calculation task and energy consumption calculation task, and triggers the corresponding recalculation task.

[0179] When the task triggering unit identifies a data node carrying a semantic tag indicating a non-critical information change, it immediately initiates a rapid exemption determination process. A non-critical information change tag indicates that the change only involves descriptive or visual attributes that do not affect building performance, such as component annotations, layer colors, line styles, and label text. The task triggering unit queries its built-in exemption rule base, which explicitly defines that non-critical information changes have no impact on any performance analysis tasks. Based on this rule, the task triggering unit directly determines that this change will not affect the calculation results of any technical analysis task.

[0180] After the determination is completed, the system will perform the following operations: First, it will remove the invalidation status of the downstream result nodes marked as invalid in step S710, restoring all result nodes marked as invalid to a valid state, ensuring that these results can continue to be used without waiting for recalculation. Second, it will record the exemption decision for this change in the change log, including the change time, change content, semantic tag, exemption reason ("non-critical information change, does not affect performance analysis"), and a list of affected but exempted result nodes. Third, it will send a notification message to the user interface, prompting the user that "a data change has been detected, but the change does not affect any analysis results and does not require recalculation." Fourth, it will update the version history of the data nodes. Although it does not trigger a calculation task, the change itself still needs to be recorded for complete version traceability. Through this rapid exemption mechanism, unnecessary consumption of computing resources is avoided, significantly saving computing time and resource costs. For example, when a designer only modifies the remarks information of a wall, although this text change will be captured by the system, because it carries the non-critical information change tag, all downstream results such as energy consumption analysis and solar radiation analysis will be determined as not needing to be updated, keeping the original calculation results valid.

[0181] When the task triggering unit detects a data node carrying a building envelope change tag, it initiates the relevant task triggering process for the building envelope. A building envelope change tag indicates that changes have occurred in the building's external envelope system (exterior walls, roof, exterior windows, exterior doors, etc.) that affect thermal or optical performance, such as changes in insulation material type, insulation layer thickness, window-to-wall ratio, or glass type. The task triggering unit queries the task-tag mapping rule base, which explicitly defines that building envelope changes will affect three types of core analysis tasks. The first type is solar radiation analysis tasks, because geometric changes to the building envelope alter the shading relationship and light-receiving area of ​​the building surface, affecting the calculation results of solar radiation hours. The second type is daylighting analysis tasks, as parameters such as window size, location, and glass transmittance directly determine the level of natural indoor lighting; building envelope changes significantly affect the daylighting coefficient and illuminance distribution. The third type is energy consumption simulation tasks, as the thermal performance parameters of the building envelope are key inputs for energy consumption calculations; any changes to materials or structures will alter the building's heating and cooling loads and energy consumption levels. After identifying these three types of tasks, the task triggering unit will search for all result nodes of these three types in the data dependency graph and add them to the list of tasks to be triggered. For each task to be triggered, a specific task instance will be generated, which will contain updated input data references to ensure that the latest version of the building envelope data is used during recalculation.

[0182] Task parameters are intelligently adjusted based on the specific changes. For example, if only the thickness of the external wall insulation layer changes, the energy consumption simulation task can use a fast recalculation mode, recalculating only the heat transfer portion of the building envelope while keeping other parameters such as internal heat sources and ventilation unchanged, thus shortening the calculation time. The task triggering unit also assesses the urgency of the task; tasks affecting critical project decisions are marked as high priority and allocated computing resources preferentially. Triggered tasks are submitted to the task queue, and a notification is sent to relevant users: "Changes to the building envelope detected; recalculating solar radiation analysis, daylighting analysis, and energy consumption simulation." When the task triggering unit detects a data node carrying a structural change tag, it initiates the relevant structural engineering task triggering process. A structural change tag indicates that changes have occurred in the building's load-bearing structural system that affect structural safety or load transfer, such as changes in component cross-sectional dimensions, reinforcement adjustments, changes in concrete strength grade, or component relocation.

[0183] The task triggering unit queries the rule base to determine that structural changes will affect three main types of analysis tasks. The first type is structural calculation tasks; any change in the geometric or material parameters of structural components will alter the structural performance, requiring recalculation of the structural design. The second type is load calculation tasks; changes in the dimensions or material density of structural components will alter the structure's self-weight, thus affecting dead load calculations. Adjustments to component layout may also change load transfer paths, necessitating re-analysis and recombination of loads. The third type is energy consumption calculation tasks. While structural changes primarily affect structural performance, some changes can indirectly impact energy consumption. For example, increasing floor slab thickness increases the building's thermal mass, affecting indoor temperature fluctuations and air conditioning load; changes in the material of the exterior wall structure layers alter the overall heat transfer coefficient of the walls. After identifying these three types of tasks, the task triggering unit performs task dependency analysis. Because structural calculations and load calculations are interdependent—the results of load calculations are the inputs for structural calculations—tasks are planned and executed in the order of "load calculation, structural calculation, energy consumption calculation." For structural calculation tasks, a local recalculation strategy can be adopted based on the specific location and scope of the change. For example, if only the cross-section of a single column changes, only the internal forces and reinforcement of that column and its directly connected beams and slabs can be recalculated, without requiring a global recalculation of the entire structure. Task instances are generated with specific structural parameters, such as load combinations, seismic fortification intensity, and code version, ensuring that the calculations comply with current code requirements. Because structural calculations involve building safety, triggered tasks are marked with the highest priority and subject to strict quality control procedures. After calculation, the system requires both automatic verification and manual review for double verification. The system will send a special notification to the structural engineer, detailing the content of the structural change and the scope of the recalculation, and generate a comparison report after calculation, showing the differences in structural performance indicators before and after the change, assisting the engineer in making technical decisions.

[0184] In this embodiment, the non-critical information change exemption mechanism intelligently identifies changes that do not affect performance, avoiding a large amount of invalid calculations and significantly saving computing time and cloud computing costs in a typical design iteration cycle. The building envelope change response mechanism accurately maps changes to three directly related analysis tasks: sunlight, daylighting, and energy consumption. This ensures timely updates of necessary analyses while avoiding accidental triggering of unrelated tasks, resulting in high task triggering accuracy. The structural change response mechanism not only triggers core structural calculation tasks but also considers the indirect impact of structural changes on energy consumption, demonstrating the depth of cross-disciplinary impact analysis. Furthermore, through task dependency analysis and local recalculation strategies, it maximizes computational efficiency while ensuring computational integrity. These three steps form a complete semantic tag-to-task trigger mapping system, covering the main change scenarios and analysis task types in green building design. Compared to the simple strategy in traditional solutions where any change triggers a recalculation of all tasks, the differentiated triggering mechanism of this invention significantly improves the efficiency of computing resource utilization and design iteration cycle. At the same time, through intelligent task screening and priority management, it ensures the timely execution of key tasks and data consistency, providing core technical support for the efficient collaborative design of green buildings and improving the work efficiency and design quality of project teams.

[0185] In one embodiment of this invention, after the task triggering unit triggers the recalculation of the task, the method further includes the following steps: S910, The task triggering unit will recalculate the task and add it to the task queue; The S920 and collaborative management platform call cloud computing resources and automatically execute recalculation tasks according to the order of the task queue; S930. After the recalculation task is completed, the data dependency management module generates a new result node and replaces the result node that is in an invalid state with the new result node. S940. The data dependency management module updates the data dependency graph and establishes directed dependency edges between data nodes and new result nodes. 950. The task management module will recalculate the task execution results and push them to the front-end interface.

[0186] Once the task triggering unit determines the analysis tasks that need to be recalculated based on semantic tags and data dependency graphs, these recalculation tasks need to be organized in an orderly manner for subsequent execution. The task triggering unit first creates a task object for each recalculation task. This task object contains rich metadata information, including the task's unique identifier, task type, task priority, task dependencies, input data references, estimated execution time, and task creation timestamp.

[0187] After a task object is created, the task triggering unit adds it to the task queue. The task queue is implemented using a priority queue data structure, supporting sorting by task priority and creation time to ensure that high-priority tasks receive computing resources first. For example, when the thermal parameters of the exterior wall material change, the energy consumption simulation task will be assigned a higher priority because it directly affects the building's energy consumption rating; while an acoustic analysis task triggered by fine-tuning the location of interior partition walls may be assigned a lower priority because its impact on the overall green building evaluation is relatively small.

[0188] The task queue also implements a persistent storage mechanism, which synchronizes the queue status to the database in real time. Even if the collaborative management platform restarts or fails, the task queue status can be restored from the database to ensure that no tasks are lost.

[0189] After receiving recalculation tasks from the task queue, the collaborative management platform needs to schedule and allocate cloud computing resources to execute these tasks. First, it retrieves the highest-priority task from the head of the queue, checks its dependencies, and confirms that all prerequisite tasks have been completed. If any prerequisite tasks are incomplete, the task is temporarily skipped, and the next task in the queue is processed. If all prerequisite tasks are completed, the task enters the resource allocation phase. The collaborative management platform integrates a cloud resource manager, which can dynamically monitor the usage of cloud computing resources, including indicators such as CPU core count, memory capacity, GPU availability, and network bandwidth. Depending on the task type, the resource manager employs different resource allocation strategies. For example, a solar radiation analysis task primarily relies on CPUs for ray tracing calculations, so the resource manager allocates a multi-core CPU instance; an energy consumption simulation task involves solving numerous heat conduction equations and fluid dynamics calculations, so the resource manager allocates a high-performance computing instance equipped with a GPU accelerator card; while a carbon emission calculation task mainly involves database queries and simple mathematical operations, so a standard general-purpose computing instance suffices.

[0190] After resource allocation, the collaborative management platform transmits the task's input data (relevant data nodes extracted from the data dependency graph) to the allocated computing instance via a high-speed network. It then calls the API interface of the corresponding performance simulation software or computing software in the green tool integration module to start the computing process. During task execution, the platform monitors the task's execution status in real time, including the percentage of computation progress, consumed computation time, and resource utilization. This information is pushed to the front-end interface via the WebSocket protocol, allowing users to understand the task's execution status in real time. If an exception occurs during task execution, the platform's exception handling mechanism automatically captures the error information, marks the task as failed, and decides whether to re-execute the task based on a preset retry strategy. For tasks that fail after multiple retries, the platform generates a detailed error report, including the error type, error stack information, and input data snapshot, and alerts relevant technical personnel for manual intervention via email or system notification.

[0191] Once the recompute task is completed on cloud computing resources, new analytical results data will be generated. This data needs to be correctly integrated into the management system of the data dependency management module. The data dependency management module first receives the results data returned from the cloud computing instance, which may be in various formats. The module then standardizes this results data, converting it into a unified internal data format, and calculates a hash value for each results data point as a data fingerprint for subsequent version management and change detection.

[0192] Next, the data dependency management module will create a new outcome node object. This object contains metadata information about the outcome data, such as the outcome node ID, outcome type, generation time, data version number, data size, storage path, hash value, and the task ID that generated the outcome. After the new outcome node is created, the data dependency management module needs to find the corresponding old outcome node that is in an invalid state in the data dependency graph.

[0193] Invalid outcome nodes are marked when upstream data nodes change. While these nodes still exist in the data dependency graph, their data is outdated and no longer valid. The data dependency management module replaces the old outcome node's reference with the new one, while retaining the old outcome node's historical version information and moving it to the historical version storage area. This version management mechanism allows users to review historical analysis results when needed and compare differences between versions, which is invaluable for iterative optimization of design schemes and decision analysis. For example, designers can compare energy consumption simulation results using different exterior wall materials to assess the impact of material selection on building energy performance. After the replacement, the new outcome node officially becomes an active node in the data dependency graph, and its data can be referenced by other downstream analysis tasks or downloaded and viewed by users through the front-end interface.

[0194] After a new outcome node replaces a failed one, the data dependency management module needs to update the data dependency graph, establishing directed dependency edges between the data nodes and the new outcome node to accurately reflect the current data dependencies. The data dependency graph is implemented using a directed acyclic graph (DAG) data structure. Nodes in the graph are divided into two categories: data nodes and outcome nodes, connected by directed edges. The direction of the edges indicates the direction of data flow, i.e., from input data to output outcome. Updating the data dependency graph first requires identifying the input data source of the new outcome node. The data dependency management module extracts input data reference information from the metadata of the recalculation task. These references point to one or more data nodes in the data dependency graph. For example, an energy consumption simulation task may depend on both a building information model (BIM) data node and a climate data node; therefore, directed dependency edges need to be established from these two data nodes to the new energy consumption analysis outcome node. When establishing directed dependency edges, the data dependency management module adds attribute information to each edge, including the edge's creation time, data transmission format type, data volume, and dependency strength level. For example, the dependency edge between the thermal parameter data node of building exterior wall materials and the energy consumption analysis result node will be marked as a strong dependency, because any change in thermal parameters will significantly affect the energy consumption calculation result; while the dependency edge between the data node of building interior decoration materials and the energy consumption analysis result node may be marked as a weak dependency, because the decoration materials have a relatively small impact on the overall energy consumption. The labeling of dependency strength provides an important reference for subsequent change impact analysis and task triggering strategies. The update operation of the data dependency graph needs to ensure atomicity and consistency. The data dependency management module adopts a transaction mechanism, locking relevant nodes and edges during the update process to ensure that concurrent operations do not lead to an inconsistent state in the data dependency graph. After the update is completed, the data dependency management module checks the entire data dependency graph for anomalies such as isolated nodes and circular dependencies. If an anomaly is found, an alarm will be triggered and the update operation will be rolled back.

[0195] Once the recalculation task is completed and the data dependency graph is updated, the task management module needs to promptly push the execution results to the front-end interface so that users can understand the task completion status and view the latest analysis results as soon as possible.

[0196] The task management module first constructs a result push message object, which contains several fields: task ID, task type, task status, execution time, result data summary, download link for the result data, preview image of the visualization chart, and task completion timestamp. For successfully executed tasks, the push message will also include compliance assessment information related to green building evaluation standards.

[0197] For tasks that fail, the push notification will include detailed error information and troubleshooting suggestions. The task management module uses WebSocket persistent connection technology to achieve real-time message push. When a user logs into the front-end interface through a browser or mobile app, a persistent connection is established with the collaborative management platform's WebSocket server. The task management module uses this connection channel to send result push messages to the user's client in real time. After receiving the push message, the front-end interface will display a pop-up or banner notification in the user interface's notification area, indicating that a new task has been completed. Clicking the notification will take the user to the task details page to view the complete execution results and output data.

[0198] By adding recalculation tasks to a task queue and executing them sequentially using cloud computing resources, automated scheduling of computational tasks and elastic allocation of resources are achieved. This avoids the cumbersome process of manually starting analysis software one by one, manually transferring data, and waiting for computation to complete, as required by traditional solutions, significantly improving task execution efficiency. Once the recalculation task is completed, the data dependency management module generates new result nodes and replaces failed nodes. Simultaneously, it updates the data dependency graph, establishing new directed dependency edges. This mechanism ensures that the data dependency graph always accurately reflects the current data state and dependencies, providing a reliable data foundation for subsequent change propagation and impact analysis. The task management module pushes execution results to the front-end interface in real time, allowing users to know the computation completion status and view the latest results immediately without actively polling the task status, significantly shortening the response time from data change to result update. The entire process forms a complete closed loop from task triggering, resource scheduling, computation execution, result generation, dependency updates to result push. It integrates multiple scattered links in traditional solutions that rely on manual intervention into a coherent process of automated system execution. This not only reduces the risk of human error, but more importantly, it supports the parallel execution of multiple recalculation tasks through the elastic scalability of cloud computing resources. In large and complex projects, when multiple data nodes change simultaneously, triggering a large number of downstream task updates, it can make full use of the computing power of the cloud to quickly complete all tasks. This avoids the problems of task queuing and project delays caused by insufficient computing resources in traditional solutions, and provides strong technical support for rapid iteration and agile decision-making in the full life cycle management of green buildings.

[0199] like Figure 3 As shown in the embodiments of this application, a green building full life cycle resource sharing and collaborative management system is also provided. The green building full life cycle resource sharing and collaborative management system includes a cloud server and a client. The client is used to send requests to the cloud server, and the cloud server is used to execute the operation steps of the above-mentioned green building full life cycle resource sharing and collaborative management method according to the requests.

[0200] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0201] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0202] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0203] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0204] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0205] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0206] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0207] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0208] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A green building full life cycle resource sharing and collaborative management method, characterized in that, The method is applied to a collaborative management platform, the collaborative management platform comprises a task management module, a green special topic review module, a green building knowledge base and a green tool integration module, the green tool integration module integrates basic modeling software, green design professional software, performance simulation software and calculation software, and the method comprises the following steps: In response to receiving a collaborative design request sent by a user through a front-end interface, the task management module creates a task tree and determines a task progress according to the task tree; The green tool integration module calls the basic modeling software based on the task progress to obtain building information model data; The green tool integration module calls the performance simulation software and the calculation software based on the building information model data to generate multi-type achievement data; The task management module performs quantitative and qualitative analysis on the multi-type achievement data to obtain analysis results and sends the analysis results to the green special topic review module; The green special topic review module generates evaluation clause data and evaluation version data based on the analysis results and integrates the evaluation clause data and the evaluation version data into an evaluation report for output.

2. The method of claim 1, wherein, The multi-type achievement data comprises sunshine analysis data, energy consumption analysis data and carbon emission calculation data, the green tool integration module calls the performance simulation software and the calculation software based on the building information model data to generate multi-type achievement data, which comprises the following steps: The green tool integration module parses the building information model data to extract building geometric information, material attribute information and spatial layout information and obtain current climate data; According to the building geometric information and the spatial layout information, the sunshine analysis tool in the performance simulation software is called and sunshine simulation is performed based on the current climate data to generate the sunshine analysis data; According to the building geometric information, the material attribute information and the spatial layout information, the energy consumption analysis tool in the performance simulation software is called to perform building energy consumption simulation to generate the energy consumption analysis data; According to the material attribute information, the carbon emission calculation tool in the calculation software is called to calculate the building full-life-cycle carbon emission amount to generate the carbon emission calculation data.

3. The method of claim 1, wherein, The green building knowledge base comprises a green building case library and a knowledge question and answer module, the knowledge question and answer module integrates an AI question and answer model, and the method further comprises the following steps: In response to a knowledge query request sent by a user through the front-end interface, the knowledge question and answer module extracts query keywords in the knowledge query request; The knowledge question and answer module searches for matched case data in the green building case library according to the query keywords; The knowledge question and answer module calls the AI question and answer model, takes the query keywords and the case data as inputs and generates knowledge answer content; The knowledge question and answer module returns the knowledge answer content and the case data to the user through the front-end interface.

4. The method of claim 1, wherein, The collaborative management platform further comprises a data dependency management module, the data dependency management module is used for constructing and maintaining a data dependency graph, and the method further comprises the following steps: When any of the building information model data or the multi-type result data is created or uploaded to the collaborative management platform, the data dependency management module registers the data as a data node; When a user calls a tool through the green tool integration module to operate at least one of the data nodes and generates new result data, the data dependency management module registers the new result data as a result node; The data dependency management module establishes a directed dependency edge between the input data node and the output result node, and the directed dependency edge represents a dependency relationship between the data node and the result node; The data dependency management module constructs the data dependency graph based on all the data nodes, the result nodes, and the directed dependency edges.

5. The method of claim 4, wherein, The data dependency management module further includes a change identification unit, and the method further includes: calculating a first hash value of target data corresponding to the data node; When the target data corresponding to the data node is modified by a user and a new version of data is uploaded, the change identification unit calculates a second hash value of the new version of data; The change identification unit compares the second hash value with the first hash value, and in the case that the second hash value is different from the first hash value, the change identification unit determines that the data node has changed, and triggers a semantic change analysis process.

6. The method of claim 5, wherein, The change identification unit includes a model comparison engine, and the semantic change analysis process includes: The model comparison engine parses the new version of data and the target data to extract data internal structure information; The model comparison engine compares the data internal structure information of the new version of data and the target data to identify specific change content; The model comparison engine determines a change type according to the specific change content, and the change type includes geometric change, attribute change, and non-functional change, wherein the geometric change includes component position change, component size change, and component thickness change, the attribute change includes material type change and thermal parameter change, and the non-functional change includes note information change and layer attribute change; The model comparison engine adds a semantic label to the data node according to the change type, and the semantic label includes a structural change label, a building envelope change label, a heating and ventilation parameter change label, and a non-key information change label.

7. The method of claim 6, wherein, The data dependency management module further includes a task triggering unit, and the method further includes: The task triggering unit starts from the data node that has changed in the data dependency graph, traverses downstream along the directed dependency edge, and marks all downstream result nodes as invalid; The task triggering unit obtains the semantic label of the data node; The task triggering unit determines an analysis task type affected by the change type according to the semantic label; The task triggering unit triggers a recalculation task only for the result nodes corresponding to the analysis task type, and other result nodes remain in the original state.

8. The method of claim 7, wherein, The task triggering unit determines an analysis task type affected by the change type according to the semantic label, including: In a case where the semantic label is the non-key information change label, the task triggering unit determines that the change does not affect any performance analysis, and does not trigger any recalculation task; In a case where the semantic label is the envelope change label, the task triggering unit determines that the analysis task type includes a sunshine analysis task, a daylighting analysis task and an energy consumption simulation task, and triggers corresponding recalculation tasks; In a case where the semantic label is the structural change label, the task triggering unit determines that the analysis task type includes a structure calculation task, a load calculation task and an energy consumption calculation task, and triggers corresponding recalculation tasks.

9. The method of claim 7, wherein, After the task triggering unit triggers the recalculation task, the method further includes: The task triggering unit adds the recalculation task to a task queue; The collaborative management platform calls cloud computing resources and automatically executes the recalculation task in the order of the task queue; When the recalculation task is executed, the data dependency management module generates a new achievement node and replaces the achievement node in the invalid state with the new achievement node; The data dependency management module updates the data dependency graph and establishes the directed dependency edge between the data node and the new achievement node; The task management module pushes the execution result of the recalculation task to the front-end interface.

10. A green building full life cycle resource sharing and collaborative management system, characterized in that, The green building full life cycle resource sharing and collaborative management system includes a cloud server and a client, the client is used to send a request to the cloud server, and the cloud server is used to execute the operation steps of the method of any one of claims 1-9 according to the request.

Images