A whole-process project eight-function management method and system based on node hierarchical response

Abstract

The application relates to a whole-process project eight-function management method and system based on node hierarchical response, and belongs to the technical field of engineering project management. The method comprises the following steps: establishing a whole-process main backbone line model, dividing a project into eight stages and setting key nodes, and configuring one to four response levels for each node; constructing an eight-function collaborative knowledge graph, taking project nodes, eight function dimension entities and resource entities as nodes, and taking preceding and succeeding relations and linkage influence paths as edges; monitoring node states, automatically activating succeeding nodes according to the preceding and succeeding relations and pushing resources according to response levels; when function dimension parameters change, automatically determining affected dimensions along the linkage influence paths and calculating adjustment amounts, and updating a total control table; and generating a remediation task list through missing comparison in response to different intervention points. The application realizes intelligent driving of a project, eight-dimensional collaborative linkage and missing remediation, and improves management efficiency and risk control capability.

Description

Technical Field

[0001] This invention belongs to the field of engineering project management technology, specifically relating to a full-process project management method and system based on node hierarchical response. Background Technology

[0002] Full-process project management involves complex coordination across multiple dimensions such as schedule, cost, functionality, and permitting. Traditional information systems that rely on manual experience and are fragmented are unable to achieve dynamic linkage and intelligent driving.

[0003] Existing technology CN117495615A discloses a construction engineering construction scheme management system, which realizes the formulation and execution of construction schemes through BIM models and construction stage module division, but only covers the construction stage and lacks full life cycle coverage and node hierarchical response mechanisms. Existing technology CN118797070A discloses a project process management method that applies knowledge graphs to software development projects for task performance evaluation, but it does not design linkage influence paths for the eight management dimensions of engineering projects and lacks intervention point omission checking functions. Existing technology CN119850155B discloses a project management method that uses knowledge graphs for task decomposition and dynamic resource allocation, but its graph traversal is mainly used for task dependencies rather than parameter influence propagation between different functional dimensions, and it does not address the omission problem during project intervention. Existing technology CN120851807A discloses a project visualization full-process management method combined with knowledge graphs, focusing on data visualization and user interaction management, but lacks automatic linkage reasoning based on predefined rules and intervention point remediation mechanisms.

[0004] In summary, there is an urgent need for an intelligent project collaborative management method and system that can simultaneously solve technical problems such as hierarchical driving of nodes throughout the entire process, automatic linkage propagation of eight-dimensional parameter changes, and inspection and remediation of omissions at different intervention points. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for managing eight aspects of a project throughout its entire process based on node-level hierarchical response, in order to solve the aforementioned problems existing in the prior art. The specific technical solution is as follows: A full-process project management method based on node-level hierarchical response includes the following steps: S101. Establish a main keel line model for the whole process management of the project. The main keel line model divides the whole process of the project into eight consecutive stages. Multiple key nodes are laid in each stage. Node attribute information is configured for each key node. The node attribute information includes node identifier, work content, deliverable requirements, response level, work group composition and planned duration. The predecessor and successor relationships between nodes are stored as a directed acyclic graph. S102. Construct an eight-function collaborative knowledge graph. The knowledge graph is constructed based on the main backbone model. Its entity types include project node entities, functional dimension entities, and resource entities. The functional dimension entities include project management dimension, cost dimension, design dimension, planning and construction permit dimension, special report dimension, bidding and procurement dimension, operation dimension, and audit dimension. The relational edges of the knowledge graph include predecessor and successor relational edges defined according to the directed acyclic graph, as well as predefined linkage and influence path edges. The linkage and influence path edges include source entity parameters, target entity parameters, influence function, direction, and priority. S103. Real-time monitoring and node-driven operation during project execution: The status of each key node is monitored by a status listener. When the status of a preceding node changes to "completed", the successor node is obtained according to the preceding-successor relationship. The successor node is checked to see if all its preceding nodes are completed. If so, the status of the successor node is automatically updated to "in progress". Resource configuration notifications are matched and pushed according to the response level of the successor node. S104. Linkage adjustment of eight functions: When the control parameters of any functional dimension entity are detected to change, a breadth-first traversal is performed along the linkage influence path edge in the knowledge graph starting from that entity. The affected target entity and its parameter change are calculated according to the influence function, a linkage adjustment scheme is generated, and the corresponding project master control table is updated in response to the user's confirmation operation of the linkage adjustment scheme. S105. Based on the intervention point, omission detection and remediation, in response to the intervention instruction of the project in a non-initial stage, extract all nodes from the main keel line model from the first stage to the stage before the intervention point as a standard node list, match the standard node list with the actual completed node results, calculate the difference set to obtain the omission node set, create a remediation task for each omission node in the omission node set, and add the remediation task to the project work breakdown structure.

[0006] As a preferred embodiment of this application, in step S103, the real-time monitoring and node driving further includes: when the response level of the successor node is level three or four, automatically reading the meeting configuration file associated with the node, automatically creating an internal meeting calendar event and sending an invitation according to the internal meeting template, automatically creating an external meeting calendar event and sending an invitation after the internal meeting resolution is passed according to the external meeting template, and requiring the minutes of the external meeting and the attached signed scanned copy as mandatory deliverables of the node.

[0007] As a preferred embodiment of this application, in step S103, the response level includes four levels: Level 1, Level 2, Level 3, and Level 4. Among them, Level 1 response corresponds to a resource allocation strategy that does not require a special meeting and is completed by the executor on its own; Level 2 response corresponds to a resource allocation strategy that is reviewed by an internal project team meeting or online; Level 3 response corresponds to a resource allocation strategy that requires an internal cross-departmental meeting and approval by the department head; and Level 4 response corresponds to a resource allocation strategy that requires a company-level review meeting and is jointly signed by the company's technical head and financial head.

[0008] As a preferred embodiment of this application, in step S104, during the breadth-first traversal process, a maximum traversal depth is set for each traversal task, a queue is used to manage the entities to be traversed, and the visited set is used to record the combination of entity parameters that have been processed. When an entity taken out of the queue has a linkage influence path edge pointing to the target entity, the influence function of that edge is executed to calculate the change, and the target entity with a non-zero calculation result is added to the queue as a new task.

[0009] As a preferred embodiment of this application, in step S105, the specific method for matching the standard node list with the actual completed node results is as follows: First, direct matching is performed through the metadata tags of the result file. For unmatched nodes, the keyword vector required by the standard node results is extracted and the cosine similarity is calculated with the keyword vector of the result file name and the header text. When the similarity exceeds a preset threshold, it is determined to be a match.

[0010] As a preferred embodiment of this application, step S105, after creating a remedial task for the missing node, further includes: using the critical path method to obtain the available window of the current project master plan, searching for a continuous time period in the available window as a suggested start and end time based on the estimated duration of each remedial task, and if the available window is insufficient to accommodate all remedial tasks, prioritizing higher-level nodes according to the response level of the missing nodes.

[0011] This application also provides a full-process project eight-function management system based on node hierarchical response, including: The main keel line modeling module is used to establish a main keel line model for the whole process management of the project. The main keel line model divides the whole process of the project into eight consecutive stages. In each stage, multiple key nodes are laid out. Node attribute information is configured for each key node, and the predecessor and successor relationships between nodes are stored as a directed acyclic graph. The knowledge graph construction and management module is used to construct an eight-function collaborative knowledge graph based on the main backbone model. The entity types of the knowledge graph include project node entities, functional dimension entities, and resource entities, and the relationship edges include predecessor and successor relationship edges and linkage influence path edges. The project status monitoring and driving module is used to monitor the completion status of each key node, automatically activate the successor node according to the predecessor and successor relationship, and push resource configuration notifications according to the response level of the successor node. The eight-function linkage analysis module is used to traverse the knowledge graph along the linkage influence path starting from the entity when the control parameters of any functional dimension entity are detected to change. It calculates the affected target entity and its parameter changes according to the influence function, generates a linkage adjustment plan, and updates the corresponding project master control table in response to user confirmation. The omission detection and remediation module is used to respond to intervention instructions in the non-initial stage of the project, extract the standard node list and match it with the actual completed node results, calculate the difference set to obtain the set of missing nodes, and create a remediation task for each missing node.

[0012] As a preferred embodiment of this application, the project status monitoring and driving module includes a meeting driving engine. The meeting driving engine is used to read the meeting configuration file associated with the node, automatically create an internal meeting based on the internal meeting template and send an invitation, automatically create an external meeting based on the external meeting template after the internal meeting resolution is passed, and verify the signed scanned copy of the external meeting minutes as a mandatory condition for the node to complete.

[0013] As a preferred embodiment of this application, the eight-function linkage analysis module includes a parameter change capture unit, a graph traversal and inference engine, a linkage scheme generator, and a master control table update executor; wherein, the graph traversal and inference engine uses a queue to manage entities to be traversed, uses visited sets to prevent duplicate processing, and executes the influence function script through a sandbox environment.

[0014] As a preferred embodiment of this application, the defect detection and remediation module includes an actual result matching unit and a remediation plan scheduling unit; the actual result matching unit determines the completed nodes by combining direct matching of metadata tags and supplementary matching of cosine similarity; the remediation plan scheduling unit calls the critical path method to obtain the main plan's idle window, allocates suggested start and end times for the remediation tasks, and sorts and handles resource conflicts according to the response level of the defective nodes.

[0015] The beneficial effects of this invention are as follows: 1) Achieve automated project advancement driven by node hierarchy. This invention configures each key node in the main backbone with a response level of one to four and clear deliverable requirements. Combined with status monitoring and an automatic activation mechanism for subsequent nodes, it pushes resource configurations and meeting templates differently according to the importance of the nodes. This achieves fully automated closed-loop management from node completion to the initiation of subsequent tasks, significantly reducing manual coordination costs and avoiding omissions and delays of key nodes.

[0016] 2) Achieving intelligent linkage and collaboration across eight dimensions. This invention constructs a knowledge graph encompassing eight dimensions: progress, cost, functional configuration, planning and approval, special reports, bidding and procurement, operation, and auditing. It also predefines the linkage impact paths and executable impact functions. When any dimension parameter changes, the system automatically propagates the change along the impact path using a breadth-first graph traversal algorithm, calculates the affected dimensions and their adjustment amounts, and updates the project master control table accordingly. This achieves automatic reasoning and dynamic closed-loop adjustment between multiple dimensions, fundamentally solving the information silo problem caused by fragmented management.

[0017] 3) Enables automatic inspection and remediation of omissions under flexible intervention. This invention automatically compares the standard node list with the actual completed results by setting intervention points, uses a similarity matching algorithm to identify missing nodes, generates a remedial task list, and automatically inserts it into the main plan's idle window. This allows project managers to quickly complete project checks and omission remediation at any stage of intervention, ensuring the integrity and continuity of the project's main framework. It is especially suitable for scenarios where consulting firms take over midway or supplementary commissions are made later.

[0018] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Of course, implementing any product or method of this application does not necessarily require achieving all of the advantages described above at the same time. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is the overall flowchart of the eight-function collaborative management method for the entire project process provided in this embodiment of the invention.

[0021] Figure 2 This is a flowchart of the main keel line model for establishing the entire project management process provided in this embodiment of the invention.

[0022] Figure 3 This is a flowchart of constructing an eight-function collaborative knowledge graph provided in an embodiment of the present invention.

[0023] Figure 4 This is a flowchart of real-time monitoring and node-driven operation provided in an embodiment of the present invention.

[0024] Figure 5 This is a flowchart of the linkage adjustment of the eight functions provided in the embodiment of the present invention.

[0025] Figure 6 This is a flowchart of the defect detection and remediation provided in the embodiments of the present invention.

[0026] Figure 7 This is an architecture diagram of the full-process project eight-function collaborative management system provided in this embodiment of the invention. Detailed Implementation

[0027] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Example 1 This invention provides a method for managing eight aspects of a project throughout its entire process based on node-level hierarchical response. For example... Figure 1 As shown, the specific steps include: S101. Establish the main framework model for project-wide management. This step aims to build a structured data model for the entire project management process. This model defines all key stages, critical nodes, and their logical relationships and attributes from project initiation to completion, serving as the data foundation and execution framework for subsequent processes. For example... Figure 2 As shown, the specific implementation is as follows: First, based on the project type selected by the user, the system retrieves the corresponding standardized phase division scheme from a template library pre-installed in a relational database. The template library stores phase division data for various project types, including residential buildings, municipal roads, industrial plants, schools, and hospitals. If the built-in templates do not meet the requirements, the system receives user-defined phase division commands through a graphical user interface, supporting operations such as adding, deleting, and modifying phases.

[0029] The project is divided into eight consecutive phases in chronological order: preliminary planning, scheme design and project approval, preliminary design and budget estimate, EPC (Engineering, Procurement, and Construction) contracting bidding, permitting, project construction, acceptance, and trial operation. Each phase has a phase identifier, phase name, planned start and end dates, and phase manager information, which are stored in a phase information table in a relational database.

[0030] Within each phase, the system further lays out multiple key nodes, and the set of key nodes across all phases constitutes the main backbone of the project. The number of key nodes can vary depending on the type of project. The arrangement of nodes follows the inherent logical dependencies of project progress; that is, the completion of one node is a necessary condition for the initiation of one or more subsequent nodes. The system stores the predecessor-successor relationships between nodes using a directed acyclic graph (DAG) data structure, where each node is a vertex in the graph, and the dependencies between nodes are directed edges. For example, the node "Complete Preliminary Design" points to the node "Preliminary Budget Preparation," indicating that "Preliminary Budget Preparation" can only begin after "Complete Preliminary Design" is completed. The system simultaneously builds this DAG when creating nodes and stores the node relationships in the node relationship table of the project database.

[0031] For each key node, the system configures the following attribute information through the node attribute configuration interface. Node attribute information includes at least: node identifier, node name, work content, deliverables, response level, working group composition, and planned duration. The node identifier is a globally unique string automatically generated by the system; the work content describes the specific work items and deliverables to be completed in a structured text list format; the deliverables describe the list of deliverables to be delivered upon completion of the node, and each deliverable can be associated with a pre-defined file template or digital verification rule. The verification rules include checking whether required fields are empty, checking whether files are uploaded to the specified path, and checking whether the electronic signature field contains a valid certificate.

[0032] The response levels are divided into four levels: Level 1, Level 2, Level 3, and Level 4. Each level corresponds to a different level of importance and resource allocation strategy. Level 1 response represents general or routine nodes, with resource allocation strategies including: no dedicated meeting required; completion and submission of deliverables by the executor; the system only records and tracks routinely. Level 2 response represents routine nodes, handled internally by the project team, with resource allocation strategies including: internal project team meetings or online reviews; approval by the project deputy manager or professional leader. Level 3 response represents important project nodes, involving cross-departmental collaboration or confirmation of important deliverables, with resource allocation strategies including: mandatory internal cross-departmental meetings; approval from department heads; and automatic system notifications to the project manager and department heads. Level 4 response represents the highest-level critical node, involving major decisions or milestones, with resource allocation strategies including: mandatory company-level review meetings; mandatory dual signatures from the company's technical and financial leaders; and automatic system notifications to the terminals of senior management. The working group composition defines the leading and key supporting roles required to complete each node. At project initiation, the system matches specific personnel from the project's human resources database and associates them with the node, simultaneously assigning corresponding operating permissions. The planned duration is the planned completion time for this node, measured in working days.

[0033] After completing the above configuration, the system serializes the main keel line model in JSON format and stores it in the model definition table of the relational database. At the same time, it calls the graphics rendering engine to generate a visual project progress Gantt chart and displays it on the front-end interface.

[0034] S102, Constructing an Eight-Function Collaborative Knowledge Graph Building upon the main framework established in S101, this step constructs a knowledge graph to support collaborative decision-making across eight functions. For example... Figure 3 As shown, the specific implementation is as follows: First, the system predefines three entity types. The first type is project node entities, corresponding to each key node in S101, with attributes including node identifier, node name, response level, planned duration, and actual completion time. The second type is functional dimension entities, corresponding to eight functional dimensions requiring collaborative management, specifically: project management, cost, design, permitting and approval, special reports, bidding and procurement, operations, and auditing. Each functional dimension entity is primarily managed by a corresponding professional role; for example, project management corresponds to the project manager, cost to the cost engineer, design to the design manager, permitting and approval to the permitting specialist, special reports to the special report specialist, bidding and procurement to the bidding specialist, operations to the operations specialist, and auditing to the audit specialist. Each functional dimension entity has a dimension identifier, dimension name, and current set of control parameters. For example, the cost dimension's control parameter set includes the total investment ceiling, sub-item cost indicators, and contingency rate; the project management dimension's control parameter set includes the critical path duration and milestone times for each stage. The third category is resource entities, which correspond to various resources involved in the project, including human resources, material resources, and financial resources. Their attributes include resource identifier, resource name, resource type, total amount of resources, allocated amount, and unit cost.

[0035] Next, the system predefines two types of relationship edges. The first type is the predecessor-successor relationship edge, which connects two project node entities, with the direction from the predecessor node to the successor node. This edge expresses the execution order dependency between nodes defined in the main backbone. The attributes of this relationship edge include dependency type and time lag. Dependency types include "end-start" (i.e., the successor node can only start after the predecessor node has completely finished) and "start-start" (i.e., the successor node can start after the predecessor node has started for a certain period). The second type is the linkage influence path edge, which connects different functional dimension entities, or connects functional dimension entities with project node entities or resource entities. This edge expresses the impact and calculation method of a change in the parameters of one entity on another entity. Each linkage influence path includes the following attributes: source entity type and parameters, target entity type and parameters, influence function, direction, and priority. The source entity type and parameters specify the entity that triggers the change and the specific parameter name. The influence function is an executable rule fragment written in a domain-specific language that calculates the change in the target parameter based on the change in the source parameter. For example, an impact function can be expressed as: "If the increase in the total investment ceiling exceeds 5%, the decoration standard level in the design dimension will be downgraded by one level, and the design change filing fee in the planning and construction dimension will be increased by 5,000 yuan." Direction indicates whether the impact is unidirectional or bidirectional, and priority is represented by integers; the smaller the value, the higher the priority. When multiple linkage paths point to the same target parameter, the path with the higher priority is used for calculation.

[0036] Based on the main keel line model established by S101, the system automatically creates all project node entities and creates corresponding predecessor-successor relationship edges according to the predecessor-successor relationships in the directed acyclic graph, storing them in the graph database. Next, the system retrieves the corresponding eight-function linkage template from the historical project knowledge base according to the project type. The linkage template is stored in XML file format, containing typical linkage impact paths summarized from historical practice for this type of project. The system parses the template file, instantiating each linkage impact path as an edge in the graph database, and simultaneously creating functional dimension entities involved in the path.

[0037] For example, for construction projects, typical linkage impact paths of system instantiation include, but are not limited to: Path 1: The source entity is "Cost Dimension Sub-item Project Cost Indicator," and the target entity is "Function Dimension Decoration Standard Level." The influence function is defined as follows: "When the cost indicator of a sub-item project decreases by 10% or more, the decoration standard level automatically decreases by one level; when the cost indicator of a sub-item project decreases by 15% or more, the decoration standard level automatically decreases by two levels; when the cost indicator of a sub-item project increases by 10% or more, the decoration standard level automatically increases by one level." The direction is unidirectional, and the priority is level one.

[0038] Path Two: The source entity is "Design Scheme Change Identifier (Design Dimension)," and the target entity is "Planning Permit Change Trigger Condition (Planning Permit Change Process) (Planning and Construction Permit Dimension)." The impact function is defined as: "If the design scheme change results in an increase or decrease in building area exceeding 2%, the planning permit re-approval flag will be set to true, and an additional 15 working days of approval time will be added." The direction is unidirectional, and the priority is level two.

[0039] Path 3: The source entity is "Delay Days of Key Nodes in Project Management Dimension," and the target entity is "Planned Release Date of Procurement Items in the Bidding and Procurement Dimension." The impact function is defined as: "If the delay days of a key node exceed 5 working days, the release date of all bidding items related to that node will be postponed by the same number of days, and the procurement urgency score for each bidding item will be recalculated." The direction is unidirectional, and the priority is level two.

[0040] All entities and relation edges are stored in a graph database, forming a queryable and traversable eight-function collaborative knowledge graph.

[0041] S103. Real-time monitoring and node-driven operation during project execution. This step aims to achieve automated project progress based on node status and meeting-driven management of key nodes. For example... Figure 4 As shown, the specific implementation is as follows: The system sets up a status listener on each project node entity. Node status includes four types: "Not Started," "In Progress," "Completed," and "Overdue." Status changes can be manually triggered by the user uploading deliverables through the front-end interface, or automatically triggered by the system through checking the digitized fields in the deliverable requirements. The automatic detection is implemented as follows: the system periodically scans the storage path of deliverables associated with the node through a scheduled task. For nodes that require uploading specific deliverable files, the system checks whether a file conforming to the naming rules exists in the specified path and reads the file metadata or specific flag bits using a file parsing tool; if a completion flag is detected, the node status is automatically changed to "Completed."

[0042] When the system detects a predecessor node's status change to "Completed," it immediately activates the successor node. The system first queries the knowledge graph, retrieving all direct successor nodes of the predecessor node through predecessor-successor relationship edges. Then, for each direct successor node, the system iterates through all predecessor-successor relationship edges pointing to that successor node in the knowledge graph, checking if the status of all its predecessor nodes is "Completed." If the check passes, the system automatically updates the successor node's status from "Not Started" to "In Progress" through a database transaction.

[0043] After the status update, the system reads the "Response Level" attribute of the successor node. If the response level is Level 1 or Level 2, the system simply adds the corresponding item to the to-do list in the front-end workbench and highlights it with a light background, without sending any additional instant messages. If the response level is Level 3 or Level 4, the system automatically generates a project task assignment notification and sends it to the leading role and key collaborating roles defined in the node's "Working Group Composition" via the integrated email and instant messaging service interfaces. The notification includes the node's work content, deliverables, planned timeline, and a preset meeting template.

[0044] The system also monitors the time taken for nodes in the "In Progress" state in real time via a timer thread. This timer thread scans all nodes in the "In Progress" state every 24 hours. For each node, the system calculates the ratio of actual time taken to the planned duration. When the actual time taken reaches 80% of the planned duration, the system generates a yellow warning; when the actual time taken reaches or exceeds 100% of the planned duration, the system generates a red warning. The scope of the warning message is determined based on the node's response level. For Level 1 and Level 2 nodes, the warning message is only sent to the node manager and project manager; for Level 3 and Level 4 nodes, the warning message is sent to the node manager, project manager, and company management simultaneously.

[0045] For critical nodes with a response level of three or four, the system incorporates a meeting-driven management mechanism. When a node's status changes to "in progress," the system automatically reads the meeting configuration file associated with that node, stored in JSON format. This configuration file defines at least internal and external meeting templates. Internal meeting templates include the meeting objective, a list of internal participants, a list of pre-meeting preparation materials, and a template for meeting resolutions. External meeting templates include the meeting objective, a list of external participating units and personnel, a list of materials to be distributed before the meeting, and templates for meeting resolutions and post-meeting action plans.

[0046] Based on the internal meeting template, the system automatically creates an event in the calendar service, schedules it for the morning of the third working day after the node is activated, and sends meeting invitations to the specific personnel corresponding to the internal participant roles. Simultaneously, the system creates a "Pre-meeting Preparation Materials" folder in the project document library and pushes a task to the person responsible for material preparation, requiring them to upload the materials 24 hours before the meeting. During the meeting, the recorder fills in the meeting minutes and uploads the resolution document in the system's meeting management interface. The system checks whether the resolution document contains essential elements such as review conclusions, a list of modification opinions, and voting results. If the verification passes, the system sets the internal meeting completion flag for that node to true. If the internal meeting resolution is "not passed," the system automatically reverts the node status to the initial "in progress" state and requires the preparation of materials to be repeated before the meeting is reconvened.

[0047] When an internal meeting resolution is "Passed" or "Passed after Modification" and the modifications are completed, the system automatically triggers an external meeting. Based on the external meeting template, the system generates a formal meeting notification and sends it to external units via email and scanned paper copies. The system ensures that pre-meeting materials have been distributed three working days prior to the meeting, and the distribution record is automatically archived. Meeting minutes generated from external meetings must include scanned copies or electronic signatures of authorized representatives from all parties; this document is a mandatory deliverable in the node requirements. The system sets a verification rule in the node completion conditions: if a fully signed external meeting minute is not uploaded, the node cannot be marked as "Completed." All materials from the two meetings are associated with this node entity and stored in the knowledge graph as node attributes. The system continuously tracks each task in the action plan, breaking down each task into sub-tasks and associating them with this node. Only when the status of all action plan sub-tasks changes to "Completed" is the node finally marked as "Completed."

[0048] S104, Coordination and Adjustment of Eight Functions This step aims to solve the information silo problem caused by traditional compartmentalized management. When the control parameters of any functional entity change, the system automatically performs linked reasoning and adjustment, and demonstrates the complete application of eight-function collaboration using quota design as an example. Figure 5 As shown, the specific implementation is as follows: The system sets change hooks for each key parameter of each functional entity. Specifically, the system provides a unified parameter update interface for each functional entity, and all parameter modification operations must be performed through this interface. Internally, before updating the database, the interface reads the current parameter value and compares it with the new value in the request. If the two are not equal, it records the values ​​before and after the change, the time of the change, and the operator's account, and encapsulates this information into a "parameter change event," which is then pushed to the system's internal event bus.

[0049] The system employs a linkage analysis engine that subscribes to parameter change events on an event bus. Upon receiving a parameter change event, the engine starts with the entity that has changed and performs a breadth-first traversal algorithm in the knowledge graph to find all reachable linkage influence paths originating from that entity. Specifically, the engine first creates a queue, adding the starting entity and the change amount as a task. Simultaneously, it creates an accessed set to record processed entity and parameter combinations, preventing duplicate processing. Then, the engine loops, retrieving a task from the head of the queue and using the entity in that task as the current node. It then queries the knowledge graph for all linkage influence paths originating from that node. For each such edge, the engine reads its influence function and calculates the change amount of the target entity's corresponding parameter based on the change amount of the source parameter. During the calculation, the engine executes the rules in the influence function using a sandbox environment. The sandbox environment restricts access to the file system and network, ensuring security. If the calculated change amount is non-zero, the engine combines the target entity, its required adjusted parameters, and the calculated change amount into a new task and adds it to the tail of the queue. Simultaneously, the engine stores this propagation record in a temporary propagation list. Before adding a new task to the queue, the engine checks the visited collection. If a record with the same entity and parameter combination already exists in the collection, the task is skipped. Traversal continues until the queue is empty. The system sets the maximum traversal depth to five levels, and users can customize this depth value in the configuration interface according to the project complexity.

[0050] For example, the "total investment ceiling" in the cost dimension increases from 100 million yuan to 110 million yuan, a change of 10 million yuan. Starting from the "cost dimension" entity, the engine finds the first linked impact path and calculates, based on the impact function, that the "decoration standard level" in the functional dimension should be upgraded from level B to level A. The engine adds the "decoration standard level" in the functional dimension and the change of one level as a new task to the queue. Simultaneously, it finds the second linked impact path; the impact function determines that the increased investment amount has exceeded the threshold, therefore increasing the "application fee" by 30,000 yuan. The engine adds the "application fee" in the application and approval dimension and the change of 30,000 yuan to the queue. Next, the engine retrieves the "functional dimension" task from the queue and continues traversing downwards, finding a path where the impact function stipulates that when the decoration standard level is upgraded, the estimated amount of the high-end decoration bidding item needs to increase by 2 million yuan. The engine adds the "estimated amount of the high-end decoration bidding item" in the bidding and procurement dimension and the change of 2 million yuan to the queue. The traversal continues until no new tasks are generated or the maximum depth is reached.

[0051] After traversing all propagation records, the engine organizes them into a "linkage adjustment plan," presenting it to the user in a structured list format. Each item in the list includes: the name of the affected functional dimension, the original parameter value, the suggested new value, and the basis for the adjustment. The system presents this plan as an interactive decision table on the user interface, providing three operation buttons: "Accept," "Reject," and "Manually Modify." Users can independently select each adjustment item. If the user clicks "Accept," the system calls the update interface of the corresponding master control table management service to write the new value of the adjustment item into the corresponding field of the master control table in the database. If the user clicks "Manually Modify," the system pops up an input box, allowing the user to enter a custom new value, which is then written to the master control table. If the user clicks "Reject," the system retains the original value but records the user's rejection operation and its reason in the log.

[0052] The project master control table includes the following tables: Project Overall Plan Table, Functional Master Control Table, Cost Master Control Table, Fund Utilization Plan Table, Planning and Construction Permit Coordination Table, and Bidding and Procurement Coordination Table. When a user accepts an adjustment, the system automatically updates one or more corresponding master control tables and recalculates the relevant summary data.

[0053] Taking the quota-based design as an example, when the system executes the quota-based design process, it first performs functional coordination, collects information on construction land and planning conditions, calculates the project construction scale based on economic indicators, collects the functional positioning of similar projects and their corresponding functional requirements and proportions, inputs the requirements confirmed by the owner, divides the project into unit projects, and generates a functional master control table. Then, it performs overall cost control. The system retrieves cost data and cost indicators for similar projects from the historical project database, matches the corresponding unit project cost indicators and total cost indicators based on the division of unit projects and the functional master control table, and determines the maximum investment ceiling for the project. Based on the project characteristics, functions, and total cost indicators, it determines the items for other construction costs and calculates their specific amounts, generating a cost master control table. Next, it performs project progress planning, planning and construction permit coordination, special report coordination, bidding and procurement coordination, operation coordination, and audit review. When any parameter changes, the system automatically triggers the aforementioned linked adjustment traversal, propagating the change to other relevant dimensions and updating the corresponding master control table accordingly.

[0054] S105. Missing Detection and Remediation Based on Intervention Timing This step aims to address the difficulty in quickly tracing and supplementing preliminary work when intervention occurs at different stages of a project. For example... Figure 6 As shown, the specific implementation is as follows: When a project manager creates a project instance in the system for the first time, the system prompts them to set the "Intervention Point" via a drop-down selection box. The possible values ​​for the intervention point are one of eight stages throughout the project lifecycle, such as "Intervention during the EPC bidding stage" or "Intervention during the preliminary design stage." This setting is stored in the intervention point field of the project configuration file.

[0055] The system performs a missing node check. First, it extracts all nodes from the main keel line model, from the first stage to the stage preceding the intervention point, and records the list of these node identifiers as the "standard node list." For example, if the intervention point is the "Engineering General Contracting Bidding Stage," the standard node list includes all nodes from the three stages: the preliminary planning stage, the scheme design and project approval stage, and the preliminary design and budget stage. Second, the system extracts a list of completed node deliverables from the project document library and deliverable database. The project document library is stored in a distributed file system, and each deliverable file is associated with one or more node identifier metadata tags. The system obtains a "set of completed node identifiers" by scanning these metadata tags. To improve matching accuracy, the system also uses natural language processing technology to supplement the analysis of deliverable file names and content. Specifically, for each standard node, the system extracts keywords from its deliverable requirement fields to form a set of keyword vectors. Simultaneously, the system performs the same keyword extraction on the name and header text of each deliverable file. Then, it calculates the cosine similarity between the two vectors; deliverables with a similarity greater than a preset threshold are considered to match the node. The matched node identifiers are also added to the "Completed Node Identifier Set". Finally, the system calculates the difference between the standard node list and the completed node identifier set to obtain the "Missing Node Set".

[0056] For each missing node in the missing node set, the system generates a remedial task. The system reads the work content, deliverables, and working group composition of the node. Then, the system automatically creates a "remedial task" and adds it to the work breakdown structure of the current project. The data structure of the remedial task is the same as that of a regular node task, but with two additional flag fields: one is a "whether it is a remedial task" flag set to true; the other is an "original node identifier" field that records the original identifier of the missing node. The system sets the priority of the remedial task to "high" and writes a suggested deadline to its attributes, which is set to 50% of the current stage milestone time.

[0057] For missing nodes with a response level of three or four, the system requires users to fill in a missing node explanation before starting remediation tasks. The explanation should include: the cause of the missing node, the responsible party, and whether any cost changes are involved. This explanation, after confirmation by the project manager, is archived and used as a basis for project auditing. All remediation tasks are summarized into a "Remediation Task List," displayed on the front end in the form of a Gantt chart or task list.

[0058] The system also provides one-click generation of "make-up plans," automatically assigning estimated durations to each make-up task, read from the planned duration field of the original node. The system first calls the critical path method calculation module to obtain the available window of the current master plan, i.e., the time period with a total float greater than zero. Then, for each make-up task, it finds a sufficiently long continuous time period within the available window and uses this time period as the suggested start and end dates for the make-up task. If the available window is insufficient to accommodate all make-up tasks, the system sorts them by priority: level 4 node make-up tasks take precedence over level 3, level 3 over level 2, and level 2 over level 1, prioritizing high-priority tasks and issuing a resource conflict warning to the project manager. The project manager can manually adjust the start and end times of make-up tasks on the system interface.

[0059] Remedial tasks are assigned to personnel with appropriate roles via the Work Breakdown Structure (WBS) interface. After logging into the system, the executor can see entries marked "Remedial Task" in their to-do list. The executor completes the work according to the node's work content and deliverable requirements and uploads the deliverables. The system also provides overdue warnings and completion status checks for remedial tasks. If, during the remedial process, it is found that missing deliverables from a node will affect the parameters of subsequent nodes, the system triggers S104 for coordinated adjustments.

[0060] Example 2 This invention also provides a full-process project eight-function management system based on node hierarchical response, such as... Figure 7 As shown, it specifically includes: Main keel line modeling module The main keel line modeling module is responsible for the creation, storage, version management, and import / export of the main keel line model. Internally, it includes a template management unit, a modeling editor unit, a node attribute configuration unit, and a version control unit. The template management unit stores default phase divisions, node preset libraries, and response level templates for different project types; data is stored in a template table within a relational database. Administrators can add or modify templates through the graphical interface provided by the template management unit; modifications support locking and approval processes. The modeling editor unit provides a graphical user interface, allowing users to create phases, add nodes, and define predecessor and successor relationships between nodes via drag-and-drop. Internally, the modeling editor unit implements a verification algorithm based on a directed acyclic graph; if a user attempts to create a relationship that leads to a circular dependency, the editor will display a warning and prohibit the operation. After editing, the modeling editor unit serializes the model into a lightweight data interchange format text and stores it in the main keel line table of the relational database. The node attribute configuration unit is used to configure detailed attribute information for each node, providing a set of input controls: a work content input box, a deliverable requirement template selector, a response level dropdown, a work team composition role selector, and a planned duration numeric input box. The deliverables template library includes pre-stored commonly used templates such as "Review Minutes Template," "Budget Estimate Template," and "Construction Drawing Template," which users can directly reference. The version control unit generates a version number for each modification to the main keel line model, consisting of a major version number, a minor version number, and a revision number. The version control unit records the time of each modification, the operator, and a summary of the changes. Users can use the version control unit to revert to any historical version and compare the differences between two versions.

[0061] Knowledge Graph Construction and Management Module The Knowledge Graph Construction and Management module is used to build, store, update, and query eight-function collaborative knowledge graphs, implemented based on a graph database. Internally, this module includes entity definition, relation definition, data import and instantiation, graph maintenance and query interfaces, and graph visualization. The entity definition unit defines the tag and attribute architecture of project node entities, functional dimension entities, and resource entities, storing this definition in an ontology file that conforms to the resource description framework specification. The relation definition unit defines the type, attributes, and constraint rules of predecessor-successor relationship edges and linkage influence path edges. The influence functions in the linkage influence path are stored as scripts in the database's script field. These scripts are written in a restricted domain-specific language that only supports arithmetic operations, logical comparisons, and conditional branches, but not loops or external calls, to ensure execution safety. The data import and instantiation unit interfaces with the main backbone modeling module and the project database. It automatically creates project node entities and predecessor-successor relationship edges by reading data from the main backbone table, and simultaneously reads the linkage template files associated with project types to instantiate functional dimension entities and linkage influence path edges. The instantiation process utilizes the graph database's batch insert interface, inserting 500 records per batch to improve efficiency. The graph maintenance and query interface provides an external application programming interface (API). The query interface supports encapsulating the graph database's query language into secure calls; upper-layer modules only need to provide entity identifiers and query depth, and the interface automatically generates the query statement and returns the results. The graph visualization unit presents the knowledge graph as an interactive network graph on the front end, requesting graph data from the back end and using an open-source graphics library to render nodes and edges. Nodes display different colors and shapes based on their type, and edges display different line types and colors based on their type. Users can zoom, drag, and click on nodes to view details within this unit.

[0062] Project Status Monitoring and Driving Module The project status monitoring and driving module is responsible for real-time monitoring of node status and automatically advancing the project process, while also implementing meeting-driven management. Internally, this module includes a status monitoring unit, a successor activation engine, a resource push unit, an overdue calculation and early warning unit, and a meeting-driven engine. The status monitoring unit uses a database trigger mechanism, creating update triggers on the node status field of the main backbone table. When the status field value changes, the trigger automatically writes the node identifier, old status, new status, and change time to a status change log table. The status monitoring unit polls this log table every five seconds, pushing newly generated events to the internal processing pipeline. When the successor activation engine receives a "completed" event from a preceding node, it calls the knowledge graph query interface to obtain a list of successor nodes. For each successor node, it checks the status of all its preceding nodes by querying the node relationship table. If the conditions are met, it changes the successor node's status to "in progress" by calling the update interface of the main backbone module. The resource push unit invokes different notification channels based on the response level of subsequent nodes: for level 1 and 2 nodes, only the front-end message push interface is called to display to-do reminders on the user's webpage; for level 3 and 4 nodes, both email service interfaces and instant messaging service interfaces are called to send notifications. The notification content template is stored in a configuration file, and variables are replaced during push notifications. The overdue calculation and early warning unit sets up a scheduled task that runs every morning at midnight, scanning all nodes with a status of "in progress" to calculate the time consumption ratio. For nodes that reach the early warning threshold, an early warning record is generated and stored in the early warning information table, and the resource push unit is called to send the early warning.

[0063] The meeting-driven engine is used to implement meeting management functions for first- and second-level nodes in the method embodiment. The engine reads the meeting configuration file associated with each node, automatically creates calendar events based on internal meeting templates, and sends meeting invitations via email. The engine monitors the completeness of pre-meeting preparation materials, checking whether files in the materials folder have been uploaded and meet template requirements. After meeting minutes are uploaded, the engine parses the review conclusion field in the resolution file. If the conclusion is "not passed," the engine calls the status rollback interface to reset the node status to "in progress." If the conclusion is "passed" or "passed after modification," the engine advances the node status to the waiting stage for external meetings. When an external meeting is triggered, the engine generates a meeting notification and records a distribution log. After external meeting minutes are uploaded, the engine verifies the existence and validity of the signed scanned copy. If the verification passes, the node is marked as "meeting completed." The engine also continuously tracks the completion status of action plan subtasks. By polling the status table of associated subtasks, a node is only allowed to be finally completed when all subtask statuses are "completed."

[0064] Eight-function linkage analysis module The eight-function linkage analysis module is the core of the system's collaborative intelligence. Internally, it includes a parameter change capture unit, a graph traversal and inference engine, a linkage solution generator, a decision support interface, and a central control table update executor. The parameter change capture unit provides a unified parameter update interface for the key control parameters of each functional dimension entity. Internally, the interface compares the new value with the current value; if they differ, it constructs a parameter change event object and publishes it to the internal event bus, finally executing a database update operation. The graph traversal and inference engine subscribes to the internal event bus, creating a new traversal task for each event and executing it according to the breadth-first traversal algorithm described in the method embodiment. The engine uses a queue data structure to manage nodes to be traversed and a hash set to record the parameter combinations of visited entities. For each linkage influence path edge, the engine executes its influence function through a script engine, which adopts a sandbox mode to limit script execution time and memory usage. After traversal, the linkage solution generator collects all propagation records into a solution object. The solution object contains a list of influence chains, each recording the complete path from the source entity to the target entity and the change amount at each step. The solution generator converts this object into a front-end renderable format. The decision support interface sends the linkage solution object to the front end and pushes real-time notifications via WebSocket protocol. After the front end displays the adjustment solution, the user's selection is returned to the linkage analysis module through this interface. The interface performs format and permission verification on the returned data. When the user confirms acceptance of an adjustment item, the master control table update executor calls the application interface of the corresponding master control table management service, passing in the adjustment item identifier and the new value. This interface internally performs data verification, permission verification, and transactional updates. The update operation itself may trigger parameter change capture again; the system avoids infinite loops by setting an anti-recursion flag in an update transaction context.

[0065] Missing item detection and remediation module The omission detection and remediation module includes an intervention timing configuration unit, a standard node extraction unit, an actual result matching unit, a difference set calculation and task generation unit, and a remediation plan scheduling unit. The intervention timing configuration unit provides a drop-down selection box on the project initialization interface, listing eight stages sequentially. After the user selects a stage, this unit stores the selected stage identifier in the intervention timing field of the project configuration table. The standard node extraction unit, based on the value of the intervention timing field, queries the main backbone table for all nodes with stage numbers less than the stage number of the intervention timing. The query results are sorted by stage number and node number to form a standard node list. The actual result matching unit scans the project document library, first querying the directly associated node identifiers in the file metadata table to obtain a direct matching set. Then, for nodes in the standard node list that are not yet matched, keyword extraction and cosine similarity calculation are used for supplementary matching. Keyword extraction employs word segmentation and stop word filtering techniques. The difference calculation and task generation unit calculates the difference between the standard node list and the set of matched node identifiers. For each node identifier in the difference set, it calls the "Create Remedial Task" interface of the Work Breakdown Structure Management Service, passing in the original node identifier. This service inserts a record into the task table, marking it as a remedial task. The remedial plan scheduling unit inserts the generated remedial task into the current project master plan. First, it calls the critical path method calculation module to obtain the free window of the current master plan. Then, for each remedial task, it finds a sufficiently long continuous time period in the free window based on its estimated duration, and uses this time period as the suggested start and end time of the remedial task. If the free window is insufficient to accommodate all remedial tasks, this unit sorts them by task priority, prioritizing high-priority tasks, and issues a resource conflict warning to the project manager.

[0066] Interaction and Display Module The Interaction and Display module provides a front-end graphical user interface, serving as the entry point for user interaction with the system. Internally, this module includes a project overview display unit, a knowledge graph visualization unit, a master control table management interface, a task dashboard unit, an early warning and message center, and a meeting management interface. The project overview display unit presents the overall project status in a "one-three-four panoramic view" format: a leading area displays the project name, project objectives, and project manager; three boundary areas display the summary data of the cost master control table, the current progress of the planning and construction permit coordination table, and the compliance score of the audit review module; four collaborative areas display the completion status of the special report coordination module, the access status of the operation coordination module, the latest results of the design module, and the ongoing items of the bidding and procurement coordination module. This display unit refreshes data every ten seconds by periodically polling the backend interface. The knowledge graph visualization unit integrates an open-source graph library, requesting knowledge graph data from the backend in node and edge lists. This unit assigns different graphic symbols based on node type: project nodes are displayed as rectangles, functional dimension entities as hexagons, and resource entities as circles; it also assigns different connection styles based on edge type: predecessor-successor relationships are displayed as solid lines with arrows, and linked influence paths are displayed as dashed lines with arrows. Users can click on nodes in this unit to view detailed attributes, drag nodes to adjust the layout, and use sliders to zoom the view. The master control table management interface displays various master control tables in the form of editable tables. Each table provides in-row editing functionality; double-clicking a cell enters edit mode, and pressing Enter submits the changes. Upon submission, the front end calls the data update interface of the corresponding master control table service and displays a success or failure message. The task dashboard unit displays the currently logged-in user's to-do task list. Task cards display the task name, node identifier, response level, and planned completion time. Task cards with a response level of level 1 have a gray background, level 2 is white, level 3 is light yellow, and level 4 is light red. Each card provides "Start Work," "Upload Results," and "Apply for Completion" buttons. The Alerts and Messages Center displays a bell icon at the top of the interface, with a badge indicating the number of unread messages. Clicking the icon expands the message list, which includes overdue alerts, linkage adjustment suggestions, and meeting notifications. Users can click on each message to jump to the corresponding details page. The Meeting Management interface is used to create meetings, upload meeting minutes, and record resolutions and action plans. This interface is linked to the meeting-driven engine; when a user opens the meeting management interface for a node, the system automatically loads the meeting configuration file associated with that node and pre-fills some fields based on whether the meeting is internal or external. The meeting minutes editing area supports rich text editing and direct image pasting. Resolutions and action plans are entered in tabular format, and each action plan can be associated with a node identifier or a subtask.

[0067] User and Permission Management Module The User and Permission Management module manages user permissions based on a role-based access control mechanism. Predefined roles include: Super Administrator, Project Manager, Cost Engineer, Design Manager, Permit Specialist, Bidding Specialist, Operations and Maintenance Specialist, Special Project Specialist, and Audit Specialist. Each role has a set of permission identifiers, such as "Edit Cost Control Table," "Approve First-Level Node Completion," and "View All Projects." The User and Permission Management module provides interfaces for user registration, login, role assignment, and permission query.

[0068] Document Management Module The document management module provides file upload, download, version management, and optical character recognition (OCR) text extraction functions. When uploading a file, the system stores it in a distributed file system and records the file identifier, original filename, storage path, uploader, upload time, and associated node identifier in the database file table. The version management function allows users to upload new versions that overwrite older versions, and the system automatically retains the three most recent versions. The OCR function extracts text from scanned documents, providing text content for the missing text check module.

[0069] Message notification module The message notification module encapsulates sending interfaces for three notification channels: email, instant messaging, and web push. External callers only need to provide the recipient identifier, message title, message content, and channel type; the message notification module automatically handles channel adaptation and sending. For email, the module uses the Simple Mail Transfer Protocol (SMT) service; for instant messaging, it calls the open interfaces of WeChat Work or DingTalk; and for web push, it pushes messages to the user's currently open page via a web socket connection.

[0070] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

Claims

1. A method for managing eight aspects of a project throughout its entire process based on node-level hierarchical response, characterized in that: Includes the following steps: S101. Establish a main keel line model for the whole process management of the project. The main keel line model divides the whole process of the project into eight consecutive stages. Multiple key nodes are laid in each stage. Node attribute information is configured for each key node. The node attribute information includes node identifier, work content, deliverable requirements, response level, work group composition and planned duration. The predecessor and successor relationships between nodes are stored as a directed acyclic graph. S102. Construct an eight-function collaborative knowledge graph. The knowledge graph is constructed based on the main backbone model. Its entity types include project node entities, functional dimension entities, and resource entities. The functional dimension entities include project management dimension, cost dimension, design dimension, planning and construction permit dimension, special report dimension, bidding and procurement dimension, operation dimension, and audit dimension. The relational edges of the knowledge graph include predecessor and successor relational edges defined according to the directed acyclic graph, as well as predefined linkage and influence path edges. The linkage and influence path edges include source entity parameters, target entity parameters, influence function, direction, and priority. S103. Real-time monitoring and node-driven operation during project execution: The status of each key node is monitored by a status listener. When the status of a preceding node changes to "completed", the successor node is obtained according to the preceding-successor relationship. The successor node is checked to see if all its preceding nodes are completed. If so, the status of the successor node is automatically updated to "in progress". Resource configuration notifications are matched and pushed according to the response level of the successor node. S104. Linkage adjustment of eight functions: When the control parameters of any functional dimension entity are detected to change, a breadth-first traversal is performed along the linkage influence path edge in the knowledge graph starting from that entity. The affected target entity and its parameter change are calculated according to the influence function, a linkage adjustment scheme is generated, and the corresponding project master control table is updated in response to the user's confirmation operation of the linkage adjustment scheme. S105. Based on the intervention point, omission detection and remediation, in response to the intervention instruction of the project in a non-initial stage, extract all nodes from the main keel line model from the first stage to the stage before the intervention point as a standard node list, match the standard node list with the actual completed node results, calculate the difference set to obtain the omission node set, create a remediation task for each omission node in the omission node set, and add the remediation task to the project work breakdown structure.

2. The method according to claim 1, characterized in that, In step S103, the real-time monitoring and node driving further includes: when the response level of the successor node is level three or four, automatically reading the meeting configuration file associated with the node, automatically creating an internal meeting calendar event and sending an invitation according to the internal meeting template, automatically creating an external meeting calendar event and sending an invitation after the internal meeting resolution is passed according to the external meeting template, and requiring the minutes of the external meeting and the attached signed scanned copy as mandatory deliverables of the node.

3. The method according to claim 2, characterized in that, The response levels are divided into four levels: Level 1, Level 2, Level 3, and Level 4. Level 1 response corresponds to resource allocation strategies that do not require a special meeting and are completed by the executor. Level 2 response corresponds to resource allocation strategies that are reviewed by the project team through internal meetings or online. Level 3 response corresponds to resource allocation strategies that require an internal cross-departmental meeting and approval by the department head. Level 4 response corresponds to resource allocation strategies that require a company-level review meeting and are jointly signed by the company's technical and financial leaders.

4. The method according to claim 1, characterized in that, In step S104, during the breadth-first traversal process, a maximum traversal depth is set for each traversal task, a queue is used to manage the entities to be traversed, and the visited set is used to record the combination of entity parameters that have been processed. When an entity taken out of the queue has a linkage influence path edge pointing to the target entity, the influence function of that edge is executed to calculate the change, and the target entity with a non-zero calculation result is added to the queue as a new task.

5. The method according to claim 1, characterized in that, In step S105, the specific method for matching the standard node list with the actual completed node results is as follows: First, direct matching is performed through the metadata tags of the result file. For unmatched nodes, the keyword vector required by the standard node results is extracted and the cosine similarity is calculated with the keyword vector of the result file name and the header text. When the similarity exceeds the preset threshold, it is determined to be a match.

6. The method according to claim 1, characterized in that, In step S105, after creating remedial tasks for the missing nodes, the method is further used to obtain the available window of the current project master plan by calling the critical path method. Based on the estimated duration of each remedial task, a continuous time period is searched in the available window as a suggested start and end time. If the available window is insufficient to accommodate all remedial tasks, the nodes with higher response levels are prioritized according to the response level of the missing nodes.

7. A full-process project eight-function management system based on node hierarchical response, wherein the system applies the method described in any one of claims 1 to 6, characterized in that, include: The main keel line modeling module is used to establish a main keel line model for the whole process management of the project. The main keel line model divides the whole process of the project into eight consecutive stages. In each stage, multiple key nodes are laid out. Node attribute information is configured for each key node, and the predecessor and successor relationships between nodes are stored as a directed acyclic graph. The knowledge graph construction and management module is used to construct an eight-function collaborative knowledge graph based on the main backbone model. The entity types of the knowledge graph include project node entities, functional dimension entities, and resource entities, and the relationship edges include predecessor and successor relationship edges and linkage influence path edges. The project status monitoring and driving module is used to monitor the completion status of each key node, automatically activate the successor node according to the predecessor and successor relationship, and push resource configuration notifications according to the response level of the successor node. The eight-function linkage analysis module is used to traverse the knowledge graph along the linkage influence path starting from the entity when the control parameters of any functional dimension entity are detected to change. It calculates the affected target entity and its parameter changes according to the influence function, generates a linkage adjustment plan, and updates the corresponding project master control table in response to user confirmation. The omission detection and remediation module is used to respond to intervention instructions in the non-initial stage of the project, extract the standard node list and match it with the actual completed node results, calculate the difference set to obtain the set of missing nodes, and create a remediation task for each missing node.

8. The system according to claim 7, characterized in that, The project status monitoring and driving module includes a meeting driving engine. The meeting driving engine is used to read the meeting configuration file associated with the node, automatically create an internal meeting based on the internal meeting template and send an invitation. After the internal meeting resolution is passed, it automatically creates an external meeting based on the external meeting template and verifies the signed scanned copy of the external meeting minutes as a mandatory condition for the node to complete.

9. The system according to claim 7, characterized in that, The eight-function linkage analysis module includes a parameter change capture unit, a graph traversal and inference engine, a linkage scheme generator, and a master control table update executor; wherein, the graph traversal and inference engine uses a queue to manage entities to be traversed, uses visited sets to prevent duplicate processing, and executes the impact function script through a sandbox environment.

10. The system according to claim 7, characterized in that, The defect detection and remediation module includes an actual result matching unit and a remediation plan scheduling unit. The actual result matching unit determines the completed nodes by combining direct matching of metadata tags and supplementary matching of cosine similarity. The remediation plan scheduling unit calls the critical path method to obtain the main plan's idle window, allocates suggested start and end times for the remediation tasks, and sorts and handles resource conflicts according to the response level of the defective nodes.

Images