A method and system for managing the design drafting stage of a ship piping

Abstract

The present application relates to the technical field of intelligent management of ship manufacturing, and particularly relates to a method and system for ship pipeline design drawing stage management, which acquires and structurally processes production schedule and process specification data, constructs a standardized input set, uses a multi-objective optimization engine to resolve conflicts, optimize progress and cost, and generates a preliminary pipeline scheme in combination with a four-dimensional model, semantic verification and dynamic scheduling, and outputs construction drawings and animations after three-dimensional simulation verification, collects feedback in real time during construction, triggers secondary scheduling and optimization, realizes continuous iteration and accurate execution of design, and through intelligent decision-making of the multi-objective optimization engine, the design drawing efficiency can be improved by more than 40%, the design error rate is reduced by 90%, and the rework in the later stage is greatly reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent management of shipbuilding, in particular to a method and system for managing the piping design stage of a ship, which is especially suitable for the fine design, intelligent decision-making and collaborative management of the piping system in the process of building large ships. BACKGROUND

[0002] The piping system of a ship is like a network of blood vessels, and its design quality directly affects the efficiency of shipbuilding and the safety of operation. The traditional management of piping design and drawing mainly relies on the experience of designers for manual coordination, which has the following technical problems:

[0003] Firstly, there is a serious disconnection between design and construction. The design department often lacks a full understanding of the actual situation of the construction site when making a drawing plan, resulting in frequent problems such as spatial conflicts and process contradictions in actual construction, with a high rate of rework.

[0004] Secondly, the multi-specialty collaboration is inefficient. Ship piping involves multiple specialties such as structure, machinery, and electrical, and there is a lack of effective information sharing and decision-making collaboration mechanism among specialties, often resulting in repeated design and resource conflicts.

[0005] Thirdly, the change management response is lagging. When problems are found in the construction site and design changes are needed, the traditional management method is slow to respond, unable to quickly assess the impact of changes and adjust subsequent plans, seriously affecting project progress.

[0006] Finally, the material management is inefficient. The division and distribution of piping materials lack scientific planning, often resulting in material overstock or shortage, increasing storage costs and construction waiting time.

[0007] With the development trend of large-scale and complex ships, the limitations of traditional management methods are becoming increasingly apparent, and it is urgent to introduce advanced technologies such as artificial intelligence to achieve intelligent management of the piping design and drawing stage. SUMMARY

[0008] The purpose of the present application is to overcome the shortcomings of the prior art and provide a method and system for managing the piping design and drawing stage of a ship, which builds an intelligent decision-making system based on artificial intelligence to achieve deep collaboration and dynamic optimization of multiple links such as design, production, and materials.

[0009] The application adopts four-dimensional space-time collision detection technology, considers not only static space interference, but also time dimension into the collision prediction range, and can identify dynamic conflicts in the construction process in advance. Through process specification detection based on semantic understanding, abstract text specifications are automatically converted into calculable constraint parameters, greatly improving the accuracy and efficiency of specification checking. The cognitive digital twin technology is introduced to realize deep perception and understanding of the construction site state, and support intelligent scheduling decisions based on causal reasoning.

[0010] The application discloses a kind of ship pipeline design drawing stage management method, comprising:

[0011] Obtain the schedule data in production and the specification parameters in process specification database, the data obtained are structured and the data timestamp mechanism is established, to form the standardized input data set;

[0012] Based on the input data set formed, through multi-objective optimization engine, the comprehensive trade-off analysis of construction conflict minimization, progress deviation minimization and material cost optimization is carried out, while executing collision detection based on four-dimensional space-time model, process specification compliance verification based on semantic understanding, dynamic scheduling optimization considering real-time deviation and tray division planning based on site constraints, to generate preliminary pipeline design scheme including construction stage division, drawing time plan and material list;

[0013] According to the generated preliminary pipeline design scheme, three-dimensional simulation verification and feasibility evaluation are carried out, potential design conflicts and construction risks are identified, and the optimized scheme after verification is formed;

[0014] Based on the optimized scheme after verification, two-dimensional drawings and construction process animation videos with construction stage marks are generated, and executable construction guidance files are output;

[0015] In the construction execution process, collect site feedback information and design change requirements, when the change influence exceeds the preset threshold, automatically trigger secondary scheduling, update drawing plan and re-execute the above analysis and optimization process, realize the iterative optimization of design scheme.

[0016] As preferred, the comprehensive trade-off analysis of the multi-objective optimization engine specifically includes:

[0017] Divide the construction space into three-dimensional grid units and mark the occupancy time window of each grid unit, calculate the space conflict severity index by identifying the case that multiple pipelines occupy the same space at the same time, analyze the equipment sharing conflict and process parallel feasibility between different pipeline construction processes, and quantify the influence degree of process conflict on overall progress;

[0018] Construction of construction network diagram and calculation of the earliest start time and the latest completion time of each activity, identification of the critical activity sequence with zero total float time to form the critical path, analysis of the construction uncertainty of various pipelines based on historical data and setting of time buffer for high-risk activities;

[0019] Analysis of the demand distribution of the same specification pipe and calculation of the unit cost under different batch sizes, comprehensive evaluation of the capital occupation cost of advance procurement and the limitation of on-site storage space, determination of the optimal material procurement batch size and timing;

[0020] Dynamic adjustment of the weight coefficients of the three optimization objectives of construction conflict, progress deviation and material cost according to the project stage, generation of multiple candidate schemes through iterative optimization and calculation of the comprehensive score of each scheme, selection of the scheme with the best comprehensive performance as the output result.

[0021] As preferred, the collision detection based on the four-dimensional space-time model specifically includes:

[0022] Construction of a four-dimensional space-time occupation tensor containing spatial three-dimensional coordinates and time dimension, wherein the first three dimensions represent the position coordinates of the pipeline in space, and the fourth dimension represents the construction time sequence;

[0023] For each pipeline section, calculate its space-time occupation trajectory in the entire construction period, including the lifting path, the space occupied during installation and the space required for welding operation;

[0024] Identify the overlapping area of different pipelines in four-dimensional space-time through tensor operation, introduce a probability distribution model to consider the uncertainty factors in the construction process, generate multiple possible construction scenarios through Monte Carlo simulation and calculate the probability distribution of collision occurrence;

[0025] When the collision probability exceeds the preset threshold, generate a collision resolution strategy based on deep reinforcement learning, including a spatial avoidance scheme for adjusting the pipeline path and a time avoidance scheme for adjusting the construction time sequence, evaluate the impact of different resolution strategies on the overall construction progress and cost, and select the optimal scheme.

[0026] As preferred, the process specification compliance verification based on semantic understanding specifically includes:

[0027] Deep semantic analysis of the process specification text using a pre-trained language model, extraction of specific control elements in the process specification text, including: constrained pipeline components (constraint objects), technical requirements that should be followed (constraint conditions), specifically defined numerical ranges (constraint values), and applicable ship areas or pipeline types (applicable range) Core semantic elements such as key semantic elements, generate high-dimensional semantic vector representation of the specification;

[0028] A standard semantic space is constructed and the semantic similarity between different standards is calculated to automatically identify the association, conflict and complementary relationship between standards, and the mapping relationship between standard semantics and design parameters is established.

[0029] Abstract standard requirements are automatically converted into specific ship pipeline design parameters, including quantifiable process indicators such as the minimum bend radius value required at the pipeline bend, the maximum flow rate variation percentage allowed when the fluid flows in the pipeline, the minimum and maximum spacing range between adjacent pipeline supports, and the required bevel angle value requirement at the welding site.

[0030] A standard dependency graph is constructed and potential standard conflict patterns are identified through a graph neural network. The standards are divided into hard constraints and soft constraints, and a multi-objective optimization algorithm is used to find a design scheme that maximizes the overall constraint satisfaction. For conflicts that cannot be completely eliminated, detailed conflict reports and resolution suggestions are generated.

[0031] As a preferred embodiment, the dynamic scheduling optimization considering real-time deviations specifically includes:

[0032] Real-time acquisition of actual drawing completion time, design change occurrence time and content, and on-site feedback problems and delay information for each pipeline, calculation of the time deviation between actual progress and planned progress, and analysis of the development trend of the deviation;

[0033] A cognitive digital twin system is constructed to integrate various sources of on-site data, including environmental and equipment parameter data collected by Internet of Things sensors installed on the construction site, construction footage data captured by on-site video monitoring systems, quality and progress feedback information submitted by construction workers through mobile terminals, and formal progress report data in the project management system. Various types of data are fused to form a unified cognitive state representation through a deep learning model.

[0034] A causal reasoning network is used to analyze the causal relationships in the construction process and predict the causal effects of different scheduling decisions.

[0035] Based on the current progress deviation, resource constraints and downstream process dependency relationships, a priority scoring model is established and the drawing priority of each pipeline system is dynamically adjusted to ensure that tasks on the critical path are completed first.

[0036] As a preferred embodiment, the tray partitioning planning based on site constraints specifically includes:

[0037] The spatial layout constraints of the construction site are analyzed, including the available storage space size, the width and height limitations of the transportation channel, the rated load of the lifting equipment, and the loading capacity of the transportation vehicle.

[0038] A digital twin model of the material flow is constructed to simulate the whole process of material flow from the warehouse to the construction site through discrete event simulation technology, and potential bottlenecks and conflicts under different tray division schemes are identified;

[0039] Pipelines of the same specification and similar construction areas are grouped for similarity, ensuring that the weight of each tray does not exceed the load limit of the equipment and the size meets the transportation constraints, and a three-dimensional packing algorithm is used to optimize the space utilization of the tray;

[0040] The resilience of the tray scheme under abnormal conditions, including transportation equipment failure and temporary unavailability of storage areas, is evaluated through Monte Carlo simulation, and a tray division scheme that is both efficient and resilient is selected, and a unique identification code containing area, batch and content information is generated for each tray.

[0041] As a preferred embodiment, the collection of field feedback information and design change requirements and the automatic triggering of secondary scheduling specifically includes:

[0042] Identify the source of the change and classify it, including active changes caused by design optimization, passive changes from field problem feedback, and mandatory changes caused by specification updates, and divide it into first-level changes affecting multiple systems, second-level changes within a single system, and third-level changes not affecting the interface according to the impact range of the change;

[0043] Analyze the direct impact of the change, including changes to the upstream and downstream connections of the pipeline, related drawings that need to be modified simultaneously, and the impact on completed construction parts, while evaluating the indirect impact, including the chain changes in space occupation, the impact on other specialties, and the impact on the overall progress of the project;

[0044] Set a change impact threshold, and when the cumulative impact exceeds the threshold, automatically trigger the rescheduling process, update the status of the affected tasks and re-run the multi-objective optimization engine to generate a revised drawing plan;

[0045] A new version number is generated for each change and the complete data before and after the change is saved, supporting version comparison and rollback, and the corresponding approval process is determined according to the change level, and the approval opinions and decision basis are recorded to ensure the traceability of the change.

[0046] As a preferred embodiment, the generation of two-dimensional drawings with construction stage identifiers and construction process animation videos specifically includes:

[0047] Key information for guiding field construction is extracted from the three-dimensional design model of the ship pipeline, including: specific location coordinates where the pipeline needs to be segmented, types and parameters of welding points used at each connection (such as butt welding, corner welding, etc.), precise installation location coordinates and types of pipeline supports, and other key technical information required by construction personnel, construction sequence numbers are generated according to the optimization results and the dependency relationship between processes is marked;

[0048] Automatic addition of segmented numbers, dimensioning, welding symbols, quality requirements, material specifications and quantity information on two-dimensional drawings, avoidance of overlapping and occlusion through layout optimization algorithms, and management of different types of annotation information using layers;

[0049] Construction of a three-dimensional scene containing ship structure models, piping models and construction equipment models, setting of animation rhythm according to actual construction time and determination of key frame positions, simulation of complete processes of piping hoisting, welding operations and inspection testing;

[0050] Use of color changes in animations to represent construction status, addition of arrow indicators to indicate installation direction, setting of translucent effects to show internal structures, superimposition of current step instructions, progress timelines and safety warning signs, and generation of multi-format output files supporting pause control, chapter jumping and field annotation functions.

[0051] As a preferred, it also includes:

[0052] Adopting a distributed collaboration framework based on blockchain, taking the design department, production department and material department as blockchain network nodes, verifying and synchronizing data through consensus mechanisms, and automatically triggering cross-department collaboration processes using smart contracts;

[0053] Building a distributed intelligent optimization mechanism based on federated learning, realizing multi-party collaborative training without sharing raw data, and protecting data privacy through differential privacy and secure multi-party computing technology;

[0054] Establishing a spiral upward mechanism among data, knowledge and intelligence, learning design patterns from historical project data to form a domain knowledge base, using knowledge to guide intelligent decision-making for current projects, and generating new data from decision-making results to update the knowledge base through continuous learning;

[0055] Setting up a hierarchical optimization architecture to balance local optimization and global coordination, with bottom-level modules performing real-time local optimization, middle-level modules performing intra-domain coordination, and top-level modules performing periodic global strategic optimization, and controlling the allocation of optimization resources to each module through an optimization budget mechanism.

[0056] A system for ship piping design drawing stage management, comprising:

[0057] A data acquisition and preprocessing module for obtaining production schedule data and specification parameters in the process specification database, structurally processing the obtained data and establishing a data timestamp mechanism to form a standardized input data set;

[0058] An intelligent analysis and decision module, including a multi-objective optimization engine, a collision detection unit, a process specification detection unit, a dynamic scheduling unit and a tray division unit, is used for comprehensive trade-off analysis of construction conflict minimization, schedule deviation minimization and material cost optimization based on the input data set, performs collision detection based on a four-dimensional space-time model, process specification compliance verification based on semantic understanding, dynamic scheduling optimization considering real-time deviation, and tray division planning based on field constraints, and generates a preliminary pipeline design scheme including construction phase division, drawing time plan and material list;

[0059] A scheme generation and verification module is used for three-dimensional simulation verification and feasibility evaluation according to the preliminary pipeline design scheme, identifies potential design conflicts and construction risks, and forms an optimized scheme after verification;

[0060] A construction guidance output module is used for generating two-dimensional drawings and construction process animation videos with construction phase identification based on the optimized scheme after verification, and outputting executable construction guidance files;

[0061] An execution feedback and iterative optimization module is used for collecting field feedback information and design change requirements during construction execution, triggering secondary scheduling automatically when the change impact exceeds a preset threshold, updating the drawing plan and calling the intelligent analysis and decision module to re-execute the comprehensive trade-off analysis of construction conflict minimization, schedule deviation minimization and material cost optimization, collision detection based on a four-dimensional space-time model, process specification compliance verification based on semantic understanding, dynamic scheduling optimization considering real-time deviation, and tray division planning based on field constraints, to realize iterative optimization of the design scheme.

[0062] Compared with the prior art, the present application has the following beneficial effects:

[0063] First, the design efficiency is significantly improved. Through intelligent decision of the multi-objective optimization engine, the design drawing efficiency can be improved by more than 40%, and the design error rate is reduced by 90%, greatly reducing the rework in the later stage.

[0064] Second, deep collaborative management is realized. The distributed collaborative framework based on blockchain ensures the consistency of data and the synchronization of decisions of each department, eliminates information silos, and improves the overall operation efficiency by 15%-20%.

[0065] Third, the dynamic response capability is enhanced. The system can sense the schedule deviation in real time and automatically trigger scheduling optimization, shortening the change response time from several days to hours, ensuring that the project is completed on schedule.

[0066] Fourth, resource allocation is optimized. Through intelligent tray division and material flow simulation, the material utilization rate is improved by 20%, the warehouse cost is reduced by 15%, and the construction waiting time is effectively reduced.

[0067] Fifth, improve the decision-making scientificity. The decision mechanism based on the knowledge graph and the causal reasoning makes the management decision more scientific and reasonable, and the overall project duration can be shortened by 20%-30%. BRIEF DESCRIPTION OF DRAWINGS

[0068] Figure 1 The overall architecture diagram of the ship pipeline design drafting stage management system provided by the embodiment of the present application is shown.

[0069] Figure 2 The workflow diagram of the multi-objective optimization engine in the embodiment of the present application is shown.

[0070] Figure 3 The principle diagram of the four-dimensional space-time collision detection in the embodiment of the present application is shown.

[0071] Figure 4 The processing flowchart of the change management in the embodiment of the present application is shown.

[0072] Figure 5 The structure diagram of the distributed collaborative framework in the embodiment of the present application is shown. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application will be further described in detail below with reference to the drawings.

[0074] Referring to Figure 1 The present application provides a kind of ship pipeline design drafting stage management method, which realizes the whole process optimization from data acquisition to construction execution by constructing intelligent management system.

[0075] Firstly, the system obtains the schedule data in production and the specification parameters in process specification database. In an embodiment of the present application, the schedule data in production is obtained through real-time interface with shipyard ERP system, including the current progress percentage of each construction area, the personnel allocation situation of each type of work, the equipment resource occupation state, etc. For example, the system automatically synchronizes data once an hour, and obtains specific information such as A area pipeline installation progress 75%, 8 welders, crane occupation rate 60%. The process specification data is extracted from the pre-constructed structured database, including technical standards, material requirements, installation specifications, etc. of various types of pipelines.

[0076] Preferably, in the data preprocessing stage, the system cleans and standardizes the acquired raw data. Specifically, the system establishes a unified data model, mapping data from different sources to a standard format. For example, abstract specification requirements are automatically converted into specific ship pipeline design parameters, including quantifiable process indicators such as the minimum bend radius value required for pipeline bends (e.g., minimum bend radius ≥ 3D, D being the pipe diameter), the maximum flow velocity variation percentage allowed for fluid flow in the pipeline, the minimum and maximum spacing range between adjacent pipeline supports, and the required bevel angle value requirement for welding. These quantitative indicators directly affect design decisions, for example, the system automatically ensures that the bend radius is greater than or equal to the minimum allowed value when designing pipeline bends. At the same time, the system establishes a timestamp mechanism for each data item, recording the acquisition time and validity period of the data, ensuring that decisions are based on the latest information.

[0077] Based on the formed standardized input data set, the system conducts comprehensive analysis through a multi-objective optimization engine. This is one of the core innovations of the invention, which simultaneously considers the three objectives of minimizing construction conflicts, minimizing schedule deviations, and optimizing material costs. In terms of minimizing construction conflicts, the system divides the construction space into three-dimensional grid cells, with each grid cell having a size of 0.5m x 0.5m x 0.5m. This precision accurately reflects the pipeline occupation without causing excessive computational burden. The system labels each grid with an occupation time window, for example, grid (10, 20, 15) is occupied by DN100 pipeline from day 5 to day 7.

[0078] In one embodiment of the invention, the calculation of the spatial conflict severity index uses the following formula:

[0079] ,

[0080] wherein, is the spatial conflict severity index; is the total number of pipelines; is the pipeline and the weight coefficient between pipeline , with a value range of [0, 1], determined according to the importance of the pipeline; is the spatial overlap degree, representing the overlapping volume proportion of two pipelines in space; is the time overlap degree, representing the overlap proportion of the construction time of two pipelines. The realization of minimizing schedule deviations is based on the Critical Path Method (CPM). The system first constructs a complete construction network diagram, calculating the earliest start time (ES), latest start time (LS), earliest finish time (EF), and latest finish time (LF) of each activity. The total float time TF calculation formula is:

[0081] ,

[0082] When TF=0, the activity is on the critical path. Preferably, the system also introduces uncertainty analysis based on historical data to set time buffer for high-risk activities. For example, according to historical data analysis, the welding operation of large-diameter pipelines (DN≥200) has a large standard deviation of completion time, and the system automatically increases a time buffer of 15% for such activities.

[0083] Material cost optimization is achieved through an improved version of the Economic Order Quantity (EOQ) model. Considering the particularity of shipbuilding, the cost function adopted by the system is:

[0084] ,

[0085] where, is the total cost; is the procurement cost, including material unit price and bulk discount; is the storage cost, calculated daily, taking into account capital occupation and site fees; is the shortage cost, i.e. the loss of construction delay due to insufficient materials. In one embodiment of the invention, the system dynamically adjusts the weights of the three objectives according to the project phase. In the early stage of the project (0%-30% progress), the progress weight is set to 0.5, the conflict weight is 0.3, and the cost weight is 0.2, focusing on ensuring the smooth start of the project; In the middle stage of the project (30-70% progress), the three weights are balanced at 0.33; In the later stage of the project (70%-100% progress), the conflict weight is increased to 0.5 to ensure the construction quality in the final stage.

[0086] The collision detection of the invention uses an innovative four-dimensional space-time model. Referring to Figure 3 , the model constructs a space-time occupancy tensor (STOT), which is a four-order tensor , where , , is the spatial dimension, is the time dimension. In specific implementation, each element of the tensor represents the occupancy probability of spatial position at time . Preferably, the system does not use binary occupancy representation, but uses continuous probability values in the range [0, 1]. This design takes into account the uncertainty of the construction process, for example, the hoisting of pipelines may have time deviation due to weather, equipment, etc.

[0087] For each pipe segment, the system calculates its complete spatiotemporal occupancy trajectory. Take a DN150, 6-meter-long straight pipe as an example, the spatiotemporal occupancy calculation of its hoisting process includes: the spatial occupancy of the hoisting point (considering the size of the hoist), the swept space of the hoisting path (considering the pipe swing), and the final occupancy of the installation position (considering the welding operation space). The system generates 1000 possible construction scenarios using Monte Carlo simulation, each scenario considering different random factors, such as hoisting speed deviation (±10%), positioning error (±50mm), etc.

[0088] The collision probability is calculated using the following formula:

[0089]

[0090] is the collision probability of pipe and pipe ; and and are the spatiotemporal occupancy tensors of the two pipes; the integration range is the entire construction period. When the collision probability exceeds the preset threshold (0.15 in this embodiment, which is determined based on shipyard safety management requirements), the system triggers the collision resolution mechanism. The generation of resolution strategies uses a deep reinforcement learning method, specifically the deep deterministic policy gradient (DDPG) algorithm. The state space contains 26 features such as the current pipe position, construction progress, and available resources; the action space includes 8 action dimensions such as path adjustment (continuous value, range ±2 meters) and delay time (discrete value, 0-7 days). The reward function is designed as:

[0091]

[0092] is the immediate reward; is the number of days of progress delay; is the additional cost (ten thousand yuan); is the number of other pipes affected; are weight coefficients, which are determined through regression analysis of historical project data.

[0093] The process specification detection module of the present invention uses deep semantic understanding technology, which is an important innovation different from traditional rule matching methods. The system first encodes the specification text using a pre-trained BERT model (fine-tuned on shipbuilding engineering corpus).

[0094] ​​​​​​A pre-trained language model is used to perform deep semantic analysis on the process specification text, extracting specific control elements from the text, including: constrained piping components (constraint objects), technical requirements to be followed (constraint conditions), specific numerical ranges (constraint values), and the ship region or piping type to which the specification applies (scope of application), etc., generating a high-dimensional semantic vector representation of the specification. For example, the system uses an attention mechanism to identify key semantic elements in the specification text "A reinforcement structure shall be set at the point where the piping passes through the cabin, and the thickness of the reinforcement plate shall not be less than 0.7 times the thickness of the cabin wall": the constraint object is "pipeline passing through the cabin wall", the constraint type is "structural requirement", and the constraint parameter is "reinforcement plate thickness ≥ 0.7 × cabin wall thickness".

[0095] The semantic space of the specifications is constructed using knowledge graph technology. Nodes in the graph represent specification clauses, and edges represent relationships between specifications. Relationship types include: dependency (e.g., welding specifications depend on material specifications), mutual exclusion (e.g., specifications of different pressure levels are mutually exclusive), and inheritance (e.g., specific specifications inherit from general specifications). The system uses a graph neural network (GNN) for specification conflict detection. The network employs three graph convolutional layers, with the message passing formula for each layer as follows:

[0096] ,

[0097] in, Represents the nodes in the graph In the Layer (first) Feature representation vectors in a graph neural network (nth iteration) For nodes In the Layer feature representation, It is a node Neighbors ; For the first The learnable weight matrix of the layer; For nodes The set of neighbors; AGG is the aggregation function, and this embodiment uses mean aggregation; For the activation function, ReLU is used. Indicates collection node All neighbors in the The characteristics of a layer are referred to as messages or neighbor information.

[0098] When a specification conflict is detected, the system models it as a constraint satisfaction problem (CSP). Preferably, the system distinguishes between hard constraints and soft constraints. Hard constraints must be satisfied, such as safety-related specifications; soft constraints should be satisfied as much as possible, such as optimization suggestion specifications. The formula for calculating constraint satisfaction is:

[0099] ,

[0100] where, is the overall satisfaction degree; is the set of hard constraints; is the set of soft constraints; is the indicator function, which is 1 if the hard constraint is satisfied, and 0 otherwise; is the satisfaction degree function of the soft constraint , which takes values in the range [0, 1].

[0101] Dynamic scheduling optimization is another core innovation of the present application. The system constructs a cognitive digital twin system to achieve deep perception and understanding of the construction site.

[0102] At the data collection level, a cognitive digital twin system is constructed to integrate various sources of field data, including: environmental data (temperature, humidity, wind speed, etc.) and equipment status (crane load, welding machine current, etc.) collected every 5 seconds by IoT sensors installed on the construction site, 25 frames of construction picture data per second provided by the field video monitoring system (personnel position, material stacking, work progress, etc. information identified by computer vision algorithm), quality and progress feedback information submitted by construction workers through mobile terminals (such as good welding quality, slight material deformation, etc.) and various types of data such as formal progress report data in the project management system, etc. Heterogeneous data is fused to form a unified cognitive state representation through a deep learning model.

[0103] Multi-modal data fusion adopts a Transformer-based architecture. Different modalities of data are first passed through their respective encoders: numerical sensor data is encoded into a 128-dimensional vector using a fully connected network; video frames are extracted by ResNet to 2048-dimensional features, then mapped to 128-dimensional features through linear transformation; text feedback is encoded by BERT to 768-dimensional, then also mapped to 128-dimensional. These unified dimension feature vectors are fused through multi-head attention mechanism:

[0104] ,

[0105] where, are the query, key, and value matrices, respectively; is the feature dimension. Through parallel computation of 8 attention heads, the system can capture complex associations between different modalities.

[0106] In one embodiment of the present application, the system builds a causal graph of the construction process through a causal discovery algorithm. Using the PC (Peter-Clark) algorithm, the system learns from historical data that the causal chain of "insufficient number of welders" "welding progress delay" "subsequent process waiting". Based on this causal relationship, when it is detected that the number of welders is less than 80% of the planned value, the system can predict that there will be a delay in progress after 3 days, and take measures in advance. The generation of scheduling decisions takes into account multiple factors. The priority scoring model is:

[0107] ,

[0108] where, is the priority score of task ; is the criticality (whether on the critical path, taking the value 0 or 1); is the urgency (the inverse of the number of days to the deadline); is the dependency (the normalized value of the number of dependent tasks); is the resource readiness (the available proportion of required resources); and the weight coefficients , , , are determined according to expert experience and historical data analysis.

[0109] The pallet partition planning module fully considers the actual constraints on site. The system first conducts a detailed constraint analysis, including spatial constraints and equipment constraints.

[0110] In terms of spatial constraints, the system obtains three-dimensional point cloud data of the site through laser scanning, with an accuracy of ±10mm. Based on the point cloud data, the system automatically identifies available storage areas and calculates their length, width, and height dimensions. For example, the size of the temporary storage yard in area A is 20m×15m×6m, but considering the requirement of safety passages, the actual available space is 18m×13m×5.5m. The analysis of transportation channels is also important, the system identifies that the main channel width is 4.5m, the branch channel width is 3m, and the minimum turning radius is 6m, which directly affects the design of the pallet size.

[0111] Equipment capacity constraints are obtained by querying equipment records and real-time monitoring. For example, the rated load of a 25-ton bridge crane is 25 tons, but considering the safety factor of 0.8, the actual use load limit is 20 tons. The fork length of a 3-ton forklift is 1.2m, and the maximum lifting height is 4m, which determines that the maximum size of the pallet cannot exceed 1.2m×1.2m×3.5m (with a safety margin of 0.5m).

[0112] The material flow digital twin model is implemented using discrete event simulation technology. The entities in the model include: pallets (attributes include size, weight, content list), transportation devices (attributes include type, capacity, speed, current location), storage locations (attributes include capacity, occupancy status), construction points (attributes include required materials, required time). The simulation process follows the following logic: materials depart from the warehouse → load onto transportation devices → move along the planned path → arrive at the designated storage location → wait for construction demand → deliver to the construction point.

[0113] In one embodiment of the invention, the system evaluates the resilience of a scheme through 1000 Monte Carlo simulations. Each simulation randomly injects fault events, such as a forklift failure probability of 5%, an average repair time of 2 hours, a temporary lane occupancy probability of 10%, and an average occupancy time of 30 minutes. The resilience index is defined as:

[0114] ,

[0115] where, is the resilience index; is the completion time under normal circumstances; is the average completion time under fault conditions; is the additional cost of handling faults; is the total cost. A scheme with a resilience index greater than 0.85 is considered acceptable. The pallet loading optimization uses an improved three-dimensional packing algorithm. The algorithm first sorts the pipes according to priority, considering the following priorities: construction order (higher priority for earlier use), vulnerability (higher priority for vulnerable components, which require better protection), weight (heavy at the bottom, light at the top). The loading process uses a "best fit decreasing" strategy, which selects the position that maximizes space utilization each time.

[0116] The change management mechanism is the key to ensuring system flexibility. The system classifies changes into three levels according to the scope of impact, each level has different processing procedures and approval requirements.

[0117] First-level changes are major modifications that affect multiple systems, such as adjustments to the main pipe path that may affect multiple branch pipe connections. Such changes require a complete impact analysis process. The system first constructs a change propagation graph, with nodes representing affected objects (pipes, drawings, construction tasks, etc.) and edges representing impact relationships. Through breadth-first search of the graph, the system can identify all directly and indirectly affected objects.

[0118] The quantification of the degree of impact uses the following formula:

[0119] ,

[0120] where, is the total impact degree; the number of affected objects; the importance weight of the object ; the direct impact degree, with a value range ; the indirect impact degree;

[0121] In an embodiment of the present application, the change impact threshold is set to 0.3. The determination of this threshold is based on historical project data analysis: when the impact degree is lower than 0.3, the chain reaction caused by the change can usually be solved by local adjustment; when it exceeds 0.3, systematic re-planning is needed. When the cumulative impact exceeds the threshold, the system automatically triggers a secondary scheduling process.

[0122] Version management adopts a mechanism similar to Git. Each change generates a new version, and the version number format is major version.minor version.revision number. The major version is updated at the project milestone, the minor version is updated at the primary change, and the revision number is updated at the secondary and tertiary changes. The system saves a complete snapshot of each version, supporting difference comparison and historical rollback between versions. Preferably, the system also records the context information of the change, including the change reason, decision basis, related personnel, etc., forming a complete change tracking chain.

[0123] The construction guidance output module converts the optimization results into site executable guidance documents. The intelligent marking process of two-dimensional drawings fully considers the usage habits of construction personnel.

[0124] When the system extracts key construction information from the three-dimensional model, it adopts a method combining rules and machine learning. For standardized information such as pipe segment position, the system directly identifies through geometric analysis; for complex information such as the best welding sequence, the system generates suggestions through the sequence model trained by analyzing historical construction data.

[0125] From the three-dimensional design model of the ship pipeline, key information for guiding on-site construction is extracted, including: the specific position coordinates of the pipeline that needs to be segmented for production, the type and parameters of the welding points used at each connection (such as butt welding, corner welding, etc.), the precise installation position coordinates and type of the pipeline support, and other key technical information necessary for construction personnel. According to the optimization results, the construction sequence number is generated and the precedence relationship between processes is marked. These extracted information is directly converted into specific annotations in the construction guidance document, ensuring that the design intent is accurately conveyed to the construction personnel.

[0126] Annotation layout optimization is an important technical detail. The system divides the drawing into a grid, and each annotation occupies a certain grid cell. The layout optimization problem is modeled as:

[0127] ,

[0128] where, is the number of annotations; is the annotation and the center distance of the two circles; is the overlap indicator variable, 1 if there is overlap, otherwise 0. By solving the problem using genetic algorithm, the system can find an annotation layout that is both clear and compact.

[0129] The construction animation is generated using procedural animation techniques. The system first constructs a sequence of key frames, each corresponding to an important construction node. The transitions between frames are automatically generated using interpolation algorithms. For example, the generation process of the pipe hoisting animation: initial position (ground) → hoisting process (Bezier curve path) → moving in the air (considering swinging) → positioning process (accurate alignment) → fixing complete. The time scale of the animation is compressed according to the actual construction time, usually using a 1:60 ratio (1 minute of animation corresponds to 1 hour of construction).

[0130] The distributed collaboration mechanism of the invention adopts a variety of advanced technologies to ensure the efficient operation of the system. The blockchain-based collaboration framework solves many problems of traditional centralized systems.

[0131] In one embodiment of the invention, the blockchain adopts a consortium chain architecture, and the participating nodes include the design department, production department, material department, quality department, etc. Each department runs a full node and maintains a complete copy of the ledger. The consensus mechanism uses the Practical Byzantine Fault Tolerance (PBFT) algorithm, which can keep the system running normally even in the case of partial node failure.

[0132] The design of the smart contract takes into account the business logic of shipbuilding. For example, the triggering conditions for the completion of the pipe design contract include: 3D model integrity check passed, collision detection no serious conflict, specification check pass rate > 95%, and relevant drawings complete. When these conditions are met, the contract is automatically executed, notifying the relevant departments and updating the task status.

[0133] The federated learning mechanism allows different shipyards to share improvements to the AI model without revealing commercial secrets. During model training, each shipyard calculates the gradient using local data:

[0134] ,

[0135] where, is the gradient of the shipyard ; is the model parameter; is the loss function of the shipyard . After collecting the gradients from each shipyard, the central server updates the global model using the federated averaging algorithm:

[0136] ,

[0137] where is the learning rate; is the number of participating shipyards. To protect privacy, each shipyard adds noise that satisfies differential privacy before uploading the gradient:

[0138] ,

[0139] where is a Gaussian noise with mean and variance ; The value of needs to balance privacy protection and model accuracy. In this embodiment, it is set to .

[0140] The continuous optimization mechanism of the system ensures the performance improvement in the long run. The spiral rise of data, knowledge, and intelligence is achieved through multiple mechanisms.

[0141] Knowledge extraction uses pattern mining technology. The system analyzes historical project data and identifies successful design patterns. For example, through association rule mining, it is found that when the pipe density is > 3 / m², the success rate of layered arrangement is 92%. These patterns are formalized as knowledge rules and stored in the knowledge base.

[0142] The application of knowledge is achieved through the reasoning engine. When facing a new design task, the system first retrieves similar cases and then applies relevant knowledge rules to generate an initial scheme. The update of knowledge uses incremental learning, and the experience of new projects is integrated into the knowledge base in a timely manner.

[0143] The layered optimization architecture effectively balances response speed and global optimization. The bottom layer of local optimization is executed every minute to handle urgent adjustment requirements; the middle layer of domain coordination is executed every hour to ensure consistency within the profession; and the top layer of global optimization is executed every day to adjust resource allocation from a strategic perspective.

[0144] The optimization budget mechanism avoids over-optimization. Each module is allocated a fixed amount of computing resources in each cycle, for example, the collision detection module can use 100 core·minutes of computing resources per hour. Modules need to allocate resources reasonably according to the importance of tasks, and perform fine detection on critical pipes and fast detection on secondary pipes.

[0145] Referring to Figure 1 and Figure 5 , the present application also provides a ship pipeline design drawing stage management system, which includes five main modules that work together to achieve overall functionality.

[0146] The data acquisition and preprocessing module 1 is responsible for data acquisition and standardization processing. This module contains multiple sub-units: the ERP interface unit realizes the docking with the shipyard management system, adopts the RESTful API method, and supports JSON format data exchange; the specification database management unit maintains structured process specification knowledge, supports version management and incremental update; the data cleaning unit identifies and processes abnormal data, and fills in missing values; the timestamp management unit adds time labels to each data item, supporting temporal query.

[0147] The intelligent analysis and decision module 2 is the core of the system, containing five key units. The multi-objective optimization engine adopts a parallel computing architecture, capable of handling multiple optimization tasks simultaneously. The collision detection unit uses GPU-accelerated tensor operations, increasing detection speed by more than 10 times. The process specification detection unit integrates natural language processing capabilities, enabling the understanding and translation of textually described specifications. The dynamic scheduling unit achieves millisecond-level response speed, ensuring timely handling of emergencies. The tray division unit predicts material flow through simulation technology, optimizing storage and transportation solutions.

[0148] The scheme generation and verification module 3 translates analysis results into executable schemes. The simulation engine of this module supports multiple simulation modes: deterministic simulation for basic feasibility verification, stochastic simulation for risk assessment, and optimization simulation for parameter tuning. Verification rules can be flexibly configured to adapt to different project needs.

[0149] The construction guidance output module 4 is responsible for generating guidance documents for on-site use. The drawing generation unit supports multiple CAD formats, ensuring compatibility with existing design software. The animation rendering unit uses real-time rendering technology to quickly generate high-quality construction animations. Output formats include PDF documents, MP4 videos, interactive 3D models, and more, meeting different usage scenarios.

[0150] The execution feedback and iterative optimization module 5 implements closed-loop management. The feedback collection unit supports multiple input methods: mobile APP entry, voice recognition, image recognition, etc. The change impact analysis unit can complete impact assessment within 30 seconds, providing support for rapid decision-making. The version control unit uses distributed storage to ensure data security and traceability.

[0151] Each module communicates through a message queue for loosely coupled communication, adopting a publish-subscribe mode to ensure system scalability and reliability. For example, when the data acquisition module obtains new data, it publishes a data update event, and the intelligent analysis module subscribes to the event and triggers the corresponding analysis process. This architecture design enables the system to flexibly adapt to demand changes, facilitating function expansion and performance optimization.

[0152] It can be seen from the above examples that the ship pipeline design drawing stage management method and system provided by the application introduces advanced technologies such as artificial intelligence, digital twinning, and blockchain, realizes deep collaboration and intelligent optimization of design, production, and materials, significantly improves the efficiency and quality of shipbuilding, and has important engineering application value.

[0153] The above examples only express the specific implementation of the application, and the description is more specific and detailed, but it cannot be understood as a limitation on the scope of the patent of the application. It should be pointed out that for ordinary skilled persons in the art, without departing from the concept of the application, a number of modifications and improvements can be made, which belong to the protection scope of the application.

Claims

1. A method of ship piping design documentation phase management, characterized by, The method comprises the following steps: Obtain production schedule data and specification parameters in the process specification database, structure the obtained data, and establish a data timestamp mechanism to form a standardized input data set; Based on the formed input data set, through a multi-objective optimization engine, comprehensive trade-off analysis of construction conflict minimization, schedule deviation minimization, and material cost optimization is performed, while collision detection based on a four-dimensional space-time model, process specification compliance verification based on semantic understanding, dynamic scheduling optimization considering real-time deviation, and tray division planning based on site constraints are executed, to generate a preliminary pipeline design scheme including construction stage division, drawing time plan, and material list; According to the generated preliminary pipeline design scheme, three-dimensional simulation verification and feasibility evaluation are performed to identify potential design conflicts and construction risks, and an optimized scheme after verification is formed; Based on the optimized scheme after verification, two-dimensional drawings with construction stage identification and construction process animation video are generated, and an executable construction guidance file is output; During the construction execution process, site feedback information and design change requirements are collected, and when the change impact exceeds a preset threshold, secondary scheduling is automatically triggered, the drawing plan is updated, and the comprehensive trade-off analysis based on the multi-objective optimization engine, the collision detection based on the four-dimensional space-time model, the process specification compliance verification based on the semantic understanding, the dynamic scheduling optimization considering the real-time deviation, and the tray division planning based on the site constraints are re-executed to realize iterative optimization of the design scheme.

2. The method for ship pipeline design drawing phase management according to claim 1, characterized in that, The comprehensive trade-off analysis by the multi-objective optimization engine specifically includes: Divide the construction space into three-dimensional grid units and mark the occupancy time window of each grid unit, calculate the space conflict severity index by identifying the case that multiple pipelines occupy the same space at the same time, analyze the equipment sharing conflict and process parallel feasibility between different pipeline construction processes, and quantify the influence degree of process conflict on the overall progress; Construct a construction network diagram and calculate the earliest start time and latest completion time of each activity, identify the critical activity sequence with zero total float time to form a critical path, analyze the construction uncertainty of various pipelines based on historical data, and set time buffer for high-risk activities; Analyze the demand distribution of the same specification pipe material and calculate the unit cost under different batch sizes, comprehensively evaluate the capital occupation cost of advance procurement and the storage space limitation on site, determine the optimal material procurement batch size and timing, and dynamically adjust the weight coefficients of the three optimization objectives of construction conflict, schedule deviation, and material cost according to the stage of the project, generate multiple candidate schemes through iterative optimization, calculate the comprehensive score of each scheme, and select the scheme with the best comprehensive performance as the output result. The collision detection based on the four-dimensional space-time model specifically includes:

3. The method for ship pipeline design drawing phase management of claim 1, wherein, Construct a four-dimensional space-time occupancy tensor containing spatial three-dimensional coordinates and time dimension, where the first three dimensions represent the position coordinates of the pipeline in space, and the fourth dimension represents the construction time sequence; For each pipeline section, calculate its space-time occupancy trajectory in the entire construction period, including the lifting path, the space occupied during installation, and the space required for welding operations; ​ The overlap area of different pipelines in four-dimensional space-time is identified by tensor operation, a probability distribution model is introduced to consider the uncertainty factors in the construction process, and Monte Carlo simulation is used to generate multiple possible construction scenarios and calculate the probability distribution of collision occurrence; When the collision probability exceeds the preset threshold, a collision resolution strategy is generated based on deep reinforcement learning, including adjusting the spatial avoidance scheme of the pipeline path and adjusting the time avoidance scheme of the construction sequence, the impact of different resolution strategies on the overall construction progress and cost is evaluated, and the optimal scheme is selected.

4. The method for ship pipeline design drawing phase management of claim 1, wherein, The process specification compliance verification based on semantic understanding specifically includes: A pre-trained language model is used to perform deep semantic analysis on the process specification text, extracting specific control elements from the process specification text, including: constrained pipeline components, technical requirements to be followed, specifically defined numerical ranges, and core semantic elements such as applicable ship area or pipeline type, generating a high-dimensional semantic vector representation of the specification; A specification semantic space is constructed and the semantic similarity between different specifications is calculated to automatically identify the association, conflict and complementary relationship between specifications, and a mapping relationship between specification semantics and design parameters is established; Abstract specification requirements are automatically converted into specific ship pipeline design parameters, including: the minimum bend radius value required at the pipeline bend, the maximum flow rate variation percentage allowed for fluid flowing in the pipeline, the minimum and maximum spacing range between adjacent pipeline supports, and the quantifiable process indicators such as the required bevel angle value at the welding location; A specification dependency graph is constructed and potential specification conflict patterns are identified using a graph neural network, the specifications are divided into hard constraints and soft constraints, a multi-objective optimization algorithm is used to find a design scheme that maximizes overall constraint satisfaction, and a detailed conflict report and resolution suggestion is generated for conflicts that cannot be completely resolved.

5. The method for ship pipeline design drawing phase management of claim 1, wherein, The dynamic scheduling optimization considering real-time deviations specifically includes: Real-time acquisition of actual drawing completion time, design change occurrence time and content, and on-site feedback problems and delay information for each pipeline, calculation of the time deviation between actual progress and planned progress, and analysis of the development trend of the deviation; A cognitive digital twin system is constructed, integrating multiple sources of on-site data, including: environmental and equipment parameter data collected by Internet of Things sensors installed on the construction site, construction footage data captured by on-site video monitoring systems, quality and progress feedback information submitted by construction workers through mobile terminals, and formal progress report data in the project management system. Different types of data are fused to form a unified cognitive state representation through a deep learning model; An attention mechanism is introduced to identify key state changes, and a causal reasoning network is used to analyze the causal relationships in the construction process to predict the causal effects of different scheduling decisions; Based on the current progress deviation, resource constraints and downstream process dependency relationships, a priority scoring model is established and the drawing priority of each pipeline system is dynamically adjusted to ensure that tasks on the critical path are completed first.

6. The method for ship pipeline design drawing phase management of claim 1, wherein, The tray division planning based on site constraints specifically includes: Analyzing the spatial layout constraints of the construction site, including the size of available storage space, the width and height restrictions of transportation passages, the rated load of lifting equipment, and the loading capacity of transportation vehicles; Building a digital twin model of the material flow to simulate the entire process of material flow from the warehouse to the construction site through discrete event simulation technology, identifying potential bottlenecks and conflicts under different tray partitioning schemes; Grouping similar specifications and similar construction areas of pipelines to ensure that the weight of each tray does not exceed the carrying capacity of the equipment and the size meets the transportation constraints, using a three-dimensional packing algorithm to optimize the space utilization of the tray; Through Monte Carlo simulation, the resilience of the tray scheme under abnormal conditions is evaluated, including transportation equipment failure and temporary unavailability of storage areas, and a tray partitioning scheme that is both efficient and resilient is selected, and a unique identification code containing area, batch, and content information is generated for each tray.

7. The method for ship pipeline design drawing phase management of claim 1, wherein, The collection of site feedback information and design change requirements and the automatic triggering of secondary scheduling specifically includes: Identifying the source of changes and classifying them, including active changes caused by design optimization, passive changes from site problem feedback, and mandatory changes caused by specification updates, dividing them into first-level changes affecting multiple systems, second-level changes within a single system, and third-level changes not affecting interfaces according to the impact range; Analyzing the direct impact of changes, including changes to upstream and downstream connections of pipelines, related drawings that need to be modified simultaneously, and impacts on completed construction parts, while evaluating indirect impacts, including chain changes in space occupation, impacts on other specialties, and impacts on the overall project schedule; Setting a change impact threshold, automatically triggering a re-scheduling process when the cumulative impact exceeds the threshold, updating the status of affected tasks and re-running the multi-objective optimization engine to generate a revised drawing plan; Generating a new version number for each change and saving the complete data before and after the change, supporting version comparison and rollback, determining the appropriate approval process according to the change level, recording approval opinions and decision-making basis to ensure the traceability of changes.

8. The method for ship pipeline design drawing phase management of claim 1, wherein, The generation of two-dimensional drawings with construction stage identification and construction process animation video specifically includes: Extracting key information from the three-dimensional design model of the ship pipeline to guide site construction, including: specific location coordinates where the pipeline needs to be segmented, types and parameters of welding points used at each connection, precise installation location coordinates and types of pipeline supports, and key technical information required by construction personnel, generating construction sequence numbers and marking the sequence dependency between processes according to the optimization results; Automatically adding segment numbers, positioning sizes, welding symbols, quality requirements, material specifications, and quantity information on two-dimensional drawings, avoiding overlapping and blocking of annotations through layout optimization algorithms, and using layer management to manage different types of annotation information; Building a three-dimensional scene containing ship structure models, pipeline models, and construction equipment models, setting animation rhythm according to actual construction time and determining keyframe positions, simulating the complete process of pipeline hoisting, welding operations, and inspection and testing; Color changes are used in the animation to represent the construction status, arrows are added to indicate the installation direction, a semi-transparent effect is set to show the internal structure, the current step description, progress timeline and safety warning signs are superimposed, and multi-format output files are generated to support pause control, chapter jumping and on-site annotation functions.

9. The method of ship piping design drawing phase management according to any one of claims 1 to 8, characterized in that, Also includes: A distributed collaboration framework based on blockchain is adopted, with the design department, production department and material department as blockchain network nodes, data verification and synchronization through consensus mechanism, and automatic triggering of cross-department collaboration processes using smart contracts; A distributed intelligent optimization mechanism based on federated learning is constructed to realize multi-party collaborative training without sharing raw data, and data privacy is protected through differential privacy and secure multi-party computation technology; A spiral upward mechanism is established among data, knowledge and intelligence, domain knowledge base is formed from learning design patterns from historical project data, knowledge guides intelligent decision-making for current projects, and decision-making results produce new data and update the knowledge base through continuous learning; A hierarchical optimization architecture is set to balance local optimization and global coordination, with real-time local optimization at the bottom, intra-domain coordination at the middle, and periodic global strategic optimization at the top, and the optimization resource allocation of each module is controlled through the optimization budget mechanism.

10. A system for ship piping design documentation phase management, characterized by It includes: A data acquisition and preprocessing module is used to obtain schedule data in production and specification parameters in the process specification database, structure the acquired data and establish a data timestamp mechanism to form a standardized input data set; An intelligent analysis and decision-making module includes a multi-objective optimization engine, a collision detection unit, a process specification detection unit, a dynamic scheduling unit and a tray division unit, which is used for comprehensive trade-off analysis of construction conflict minimization, progress deviation minimization and material cost optimization based on the input data set, collision detection based on a four-dimensional space-time model, process specification compliance verification based on semantic understanding, dynamic scheduling optimization considering real-time deviation, and tray division planning based on site constraints, to generate a preliminary pipe design scheme including construction phase division, drawing time plan and material list; A scheme generation and verification module is used to perform three-dimensional simulation verification and feasibility evaluation based on the preliminary pipe design scheme, identify potential design conflicts and construction risks, and form an optimized scheme after verification; A construction guidance output module is used to generate two-dimensional drawings with construction phase identification and construction process animation videos based on the optimized scheme after verification, and output executable construction guidance files; An execution feedback and iterative optimization module is used to collect on-site feedback information and design change requirements during construction execution, automatically trigger secondary scheduling when the change impact exceeds the preset threshold, update the drawing plan and call the intelligent analysis and decision-making module to re-execute the comprehensive trade-off analysis of construction conflict minimization, progress deviation minimization and material cost optimization, collision detection based on a four-dimensional space-time model, process specification compliance verification based on semantic understanding, dynamic scheduling optimization considering real-time deviation, and tray division planning based on site constraints, to realize iterative optimization of the design scheme.

Images