Construction and application method of ship assembly construction digital upgrading strategy inference model

Abstract

The application discloses a ship assembly construction digital upgrading strategy reasoning model construction and application method, comprising: collecting segmented assembly, assembly energy consumption, welding process and related data of outfitting assembly, and building a collaborative assembly simulation platform containing multiple function modules based on the data; then, through corresponding optimization models, prediction algorithms and analysis models, the assembly process scheduling, assembly energy consumption trend and welding quality influence factor are dynamically adjusted, multi-dimensionally deduced and quantitatively analyzed; wherein the assembly scheduling, energy consumption prediction and welding quality analysis steps all comprise sub-step processes of classified data, input corresponding model, parameter adjustment and output result, finally, the output results of each module are integrated to form a multi-scenario adaptive digital upgrading execution scheme, realizing ship assembly construction whole-process data collaborative processing and systematic upgrading, and providing technical support for digital transformation of ship industry.

Description

Technical Field

[0001] This invention relates to the field of digital upgrade technology for ship assembly and construction, and in particular to the construction and application method of a strategy reasoning model for digital upgrade of ship assembly and construction. Background Technology

[0002] Shipbuilding assembly is a complex manufacturing process involving multiple stages and elements, including core processes such as section assembly, final assembly, welding operations, and outfitting assembly. These processes are closely interconnected, with complex technological logic, continuously increasing demands for construction efficiency, resource allocation, and process coordination. As the shipbuilding industry transforms towards digitalization and intelligence, the traditional experience-driven construction model is no longer sufficient to meet the needs of full-process optimization. There is an urgent need for systematic technological integration to achieve effective linkage and efficient processing of data across all stages. Currently, the industry has an increasingly pressing need for multi-process collaborative optimization, full-cycle data utilization, and the development of targeted upgrade solutions. Promoting the digital upgrade of the entire shipbuilding assembly process has become a key direction for enhancing the industry's core competitiveness, and building a systematic upgrade method adaptable to multiple scenarios has become an important support for breaking through the bottlenecks of the traditional model.

[0003] Existing technologies in the field of ship assembly and construction have two significant shortcomings: First, the application of related technologies in various construction stages is fragmented. Processes such as segmented assembly, energy consumption control, welding operations, and outfitting assembly lack a unified collaborative processing mechanism, resulting in poor data flow and an inability to integrate and utilize data across the entire process, making it difficult to form effective linkages in related processing results. Second, there is a lack of systematic upgrade planning methods for the entire ship assembly and construction process. The coupling effects of key elements in each stage are not fully considered, and the adaptability to dynamic changes during the construction process is insufficient. It is difficult to form a comprehensive and multi-dimensional upgrade solution, and thus cannot provide coherent and effective technical support for the digital transformation of ship assembly and construction. Summary of the Invention

[0004] In order to overcome the shortcomings and deficiencies of existing technologies, this invention provides a method for constructing and applying a reasoning model for digital upgrade strategies in ship assembly and construction.

[0005] The technical solution adopted in this invention is a method for constructing and applying a digital upgrade strategy reasoning model for ship assembly and construction, comprising the following steps: S1, collecting data on the operating status of tooling equipment, the correlation data of segment docking accuracy, and the timing data of assembly processes during the ship segment assembly process, and simultaneously acquiring energy consumption monitoring data, hull structure welding process parameters, and outfitting component assembly position correlation data during the final assembly stage; S2, building a ship outfitting collaborative assembly simulation platform based on the collected data, dividing it into a segment assembly collaborative scheduling module, an energy consumption dynamic monitoring module, a welding quality analysis module, and an upgrade strategy reasoning module; S3, dynamically adjusting the priority of assembly processes, resource allocation ratios, and process connection logic through a ship segment assembly collaborative scheduling optimization model; S4, using a final assembly energy consumption dynamic optimization prediction algorithm to perform multi-dimensional extrapolation of energy consumption trends in different processes during the final assembly process; S5, using a hull structure welding quality prediction model to quantitatively analyze the quality influencing factors related to welding current, voltage, and welding speed; S6, integrating the output results of different steps through the ship assembly digital upgrade strategy reasoning model to form a digital upgrade execution plan adaptable to multiple scenarios.

[0006] Furthermore, the expression for the ship section assembly collaborative scheduling optimization model is as follows: ,in, To optimize the target value for segmented assembly collaborative scheduling, For scheduling weight coefficients, Let be the priority coefficient for the i-th assembly process. Let i be the standard execution time of the i-th process. Let j be the allocation proportion of the j-th type of resource. Let the utilization rate of resource type j be , Let be the docking accuracy deviation value of the k-th segment. The stability coefficient of the first tooling device. The load factor for the first tooling device. These are the number of processes, the number of resource types, and the number of tooling and equipment.

[0007] Furthermore, the expression for the dynamic optimization prediction algorithm for final assembly energy consumption is: ,in, Predicted energy consumption values, Energy consumption prediction coefficient, To predict the length of time, The baseline energy consumption value at time t. The weight of the s-th type of energy consumption influencing factor, Let be the value of the s-th type of energy consumption influence factor at time t. The number of energy consumption monitoring nodes, Let be the weight of the u-th monitoring node. for The energy consumption fluctuation value of the u-th monitoring node at time u. S This refers to the total number of categories of all energy consumption influencing factors in the ship assembly and construction process.

[0008] Furthermore, the expression for the prediction model of the hull structure welding quality is as follows: ,in, This is a quantitative value for welding quality. For quality prediction coefficient, The influence index of welding parameters, For welding current, For welding voltage, For welding speed, The number of welded joint types, Let be the quality influence coefficient of the d-th type of welded joint. For the angular deviation of the d-th type of welded joint, The number of welding environmental factors, The weight of the e-th environmental factor, This represents the actual value of the e-th environmental factor.

[0009] Furthermore, the simulation model expression of the ship outfitting collaborative assembly simulation platform is as follows: ,in, The values ​​represent the simulation results of outfitting and collaborative assembly. These are the simulation coefficients. The number of outfitting components. For the dimensional matching degree of the f-th outfitting component, Let f be the installation difficulty coefficient of the f-th outfitting component. Let f be the assembly timing coefficient for the f-th outfitting component. This refers to the number of assembly stations. Let be the collaborative efficiency coefficient of the g-th workstation. Let g be the equipment configuration coefficient for the g-th workstation. The number of outfitting process steps. Let h be the simulation weight of the h-th process step. is the deviation correction coefficient for the h-th process step.

[0010] Furthermore, the expression for the reasoning model of the digital upgrade strategy for ship assembly and construction is as follows: ,in, The result value of the reasoning for the digital upgrade strategy. For strategy reasoning coefficients, To optimize the target value for segmented assembly collaborative scheduling, This is the predicted energy consumption value for final assembly. This is a quantitative value for welding quality. The values ​​represent the simulation results of outfitting and collaborative assembly. To increase the number of dimensions for strategy evaluation, Let z be the weight of the z-th evaluation dimension. For the z-th evaluation dimension, To determine the number of stages in the ship assembly and construction process, Let q be the digital potential coefficient of the q-th calibration step. Let q be the upgrade feasibility coefficient for the q-th calibration stage.

[0011] Further, S3 includes the following sub-steps: S31, classifying and sorting the collected ship section assembly process data, extracting the execution time, resource requirement type, and connection logic information of each process, and establishing a process data association matrix; S32, inputting the association matrix data into the ship section assembly collaborative scheduling optimization model, and using the multi-objective optimization logic built into the model to initially sort the process priorities and determine the boundary between core processes and auxiliary processes; S33, adjusting the resource allocation ratio based on the sorting results, and dynamically correcting the resource allocation scheme in combination with the tooling equipment operation status data to ensure that the resource supply and demand of each process are matched; S34, outputting the optimized scheduling scheme, including the process execution order, resource allocation details, and process connection time nodes, to provide a scheduling basis for the subsequent final assembly process.

[0012] Further, step S4 includes the following sub-steps: S41, screening the calibrated energy consumption nodes in the energy consumption monitoring data of the final assembly stage, extracting the energy consumption data of each node in different process stages, and establishing an energy consumption time series database; S42, dividing the time series database data into a training dataset and a validation dataset according to the time dimension, and setting the initial parameters of the final assembly energy consumption dynamic optimization prediction algorithm; S43, iteratively training the algorithm using the training dataset, adjusting the weights of energy consumption influencing factors and prediction coefficients in the algorithm, so that the energy consumption trend output by the algorithm is consistent with the actual energy consumption data; S44, testing the trained algorithm using the validation dataset, optimizing the algorithm parameters through multiple rounds of testing, and finally outputting a stable final assembly energy consumption prediction result.

[0013] Further, S5 includes the following sub-steps: S51, collecting current, voltage, and welding speed calibration process parameters during the hull structure welding process, simultaneously recording weld joint quality inspection data, and establishing the correspondence between process parameters and quality data; S52, filtering and processing the corresponding data, removing abnormal data points, retaining valid data for model training, and determining the input and output variables of the hull structure welding quality prediction model; S53, inputting valid data into the model, establishing a correlation model between process parameters and welding quality quantification values ​​through the model's nonlinear mapping logic, and adjusting the quality prediction coefficient and influence index in the model; S54, performing quantitative analysis of welding quality under different process parameter combinations based on the correlation model, outputting the quality quantification results corresponding to each parameter combination, and providing data support for welding process optimization.

[0014] The method for constructing and applying a digital upgrade strategy reasoning model for ship assembly construction is implemented through different units, including: a ship section assembly data acquisition and preprocessing unit, a ship outfitting collaborative assembly simulation calculation unit, a section assembly collaborative scheduling optimization analysis unit, a general assembly energy consumption dynamic optimization prediction calculation unit, a hull structure welding quality quantitative evaluation unit, and a digital upgrade strategy reasoning output unit. The ship section assembly data acquisition and preprocessing unit and the ship outfitting collaborative assembly simulation calculation unit are bidirectionally connected, used to transmit the acquired assembly data to the simulation calculation unit and receive simulation feedback data. The ship outfitting collaborative assembly simulation calculation unit... The system is unidirectionally connected to the segmented assembly collaborative scheduling optimization analysis unit, the general assembly energy consumption dynamic optimization prediction calculation unit, and the hull structure welding quality quantitative evaluation unit, transmitting simulation data to the analysis unit respectively. The segmented assembly collaborative scheduling optimization analysis unit, the general assembly energy consumption dynamic optimization prediction calculation unit, and the hull structure welding quality quantitative evaluation unit are all unidirectionally connected to the digital upgrade strategy reasoning output unit, transmitting their respective analysis results to the reasoning output unit. The digital upgrade strategy reasoning output unit integrates the data from each unit, generates an upgrade plan through the ship general assembly construction digital upgrade strategy reasoning model, and performs collaborative processing and strategy output of the data from each unit.

[0015] Beneficial Effects: This invention proposes a method for constructing and applying a reasoning model for the digital upgrade strategy of ship assembly and construction. By systematically collecting core data from each stage and building a dedicated simulation platform, it divides the data into clearly defined collaborative modules, breaking the fragmented limitations of traditional technology applications in each stage. It establishes a data interaction and collaboration mechanism, achieving integrated utilization of data throughout the entire process, effectively overcoming the shortcomings of existing technologies such as a lack of unified collaborative processing and poor data flow. Simultaneously, by dynamically optimizing and quantitatively analyzing key processes such as assembly scheduling, energy consumption prediction, and welding quality assessment, it fully integrates the coupling effects of key elements in each stage, improving adaptability to dynamic changes in the construction process. This forms a comprehensive and multi-dimensional upgrade solution, solving the problems of lacking a systematic upgrade plan for the entire process and difficulty in providing coherent and effective technical support in existing technologies. Finally, by integrating the output results of each module, it generates a targeted upgrade execution plan, promoting the digital transformation of ship assembly and construction towards efficiency, collaboration, and precision, providing strong technical support for the high-quality development of the industry. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention. Figure 2 This is a flowchart of method step S3 of the present invention; Figure 3 This is a flowchart of method step S4 of the present invention; Figure 4 This is a flowchart of step S5 of the method of the present invention; Figure 5 This is a diagram showing the unit composition for implementing the method of the present invention. Detailed Implementation

[0017] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1As shown, the construction and application method of the digital upgrade strategy reasoning model for ship assembly construction includes the following steps: S1, collecting data on the operating status of tooling equipment, the correlation data of segment docking accuracy, and the timing data of assembly processes during the ship segment assembly process, and simultaneously acquiring energy consumption monitoring data, hull structure welding process parameters, and outfitting component assembly position correlation data during the final assembly stage; S2, building a ship outfitting collaborative assembly simulation platform based on the collected data, dividing it into a segment assembly collaborative scheduling module, an energy consumption dynamic monitoring module, a welding quality analysis module, and an upgrade strategy reasoning module; S3, dynamically adjusting the priority of assembly processes, resource allocation ratios, and process connection logic through the ship segment assembly collaborative scheduling optimization model; S4, using a final assembly energy consumption dynamic optimization prediction algorithm to perform multi-dimensional extrapolation of energy consumption trends in different processes during the final assembly process; S5, using a hull structure welding quality prediction model to quantitatively analyze the quality influencing factors related to welding current, voltage, and welding speed; S6, integrating the output results of different steps through the ship assembly digital upgrade strategy reasoning model to form a digital upgrade execution plan adapted to multiple scenarios.

[0019] Step S1 is the foundational step of data acquisition, providing comprehensive and accurate raw data support for subsequent model building and simulation analysis. The implementation process covers key processes in ship assembly and construction through a multi-dimensional data acquisition system. By deploying a distributed sensor network, industrial-grade data acquisition terminals, and production management system interfaces, data from four core processes are collected simultaneously: In the ship section assembly stage, status data such as the operating speed, load pressure, and continuous operating time of tooling equipment are collected at a frequency of once per second. Simultaneously, precision data such as horizontal deviation, vertical deviation, and gap dimensions of section docking are collected using laser rangefinders and 3D scanners, with over 20 measurement points set at each docking surface; In the final assembly stage, energy consumption data such as electricity and gas are collected using energy consumption monitoring instruments at a frequency of once every 5 minutes; In the hull structure welding stage, process parameters such as current, voltage, and welding speed are collected, while simultaneously recording data such as welding environment temperature, humidity, and wind speed; In the outfitting and assembly stage, positional correlation data such as component installation coordinates, gap dimensions with adjacent components, and assembly sequence are collected. All data is initially processed by the acquisition terminals and stored in a dedicated database, ensuring data coverage of all key nodes in the process and laying a data foundation for subsequent work.

[0020] Step S2 establishes a unified collaborative simulation platform to achieve multi-module data integration and efficient processing, breaking the limitations of traditional data silos in different stages. The implementation process is based on multi-source data collected in S1, using digital modeling technology to construct a ship outfitting collaborative assembly simulation platform. First, data interface standardization is performed, converting data from different sources and formats into a unified format, establishing a data dictionary and association rules to ensure smooth data flow. Then, four functional modules are divided: the segmented assembly collaborative scheduling module receives and processes assembly-related data, possessing data classification storage, real-time updates, and query statistics functions; the energy consumption dynamic monitoring module focuses on processing final assembly energy consumption data, setting up data filtering, anomaly identification, and trend analysis sub-functions; the welding quality analysis module constructs a correlation analysis framework between process parameters and quality indicators; and the upgraded strategy reasoning module, as the core, reserves data interfaces with other modules, responsible for integrating processing results and building a strategy reasoning logic framework. The platform adopts a distributed architecture design, supports multi-module parallel computing, and synchronously sets up security mechanisms such as encrypted data storage and access control. The computing efficiency can reach over a thousand data entries per second, meeting the simulation needs of complex scenarios.

[0021] Step S3 optimizes the segmented assembly process to improve assembly efficiency and seamless integration. The implementation process utilizes a collaborative scheduling optimization model for dynamic adjustments. First, it receives process data, equipment data, and docking accuracy data from the segmented assembly collaborative scheduling module. All assembly processes are comprehensively reviewed, clarifying the preconditions, subsequent processes, and resource requirements for each process. A multi-dimensional priority evaluation system is established, setting evaluation indicators and assigning corresponding weights based on three dimensions: critical path attributes, resource consumption intensity, and quality impact weight. Through comprehensive scoring, process priorities are divided into five levels, prioritizing the execution of core processes on the critical path. Based on priority and resource utilization, the allocation ratio of manpower, equipment, and materials is dynamically adjusted, allocating a higher proportion of critical resources to core processes. Simultaneously, a resource allocation conflict coordination mechanism is established to resolve conflicts by adjusting process order and optimizing scheduling paths. Dependencies between processes are analyzed to identify bottlenecks in the integration process. Seamless integration is improved by optimizing waiting time, adjusting integration methods, and adding parallel processes. Finally, an optimized scheduling plan is output, including process execution order, resource allocation details, start and end times, and integration requirements.

[0022] Step S4 accurately predicts the energy consumption trend of final assembly, providing data support for energy consumption optimization and resource allocation. The implementation process employs a dynamic optimization prediction algorithm to conduct multi-dimensional simulations. It receives historical energy consumption data, real-time energy consumption data, and environmental impact data transmitted from the energy consumption dynamic monitoring module. The data is divided into three time periods: hour, day, and week, and into three spatial regions: workshop, assembly line, and individual equipment. A multi-factor energy consumption prediction model is constructed, identifying key influencing factors such as process type, equipment operating parameters, production load, and ambient temperature. Differentiated prediction factors and weights are set for different process characteristics. Time series analysis is used to decompose historical data into trends, identifying long-term trends and seasonal fluctuations. Combined with real-time data and dynamic changes in influencing factors, energy consumption trends for the next 12 hours, 24 hours, and 72 hours are predicted in different time periods. Simultaneously, multi-scenario simulations are conducted to predict corresponding energy consumption for different production plans, equipment operating modes, and process parameter combinations. New data is received every 30 minutes to dynamically correct model parameters, keeping prediction errors within a low range. Finally, an energy consumption trend prediction report for each process, time period, and scenario is output.

[0023] Step S5 quantifies the impact of welding process parameters on quality, providing a precise basis for process optimization and ensuring the welding stability of the hull structure. The implementation process involves quantitative analysis through a welding quality prediction model. First, the system receives process parameter data, environmental data, and quality inspection data transmitted from the welding quality analysis module. Process parameters include real-time changes in welding current, voltage, and welding speed, sampled 5 times per second. Environmental data includes temperature, humidity, wind speed, and gas shielding concentration. Quality inspection data includes weld joint strength, hardness, toughness, weld formation dimensions, and the number and type of defects. The system performs correlation processing on the input data to establish a nonlinear mapping relationship between process parameters, environmental factors, and quality indicators. Correlation analysis clarifies the degree of influence of each parameter on different quality indicators. For welding current, the system analyzes the impact of different value ranges on weld penetration, weld width, and joint strength, determining the correlation between current changes and quality indicator changes. For welding voltage, the system quantifies the impact of different ranges on arc stability, weld formation, and defect incidence. For welding speed, the system studies its balance between welding efficiency and quality, clarifying the patterns of microstructure changes. Simultaneously, the system quantifies environmental interference, corrects the correlation between process parameters and quality indicators, and finally outputs the quantitative results of quality influence factors corresponding to each combination of process parameters.

[0024] Step S6 integrates the full-process analysis results to form a targeted digital upgrade plan, promotes the systematic transformation of the final assembly and construction system, and achieves multi-scenario adaptation through the upgrade strategy reasoning model during implementation. First, data such as segmented assembly scheduling optimization plans, final assembly energy consumption prediction results, and welding quality quantitative analysis reports are received. A data integration framework is established, and different types of result data are standardized. The correlation logic and complementary relationships between data are clarified, achieving an effective combination of process adjustment information and energy consumption data, process optimization suggestions and resource allocation. A multi-dimensional upgrade strategy evaluation system is built, setting six evaluation dimensions: efficiency improvement, energy consumption reduction, quality improvement, cost control, resource optimization, and risk prevention. Specific evaluation indicators and quantitative standards are set for each dimension. The feasibility and effectiveness of different upgrade directions are comprehensively evaluated through a weighted scoring method. Combining the actual needs of ship final assembly construction and digital transformation goals, differentiated strategy reasoning logic is constructed for different scenarios such as batch construction, customized construction, and emergency order production. Through multiple rounds of iterative reasoning, specific upgrade measures including assembly scheduling optimization, energy consumption control, process improvement, and outfitting collaboration optimization are generated. The implementation steps, responsible entities, time nodes, and expected effects of each measure are clarified. Finally, a digital upgrade execution plan that is adaptable to multiple scenarios and covers the entire process is formed, providing a clear and operable implementation guide for transformation.

[0025] Preferably, the expression for the ship section assembly collaborative scheduling optimization model is: ,in, To optimize the target value for segmented assembly collaborative scheduling, For scheduling weight coefficients, Let be the priority coefficient for the i-th assembly process. Let i be the standard execution time of the i-th process. Let j be the allocation proportion of the j-th type of resource. Let the utilization rate of resource type j be , Let be the docking accuracy deviation value of the k-th segment. The stability coefficient of the first tooling device. The load factor for the first tooling device. These are the number of processes, the number of resource types, and the number of tooling and equipment.

[0026] Specifically, the ship section assembly collaborative scheduling optimization model takes the multi-objective collaborative constraints of section assembly as its core and is based on the coupling relationship between process priority, resource allocation, docking accuracy and equipment stability. Since it is necessary to balance scheduling efficiency, resource utilization and docking quality at the same time, a single variable or linear relationship cannot characterize the complex relationship. Therefore, a combination of summation term and product term with logarithmic operation is used to construct the model. The summation term is used to accumulate the basic indicators of different process priorities and resource allocation. The square root operation amplifies the difference characteristics of resource utilization. The logarithmic operation weakens the interference of extreme values ​​of docking accuracy deviation on the overall result. The product term reflects the synergistic effect of tooling equipment operation stability and load status. The parameter values ​​are determined by combining actual production data from ship section assembly. The scheduling weight coefficients are set to four gradients (0.35, 0.28, 0.22, and 0.15) based on process importance, resource scarcity, accuracy requirements, and equipment reliability. Process priority coefficients are assigned values ​​from 1 to 5 according to the proportion of the critical path. The resource allocation proportion is limited to a range of 0.1-0.8. The docking accuracy deviation value is set within a reasonable range of 0-5, referencing industry standards. The equipment operation stability coefficient is determined to be 0.7-0.98 based on statistical analysis of historical fault data from the past three years. This model breaks through the experience-based reliance of traditional scheduling, achieving dynamic coordination of processes, resources, and equipment. During implementation, it receives real-time data transmitted from the section assembly collaborative scheduling module, inputs process information, resource status, accuracy data, and equipment parameters into the model for calculation, and dynamically outputs the optimal scheduling scheme. It clarifies the process execution sequence, resource allocation details, and connection time nodes, providing quantitative guidance for the orderly advancement of the assembly process.

[0027] Preferably, the expression for the dynamic optimization prediction algorithm for final assembly energy consumption is: ,in, Predicted energy consumption values, Energy consumption prediction coefficient, To predict the length of time, The baseline energy consumption value at time t. The weight of the s-th type of energy consumption influencing factor, Let be the value of the s-th type of energy consumption influence factor at time t. The number of energy consumption monitoring nodes, Let be the weight of the u-th monitoring node. for The energy consumption fluctuation value of the u-th monitoring node at time u. S This refers to the total number of categories of all energy consumption influencing factors in the ship assembly and construction process.

[0028] Specifically, the dynamic optimization prediction algorithm for final assembly energy consumption is based on the spatiotemporal distribution characteristics and multi-factor coupling patterns of final assembly energy consumption. Because energy consumption is dynamically affected by multiple dimensions such as process type, equipment parameters, production load, and environmental conditions, and it needs to capture consumption trends over continuous time, linear models cannot meet the required prediction accuracy. Therefore, a model is constructed using a combination of summation terms, product terms, and integral operations. The summation term accumulates the baseline energy consumption at different time dimensions and the cumulative energy consumption fluctuations at different monitoring nodes, respectively. The product term characterizes the nonlinear relationship between energy consumption influencing factors and baseline energy consumption. The integral operation captures the continuous cumulative effect of energy consumption fluctuations over time. The energy consumption prediction coefficients are set to two core values ​​of 0.65 and 0.35 based on data fit and real-time update frequency. The prediction time is divided into three periods: 12 hours, 24 hours, and 72 hours, according to the process cycle. The baseline energy consumption value is determined through statistical analysis of the historical average energy consumption of each process. The weights of energy consumption influencing factors are assigned values ​​of 0.1-0.4 based on their impact on energy consumption, and the weights of monitoring nodes are set to 0.05-0.2 based on coverage and data importance. This algorithm enables accurate prediction of energy consumption in multiple scenarios and dynamics, providing data support for energy consumption optimization and resource allocation. During implementation, it receives historical and real-time energy consumption data transmitted by the energy consumption dynamic monitoring module, divides the data according to time and space dimensions, inputs it into the algorithm, iteratively adjusts parameters through the training dataset, tests and optimizes using the validation dataset, and finally outputs the predicted energy consumption trend results for each process, time period, and scenario.

[0029] Preferably, the expression for the hull structure welding quality prediction model is: ,in, This is a quantitative value for welding quality. For quality prediction coefficient, The influence index of welding parameters, For welding current, For welding voltage, For welding speed, The number of welded joint types, Let be the quality influence coefficient of the d-th type of welded joint. For the angular deviation of the d-th type of welded joint, The number of welding environmental factors, The weight of the e-th environmental factor, This represents the actual value of the e-th environmental factor.

[0030] Specifically, the hull structure welding quality prediction model is based on the inherent nonlinear correlation between welding process parameters, environmental factors, and welding quality indicators. Since welding quality is comprehensively affected by core process parameters such as current, voltage, and welding speed, as well as environmental factors such as temperature and humidity, and there is a synergistic effect among these parameters, a model is constructed using fractional expressions, summation terms, and square root operations. The fractional expressions characterize the core impact of the synergistic effect of welding current, voltage, and welding speed on quality. The summation terms accumulate the influence weights of different weld joint types and environmental factors. The square root operation performs nonlinear adjustments to account for the comprehensive interference of multiple environmental factors. The quality prediction coefficients are set to 0.4, 0.3, and 0.25 based on the degree of influence of process parameters and environmental factors on quality indicators, respectively. The influence index of welding parameters is fitted to the quality change amplitude under different parameter combinations, yielding values ​​of 1.2, 1.1, and 0.9. The influence coefficient of weld joint quality is set to 0.6-0.9 based on the joint structural characteristics and welding difficulty. The weight of environmental factors is assigned from 0.08 to 0.22 based on their degree of interference with the welding process. This model accurately quantifies the impact of various factors on welding quality. During implementation, it receives process parameters, environmental data, and quality inspection data transmitted from the welding quality analysis module, substitutes them into the model for quantitative calculation, clarifies the quality influencing factors corresponding to different combinations of process parameters, provides a scientific basis for optimizing welding process parameters, and improves the stability of welding quality.

[0031] Preferably, the simulation model expression of the ship outfitting collaborative assembly simulation platform is: ,in, The values ​​represent the simulation results of outfitting and collaborative assembly. These are the simulation coefficients. The number of outfitting components. For the dimensional matching degree of the f-th outfitting component, Let f be the installation difficulty coefficient of the f-th outfitting component. Let f be the assembly timing coefficient for the f-th outfitting component. This refers to the number of assembly stations. Let be the collaborative efficiency coefficient of the g-th workstation. Let g be the equipment configuration coefficient for the g-th workstation. The number of outfitting process steps. Let h be the simulation weight of the h-th process step. is the deviation correction coefficient for the h-th process step.

[0032] Specifically, the simulation model of the ship outfitting collaborative assembly simulation platform revolves around the coupling relationship of three core elements in outfitting assembly: component characteristics, workstation collaboration, and process links. Since outfitting assembly involves more than a hundred types of components, dozens of workstations, and multiple process links, it is necessary to reproduce the complex process of multi-element collaboration. Therefore, the model is constructed by combining summation terms, product terms, and arctangent operations. The summation term accumulates the comprehensive characteristic values ​​of the size matching degree, installation difficulty, and assembly sequence of different outfitting components. The product term characterizes the collaborative efficiency of assembly workstations and the synergistic effect of equipment configuration. The arctangent operation is used to perform nonlinear processing on the deviation correction of process links to avoid the excessive influence of extreme deviations. Based on outfitting and assembly design standards and production data from the past three years, simulation coefficients were set to 0.42, 0.33, and 0.25 respectively, considering component assembly difficulty, workstation collaboration level, and the importance of process steps. The dimensional matching degree of outfitting components ranged from 0.85 to 0.99, the installation difficulty coefficient was set to 0.3 to 0.9, the assembly sequence coefficient was set to 0.1 to 0.7, the workstation collaboration efficiency coefficient and equipment configuration coefficient were determined through on-site testing to be 0.6 to 0.95, the simulation weight of process steps was 0.15 to 0.4, and the deviation correction coefficient was 0.05 to 0.2. This model achieves digital simulation of the entire outfitting and assembly process. During implementation, collected outfitting component data, workstation data, and process data are input into the simulation platform. The model completes multi-scenario assembly simulations, identifies collaboration conflicts and deviation risks in advance, and outputs optimized assembly scheme suggestions.

[0033] Preferably, the expression for the reasoning model of the digital upgrade strategy for ship assembly and construction is: ,in, The result value of the reasoning for the digital upgrade strategy. For strategy reasoning coefficients, To optimize the target value for segmented assembly collaborative scheduling, This is the predicted energy consumption value for final assembly. This is a quantitative value for welding quality. The values ​​represent the simulation results of outfitting and collaborative assembly. To increase the number of dimensions for strategy evaluation, Let z be the weight of the z-th evaluation dimension. For the z-th evaluation dimension, To determine the number of stages in the ship assembly and construction process, Let q be the digital potential coefficient of the q-th calibration step. Let q be the upgrade feasibility coefficient for the q-th calibration stage.

[0034] Specifically, the reasoning model for the digital upgrade strategy of ship assembly and construction is based on the coupling relationship between the output results of each module in the entire process and the digital upgrade goals. Because it needs to integrate multi-dimensional results from segmented assembly scheduling, assembly energy consumption prediction, welding quality analysis, and outfitting and assembly simulation to generate a systematic upgrade plan, the results of a single module or simple superposition cannot meet the global optimization requirements. Therefore, a reasoning model is constructed using a combination of fractional and logarithmic operations and a summation term. The fractional operation is used to characterize the synergistic integration and weight allocation of the results of each module, while the logarithmic operation is used to non-linearly amplify the digital potential of key links, highlighting the upgrade value of core links. Based on the shipbuilding industry's digital transformation standards and the actual needs of enterprises, the strategy reasoning coefficients are set to 0.7 and 0.3 according to the priority of upgrade goals and the importance of each module's results. The weights of the upgrade strategy evaluation dimensions are set to 0.22, 0.2, 0.2, 0.18, 0.12, and 0.08 for efficiency improvement, energy consumption reduction, quality improvement, cost control, resource optimization, and risk prevention, respectively. The digital potential coefficient for key links is set to 0.5-0.9, and the upgrade feasibility coefficient is 0.4-0.85. This model breaks through the limitations of independent operation of each module, realizes the collaborative integration of data throughout the entire process, and receives the optimization results and quantitative data output by each module during implementation. These are then substituted into the model for multi-dimensional evaluation and iterative reasoning, ultimately forming a multi-scenario adaptable upgrade execution solution covering all aspects of assembly, energy consumption, welding, and outfitting, providing a clear path for digital transformation.

[0035] Preferred, such as Figure 2 As shown, S3 includes the following sub-steps: S31, classify and sort the collected ship section assembly process data, extract the execution time, resource demand type and connection logic information of each process, and establish a process data association matrix; S32, input the association matrix data into the ship section assembly collaborative scheduling optimization model, and preliminarily sort the process priorities through the multi-objective optimization logic built into the model to determine the boundary between core processes and auxiliary processes; S33, adjust the resource allocation ratio based on the sorting results, and dynamically correct the resource allocation plan in combination with the tooling equipment operation status data to ensure that the resource supply and demand of each process are matched; S34, output the optimized scheduling plan, including the process execution order, resource allocation details and process connection time nodes, to provide a scheduling basis for the subsequent final assembly process.

[0036] Specifically, step S3 involves four sub-steps that achieve precise optimization of segmented assembly collaborative scheduling through systematic phased operations, providing efficient and orderly scheduling support for the final assembly process. Sub-step S31 first categorizes and organizes the collected ship segmented assembly process data according to process type, resource requirements, and connection relationships. Key information such as the execution time of each process, the type and quantity of resources involved, and the numbers of preceding and following processes are extracted. A process data association matrix containing hundreds of records is established using data association technology to ensure data completeness and logical clarity. Step S32 inputs the association matrix data into the ship segmented assembly collaborative scheduling optimization model according to a preset format. The model has built-in multi-objective optimization logic, and based on indicators such as the impact of processes on the overall project duration and resource consumption intensity, it initially prioritizes processes, clarifying the boundaries between core and auxiliary processes. The proportion of core processes is controlled within the overall project duration. S33 Based on the sorting results and combined with the real-time operating status data of tooling equipment, the allocation ratio of human resources, equipment resources, and material resources is dynamically adjusted. The proportion of key equipment in the core process allocation is not less than 70%, and the proportion of skilled human resources is not less than 80%. The resource allocation plan is corrected through multiple rounds of data feedback to ensure that the resource supply and process demand are accurately matched. After completing the scheduling optimization calculation, S34 outputs an optimized scheduling plan including the process execution sequence, details of various resource allocations, and process connection time nodes. The process connection time error in the plan is controlled within 5 minutes, providing a quantitative and operable scheduling basis for the smooth progress of the subsequent final assembly process.

[0037] Preferred, such as Figure 3 As shown, step S4 includes the following sub-steps: S41, screening the calibrated energy consumption nodes in the energy consumption monitoring data of the final assembly stage, extracting the energy consumption data of each node in different process stages, and establishing an energy consumption time series database; S42, dividing the time series database data into a training dataset and a validation dataset according to the time dimension, and setting the initial parameters of the final assembly energy consumption dynamic optimization prediction algorithm; S43, iteratively training the algorithm using the training dataset, adjusting the weights of energy consumption influencing factors and prediction coefficients in the algorithm, so that the energy consumption trend output by the algorithm is consistent with the actual energy consumption data; S44, testing the trained algorithm using the validation dataset, optimizing the algorithm parameters through multiple rounds of testing, and finally outputting a stable final assembly energy consumption prediction result.

[0038] Specifically, step S4 involves four sub-steps that improve the accuracy and stability of final assembly energy consumption prediction through phased data processing and algorithm training, providing reliable data support for energy consumption optimization. Sub-step S41 first filters the energy consumption monitoring data from the final assembly stage, identifying and marking key energy consumption nodes. These nodes cover major production processes and core equipment. Energy consumption data for each node at different process stages is extracted and organized by timestamp to form a continuous energy consumption time-series database. The database data sampling interval is 5 minutes to ensure data continuity and integrity. S42 divides the time-series database data into training and validation datasets in a 7:3 ratio. Based on the characteristics of the final assembly process and energy consumption change patterns, initial parameters for the dynamic optimization prediction algorithm for final assembly energy consumption are set, including an energy consumption influence factor weight range of 0.1-0.4 and an initial prediction coefficient value of 0.65. 43. The algorithm is iteratively trained using a training dataset, with 500 iterations. After each training iteration, the prediction error is calculated, and the weights of energy consumption influencing factors and prediction coefficients in the algorithm are adjusted based on the error results to keep the deviation between the energy consumption trend output by the algorithm and the actual energy consumption data within 5%. S44. The trained algorithm is tested using a validation dataset, with the test sample size covering different process combinations and production load scenarios. The algorithm parameters are optimized through 10 rounds of testing, and finally, a stable final assembly energy consumption prediction result is output. The prediction result includes the energy consumption trend of each process in the next 12 hours, 24 hours, and 72 hours, providing accurate data support for energy consumption control in the final assembly process.

[0039] Preferred, such as Figure 4 As shown, step S5 includes the following sub-steps: S51, collecting current, voltage, and welding speed during the hull structure welding process to calibrate process parameters, simultaneously recording weld joint quality inspection data, and establishing a correspondence between process parameters and quality data; S52, filtering and processing the corresponding data, removing abnormal data points, retaining valid data for model training, and determining the input and output variables of the hull structure welding quality prediction model; S53, inputting valid data into the model, establishing a correlation model between process parameters and welding quality quantification values ​​through the model's nonlinear mapping logic, and adjusting the quality prediction coefficient and influence index in the model; S54, performing quantitative analysis of welding quality under different process parameter combinations based on the correlation model, outputting the quality quantification results corresponding to each parameter combination, and providing data support for welding process optimization.

[0040] Specifically, step S5, consisting of four sub-steps, quantifies the impact of welding process parameters on quality through phased data processing and model calculations, providing a scientific basis for welding process optimization. Sub-step S51 collects key process parameters during the hull structure welding process, including welding current (50-500A), voltage (10-40V), and welding speed (5-50mm / s). Simultaneously, it records weld joint quality inspection data, including strength (200-800MPa), hardness (100-300HB), and defect quantity. The dynamic changes of these parameters are recorded at a frequency of 5 times per second using a data acquisition terminal, establishing a one-to-one correspondence between process parameters and quality data. Sub-step S52 filters the corresponding data, using the 3σ criterion to remove outlier data points, retaining at least 95% of the total collected data as valid samples, thus clarifying the input variables for the hull structure welding quality prediction model. The parameters are process parameters and environmental data, and the output variable is the quantitative value of welding quality. S53 inputs the effective data into the model, establishes the correlation between parameters and quality through the nonlinear mapping logic of the model, and adjusts the quality prediction coefficient in the model to a range of 0.3-0.4 and the welding parameter influence index to 1.0-1.2, so that the model's fit reaches above 0.9. S54 performs quantitative analysis on the welding quality under different combinations of process parameters based on the correlation model. The analysis covers 20 typical parameter combinations, outputs the quantitative results of quality corresponding to each combination, clarifies the correlation between parameter changes and quality indicators, provides a quantitative basis for the precise adjustment of welding process parameters, and improves the stability and consistency of welding quality.

[0041] like Figure 5 As shown, the construction and application method of the digital upgrade strategy reasoning model for ship assembly construction is implemented through different units, including: a ship section assembly data acquisition and preprocessing unit, a ship outfitting collaborative assembly simulation calculation unit, a section assembly collaborative scheduling optimization analysis unit, a general assembly energy consumption dynamic optimization prediction calculation unit, a hull structure welding quality quantitative evaluation unit, and a digital upgrade strategy reasoning output unit. The ship section assembly data acquisition and preprocessing unit and the ship outfitting collaborative assembly simulation calculation unit are bidirectionally connected, used to transmit the acquired assembly data to the simulation calculation unit and receive simulation feedback data. The ship outfitting collaborative assembly simulation calculation unit is divided into... The simulation data is transmitted to the analysis unit via a one-way data connection with the segmented assembly collaborative scheduling optimization analysis unit, the general assembly energy consumption dynamic optimization prediction calculation unit, and the hull structure welding quality quantitative evaluation unit. The segmented assembly collaborative scheduling optimization analysis unit, the general assembly energy consumption dynamic optimization prediction calculation unit, and the hull structure welding quality quantitative evaluation unit are all connected to the digital upgrade strategy reasoning output unit via a one-way data connection, transmitting their respective analysis results to the reasoning output unit. The digital upgrade strategy reasoning output unit integrates the data from each unit, generates an upgrade plan through the ship general assembly construction digital upgrade strategy reasoning model, and performs collaborative processing and strategy output of the data from each unit.

[0042] This invention integrates different scalar and vector parameters for unified calculation, establishing a parameter collaboration framework that aligns with the logic of ship assembly and construction processes. Through dimensional normalization, weight assignment, and nonlinear mapping transformation, it eliminates computational barriers between different types of parameters. Scalar parameters focus on single-dimensional quantitative characteristics, such as process priority coefficients, resource allocation ratios, and welding current magnitudes, directly reflecting the state attributes of individual production factors. Vector parameters, on the other hand, encompass multi-dimensional correlation information, such as deviations in multiple directions related to segmented docking accuracy, energy consumption fluctuations involving energy changes at various time periods and nodes, and multi-dimensional data sets formed by combinations of process parameters, comprehensively depicting the overall state of complex production scenarios. During formula construction, based on the physical correlations and data logic of ship assembly and construction, scalar and vector parameters are first assigned weight coefficients appropriate to their degree of influence. Then, through summation, multiplication, and other operations, precise matching between the dimensions of vector parameters and scalar weights is achieved. Simultaneously, nonlinear operations such as logarithms and square roots are used to adjust the order-of-magnitude differences of different parameters, ensuring comparability and synergy among various parameters within the same calculation system. For example, in the segmented assembly collaborative scheduling optimization model, the scalar process priority coefficient and the vector docking accuracy deviation data are integrated through weight fusion and calculation transformation to achieve synergy, which not only reflects the importance of a single process, but also takes into account the multi-dimensional accuracy requirements of segment docking; in the final assembly energy consumption prediction algorithm, the scalar energy consumption influence factor weight and the vector energy consumption time series data are integrated through integration and summation operations to accurately capture the dynamic trend of energy consumption, and finally enable the formula to integrate the core information of various parameters and output optimization results that fit the actual production.

[0043] The construction and application method of the reasoning model for the digital upgrade strategy of ship assembly and construction systematically collects key data from all stages of ship assembly and construction, builds a dedicated simulation platform integrating multiple modules, realizes the technical integration and data interoperability of processes such as segmented assembly, energy consumption control, and welding quality assessment, and breaks the limitations of the traditional isolated operation of each stage. At the same time, by conducting targeted analysis and dynamic adjustment of assembly scheduling logic, energy consumption trends, and welding quality influencing factors, the coupling effect of key elements in each stage is fully considered, improving the adaptability to complex construction scenarios and providing accurate and comprehensive technical support for digital upgrade.

[0044] This method utilizes a unified simulation platform and modular design to establish a data interaction and collaborative processing mechanism for each stage, enabling the integrated utilization of data throughout the entire process. This solves the problems of fragmented application of technology and poor data flow in existing technologies. By constructing a systematic upgrade strategy reasoning system covering the entire process, it fully integrates the analysis results of each stage to form an upgrade execution plan that adapts to multiple scenarios. This compensates for the shortcomings of existing technologies, such as the lack of overall planning for the entire process and insufficient adaptation to dynamic changes. It promotes the transformation of the digital upgrade of ship assembly and construction from decentralized and experience-based to systematic and precise, providing a coherent and efficient technical solution for the industry's digital transformation.

[0045] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0046] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for constructing and applying a reasoning model for digital upgrade strategies in ship assembly and construction, characterized in that, Includes the following steps: S1 collects data on the operating status of tooling equipment, the correlation data on the docking accuracy of sections, and the timing data of assembly processes during the assembly of ship sections. Simultaneously, it acquires energy consumption monitoring data, hull structure welding process parameters, and correlation data on the assembly position of outfitting components during the final assembly stage. S2, based on the collected data, build a ship outfitting collaborative assembly simulation platform, which is divided into a segmented assembly collaborative scheduling module, an energy consumption dynamic monitoring module, a welding quality analysis module, and an upgrade strategy reasoning module; S3 uses a ship section assembly collaborative scheduling optimization model to dynamically adjust the priority of assembly processes, resource allocation ratios, and process connection logic. S4, adopts the dynamic optimization and prediction algorithm for energy consumption of final assembly to perform multi-dimensional extrapolation of the energy consumption trend of different processes in the final assembly process; S5. The quality influencing factors related to welding current, voltage and welding speed are quantitatively analyzed using a ship structure welding quality prediction model. S6 integrates the output results of different steps through the reasoning model of digital upgrade strategy for ship assembly and construction, forming a digital upgrade execution plan that is adaptable to multiple scenarios. The expression for the ship section assembly collaborative scheduling optimization model is as follows: ， in, To optimize the target value for segmented assembly collaborative scheduling, For scheduling weight coefficients, Let be the priority coefficient for the i-th assembly process. Let i be the standard execution time of the i-th process. Let j be the allocation proportion of the j-th type of resource. Let the utilization rate of resource type j be , Let be the docking accuracy deviation value of the k-th segment. The stability coefficient of the first tooling device. The load factor for the first tooling device. These are the number of processes, the number of resource types, and the number of tooling and equipment; The expression for the dynamic optimization prediction algorithm for final assembly energy consumption is: ， in, Predicted energy consumption values, Energy consumption prediction coefficient, To predict the length of time, Let t be the baseline energy consumption value. The weight of the s-th type of energy consumption influencing factor is... Let be the value of the s-th type of energy consumption influence factor at time t. The number of energy consumption monitoring nodes, Let be the weight of the u-th monitoring node. for The energy consumption fluctuation value of the u-th monitoring node at time u. S The total number of categories for all energy consumption influencing factors in the ship assembly and construction process; The expression for the prediction model of the hull structure welding quality is as follows: ， in, This is a quantitative value for welding quality. For quality prediction coefficient, The influence index of welding parameters, For welding current, For welding voltage, For welding speed, The number of welded joint types, Let be the quality influence coefficient of the d-th type of welded joint. For the angular deviation of the d-th type of welded joint, The number of welding environmental factors, The weight of the e-th environmental factor, This represents the actual value of the e-th environmental factor.

2. The method for constructing and applying the reasoning model for the digital upgrade strategy of ship assembly and construction according to claim 1, characterized in that, The simulation model expression of the ship outfitting collaborative assembly simulation platform is as follows: ， in, The values ​​represent the simulation results of outfitting and collaborative assembly. For simulation coefficients, The number of outfitting components. For the dimensional matching degree of the f-th outfitting component, Let f be the installation difficulty coefficient of the f-th outfitting component. Let f be the assembly timing coefficient for the f-th outfitting component. This refers to the number of assembly stations. Let be the collaborative efficiency coefficient of the g-th workstation. Let g be the equipment configuration coefficient for the g-th workstation. The number of outfitting process steps. Let h be the simulation weight of the h-th process step. is the deviation correction coefficient for the h-th process step.

3. The method for constructing and applying the reasoning model for the digital upgrade strategy of ship assembly and construction according to claim 1, characterized in that, The expression for the reasoning model of the digital upgrade strategy for ship assembly and construction is as follows: ， in, The result value of the reasoning for the digital upgrade strategy. For strategy reasoning coefficients, To optimize the target value for segmented assembly collaborative scheduling, This is the predicted energy consumption value for final assembly. This is a quantitative value for welding quality. The values ​​represent the simulation results of outfitting and collaborative assembly. To increase the number of dimensions for strategy evaluation, Let z be the weight of the z-th evaluation dimension. For the z-th evaluation dimension, To determine the number of stages in the ship assembly and construction process, Let q be the digital potential coefficient of the q-th calibration step. Let q be the upgrade feasibility coefficient for the q-th calibration stage.

4. The method for constructing and applying the reasoning model for the digital upgrade strategy of ship assembly and construction according to claim 1, characterized in that, S3 includes the following steps: S31. Classify and sort the collected ship section assembly process data, extract the execution time, resource requirement type and connection logic of each process, and establish a process data association matrix. S32, input the correlation matrix data into the ship section assembly collaborative scheduling optimization model, and use the multi-objective optimization logic built into the model to initially sort the process priorities and determine the boundary between core processes and auxiliary processes; S33, adjust the resource allocation ratio based on the sorting results, and dynamically correct the resource allocation plan in combination with the tooling and equipment operation status data to ensure that the resource supply and demand of each process are matched. S34 outputs the optimized scheduling scheme, including the process execution sequence, resource allocation details, and process connection time nodes, providing a scheduling basis for the subsequent final assembly process.

5. The method for constructing and applying the reasoning model for the digital upgrade strategy of ship assembly and construction according to claim 1, characterized in that, S4 includes the following steps: S41, screen the calibration energy consumption nodes in the energy consumption monitoring data of the final assembly stage, extract the energy consumption data of each node in different process stages, and establish an energy consumption time series database. S42, divide the time series database data into training dataset and validation dataset according to the time dimension, and set the initial parameters of the dynamic optimization prediction algorithm for total assembly energy consumption; S43, the algorithm is iteratively trained using the training dataset, and the weights and prediction coefficients of the energy consumption influencing factors in the algorithm are adjusted to ensure that the energy consumption trend output by the algorithm is consistent with the actual energy consumption data. S44 uses a validation dataset to test the trained algorithm, optimizes the algorithm parameters through multiple rounds of testing, and finally outputs a stable prediction result for total assembly energy consumption.

6. The method for constructing and applying the reasoning model for the digital upgrade strategy of ship assembly and construction according to claim 1, characterized in that, S5 includes the following steps: S51, collect current, voltage, and welding speed during the hull structure welding process to calibrate process parameters, simultaneously record weld joint quality inspection data, and establish the correspondence between process parameters and quality data; S52, the corresponding relationship data is filtered and processed to remove abnormal data points and retain valid data for model training, and the input and output variables of the ship structure welding quality prediction model are determined. S53. Input the effective data into the model, establish the correlation model between process parameters and welding quality quantification value through the nonlinear mapping logic of the model, and adjust the quality prediction coefficient and influence index in the model. S54 quantifies the welding quality under different combinations of process parameters based on the correlation model, and outputs the quality quantification results corresponding to each parameter combination, providing data support for welding process optimization.

7. The method for constructing and applying the reasoning model for the digital upgrade strategy of ship assembly and construction according to any one of claims 1-6, characterized in that, This method is implemented through different units, including: a ship section assembly data acquisition and preprocessing unit, a ship outfitting collaborative assembly simulation calculation unit, a section assembly collaborative scheduling optimization analysis unit, a general assembly energy consumption dynamic optimization prediction calculation unit, a hull structure welding quality quantitative evaluation unit, and a digital upgrade strategy reasoning output unit. The ship section assembly data acquisition and preprocessing unit is bidirectionally connected to the ship outfitting collaborative assembly simulation calculation unit, used to transmit the acquired assembly data to the simulation calculation unit and receive simulation feedback data. The ship outfitting collaborative assembly simulation calculation unit is unidirectionally connected to the section assembly collaborative scheduling optimization analysis unit, the final assembly energy consumption dynamic optimization prediction calculation unit, and the hull structure welding quality quantitative evaluation unit, respectively, transmitting simulation data to the analysis unit. The section assembly collaborative scheduling optimization analysis unit, the final assembly energy consumption dynamic optimization prediction calculation unit, and the hull structure welding quality quantitative evaluation unit are all unidirectionally connected to the digital upgrade strategy reasoning output unit, transmitting their respective analysis results to the reasoning output unit. The digital upgrade strategy reasoning output unit integrates the data from each unit, generates an upgrade plan through the ship final assembly construction digital upgrade strategy reasoning model, and performs collaborative processing and strategy output of the data from each unit.

Images