Large ship transformation time estimation method based on spatial constraint critical path

Abstract

The invention discloses a large ship reconstruction time estimation method based on a spatial constraint critical path, relates to the technical field of shipbuilding, and proposes the following scheme: obtaining large ship reconstruction demand data including a subtask set composed of a plurality of subtasks, the subtasks including time consumption, three-dimensional space coordinates and resource demand vectors, and the time consumption, the three-dimensional space coordinates and the resource demand vectors being equal to the time consumption, the three-dimensional space coordinates and the resource demand vectors; a dependency relationship set, a spatial conflict rule base and a global resource pool among all the sub-tasks in the sub-task set are obtained, the dependency relationship is a time sequence constraint edge among the sub-tasks, and three-dimensional space coordinate conflict detection is carried out based on the sub-task set and the dependency relationship set in combination with the spatial conflict rule base. According to the method, the directed acyclic graph is constructed, and the key path is judged based on the double-time equality criterion, so that the problem of key path distortion caused by neglecting space and resource constraints in a traditional key path method is solved, time, space and resource triple constraint coupled key path accurate identification is realized, and the theoretical value and actual deviation of a construction period are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of shipbuilding, more particularly, the present application relates to a large ship reconstruction time estimation method based on a space constraint critical path. BACKGROUND

[0002] The reconstruction of large ships is a kind of shipbuilding technology, which involves a large number of parallel and space-intensive sub-task collaborative scheduling. The existing mainstream method mainly relies on the critical path method (CPM) and project management tools for construction period estimation. Its typical operation mode is as follows: first, a time sequence dependent network is established based on task decomposition, the task float time is calculated by forward / backward propagation, and the longest time-consuming critical path is taken as the total construction period benchmark; the resource allocation adopts a static pooling model, and in the initial scheduling stage, manpower, equipment and other resources are allocated to sub-tasks according to the preset capacity; although this architecture can meet the basic time sequence constraint requirements, when dealing with three-dimensional space coupled tasks (such as cabin welding and pipeline laying) and dynamic resource competition, it still relies heavily on the experience of engineers for manual intervention in conflicts. Especially in large-scale reconstruction projects, the three-dimensional intersection of physical space and the real-time flow of resources make the limitations of traditional static models increasingly prominent.

[0003] The existing large ship reconstruction time estimation method lacks three-dimensional space conflict modeling and dynamic resource constraint processing, resulting in task scheduling rigidity, critical path distortion and construction period estimation deviation. The core defects are as follows: on the one hand, the space constraint edges are not constructed based on the space coordinate conflict rules, resulting in the lack of time sequence dependency between tasks with overlapping physical space or insufficient safety margin, causing scheduling conflicts and efficiency loss; on the other hand, the resource competition is not integrated into the topological sorting process in real time, and the resource shortage lacks adaptive adjustment capability of the dependency relationship, causing task blocking and inaccurate critical path identification. Therefore, the large ship reconstruction time estimation method based on space constraint critical path is proposed to solve this problem. SUMMARY

[0004] To solve the above technical problems, the large ship reconstruction time estimation method based on space constraint critical path is provided, which solves the problems raised in the background technology.

[0005] In order to achieve the above purpose, the technical scheme of the present application is as follows:

[0006] In a first aspect, the present application provides a large ship reconstruction time estimation method based on a space constraint critical path, which comprises:

[0007] Obtaining reconstruction requirement data of a large ship, including a sub-task set composed of a plurality of sub-tasks, wherein the sub-task contains time consumption, three-dimensional space coordinates and resource demand vector;

[0008] Obtain the dependency set, spatial conflict rule base, and global resource pool of all subtasks in the subtask set, where the dependency relationship is the temporal constraint edge between subtasks;

[0009] Based on the set of subtasks and the set of dependencies, and combined with the spatial conflict rule base, 3D spatial coordinate conflict detection is performed to generate spatial constraint edges between subtasks.

[0010] Construct a directed acyclic graph, where the vertices of the directed acyclic graph are subtasks and the directed edges are spatial constraint edges;

[0011] A topological sorting of the directed acyclic graph is performed using an in-degree queue algorithm, and the resource requirement vector of the subtask is matched with the current available amount of the global resource pool during the queue processing.

[0012] If resources are sufficient, allocate resources to the subtask and remove it from the scheduling queue, then add it to the topological sorting sequence. Otherwise, generate a virtual resource constraint edge from the currently resource-occupying subtask to the subtask, update the in-degree count, and add it back to the scheduling queue. Output a linear task sequence that satisfies both the spatial constraint edge and the resource constraint edge.

[0013] Based on the linear task sequence, the earliest start time and earliest finish time of each sub-task are calculated through forward propagation, and the latest start time and latest finish time of each sub-task are calculated through backward propagation starting from these two times.

[0014] If a subtask satisfies the condition that the earliest start time equals the latest start time and the earliest finish time equals the latest finish time, it is included in the critical path. The time consumed by all subtasks is added together as the base time, and then multiplied by the application efficiency factor to obtain the time required for the modification of the large ship.

[0015] Secondly, this application provides a large ship conversion time estimation system based on spatially constrained critical path, used to implement the large ship conversion time estimation method based on spatially constrained critical path described in any of the above claims, including:

[0016] The modification requirement data acquisition module is used to acquire modification requirement data for large ships, including a set of subtasks consisting of multiple subtasks. Each subtask contains time consumption, three-dimensional spatial coordinates, and resource requirement vectors. The module also acquires the dependency set, spatial conflict rule base, and global resource pool among all subtasks in the subtask set. The dependency relationship is a temporal constraint edge between subtasks.

[0017] The directed acyclic graph (DAG) construction module is used to perform 3D spatial coordinate conflict detection based on the set of subtasks and the set of dependencies, combined with the spatial conflict rule base, to generate spatial constraint edges between subtasks and construct a directed acyclic graph. The vertices of the DAG are subtasks, and the directed edges are spatial constraint edges.

[0018] The resource-aware topology sorting module is used to perform topology sorting on a directed acyclic graph using an in-degree queue algorithm. During queue processing, the resource requirement vector of a subtask is matched with the current availability of the global resource pool. If resources are sufficient, resources are allocated to the subtask and it is removed from the scheduling queue and added to the topology sorting sequence. Otherwise, a virtual resource constraint edge is generated from the currently resource-occupying subtask to the subtask, the in-degree count is updated, and the subtask is re-added to the scheduling queue. The module outputs a linear task sequence that satisfies both spatial and resource constraint edges.

[0019] The critical path calculation module is used to calculate the earliest start time and earliest finish time of each sub-task based on the linear task sequence by forward propagation, and to calculate the latest start time and latest finish time of each sub-task by backward propagation starting from the two.

[0020] The modification time prediction module is used to include subtasks in the critical path if the earliest start time equals the latest start time and the earliest finish time equals the latest finish time. The time of all subtasks in the critical path is accumulated as the base time, and multiplied by the application efficiency factor to obtain the modification time requirement for the large ship.

[0021] Thirdly, this application provides a computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described method for estimating the time of large ship conversion based on spatially constrained critical paths.

[0022] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for estimating the conversion time of large ships based on spatially constrained critical paths.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0024] This application combines a spatial conflict rule base to perform Euclidean distance analysis on the three-dimensional coordinates of subtasks and dynamically generates bidirectional spatial constraint edges. This solves the problem of missing temporal dependencies caused by physical space overlap / insufficient safety margin in existing methods, realizes automatic modeling of the temporal relationship of spatially mutually exclusive task pairs, and avoids scheduling deadlock and loss of job efficiency caused by spatial conflicts.

[0025] This application solves the deadlock problem caused by resource contention by dynamically generating virtual resource constraint edges and updating the in-degree queue in real time by comparing the task resource demand vector with the global resource pool availability in topology sorting. It realizes dynamic reconstruction of task dependencies in resource contention scenarios, eliminates scheduling delays caused by resource blocking, and improves task queue fluidity.

[0026] This application solves the problem of critical path distortion caused by neglecting spatial and resource constraints in the traditional critical path method by constructing a directed acyclic graph and determining the critical path based on the dual time equality criterion. It achieves accurate identification of critical paths coupled with time, space and resource constraints, and reduces the deviation between the theoretical and actual project duration. Attached Figure Description

[0027] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. Wherein:

[0028] Figure 1 This is a flowchart of the large ship modification time estimation method based on spatially constrained critical path proposed in this invention.

[0029] Figure 2 This is a structural block diagram of the large ship modification time prediction system based on spatially constrained critical path proposed in this invention. Detailed Implementation

[0030] It is readily understood that, based on the technical solution of this invention, those skilled in the art can propose various interchangeable structural methods and implementations without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention.

[0031] In existing technologies, the retrofitting of large ships involves the collaborative scheduling of a large number of parallel and spatially dense sub-tasks. Traditional methods rely on the critical path method and static resource allocation models. These methods preset resource capacity in the initial scheduling phase, establish a temporal dependency network based on task decomposition, and calculate task float time through forward and backward propagation. However, spatial coordinate conflict rules are not used to construct spatial constraint edges, which leads to the inability to establish correct temporal dependencies when physical space overlaps or safety margins are insufficient. At the same time, dynamic resource constraints are not integrated into the topology sorting process in real time, and the dependency relationship lacks adaptive adjustment capability when resource competition occurs. Ultimately, this leads to three core problems: rigid task scheduling, critical path distortion, and deviation in project duration prediction.

[0032] To solve the above problems, refer to Figure 1 As shown, this application proposes a method for estimating the conversion time of large ships based on spatially constrained critical paths, including:

[0033] Obtain data on the modification requirements of large ships, including a set of subtasks consisting of multiple subtasks, where each subtask contains time consumption, three-dimensional spatial coordinates, and resource requirement vectors.

[0034] It should be noted that the modification of large ships requires multi-deck collaborative operations (adding new entertainment facilities to cruise ships involves nested sub-tasks of electrical, piping and structural engineering), modification of high-risk systems (maintenance of LNG ship cargo tanks requires linkage of sub-tasks of cryogenic welding and airtightness testing), hoisting of giant components, etc. The number of sub-tasks, spatial conflict dimensions and resource competition intensity are significantly higher than those of small and medium-sized ships. Common large ships include cruise ships, LNG carriers, container ships and roll-on / roll-off ferries.

[0035] Obtain the dependency set, spatial conflict rule base, and global resource pool of all subtasks in the subtask set, where the dependency relationship is the temporal constraint edge between subtasks;

[0036] Specifically, the spatial conflict rule base is a structured dataset that quantifies the physical conflict risk of parallel sub-tasks in three-dimensional space by defining safe space thresholds and assembly process constraint rules between different work types. Its construction process is as follows: according to the large ship modification process manual, work types (such as welding, painting, hoisting, pipeline laying, etc.) are assigned to sub-tasks, and a unique code is assigned to each work type; safe distance thresholds are set based on industry safety standards; and several core parameters are pre-set for each work type, which are then collected into three-dimensional structure influence parameters.

[0037] A dependency set is a set of directed edges that describe the mandatory temporal logical constraints between subtasks. Essentially, it is the basic topology in the project scheduling network, which is predetermined by process specifications and engineering logic. Each dependency represents the sequential execution constraint of two subtasks, such as task A must be completed before task B begins (A→B). This temporal constraint originates from the physical assembly sequence (e.g., painting can only be done after welding is completed), safety procedures (e.g., work can only begin after the cabin ventilation test is passed) or technical dependencies (e.g., equipment can only be installed after pipeline laying is completed) in the modification of large ships.

[0038] The global resource pool is a dynamic quantitative model for centralized management of all shared resources in a renovation project. It integrates the types of scarce resources required for large ship renovation and their real-time availability, including but not limited to: human resources (such as the number of specialized welders and assembly technicians) and equipment resources (such as crane hours and the available time of high-pressure sprayers).

[0039] Spatial resources (such as area quotas for enclosed work areas); resource status is dynamically updated during topology sorting: when a subtask is scheduled, its resource requirement vector (e.g., <welders: 3 people, cranes: 2 units>) is deducted from the global resource pool; resources are released back to the pool after the subtask is completed;

[0040] Based on the set of subtasks and the set of dependencies, and combined with the spatial conflict rule base, 3D spatial coordinate conflict detection is performed to generate spatial constraint edges between subtasks.

[0041] Construct a directed acyclic graph, where the vertices of the directed acyclic graph are subtasks and the directed edges are spatial constraint edges;

[0042] A topological sorting of the directed acyclic graph is performed using an in-degree queue algorithm, and the resource requirement vector of the subtask is matched with the current available amount of the global resource pool during the queue processing.

[0043] If resources are sufficient, allocate resources to the subtask and remove it from the scheduling queue, then add it to the topological sorting sequence. Otherwise, generate a virtual resource constraint edge from the currently resource-occupying subtask to the subtask, update the in-degree count, and add it back to the scheduling queue. Output a linear task sequence that satisfies both the spatial constraint edge and the resource constraint edge.

[0044] It should be noted that each dimension of the resource demand vector is the number of special jobs, the time of heavy equipment, and the area of ​​the enclosed work area. During the topology sorting process, the current available amount of the global resource pool is collected in real time, and the sum of the absolute values ​​of the differences between the resource demand vector of the subtask and each dimension of the current available amount is calculated. When the sum of the absolute values ​​is less than the preset tolerance threshold, it is determined that the resources are sufficient.

[0045] For example, the preset tolerance threshold = (number of special jobs required × 8%) + (hours of heavy equipment × 5%) + (area of ​​enclosed work area × 2%); when the environmental risk index > 7.0, the preset tolerance threshold increases by 20%, and when the supply chain risk index > 15%, the preset tolerance threshold decreases by 10%.

[0046] Based on the linear task sequence, the earliest start time and earliest finish time of each sub-task are calculated through forward propagation, and the latest start time and latest finish time of each sub-task are calculated through backward propagation starting from these two times.

[0047] If a subtask satisfies the condition that the earliest start time equals the latest start time and the earliest finish time equals the latest finish time, it is included in the critical path. The time consumed by all subtasks is added together as the base time, and then multiplied by the application efficiency factor to obtain the modification time required for the large ship.

[0048] Therefore, this application deeply integrates three-dimensional spatial coordinate conflict detection with dynamic resource constraint processing mechanisms, incorporating physical spatial location information and resource consumption quantification parameters during the acquisition stage to provide a structured data foundation for subsequent constraint generation; it uniformly expresses temporal, spatial, and resource constraints when constructing a directed acyclic graph, avoiding the reliance on human experience in traditional static models; it embeds dynamic resource verification in real time during the topology sorting stage, using virtual resource constraint edges to transform resource competition into graph structure dependence, enabling the task sequence to adapt to resource flow status; and it combines time parameter consistency to identify the real critical task chain in critical path calculation, and introduces an application efficiency factor to adjust the base time.

[0049] Specifically, the automatic generation of spatial constraint edges effectively eliminates scheduling conflicts caused by physical space overlap, the dynamic establishment mechanism of virtual resource constraint edges solves the task blocking problem when resources are scarce, and the bidirectional propagation of time parameters ensures the accuracy of floating time calculation. Thus, it overcomes the core defects of rigid task scheduling, critical path distortion and project duration prediction deviation, and improves the reliability and engineering applicability of large ship conversion time prediction.

[0050] In an optional embodiment, 3D spatial coordinate conflict detection is performed using a spatial conflict rule base to generate spatial constraint edges between subtasks, specifically including:

[0051] Get the job type of each subtask in the subtask set;

[0052] Extract the safe distance threshold for the corresponding job type from the spatial conflict rule base;

[0053] It should be noted that the setting of safety distance thresholds follows international industry standards, including ship safety distance standards, welding operation specifications, and historical accident data statistics, and the basic values ​​are set differently according to the risk level of different operation types. For example, the range of safety distance thresholds can be: 1.8-2.5m for high-temperature welding, 3.0-3.5m for high-pressure water rust removal, 2.5-3.0m for painting of enclosed compartments, and 4.0-5.0m for hoisting heavy equipment.

[0054] Iterate through the three-dimensional spatial coordinates of all subtasks and calculate the Euclidean distance between every two sets of three-dimensional spatial coordinates.

[0055] Furthermore, when the sub-task coordinates are located within the curved surface region of the hull, the safety distance threshold needs to be adjusted based on the surface inclination angle (the angle between the local tangent plane and the horizontal plane). The update rule is: if the surface inclination angle... The safe distance threshold remains unchanged; if Curved surface angle The safe distance threshold is increased to 1.2 times; if the surface inclination angle The safety distance threshold is increased to 1.4 times; before traversing the three-dimensional coordinates of subtasks to calculate the Euclidean distance, the surface tilt angle of each subtask coordinate point is detected in real time and the safety distance threshold of its respective job type is dynamically adjusted; through surface space coordinate transformation and adjustment of safety distance threshold, the false alarm rate of spatial conflict detection is reduced, and scheduling rigidity caused by redundant spatial constraint edges is avoided.

[0056] When the Euclidean distance is less than the corresponding safe distance threshold, it is marked as a spatial conflict task pair, and two spatial constraint edges in opposite directions are generated for each conflict task pair. Specifically, the system will create two directed edges in opposite directions in the constructed directed acyclic graph: one edge indicates that the first task must be executed before the second task, and the other edge indicates that the second task must be executed before the first task, thereby forcing the two tasks to not be executed at the same time, but only to choose one order of execution to eliminate the risk of spatial conflict.

[0057] Through the above scheme, this application achieves refined and dynamic spatial conflict detection, solves the problem of conflict judgment bias caused by the single safety distance threshold, ensures the accuracy of spatial constraint edge generation, and improves the physical feasibility of task scheduling sequence and the accuracy of critical path calculation.

[0058] In an optional embodiment, after generating the spatial constraint edge, it is determined whether the subtasks at both ends of the spatial constraint edge can perform operations in the target region at the same time.

[0059] Extract the job types of the two subtasks and match them according to the 3D structure influence parameters in the spatial conflict rule base;

[0060] The two sub-tasks are alternately selected as the preceding and following tasks, and the corresponding process assemblability evaluation values ​​are generated by combining the three-dimensional structural influence parameters.

[0061] It should be noted that each job type has 2-4 core parameters pre-set in the spatial conflict rule base, and a parameter weight value is set for each core parameter. For example, welding tasks are configured with heat radiation coefficient and structural deformation risk, and equipment installation tasks are configured with heat sensitivity level and vibration tolerance.

[0062] The calculation process for the process assemblability evaluation value is as follows: Based on the operation type of the preceding sub-task, extract the set of three-dimensional structural influence parameters from the spatial conflict rule base; normalize each three-dimensional structural influence parameter into a risk value in the [0,1] interval; and use the parameter weight values ​​to weightedly sum the risk values ​​in each interval to obtain the process assemblability evaluation value.

[0063] For example, the normalization of 3D structure influence parameters can be broadly categorized into distance-based parameters and threshold-based parameters:

[0064] For distance-related parameters (such as thermal radiation distance), the interval risk value can be obtained by subtracting the ratio of the distance-related parameter to its maximum safety threshold from 1. For threshold-related parameters (such as vibration acceleration), the interval risk value can be obtained by subtracting the ratio of the threshold-related parameter to its critical value.

[0065] Common parameters affecting three-dimensional structures include:

[0066] Thermodynamic parameters: thermal radiation coefficient (welding task), heat sensitivity level (equipment installation task);

[0067] Mechanical parameters: vibration tolerance (precision assembly tasks), structural deformation risk (large component modification tasks);

[0068] Environmental parameters: chemical contamination diffusion radius (painting tasks), ventilation dependence (confined space tasks).

[0069] Spatial parameters: minimum safe operating radius (heavy equipment hoisting tasks), electromagnetic shielding distance (electronic system tasks);

[0070] The feasibility of parallel operation is determined based on the assemblability evaluation value of the process. If the condition is met, the spatial constraint edge is marked as a pseudo constraint edge; otherwise, it is marked as a true constraint edge. The two sets of constraint edges are stored in the directed acyclic graph in partitions.

[0071] It should be noted that the feasibility of parallel operation is determined based on the process assemblability evaluation value as follows: the process assemblability evaluation value is compared with the preset feasibility threshold. If it exceeds the threshold, it is deemed feasible. The preset feasibility threshold is determined by analyzing no less than 300 historical modification conflict cases of large ships through ROC curve analysis. The optimization objective is to maximize the Youden index, balance the true constraint edge recognition rate (TPR) and the false recognition rate (FPR), and reduce the risk of conflict false recognition through engineering verification. Therefore, the preset feasibility threshold can be specifically set to 60 points.

[0072] To facilitate understanding of the above embodiments, a specific application scenario of the above embodiments will be used as an example for illustration below:

[0073] In the refitting of a large ship, the welding task at coordinates (35,12,-4) is 1.8m away from the navigation installation task at coordinates (36,13,-4), and taking the welding task as a prerequisite task as an example:

[0074] Its three-dimensional structural influence parameters are distance-related parameters: thermal radiation risk: Electromagnetic interference risk: The weight values ​​of the parameters associated with the two are 0.7 and 0.3, respectively.

[0075] The process assemblability evaluation value is: ;

[0076] Therefore, the process assemblability evaluation value is 69.2, which is greater than 60 and is judged as a pseudo-constraint edge. Similarly, the process assemblability evaluation value is calculated again with the navigation installation task as the prerequisite task. If and only if both constraint edges are pseudo-constraint edges, it is determined that the two sub-tasks can work in the same area.

[0077] Through the above scheme, this application can accurately identify spatial conflict scenarios that can be parallelized, avoid forcing parallelizable tasks to be executed sequentially, thereby reducing redundant time on the critical path and improving the accuracy of the transformation time prediction.

[0078] In an optional embodiment, generating a virtual resource constraint edge from the currently resource-occupying subtask to that subtask specifically includes:

[0079] Extract the set of all subtasks that occupy the same current resource, and calculate the resource release time of each subtask;

[0080] It should be noted that the resource release time = the earliest completion time of the subtask + the resource recovery buffer time; the resource recovery buffer time is to ensure that the resources are safe and usable after release and comply with industry safety operation standards. The specific buffer time can be set in advance according to the operation type of the subtask. For example, the buffer time for heavy equipment (cranes) is 15 minutes, the buffer time for special types of work (welders) is 5 minutes, and the buffer time for enclosed work areas is 30 minutes.

[0081] To further clarify, when the difference in resource release time among multiple subtasks is ≤5 minutes, they are considered to be released simultaneously. In this case, the subtask with the smallest task float time is selected as the dependent node.

[0082] Select the subtask with the earliest resource release time as the dependent node, and generate a virtual resource constraint edge from the dependent node to the subtask to be scheduled.

[0083] The above technical solution avoids the problem of extended resource waiting time caused by randomly selecting dependent nodes, optimizes the rationality of resource scheduling sequence, thereby improving the accuracy of critical path calculation and reducing the deviation of project schedule prediction.

[0084] In an optional embodiment, after updating the in-degree count, the item is re-added to the scheduling queue, specifically including:

[0085] Extract the scheduled subtasks connected by the newly generated virtual resource constraint edges;

[0086] Increment the in-degree count of the subtask to be scheduled by one unit and extract its resource release time;

[0087] The ratio of the resource release time to the task float time is used to adjust the queue weight, where the task float time is the difference between the latest start time and the earliest start time; the minimum value of the task float time is forcibly set to 5 minutes.

[0088] The subtask to be scheduled is re-inserted into the queue in descending order of the adjusted queue weights.

[0089] Specifically, when the difference between the weights of two subtask queues is ≤0.3, they are considered to have equal weights. In this case, they are sorted in descending order of total resource demand > ascending order of task time > ascending order of task number.

[0090] Real-time monitoring of the waiting time of each subtask in the queue to be scheduled; recalculation of queue weights and reordering according to a dynamic decay formula at fixed intervals.

[0091] The dynamic attenuation formula is as follows:

[0092] ;

[0093] In the formula, For real-time queue weights, The queue weights before adjustment. The attenuation coefficient is... Waiting time;

[0094] It should be noted that the queue weight decays exponentially over time, which conforms to the law of decreasing priority in resource scheduling. The decay coefficient can be 0.02. The dynamic weight decay mechanism of the queue prevents subtasks with excessive task float time from occupying queue resources for a long time due to high initial queue weight. This breaks the rigid logic of "low float time must be given priority" in traditional scheduling and maximizes resource utilization while ensuring the critical path.

[0095] The above solution addresses the lack of a dynamic priority determination mechanism when tasks are re-inserted into the queue, preventing critical tasks from being delayed due to resource congestion, thereby reducing critical path distortion and preventing unnecessary delays in the overall project duration of large ship conversion.

[0096] In an optional embodiment, the dynamic generation process of the application efficiency factor is as follows:

[0097] Obtain the actual and theoretical duration of historical large ship renovation projects, where the theoretical duration is the total time consumed by the subtasks of the critical path corresponding to the historical large ship renovation project.

[0098] The real-time environmental risk index, supply chain risk index, and technology complexity index are obtained, and a comprehensive risk value is calculated by weighting them. The weight ratio for the weighted calculation can be 5:3:2, with environmental risk having the highest weight because it has the greatest impact on the project schedule.

[0099] It should be noted that the environmental risk index quantifies the safety risk level through real-time meteorological data (wind speed, wave height, precipitation), and is assessed using a 0-10 point system. For example, a typhoon (wind speed 20m / s, wave height 3.0m) is scored as 8.5 points, and a clear day (wind speed 5m / s, wave height 0.5m) is scored as 2.0 points. The supply chain risk index quantifies the supply interruption risk level by calculating the delay rate of key materials in sub-tasks, and is assessed using a 0-100% percentage format. The technical complexity index is set based on the operational difficulty classification specification defined by the national standard. This standard classifies technical levels according to dimensions such as process precision requirements, structural complexity, and assembly tolerance. The specific value is determined through three-dimensional model simulation analysis or parameterized calculation of process documents (e.g., curved surface welding tasks are quantified based on conditions such as curvature radius ≤0.5m and weld length ≥10m). The baseline value for conventional pipeline installation is 5.0, and for three-dimensional irregular structure welding, it is 9.5.

[0100] Calculate the total time consumption of subtasks on the current critical path as the total resource requirement, and record the ratio of this to the baseline capacity of the global resource pool as the resource ratio.

[0101] Input the comprehensive risk value and resource ratio into the regression model, and output the application efficiency factor;

[0102] The regression model is as follows:

[0103] ;

[0104] In the formula, To apply the efficiency factor, For resource ratio, For the comprehensive risk value, , , These are the regression coefficients for historical data;

[0105] Specifically, the regression coefficients are set based on the following: data on the deviation between the actual and theoretical construction periods of no fewer than 50 historical large ship renovation projects are used, and the coefficients are obtained through least squares linear regression fitting. The goodness-of-fit verification index is ≥0.88. For example, , , Specifically, these values ​​can be -0.15, 0.25, and 1.05.

[0106] Through the above technical solution, this application can dynamically generate application efficiency factors, so that the transformation time prediction accurately reflects the combined impact of historical project patterns and real-time risk factors, thereby improving the adaptability of the prediction results to the uncertainty of actual implementation.

[0107] In an optional embodiment, a dynamic update mechanism is also included after the critical path is generated:

[0108] Extract the actual start time of non-critical path subtasks. If it is later than the corresponding earliest start time, use the difference between the actual start time and the earliest start time as the delay.

[0109] The difference between the latest start time and the earliest start time corresponding to the subtask is used as the time margin;

[0110] If the delay exceeds the time margin, a critical path recalculation is triggered:

[0111] Release the resources occupied by the subtask to the global resource pool, delete the virtual resource constraint edges derived from it, re-execute the topology sort, and update the critical path;

[0112] It should be noted that critical path calculation is used to determine the transformation time. However, in the implementation process, the actual execution delay of non-critical path subtasks may exceed their time margin, causing the critical path to be distorted and unable to be automatically updated, resulting in deviations in the project schedule. Therefore, this embodiment is proposed to ensure that critical path calculation always reflects the latest execution environment and avoid the prediction model from deviating from the actual progress.

[0113] Through the above technical solution, this application can capture on-site progress deviations in real time, automatically adjust the critical path model, solve the problem of critical path distortion caused by delays in non-critical path sub-tasks, and improve the accuracy and reliability of large ship conversion time estimation.

[0114] See Figure 2 As shown, this solution proposes a large ship retrofit time prediction system based on spatially constrained critical path, used to implement the aforementioned large ship retrofit time prediction method based on spatially constrained critical path, including:

[0115] The modification requirement data acquisition module is used to acquire modification requirement data for large ships, including a set of subtasks consisting of multiple subtasks. Each subtask contains time consumption, three-dimensional spatial coordinates, and resource requirement vectors. The module also acquires the dependency set, spatial conflict rule base, and global resource pool among all subtasks in the subtask set. The dependency relationship is a temporal constraint edge between subtasks.

[0116] The directed acyclic graph (DAG) construction module is used to perform 3D spatial coordinate conflict detection based on the set of subtasks and the set of dependencies, combined with the spatial conflict rule base, to generate spatial constraint edges between subtasks and construct a directed acyclic graph. The vertices of the DAG are subtasks, and the directed edges are spatial constraint edges.

[0117] The resource-aware topology sorting module is used to perform topology sorting on a directed acyclic graph using an in-degree queue algorithm. During queue processing, the resource requirement vector of a subtask is matched with the current availability of the global resource pool. If resources are sufficient, resources are allocated to the subtask and it is removed from the scheduling queue and added to the topology sorting sequence. Otherwise, a virtual resource constraint edge is generated from the currently resource-occupying subtask to the subtask, the in-degree count is updated, and the subtask is re-added to the scheduling queue. The module outputs a linear task sequence that satisfies both spatial and resource constraint edges.

[0118] The critical path calculation module is used to calculate the earliest start time and earliest finish time of each sub-task based on the linear task sequence by forward propagation, and to calculate the latest start time and latest finish time of each sub-task by backward propagation starting from the two.

[0119] The modification time prediction module is used to include subtasks in the critical path if the earliest start time equals the latest start time and the earliest finish time equals the latest finish time. The time of all subtasks in the critical path is accumulated as the base time, and multiplied by the application efficiency factor to obtain the modification time requirement for the large ship.

[0120] In another embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above embodiments.

[0121] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps described above.

[0122] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps described above.

[0123] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.

Claims

1. A method for estimating the refitting time of a large ship based on spatially constrained critical path, characterized in that, The method comprises: Obtaining the reconstruction requirement data of the large ship, including a subtask set composed of multiple subtasks, wherein the subtask contains time consumption, three-dimensional space coordinates and resource requirement vector; Obtaining the dependency set between all subtasks in the subtask set, the space conflict rule library and the global resource pool, and the dependency is the time sequence constraint edge between subtasks; Based on the subtask set and the dependency set, three-dimensional space coordinate conflict detection is performed in combination with the space conflict rule library to generate space constraint edges between subtasks; A directed acyclic graph is constructed, the vertices of the directed acyclic graph are subtasks, and the directed edges are space constraint edges; Topological sorting of the directed acyclic graph is performed by using the in-degree queue algorithm, and the resource requirement vector of the subtask is matched with the current available amount of the global resource pool during the queue processing; If the resources are sufficient, the subtask is allocated resources and removed from the scheduling queue, added to the topological sorting sequence, otherwise a virtual resource constraint edge is generated from the current resource-occupied subtask to the subtask, the in-degree count is updated and then the subtask is added to the scheduling queue again, and a linear task sequence satisfying the space constraint edges and the resource constraint edges is output; Based on the linear task sequence, the earliest start time and the earliest completion time of each subtask are calculated by forward propagation, and the latest start time and the latest completion time of each subtask are calculated by reverse propagation from the two times as starting points; If the subtask satisfies the conditions that the earliest start time is equal to the latest start time and the earliest completion time is equal to the latest completion time, the subtask is included in the critical path, the time consumption of all subtasks in the critical path is accumulated as the base time, and the base time is multiplied by the application efficiency factor to obtain the reconstruction requirement time of the large ship.

2. The method of claim 1, wherein, Three-dimensional space coordinate conflict detection is performed in combination with the space conflict rule library to generate space constraint edges between subtasks, which specifically comprises: Obtaining the operation type of each subtask in the subtask set; Extracting the safety distance threshold corresponding to the operation type from the space conflict rule library; Traversing the three-dimensional space coordinates of all subtasks, calculating the Euclidean distance between each two sets of three-dimensional space coordinates; When the Euclidean distance is less than the corresponding safety distance threshold, mark the two subtasks as a space conflict task pair, and generate two space constraint edges with opposite directions for each conflict task pair.

3. The method of claim 2, wherein, After generating the space constraint edge, it is determined whether the two subtasks at the ends of the space constraint edge can cross and operate in the target area: Extracting the operation type of the two subtasks, and matching the three-dimensional structure influence parameter in the space conflict rule library according to the operation type; Alternately selecting the two subtasks as the preceding task and the subsequent task respectively, generating the corresponding process assembly evaluation value according to the three-dimensional structure influence parameter; According to the process assembly evaluation value, it is determined whether the parallel operation is feasible, if yes, the space constraint edge is marked as a pseudo constraint edge, if not, the space constraint edge is marked as a true constraint edge, and the two types of constraint edge sets are stored in the directed acyclic graph.

4. The method of claim 1, wherein, Generating a virtual resource constraint edge from the current resource-occupied subtask to the subtask, specifically comprising: Extracting all subtask sets occupying the same current resource, and calculating the resource release time of each subtask in the subtask set; Select the subtask with the earliest resource release time as the dependent node, and generate a virtual resource constraint edge from the dependent node to the to-be-scheduled subtask.

5. The method of claim 4, wherein, After updating the in-degree count, rejoin the to-be-scheduled queue, which specifically includes: Extract the to-be-scheduled subtask connected by the newly generated virtual resource constraint edge; Increase the in-degree count value of the to-be-scheduled subtask by one unit, and extract its resource release time; Take the ratio of the resource release time to the task float time as the adjusted queue weight, where the task float time is the difference between the latest start time and the earliest start time; Reinsert the to-be-scheduled subtask into the to-be-scheduled queue in descending order of the adjusted queue weight; Real-time monitor the waiting time of each subtask in the to-be-scheduled queue, and recompute the queue weight and reorder according to the fixed interval and the dynamic decay formula; The dynamic decay formula is: ； wherein is the real-time queue weight, is the queue weight before adjustment, is the decay coefficient, is the latency.

6. The method of claim 1, wherein, The dynamic generation process of the application efficiency factor is: Obtain the actual duration and theoretical duration of a historical large ship modification project, where the theoretical duration is the total time consumption of the subtasks of the corresponding critical path of the historical large ship modification project; Obtain the real-time environmental risk index, supply chain risk index, and technical complexity index, and calculate the comprehensive risk value by weighting; Calculate the total time consumption of the subtasks of the current critical path as the total resource demand, and record the ratio of the total resource demand to the benchmark capacity of the global resource pool as the resource ratio; Input the comprehensive risk value and the resource ratio into the regression model to output the application efficiency factor; The regression model is: ； wherein is an application efficiency factor, is a resource ratio, is a composite risk value, , , is a historical data regression coefficient.

7. The method of claim 1, wherein, After generating the critical path, a dynamic updating mechanism is also included: Extract the actual start time of the non-critical path subtask, and if it is later than the corresponding earliest start time, take the difference between the actual start time and the earliest start time as the delay amount; Take the difference between the latest start time and the earliest start time of the subtask as the time margin; If the delay amount is greater than the time margin, trigger the critical path recalculation: Release the resources occupied by the subtask to the global resource pool, delete the virtual resource constraint edge derived from it, and reexecute the topological sorting to update the critical path.

8. A large ship refit time estimation system based on spatial constraint critical path, characterized in that, A large ship modification time estimation method based on spatial constraint critical path is implemented, including: A modification requirement data acquisition module is used to obtain large ship modification requirement data, including a subtask set composed of multiple subtasks, where a subtask contains time consumption, three-dimensional spatial coordinates, and a resource demand vector, obtain the dependency relationship set between all subtasks in the subtask set, a spatial conflict rule library, and a global resource pool, and the dependency relationship is a temporal constraint edge between subtasks; A directed acyclic graph construction module is used to perform three-dimensional spatial coordinate conflict detection based on the subtask set and the dependency relationship set, in combination with the spatial conflict rule library, generate spatial constraint edges between subtasks, and construct a directed acyclic graph, where the vertices of the directed acyclic graph are subtasks, and the directed edges are spatial constraint edges. The resource-aware topological sorting module is configured to perform topological sorting on the directed acyclic graph by using an in-degree queue algorithm, match the resource requirement vector of a subtask with the current available amount of the global resource pool during queue processing, allocate resources to the subtask and move it out of the queue for scheduling if the resources are sufficient, add it to the topological sorting sequence, otherwise generate a virtual resource constraint edge from the current resource-occupied subtask to the subtask, update the in-degree count, and re-add it to the queue for scheduling, and output a linear task sequence that meets the spatial constraint edge and the resource constraint edge. The critical path calculation module is configured to calculate the earliest start time and the earliest completion time of each subtask by forward propagation based on the linear task sequence, and calculate the latest start time and the latest completion time of each subtask by reverse propagation from the two times as the starting points. The refitting time prediction module is configured to include a subtask in a critical path if the subtask meets the condition that the earliest start time is equal to the latest start time and the earliest completion time is equal to the latest completion time, accumulate the time consumption of all subtasks in the critical path as a base time, and multiply the base time by an application efficiency factor as the refitting requirement time of the large ship. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to implement the method of any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising: The computer program is executed by the processor to implement the method of any one of claims 1-7.

Images