Water conservancy construction progress intelligent management method and system

Abstract

The invention provides a water conservancy construction progress intelligent management method and system. Wherein hydraulic engineering project plans are collected, dynamic changes in a life cycle are obtained through an integrated building information model technology, and a hydraulic engineering construction progress management framework is generated; extracting multi-source heterogeneous data, performing cross-modal information integration by applying a multi-modal fusion network algorithm, evaluating specific influence of external environment factors by adopting a causal forest technology, and generating a water conservancy construction progress and risk early warning report; establishing a kinetic equation by applying a graph neural differential equation algorithm, performing dynamic system simulation, and constructing a virtual water conservancy construction environment by adopting an enhanced digital twinborn simulation technology to capture construction efficiency and resource utilization conditions and generate a water conservancy construction resource allocation optimization report; and a lean six-sigma methodology is introduced to set a lean improvement plan, project feedback is collected, and a water conservancy construction progress intelligent management scheme is generated. According to the technical scheme provided by the invention, the intelligent level of water conservancy construction progress management is remarkably improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of water conservancy projects, and in particular, to an intelligent management method and system for water conservancy construction progress. Background Technique

[0002] With the continuous increase in the scale and complexity of water conservancy project items, project planning and management are facing unprecedented challenges. Modern water conservancy projects not only require precise construction progress control but also need to comprehensively capture the dynamic changes within the project life cycle to cope with the ever-changing external environment. Traditional management methods are difficult to provide sufficient flexibility and foresight. Therefore, there is an urgent need for an intelligent management system that can integrate building information model technology to obtain the dynamic changes of the entire project life cycle and generate a water conservancy construction progress management framework with a comprehensive perspective. In addition, the system also needs to have the ability to process multi-source heterogeneous data, perform cross-modal information integration through advanced algorithms, evaluate the specific impact of external environmental factors on the construction progress, so as to achieve accurate risk warning and optimized resource allocation.

[0003] Currently, most water conservancy project construction management relies on traditional methods and technologies, such as progress tracking based on Excel spreadsheets, simple linear programming models, and manual experience judgment. Although these methods meet the basic management needs to a certain extent, they are overwhelmed when facing complex multi-source heterogeneous data. In recent years, some projects have started to introduce building information model technology in an attempt to improve management efficiency through digital means. However, the existing applications of building information model technology mainly focus on the design and initial planning stages, and the support for dynamic changes and risk warning during the construction process is still limited.

[0004] The existing water conservancy construction progress management lacks effective tools and technologies to process and integrate multi-source heterogeneous data, resulting in a serious information island phenomenon, being unable to comprehensively capture the dynamic changes within the project life cycle, and the intelligent level of water conservancy construction progress management is not high. In addition, most traditional risk warning methods are based on historical data and static analysis, and do not fully consider the influence of external environmental factors, with insufficient accuracy and timeliness of warning. In terms of resource allocation, existing solutions usually adopt fixed scheduling plans, lack the ability to flexibly adjust, and have low resource utilization efficiency. These problems limit the efficiency and quality of water conservancy project construction management and increase the risk of delays and cost overruns. Summary of the Invention

[0005] The embodiments of the present application provide an intelligent management method and system for water conservancy construction progress to solve the problem of low intelligent level in existing water conservancy construction progress management.

[0006] In a first aspect, the embodiments of the present application provide an intelligent management method for water conservancy construction progress, including:

[0007] Collect the water conservancy project plan, obtain the dynamic changes during the life cycle of the water conservancy project by integrating building information model technology, provide a comprehensive perspective for capturing the dynamic changes, and generate a water conservancy construction progress management framework;

[0008] Extract the multi-source heterogeneous data in the water conservancy construction progress management framework, apply the multi-modal fusion network algorithm, and perform cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data. Adopt the causal forest technology, add external environmental factors through the random forest expansion in the causal forest technology, and evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay and generate a water conservancy construction progress and risk warning report;

[0009] Apply the graph neural differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and use the enhanced digital twin simulation technology to construct a virtual water conservancy construction environment. Map the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation and generate a water conservancy construction resource allocation optimization report;

[0010] According to the latest optimization results in the water conservancy construction resource allocation optimization report, combine with the historical construction data in the water conservancy construction progress management framework, introduce the lean six sigma methodology, set a lean improvement plan for the latest optimization results, collect project feedback on the lean improvement plan to dynamically adjust the lean improvement plan, and generate a water conservancy construction progress intelligent management plan.

[0011] Optionally, the extracting the multi-source heterogeneous data in the water conservancy construction progress management framework, applying the multi-modal fusion network algorithm, performing cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data, adopting the causal forest technology, adding external environmental factors through the random forest expansion in the causal forest technology, evaluating the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay and generating a water conservancy construction progress and risk warning report includes:

[0012] Extract the multi-source heterogeneous data in the water conservancy construction progress management framework, perform data cleaning and noise filtering on the multi-source heterogeneous data, adopt time series analysis technology to identify the periodic and trend components in the results of data cleaning and noise filtering, and generate a preprocessed multi-source heterogeneous data set;

[0013] Apply the multi-modal fusion network algorithm, and perform cross-modal information integration on the preprocessed multi-source heterogeneous data set through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the mutual relationship between each data modality in the multi-source heterogeneous data, and generate a comprehensive multi-modal data representation;

[0014] Adopt the causal forest technology, add external environmental factors to the comprehensive multi-modal data representation through the random forest extension in the causal forest technology, and evaluate the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay and generate a causal effect evaluation report;

[0015] Use the causal effect evaluation report as an input variable, apply a time series prediction model to deeply analyze the historical progress data of the water conservancy construction progress, and predict the future construction progress according to the deep analysis results to identify the potential risk points and occurrence probabilities in the water conservancy construction progress, and generate a water conservancy construction progress and risk warning report.

[0016] Optionally, the application of the multi-modal fusion network algorithm to perform cross-modal information integration on the preprocessed multi-source heterogeneous data set through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the mutual relationship between each data modality in the multi-source heterogeneous data and generate a comprehensive multi-modal data representation includes:

[0017] Adopt data alignment technology, set a specific time dimension, and synchronize the data from different sources in the preprocessed multi-source heterogeneous data set on the specific time dimension to generate a synchronized multi-source data set;

[0018] Apply the multi-modal fusion network algorithm, and perform cross-modal information integration on the synchronized multi-source data set and the preprocessed multi-source heterogeneous data set through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the mutual relationship between each data modality in the multi-source heterogeneous data and generate a cross-modal information set;

[0019] Through feature extraction technology, extract cross-modal feature vectors from the cross-modal information set, apply cluster analysis to the cross-modal feature vectors, and perform clustering processing according to the association strength between the cross-modal feature vectors to generate cross-modal feature clusters;

[0020] Introduce the principal component analysis method and preset a low-dimensional space, map the high-dimensional features in the cross-modal feature cluster into the low-dimensional space, reduce the data redundancy of the cross-modal feature cluster, and generate a comprehensive multi-modal data representation.

[0021] Optionally, adopt the causal forest technology, through the random forest expansion in the causal forest technology, add external environmental factors to the comprehensive multi-modal data representation, and evaluate the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn of the risk of water conservancy construction progress delay and generate a causal effect evaluation report, including:

[0022] Apply a feature selection algorithm to the comprehensive multi-modal data representation for feature selection, identify the significant impact features for the water conservancy construction progress, and through time series decomposition technology, separate the seasonal and random fluctuation components of the significant impact features to generate a refined feature set;

[0023] Adopt the causal forest technology, through the random forest expansion in the causal forest technology, import the refined feature set into the comprehensive multi-modal data representation, and add external environmental factors, evaluate the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn of the risk of water conservancy construction progress delay and generate a preliminary causal effect evaluation result;

[0024] Based on the preliminary causal effect evaluation result, introduce the Bayesian optimization algorithm, systematically search for the optimal hyperparameter combination in the evaluation process, and apply sensitivity analysis to evaluate the risk level of different environmental factor combinations in the optimal hyperparameter combination to generate an optimized causal effect evaluation model;

[0025] Apply the decision tree analysis method to classify and prioritize the key risk factors in the optimized causal effect evaluation model, and formulate a risk avoidance plan according to the risk classification and prioritization results to generate a causal effect evaluation report.

[0026] Optionally, use the graph neural differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, adopt the enhanced digital twin simulation technology to construct a virtual water conservancy construction environment, and map the construction resource allocation into the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation and generate a water conservancy construction resource allocation optimization report, including:

[0027] Introduce an anomaly detection algorithm to label the anomalies in the water conservancy construction progress and risk early warning report, and use fault tree analysis to construct a causal relationship diagram for the anomalies to reveal the root causes of the anomalies and generate a causal relationship diagram of anomaly events;

[0028] Apply the graph neural differential equation algorithm, with the causal relationship diagram of anomaly events as the constraint condition, establish a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk early warning report, and perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation and generate a resource optimization dynamic model;

[0029] Adopt enhanced digital twin simulation technology, construct a virtual water conservancy construction environment according to the resource optimization dynamic model, map the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation, and generate a virtual construction simulation environment;

[0030] Apply multi-criteria decision analysis to quantitatively evaluate the simulation data in the virtual construction simulation environment, and optimize the construction efficiency and resource utilization of the construction resource allocation according to the quantitative evaluation results to generate a water conservancy construction resource allocation optimization report.

[0031] Optionally, the application of the graph neural differential equation algorithm, with the causal relationship diagram of anomaly events as the constraint condition, establishing a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk early warning report, performing dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and generating a resource optimization dynamic model includes:

[0032] Extract key node and edge information from the causal relationship diagram of anomaly events, apply the shortest path algorithm in graph theory to calculate the maximum carrying capacity of each node in the key node and edge information, and generate the maximum path carrying capacity;

[0033] Use the maximum path carrying capacity as the input variable to construct a graph structure network for the water conservancy construction progress, apply the graph neural differential equation algorithm to the graph structure network, with the causal relationship diagram of anomaly events as the constraint condition, establish a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk early warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and generate a construction progress dynamic model;

[0034] Train an intelligent agent using reinforcement learning technology, introduce construction safety and quality assurance as additional constraint conditions, and let the intelligent agent learn the best resource allocation strategy in the construction progress dynamic model to generate the best resource allocation strategy for construction progress;

[0035] Using a mathematical programming method for the construction progress dynamic model, the optimal parameters are solved under the conditions of the best resource allocation strategy for the construction progress and the maximum carrying capacity of the path, and a resource optimization dynamics model is generated according to the optimal parameters.

[0036] Optionally, based on the latest optimization results in the water conservancy construction resource allocation optimization report, combined with the historical construction data in the water conservancy construction progress management framework, introducing the Lean Six Sigma methodology, setting a lean improvement plan for the latest optimization results, collecting project feedback of the lean improvement plan to dynamically adjust the lean improvement plan, and generating a water conservancy construction progress intelligent management plan, including:

[0037] Using the difference analysis method, combined with the historical construction data in the water conservancy construction progress management framework, performing difference analysis on the latest optimization results in the water conservancy construction resource allocation optimization report, quantifying the contribution of the construction period shortening of the resource optimization strategy in the water conservancy construction resource allocation optimization report, and generating a resource optimization strategy contribution table;

[0038] Introducing the Lean Six Sigma methodology, taking the strategy contribution value in the resource optimization strategy contribution table as a reference value, defining improving construction efficiency and reducing delay risk as improvement objectives, and generating comprehensive improvement objectives;

[0039] Designing a lean improvement plan according to the comprehensive improvement objectives, verifying the effectiveness of the lean improvement plan by running a pilot project, and collecting feedback during the process of running the pilot project to dynamically adjust the lean improvement plan, and generating a pilot evaluation result;

[0040] Applying discrete event simulation technology to set multiple simulation scenarios, evaluating the execution effects of the water conservancy construction progress management framework under the multiple simulation scenarios, introducing a feedback mechanism, regularly collecting the simulation prediction values in the execution effects, making a comparative analysis with the comprehensive improvement objectives as the standard, and regularly adjusting the latest optimization results to generate a water conservancy construction progress intelligent management plan.

[0041] In a second aspect, an embodiment of the present application provides a water conservancy construction progress intelligent management system, including:

[0042] A collection module, configured to collect the water conservancy project plan, obtain the dynamic changes during the life cycle of the water conservancy project by integrating building information model technology, provide a comprehensive perspective for capturing the dynamic changes, and generate a water conservancy construction progress management framework;

[0043] An integration module, configured to extract multi-source heterogeneous data in the water conservancy construction progress management framework, apply a multi-modal fusion network algorithm, perform cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data, adopt a causal forest technology, add external environmental factors through the random forest expansion in the causal forest technology, evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay, and generate a water conservancy construction progress and risk warning report;

[0044] A simulation module, configured to apply a graph neural differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, adopt an enhanced digital twin simulation technology to construct a virtual water conservancy construction environment, and map the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation, and generate a water conservancy construction resource allocation optimization report;

[0045] An adjustment module, configured to, according to the latest optimization result in the water conservancy construction resource allocation optimization report, combine the historical construction data in the water conservancy construction progress management framework, introduce the lean six sigma methodology, set a lean improvement plan for the latest optimization result, collect project feedback of the lean improvement plan to dynamically adjust the lean improvement plan, and generate a water conservancy construction progress intelligent management plan.

[0046] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a water conservancy construction progress intelligent management method as described in the first aspect.

[0047] In a fourth aspect, an embodiment of the present application provides a computer storage medium, storing a computer program, and when the computer program is executed by a computer, it implements a water conservancy construction progress intelligent management method as described in the first aspect.

[0048] In the embodiments of the present application, the planning of water conservancy engineering projects is collected, the dynamic changes during the life cycle of the water conservancy engineering project are obtained by integrating building information modeling technology, a comprehensive perspective for capturing the dynamic changes is provided, and a water conservancy construction progress management framework is generated; the multi-source heterogeneous data in the water conservancy construction progress management framework is extracted, the multi-modal fusion network algorithm is used, and cross-modal information integration of the multi-source heterogeneous data is performed through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data. The causal forest technology is adopted, and external environmental factors are added through the random forest expansion in the causal forest technology to evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay, and a water conservancy construction progress and risk warning report is generated; the graph neural differential equation algorithm is used to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, and dynamic system simulation is performed on the dynamic equation to find the optimal solution of the dynamic equation. The enhanced digital twin simulation technology is adopted to construct a virtual water conservancy construction environment, and the construction resource allocation is mapped into the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation, and a water conservancy construction resource allocation optimization report is generated; according to the latest optimization results in the water conservancy construction resource allocation optimization report, combined with the historical construction data in the water conservancy construction progress management framework, the lean six sigma methodology is introduced, a lean improvement plan is set for the latest optimization results, the project feedback of the lean improvement plan is collected to dynamically adjust the lean improvement plan, and a water conservancy construction progress intelligent management plan is generated. By integrating a variety of advanced technologies, the accuracy and efficiency of construction progress management and risk warning are significantly improved. Using building information modeling technology (BIM) to capture the dynamic changes during the project life cycle provides a comprehensive perspective and ensures the comprehensiveness and accuracy of the management framework. The application of the multi-modal fusion network algorithm and the causal forest technology enhances the processing ability of multi-source heterogeneous data and realizes more accurate risk assessment and warning. The combination of the graph neural differential equation algorithm and the enhanced digital twin simulation technology optimizes resource allocation and improves resource utilization efficiency. The introduction of the lean six sigma methodology realizes continuous improvement and ensures the flexibility and adaptability of the construction plan.

[0049] Furthermore, through detailed data preprocessing and time series analysis, high-quality basic data was ensured, providing a reliable basis for subsequent risk assessment. By adopting the multi-modal fusion network algorithm and causal forest technology, not only was the effect of cross-modal information integration enhanced, but also the impact of external environmental factors was accurately evaluated, generating a detailed causal effect assessment report. By applying the time series prediction model, potential risk points and their occurrence probabilities were identified in advance, making the construction schedule and risk warning more accurate and timely. This method effectively improved the predictability and accuracy of risk warning, helped project managers formulate countermeasures in advance, and significantly reduced the risk of construction delays.

[0050] Furthermore, through anomaly detection algorithms and fault tree analysis, a causal relationship diagram of abnormal events was constructed, revealing the root causes and providing reliable constraint conditions for the establishment of dynamic equations. The application of the graph neural differential equation algorithm ensured the scientificity and rationality of the dynamic equation simulation, found the optimal solution, and generated an optimized dynamic model of resources. The combination of enhanced digital twin simulation technology and multi-criteria decision analysis not only intuitively demonstrated the construction efficiency and resource utilization of construction resource allocation, but also optimized resource allocation through quantitative evaluation, greatly improving the flexibility and response speed of construction management. This method significantly improved the scientificity and rationality of resource allocation, reduced resource waste, increased construction efficiency, and thus reduced the total cost of the project.

[0051] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0053] Figure 1 It is a flowchart of an intelligent management method for water conservancy construction progress provided by an embodiment of the present application;

[0054] Figure 2 It is a schematic structural diagram of an intelligent management system for water conservancy construction progress provided by an embodiment of the present application;

[0055] Figure 3 It is a schematic structural diagram of a computing device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] To enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application.

[0057] In some processes described in the specification, claims and the above-mentioned drawings of this application, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this article or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit that "first" and "second" are of different types.

[0058] The technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of this application.

[0059] Figure 1 A flowchart of an intelligent management method for the construction progress of water conservancy projects is provided for the embodiments of this application, as Figure 1 shown. The method includes:

[0060] 101. Collect the water conservancy project plan, obtain the dynamic changes during the life cycle of the water conservancy project through integrating building information modeling technology, provide a comprehensive perspective for capturing the dynamic changes, and generate a water conservancy construction progress management framework;

[0061] In this step, the water conservancy project plan refers to the detailed preliminary planning of the overall design, construction and operation of the water conservancy project, including project goal setting, resource allocation and time arrangement.

[0062] Building information modeling technology is a digital modeling tool that can create and manage information throughout the project life cycle from design to construction to operation.

[0063] The water conservancy construction progress management framework is a systematic management platform established based on the collected data and information for monitoring and managing the construction progress. This framework integrates all relevant data to ensure the coherence and consistency of information in each stage.

[0064] In the embodiments of the present application, first, a three-dimensional digital model of a water conservancy project is constructed using building information modeling technology to ensure that all relevant information is visually displayed; second, multi-source data such as construction plans, material lists, and human resources are integrated into the model to form a unified data platform; third, data at the construction site is collected in real time through sensors and monitoring devices to ensure real-time information update; finally, a dynamically updated construction progress management framework is generated by combining historical data and predictive analysis, providing a solid foundation for subsequent steps.

[0065] First, assume that a large water conservancy project first constructs a digital model of the whole life cycle using building information modeling technology; second, integrate multi-source data of design drawings, construction logs, and material procurement records; third, install multiple sensors to monitor the on-site environment and construction progress; finally, develop a comprehensive platform to display the project progress in real time and compare it with the prediction model to ensure the accuracy and timeliness of the management framework.

[0066] 102. Extract the multi-source heterogeneous data in the water conservancy construction progress management framework, apply the multi-modal fusion network algorithm, and perform cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data. Adopt the causal forest technology, and add external environmental factors through the random forest expansion in the causal forest technology to evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay, and generate a water conservancy construction progress and risk warning report;

[0067] In this step, multi-source heterogeneous data refers to data from different sources and with different formats, which usually have different structures and semantics.

[0068] The multi-modal fusion network algorithm is a technology for processing multi-source heterogeneous data, aiming to enhance the integrity and accuracy of information by fusing data of different modalities. This algorithm can capture the correlation between different data types.

[0069] The cross-modal attention mechanism is part of the multi-modal fusion network algorithm, used to evaluate the importance between different data modalities, thereby optimizing the information integration effect. It can highlight key information and improve the accuracy of data analysis.

[0070] The causal forest technology is a method for expanding the random forest algorithm, used to identify the causal relationship between variables. By introducing external environmental factors, it can evaluate the specific impact of these factors on the construction progress.

[0071] External environmental factors refer to various external factors that may affect the construction progress, including weather conditions, changes in policies and regulations, and market fluctuations.

[0072] The water conservancy construction progress and risk early warning report evaluates the impact of external environmental factors through the analysis of multi-source heterogeneous data to early warn of the risk of construction progress delays and propose corresponding countermeasures.

[0073] In the embodiments of this application, first, multi-source heterogeneous data in the water conservancy construction progress management framework is extracted and preprocessed to eliminate noise and improve data quality; second, the multi-modal fusion network algorithm is used to integrate the multi-source heterogeneous data through a cross-modal attention mechanism to highlight the important associations between different data modalities; third, the causal forest technology is adopted to evaluate the specific impact of external environmental factors on the construction progress and reveal potential risk points; finally, a water conservancy construction progress and risk early warning report is generated to provide decision-making support for project managers.

[0074] For example, continuing with the previous example, assume that a large-scale water conservancy project has generated a water conservancy construction progress management framework. First, the system extracts multi-source heterogeneous data from the framework and preprocesses it to eliminate noise and improve data quality; second, the multi-modal fusion network algorithm is used to integrate the multi-source heterogeneous data through a cross-modal attention mechanism to highlight the important associations between different data modalities; third, the causal forest technology is adopted to evaluate the specific impact of external environmental factors such as weather conditions and market fluctuations on the construction progress and identify potential risk points; finally, the system generates a detailed water conservancy construction progress and risk early warning report, early warning of possible delay risks and proposing specific countermeasures.

[0075] 103. Use the graph neural ordinary differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk early warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, use the enhanced digital twin simulation technology to construct a virtual water conservancy construction environment, map the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation, and generate a water conservancy construction resource allocation optimization report;

[0076] In this step, the graph neural ordinary differential equation algorithm is an advanced algorithm that combines graph neural networks and ordinary differential equation solvers and is used to handle dynamic problems in complex networks. It can simulate the relationships and dynamic changes between nodes.

[0077] The dynamic equation is a mathematical expression that describes the change of a system over time and can be used to simulate the construction resource allocation and scheduling problem to find the optimal solution.

[0078] The enhanced digital twin simulation technology is a virtual reality technology that realizes the synchronous mapping of the physical world and the digital world by constructing a virtual construction environment to help optimize resource allocation.

[0079] The construction resource allocation optimization report is the best resource allocation recommendation generated by simulating and evaluating different resource allocation scenarios, aiming to improve construction efficiency and resource utilization.

[0080] In the embodiments of the present application, first, an anomaly detection algorithm is introduced to label the anomalies in the water conservancy construction progress and risk early warning report, and a causal relationship diagram is constructed using fault tree analysis; second, the graph neural differential equation algorithm is used to establish a dynamic equation for the construction resource allocation and scheduling problem to simulate the dynamic changes under different configurations; third, the dynamic system simulation is performed on the dynamic equation to find the optimal solution and generate a resource optimization dynamic model; finally, the enhanced digital twin simulation technology is adopted to map the construction resource allocation to the virtual water conservancy construction environment, capture the construction efficiency and resource utilization, and generate the construction resource allocation optimization report.

[0081] For example, continuing with the above example, assume that a large-scale water conservancy project has generated a water conservancy construction progress and risk early warning report. First, the system introduces an anomaly detection algorithm to label the anomalies in the report and constructs a causal relationship diagram of the abnormal events using fault tree analysis; second, the graph neural differential equation algorithm is used to establish a dynamic equation for the construction resource allocation and scheduling problem and simulate the dynamic changes under different configurations; third, the dynamic system simulation is performed on the dynamic equation to find the optimal solution and generate a resource optimization dynamic model; finally, the system adopts the enhanced digital twin simulation technology to map the construction resource allocation to the virtual water conservancy construction environment, intuitively displays the construction efficiency and resource utilization, and generates the construction resource allocation optimization report.

[0082] 104. According to the latest optimization results in the water conservancy construction resource allocation optimization report, combined with the historical construction data in the water conservancy construction progress management framework, introduce the Lean Six Sigma methodology, set a Lean improvement plan for the latest optimization results, collect the project feedback of the Lean improvement plan to dynamically adjust the Lean improvement plan, and generate a water conservancy construction progress intelligent management plan.

[0083] In this step, the Lean Six Sigma methodology is a quality management tool aiming to improve product quality and customer satisfaction by reducing variation and improving processes.

[0084] The Lean improvement plan is an improvement plan formulated based on the latest optimization results, aiming to continuously improve construction efficiency and resource utilization.

[0085] The water conservancy construction progress intelligent management plan is the final management plan formulated through the Lean Six Sigma methodology by integrating the latest optimization results and historical construction data, ensuring the intelligent management and continuous optimization of the construction progress.

[0086] In the embodiments of the present application, first, according to the latest optimization results in the construction resource allocation optimization report, combined with historical construction data, the Lean Six Sigma methodology is introduced to formulate a lean improvement plan; second, project feedback on the lean improvement plan is collected to evaluate the actual implementation effect; third, the lean improvement plan is dynamically adjusted according to the feedback information to ensure its adaptability and effectiveness; finally, an intelligent management plan for the water conservancy construction progress is generated to provide comprehensive management guidance and support.

[0087] For example, continuing with the above example, assume that a large-scale water conservancy project has generated a construction resource allocation optimization report. First, the system, based on the latest optimization results in the report and combined with historical construction data, introduces the Lean Six Sigma methodology to formulate a lean improvement plan to ensure that each step conforms to the actual situation of the project; second, the system collects project feedback on the lean improvement plan, including actual construction data and employee opinions, to understand the implementation effect and the feelings of all parties; third, the lean improvement plan is dynamically adjusted according to the collected feedback information to ensure its adaptability and effectiveness; finally, the system generates an intelligent management plan for the water conservancy construction progress to provide comprehensive management guidance and support to ensure the intelligent management and continuous optimization of the construction progress.

[0088] To solve the problems of noise and periodic component identification in multi-source heterogeneous data processing, in some embodiments, the extraction of multi-source heterogeneous data in the water conservancy construction progress management framework in step 102 includes: extracting the multi-source heterogeneous data in the water conservancy construction progress management framework, performing data cleaning and noise filtering on the multi-source heterogeneous data, using time series analysis techniques to identify the periodic and trend components in the results of data cleaning and noise filtering, and generating a preprocessed multi-source heterogeneous data set; applying a multi-modal fusion network algorithm, and through the cross-modal attention mechanism in the multi-modal fusion network algorithm, performing cross-modal information integration on the preprocessed multi-source heterogeneous data set to focus on the mutual relationship between each data modality in the multi-source heterogeneous data, and generating a comprehensive multi-modal data representation; using a causal forest technique, and through the random forest extension in the causal forest technique, adding external environmental factors to the comprehensive multi-modal data representation, and evaluating the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay and generate a causal effect evaluation report; using the causal effect evaluation report as an input variable, applying a time series prediction model to deeply analyze the historical progress data of the water conservancy construction progress, and predicting the future construction progress according to the deep analysis results to identify the potential risk points and occurrence probabilities in the water conservancy construction progress and generate a water conservancy construction progress and risk warning report.

[0089] In this embodiment, the time series analysis technique is a method for identifying periodic and trend components in data, which can help understand the temporal characteristics of the data and predict future change trends.

[0090] The preprocessed multi-source heterogeneous data set is a data set generated after data cleaning and noise filtering. Further optimized through time series analysis, it provides high-quality basic data for subsequent analysis.

[0091] The comprehensive multi-modal data representation is a comprehensive and accurate data representation form formed by integrating cross-modal information from the preprocessed multi-source heterogeneous data, which helps in in-depth analysis and modeling.

[0092] The causal effect evaluation report is a detailed report generated by adding external environmental factors to the comprehensive multi-modal data representation and evaluating their impacts, revealing potential risk points and their causal effects.

[0093] The time series prediction model is a mathematical model for predicting future trends based on historical data. It can predict the future construction progress according to the in-depth analysis results, and identify potential risk points and occurrence probabilities.

[0094] In the embodiment of this application, first, extract the multi-source heterogeneous data in the water conservancy construction progress management framework, perform data cleaning and noise filtering processing, and use the time series analysis technique to identify the periodic and trend components in the results of data cleaning and noise filtering processing, and generate a preprocessed multi-source heterogeneous data set; secondly, apply the multi-modal fusion network algorithm to perform cross-modal information integration on the preprocessed multi-source heterogeneous data set through a cross-modal attention mechanism, paying attention to the mutual relationships between different data modalities, and generate a comprehensive multi-modal data representation; thirdly, adopt the causal forest technique to add external environmental factors to the comprehensive multi-modal data representation through random forest extension, evaluate the specific impacts and causal effects of external environmental factors on the construction progress, and generate a causal effect evaluation report; finally, use the causal effect evaluation report as an input variable, apply the time series prediction model to perform in-depth analysis on the historical progress data of the water conservancy construction progress, predict the future construction progress according to the in-depth analysis results, identify potential risk points and occurrence probabilities, and generate a water conservancy construction progress and risk warning report.

[0095] For example, in a large-scale water conservancy project, assume that the system needs to enhance its risk warning ability for the construction progress. First, the system extracts multi-source heterogeneous data including weather forecasts, material supply records, and equipment operation status from multiple data sources. Second, the system cleans and filters the noise of these data, uses time series analysis technology to identify periodic and trend components, and generates a high-quality preprocessed multi-source heterogeneous data set. Third, the multi-modal fusion network algorithm is used to integrate the preprocessed data through the cross-modal attention mechanism, highlighting the mutual relationship between different data modalities and generating a comprehensive multi-modal data representation. Finally, the system uses the causal forest technology to evaluate the specific impact of external environmental factors (such as weather changes and supply chain fluctuations) on the construction progress, generates a detailed causal effect evaluation report, and uses the time series prediction model to deeply analyze the future construction progress, identify potential risk points and their occurrence probabilities, and generate a water conservancy construction progress and risk warning report.

[0096] To solve the problems of time synchronization and feature redundancy among multi-source heterogeneous data, in some embodiments, the use of the multi-modal fusion network algorithm in step 102 includes: adopting data alignment technology, setting a specific time dimension, and synchronizing different-source data in the preprocessed multi-source heterogeneous data set on the specific time dimension to generate a synchronized multi-source data set; using the multi-modal fusion network algorithm, through the cross-modal attention mechanism in the multi-modal fusion network algorithm, performing cross-modal information integration on the synchronized multi-source data set and the preprocessed multi-source heterogeneous data set to focus on the mutual relationship between data modalities in the multi-source heterogeneous data and generate a cross-modal information set; through feature extraction technology, extracting cross-modal feature vectors from the cross-modal information set, applying clustering analysis to the cross-modal feature vectors, and performing clustering processing according to the association strength between the cross-modal feature vectors to generate cross-modal feature clusters; introducing the principal component analysis method and presetting a low-dimensional space, mapping the high-dimensional features in the cross-modal feature clusters to the low-dimensional space to reduce the data redundancy of the cross-modal feature clusters and generate a comprehensive multi-modal data representation.

[0097] In this embodiment, the data alignment technology refers to a method of synchronizing different-source data on a specific time dimension. By matching the timestamps of different data sources, it ensures that all data is analyzed within the same time frame, eliminating the impact of time differences.

[0098] The synchronized multi-source data set is a data set generated after setting a specific time dimension. After being processed by the data alignment technology, it ensures the time consistency of multi-source data and provides a reliable basis for subsequent analysis.

[0099] The cross-modal information set is a data set generated after integrating cross-modal information from synchronized multi-source data sets and preprocessed multi-source heterogeneous data sets. Through the cross-modal attention mechanism, the mutual relationships between different data modalities are highlighted, forming a more comprehensive information representation.

[0100] Feature extraction technology is a method for identifying and extracting meaningful features from data. By extracting features from the data in the cross-modal information set, cross-modal feature vectors are generated, which contain the key features of each modality data.

[0101] Cross-modal feature vectors are high-dimensional data representations generated from the cross-modal information set through feature extraction technology. Each vector contains key features from different data modalities, which helps to reveal the internal relationships between the data.

[0102] Cluster analysis is an unsupervised learning method that groups data points based on their similarity. By applying cluster analysis to the cross-modal feature vectors, clustering is performed according to the association strength between the feature vectors, generating cross-modal feature clusters to reveal the internal structure of the data.

[0103] Cross-modal feature clusters are data groupings generated by cluster analysis. Each cluster contains a set of cross-modal feature vectors with similar features, which can help identify patterns and trends in the data and provide a deeper understanding.

[0104] Principal component analysis method is a dimensionality reduction technique used to map high-dimensional data into a low-dimensional space, reducing data redundancy and retaining the main information.

[0105] Integrated multi-modal data representation is a data representation form generated by mapping the high-dimensional features in the cross-modal feature clusters into a low-dimensional space through the principal component analysis method, reducing data redundancy and improving the interpretability and processing efficiency of the data.

[0106] In the embodiments of this application, first, a data alignment technology is adopted to set a specific time dimension and synchronize the data from different sources in the preprocessed multi-source heterogeneous data set on the specific time dimension to generate a synchronized multi-source data set; second, a multi-modal fusion network algorithm is used to perform cross-modal information integration on the synchronized multi-source data set and the preprocessed multi-source heterogeneous data set through the cross-modal attention mechanism, focusing on the mutual relationships between different data modalities to generate a cross-modal information set; third, through feature extraction technology, cross-modal feature vectors are extracted from the cross-modal information set, and cluster analysis is applied to perform clustering according to the association strength between the cross-modal feature vectors to generate cross-modal feature clusters; finally, the principal component analysis method is introduced and a low-dimensional space is preset to map the high-dimensional features in the cross-modal feature clusters into a low-dimensional space, reducing data redundancy and generating an integrated multi-modal data representation.

[0107] The following is a specific example:

[0108] For example, in a large - scale hydropower station construction project, assume that the system needs to optimize the construction schedule management to cope with complex geological conditions and variable weather conditions. First, the system collects multi - source heterogeneous data from seismic monitoring, groundwater level changes, weather forecasts, etc., and uses data alignment technology to set a specific time dimension and synchronize this data to generate a synchronized multi - source dataset. Second, the system uses a multi - modal fusion network algorithm and, through a cross - modal attention mechanism, integrates cross - modal information between the synchronized multi - source dataset and the pre - processed multi - source heterogeneous dataset to generate a cross - modal information set. Third, the system extracts cross - modal feature vectors from the cross - modal information set through feature extraction technology and applies clustering analysis to cluster according to the correlation strength between feature vectors to generate cross - modal feature clusters. Finally, the system introduces the principal component analysis method to map the high - dimensional features in the cross - modal feature clusters to a low - dimensional space, reducing data redundancy and generating a comprehensive multi - modal data representation, thus providing more accurate risk warnings and resource allocation suggestions for project managers.

[0109] To solve the problem of specifically evaluating the impact of external environmental factors in water conservancy construction schedule management, in some embodiments, the causal forest technology adopted in step 102 includes: applying a feature selection algorithm to perform feature selection on the comprehensive multi - modal data representation, identifying significant impact features for the water conservancy construction schedule, separating the seasonal and random fluctuation components of the significant impact features through time - series decomposition technology to generate a refined feature set; adopting the causal forest technology, through the random forest extension in the causal forest technology, importing the refined feature set into the comprehensive multi - modal data representation and adding external environmental factors to evaluate the specific impact and causal effect of the external environmental factors on the water conservancy construction schedule in the water conservancy construction schedule management framework to warn of the risk of water conservancy construction schedule delay and generate a preliminary causal effect evaluation result; based on the preliminary causal effect evaluation result, introducing the Bayesian optimization algorithm to systematically search for the optimal hyperparameter combination in the evaluation process and applying sensitivity analysis to evaluate the risk levels of different environmental factor combinations in the optimal hyperparameter combination to generate an optimized causal effect evaluation model; applying the decision tree analysis method to classify and prioritize the key risk factors in the optimized causal effect evaluation model and formulating a risk avoidance plan according to the risk classification and prioritization results to generate a causal effect evaluation report.

[0110] In this embodiment, the feature selection algorithm is a method for identifying the most influential features from a dataset. By performing feature selection on the comprehensive multi - modal data representation, features that have a significant impact on the water conservancy construction schedule can be identified, thereby improving the accuracy and efficiency of subsequent analysis.

[0111] The refined feature set is a data set generated through feature selection algorithms and time series decomposition techniques. It contains significant impact features for the water conservancy construction progress, separates the seasonal and random fluctuation components of these features, and provides a more refined data basis.

[0112] The preliminary evaluation result of the causal effect is the result generated by applying the causal forest technique to the integrated multi-modal data representation after importing the refined feature set and adding external environmental factors. It reveals the specific impact and causal effect of external environmental factors on the water conservancy construction progress, providing a preliminary basis for risk warning.

[0113] The Bayesian optimization algorithm is an efficient hyperparameter optimization method used to systematically search for the optimal hyperparameter combination during the evaluation process. Through iterative optimization, it finds the best model configuration, improving the accuracy and reliability of the evaluation.

[0114] Sensitivity analysis is a technique for evaluating the impact of different variables on the model output. By performing sensitivity analysis on the risk levels of different combinations of environmental factors in the optimal hyperparameter combination, it is possible to better understand the degree of influence of each factor on the construction progress.

[0115] The optimized causal effect evaluation model is the final evaluation model generated through the Bayesian optimization algorithm and sensitivity analysis. It not only considers the impact of external environmental factors but also optimizes the model parameters, improving the precision and robustness of the evaluation.

[0116] The decision tree analysis method is a classification and regression tool used to classify and prioritize the key risk factors in the optimized causal effect evaluation model. It can help project managers formulate targeted risk avoidance plans to ensure that the construction progress is not delayed.

[0117] The causal effect evaluation report is the final report generated based on the optimized causal effect evaluation model and the decision tree analysis method. It details the key risk factors and their priorities and proposes specific risk avoidance measures, providing a scientific basis for project management.

[0118] In the embodiments of the present application, first, a feature selection algorithm is applied to perform feature selection on the comprehensive multi-modal data representation, identify the significant influencing features for the water conservancy construction progress, and separate the seasonal and random fluctuation components of the significant influencing features through time series decomposition technology to generate a refined feature set; second, the causal forest technology is adopted, through random forest expansion, the refined feature set is imported into the comprehensive multi-modal data representation and external environmental factors are added to evaluate the specific impact and causal effect of the external environmental factors on the water conservancy construction progress, and a preliminary causal effect evaluation result is generated; third, based on the preliminary causal effect evaluation result, the Bayesian optimization algorithm is introduced to systematically search for the optimal hyperparameter combination in the evaluation process, and sensitivity analysis is applied to evaluate the risk levels of different environmental factor combinations in the optimal hyperparameter combination to generate an optimized causal effect evaluation model; finally, the decision tree analysis method is applied to classify and prioritize the key risk factors in the optimized causal effect evaluation model, formulate a risk avoidance plan according to the risk classification and prioritization results, and generate a causal effect evaluation report.

[0119] The following is a specific example:

[0120] For example, in a large-scale water diversion project, assume that the system needs to accurately evaluate the impact of external environmental factors on the construction progress to optimize resource allocation and risk control. First, the system applies a feature selection algorithm to identify the features that have a significant impact on the construction progress from the comprehensive multi-modal data representation, and separates the seasonal and random fluctuation components of these features through time series decomposition technology to generate a refined feature set; second, the system adopts the causal forest technology, through random forest expansion, imports the refined feature set into the comprehensive multi-modal data representation and adds external environmental factors (such as rainfall, temperature change) to evaluate the specific impact and causal effect of these factors on the construction progress, and generates a preliminary causal effect evaluation result; third, based on the preliminary evaluation result, the system introduces the Bayesian optimization algorithm to systematically search for the optimal hyperparameter combination in the evaluation process, and applies sensitivity analysis to evaluate the risk levels of different environmental factor combinations to generate an optimized causal effect evaluation model; finally, the system applies the decision tree analysis method to classify and prioritize the key risk factors in the optimized causal effect evaluation model, formulates a specific risk avoidance plan, and generates a detailed causal effect evaluation report to ensure the smooth progress of the construction progress and the effective utilization of resources.

[0121] To solve the optimization problem of resource allocation and scheduling in water conservancy construction progress management, in some embodiments, the application of the graph neural differential equation algorithm in step 103 includes introducing an anomaly detection algorithm to label the anomalies in the water conservancy construction progress and risk warning report, and using fault tree analysis to construct a causal relationship graph for the anomalies to reveal the root causes of the anomalies and generate a causal relationship graph of anomaly events; applying the graph neural differential equation algorithm, with the causal relationship graph of anomaly events as a constraint condition, establishing a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, performing dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation and generate a resource optimization dynamic model; adopting an enhanced digital twin simulation technology to construct a virtual water conservancy construction environment according to the resource optimization dynamic model, mapping the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation and generate a virtual construction simulation environment; applying multi-criteria decision analysis to quantitatively evaluate the simulation data in the virtual construction simulation environment, and optimizing the construction efficiency and resource utilization of the construction resource allocation according to the quantitative evaluation results to generate a water conservancy construction resource allocation optimization report.

[0122] In this embodiment, the anomaly detection algorithm is a technology for identifying outliers and unusual patterns in a dataset, which can help the system timely detect and label the anomalies in the construction progress and risk warning report, ensuring the accuracy and reliability of subsequent analysis.

[0123] Fault tree analysis is a system safety analysis method that reveals the root causes of anomalies by constructing a causal relationship graph.

[0124] The causal relationship graph of anomaly events is a chart generated by fault tree analysis, which shows the root causes of anomalies and their propagation paths. Such a graph can intuitively reveal the causes of anomalies and provide support for formulating countermeasures.

[0125] The resource optimization dynamic model is a model generated after performing dynamic system simulation on the dynamic equation, which reveals the best path for construction resource allocation and scheduling. This model takes into account various constraint conditions and provides an optimized resource allocation plan.

[0126] Multi-criteria decision analysis is a comprehensive evaluation method that selects the best resource allocation plan by quantitatively evaluating multiple indicators. It can balance the conflicts between different objectives and ensure the comprehensiveness and rationality of the optimization results.

[0127] The water conservancy construction resource allocation optimization report is the final report generated based on the simulation data in the virtual construction simulation environment. It details the optimized construction resource allocation plan and provides scientific decision-making support.

[0128] In the embodiments of the present application, first, an anomaly detection algorithm is introduced to label the anomalies in the water conservancy construction progress and risk early warning report, and a fault tree analysis is used to construct a causal relationship diagram for the anomalies to reveal the root causes of the anomalies and generate a causal relationship diagram of anomaly events; second, the graph neural differential equation algorithm is used to establish a dynamic equation for the construction resource allocation and scheduling problem with the causal relationship diagram of anomaly events as the constraint condition, perform dynamic system simulation on the dynamic equation to find the optimal solution, and generate a resource optimization dynamic model; third, an enhanced digital twin simulation technology is adopted to construct a virtual water conservancy construction environment according to the resource optimization dynamic model, map the construction resource allocation to the virtual environment, capture the construction efficiency and resource utilization, and generate a virtual construction simulation environment; finally, multi-criteria decision analysis is applied to quantitatively evaluate the simulation data in the virtual construction simulation environment, and optimize the construction efficiency and resource utilization of the construction resource allocation according to the quantitative evaluation results to generate a water conservancy construction resource allocation optimization report.

[0129] The following is a specific example:

[0130] For example, in a large irrigation project, assume that the system needs to optimize the resource allocation and scheduling in the construction progress management to improve the construction efficiency and resource utilization rate. First, the system introduces an anomaly detection algorithm to label the anomalies in the construction progress and risk early warning report, and constructs a causal relationship diagram of anomaly events using fault tree analysis to reveal the root causes of the anomalies; second, the system uses the graph neural differential equation algorithm to establish a dynamic equation with the causal relationship diagram of anomaly events as the constraint condition, performs dynamic system simulation on the dynamic equation to find the optimal solution, and generates a resource optimization dynamic model; third, the system adopts an enhanced digital twin simulation technology to construct a virtual water conservancy construction environment according to the resource optimization dynamic model, maps the construction resource allocation to the virtual environment, captures the construction efficiency and resource utilization, and generates a virtual construction simulation environment; finally, the system applies multi-criteria decision analysis to quantitatively evaluate the simulation data in the virtual construction simulation environment, and optimizes the construction efficiency and resource utilization of the construction resource allocation according to the quantitative evaluation results to generate a detailed water conservancy construction resource allocation optimization report, ensuring the efficient progress of the project and the effective utilization of resources.

[0131] To address the complexity of resource allocation and scheduling in the water conservancy construction progress and risk warning report, in some embodiments, the application of the graph neural differential equation algorithm in step 103 includes: extracting key node and edge information from the abnormal event causal relationship graph, applying the shortest path algorithm in graph theory to calculate the maximum bearing capacity of each node in the key node and edge information, and generating the maximum bearing capacity of the path; using the maximum bearing capacity of the path as an input variable to construct the graph structure network of the water conservancy construction progress, applying the graph neural differential equation algorithm to the graph structure network, establishing a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report with the abnormal event causal relationship graph as a constraint condition, performing dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and generating a construction progress dynamic model; training an intelligent agent using reinforcement learning technology, introducing construction safety and quality assurance as additional constraint conditions, and having the intelligent agent learn the optimal resource allocation strategy in the construction progress dynamic model to generate the optimal resource allocation strategy for the construction progress; using a mathematical programming method for the construction progress dynamic model to solve for the optimal parameters that satisfy the optimal resource allocation strategy for the construction progress and the maximum bearing capacity of the path, and generating a resource optimization dynamic model based on the optimal parameters.

[0132] In this embodiment, the abnormal event causal relationship graph is a graphical tool constructed through fault tree analysis. It depicts all possible paths from the starting event to the final result and is used to identify and understand the root causes leading to construction problems.

[0133] The key node and edge information refers to the elements in the abnormal event causal relationship graph that have a decisive impact on the system behavior. These nodes and edges represent the most sensitive or critical links in the construction process, and their maximum bearing capacity determines the performance limit of the entire system.

[0134] The shortest path algorithm is a graph theory algorithm used to find the shortest path between two nodes. In this solution, it is used to calculate the maximum bearing capacity between key nodes to ensure that resource allocation does not exceed the maximum tolerance of any single point.

[0135] The graph structure network is an abstract model used to represent various associations and dependencies in the water conservancy construction progress. It maps different tasks, activities, and their interactions in the construction process into a graph, facilitating analysis using advanced algorithms.

[0136] The construction progress dynamic model is a simulation model established based on a dynamic equation. It can reflect the changes in the construction progress in real time and help managers make more accurate decisions.

[0137] Reinforcement learning technology is a machine learning method that optimizes the behavior of an intelligent agent by having it perform actions in an environment and adjust its strategy based on feedback. In this context, the intelligent agent is trained to learn the optimal resource allocation strategy while considering construction safety and quality assurance as additional constraints.

[0138] Mathematical programming methods are a series of techniques aimed at finding the optimal solution under specific conditions, such as linear programming, non - linear programming, etc. It is used to solve the optimal parameters in the construction progress dynamic model under the conditions of the best resource allocation strategy and the maximum carrying capacity of the path.

[0139] In the embodiments of this application, first, extract the key node and edge information from the causal relationship graph of abnormal events, apply the shortest path algorithm in graph theory to calculate the maximum carrying capacity of each node, and generate the maximum carrying capacity of the path; second, use the maximum carrying capacity of the path as an input variable to construct a graph - structured network for the water conservancy construction progress, apply the graph neural differential equation algorithm to the graph - structured network, establish a dynamic equation for the construction resource allocation and scheduling problem with the causal relationship graph of abnormal events as a constraint condition, and perform dynamic system simulation to generate a construction progress dynamic model; third, use reinforcement learning technology to train the intelligent agent, introduce construction safety and quality assurance as additional constraints, so that the intelligent agent learns the best resource allocation strategy in the construction progress dynamic model and generates the best resource allocation strategy for the construction progress; finally, use the mathematical programming method for the construction progress dynamic model to solve the optimal parameters that meet the conditions of the best resource allocation strategy for the construction progress and the maximum carrying capacity of the path, and generate a resource optimization dynamic model.

[0140] The following is a specific example:

[0141] Suppose in a project of building a flood control dyke in a city, to further improve the intelligent level of construction progress management, first, engineers extract key node and edge information from the causal relationship diagram of abnormal events constructed previously, and use the shortest path algorithm to calculate the maximum bearing capacity of these nodes, and generate the maximum bearing capacity of the path based on this; second, they use the obtained maximum bearing capacity of the path as an input variable to construct the graph structure network of this project, and apply the graph neural differential equation algorithm to this network, combine the causal relationship diagram of abnormal events as a constraint condition, establish a dynamic equation, conduct dynamic system simulation, and generate a detailed dynamic model of construction progress; third, the team trains an intelligent agent using reinforcement learning technology, taking construction safety and quality assurance into consideration, so that this intelligent agent can learn in the dynamic model of construction progress and propose the best resource allocation strategy; finally, they use the mathematical programming method to solve the optimal parameters according to the best resource allocation strategy proposed by the intelligent agent and the condition of the maximum bearing capacity of the path, so as to generate the final resource optimization dynamic model, ensuring the smooth progress of the project and the effective utilization of resources.

[0142] To solve the problems of the application and continuous improvement of the resource allocation optimization results in water conservancy construction progress management, in some embodiments, according to the latest optimization results in the water conservancy construction resource allocation optimization report in step 104, combined with historical construction data, the lean six sigma methodology is introduced, including: using the differential analysis method, combined with the historical construction data in the water conservancy construction progress management framework, to conduct differential analysis on the latest optimization results in the water conservancy construction resource allocation optimization report, quantify the contribution of the construction period shortening of the resource optimization strategy in the water conservancy construction resource allocation optimization report, and generate a resource optimization strategy contribution table; introducing the lean six sigma methodology, taking the strategy contribution value in the resource optimization strategy contribution table as a reference value, defining improving construction efficiency and reducing delay risk as improvement goals, and generating a comprehensive improvement goal; designing a lean improvement plan according to the comprehensive improvement goal, verifying the effectiveness of the lean improvement plan by running a pilot project, and collecting feedback during the process of running the pilot project to dynamically adjust the lean improvement plan, and generating a pilot evaluation result; applying discrete event simulation technology to set multiple simulation scenarios, evaluating the execution effect of the water conservancy construction progress management framework under the multiple simulation scenarios, introducing a feedback mechanism, regularly collecting the simulation prediction values in the execution effect, and conducting comparative analysis with the comprehensive improvement goal as a standard, and regularly adjusting the latest optimization results to generate a water conservancy construction progress intelligent management plan.

[0143] In this embodiment, the differential analysis method is a technique used to compare data changes at different time points or under different conditions.

[0144] The resource optimization strategy contribution table is a table generated by performing differential analysis on the latest optimization results in the water conservancy construction resource allocation optimization report. It quantifies the specific contributions of various resource optimization strategies to shortening the construction period and provides a clear basis for performance evaluation.

[0145] The comprehensive improvement goal is a goal defined based on the strategy contribution values in the resource optimization strategy contribution table, specifically including two aspects: improving construction efficiency and reducing delay risks. These goals guide the design of subsequent lean improvement plans.

[0146] Running a pilot project means implementing a lean improvement plan on a small scale to verify its feasibility and effectiveness. The success or failure of the pilot project is directly related to whether the plan can be promoted on a larger scale.

[0147] Discrete event simulation technology is a method for simulating the occurrence process of discrete events within a system and is applicable to evaluating the performance of complex systems. It is used to set up multiple simulation scenarios to evaluate the implementation effect of the water conservancy construction progress management framework.

[0148] The simulation prediction value is a predicted value of the future state generated during the discrete event simulation process. Regularly collecting these prediction values and comparing them with the actual implementation situation helps to evaluate the effect of improvement measures and make necessary adjustments.

[0149] In the embodiments of this application, first, using the differential analysis method and combining with the historical construction data in the water conservancy construction progress management framework, perform differential analysis on the latest optimization results in the water conservancy construction resource allocation optimization report, quantify the contributions of resource optimization strategies to shortening the construction period, and generate a resource optimization strategy contribution table; second, introduce the lean six sigma methodology, use the strategy contribution values in the resource optimization strategy contribution table as reference values, define improving construction efficiency and reducing delay risks as improvement goals, and generate comprehensive improvement goals; third, design a lean improvement plan according to the comprehensive improvement goals, verify its effectiveness by running a pilot project, and collect feedback during the pilot project to dynamically adjust the lean improvement plan to generate pilot evaluation results; finally, apply discrete event simulation technology to set up multiple simulation scenarios, evaluate the implementation effect of the water conservancy construction progress management framework under each simulation scenario, introduce a feedback mechanism, regularly collect the simulation prediction values in the implementation effect, conduct comparative analysis based on the comprehensive improvement goals, and regularly adjust the latest optimization results to generate an intelligent management plan for water conservancy construction progress.

[0150] The following is a specific example:

[0151] For example, in a large reservoir expansion project, assume that the system needs to further optimize the construction schedule management to ensure the effectiveness of resource allocation and the improvement of construction efficiency. First, the engineers used the differential analysis method and combined with historical construction data to conduct a detailed analysis of the latest optimization results in the water conservancy construction resource allocation optimization report, quantified the specific contributions of various resource optimization strategies to the shortening of the construction period, and generated a resource optimization strategy contribution table. Second, the team introduced the Lean Six Sigma methodology, set comprehensive improvement goals for improving construction efficiency and reducing delay risks based on the strategy contribution values in the resource optimization strategy contribution table. Third, they designed a Lean improvement plan according to the comprehensive improvement goals, verified the effectiveness of the plan through a small-scale pilot project, collected feedback information during the pilot project process, dynamically adjusted the Lean improvement plan, and generated pilot evaluation results. Finally, the team applied discrete event simulation technology to set up multiple different simulation scenarios, evaluated the implementation effects of the water conservancy construction schedule management framework under these scenarios, regularly collected simulation prediction values, compared and analyzed them with the actual implementation situation, and made necessary adjustments according to the comprehensive improvement goals, and finally generated a detailed intelligent management plan for water conservancy construction schedule, ensuring the efficient promotion and continuous optimization of the project.

[0152] This application considers that in order to solve the problem of cross-modal information integration of multi-source heterogeneous data in water conservancy construction schedule management, in the prior art, because it is difficult to effectively capture the information correlation between different data modalities, the invention embodiment proposes this alternative solution. To solve the technical problems of information loss and incoherence in the process of multi-source data fusion, a new alternative solution is proposed, and this solution includes:

[0153] Using the multi-modal fusion network algorithm, through the cross-modal attention mechanism in the multi-modal fusion network algorithm, perform cross-modal information integration on the synchronous multi-source data set and the preprocessed multi-source heterogeneous data set to focus on the mutual relationship between each data modality in the multi-source heterogeneous data, and generate a cross-modal information set, including:

[0154] Perform noise removal processing on the synchronous multi-source data set through an adaptive filter, introduce mutual information analysis, evaluate the data before and after dependence relationship in the noise removal processing result, introduce a cross-modal attention mechanism to quantify the data before and after dependence relationship, and generate attention weights;

[0155] Calculate the attention weights through the following formula:

[0156]

[0157] where a ij is the attention weight of the i-th data modality to the j-th data modality; W a is the attention weight matrix; hi and h j are the hidden state vectors of the i-th and j-th data modalities in the synchronous multi-source dataset and the preprocessed multi-source heterogeneous dataset respectively; U a is a linear transformation matrix; b a is a bias term used to adjust the activation function threshold; V is a non-linear transformation matrix; c a is a non-linear bias term used to adjust the starting point of the non-linear function; Z a is a feature mapping matrix; N is the number of data modalities; j is the index of the data modality, ranging from 1 to N - 1;

[0158] Apply the attention weights to the cross-modal information integration process for weighted summation, introduce a non-linear activation function to enhance the feature expression ability in the weighted summation process, and through the residual connection mechanism, improve the information transfer consistency in the weighted summation process to generate a cross-modal information integration feature representation;

[0159] Calculate the cross-modal information integration feature representation through the following formula:

[0160]

[0161] where, f i is the cross-modal information integration feature representation of the i-th data modality; a ij is the attention weight of the i-th data modality to the j-th data modality; W f is a feature transformation matrix; b f is a bias term used to adjust the feature transformation baseline; γ is a non-linear adjustment coefficient used to control the contribution of the non-linear part; V is a non-linear transformation matrix; c f is a non-linear activation bias term used to adjust the starting point of the non-linear function after feature transformation; σ is an activation function; α is a logarithmic transformation coefficient; M f is a logarithmic transformation matrix; d f is a logarithmic transformation bias term; N is the number of data modalities; h i and h j are the hidden state vectors of the i-th and j-th data modalities in the synchronous multi-source dataset and the preprocessed multi-source heterogeneous dataset respectively; j is the index of the data modality, ranging from 1 to N - 1;

[0162] Use a variational autoencoder to perform probability modeling on the cross-modal information integration feature representation, capture the complex interaction relationships between the cross-modal information integration feature representations, so as to strengthen the expression coherence of the cross-modal information in the cross-modal information integration feature representation and generate a cross-modal information set.

[0163] This method aims to improve the accuracy and reliability of cross-modal information integration. By combining adaptive filters, mutual information analysis, and cross-modal attention mechanisms, it ensures the extraction of the most valuable information from multi-source heterogeneous data and captures the potential relationships between these information through complex mathematical models.

[0164] Suppose in the construction of a large water conservancy project bridge, the system needs to optimize the construction progress management to ensure the effectiveness of resource allocation and the improvement of construction efficiency. Suppose W a = 0.8, U a = 0.6, b a = -0.2, V = 0.5, Z a = 0.4, c a = -0.3, and select h i and h j as two typical data modality hidden state vectors;

[0165]

[0166] Suppose W f = 0.9, b f = 0.1, γ = 0.5, V = 0.4, c f = -0.1, α = 0.6, M f = 0.3, d f = -0.2;

[0167]

[0168] Suppose the set threshold is 0.8. Since the calculation result 0.82 is greater than the set threshold, it indicates that the intelligent management scheme for the water conservancy construction progress has high accuracy and reliability, and can ensure the effective monitoring of the construction progress and the optimal allocation of resources. This is because the higher attention weight index reflects that the important relevance between different data modalities has been fully considered, and the result of the cross-modal information integration feature representation also verifies the effectiveness of this relevance. Through the above steps, the precise management of the water conservancy construction progress and the efficient utilization of resources are ensured, the scientific nature and response speed of the entire project are improved, and the overall quality control level of the project is enhanced.

[0169] This application considers that in order to solve the optimization problem of resource allocation and scheduling in water conservancy construction progress management, in the existing technology, due to the failure to fully consider the maximum carrying capacity of each path and the impact of abnormal events during path planning and resource allocation, problems such as low resource utilization efficiency and increased technical risks occur. Therefore, the invention embodiment proposes this alternative solution to solve these problems and ensure the effective monitoring of the construction progress and the optimal allocation of resources. Therefore, a new alternative solution is proposed, and this solution includes:

[0170] Using the maximum load capacity of the path as the input variable, construct the graph structure network of the water conservancy construction progress, apply the graph neural differential equation algorithm to the graph structure network, establish a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk warning report with the causal relationship graph of abnormal events as the constraint condition, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and generate a dynamic model of construction progress, including:

[0171] Apply network traffic analysis technology to evaluate the current load conditions of each path in the maximum load capacity of the path, re-plan the data flow of each path through the topology optimization algorithm, and improve the maximum load range between all nodes and neighbor nodes of each path to generate an information aggregation result;

[0172] Calculate the information aggregation result through the following formula:

[0173]

[0174] where, z (t+1) is the information aggregation result at the next moment; n is the index of the neighbor node; P is the set of neighbor nodes; A is the aggregation weight matrix used to adjust the importance of information from neighbors; w n is the weight of each neighbor node; is the state of the neighbor node at the current moment; B is the self-feedback weight matrix used to maintain the state of the node itself; y (t) is the state of the node at the current moment; c is the bias term used to adjust the threshold of the activation function; σ is the activation function; t is the time variable;

[0175] Define a cost function with the information aggregation result and the cost-benefit ratio in the process of re-planning the data flow, and explore all possible path combinations in the process of re-planning the data flow through the iterative search algorithm until the algorithm converges to generate an optimal solution path;

[0176] Calculate the optimal solution path through the following formula:

[0177]

[0178] where, s(t) is the optimal solution path; T is the total time; L(q(t), v(t), t) is the loss function; λ is the regularization coefficient; q(t) is the system state vector; v(t) is the control input vector; g(q(t), v(t), t) is the state transition function; η(t) is the external disturbance term; Φ(q(T)) is the terminal cost function; is the rate of change of the state; μ is the information aggregation result balance coefficient; z (t) is the information aggregation result at the current moment; is the information aggregation result based on system state prediction;

[0179] Apply the finite element analysis method to simulate the soil settlement and structural stress changes during the water conservancy construction process based on the optimal solution path, introduce an expert system rule engine to dynamically adjust the task scheduling parameters in the optimal solution path, and import the dynamic adjustment results into the kinetic equation to generate a dynamic construction progress model.

[0180] This method aims to improve the accuracy and reliability of water conservancy construction progress management. By combining network traffic analysis technology, topology optimization algorithm, iterative search algorithm, finite element analysis method and expert system rule engine, it captures the non-linear relationship between various types of construction paths and resource allocation, and enhances the sensitivity to construction environment changes through a dynamic adjustment mechanism, ultimately achieving accurate construction progress prediction and optimal resource allocation.

[0181] In the information aggregation result, the aggregation weight matrix term is used to adjust the importance of information from neighbors to ensure that key information is taken seriously; the self-feedback weight matrix term B·y (t) : is used to maintain the state of the node itself, prevent information loss, and ensure that the data features learned by the model are more complete; the bias term c: is used to adjust the threshold of the activation function to ensure output stability;

[0182] Among them, A, B, w n , c: are optimized through the model training stage. The specific method is to minimize the loss function using the backpropagation algorithm; and y (t) are obtained through real-time monitoring or historical data analysis; P is determined according to the actual application scenario;

[0183] In the optimal solution path, the loss function term measures the cost of the system state and control input changing over time; the regularization term ensures that the state transition conforms to physical laws and at the same time reduces the impact of external disturbances; the information aggregation result balance term μ· maintains the consistency of the information aggregation result and ensures the accuracy of the prediction result; the terminal cost function term Φ(q(T)): evaluates the cost of the final state and ensures that the terminal goal of the system is achieved;

[0184] Among them, L(q(t), v(t), t) is obtained through simulation or actual measurement; λ, μ are optimized through the model training stage; q(t), v(t), g(q(t), v(t), t), η(t), are set according to the actual application scenario; z (t) , From the information aggregation results and the information aggregation results based on system state prediction; Φ(q(T)) is set according to project requirements;

[0185] In a dam construction project of a medium-sized hydropower station, it is necessary to optimize the construction schedule management and ensure the effectiveness of resource allocation; assume A = 0.6, B = 0.4, w n = 0.5, and the bias term c = 0.2; select two typical data modal state vectors (representing the current load situation of each path) and y (t) (the state of the node itself);

[0186]

[0187] Assume λ = 0.5, μ = 0.3; and given L(q(t), v(t), t), η(t), Φ(q(T)), z (t) ,

[0188]

[0189] Set the threshold to 0.9. Since the calculation result s(t) = 0.91 is greater than the set threshold, this indicates that the application of the intelligent management scheme for water conservancy construction progress in dam construction is effective. Because a result higher than the threshold means that the system can accurately capture the key information during the construction process and generate the optimal solution path accordingly, thus ensuring the rationality of resource allocation and the efficiency of construction progress. At the same time, such a conclusion also proves the reliability and superiority of the model in dealing with complex engineering problems, which helps to improve the management level and technical content of the entire project. Through the above steps, the intelligent management and optimal allocation of the construction progress of the dam construction project are realized, which not only improves work efficiency but also effectively reduces potential risks and ensures project quality.

[0190] Figure 2 The structure diagram of an intelligent management system for water conservancy construction progress is provided for the embodiments of this application. As Figure 2 shown, the device includes:

[0191] A collection module 21, configured to collect the water conservancy project plan, obtain the dynamic changes during the life cycle of the water conservancy project by integrating building information model technology, provide a comprehensive perspective for capturing the dynamic changes, and generate a water conservancy construction progress management framework;

[0192] An integration module 22 is used to extract multi-source heterogeneous data in the water conservancy construction progress management framework, apply a multi-modal fusion network algorithm, and perform cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data. The causal forest technology is adopted, and external environmental factors are added through the random forest expansion in the causal forest technology to evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn of the risk of water conservancy construction progress delay and generate a water conservancy construction progress and risk warning report;

[0193] A simulation module 23 is used to apply the graph neural differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and adopt an enhanced digital twin simulation technology to construct a virtual water conservancy construction environment, and map the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation, and generate a water conservancy construction resource allocation optimization report;

[0194] An adjustment module 24 is used to, according to the latest optimization result in the water conservancy construction resource allocation optimization report, combine the historical construction data in the water conservancy construction progress management framework, introduce the lean six sigma methodology, set a lean improvement plan for the latest optimization result, collect project feedback of the lean improvement plan to dynamically adjust the lean improvement plan, and generate a water conservancy construction progress intelligent management plan.

[0195] Figure 2 The described intelligent water conservancy construction progress management system can execute Figure 1 The intelligent water conservancy construction progress management method described in the illustrated embodiment, and its implementation principle and technical effects will not be elaborated further. For the intelligent water conservancy construction progress management system in the above embodiment, the specific ways for each module and unit to execute operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0196] In a possible design, Figure 2 The intelligent water conservancy construction progress management system of the illustrated embodiment can be implemented as a computing device, as Figure 3 shown, and this computing device can include a storage component 31 and a processing component 32;

[0197] The storage component 31 stores one or more computer instructions, and among them, the one or more computer instructions are called and executed by the processing component 32.

[0198] The processing component 32 is used for: collecting the planning of water conservancy project, obtaining the dynamic changes during the life cycle of the water conservancy project by integrating building information modeling technology, providing a comprehensive perspective for capturing the dynamic changes, and generating a water conservancy construction progress management framework; extracting multi-source heterogeneous data from the water conservancy construction progress management framework, applying a multi-modal fusion network algorithm, and performing cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multi-modal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data, adopting the causal forest technology, adding external environmental factors through the random forest expansion in the causal forest technology, evaluating the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework to warn of the risk of water conservancy construction progress delay, and generating a water conservancy construction progress and risk warning report; applying the graph neural differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, performing dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, adopting an enhanced digital twin simulation technology to construct a virtual water conservancy construction environment, and mapping the construction resource allocation to the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource allocation, and generating a water conservancy construction resource allocation optimization report; according to the latest optimization result in the water conservancy construction resource allocation optimization report, combining the historical construction data in the water conservancy construction progress management framework, introducing the lean six sigma methodology, setting a lean improvement plan for the latest optimization result, collecting the project feedback of the lean improvement plan to dynamically adjust the lean improvement plan, and generating a water conservancy construction progress intelligent management plan.

[0199] Among them, the processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above method.

[0200] The storage component 31 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0201] Of course, the computing device will necessarily also include other components, such as input / output interfaces, display components, communication components, etc.

[0202] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module can be an output device, an input device, etc.

[0203] The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0204] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. At this time, the computing device can refer to a cloud server, and the above processing component, storage component, etc. can be basic server resources rented or purchased from a cloud computing platform.

[0205] The embodiment of the present application also provides a computer storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the above Figure 1 shown embodiment of a method for intelligent management of water conservancy construction progress.

[0206] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0207] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement without creative efforts.

[0208] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0209] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for intelligent management of water conservancy construction progress, characterized in that: include: Collect water conservancy project plans, obtain dynamic changes in the life cycle of the water conservancy project through integrated building information modeling technology, provide a comprehensive perspective to capture the dynamic changes, and generate a water conservancy construction progress management framework; Extract multi-source heterogeneous data in the water conservancy construction progress management framework, use a multimodal fusion network algorithm, and integrate the multi-source heterogeneous data through the cross-modal attention mechanism in the multimodal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data. Use causal forest technology to add external environmental factors through the random forest extension in the causal forest technology to evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn of the risk of delay in the water conservancy construction progress and generate a water conservancy construction progress and risk warning report; Using the graph neural differential equation algorithm, a dynamic equation is established for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk warning report, and a dynamic system simulation is performed on the dynamic equation to find the optimal solution of the dynamic equation. The virtual water conservancy construction environment is constructed using enhanced digital twin simulation technology, and the construction resource configuration is mapped into the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource configuration, and generate a water conservancy construction resource configuration optimization report; According to the latest optimization results in the water conservancy construction resource allocation optimization report, combined with the historical construction data in the water conservancy construction progress management framework, the Lean Six Sigma methodology is introduced, a lean improvement plan is set for the latest optimization results, and project feedback of the lean improvement plan is collected to dynamically adjust the lean improvement plan to generate an intelligent management plan for the water conservancy construction progress.

2. The method according to claim 1, characterized in that The multi-source heterogeneous data in the water conservancy construction progress management framework is extracted, and a multimodal fusion network algorithm is used to integrate the multi-source heterogeneous data through the cross-modal attention mechanism in the multimodal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data. The causal forest technology is used to add external environmental factors through the random forest extension in the causal forest technology to evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn of the risk of delay in the water conservancy construction progress and generate a water conservancy construction progress and risk warning report, including: Extracting multi-source heterogeneous data in the water conservancy construction progress management framework, performing data cleaning and noise filtering on the multi-source heterogeneous data, using time series analysis technology to identify periodicity and trend components in the data cleaning and noise filtering processing results, and generating a pre-processed multi-source heterogeneous data set; Using a multimodal fusion network algorithm, cross-modal information integration is performed on the pre-processed multi-source heterogeneous data set through a cross-modal attention mechanism in the multimodal fusion network algorithm, so as to focus on the relationship between each data modality in the multi-source heterogeneous data and generate a comprehensive multimodal data representation; Using causal forest technology, through the random forest extension in the causal forest technology, external environmental factors are added to the comprehensive multimodal data representation, and the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework are evaluated to warn of the risk of delay in the water conservancy construction progress and generate a causal effect evaluation report; The causal effect assessment report is used as an input variable, and a time series prediction model is applied to conduct an in-depth analysis of the historical progress data of the water conservancy construction progress. Based on the in-depth analysis results, future construction progress is predicted to identify potential risk points and occurrence probabilities in the water conservancy construction progress, and generate a water conservancy construction progress and risk warning report.

3. The method according to claim 2, characterized in that The multimodal fusion network algorithm is used to integrate cross-modal information of the pre-processed multi-source heterogeneous data set through a cross-modal attention mechanism in the multimodal fusion network algorithm, so as to focus on the mutual relationship between each data modality in the multi-source heterogeneous data and generate a comprehensive multimodal data representation, including: Using data alignment technology, setting a specific time dimension, synchronously processing different source data in the preprocessed multi-source heterogeneous data set on the specific time dimension, and generating a synchronized multi-source data set; Using a multimodal fusion network algorithm, through a cross-modal attention mechanism in the multimodal fusion network algorithm, cross-modal information integration is performed on the synchronized multi-source data set and the pre-processed multi-source heterogeneous data set, so as to focus on the mutual relationship between each data modality in the multi-source heterogeneous data and generate a cross-modal information set; Extracting a cross-modal feature vector from the cross-modal information set by using a feature extraction technique, applying cluster analysis to the cross-modal feature vector, performing clustering processing according to the correlation strength between the cross-modal feature vectors, and generating a cross-modal feature cluster; The principal component analysis method is introduced and a low-dimensional space is preset to map the high-dimensional features in the cross-modal feature cluster to the low-dimensional space, thereby reducing the data redundancy of the cross-modal feature cluster and generating a comprehensive multimodal data representation.

4. The method according to claim 2, characterized in that: The causal forest technology is adopted, and the random forest extension in the causal forest technology is used to add external environmental factors to the comprehensive multimodal data representation, and the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework are evaluated to warn of the risk of delay in the water conservancy construction progress and generate a causal effect evaluation report, including: Applying a feature selection algorithm to perform feature selection on the comprehensive multimodal data representation, identifying the significant influencing features on the progress of the water conservancy construction, separating the seasonality and random fluctuation components of the significant influencing features through time series decomposition technology, and generating a refined feature set; Using causal forest technology, through the random forest extension in the causal forest technology, the refined feature set is introduced into the comprehensive multimodal data representation, and external environmental factors are added to evaluate the specific impact and causal effect of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn the risk of delay in the water conservancy construction progress and generate preliminary causal effect evaluation results; Based on the preliminary causal effect evaluation results, a Bayesian optimization algorithm is introduced to systematically search for the optimal hyperparameter combination in the evaluation process, and a sensitivity analysis is applied to evaluate the risk level of different environmental factor combinations in the optimal hyperparameter combination to generate an optimized causal effect evaluation model; The decision tree analysis method is applied to classify and prioritize the key risk factors in the optimized causal effect assessment model, and a risk avoidance plan is formulated based on the risk classification and priority ranking results to generate a causal effect assessment report.

5. The method according to claim 1, characterized in that The graph neural differential equation algorithm is used to establish a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk warning report, and a dynamic system simulation is performed on the dynamic equation to find the optimal solution of the dynamic equation. The virtual water conservancy construction environment is constructed using enhanced digital twin simulation technology, and the construction resource configuration is mapped into the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource configuration, and generate a water conservancy construction resource configuration optimization report, including: An anomaly detection algorithm is introduced to mark the abnormal situations in the water conservancy construction progress and risk warning report, and a cause-effect relationship diagram is constructed for the abnormal situation using fault tree analysis to reveal the root cause of the abnormal situation and generate an abnormal event cause-effect relationship diagram; Using the graph neural differential equation algorithm, taking the abnormal event causal relationship graph as a constraint condition, a dynamic equation is established for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk warning report, and a dynamic system simulation is performed on the dynamic equation to find the optimal solution of the dynamic equation and generate a resource optimization dynamic model; Adopting enhanced digital twin simulation technology, constructing a virtual water conservancy construction environment according to the resource optimization dynamics model, mapping the construction resource configuration into the virtual water conservancy construction environment, so as to capture the construction efficiency and resource utilization of the construction resource configuration, and generate a virtual construction simulation environment; Multi-criteria decision analysis is applied to quantitatively evaluate the simulation data in the virtual construction simulation environment, and the construction efficiency and resource utilization of the construction resource configuration are optimized according to the quantitative evaluation results to generate a water conservancy construction resource configuration optimization report.

6. The method according to claim 5, characterized in that The application of the graph neural differential equation algorithm, taking the abnormal event causal relationship diagram as a constraint condition, establishes a dynamic equation for the construction resource allocation and scheduling problem in the water conservancy construction progress and risk warning report, performs a dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and generates a resource optimization dynamic model, including: Extract key nodes and edge information from the abnormal event causal relationship graph, apply the shortest path algorithm in graph theory, calculate the maximum carrying capacity of each node in the key nodes and edge information, and generate the maximum carrying capacity of the path; Using the maximum carrying capacity of the path as an input variable, constructing a graph structure network of the water conservancy construction progress, applying a graph neural differential equation algorithm to the graph structure network, taking the abnormal event causal relationship graph as a constraint condition, establishing a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk warning report, performing a dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, and generating a dynamic model of the construction progress; Reinforcement learning technology is used to train the intelligent agent, and construction safety and quality assurance are introduced as additional constraints. The intelligent agent learns the optimal resource allocation strategy in the construction progress dynamic model to generate the optimal resource allocation strategy for the construction progress; A mathematical programming method is used for the construction progress dynamic model to solve the optimal parameters that meet the best resource allocation strategy for the construction progress and the maximum carrying capacity of the path, and a resource optimization dynamic model is generated based on the optimal parameters.

7. The method according to claim 1, characterized in that According to the latest optimization results in the water conservancy construction resource allocation optimization report, combined with the historical construction data in the water conservancy construction progress management framework, the Lean Six Sigma methodology is introduced, a lean improvement plan is set for the latest optimization results, project feedback of the lean improvement plan is collected to dynamically adjust the lean improvement plan, and a water conservancy construction progress intelligent management plan is generated, including: Using the differential analysis method, combined with the historical construction data in the water conservancy construction progress management framework, differential analysis is performed on the latest optimization results in the water conservancy construction resource allocation optimization report, the construction period shortening contribution of the resource optimization strategy in the water conservancy construction resource allocation optimization report is quantified, and a resource optimization strategy contribution table is generated; Introducing the Lean Six Sigma methodology, taking the strategy contribution value in the resource optimization strategy contribution table as a reference value, defining improving construction efficiency and reducing delay risks as improvement goals, and generating comprehensive improvement goals; Design a lean improvement plan based on the comprehensive improvement goal, verify the effectiveness of the lean improvement plan by running a pilot project, collect feedback during the running of the pilot project to dynamically adjust the lean improvement plan, and generate pilot evaluation results; Discrete event simulation technology is used to set up multiple simulation scenarios, and the execution effect of the water conservancy construction progress management framework under the multiple simulation scenarios is evaluated. A feedback mechanism is introduced to regularly collect simulation prediction values ​​from the execution effects, and comparative analysis is performed based on the comprehensive improvement goals. The latest optimization results are regularly adjusted to generate an intelligent management plan for water conservancy construction progress.

8. An intelligent management system for water conservancy construction progress, characterized in that: include: A collection module is used to collect water conservancy project planning, obtain dynamic changes in the life cycle of the water conservancy project through integrated building information modeling technology, provide a comprehensive perspective to capture the dynamic changes, and generate a water conservancy construction progress management framework; An integration module is used to extract multi-source heterogeneous data in the water conservancy construction progress management framework, use a multimodal fusion network algorithm, and perform cross-modal information integration on the multi-source heterogeneous data through the cross-modal attention mechanism in the multimodal fusion network algorithm to focus on the data correlation between the multi-source heterogeneous data, use causal forest technology, and add external environmental factors through the random forest extension in the causal forest technology to evaluate the specific impact of the external environmental factors on the water conservancy construction progress in the water conservancy construction progress management framework, so as to warn of the risk of delay in the water conservancy construction progress and generate a water conservancy construction progress and risk warning report; A simulation module is used to use a graph neural differential equation algorithm to establish a dynamic equation for the construction resource allocation and scheduling problems in the water conservancy construction progress and risk warning report, perform dynamic system simulation on the dynamic equation to find the optimal solution of the dynamic equation, use enhanced digital twin simulation technology to build a virtual water conservancy construction environment, map the construction resource configuration into the virtual water conservancy construction environment to capture the construction efficiency and resource utilization of the construction resource configuration, and generate a water conservancy construction resource configuration optimization report; The adjustment module is used to introduce the Lean Six Sigma methodology based on the latest optimization results in the water conservancy construction resource allocation optimization report and the historical construction data in the water conservancy construction progress management framework, set a lean improvement plan for the latest optimization results, collect project feedback of the lean improvement plan to dynamically adjust the lean improvement plan, and generate an intelligent management plan for the water conservancy construction progress.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a water conservancy construction progress intelligent management method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, an intelligent management method for water conservancy construction progress as described in any one of claims 1 to 7 is implemented.

Images