Intelligent water conservancy construction management platform

Abstract

The invention belongs to the technical field of digital management, and provides an intelligent water conservancy construction management platform which comprises a database module, an AI data analysis module and a construction management module. The database module is used for acquiring multi-source hydraulic engineering information; each processing sub-block in the AI data analysis module is used for performing standardization processing on various types of data; the association analysis block is used for performing association analysis on various data to obtain association features; and the construction management module is used for constructing a water conservancy construction site simulation model according to various data and associated features after standardization processing, and obtaining a construction simulation result according to the current construction data. Data association features are mined and model parameters are adaptively adjusted by using a space-time convolutional neural network through association analysis blocks, and a construction management module constructs an accurate model and updates simulation in real time, so that data availability, model accuracy and construction management scientificity and efficiency are improved, resource allocation can be effectively optimized, and construction quality and safety are guaranteed.

Description

Technical Field

[0001] The invention belongs to the technical field of digital management, and in particular relates to an intelligent water conservancy construction management platform. Background Art

[0002] The construction of water conservancy projects covers multiple disciplines and multiple links, and construction management requires comprehensive consideration of many factors. With the widespread application of digital and intelligent technologies in the field of water conservancy, technologies such as digital twins, sensors, and data analysis have brought new ways to construction site management and simulation. Water conservancy project data comes from a wide range of sources, including construction sites, water bodies, geology, ecological environment and other types of data, with characteristics such as multi-source heterogeneity, large volume, and dynamic changes. At the same time, in construction management, it is necessary to grasp the construction progress, resource utilization, equipment status and other information in real time and make timely decisions, as well as to consider the impact of the project on the environment to achieve sustainable development. This makes it inevitable to establish an accurate and reliable construction site simulation model.

[0003] At present, water conservancy projects use various sensors to collect data in real time and manually input data, or use data processing and analysis, model construction, model-based simulation and prediction to assist decision-making. However, existing technologies still have many shortcomings. For example, at the data level, existing technologies have quality problems, integration difficulties and safety risks; in terms of models, their accuracy, computational efficiency and adaptability need to be improved; in terms of technical integration, technical synergy is poor and there is a lack of unified standards. These have led to difficulties in sharing and reusing data and models between different projects, increased system costs, and affected overall operational efficiency and performance. Summary of the invention

[0004] In view of this, the present invention provides an intelligent water conservancy construction management platform, aiming to solve the problems that the prior art still has many shortcomings, increases system costs, and affects overall operating efficiency and performance.

[0005] A first aspect of an embodiment of the present invention provides an intelligent water conservancy construction management platform, including: Database module, AI data analysis module and construction management module; The database module is used to obtain multi-source water conservancy project information; wherein, the multi-source water conservancy project information includes the following types of data: water conservancy construction site data, water body data, geological data and ecological environment data; The AI ​​data analysis module includes multiple processing sub-blocks and correlation analysis blocks; Each processing sub-block is used to standardize various types of data; The association analysis block is used to perform association analysis on various types of data to obtain association features; wherein the association features include: direct time association features, direct space association features, indirect time association features and indirect space association features; The construction management module is used to build a water conservancy construction site simulation model based on various types of data and related features after standardized processing, and to obtain current construction data in real time, and to obtain construction simulation results based on current construction data and the water conservancy construction site simulation model.

[0006] In some possible implementations, the association analysis block is used to: input various types of data into a pre-established spatiotemporal convolutional neural network model to obtain association features.

[0007] In some possible implementations, the spatiotemporal convolutional neural network model includes an input layer, a spatiotemporal convolution layer, a feature fusion layer, a feature separation layer, a correlation extraction layer, and an output layer. The correlation analysis block is used to: Input various types of data from the input layer into the spatiotemporal convolution layer to obtain multiple first feature maps; Inputting the multiple first feature maps into the feature fusion layer to obtain fused features; Input the fused features into the feature separation layer to obtain direct temporal correlation features and direct spatial correlation features; The fusion features, direct temporal correlation features and direct spatial correlation features are input into the correlation extraction layer to obtain indirect temporal correlation features and indirect spatial correlation features; The direct time association features, direct space association features, indirect time association features and indirect space association features are output through the output layer.

[0008] In some possible implementations, each processing sub-block is used to: Each type of data is cleaned, standardized, converted and stored to obtain various types of data after annotation.

[0009] In some possible implementations, the spatiotemporal convolutional neural network model further includes a parameter adjustment layer; and the association analysis block is used to: Obtaining characteristic information of the standardized processing in each processing sub-block; Inputting the feature information into the parameter adjustment layer to obtain parameter adjustment information; According to the parameter adjustment information, the parameters of the spatiotemporal convolutional neural network model are adjusted.

[0010] In some possible implementations, the construction management module is used to: Establish a physical model of the water conservancy construction site; Map the water conservancy construction site based on various data and physical models to establish a digital twin model; According to the correlation characteristics, the digital twin model is dynamically simulated to obtain a simulation model of the water conservancy construction site.

[0011] In some possible implementations, the construction management module is used to: Determine the dynamic information of the construction site based on the correlation features and behavior prediction model; According to the dynamic information, the behavior mapping of the water conservancy construction site simulation model is adjusted.

[0012] In some possible implementations, the construction management module is used to: Update the model parameters of the water conservancy construction site simulation model according to the current construction data; Start the updated water conservancy construction site simulation model to perform simulation calculations and obtain the construction simulation results.

[0013] In some possible implementations, the platform further includes an evaluation module, which is used to: The construction simulation results are evaluated to obtain evaluation results.

[0014] In some possible implementations, the evaluation module is further used to: Determine the water conservancy project construction management plan based on the evaluation results and current construction data.

[0015] The intelligent water conservancy construction management platform provided by the embodiment of the present invention comprises a database module, an AI data analysis module and a construction management module; the database module is used to obtain multi-source water conservancy project information; the AI ​​data analysis module comprises a plurality of processing sub-blocks and an association analysis block; each processing sub-block is used to perform standardized processing on various types of data; the association analysis block is used to perform association analysis on various types of data to obtain association features; the construction management module is used to construct a water conservancy construction site simulation model according to various types of data and association features after standardized processing, and obtain current construction data in real time, and obtain construction simulation results according to current construction data and the water conservancy construction site simulation model. The database module integrates multi-source data to break data silos, the processing sub-block performs standardized processing to improve data quality, the association analysis block uses a spatiotemporal convolutional neural network to mine data association features and adaptively adjust model parameters, the construction management module constructs an accurate model and updates the simulation in real time, and determines a scientific construction management plan in combination with the results of the evaluation module, realizing full process optimization from data integration, analysis to model construction, simulation and decision support, improving data availability, model accuracy, scientificity and efficiency of construction management, effectively optimizing resource allocation, ensuring construction quality and safety, and promoting the sustainable development of water conservancy projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0017] Figure 1 It is a structural diagram of an intelligent water conservancy construction management platform provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0019] Figure 1 Schematic diagram of the structure of the intelligent water conservancy construction management platform provided by the embodiment of the present invention. Figure 1 As shown, in some embodiments, the intelligent water conservancy construction management platform includes: Database module 11, AI data analysis module 12 and construction management module 13; The database module 11 is used to obtain multi-source water conservancy project information; wherein, the multi-source water conservancy project information includes the following types of data: water conservancy construction site data, water body data, geological data and ecological environment data; the AI ​​data analysis module 12 includes multiple processing sub-blocks and correlation analysis blocks; each processing sub-block is used to perform standardized processing on various types of data; the correlation analysis block is used to perform correlation analysis on various types of data to obtain correlation features; wherein, the correlation features include: direct time correlation features, direct space correlation features, indirect time correlation features and indirect space correlation features; the construction management module 13 is used to construct a water conservancy construction site simulation model based on various types of data and correlation features after standardized processing, and obtain current construction data in real time, and obtain construction simulation results based on current construction data and the water conservancy construction site simulation model.

[0020] In the embodiment of the present invention, the database module plays a vital role in the intelligent water conservancy construction management platform. It is responsible for acquiring multi-source water conservancy project information, which covers water conservancy construction site data, water body data, geological data, ecological environment data and other aspects.

[0021] For construction equipment operation data: The database module obtains the equipment's operating parameters in real time by connecting to various sensors installed on construction equipment (such as excavators, cranes, concrete mixers, etc.). These parameters include engine speed, temperature, oil pressure, equipment working hours, load, etc. For example, for an excavator, the sensor can monitor the working angle of its digging arm, digging depth, and engine fuel consumption in real time, and transmit these data to the database module for storage and management. By analyzing these data, we can understand the equipment's operating status, discover equipment failure hazards in a timely manner, and reasonably arrange equipment maintenance and maintenance plans to improve the equipment's efficiency and lifespan.

[0022] Regarding construction personnel information: With the help of the positioning system and attendance equipment, the database module can obtain information such as the location and attendance of construction personnel. The location information of construction personnel can help managers to grasp the distribution of personnel in real time, reasonably allocate human resources, and improve construction efficiency. For example, when a construction area needs to increase manpower, managers can quickly find nearby and idle construction personnel for deployment by checking the location information of the personnel. The attendance recorded by the attendance equipment can be used to count the working hours of personnel and evaluate work performance, which also helps to protect the labor rights and interests of construction personnel.

[0023] For construction progress data: Using cameras, laser scanners and other equipment, the database module can monitor and record the actual progress of the construction site. For example, by regularly performing three-dimensional laser scanning on the construction site, the shape, size and other information of the building can be obtained, and compared with the design model to accurately calculate the construction progress. For large-scale water conservancy projects such as dam construction, displacement sensors installed in key locations can also be used to monitor the deformation of the dam body in real time and promptly identify safety hazards during the construction process. In addition, construction personnel can also manually enter relevant information on the construction progress, such as the amount of work completed, the progress of the construction area, etc., through mobile terminals and other equipment to ensure that the database module can fully and accurately grasp the construction progress.

[0024] For water level and flow data: the database module is connected to equipment such as water level meters and flow meters installed in the water body to obtain real-time data on water level height and flow size. These data are crucial for the scheduling and operation of water conservancy projects. For example, in reservoir management, accurate knowledge of water level and flow data can reasonably control the reservoir's water storage and flood discharge, ensure the safe operation of the reservoir, and also help to reasonably allocate water resources to meet different needs such as irrigation and power generation. In addition, changes in water level and flow data can also reflect the dynamic changes of water bodies, providing an important basis for flood control, drought relief and other work.

[0025] For water quality parameter data: Through the water quality monitor, the database module can measure water quality parameters such as pH, dissolved oxygen, and chemical oxygen demand. Changes in water quality data can reflect the pollution status of the water body and the health of the ecological environment. For example, when the chemical oxygen demand in the water body exceeds the standard, it may mean that the water body is polluted by organic matter and timely measures need to be taken to treat it. The long-term monitoring and recording of these water quality parameters by the database module helps to analyze the trend of water quality changes, evaluate the impact of water conservancy projects on the water environment, and provide data support for water resource protection and ecological restoration.

[0026] For geological structure data: The database module collects underground geological structure information obtained through geological exploration (such as drilling, geophysical exploration, etc.), including the distribution of rock layers, soil types, faults, etc. These data are of great significance to the foundation design and construction safety of water conservancy projects. For example, when building a dam, understanding the geological structure at the dam site can determine the type and depth of the foundation and ensure the stability of the dam. Through the analysis of geological structure data, it is also possible to predict geological problems that may be encountered during construction, such as landslides and collapses, and take corresponding preventive measures in advance.

[0027] For geotechnical mechanics parameter data: laboratory tests and on-site in-situ tests are used to obtain geotechnical mechanical property data, such as compressive strength, shear strength, deformation modulus, etc. These parameters are the basis for the construction of geological models and play a key role in evaluating the bearing capacity of the foundation, calculating the deformation and stability of the soil, etc. The database module stores these geotechnical mechanics parameter data, providing accurate mechanical basis for the structural design and construction of water conservancy projects, and ensuring the safety and reliability of the project.

[0028] For meteorological data: the database module exchanges data with the meteorological station to collect meteorological information of the construction site and surrounding areas, such as temperature, humidity, wind speed, precipitation, etc. Meteorological conditions have an important impact on the construction progress and ecological environment. For example, extreme weather such as high temperature and heavy rain may cause construction to be suspended, affecting the construction progress; at the same time, meteorological factors such as precipitation and wind speed will also affect soil erosion and the balance of the ecosystem. By acquiring meteorological data in real time, the database module can provide weather warning information for construction management, help managers to reasonably arrange construction plans, take corresponding protective measures, and reduce the adverse effects of meteorological factors on construction and the ecological environment.

[0029] For biodiversity data: Through field surveys and monitoring, the database module records biodiversity data such as the species and number of plants and animals at the construction site and surrounding areas. Biodiversity data can reflect the stability and health of the ecosystem. For example, in the process of water conservancy project construction, protecting biodiversity is crucial to maintaining ecological balance and sustainable development. The database module's recording and analysis of biodiversity data helps to assess the impact of project construction on the ecological environment, take corresponding ecological protection measures, protect biological habitats, and reduce damage to biodiversity.

[0030] In some embodiments, the association analysis block is used to: input various types of data into a pre-established spatiotemporal convolutional neural network model to obtain association features. In some embodiments, the spatiotemporal convolutional neural network model includes an input layer, a spatiotemporal convolutional layer, a feature fusion layer, a feature separation layer, an association extraction layer, and an output layer, and the association analysis block is used to: input various types of data from the input layer into the spatiotemporal convolutional layer to obtain multiple first feature maps; input multiple first feature maps into the feature fusion layer to obtain fusion features; input the fusion features into the feature separation layer to obtain direct time association features and direct space association features; input the fusion features, direct time association features, and direct space association features into the association extraction layer to obtain indirect time association features and indirect space association features; output the direct time association features, direct space association features, indirect time association features, and indirect space association features through the output layer.

[0031] In the embodiment of the present invention, the spatiotemporal convolutional neural network model is designed for processing data with spatiotemporal characteristics, and is composed of an input layer, a spatiotemporal convolution layer, a feature fusion layer, a feature separation layer, a correlation extraction layer, and an output layer. This structure can dig deep into the spatiotemporal correlation information in the data layer by layer, from basic feature extraction to the acquisition of complex correlation features, and provide comprehensive and accurate correlation features for water conservancy construction management.

[0032] In the embodiment of the present invention, the spatiotemporal convolution layer is the core component of the spatiotemporal convolutional neural network, which is specifically used to process data with time and space dimension information. In the water conservancy construction management scenario, water conservancy construction site data, water body data, geological data, and ecological environment data all contain rich spatiotemporal characteristics. The spatiotemporal convolution layer can effectively capture the local characteristics of these data in the spatiotemporal dimension, and mine the spatiotemporal correlation patterns in the data, providing an important basis for subsequent feature fusion, analysis, and final decision-making.

[0033] The convolution operation is the basis of the spatiotemporal convolution layer, which slides the convolution kernel (also called filter) on the input data and performs a weighted sum calculation on the data. In traditional two-dimensional convolution, the convolution kernel slides on the spatial dimension (length and width) of the image; in spatiotemporal convolution, the convolution kernel not only slides on the spatial dimension, but also moves on the temporal dimension, thereby considering the temporal and spatial information of the data at the same time. The spatiotemporal convolution kernel is a three-dimensional filter, which has a certain size in the three dimensions of time, length and width of space. At the water conservancy construction site, the construction progress will change in different spatial areas over time. The spatiotemporal convolution layer can capture this spatiotemporal change feature. For example, by performing convolution operations on the construction completion data at different time points and different construction areas, the convolution kernel can learn the advancement mode of the construction progress, such as whether there is periodic acceleration or deceleration of construction in certain areas, and the mutual influence relationship between the construction progress of different areas. If the basic construction progress of a certain area is accelerated, it may promote the subsequent construction of adjacent areas. This related information can be discovered through the feature extraction of the spatiotemporal convolution layer.

[0034] Water body data (such as water level, flow, water quality, etc.) also has obvious spatiotemporal characteristics. The spatiotemporal convolution layer can process water body data at different time and spatial locations to explore the laws of water body changes. For example, for water level data in rivers, the spatiotemporal convolution layer can identify the change patterns of water levels in different seasons and different river sections, as well as the propagation characteristics of extreme events such as floods in time and space. By analyzing these characteristics, the trend of water level changes can be predicted in advance, providing a decision-making basis for the scheduling of water conservancy projects and flood prevention and disaster reduction. There are also complex associations between geological data (such as soil type, geological structure, etc.) and ecological environment data (such as vegetation coverage, biodiversity, etc.) in time and space. The spatiotemporal convolution layer can extract the spatiotemporal correlation characteristics between these data. For example, changes in geological conditions may affect the growth and distribution of vegetation in the ecological environment, and changes in vegetation may in turn affect soil erosion and water quality. The spatiotemporal convolution layer can capture this complex interaction relationship and provide support for the ecological environment protection and sustainable development of water conservancy projects.

[0035] Unlike traditional convolutional neural networks that can only process spatial information or simply flatten the time dimension, the spatiotemporal convolution layer can simultaneously consider the temporal and spatial dimensions of the data and capture the characteristics of the data more comprehensively. This makes it more accurate and effective when processing data with spatiotemporal characteristics. The spatiotemporal convolution layer can automatically learn spatiotemporal patterns and associated features from the data through the weight learning mechanism of the convolution kernel, without the need for manual extraction of complex features. This automatic learning capability enables the model to adapt to different water conservancy construction management scenarios and data changes, improving the generalization ability of the model.

[0036] In an embodiment of the present invention, the spatiotemporal convolution layer extracts multiple first feature maps from the input multi-source water conservancy project data (water conservancy construction site data, water body data, geological data, and ecological environment data, etc.) through a convolution operation. Each of these first feature maps contains information about the data at different spatiotemporal dimensions, different local areas, or different feature levels. However, the information expressed by a single first feature map is relatively limited and cannot fully reflect the overall picture and internal connection of the data. The task of the feature fusion layer is to fuse these scattered first feature maps with different focuses and organically combine the information they contain. In essence, the feature fusion layer is a process of comprehensive processing of multiple first feature maps, and through certain calculations or operations, they are converted into a new, more comprehensive feature representation, namely, a fusion feature. This fusion feature can more comprehensively capture the spatiotemporal characteristics of the data and the potential relationship between different data types, providing a richer information basis for subsequent analysis and decision-making.

[0037] Specifically, fusion based on the attention mechanism can be used. The attention mechanism can automatically learn the importance of each feature map at different positions or channels based on the content of the input feature map, and then fuse them based on these importances. This method can more flexibly capture the relationship between different feature maps and improve the quality of fused features.

[0038] In an embodiment of the present invention, the feature fusion layer can effectively integrate feature information from different data sources. For example, the equipment operation data (such as the working status and operating parameters of the equipment) at the water conservancy construction site are fused with the feature graphs of water body data (such as water level, flow, water quality, etc.). Through fusion, the potential impact of equipment operation on the water environment can be discovered, such as whether the operation of certain construction equipment leads to water pollution or water flow changes; or whether changes in the water environment will affect the normal operation of the equipment. Similarly, the fusion of feature graphs of geological data (such as geological structure, rock and soil mechanics parameters, etc.) and ecological environment data (such as vegetation coverage, biodiversity, etc.) helps to analyze the impact of geological conditions on the ecological environment, as well as the feedback effect of ecological environment changes on geological stability. For example, by fusion features, it can be determined whether the destruction of vegetation will cause soil erosion under specific geological conditions, thereby affecting the basic stability of water conservancy projects.

[0039] In an embodiment of the present invention, the feature separation layer receives fused features from the feature fusion layer. Although the fused features integrate the information of multiple first feature maps, the time and space correlation features therein are intertwined. The core principle of the feature separation layer is to use specific algorithms and models to conduct in-depth analysis and processing of the fused features, identify and extract the direct correlation of the data in the time dimension and the direct correlation in the space dimension. The time correlation reflects the law and relationship of data changes over time, such as the progress of construction progress over time, the trend of water level changes at different time points, etc. The spatial correlation reflects the mutual connection of data in spatial distribution, such as the differences in construction conditions in different areas of the construction site, the distribution of water quality in different geographical locations, etc. The role of the feature separation layer is to separate the two directly related features of different dimensions from the fused features, so as to conduct more detailed analysis and application of them later.

[0040] In an embodiment of the present invention, the feature separation layer can extract direct time-related features from the fused features. It is usually composed of a series of operation units designed for the time dimension, and common structures include one-dimensional convolutional layers, recurrent neural networks and their variants (such as long short-term memory networks, gated recurrent units), etc. Among them, in the time feature separation, the convolution kernel of the one-dimensional convolutional layer slides on the time dimension to perform convolution operations on the fused features. Through different convolution kernel settings, different patterns and features of the fused features in the time series can be captured. For example, a smaller convolution kernel can capture short-term time change features, while a larger convolution kernel is more suitable for extracting long-term time trend features. Network structures such as recurrent neural networks can effectively process data with time series characteristics. They control the flow and memory of information through a gating mechanism, and can better capture long-term dependencies in the data. In the feature separation layer, these network structures can conduct in-depth analysis of the time dimension of the fused features and extract direct time-related features, such as the law of changes in the construction progress of water conservancy projects over time, and the correlation of equipment operating status at different time points.

[0041] If a one-dimensional convolution layer is used, the convolution kernel slides in the time dimension according to the set step size to perform convolution calculations on the fused features. Each convolution operation obtains a scalar value, which constitutes a time-related feature representation. Through the operation of multiple convolution kernels, multiple time-related feature maps can be obtained. These feature maps contain the responses of the fused features in different time modes, and finally the direct time-related features are extracted after further processing.

[0042] If a network structure such as a recurrent neural network is used, the fused features are input into the network in chronological order. The gate control units (input gate, forget gate, output gate) in the network will decide which information to keep, which information to forget, and which information to output based on the input information and the previous memory state. In this way, the dependency of the fused features on the time series can be learned, thereby separating the directly time-related features.

[0043] In an embodiment of the present invention, the feature separation layer can also separate direct spatial correlation features from the fused features. It is generally composed of a two-dimensional convolution layer, a pooling layer, etc. Among them, the convolution kernel of the two-dimensional convolution layer performs a sliding operation on the two dimensions of height and width of the space, and performs a convolution operation on the fused features. Through the convolution operation, the two-dimensional convolution layer can extract the local features and patterns of the fused features in space, such as the spatial relationship between different areas of the water conservancy construction site, the distribution characteristics of water quality in different geographical locations, etc. Convolution kernels of different sizes and numbers can capture spatial features of different scales. The pooling layer is usually used in conjunction with the two-dimensional convolution layer, and the most common ones are maximum pooling and average pooling. The role of the pooling layer is to downsample the feature map output by the convolution layer, reducing the amount of data while retaining important spatial features. Through the pooling operation, the dimension of the feature map can be reduced, the computational efficiency of the model can be improved, and overfitting can be prevented to a certain extent.

[0044] The convolution kernel slides in the height and width directions of the space, and performs convolution calculations on each position of the fused feature to obtain a series of feature maps. These feature maps reflect the local features of the fused features at different spatial positions. Next, the pooling layer downsamples the feature maps output by the convolution layer. Maximum pooling selects the maximum value in each pooling area as the output, while average pooling calculates the average value of each pooling area as the output. Through the pooling operation, the dimension of the feature map is reduced while retaining important spatial features, and finally the direct spatial correlation features are obtained.

[0045] In an embodiment of the present invention, in the spatiotemporal convolutional neural network, the association extraction layer is a key component, which inherits the processing results of the feature fusion layer and the feature separation layer, and further mines the indirect time correlation features and indirect space correlation features in the data, providing more comprehensive and in-depth information support for water conservancy construction management.

[0046] The association extraction layer receives the fusion features from the feature fusion layer, as well as the direct time association features and direct space association features output by the feature separation layer. The fusion features integrate the information of multiple first feature maps output by the spatiotemporal convolution layer, and contain rich but complex spatiotemporal feature information; the direct time association features reflect the direct connection of data in the time dimension, such as the trend of construction progress over time, the association of equipment operating parameters at different time points, etc.; the direct space association features reflect the direct relationship of data in the spatial dimension, such as the spatial position relationship between different construction areas, the spatial distribution characteristics of water quality, etc. These input data provide a rich information basis for the association extraction layer to mine indirect association features.

[0047] The association extraction layer uses deep learning models, such as long short-term memory networks (LSTM), gated recurrent units (GRU), graph neural networks (GNN), etc., to mine indirect association features in the data. The association extraction layer interacts and combines the input fusion features, direct time association features, and direct space association features to mine the indirect associations between them. It is not just an analysis of a single feature, but also discovers potential and complex associations through the interaction and influence between features. For example, the water body data features in the fusion features are combined with the construction area location information in the direct space association features to analyze how the flow and changes of the water body indirectly affect the construction progress and quality of different construction areas; or the equipment running time information in the direct time association features is interacted with the construction environment data features in the fusion features to mine the indirect impact of the equipment on its performance after long-term operation under different environmental conditions.

[0048] The deep learning model of the association extraction layer needs to be trained with a large amount of data to learn the indirect association feature patterns in the data. During the training process, the model parameters are adjusted to optimize the model performance based on the difference between the model's prediction results and the actual data. Through continuous training and parameter adjustment, the association extraction layer can more accurately extract the indirect temporal association features and indirect spatial association features in the data, and improve the model's ability to understand and express complex associations in water conservancy construction management.

[0049] In some embodiments, each processing sub-block is used to: perform data cleaning, standardization conversion and storage on each type of data to obtain each type of data after labeling.

[0050] In an embodiment of the present invention, in the intelligent water conservancy construction management platform, the processing sub-block undertakes the important task of pre-processing each type of data in the multi-source water conservancy project information, specifically including three key steps of data cleaning, standardized conversion and storage, so as to obtain various types of data after standardized processing, and provide a high-quality data foundation for subsequent analysis and model building.

[0051] The sources of water conservancy project data are extensive and complex. In the process of data collection, noise, missing values, and outliers are inevitably introduced. These problems will affect the quality of the data and the accuracy of subsequent analysis. Therefore, data cleaning is required. Data cleaning specifically includes noise removal, missing value processing, outlier identification and processing, etc.

[0052] Noise removal: Due to the accuracy limitation of sensors, environmental interference and other factors, the collected data may contain noise. The processing sub-block will use filtering algorithms to remove noise, such as mean filtering, median filtering, Gaussian filtering, etc. For example, for the vibration noise in the equipment operation data at the water conservancy construction site, the mean filter can smooth the data and reduce the interference of noise on the judgment of the equipment operation status.

[0053] Missing value processing: Missing data is a common problem, which may be caused by sensor failure, data transmission interruption, etc. The processing subblock will select an appropriate method to handle missing values ​​based on the characteristics of the data. For a small number of missing values, interpolation methods such as linear interpolation and spline interpolation can be used to estimate the missing values ​​based on the trend of existing data; for a large number of missing values, it may be necessary to re-collect data or use other statistical methods for processing. For example, in water body data, if the water level data for a certain period of time is missing, the missing values ​​can be filled by linear interpolation of the water level data at the previous and next time points.

[0054] Identification and processing of outliers: Outliers refer to observations that are significantly deviated from other data, which may be caused by measurement errors, equipment failures, or other special reasons. The processing subblock will identify outliers by setting a reasonable threshold range or using statistical methods (such as box plots, 3σ principle, etc.). For outliers, they are generally corrected or eliminated according to the specific situation. For example, in geological data, if the rock hardness value of a sample is significantly beyond the normal range, and it is found to be caused by measurement errors after inspection, the outlier can be eliminated.

[0055] In the embodiment of the present invention, since the water conservancy project data includes multiple types, such as numerical type, categorical type, etc., and the dimensions and value ranges of different types of data vary greatly, in order to make the data comparable and facilitate subsequent processing, standardized conversion is required.

[0056] In an embodiment of the present invention, the data after data cleaning and standardization conversion needs to be stored for subsequent analysis and use. The processing sub-block will store various types of data after standardization processing in the database. The database can be a relational database (such as MySQL, Oracle) or a non-relational database (such as MongoDB, Redis), depending on the characteristics of the data and application requirements. The stored data should be well organized and indexed for fast retrieval and query. For example, water conservancy construction site data, water body data, geological data, and ecological environment data are stored in different tables, and corresponding indexes are established to facilitate data reading during association analysis and model construction.

[0057] In some embodiments, the spatiotemporal convolutional neural network model also includes a parameter adjustment layer; the association analysis block is used to: obtain feature information of standardized processing in each processing sub-block; input the feature information into the parameter adjustment layer to obtain parameter adjustment information; and adjust the parameters of the spatiotemporal convolutional neural network model according to the parameter adjustment information.

[0058] In the embodiment of the present invention, the operation of the parameter adjustment layer is specifically as follows: 1. Feature information acquisition. The first operation of the parameter adjustment layer is to obtain the standardized feature information from each processing sub-block. This feature information covers the key features of multi-source data in the field of water conservancy projects after processing, as follows: Water conservancy construction site data features: Contains the standardized numerical features of the operating parameters of the construction equipment (such as speed, temperature, load, etc.), as well as the processed features of the attendance and location information of the construction personnel. For example, the statistical features such as the mean and variance of the equipment operating parameters can reflect the stability of the equipment operation and the changing trend of the working status. This information is crucial for the parameter adjustment layer to determine whether the model has accurately learned the equipment operation status.

[0059] Water body data characteristics: including the standardized characteristic values ​​of water level, flow, and water quality indicators (such as pH, dissolved oxygen, etc.). For example, the amplitude and frequency characteristics of water level changes, as well as the correlation characteristics between different water quality indicators, can help the parameter adjustment layer understand the performance of the model when processing water body data and determine whether parameters need to be adjusted to better capture the dynamic changes of the water body.

[0060] Geological data characteristics: It involves the standardized characteristic information of geological structure (such as rock layer distribution, soil type, etc.) and geotechnical parameters (such as compressive strength, shear strength, etc.). For example, the spatial distribution characteristics of geological structure and the degree of variation of geotechnical parameters are of great significance for the parameter adjustment layer to evaluate the model's simulation and analysis capabilities of geological conditions, thereby determining whether to adjust the model parameters accordingly.

[0061] Ecological and environmental data features: including the processed features of meteorological data (such as temperature, humidity, wind speed, etc.) and biodiversity data (such as species and quantity of plants and animals, etc.). For example, the seasonal variation characteristics of temperature and humidity, and the changing trends of biodiversity indicators, etc. The parameter adjustment layer uses these feature information to judge the accuracy and adaptability of the model in processing ecological and environmental data, and then adjusts the model parameters.

[0062] 2. Feature information input and analysis After the acquired characteristic information of various standardized processes is input into the parameter adjustment layer, the layer will conduct in-depth analysis of this information, mainly through the following methods: Statistical analysis: Calculate various statistics of input feature information, such as mean, median, standard deviation, coefficient of variation, etc. Through these statistics, the parameter adjustment layer can understand the distribution of data and determine whether the data is abnormal or deviates from expectations. For example, if the standard deviation of a certain type of data is found to be too large, it may mean that the model has a large uncertainty when processing the data, and the parameters need to be adjusted to improve the stability of the model.

[0063] 3. Generate parameter adjustment information Based on the analysis results of the input feature information, the parameter adjustment layer will generate corresponding parameter adjustment information, including the following aspects: Parameter adjustment direction: Determine the direction in which the model parameters need to be adjusted, whether to increase or decrease the value of a parameter. For example, if the analysis finds that the model generally underestimates the predicted values ​​when processing water flow data, the parameter adjustment layer may decide to increase the value of the flow-related parameters to improve the model's prediction accuracy.

[0064] Parameter adjustment range: Calculate the specific range of adjustment required for each parameter. This is usually determined based on the analysis of feature information and the performance evaluation of the model. For example, by comparing the error between the model's prediction results and the actual data, combined with the changes in feature information, the parameter adjustment layer can determine the appropriate parameter adjustment range to improve the performance of the model without destroying the stability of the model.

[0065] Parameter adjustment priority: Determine the adjustment priority of different parameters based on the degree of influence of feature information on model performance. Key parameters that have a greater impact on model performance will be adjusted first; while for some minor parameters, they can be fine-tuned based on the performance of the model after the key parameters are adjusted. For example, when processing water conservancy construction site data, parameters directly related to the construction progress may have a higher adjustment priority, while some parameters related to the operation of auxiliary equipment may have a relatively low adjustment priority.

[0066] 4. Adjust model parameters based on parameter adjustment information The parameter adjustment layer passes the generated parameter adjustment information to other layers of the spatiotemporal convolutional neural network model to adjust the model parameters. The specific operations are as follows: Parameter adjustment of spatiotemporal convolution layer: The parameters of the spatiotemporal convolution layer, such as the weight and bias of the convolution kernel, are updated according to the parameter adjustment information. For example, if the parameter adjustment information indicates that the model needs to enhance its ability to extract a certain type of feature, the weight value of the convolution kernel may be increased to make it pay more attention to this type of feature. At the same time, adjusting the bias parameter can change the output range of the convolution layer to adapt to changes in the data.

[0067] In some embodiments, the construction management module is used to: establish a physical model of the water conservancy construction site; map the water conservancy construction site according to various data and physical models to establish a digital twin model; dynamically simulate the digital twin model according to associated characteristics to obtain a simulation model of the water conservancy construction site.

[0068] In an embodiment of the present invention, various types of real-time data collected (such as current construction progress data, equipment operation status data, environmental monitoring data, etc.) are integrated and fused with data related to the physical model. The quality and consistency of the data are ensured through pre-processing operations such as data cleaning and standardization. For example, the real-time operation data of the construction equipment is fused with the design parameters of the equipment in the physical model to more accurately evaluate the operation status and performance of the equipment. A mapping relationship can then be established, where the mapping relationship includes geometric mapping, attribute mapping, and behavior mapping.

[0069] Geometric mapping: Using geographic information system (GIS) technology and 3D modeling software, the geometric shape and spatial position of the physical model are accurately mapped to the actual water conservancy construction site. This ensures that the buildings, topography, and other features in the digital twin model are geometrically consistent with the actual scene, providing accurate spatial reference for subsequent simulation and analysis.

[0070] Attribute mapping: associate and map various attributes in the physical model (such as material attributes, mechanical parameters, construction technology, etc.) with the corresponding attributes of the actual construction site. At the same time, dynamically update the attribute information in the real-time monitoring data (such as the real-time status of the equipment, the working conditions of the construction personnel, etc.) to the digital twin model, so that the model can reflect the actual situation in real time.

[0071] Behavior mapping: Establish the mapping relationship between various physical processes in the physical model (such as water flow, structural stress, construction activities, etc.) and the corresponding behaviors in the actual construction site. Through the simulation of the physical model and the verification of actual data, ensure that the digital twin model can accurately simulate various behaviors and changes in the actual construction process.

[0072] In an embodiment of the present invention, the digital twin model architecture constructed based on the above mapping includes a data layer, a model layer and an application layer.

[0073] Data layer: Establish data storage and management system to store various types of data, including physical model data, real-time monitoring data, historical data, etc. Use database management system (such as MySQL, Oracle, etc.) and data warehouse technology to achieve efficient storage, retrieval and management of data.

[0074] Model layer: Integrate the physical model and real-time data processing module to build the core algorithms and modules of the digital twin model. By updating and adjusting the physical model in real time, it can accurately reflect the changes in the actual construction site. At the same time, data analysis and machine learning algorithms are used to analyze and predict real-time data to provide decision support for construction management.

[0075] Application layer: Develop various application functions, such as construction progress monitoring, equipment fault diagnosis, quality control, safety warning, etc. Through the visual interface, the simulation results and analysis information of the digital twin model are intuitively displayed to construction managers to facilitate their decision-making and management.

[0076] In an embodiment of the present invention, various types of correlation features are obtained from the correlation analysis block, including direct time correlation features, direct space correlation features, indirect time correlation features, and indirect space correlation features. These correlation features are analyzed in depth to understand their role and influence mechanism in the water conservancy construction process. For example, the direct time correlation features between construction progress and equipment operation time, as well as the indirect space correlation features between geological conditions and construction safety are analyzed to provide a basis for dynamic simulation.

[0077] The dynamic simulation process includes initial condition setting, model updating and simulation calculation.

[0078] Initial condition setting: Set initial conditions for the digital twin model based on the current construction status and real-time data. This includes the current status of the construction progress, the operating status of the equipment, environmental parameters, etc. For example, input the current construction progress data into the model to determine the completion status and remaining workload of each construction task; input the real-time operating parameters of the equipment into the model to set the initial status of the equipment.

[0079] Model update: The digital twin model is updated in real time based on the associated features and real-time data. For example, if the geological conditions of a construction area are found to have changed, the geological parameters and construction difficulty assessment of the area in the model are updated based on the indirect spatial association features; if the construction progress is delayed, the time schedule and resource allocation of subsequent construction tasks are adjusted based on the direct time association features.

[0080] Simulation calculation: Use the updated digital twin model to perform simulation calculations to predict the development of the construction process in the future. By solving various equations and logical relationships in the model, simulate the advancement of construction progress, changes in equipment operating status, environmental impact, etc. For example, simulate the construction progress and resource requirements under different construction plans to evaluate the feasibility and efficiency of the construction plan.

[0081] In some embodiments, the construction management module is used to: determine the dynamic information of the construction site based on the associated features and the behavior prediction model; and adjust the behavior mapping of the water conservancy construction site simulation model based on the dynamic information.

[0082] In the embodiments of the present invention, in some embodiments of the intelligent water conservancy construction management platform, the construction management module uses correlation features and behavior prediction models to determine the dynamic information of the construction site, and adjusts the behavior mapping of the water conservancy construction site simulation model based on this information. This process is of great significance to the realization of precise and intelligent construction management.

[0083] The construction management module first obtains various types of association features from the association analysis block, including direct time association features, direct space association features, indirect time association features, and indirect space association features. These association features reveal the inherent connection between various factors in the water conservancy construction process. For example, the direct time association features between the construction progress and the equipment operation time can help the construction management module understand the impact of the continuous operation of the equipment on the progress of the construction. If it is found that the construction progress shows a significant trend of acceleration or slowdown after the equipment has been running at high load for a long time, this association feature can be used to predict the changes in the subsequent construction progress. The construction sequence and progress of adjacent construction areas restrict each other. By analyzing the direct space association features, the impact of changes in the construction progress of a certain area on the surrounding areas can be predicted in advance. For example, the indirect time association features of the early construction decision (such as the selection of construction technology) on the quality of the later project, the construction management module can predict the possible problems in the quality of the future project based on this feature and the current implementation of the construction technology, and take corresponding quality control measures in advance. The geological conditions affect the stability of the surrounding construction areas by affecting the flow of groundwater, and other indirect space association features. The construction management module can use these features to evaluate the potential impact of changes in geological conditions on the entire construction site and adjust the construction plan in time.

[0084] The behavior prediction model is built based on historical data and machine learning algorithms to predict the probability of occurrence and development trend of various behaviors on the construction site. The behavior prediction model extracts patterns and rules by learning a large amount of historical construction data (including construction progress, equipment operation data, personnel activity data, etc.). For example, learn the time and cause of equipment failures in similar water conservancy projects in the past, as well as related construction environment factors. Combined with the associated features, the behavior prediction model can predict various behaviors on the construction site. For example, predict the possibility of construction equipment failure in a certain period of time in the future, or predict the probability of construction progress delays in a certain construction area due to weather changes. At the same time, it can also predict the behavior patterns of construction personnel, such as changes in work efficiency and the probability of safety violations.

[0085] The construction management module combines the correlation features with the prediction results of the behavior prediction model to comprehensively determine the dynamic information of the construction site. This dynamic information includes but is not limited to the changing trend of the construction progress, the changes in the operating status of the equipment (such as fault warnings), the fluctuations in the work efficiency of personnel, the impact of environmental factors on construction (such as changes in construction conditions caused by weather changes), etc. For example, by analyzing the correlation features, it is found that the operating parameters of a certain device have similar correlation patterns with historical fault data, and the behavior prediction model predicts that the probability of failure of the device in the next few days is high. In this case, the construction management module can determine the dynamic information that the device has a risk of failure.

[0086] The behavior mapping of the hydraulic construction site simulation model refers to the simulation and representation of various behaviors of the actual construction site (such as the operation of construction equipment, the operation of construction personnel, the movement of water flow, etc.) in the model. The accuracy of the behavior mapping directly affects the degree to which the simulation model reflects the actual construction situation. If the dynamic information shows that the construction progress may be delayed, the construction management module will adjust the time arrangement of the relevant construction tasks in the simulation model, extend the duration of the corresponding tasks, or re-plan the construction sequence to reflect the actual possible progress changes. For example, due to weather reasons, the progress of the earthwork project in a certain construction area is delayed, and the time of the earthwork project task in this area in the simulation model will be extended accordingly, and it may affect the start time of subsequent related construction tasks. When the dynamic information that the equipment has a risk of failure is determined, the construction management module will update the operating status of the equipment in the simulation model, mark it as a possible failure state, and adjust the construction tasks related to the equipment. For example, reduce the workload of the equipment, or arrange other equipment to replace it, and adjust the construction schedule to adapt to the possible impact of equipment failure.

[0087] If the dynamic information indicates that the work efficiency of the construction personnel fluctuates, the construction management module will adjust the work parameters of the personnel in the simulation model, such as working hours, work intensity, etc. For example, if it is found that the work efficiency of a construction team has decreased, the amount of work completed by the team per unit time in the simulation model will be reduced accordingly, thereby affecting the simulation results of the entire construction progress. For dynamic information such as changes in construction conditions caused by environmental factors, such as rising water levels and deteriorating geological conditions, the construction management module will adjust the relevant environmental parameters and construction behavior mapping in the simulation model. For example, when the rising water level may affect the foundation construction of water conservancy facilities, the simulation model will add simulation of flood control measures and adjust the methods and schedule of foundation construction to cope with the impact of environmental changes.

[0088] In some embodiments, the construction management module is used to: update the model parameters of the water conservancy construction site simulation model according to the current construction data; start the updated water conservancy construction site simulation model to perform simulation calculations to obtain construction simulation results.

[0089] In the process of constructing the water conservancy construction site simulation model, the data used are all historical data in the database, so it is necessary to collect real-time data to fine-tune the model to ensure accuracy. In the embodiments of the present invention, in some embodiments of the intelligent water conservancy construction management platform, the construction management module updates the parameters of the water conservancy construction site simulation model based on the current construction data and simulates the calculation process, which is a key link in achieving precise construction management and scientific decision-making.

[0090] The construction management module first obtains the current construction data from the database module. These data cover all aspects of the water conservancy construction site, including but not limited to construction equipment operation data (such as engine speed, working hours, fuel consumption, etc.), construction personnel information (attendance, work location, etc.), construction progress data (completed engineering volume, progress of each construction link, etc.), water body data (water level, flow, water quality, etc.), geological data (real-time monitoring of geological changes, etc.) and ecological environment data (meteorological conditions, biodiversity changes, etc.). After obtaining the data, the construction management module will pre-process these data, including data cleaning (noise removal, processing of missing values ​​and outliers), data standardization (unified data format and dimension) and other operations to ensure the quality and availability of the data. For example, the noise interference in the equipment operation data is removed by filtering algorithm; for the missing construction progress data, a reasonable interpolation method is used to fill it.

[0091] If the current construction data shows that there is a difference between the actual progress of a construction task and the progress predicted by the simulation model, the construction management module will update the duration parameters, resource allocation parameters, etc. of the task. For example, if it is found in actual construction that the concrete pouring speed is faster than the model prediction, the duration parameters of the concrete pouring task will be shortened accordingly, and the relevant human and equipment resource parameters will be adjusted. When the actual stress conditions of the building structure are inconsistent with the results calculated by the simulation model, the construction management module will update the parameters of the structural mechanics model, such as the elastic modulus and Poisson's ratio of the material. For example, if the actual stress of a certain part of the dam body monitored by the stress sensor exceeds the model prediction range, it may be necessary to adjust the mechanical parameters of the material at that part to more accurately reflect the stress state of the structure. Based on the current water body data, such as changes in water level and flow, the construction management module will update the parameters of the hydraulic model, such as the water flow resistance coefficient, boundary condition parameters, etc. For example, when the actual water level is higher than the model predicted water level, it may be necessary to adjust the water flow resistance coefficient to more accurately simulate the movement of the water flow and its impact on the surrounding environment. After completing the model parameter update, the construction management module starts the updated water conservancy construction site simulation model for simulation calculation. The simulation calculation process is based on the updated model parameters, by solving the mathematical equations and logical relationships of each sub-model to simulate various physical processes and construction activities at the water conservancy construction site. Construction process simulation: According to the updated construction process model parameters, simulate the execution order of construction tasks, the allocation and use of resources, and predict the development trend of construction progress. For example, calculate the completion time of each construction task after the updated parameters, and the estimated completion time of the entire project. Use the updated structural mechanics model parameters to simulate the stress and deformation of the building structure under different working conditions. By solving the mechanical equations, the stress and strain distribution of the structure is obtained, and the safety and stability of the structure are evaluated. Based on the updated hydraulic model parameters, simulate the changes in parameters such as water flow speed, pressure, flow, and the impact of water flow on water conservancy facilities and the surrounding environment. For example, simulate the evolution of water flow under flood conditions to evaluate the effectiveness of flood control facilities.

[0092] After the simulation calculation is completed, the construction management module obtains the construction simulation results, which include but are not limited to information on construction progress forecast, resource demand forecast, equipment status forecast, quality risk assessment, environmental impact assessment, etc.

[0093] In some embodiments, the platform further comprises an evaluation module, which is used to evaluate the construction simulation results to obtain evaluation results. In some embodiments, the evaluation module is further used to determine a water conservancy project construction management plan based on the evaluation results and current construction data.

[0094] In an embodiment of the present invention, the evaluation module first needs to construct a comprehensive and scientific evaluation index system to quantitatively evaluate the construction simulation results. The evaluation module selects a suitable evaluation method for analysis and calculation based on the characteristics and data types of the evaluation index. After obtaining the evaluation results, the evaluation module deeply analyzes the relationship between the evaluation results and the current construction data. By comparing the evaluation results with the current construction data, existing problems and potential risks are found. For example, if the evaluation results show that the construction progress is lagging behind, and at the same time, a high failure rate of a key equipment is found in the current construction data, it can be inferred that the equipment failure may be one of the reasons for the delay in progress.

[0095] Based on the analysis results, the evaluation module formulates a targeted water conservancy project construction management plan. Specific measures include: Construction progress management: If the construction progress is delayed, measures such as increasing construction personnel and equipment, adjusting the construction sequence, and optimizing the construction process can be taken to speed up the progress; if the progress is ahead of schedule, resource allocation can be appropriately adjusted to avoid resource waste. For example, for the dam concrete pouring task with delayed progress, the number of concrete mixers and transport vehicles can be increased, and the pouring process can be optimized to improve construction efficiency.

[0096] Resource management: Re-plan resource allocation and scheduling in response to unreasonable resource utilization. For example, reasonably arrange the use time of equipment to avoid equipment idleness; optimize the material procurement plan to ensure that the supply of materials matches the construction progress. For example, according to the construction progress forecast, purchase the required building materials in advance to avoid construction interruptions due to material shortages.

[0097] Quality management: If there are problems with the project quality, strengthen quality inspection and testing, strictly implement construction specifications and standards, and promptly rectify unqualified construction sites. For example, increase the frequency of concrete strength testing and rework sites that do not meet strength standards.

[0098] Safety management: For situations with high safety risks, strengthen safety training and education, improve safety management systems, and add safety protection facilities. For example, provide construction workers with safety operating procedures training and set up obvious safety warning signs at the construction site.

[0099] Environmental management: If the environmental impact is significant, take corresponding environmental protection measures, such as strengthening soil and water loss control, controlling water pollution, protecting the ecological environment, etc. For example, set up a sedimentation tank at the construction site to treat construction wastewater before discharging it, reducing pollution to surrounding water bodies.

[0100] In addition, the evaluation module also needs to optimize and adjust the formulated construction management plan, and update the management plan in a timely manner according to the actual construction situation and new problems. For example, in the process of implementing the construction management plan, if new safety hazards or quality problems are found, the management plan should be adjusted in a timely manner and corresponding measures should be added to solve the problems. At the same time, the implementation effect of the management plan should be evaluated regularly, and the management plan should be continuously improved and perfected to improve the level and efficiency of water conservancy project construction management.

[0101] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0102] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. An intelligent water conservancy construction management platform, characterized in that: include: Database module, AI data analysis module and construction management module; The database module is used to obtain multi-source water conservancy project information; wherein the multi-source water conservancy project information includes the following types of data: water conservancy construction site data, water body data, geological data and ecological environment data; The AI ​​data analysis module includes multiple processing sub-blocks and an associated analysis block; Each processing sub-block is used to standardize various types of data; The association analysis block is used to perform association analysis on various types of data to obtain association features; wherein the association features include: direct time association features, direct space association features, indirect time association features and indirect space association features; The construction management module is used to construct a water conservancy construction site simulation model based on various types of standardized data and the associated features, and to obtain current construction data in real time, and to obtain construction simulation results based on the current construction data and the water conservancy construction site simulation model.

2. The intelligent water conservancy construction management platform according to claim 1 is characterized in that: The association analysis block is used to: Various types of data are input into a pre-established spatiotemporal convolutional neural network model to obtain the associated features.

3. The intelligent water conservancy construction management platform according to claim 2 is characterized in that: The spatiotemporal convolutional neural network model includes an input layer, a spatiotemporal convolution layer, a feature fusion layer, a feature separation layer, a correlation extraction layer, and an output layer. The correlation analysis block is used to: Input various types of data from the input layer into the spatiotemporal convolution layer to obtain multiple first feature maps; Inputting the multiple first feature maps into a feature fusion layer to obtain fused features; Inputting the fused features into a feature separation layer to obtain direct temporal correlation features and direct spatial correlation features; Inputting the fusion feature, the direct time association feature and the direct space association feature into the association extraction layer to obtain an indirect time association feature and an indirect space association feature; The direct time association features, direct space association features, indirect time association features and indirect space association features are output via an output layer.

4. The intelligent water conservancy construction management platform according to claim 3 is characterized in that: Each processing sub-block is used to: Each type of data is cleaned, standardized, converted and stored to obtain various types of data after annotation.

5. The intelligent water conservancy construction management platform according to claim 4 is characterized in that: The spatiotemporal convolutional neural network model further includes a parameter adjustment layer; the association analysis block is used to: Obtaining characteristic information of the standardized processing in each processing sub-block; Inputting the characteristic information into the parameter adjustment layer to obtain parameter adjustment information; According to the parameter adjustment information, the parameters of the spatiotemporal convolutional neural network model are adjusted.

6. The intelligent water conservancy construction management platform according to claim 5 is characterized in that: The Construction Management module is used to: Establish a physical model of the water conservancy construction site; Map the water conservancy construction site based on various data and the physical model to establish a digital twin model; According to the associated features, the digital twin model is dynamically simulated to obtain a water conservancy construction site simulation model.

7. The intelligent water conservancy construction management platform according to claim 6 is characterized in that: The Construction Management module is used to: Determining dynamic information of the construction site based on the association characteristics and the behavior prediction model; According to the dynamic information, the behavior mapping of the water conservancy construction site simulation model is adjusted.

8. The intelligent water conservancy construction management platform according to claim 7 is characterized in that: The Construction Management module is used to: According to the current construction data, updating the model parameters of the water conservancy construction site simulation model; Start the updated water conservancy construction site simulation model to perform simulation calculations and obtain the construction simulation results.

9. The intelligent water conservancy construction management platform according to claim 6 is characterized in that: The platform also includes an evaluation module, which is used to: The construction simulation result is evaluated to obtain an evaluation result.

10. The intelligent water conservancy construction management platform according to claim 7 is characterized in that: The evaluation module is also used to: A water conservancy project construction management plan is determined based on the evaluation results and the current construction data.

Images