Cognitive geologic modeling method and system based on BIM system

Abstract

The invention relates to the technical field of building information models (BIM) and geological modeling, in particular to a cognitive geological modeling method and system based on a BIM system, and the system comprises a data collection and preprocessing module and a mode recognition and feature extraction module. A modeling scheme and parameter optimization module, an automatic model generation module and a cross-chain data sharing module; the data collecting and preprocessing module is used for collecting historical data, collecting real-time data, filtering and cleaning the data, and integrating and standardizing the data; the mode recognition and feature extraction module is used for realizing mode recognition and feature extraction; the modeling scheme and parameter optimization module is used for realizing data-driven model generation, parameter adjustment and optimization and a real-time feedback mechanism; the model automatic generation module is used for realizing model automatic generation, and the model is integrated to the BIM system; and the cross-chain data sharing module is used for realizing data preparation and packaging, data encryption and verification, cross-chain data transmission and data decryption and verification.

Description

Technical Field

[0001] The present invention relates to the technical field of building information modeling (BIM) and geological modeling, and in particular to a cognitive geological modeling method and system based on a BIM system. Background Art

[0002] In construction engineering and geological exploration, geological modeling is an indispensable key link. Traditional geological modeling methods are usually based on fixed model parameters and rules. These methods have the following shortcomings:

[0003] (1) Lack of dynamic response capability: Existing geological modeling methods rely on static modeling parameters and cannot effectively respond to real-time changes in geological conditions. As the construction environment becomes more complex, geological conditions often change rapidly. Traditional methods find it difficult to adjust model parameters in real time, resulting in insufficient model accuracy and an inability to provide reliable engineering support.

[0004] (2) Low data utilization: With the development of Internet of Things technology, field sensors and equipment can collect a large amount of geological data in real time. However, existing technologies cannot fully utilize these real-time data for model optimization and only rely on historical data for modeling, ignoring the importance of real-time data, resulting in a disconnect between the model and the actual geological conditions.

[0005] (3) Data sharing and security issues: In large-scale engineering projects, cross-project data sharing is necessary. However, traditional data sharing methods often have security and consistency issues, and are prone to risks such as data tampering and leakage. In addition, the lack of unified data standards between different projects further increases the difficulty of data sharing.

[0006] Therefore, the existing technology urgently needs a geological modeling method that can dynamically adjust model parameters, make full use of real-time data, and ensure the security and consistency of data sharing. Summary of the invention

[0007] The purpose of the present invention is to provide a cognitive geological modeling method and system based on a BIM system to solve the technical problems existing in the prior art.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] The present invention provides a cognitive geological modeling method based on a BIM system, which comprises:

[0010] S1. Data collection and preprocessing, including the following steps:

[0011] S11. Collection of historical data; S12. Collection of real-time data; S13. Data filtering and cleaning; S14. Data integration and standardization;

[0012] S2, pattern recognition and feature extraction, including the following steps:

[0013] S21, pattern recognition; S22, feature extraction; S3, modeling scheme and parameter optimization, including the following steps:

[0014] S31. Data-driven model generation; S32. Parameter adjustment and optimization; S33. Real-time feedback mechanism;

[0015] S4, automatic model generation, including the following steps:

[0016] S41, automatic model generation; S42, model integration into BIM system;

[0017] S5. Cross-chain data sharing includes the following steps:

[0018] S51, data preparation and packaging; S52, data encryption and verification; S53, cross-chain data transmission;

[0019] S54. Data decryption and verification.

[0020] Preferably, S11, collection of historical data, specifically includes:

[0021] Data query and extraction: The system obtains a large amount of historical modeling data from the BIM database, and extracts data related to geological modeling from the relational database MySQL. Through standard SQL query statements, the system can automatically retrieve and extract data that is highly relevant to the current modeling task;

[0022] Data extraction and storage: The queried data will be stored in a temporary data storage area and prepared for further processing. The system ensures the efficiency of data extraction and the integrity of the data set through a unified interface;

[0023] S12. Collection of real-time data, including:

[0024] Real-time data collection: The system collects geological data in real time through cognitive IoT devices. The cognitive IoT devices are equipped with high-precision sensors that can monitor key geological parameters including soil moisture, formation pressure, and groundwater level. The monitored data is transmitted to the BIM system through a wireless sensor network;

[0025] Data transmission and processing: During the data transmission process, the system uses the transport layer security protocol for encryption to ensure the integrity and confidentiality of the data. After the data arrives, the system uses the real-time stream processing framework Kafka to process the data in real time, including data denoising and data compression;

[0026] S13. Data filtering and cleaning, including:

[0027] Data cleaning technology: The system uses the triple standard deviation method to identify and process outliers. For missing values, the system uses a linear interpolation algorithm to fill in the missing values ​​through the weighted average of adjacent data points to ensure the continuity and integrity of the data;

[0028] Data denoising: Before entering the analysis phase, the system denoises the data using a Gaussian filter algorithm;

[0029] S14. Data integration and standardization, including:

[0030] Data integration: unified processing and management of geological data from different sources and time periods to form a comprehensive data set, including data alignment and integration, data conversion and integration. In the data alignment and integration, historical data and real-time data are aligned and integrated through timestamp alignment and spatial coordinate matching technology; in the data conversion and integration, data from different data sources are extracted, converted and integrated through the ETL process;

[0031] Data standardization: The process of unifying data from different sources in terms of dimensions and units to ensure that the data is comparable and consistent during analysis. The standardization process includes two aspects: unit conversion and data normalization.

[0032] Preferably, S21, pattern recognition, specifically includes:

[0033] Data preparation and input: The system first inputs the preprocessed historical data into the random forest model. The data includes geological parameters and corresponding model accuracy indicators. During the input process, the system groups the data to ensure that each input sample contains a complete set of geological parameters.

[0034] Model training and pattern recognition: During the training process, the random forest algorithm constructs multiple decision trees through multiple random samplings. Each decision tree is split according to the input geological parameters, and finally forms a set of leaf nodes. Each leaf node represents the model accuracy under a geological condition.

[0035] S22, feature extraction, specifically including:

[0036] Convolution operation and feature mapping: The system inputs the geological data collected in real time by cognitive IoT devices into the convolutional neural network. In the convolutional neural network, the convolution layer performs convolution operation on the input data through the convolution kernel to extract the feature map.

[0037] Pooling and dimensionality reduction: After the convolution operation, the system uses a pooling layer to reduce the dimensionality of the feature map, retaining important features and reducing the data dimension;

[0038] Fully connected layer and feature extraction: After completing the convolution and pooling operations, the system integrates the feature maps through the fully connected layer. Through the fully connected layer, the system can generate the final feature vector, which represents the high-level features of the input data;

[0039] Feature output and model optimization: The extracted feature vectors will be used in the subsequent model optimization process to optimize the parameter settings and structural design of geological modeling by analyzing the features.

[0040] Preferably, S31, data-driven model generation, specifically includes:

[0041] Model Generation: XGBoost gradually optimizes the model’s predictive power throughout the training process by gradually building multiple decision trees, each of which attempts to correct the errors of the previous tree. In each iteration, XGBoost minimizes the loss function by calculating the residual and fitting a new tree.

[0042] Feature input: The system first inputs the feature vector generated from the pattern recognition and feature extraction steps into the XGBoost algorithm;

[0043] Modeling process: During the training process, XGBoost generates geological modeling solutions through the following steps:

[0044] Among them, initialization model: starting from the initial model structure and parameters, XGBoost initializes a simple model; among them, iterative training: in each iteration, XGBoost uses the residual between the predicted result and the actual result of the current model as the target value to fit a new tree, and the construction of each new tree aims to reduce the overall error; among them, update model: add the newly fitted tree to the existing model structure, update the overall model, and repeat this process multiple times until the predetermined stop condition is reached; among them, model output: the final generated model structure includes several decision trees, as well as the parameter settings and calculation strategies related to them;

[0045] Model structure and calculation strategy: The model output by XGBoost contains not only the structure of the tree, but also the split points associated with each tree, the values ​​of the leaf nodes, and the combined weights of the trees;

[0046] S32, parameter adjustment and optimization. After initially generating the geological modeling scheme, the system further optimizes the model parameters through the particle swarm optimization algorithm, including:

[0047] Initialize particle swarm: initialize a group of particles in the parameter space, each particle corresponds to a set of initial model parameters;

[0048] Evaluate fitness: Each particle generates a model based on the current parameter combination and uses historical data and real-time data to evaluate the model's prediction error. The smaller the model error, the higher the fitness.

[0049] Update particle position: Update the particle position according to the PSO algorithm formula, gradually approaching the area with the minimum error;

[0050] Iterative optimization: Repeat fitness evaluation and position update until a globally optimal parameter combination is found or a preset stop condition is reached;

[0051] S33. Real-time feedback mechanism, specifically including:

[0052] Real-time data monitoring: Real-time data monitoring is the basis of the real-time feedback mechanism. The system relies on cognitive IoT devices to continuously monitor and collect data on the geological environment;

[0053] Dynamic adjustment of model parameters: After identifying changes in geological conditions, the system automatically starts the dynamic adjustment mechanism of parameters to ensure that the model can adapt to these changes in a timely manner.

[0054] Preferably, S41, automatic model generation: the system converts the optimized geological modeling scheme into a complete three-dimensional geological model through an automated process, and implements it using GOCAD modeling software, specifically including:

[0055] Automatic conversion of modeling schemes: The system first converts the geological modeling scheme optimized in the early stage, including model structure, parameter settings, and calculation strategies, into an input format recognizable by GOCAD software;

[0056] Model verification and correction: After GOCAD generates the geological model, the system automatically performs preliminary model verification. If there are inconsistencies or errors in the model, the system automatically triggers the correction mechanism, readjusts the input parameters or data, and regenerates the model;

[0057] Execution and monitoring of automated processes: The entire modeling process is managed by the system's automated scheduling module, which is responsible for scheduling modeling tasks, allocating computing resources, and monitoring modeling progress. The system will generate detailed log records, covering the execution of each step, the parameters used, and the final generated model information. The generated logs are used for subsequent model review and traceability;

[0058] S42. Model integration into the BIM system, including:

[0059] Geological model import: convert the 3D geological model generated by GOCAD into a format that can be processed by the BIM system, and introduce it into the BIM environment to ensure the integrity and accuracy of the model;

[0060] Data exchange and integration: seamlessly integrate the imported geological model with other design modules in the BIM system to ensure the coordination of the overall design;

[0061] Use of integrated tools: Various integrated tools are used on the BIM platform to connect the geological model with other design modules to achieve collaborative work of multiple models.

[0062] Preferably, S51, data preparation and packaging, specifically includes:

[0063] Data extraction: The system first extracts the geological data to be shared from the BIM database;

[0064] Data packaging: The extracted data will be packaged into a standardized data package;

[0065] S52. Data encryption and verification, specifically including:

[0066] Data encryption: Before data transmission, cryptographic algorithms are used to convert plaintext data into ciphertext;

[0067] Data verification: After receiving cross-chain data, the system verifies the integrity and authenticity of the data through smart contracts;

[0068] S53. Cross-chain data transmission, including:

[0069] Selection of relay nodes: Responsible for transmitting data between different blockchain networks. Relay nodes capture and transmit relevant data to another blockchain network by monitoring events on one blockchain.

[0070] Data broadcast: The process by which a relay node transmits encrypted data packets from a source blockchain network to a target blockchain network;

[0071] Hash Time Lock Contract: A key mechanism to ensure the security and atomicity of cross-chain data transmission. HTLC uses smart contract technology to ensure the integrity of data transmission and avoid data loss or transaction failure due to network delays or node failures.

[0072] S54, data decryption and verification, specifically including: when the receiving blockchain network receives the data packet, the system will decrypt and verify the data to ensure the integrity and authenticity of the data.

[0073] In the second aspect, the present invention provides a cognitive geological modeling system based on a BIM system, which adopts the cognitive geological modeling method based on the BIM system, including: a data collection and preprocessing module, a pattern recognition and feature extraction module; a modeling scheme and parameter optimization module, a model automatic generation module, and a cross-chain data sharing module; the data collection and preprocessing module is used to realize the collection of historical data, the collection of real-time data, data filtering and cleaning, data integration and standardization; the pattern recognition and feature extraction module is used to realize pattern recognition and feature extraction; the modeling scheme and parameter optimization module is used to realize data-driven model generation, parameter adjustment and optimization, and real-time feedback mechanism; the model automatic generation module is used to realize model automatic generation and model integration into the BIM system; the cross-chain data sharing module is used to realize data preparation and packaging, data encryption and verification, cross-chain data transmission, and data decryption and verification.

[0074] By adopting the above technical solution, the present invention has the following beneficial effects:

[0075] The present invention introduces the XGBoost algorithm to automatically generate geological model solutions and optimize parameters in real time based on historical data and real-time data. Compared with the existing modeling methods that rely on manual adjustment, the present invention can generate models more quickly and accurately, significantly improving the reliability and applicability of geological models.

[0076] Through real-time data collection of cognitive IoT devices and dynamic parameter adjustment of the PSO algorithm, the present invention can continuously monitor changes in geological conditions and adjust model parameters in real time during the modeling process. This real-time feedback mechanism ensures the adaptability of the model throughout the life cycle of the project and reduces the project risks caused by changes in geological conditions.

[0077] The present invention combines cross-chain technology and smart contracts to achieve secure sharing and collaborative operation of geological data between different BIM projects. Compared with the traditional single-project data management model, the present invention provides a cross-project data integration and verification mechanism to ensure data consistency and integrity, and enhance the collaboration capabilities between engineering projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0079] Figure 1 A diagram of the architecture of a cognitive geological modeling system based on a BIM system provided in an embodiment of the present invention;

[0080] Figure 2 A flow chart of data collection and preprocessing provided by an embodiment of the present invention;

[0081] Figure 3 A flow chart of PSO optimization model parameters provided by an embodiment of the present invention;

[0082] Figure 4 A flow chart of cross-chain data sharing provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0083] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0084] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0085] Embodiment 1

[0086] Combination Figure 1 to Figure 2 As shown, this embodiment 1 provides a cognitive geological modeling method based on a BIM system, which includes:

[0087] S1. Data collection and preprocessing. Data collection and preprocessing are key steps in the geological modeling process. Ensuring the quality and consistency of input data directly affects the accuracy and reliability of the model, including the following steps:

[0088] S11. Collection of historical data;

[0089] The system first obtains a large amount of historical modeling data from the BIM database, including past geological modeling parameters, model building results, relevant geological environment data, and post-modeling verification results.

[0090] 1. Data query and extraction:

[0091] The system first obtains a large amount of historical modeling data from the BIM database, and then extracts data related to geological modeling from the relational database MySQL. Through standard SQL query statements, the system can automatically retrieve and extract data that is highly relevant to the current modeling task.

[0092] 2. Database structure: The MySQL database contains multiple tables, such as Geological_Parameters (geological parameter table), Model_Results (model result table), Environmental_Data (environmental data table) and Validation_Results (validation result table).

[0093] Query example: For a relational database, use the SQL statement to query:

[0094] {

[0095] sql

[0096] SELECT gp.structure, gp.soil_composition, ed.climate_conditions,ed.seismic_activity, mr.model_id, mr.parameters, vr.validation_score

[0097] FROM Geological_Parameters gp

[0098] JOIN Environmental_Data ed ON gp.location_id = ed.location_id

[0099] JOIN Model_Results mr ON gp.model_id = mr.model_id

[0100] JOIN Validation_Results vr ON mr.model_id = vr.model_id

[0101] WHERE gp.location_id = 'specific_location'AND mr.model_date BETWEEN 'start_date' AND 'end_date';

[0102] }

[0103] 3. Data extraction and storage: The data obtained from the query will be stored in a temporary data storage area and prepared for further processing. The system ensures the efficiency of data extraction and the integrity of the data set through a unified interface.

[0104] S12, real-time data collection;

[0105] Real-time data is an important basis for dynamic adjustment of the model. The system ensures that the model can reflect the current geological conditions through real-time collection and processing.

[0106] 1. Real-time data collection: The system collects geological data in real time through cognitive Internet of Things (CIoT) devices, which are equipped with high-precision sensors that can monitor key geological parameters including soil moisture, formation pressure, groundwater level, etc. These data are transmitted to the BIM system through wireless sensor networks (WSN).

[0107] Collection process: The sensor collects data at a preset sampling frequency (such as once per second) and transmits it to a centralized data server via a wireless network, and finally forwards it to the real-time data warehouse in the BIM system.

[0108] 2. Data transmission and processing: During the data transmission process, the system uses the Transport Layer Security Protocol (TLS) for encryption to ensure the integrity and confidentiality of the data. After the data arrives, the system uses the real-time stream processing framework Kafka to process the data in real time, including data denoising and data compression.

[0109] Data denoising: The system uses Fourier Transform to perform frequency domain analysis to remove high-frequency noise in the data and retain effective low-frequency signals. The processed data is further reduced in size through data compression algorithms to ensure efficient processing.

[0110] S13, data filtering and cleaning;

[0111] After acquiring the data, the system filters and cleans the data according to the set rules. It uses rule-based filtering algorithms and data cleaning techniques to process redundant or irrelevant data and handle missing values ​​and outliers.

[0112] 1. Data cleaning technology:

[0113] The system uses the 3-sigma Rule to identify and handle outliers. The formula is as follows:

[0114]

[0115]

[0116] Where μ is the mean of the data set and σ is the standard deviation. Values ​​outside the upper and lower bounds will be marked as outliers.

[0117] For missing values, the system uses a linear interpolation algorithm to fill in the missing values ​​through the weighted average of adjacent data points to ensure the continuity and integrity of the data.

[0118] 2. Data denoising: Before entering the analysis phase, the system denoises the data using the Gaussian filter algorithm.

[0119] Denoising formula: Using Gaussian filter formula:

[0120]

[0121] in, is the filtered value, is the standard deviation of the filter. The algorithm smoothes the data and removes noise to ensure the quality and continuity of the data.

[0122] S14. Data integration and standardization;

[0123] Data integration and standardization are key steps in the geological modeling process, which aims to unify the historical data and real-time data extracted from different sources or time periods to form a structured data set that can be used for analysis. This step ensures that the data from different sources are consistent in dimension, unit and scale, so that the model can consistently interpret and process the data. The specific process of data integration and standardization is as follows:

[0124] 1. Data Integration

[0125] Data integration refers to the unified processing and management of geological data from different sources and time periods to form a comprehensive data set. In this process, the system needs to solve the heterogeneity problem between data, including differences in data format, unit, dimension, etc.

[0126] Data source and type:

[0127] Historical data: including geological modeling parameters, model construction results, environmental data, and verification results, etc., are usually stored in the database of the BIM system.

[0128] Real-time data: On-site geological parameters such as soil moisture, ground pressure, and groundwater level collected in real time through cognitive Internet of Things (CIoT) devices.

[0129] Data alignment and integration: The system first aligns and integrates historical data and real-time data through timestamp alignment and spatial coordinate matching technology. Timestamp alignment ensures the consistency of data in the time dimension, that is, data at different time points can be integrated according to the time series; spatial coordinate matching ensures the consistency of data in the spatial dimension, that is, data at different locations can be accurately matched to the corresponding geographical location.

[0130] Timestamp alignment: For each set of data, the system uses linear interpolation to align data collected at different times to the same time node based on its collection timestamp. For example, if the historical data collection frequency is once an hour, and the real-time data is once a minute, the system will align the historical data to the minute level through interpolation to ensure the time consistency between the data.

[0131] Spatial coordinate matching: The system matches spatial data from different data sources based on the spatial coordinates of the data. For example, formation pressure data and groundwater level data need to be matched to the same geographical location to ensure that the model can correctly reflect the spatial geological conditions during analysis.

[0132] Data conversion and integration: The system uses the ETL (Extract, Transform, Load) process to extract, transform and integrate data from different data sources. The extraction process involves obtaining data from various databases or sensor interfaces, the conversion process involves converting data in different formats into a unified standard format, and the integration process is to store the converted data in a unified database.

[0133] 2. Data Standardization

[0134] Data standardization is the process of unifying data from different sources in terms of dimensions and units to ensure that the data is comparable and consistent when analyzed. The standardization process mainly involves two aspects: unit conversion and data normalization.

[0135] Unit conversion: In geological modeling, data may be expressed in different units, for example, formation pressure may be expressed in Pascal (Pa) or kilopascal (kPa). The system uses a unit conversion algorithm to convert all data into a unified unit to eliminate the impact of unit differences on data analysis.

[0136] Unit conversion formula: When the system converts kilopascals (kPa) to pascals (Pa), it uses the following conversion formula:

[0137]

[0138] Through this formula, all values ​​representing formation pressure will be unified into Pascal (Pa) to ensure data consistency.

[0139] Data normalization: Data normalization is to compress data of different scales into the same numerical range (usually 0 to 1) to eliminate differences in magnitude so that different types of data can be treated equally during analysis.

[0140] The normalization formula is as follows:

[0141]

[0142] Among them, x is the original data, and are the minimum and maximum values ​​in the data set, respectively. are normalized values. This step ensures that the data have consistent magnitude and proportion when analyzed.

[0143] Standardized data storage: After completing the data standardization process, the system stores all data in a unified data warehouse. The data in the warehouse has been standardized into a unified format and dimension, ensuring that all data can be interpreted and processed consistently during model analysis and processing.

[0144] Through the process of data integration and standardization, the system can unify data from different sources and time periods into a structured and standardized data set, ensuring consistency in the dimensions, units and scales of the data in the subsequent modeling process, thereby improving the accuracy and reliability of the geological model.

[0145] S2. Pattern recognition and feature extraction. Pattern recognition and feature extraction are important steps in geological modeling. The system can identify key patterns and features from a large amount of data, providing a basis for subsequent model construction and optimization. The following is the specific implementation process of pattern recognition and feature extraction:

[0146] S21, pattern recognition;

[0147] Pattern recognition is to identify the relationship pattern between geological parameters and modeling accuracy by analyzing historical data. In the present invention, the system uses the random forest algorithm to analyze historical data and extract key features that affect the accuracy of geological modeling. Random Forest is an integrated learning method that performs classification or regression by constructing multiple decision trees and combining the prediction results of these trees. In geological modeling, the random forest algorithm can effectively process high-dimensional data and identify the complex relationship between geological parameters and model accuracy.

[0148] 1. Data preparation and input:

[0149] The system first inputs the preprocessed historical data into the random forest model. These data include geological parameters (such as stratum structure, soil composition, groundwater level, etc.) and corresponding model accuracy indicators (such as model error, accuracy score, etc.). During the input process, the system groups the data to ensure that each input sample contains a complete set of geological parameters.

[0150] 2. Model training and pattern recognition:

[0151] During the training process, the random forest algorithm constructs multiple decision trees through multiple random samplings. Each decision tree is split according to the input geological parameters, and finally forms a set of leaf nodes, each of which represents the model accuracy under a geological condition.

[0152] Decision tree construction formula:

[0153] The decision tree construction process uses information gain or Gini Index as the splitting criterion. The information gain calculation formula is:

[0154]

[0155] in, Representation dataset The entropy of is the ith subset after splitting, and are the sizes of the data sets before and after the split, respectively. By maximizing information gain, the system selects the best split point to build the optimal decision tree.

[0156] Pattern Recognition:

[0157] Random forests identify patterns in the relationship between geological parameters and modeling accuracy by integrating multiple decision trees. For example, certain stratigraphic structures are significantly correlated with model errors, while other parameters (such as soil composition) may have a greater impact on model accuracy. Through these patterns, the system is able to determine which geological parameters are the key factors affecting model accuracy.

[0158] Output and result explanation:

[0159] The trained random forest model will output a set of feature importance scores, each of which represents the degree of influence of a geological parameter on the model accuracy. Based on these scores, the system will extract key features, such as stratigraphic structure and physical properties, as the basis for subsequent feature extraction and model optimization.

[0160] S22, feature extraction;

[0161] Feature extraction is the process of extracting features from real-time data that have an important impact on model building. This process uses deep learning technology, especially convolutional neural networks (CNN), to automatically extract high-level features from complex geological data to support further optimization of the model. Deep learning is a technology that uses multi-layer neural networks to extract features and recognize patterns. Convolutional neural networks (CNN) are an important model of deep learning and are particularly suitable for processing data with spatial structures. In geological modeling, CNN can effectively process multi-dimensional data from sensors (such as two-dimensional or three-dimensional geological profiles) and automatically extract key features.

[0162] 1. Convolution operation and feature mapping:

[0163] The system inputs geological data collected in real time by cognitive Internet of Things (CIoT) devices into the convolutional neural network. These data include two-dimensional geological profiles, three-dimensional geological structures, etc.

[0164] In a convolutional neural network, the convolution layer performs convolution operations on the input data through convolution kernels (Filters) to extract feature maps. The mathematical representation of the convolution operation is:

[0165]

[0166] in, represents the i-th channel of the input data, is the corresponding convolution kernel, * indicates the convolution operation, is the bias term, It is an activation function, and the ReLU function is used as the activation function in the present invention. Through multi-layer convolution operations, CNN can extract features of different levels from the input data.

[0167] 2. Pooling and dimensionality reduction:

[0168] After the convolution operation, the system uses the pooling layer to reduce the dimension of the feature map, retaining important features and reducing the data dimension. The pooling operation usually uses the maximum pooling, which is mathematically expressed as:

[0169]

[0170] in, The function takes the maximum value in the local area, representing the most significant feature.

[0171] 3. Fully connected layer and feature extraction:

[0172] After completing the convolution and pooling operations, the system integrates the feature maps through the fully connected layer. In the deep learning model, the fully connected layer is a key component in the convolutional neural network, which is used to integrate and classify the features extracted by the previous convolutional layer and pooling layer. Through the fully connected layer, the system can generate the final feature vector. This feature vector represents the high-level features of the input data, which are critical to the accuracy and precision of geological modeling.

[0173] The fully connected layer is a neural network layer that connects every node in the input layer to every node in the output layer. The main function of this layer is to integrate the features extracted by the previous layer to generate a comprehensive feature vector, thereby providing input for the final classification or regression task.

[0174] Mathematical expression of fully connected layer: The operation of fully connected layer can be expressed as a linear transformation followed by a nonlinear activation function. The specific expression is:

[0175]

[0176] in:

[0177] represents the output of the jth node in the fully connected layer.

[0178] is the weight connecting input node i and output node j.

[0179] is the i-th element of the input feature map.

[0180] is the bias term of the jth node.

[0181] Is the activation function. Commonly used activation functions are ReLU and Sigmoid.

[0182] Generation of feature vector: The output of the fully connected layer is usually a one-dimensional vector, namely the feature vector (FeatureVector). This vector contains all the important features of the input data and is the result of extraction and integration through multiple layers of convolution and pooling operations. The feature vector represents the high-level information of the input data and can be used to describe key features in geological modeling, such as stratum structure, soil composition, etc. Through the feature vector generated by the fully connected layer, the system can efficiently model and predict geological data.

[0183] 4. Feature output and model optimization:

[0184] The extracted feature vectors will be used in the subsequent model optimization process. By analyzing these features, the system optimizes the parameter settings and structural design of geological modeling to ensure that the model can accurately reflect the changes in geological conditions and improve prediction accuracy.

[0185] Based on pattern recognition and feature extraction modules, the system can extract the most critical features for geological modeling from historical data and real-time data, and use the features for subsequent model optimization to ensure the accuracy and adaptability of the geological model to meet actual engineering needs.

[0186] S3. Modeling scheme and parameter optimization. In the geological modeling process, data-driven model generation, parameter adjustment and optimization, and real-time feedback mechanism are key steps to ensure that the model can accurately reflect the changes in geological conditions and improve the prediction accuracy. The following is the specific implementation process of modeling scheme and parameter optimization:

[0187] S31, data-driven model generation;

[0188] XGBoost (Extreme Gradient Boosting) algorithm is the core algorithm of this system for automatically generating preliminary geological modeling schemes. XGBoost is an ensemble learning algorithm based on gradient boosting, which is efficient, accurate, and has strong parallel computing capabilities. It is particularly suitable for processing large-scale and complex data sets.

[0189] 1. Model generation: XGBoost gradually optimizes the model's predictive power throughout the training process by gradually building multiple decision trees, each of which attempts to correct the error of the previous tree. In each iteration, XGBoost minimizes the loss function by calculating the residual and fitting a new tree. The algorithm introduces regularization terms in the process of building the tree to prevent overfitting, and adopts a more efficient tree structure search algorithm and node splitting strategy to ensure the efficiency and stability of the model.

[0190] 2. Feature input: The system first inputs the feature vector generated from the pattern recognition and feature extraction steps into the XGBoost algorithm. The feature vector contains key geological parameters extracted through deep learning and statistical analysis, such as stratigraphic structure, soil composition, groundwater level, etc.

[0191] 3. Modeling process: During the training process, XGBoost generates a geological modeling solution through the following steps:

[0192] Initialize the model: Using the initial model structure and parameters as a starting point, XGBoost initializes a simple model.

[0193] Iterative training: In each iteration, XGBoost uses the residual between the current model's prediction and the actual result as the target value to fit a new tree. Each new tree is constructed with the goal of reducing the overall error.

[0194] Update model: Add the newly fitted tree to the existing model structure and update the overall model. This process is repeated multiple times until the predetermined stopping condition is reached (such as reaching the maximum number of iterations or the error is no longer significantly reduced).

[0195] Model output: The final generated model structure includes several decision trees, as well as the associated parameter settings and calculation strategies. These structures and parameters together constitute the preliminary geological modeling scheme.

[0196] 4. Model structure and calculation strategy: The model output by XGBoost contains not only the structure of the tree, but also the split points associated with each tree, the values ​​of the leaf nodes, and the combined weights of the trees. These model parameters determine how to predict new geological data and provide the calculation strategy for the model in practical applications.

[0197] S32, parameter adjustment and optimization;

[0198] After the geological modeling scheme is initially generated, the system further optimizes the model parameters through the particle swarm optimization algorithm (PSO). PSO is a global optimization algorithm based on swarm intelligence, which simulates the foraging process of bird flocks in nature and can effectively search for the optimal solution in the parameter space.

[0199] 1. PSO algorithm: PSO moves in the search space through a group of "particles", each particle represents a possible parameter combination. Particles adjust their positions based on their own experience and group experience, gradually approaching the optimal solution of the objective function. The position update formula of each particle is:

[0200]

[0201]

[0202] in, represents the velocity of particle i in the tth generation, is the position of the particle, is the best position in the particle's history, is the global best position, is the inertia weight, and is the acceleration constant, and is a random number.

[0203] 2. Parameter optimization process:

[0204] Initialize particle swarm: Initialize a group of particles in the parameter space, each particle corresponds to a set of initial model parameters.

[0205] Evaluate fitness: Each particle generates a model based on the current parameter combination and uses historical data and real-time data to evaluate the model's prediction error. The smaller the model error, the higher the fitness.

[0206] Update particle position: Update the particle position according to the PSO algorithm formula, gradually approaching the area with the smallest error.

[0207] Iterative optimization: Repeat fitness evaluation and position update until a globally optimal parameter combination is found, or a preset stopping condition is reached (such as reaching the maximum number of iterations or the model error is lower than a threshold).

[0208] 3. Application of optimization results: The optimized parameters will be used to adjust the geological modeling scheme to ensure that the model can better adapt to the actual geological conditions. The optimization process can include the model's hyperparameters (such as tree depth, number of leaf nodes, learning rate, etc.) to further improve the model's prediction accuracy and stability.

[0209] S33, real-time feedback mechanism;

[0210] To ensure that the geological model can continue to adapt to environmental changes, the system has designed a real-time feedback mechanism to dynamically adjust the model parameters during the modeling process. The real-time feedback mechanism enables the model to maintain its accuracy and reliability in the face of changing geological conditions.

[0211] 1. Real-time data monitoring: Real-time data monitoring is the basis of the real-time feedback mechanism. The system relies on cognitive Internet of Things (CIoT) devices to continuously monitor and collect data on the geological environment.

[0212] Data collection of CIoT devices: The CIoT devices deployed in the system include a variety of high-precision sensors that can monitor key geological parameters such as groundwater level, formation pressure, soil moisture, etc. in real time. These sensors collect data at a predetermined frequency (such as once per second) and transmit the data to the central data processing unit through a wireless network or a wired network.

[0213] Real-time data input and analysis: The system inputs the collected real-time data into the geological model for immediate analysis. The data stream processing framework Kafka is used to process large-scale real-time data, ensuring that the data can be received and processed by the model with low latency. The system continuously monitors this data to identify abnormal changes or trends in geological conditions. For example, when the groundwater level suddenly rises or the formation pressure fluctuates abnormally, the system will immediately mark these changes and prepare to adjust the model parameters.

[0214] 2. Dynamic adjustment of model parameters: After identifying changes in geological conditions, the system automatically starts the dynamic adjustment mechanism of parameters to ensure that the model can adapt to these changes in a timely manner.

[0215] Deviation analysis and trigger mechanism: The system compares the real-time data with the model prediction results and calculates the deviation between the two. If the deviation exceeds the preset threshold, the system will automatically trigger the parameter adjustment mechanism. This trigger mechanism is based on the set response strategy to ensure that the model can quickly adjust in the face of significant environmental changes.

[0216] PSO algorithm adjusts parameters: The system uses the particle swarm optimization algorithm (PSO) to readjust the model parameters. The PSO algorithm gradually optimizes the model parameters by simulating the movement of the particle swarm in the parameter space to minimize the error between the real-time data and the model prediction results. The adjustment process is continuous and automatic. The PSO algorithm continuously updates the model parameters through multiple iterations until the best parameter combination is found to ensure that the model can respond to environmental changes in a timely manner.

[0217] Automated adjustment process: The dynamic parameter adjustment process is fully automated. The system can optimize parameters based on real-time monitoring results without human intervention. This automation ensures that the model can maintain the best state at all times and adapt to dynamic changes in the environment.

[0218] S4, automatic model generation, including the following steps:

[0219] S41, automatic model generation;

[0220] Automatic model generation means that the system converts the optimized geological modeling scheme into a complete three-dimensional geological model through an automated process, and implements it using GOCAD modeling software.

[0221] 1. Automatic conversion of modeling scheme: The system first converts the geological modeling scheme optimized in the early stage, including model structure, parameter settings, calculation strategy, etc., into an input format recognizable by GOCAD software. GOCAD is a powerful 3D geological modeling software widely used in geological exploration and underground structure analysis.

[0222] Data input and format conversion: The system converts geological data and optimization parameters into file formats supported by GOCAD, such as TSurf, Voxet, GOCAD Object, etc. These formats can describe complex three-dimensional structures such as stratigraphic interfaces, geological bodies, attribute distribution, etc. The data conversion process involves converting the system-optimized modeling scheme into GOCAD input files and ensuring that the data format is consistent with GOCAD requirements.

[0223] Generation of automated modeling scripts: In order to achieve automation, the system generates a set of GOCAD scripting language (GSLib) scripts. These scripts can automatically call GOCAD's modeling functions to complete the generation of models. The script content includes the construction of three-dimensional geological bodies, attribute assignment, model grid generation, etc. By calling these scripts, GOCAD can automatically complete the modeling task without human intervention. The system generates GOCAD scripts through the following steps to achieve automated modeling:

[0224] 1.1 Preparation and formatting of input data

[0225] Data Collection: The system first extracts data related to the current geological modeling task from the BIM database, including stratigraphic boundaries, geological properties, measurement data, etc. These data are usually stored in standard formats such as CSV, TXT or dedicated geological data formats (such as Well Logs, Seismic Data, etc.).

[0226] Data formatting: The extracted data needs to be converted into a format that GOCAD can handle. For example, geological boundary data can be converted into TSuRF format (used to define stratigraphic interfaces), while geological attribute data can be converted into Voxet format (used to represent three-dimensional attribute grids). The system automatically completes the data format conversion through the data preprocessing module and generates data files corresponding to the GSLib script.

[0227] 1.2 Construction of 3D geological bodies

[0228] Strata Construction: The GSLib script first defines the 3D geometry of the strata. The system generates TSuRF objects based on the strata boundary data, which represent the upper and lower interfaces of the geological layers. The script uses GOCAD's TSURF command to create these objects and set names, colors, and other properties for each stratum.

[0229] TSurf MyGeologicalLayer {

[0230] NAME "Layer1"

[0231] COLOR 1 0 0 1

[0232] VERTEX_FILE "Layer1_vertices.xyz"

[0233] TRIANGLE_FILE "Layer1_triangles.tri"

[0234] }

[0235] Geological body generation: After the formation is constructed, the GSLib script uses these interface data to generate three-dimensional geological bodies. These geological bodies can be generated through Boolean operations between formations (such as intersection, union, etc.), or through polygon stretching and rotation to generate complex geometric shapes.

[0236] GEOLOGICAL_OBJECT MyGeoBody {

[0237] BASE TSurf "Layer1"

[0238] EXTRUDE 100

[0239] NAME "GeoBody1"

[0240] }

[0241] 1.3 Attribute Assignment

[0242] Attribute interpolation and assignment: GSLib scripts include interpolation and assignment operations for geological attributes. The system generates Voxet objects based on geological attribute data (such as density and porosity). These objects represent the distribution of attributes in three-dimensional space. GSLib scripts use VOXET commands to define these objects and assign attributes through interpolation algorithms.

[0243] VOXET MyVoxet

[0244] DIMENSIONS 100 100 100

[0245] ORIGIN 0 0 0

[0246] SPACING 10 10 10

[0247] PROPERTY "Density"

[0248] INTERPOLATE INVERSE_DISTANCE

[0249] }

[0250] 1.4 Model mesh generation

[0251] Mesh generation and optimization: GSLib scripts generate computational meshes for geological models by defining mesh parameters. Mesh generation involves discretization of three-dimensional geological bodies, usually including vertex generation, triangular facet division, etc. GSLib scripts allow users to define mesh density, node distribution and other parameters to optimize computational performance.

[0252] GRID MyMesh {

[0253] FROM TSurf "Layer1"

[0254] SPACING 5 5 5

[0255] GENERATE

[0256] }

[0257] Mesh optimization: In order to ensure the quality and computational efficiency of the mesh, the system will call GOCAD's mesh optimization function through the GSLib script. This includes mesh smoothing, node merging, boundary adjustment and other operations to ensure that the generated mesh meets the computational requirements.

[0258] MESH_OPTIMIZATION MyMesh {

[0259] SMOOTH

[0260] MERGE_NODES TOLERANCE 0.01

[0261] }

[0262] 2. Model verification and correction: After GOCAD generates the geological model, the system automatically performs preliminary model verification. The verification includes geometric consistency check (such as no intersection between strata and no gaps in geological bodies), parameter integrity check (such as all attributes have been correctly assigned), etc. If there are inconsistencies or errors in the model, the system will automatically trigger the correction mechanism, readjust the input parameters or data, and regenerate the model.

[0263] 2.1 Geometric consistency check

[0264] Geometric consistency check is the first step of model verification, which is used to ensure that the generated 3D geological model has no errors in geometric structure. The check mainly includes the following two aspects:

[0265] Stratigraphic cross-check: In geological modeling, different strata should be distributed strictly according to geological laws to avoid crossing or overlapping each other. The system uses Boolean operations to detect the intersections between strata. If any intersections are detected, the system will mark these areas and output a warning message.

[0266] The system checks the stratum interface layer by layer through the Boolean operation module, and uses GOCAD's built-in geometric operation functions, such as Intersect, Union, etc., to calculate and detect the intersections between strata. The system will output the volume or area of ​​the intersection and mark the error location.

[0267] INTERSECT Layer1 Layer2 OUTPUT IntersectVolume

[0268] Geological body gap check: The geological body in the geological model should be continuous and void-free to ensure the integrity of the model. The system uses mesh analysis technology to check whether there are unfilled gaps or caves inside the geological body.

[0269] The system uses GOCAD's mesh detection function to perform a closed check (MeshClosure) on the mesh of each geological body. The system checks whether the vertices, edges, and faces of the mesh are connected and whether a closed three-dimensional geometry is formed. If a gap is detected, the system will output the location and volume of the gap and mark it as an error.

[0270] MESH_CHECK CLOSURE MyGeoBody OUTPUT Errors

[0271] 2.2 Parameter integrity check

[0272] Parameter integrity checking aims to ensure that every part of the geological model is correctly assigned values ​​and that all physical properties (e.g., density, porosity) are configured as per the design requirements.

[0273] Attribute assignment check: The system checks whether each geological body in the model is correctly assigned the necessary attributes, such as density, porosity, permeability, etc. If a geological body is missing an attribute value, the system will mark the geological body and prompt the user.

[0274] The system checks the properties of each geological body one by one through the Property Validation Module. The system calls the property query functions in GOCAD, such as GetProperty, to check whether all required properties have been correctly assigned. For properties that are not assigned or assigned incorrectly, the system records relevant information and generates an error report.

[0275] CHECK_PROPERTY MyGeoBody "Density" OUTPUT Errors

[0276] Attribute consistency verification: In addition to checking whether the attribute has been assigned a value, the system also verifies the consistency of the attribute value. For example, the attribute values ​​in the same stratum should conform to a certain geological law or value range. The system will perform statistical analysis to check whether there are abnormal fluctuations or incorrect distribution of attribute values.

[0277] The system uses statistical analysis tools to check the consistency of attribute values. By calculating the mean, standard deviation and distribution range of attribute values, the system is able to detect outliers. The system uses Z-score detection to identify and mark outliers in attribute values.

[0278] STATISTICS MyGeoBody "Density" OUTPUT Stats

[0279] DETECT_OUTLIERS Stats OUTPUT Errors

[0280] 2.3 Automated Correction Mechanism

[0281] If problems are found during model verification, the system will automatically trigger a correction mechanism to readjust input parameters or data and regenerate the model. The purpose of the correction mechanism is to fix errors in the model in an automated way to ensure that the model meets the design requirements.

[0282] Automatic parameter adjustment: When geometric errors or attribute inconsistencies are found, the system automatically adjusts the relevant input parameters. If a stratum intersection problem is found, the system will adjust the depth or dip parameters of the stratum; if an attribute inconsistency is found, the system will re-execute the attribute interpolation operation.

[0283] The system calls the parameter adjustment module and automatically modifies the input parameters according to the verification results. The adjusted parameters are re-entered into GOCAD to trigger a new modeling process. The system uses the optimization algorithm particle swarm optimization (PSO) to find the optimal parameter combination to avoid the same error again.

[0284] ADJUST_PARAMETER Layer1 "Depth" NEW_VALUE 150

[0285] REGENERATE_MODEL

[0286] Remodeling: After completing the parameter adjustment, the system will call GOCAD's modeling function to generate a new model. The new model will go through the validation process again to ensure that all errors have been fixed. If the problem still exists, the system will continue to adjust the parameters and repeat the modeling until the model passes all validation checks.

[0287] The system performs the remodeling process through an automated modeling script (GSLib), ensuring that the adjusted model meets the design requirements. The regenerated model undergoes a complete geometric and parametric validation until the model passes all checks.

[0288] REGENERATE_MODEL MyGeoBody NEW_PARAMETERS

[0289] 3. Execution and monitoring of automated processes: The entire modeling process is managed by the system's automated scheduling module, which is responsible for scheduling modeling tasks, allocating computing resources, and monitoring modeling progress. The system generates detailed log records covering the execution of each step, the parameters used, and the final generated model information. These logs are used for subsequent model review and traceability.

[0290] Script execution: After the script is generated, the system will automatically start GOCAD and execute the GSLib script. The script execution process includes model creation, attribute assignment and mesh generation. All operations are automatically completed by the script without manual intervention.

[0291] Execution monitoring and logging: The system monitors the running status of the script in real time during execution and records logs. These logs include every step of the execution, the intermediate results generated, and the final model information. If an error or exception is encountered, the system will automatically pause the script execution and generate an error report for subsequent analysis.

[0292] GOCAD modeling process: The system automatically performs the following modeling process through GOCAD modeling software:

[0293] Geological layer division: GOCAD generates the three-dimensional interface of each stratum based on the input geological data. The system generates the geometric shape and spatial distribution of the stratum based on geological parameters (such as stratum depth and geological contact relationship).

[0294] Geological body construction: The system uses a variety of construction tools (such as polygon construction and 3D interpolation) in GOCAD to generate geological body models. Each geological body represents an independent geological unit, such as a rock layer, fault or ore body.

[0295] Attribute assignment: The system assigns physical properties to the generated geological bodies, such as density, porosity, permeability, etc. These attributes are usually calculated and distributed in GOCAD based on the collected data through interpolation algorithms (such as Kriging interpolation and inverse distance weighted interpolation).

[0296] S42, model integration into BIM system;

[0297] During the project design and construction process, importing and integrating the 3D geological model generated by GOCAD into the BIM system is a key step to ensure the consistency and coordination of the overall design.

[0298] 1. Geological model import

[0299] Geological model import refers to converting the three-dimensional geological model generated by GOCAD into a format that can be processed by the BIM system, and introducing it into the BIM environment to ensure the integrity and accuracy of the model.

[0300] 1.1 Data format conversion: The three-dimensional geological models generated by GOCAD usually use specific formats (such as TSurf, Voxet, GOCAD Object, etc.), which need to be converted into standardized data formats supported by the BIM system to achieve seamless integration.

[0301] Technology used:

[0302] Data format conversion tool: The system uses a dedicated data format conversion tool FME (FeatureManipulation Engine). FME is a powerful data integration and conversion platform that supports the conversion and processing of hundreds of geospatial data formats. It can be used to convert GOCAD models (such as TSuRF or Voxet formats) into formats supported by BIM systems (such as IFC or CityGML). These formats are standards in the fields of architecture and infrastructure and support extensive interoperability.

[0303] Model data mapping: During the conversion process, the system performs model data mapping to ensure that the geometry, topology and attribute information in the GOCAD model are accurately converted to the corresponding data structure in the BIM system. For example, the stratum interface in GOCAD will be mapped to IfcGeometricRepresentationItem in IFC, and the attribute data will be mapped to IFC Property Sets.

[0304] Convert GOCAD models to IFC format:

[0305] convert_gocad_to_ifc(input_file="gocad_model.ts", output_file="bim_model.ifc")

[0306] 1.2 Model import: After completing the format conversion, the system imports the geological model into the BIM environment. The import process involves loading the converted model file into the BIM platform and docking it with the existing building and infrastructure models.

[0307] Technology used:

[0308] BIM data import interface: The system uses the data import interface provided by the BIM platform (such as Revit's IFCImporter) to import the geological model into the BIM environment. The import interface supports the complete import of the geological model's data structure, attribute information, spatial relationships, etc. into the BIM platform.

[0309] Model verification and correction: During the model import process, the system automatically verifies whether the model's geometry and attribute data are consistent with the existing data in the BIM environment, and checks for data loss or errors. If necessary, the system automatically triggers the correction mechanism to adjust the model data to ensure its compatibility with the BIM system.

[0310] Import IFC models using BIM API

[0311] bim_api.import_model("bim_model.ifc")

[0312] 2. Data exchange and integration

[0313] Data exchange and integration refers to the seamless integration of the imported geological model with other design modules (such as buildings, structures, and pipelines) in the BIM system to ensure the coordination of the overall design.

[0314] 2.1 Data exchange: The imported geological model needs to exchange data with other modules in the BIM system. This process ensures that the geographical location, size, and attribute information of the geological model are consistent with the data of the building, structure and other modules.

[0315] Technology used:

[0316] Data exchange standards: The system uses standardized data exchange formats (such as IFC and CityGML) for data exchange, which support extensive interoperability and data consistency. Among them, the IFC format can describe complex 3D geometry, attribute information and spatial relationships, and supports the exchange of geological and architectural information between BIM platforms.

[0317] Model Coordinate System Conversion: The system performs coordinate system conversion, converting the geographic coordinate system of the GOCAD model to the engineering coordinate system used by the BIM system to ensure that the geological model is spatially aligned with the architectural model.

[0318] Convert coordinate system:

[0319] convert_coordinates(input_model="bim_model.ifc", target_cs="engineering_cs")

[0320] 2.2 Model integration: After the data exchange is completed, the system integrates the geological model with other design modules in the BIM system to ensure consistency and collaborative work between models at the geometric and attribute levels.

[0321] Technology used:

[0322] Integration tools of BIM platform: The system uses the integration tools provided by BIM platform (such as Link IFC tool in Revit) for model integration. The integration tools can align the imported geological model with the existing building and structural models and map the attributes to ensure the coordination between the models.

[0323] Geometric consistency check and adjustment: The system performs a geometric consistency check to ensure that the geological model is spatially aligned with the building model. The system checks the interface between the geological layer and the building foundation to ensure that there is no conflict in space. If inconsistency is found, the system automatically adjusts the geometric position or shape of the model to achieve perfect integration.

[0324] Model integration using BIM tools:

[0325] bim_platform.integrate_models(geological_model="bim_model.ifc",architectural_model="arch_model.ifc")

[0326] 3. Use of integrated tools

[0327] The use of integrated tools refers to the use of various integrated tools on the BIM platform to connect the geological model with other design modules to achieve collaborative work of multiple models.

[0328] 3.1 Integration tool selection: According to the characteristics of the BIM platform, the system selects appropriate integration tools for model integration. Different BIM platforms may provide different integration functions, such as Revit's Link IFC tool, Navisworks' DataTools, etc.

[0329] Technology used:

[0330] Link IFC tool for Revit: If you use Revit for integration, the system will use its Link IFC function to link the geological model into the project. Link IFC allows users to import the IFC model as a reference model and display and operate it in the BIM project.

[0331] Data Tools for Navisworks: If Navisworks is used, the system will use the Data Tools function to associate and query the data attributes of the geological model with other models, supporting multi-model integration operations.

[0332] Using Revit's Link IFC tool:

[0333] revit_app.link_ifc("bim_model.ifc")

[0334] 3.2 Attribute mapping and collaboration: The integrated tool also supports attribute mapping and collaboration between different modules. For example, the system can associate the soil bearing capacity attributes in the geological model with the basic design parameters in the structural model to ensure the coordination and consistency of the design.

[0335] Technology used:

[0336] Attribute mapping tool: The system uses the attribute mapping tool in the BIM platform to map the attributes of the geological model with the attributes of the building and structural models. For example, the stratigraphic information in the geological model is mapped to the basic support information in the building model to ensure the close connection between geological conditions and building design.

[0337] Multi-model collaboration: The system supports multi-model collaboration. Designers can view and operate geological models and architectural models simultaneously on the BIM platform, and achieve cross-disciplinary coordinated design through multi-view and multi-mode collaboration tools.

[0338] Property mapping: bim_platform.map_properties(geological_property.

[0339] S5. Cross-chain data sharing;

[0340] Cross-chain technology refers to a series of technologies and protocols that enable interoperability of data and assets between different blockchain networks. In this system, cross-chain technology is used to achieve geological data sharing and verification between different BIM projects, ensuring the reliability and consistency of data when transmitted between projects. Cross-chain communication protocols are the basis of cross-chain technology, through which blockchain networks can exchange data securely. Common cross-chain communication protocols include: Hash Time Lock Contract (HTLC): HTLC is a smart contract used to ensure the security and atomicity of cross-chain transactions. Through HTLC, the system can set a time lock, and within a specified time, only nodes that provide the correct hash value can complete data exchange, ensuring the security of data transmission. Relay: Relay refers to an intermediary node set between different blockchain networks, which is used to listen to events on one blockchain and propagate them to another blockchain. The relay mechanism enables different blockchain networks to share data in real time and achieve interoperability. Cross-Chain Verifier: A cross-chain verifier is a node or module used to verify the validity of cross-chain transactions. Through the verifier, the system can ensure the authenticity and integrity of cross-chain data and avoid data tampering and loss.

[0341] Cross-chain data sharing includes the following steps:

[0342] S51, data preparation and packaging;

[0343] Data preparation is the first step in cross-chain communication. The system needs to extract the geological data to be shared from the BIM project and perform necessary preprocessing.

[0344] Data extraction: The system first extracts the geological data to be shared from the BIM database. This data may include stratigraphic structure information, groundwater level monitoring results, soil composition analysis, etc. The extracted data is usually stored in a structured form, such as JSON or XML format, to facilitate subsequent processing and transmission.

[0345] Data Packaging: The extracted data will be packaged into a standardized data package. The data package includes the following key parts:

[0346] Data subject: The geological data itself to be shared.

[0347] Metadata: additional information describing the data, such as the time when the data was generated, the source of the data, the unit and dimension, etc.

[0348] Hash value: A data hash value generated using a hash algorithm such as SHA-256, which is used for subsequent integrity verification.

[0349] Digital signature: The data sender uses its private key to digitally sign the data packet to ensure that the source of the data is credible and cannot be tampered with.

[0350] S52, data encryption and verification;

[0351] In order to ensure the security and transparency of cross-chain data sharing, the system adopts a data encryption and verification mechanism. This mechanism uses smart contracts to encrypt and verify cross-chain shared data to ensure the data is tamper-proof and transparent.

[0352] 1. Data encryption: Data encryption refers to the use of cryptographic algorithms to convert plaintext data into ciphertext before data transmission to prevent unauthorized access and tampering. In cross-chain data sharing, the system uses the following encryption technologies:

[0353] Symmetric encryption algorithm: AES (Advanced Encryption Standard), used to encrypt geological data in transmission. The characteristic of symmetric encryption algorithm is that the same key is used for encryption and decryption, which can quickly process large-scale data and is suitable for efficient data transmission.

[0354] Asymmetric encryption algorithm: RSA, used to encrypt symmetric keys and digital signatures. The characteristic of asymmetric encryption algorithm is that different keys (public key and private key) are used for encryption and decryption, which is suitable for scenarios with higher security requirements. The system usually uses asymmetric encryption to protect symmetric encryption keys and ensure that only authorized nodes can decrypt data.

[0355] Digital signature: By using an asymmetric encryption algorithm to generate a digital signature, the system can verify the identity of the data sender and ensure that the data has not been tampered with during transmission. The formula for generating a digital signature is:

[0356]

[0357] in, It is a summary generated by a hash function on the data. The sender's private key is used to encrypt the hash value. The receiver uses the sender's public key to verify the signature.

[0358] 2. Data verification: Data verification means that after receiving cross-chain data, the system verifies the integrity and authenticity of the data through smart contracts. Smart contracts are automated programs running on the blockchain that can execute pre-defined rules and logic to ensure the security of data transmission.

[0359] Smart contract verification: In cross-chain data sharing, smart contracts verify data through the following steps:

[0360] Hash value verification: The smart contract first calculates the hash value of the received data and compares it with the hash value provided by the sender. If the two hash values ​​are consistent, it means that the data has not been tampered with during transmission.

[0361] Digital signature verification: The smart contract uses the sender’s public key to verify the digital signature to ensure that the sender of the data is legitimate and that the integrity of the data has not been compromised.

[0362] Timestamp verification: The smart contract verifies the timestamp in the data packet to ensure that the data is transmitted within the specified time window to prevent replay attacks.

[0363] Transparency guarantee: All data encryption and verification operations are recorded on the blockchain, forming an audit log that cannot be tampered with. Through the openness and transparency of the blockchain, all participants can verify the data transmission process and ensure the transparency of cross-project data collaboration.

[0364] S53, cross-chain data transmission;

[0365] After data encryption and signature are completed, the system starts cross-chain data transmission. During the cross-chain data transmission process, the system uses a series of secure technical means to ensure that geological data can be safely and efficiently transmitted between blockchain networks of different BIM projects.

[0366] 1. Selection of relay nodes

[0367] Relay nodes are the core components of cross-chain communication. They are responsible for transmitting data between different blockchain networks. Relay nodes capture and transmit relevant data to another blockchain network by monitoring events on one blockchain.

[0368] Node configuration: Relay nodes are usually specially configured by network administrators and can access multiple blockchain networks. These nodes have sufficient computing and storage capabilities to support real-time data transmission and processing. The selection of relay nodes needs to consider the network topology, node performance, and the optimization of data transmission paths.

[0369] Monitoring and Capture: The relay node continuously monitors the occurrence of specific events on the source blockchain, such as the generation of data packets, confirmation of transactions, etc. Once these events are detected, the relay node immediately captures the relevant data and prepares to transmit it to the target blockchain network.

[0370] Cross-chain event triggering: When a relay node captures a cross-chain transmission event, it triggers a corresponding event on the target blockchain network, such as a cross-chain transaction request or a data reception request. This process usually involves the standardization of cross-chain communication protocols to ensure the consistency of data formats and communication processes.

[0371] 2. Data broadcasting

[0372] Data broadcasting refers to the process in which a relay node transmits encrypted data packets from a source blockchain network to a target blockchain network.

[0373] Packet splitting: Before data is broadcast, relay nodes usually split the data packet into smaller chunks. This splitting can improve transmission efficiency, especially when network bandwidth is limited or needs to be transmitted through multiple relay nodes. Each data block is marked as part of a complete data packet to ensure that it can be correctly reassembled in the target blockchain network.

[0374] Segmented transmission: Data blocks are gradually transmitted to the target blockchain network. When the relay node transmits each data block, it will record the transmission status so that it can be retransmitted in case of network failure or data loss. This mechanism ensures the reliability of data transmission.

[0375] Cross-chain transaction request: In the target blockchain network, the relay node will create a cross-chain transaction request. This request contains the identification information and verification data (such as hash value) of all data blocks, waiting for the recipient's confirmation. The node of the receiving blockchain network will verify the data according to the information in the request and receive the data block after confirmation.

[0376] 3. Hashed Time Lock Contract (HTLC)

[0377] Hash Time Lock Contract (HTLC) is a key mechanism to ensure the security and atomicity of cross-chain data transmission. HTLC guarantees the integrity of data transmission through smart contract technology and avoids data loss or transaction failure caused by network delays or node failures.

[0378] 1. Set a time lock: The HTLC contract sets a time lock on the target blockchain network. The length of the time lock is usually determined based on network latency and packet size, ensuring that the receiver has enough time to verify the hash value and receive the data. Within the set time, the receiver must provide the correct hash value to unlock the data packet.

[0379] Time lock mechanism: The time lock mechanism ensures the timeliness of cross-chain transactions. If the recipient fails to complete the hash verification within the specified time, the transaction is automatically canceled and the data packet is returned to the sender to prevent data loss or malicious interception.

[0380] 2. Verify the hash value: In the receiving blockchain network, after the recipient provides a hash value that matches the data packet, the HTLC contract verifies the validity of the hash value. After the verification is passed, the HTLC contract unlocks the data packet and delivers it to the recipient.

[0381] Hash value generation and verification: The hash value is a unique identifier generated by the content of the data packet using a hash function (SHA-256). The receiver needs to provide a hash value that is consistent with the sender in order to successfully unlock the data packet. This process ensures that the data has not been tampered with during transmission.

[0382] 3. Atomicity guarantee: HTLC contract guarantees the atomicity of cross-chain transactions, that is, transactions are either completed in full or canceled completely. If the recipient fails to provide the correct hash value within the specified time, the HTLC contract will automatically cancel the transaction and the data packet will be returned to the sender, ensuring the security and reliability of data transmission.

[0383] Transaction rollback: If the HTLC contract cancels a transaction, all operations related to the transaction will be rolled back and restored to the state before the transaction. This mechanism avoids data loss or inconsistency and ensures the integrity and security of cross-chain data transmission.

[0384] S54. Data decryption and verification.

[0385] When the receiving blockchain network receives the data packet, the system will decrypt and verify the data to ensure its integrity and authenticity.

[0386] Data Decryption:

[0387] Decrypting the symmetric key: The recipient uses his own private key to decrypt the encrypted symmetric key.

[0388] Decrypt data packet: Use the decrypted symmetric key to decrypt the data packet and restore the original geological data.

[0389] Data Verification: After decryption, the system verifies the data through the following steps:

[0390] Hash value verification: The system recalculates the hash value of the decrypted data packet and compares it with the hash value provided by the sender to ensure that the data has not been tampered with during transmission.

[0391] Digital signature verification: The system uses the sender's public key to verify the digital signature and confirm that the data does come from the legitimate sender.

[0392] Timestamp verification: The system checks the timestamp in the data packet to ensure that the data is transmitted within a valid time window to prevent replay attacks.

[0393] In summary, the innovative features of the present invention are as follows:

[0394] 1. Data-driven model generation and optimization

[0395] The present invention introduces the XGBoost algorithm to automatically generate geological model solutions and optimize parameters in real time based on historical data and real-time data. Compared with the existing modeling methods that rely on manual adjustment, the present invention can generate models more quickly and accurately, significantly improving the reliability and applicability of geological models.

[0396] 2. Real-time feedback and adaptive adjustment

[0397] Through real-time data collection of cognitive IoT devices and dynamic parameter adjustment of the PSO algorithm, the present invention can continuously monitor changes in geological conditions and adjust model parameters in real time during the modeling process. This real-time feedback mechanism ensures the adaptability of the model throughout the project life cycle and reduces the project risks caused by changes in geological conditions.

[0398] 3. Cross-chain data sharing and collaboration

[0399] The present invention combines cross-chain technology and smart contracts to achieve secure sharing and collaborative operation of geological data between different BIM projects. Compared with the traditional single-project data management model, the present invention provides a cross-project data integration and verification mechanism to ensure data consistency and integrity, and enhance the collaboration capabilities between engineering projects.

[0400] Embodiment 2

[0401] Combination Figure 1 to Figure 2 As shown, the second embodiment provides a cognitive geological modeling system based on the BIM system on the basis of the first embodiment. The second embodiment includes the technical content disclosed in the first embodiment. The technical content of the second embodiment that is the same as that of the first embodiment will not be repeated. The difference between the second embodiment and the first embodiment is described below. A cognitive geological modeling system based on the BIM system includes: a data collection and preprocessing module, a pattern recognition and feature extraction module; a modeling scheme and parameter optimization module, a model automatic generation module, and a cross-chain data sharing module; the data collection and preprocessing module is used to realize the collection of historical data, the collection of real-time data, data filtering and cleaning, data integration and standardization; the pattern recognition and feature extraction module is used to realize pattern recognition and feature extraction; the modeling scheme and parameter optimization module is used to realize data-driven model generation, parameter adjustment and optimization, and real-time feedback mechanism; the model automatic generation module is used to realize model automatic generation and model integration into the BIM system; the cross-chain data sharing module is used to realize data preparation and packaging, data encryption and verification, cross-chain data transmission, and data decryption and verification. Each module performs specific tasks to ensure that the entire process from data collection, processing to model generation, optimization and deployment is efficient and accurate. The architecture integrates a variety of advanced technologies, including XGBoost algorithm, PSO optimization algorithm, cognitive Internet of Things, GOCAD modeling tools, etc., to support the automation and intelligence of geological modeling.

[0402] Embodiment 3

[0403] This embodiment three provides a specific application example based on the embodiment one. Taking the application of geological modeling in a highway engineering project as an example, in a highway construction project, it is necessary to conduct detailed exploration and modeling of the geological conditions along the route. The project is located in a mountainous area with complex geological conditions, including multi-layer geological structures, large fluctuations in groundwater levels, and potential landslide risks. In order to ensure the safety of the project and the rationality of the design, it is decided to adopt the cognitive geological modeling method and system based on the BIM system provided by the present invention to conduct a comprehensive analysis and modeling of the geological conditions along the project.

[0404] 1. Data collection and preprocessing:

[0405] Collection of historical data:

[0406] The geological exploration data of the area over the past ten years were extracted from the project database based on MySQL. The data include:

[0407] Stratigraphic structure data (e.g. thickness and distribution of sand, clay, and rock layers).

[0408] Environmental data (e.g. annual rainfall, groundwater level changes).

[0409] Previous geological model results and their subsequent engineering validation data.

[0410] Get this data from the MySQL database through SQL query:

[0411] SELECT gp.layer_thickness, gp.soil_type, ed.annual_rainfall,ed.groundwater_level, mr.model_accuracy

[0412] FROM Geological_Parameters gp

[0413] JOIN Environmental_Data ed ON gp.location_id = ed.location_id

[0414] JOIN Model_Results mr ON gp.model_id = mr.model_id

[0415] WHERE gp.location_id = 'mountain_section'

[0416] AND mr.model_date BETWEEN '2010-01-01' AND '2020-12-31';

[0417] Real-time data collection:

[0418] Cognitive Internet of Things (CIoT) devices were deployed at the project site, including soil moisture sensors, groundwater level monitors, and formation pressure sensors. The sensors collect real-time geological parameter data at a sampling frequency of once an hour.

[0419] These data are encrypted via the Transport Layer Security (TLS) protocol, transmitted in real time to the data warehouse in the BIM system, and processed in real time using the Kafka framework.

[0420] Data preprocessing:

[0421] The system removes noise, cleans and standardizes the collected historical data and real-time data, including using the triple standard deviation method to clean up outliers in historical data and using Gaussian filtering to filter noise in real-time data.

[0422] All data are integrated and standardized into a unified format and stored in the project's data warehouse for further analysis and modeling.

[0423] 2. Pattern recognition and feature extraction:

[0424] Pattern Recognition:

[0425] The system used the random forest algorithm to perform pattern recognition on historical data and found that the thickness of the strata and the change of groundwater levels were significantly correlated with the prediction accuracy of the geological model. These key features were extracted as core parameters for subsequent modeling.

[0426] Feature extraction:

[0427] Using the convolutional neural network (CNN) in deep learning, the system extracts high-level features from the real-time geological profile data, such as the changing trend of the stratum structure and the dynamic changes of soil moisture. These features provide an important basis for the dynamic adjustment of the model.

[0428] 3. Modeling scheme and parameter optimization:

[0429] Data-driven model generation:

[0430] The system inputs the extracted features into the XGBoost algorithm to automatically generate a preliminary geological modeling solution. The generated model includes a three-dimensional representation of the multi-layer geological structure, a predictive model for groundwater level changes, and a prediction of possible landslide areas.

[0431] Parameter adjustment and optimization:

[0432] In order to optimize the accuracy of the model, the system used the particle swarm optimization algorithm (PSO) to iteratively adjust the model parameters generated initially. By comparing the model accuracy under different parameter combinations, the optimal modeling parameters were finally determined to ensure that the model can accurately reflect the complex geological conditions on site.

[0433] 4. Real-time feedback mechanism:

[0434] Real-time data monitoring and dynamic parameter adjustment:

[0435] During the construction process, the system continuously monitors real-time data. When it detects an abnormal rise in groundwater levels, the system automatically triggers the PSO algorithm to readjust model parameters and update the prediction model for landslide risk areas. The construction team receives updated data from the system in real time and adjusts the construction plan to avoid risks.

[0436] 5. Model generation and integration:

[0437] Automatic generation of GOCAD models:

[0438] The system uses the GOCAD software through the automatically generated GSLib script to construct a 3D geological model along the highway. The model describes the stratigraphic structure, groundwater level, and possible landslide areas in detail.

[0439] Model integration into BIM system:

[0440] The generated geological model was converted into IFC format and imported into the BIM system. The project design team used Revit's Link IFC tool to connect the geological model with the architectural and structural design models, achieving seamless integration of the design scheme and geological conditions.

[0441] 6. Cross-chain data sharing:

[0442] Cross-project data collaboration:

[0443] Since the highway project overlaps with another nearby tunnel project, the project team shared geological data to the BIM system of the tunnel project through cross-chain technology. The HTLC smart contract ensures the security and reliability of data sharing, and enables real-time data collaboration between the two projects, guiding decision-making during the construction process.

[0444] The implementation of the present invention ensures that the project team can accurately and in real time grasp the complex geological conditions, optimize the highway design and construction plan, significantly reduce the engineering risk, and ensure the smooth progress of the construction. The cross-project data sharing mechanism also improves the efficiency of collaboration between different projects, fully demonstrating the practical application value of the present invention.

[0445] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cognitive geological modeling method based on BIM system, characterized in that: include: S1. Data collection and preprocessing, including the following steps: S11. Collection of historical data; S12, real-time data collection; S13, data filtering and cleaning; S14. Data integration and standardization; S2, pattern recognition and feature extraction, including the following steps: S21, pattern recognition; S22, feature extraction; S3, modeling scheme and parameter optimization, including the following steps: S31, data-driven model generation; S32, parameter adjustment and optimization; S33, real-time feedback mechanism; S4, automatic model generation, including the following steps: S41, automatic model generation; S42, model integration into BIM system; S5. Cross-chain data sharing includes the following steps: S51, data preparation and packaging; S52, data encryption and verification; S53, cross-chain data transmission; S54. Data decryption and verification.

2. The cognitive geological modeling method based on the BIM system according to claim 1 is characterized in that: S11. Collection of historical data, including: Data query and extraction: The system obtains a large amount of historical modeling data from the BIM database, and extracts data related to geological modeling from the relational database MySQL. Through standard SQL query statements, the system can automatically retrieve and extract data that is highly relevant to the current modeling task; Data extraction and storage: The queried data will be stored in a temporary data storage area and prepared for further processing. The system ensures the efficiency of data extraction and the integrity of the data set through a unified interface; S12. Real-time data collection, including: Real-time data collection: The system collects geological data in real time through cognitive IoT devices. The cognitive IoT devices are equipped with high-precision sensors that can monitor key geological parameters including soil moisture, formation pressure, and groundwater level. The monitored data is transmitted to the BIM system through a wireless sensor network; Data transmission and processing: During the data transmission process, the system uses the transport layer security protocol for encryption to ensure the integrity and confidentiality of the data. After the data arrives, the system uses the real-time stream processing framework Kafka to process the data in real time, including data denoising and data compression; S13. Data filtering and cleaning, including: Data cleaning technology: The system uses the triple standard deviation method to identify and process outliers. For missing values, the system uses a linear interpolation algorithm to fill in the missing values ​​through the weighted average of adjacent data points to ensure the continuity and integrity of the data; Data denoising: Before entering the analysis phase, the system denoises the data using a Gaussian filter algorithm; S14. Data integration and standardization, including: Data integration: unified processing and management of geological data from different sources and time periods to form a comprehensive data set, including data alignment and integration, data conversion and integration. In the data alignment and integration, historical data and real-time data are aligned and integrated through timestamp alignment and spatial coordinate matching technology; in the data conversion and integration, data from different data sources are extracted, converted and integrated through the ETL process; Data standardization: The process of unifying data from different sources in terms of dimensions and units to ensure that the data is comparable and consistent during analysis. The standardization process includes two aspects: unit conversion and data normalization.

3. The cognitive geological modeling method based on the BIM system according to claim 1 is characterized in that: S21. Pattern recognition, specifically including: Data preparation and input: The system first inputs the preprocessed historical data into the random forest model. The data includes geological parameters and corresponding model accuracy indicators. During the input process, the system groups the data to ensure that each input sample contains a complete set of geological parameters. Model training and pattern recognition: During the training process, the random forest algorithm constructs multiple decision trees through multiple random samplings. Each decision tree is split according to the input geological parameters, and finally forms a set of leaf nodes. Each leaf node represents the model accuracy under a geological condition. S22, feature extraction, specifically including: Convolution operation and feature mapping: The system inputs the geological data collected in real time by cognitive IoT devices into the convolutional neural network. In the convolutional neural network, the convolution layer performs convolution operation on the input data through the convolution kernel to extract the feature map. Pooling and dimensionality reduction: After the convolution operation, the system uses a pooling layer to reduce the dimensionality of the feature map, retaining important features and reducing the data dimension; Fully connected layer and feature extraction: After completing the convolution and pooling operations, the system integrates the feature maps through the fully connected layer. Through the fully connected layer, the system can generate the final feature vector, which represents the high-level features of the input data; Feature output and model optimization: The extracted feature vectors will be used in the subsequent model optimization process to optimize the parameter settings and structural design of geological modeling by analyzing the features.

4. The cognitive geological modeling method based on the BIM system according to claim 1, characterized in that: S31. Data-driven model generation, including: Model Generation: XGBoost gradually optimizes the model’s predictive power throughout the training process by gradually building multiple decision trees, each of which attempts to correct the errors of the previous tree. In each iteration, XGBoost minimizes the loss function by calculating the residual and fitting a new tree. Feature input: The system first inputs the feature vector generated from the pattern recognition and feature extraction steps into the XGBoost algorithm; Modeling process: During the training process, XGBoost generates geological modeling solutions through the following steps: Among them, initialization model: starting from the initial model structure and parameters, XGBoost initializes a simple model; among them, iterative training: in each iteration, XGBoost uses the residual between the predicted result and the actual result of the current model as the target value to fit a new tree, and the construction of each new tree aims to reduce the overall error; among them, update model: add the newly fitted tree to the existing model structure, update the overall model, and repeat this process multiple times until the predetermined stop condition is reached; among them, model output: the final generated model structure includes several decision trees, as well as the parameter settings and calculation strategies related to them; Model structure and calculation strategy: The model output by XGBoost contains not only the structure of the tree, but also the split points associated with each tree, the values ​​of the leaf nodes, and the combined weights of the trees; S32, parameter adjustment and optimization. After initially generating the geological modeling scheme, the system further optimizes the model parameters through the particle swarm optimization algorithm, including: Initialize particle swarm: initialize a group of particles in the parameter space, each particle corresponds to a set of initial model parameters; Evaluate fitness: Each particle generates a model based on the current parameter combination and uses historical data and real-time data to evaluate the model's prediction error. The smaller the model error, the higher the fitness. Update particle position: Update the particle position according to the PSO algorithm formula, gradually approaching the area with the minimum error; Iterative optimization: Repeat fitness evaluation and position update until a globally optimal parameter combination is found or a preset stop condition is reached; S33. Real-time feedback mechanism, specifically including: Real-time data monitoring: Real-time data monitoring is the basis of the real-time feedback mechanism. The system relies on cognitive IoT devices to continuously monitor and collect data on the geological environment; Dynamic adjustment of model parameters: After identifying changes in geological conditions, the system automatically starts the dynamic adjustment mechanism of parameters to ensure that the model can adapt to these changes in a timely manner.

5. The cognitive geological modeling method based on the BIM system according to claim 1, characterized in that: S41. Automatic model generation: The system converts the optimized geological modeling scheme into a complete three-dimensional geological model through an automated process and implements it using GOCAD modeling software, including: Automatic conversion of modeling schemes: The system first converts the geological modeling scheme optimized in the early stage, including model structure, parameter settings, and calculation strategies, into an input format recognizable by GOCAD software; Model verification and correction: After GOCAD generates the geological model, the system automatically performs preliminary model verification. If there are inconsistencies or errors in the model, the system automatically triggers the correction mechanism, readjusts the input parameters or data, and regenerates the model; Execution and monitoring of automated processes: The entire modeling process is managed by the system's automated scheduling module, which is responsible for scheduling modeling tasks, allocating computing resources, and monitoring modeling progress. The system will generate detailed log records, covering the execution of each step, the parameters used, and the final generated model information. The generated logs are used for subsequent model review and traceability; S42. Model integration into the BIM system, including: Geological model import: convert the 3D geological model generated by GOCAD into a format that can be processed by the BIM system, and introduce it into the BIM environment to ensure the integrity and accuracy of the model; Data exchange and integration: seamlessly integrate the imported geological model with other design modules in the BIM system to ensure the coordination of the overall design; Use of integrated tools: Various integrated tools are used on the BIM platform to connect the geological model with other design modules to achieve collaborative work of multiple models.

6. The cognitive geological modeling method based on the BIM system according to claim 1, characterized in that: S51. Data preparation and packaging, including: Data extraction: The system first extracts the geological data to be shared from the BIM database; Data packaging: The extracted data will be packaged into a standardized data package; S52. Data encryption and verification, including: Data encryption: Before data transmission, cryptographic algorithms are used to convert plaintext data into ciphertext; Data verification: After receiving cross-chain data, the system verifies the integrity and authenticity of the data through smart contracts; S53. Cross-chain data transmission, including: Selection of relay nodes: Responsible for transmitting data between different blockchain networks. Relay nodes capture and transmit relevant data to another blockchain network by monitoring events on one blockchain. Data broadcast: The process by which a relay node transmits encrypted data packets from a source blockchain network to a target blockchain network; Hash Time Lock Contract: A key mechanism to ensure the security and atomicity of cross-chain data transmission. HTLC uses smart contract technology to ensure the integrity of data transmission and avoid data loss or transaction failure due to network delays or node failures. S54, data decryption and verification, specifically including: when the receiving blockchain network receives the data packet, the system will decrypt and verify the data to ensure the integrity and authenticity of the data.

7. A cognitive geological modeling system based on a BIM system, characterized in that: The cognitive geological modeling method based on the BIM system described in claim 1 includes: a data collection and preprocessing module, a pattern recognition and feature extraction module; a modeling scheme and parameter optimization module, a model automatic generation module, and a cross-chain data sharing module; the data collection and preprocessing module is used to realize the collection of historical data, the collection of real-time data, data filtering and cleaning, data integration and standardization; the pattern recognition and feature extraction module is used to realize pattern recognition and feature extraction; the modeling scheme and parameter optimization module is used to realize data-driven model generation, parameter adjustment and optimization, and real-time feedback mechanism; the model automatic generation module is used to realize model automatic generation and model integration into the BIM system; the cross-chain data sharing module is used to realize data preparation and packaging, data encryption and verification, cross-chain data transmission, and data decryption and verification.

Images