Foundation pit blasting intelligent vibration reduction method and system based on digital twinning, and medium

Abstract

The application discloses a kind of based on digital twinning foundation pit blasting intelligent damping method, system and medium, the method is by constructing an integrated " perception-prediction-generation-regulation " intelligent closed loop, built-in online correction neural network intelligent fluctuation theory model, realizes the dynamic high-precision deduction of wave propagation;Proposed fusion strain energy and kinetic energy dissipation topology optimization objective function, realizes the automatic generation of damping structure from macro layout to micro composite material distribution;Adopt the edge reinforcement learning based on attention mechanism, realize millisecond level self-adaptive precision control.System, through the cooperation of digital twinning engine, intelligent design brain, edge control node and block chain storage platform, realize the digitization, intelligentization and credibility of whole process.The application significantly improves the scientificity, precision and adaptive ability of blasting vibration control.

Description

Technical Field

[0001] This invention relates to the field of blasting vibration control technology in geotechnical engineering, specifically to an intelligent vibration reduction method, system, and medium for foundation pit blasting based on digital twins. Background Technology

[0002] As urban underground space development progresses towards greater depth, scale, and density, vibration control during foundation pit blasting excavation has become a core challenge concerning engineering safety, social stability, and economic benefits, especially when facing sensitive targets such as adjacent existing tunnels, historical buildings, and precision instruments. Currently, installing vibration damping hole arrays is a commonly used passive control measure to control the propagation of blasting vibrations. Its classic mechanism lies in utilizing the reflection, scattering, and diffraction effects of stress waves at the hole interface to dissipate wave energy, thereby achieving the purpose of attenuating vibration.

[0003] However, after in-depth analysis of existing technologies and academic research, it has been found that the current vibration damping hole layout technology system has the following four fundamental limitations, which can no longer meet the high-level requirements of modern complex engineering for precision, intelligence and greenness:

[0004] 1. Limitations of Idealized Theoretical Models: Existing design theories are generally based on the assumption of a homogeneous, continuous, and isotropic ideal medium, and only consider a single elastic dynamic process. This makes it impossible to accurately describe the complex propagation behavior of stress waves in non-homogeneous rock masses with joints and fissures, as well as under the coupling of multiple physical fields such as groundwater seepage and explosive thermal effects. The low fidelity of theoretical models is the root cause of systematic deviations in vibration reduction effect predictions and the tendency for conservative or overly aggressive designs.

[0005] 2. Static and Empirical Design Methods: Current methods heavily rely on individual engineers' experience and simplified empirical formulas, essentially constituting a form of "static design." They lack a systematic quantitative understanding and efficient exploration methods for the high-dimensional, nonlinear coupling relationships between geological parameters, blasting parameters, and vibration damping hole parameters. Parameter selection is highly arbitrary, failing to achieve "precise prescriptions" for specific sites, and even less enabling "dynamic optimization" based on real-time feedback during construction.

[0006] 3. Structural uniformity and rigidity: The vibration damping holes commonly used in the industry are regularly arranged circular holes, which have a single structural function and have reached a bottleneck in adaptability and dissipation efficiency for complex spectrum stress waves. Existing technologies lack exploration and application of innovative structural forms such as irregular hole array topologies and functionally graded material filling, and have failed to tap the potential of wave energy dissipation from the source of structural innovation.

[0007] 4. Isolation and Delay in the Control System: The traditional "design-construction-monitoring" process is linear and open-loop. The design scheme is disconnected from the dynamically changing construction process, and on-site monitoring data is only used for post-event verification, unable to provide real-time and reliable feedback to the decision-making end to trigger autonomous optimization of the scheme. This isolated system, lacking an intelligent closed loop of "perception-decision-execution," makes the entire control process lagging and rigid, unable to cope with the many uncertainties on the construction site.

[0008] In summary, existing technologies are confined by idealized theories, static designs, simplistic structures, and isolated controls, urgently requiring a paradigm shift from underlying theory to top-level architecture. This invention, based on a profound understanding of these deep-seated contradictions, aims to provide a completely new systemic solution. Summary of the Invention

[0009] The present invention proposes an intelligent vibration reduction method and system for foundation pit blasting based on digital twins, which can at least solve one of the technical problems in the background art.

[0010] To achieve the above objectives, the present invention adopts the following technical solution:

[0011] A digital twin-based intelligent vibration reduction method for foundation pit blasting includes the following steps:

[0012] S100, constructing a digital twin of the foundation pit blasting site;

[0013] S200, based on a digital twin, runs a two-stage optimization strategy;

[0014] S300 deploys a lightweight reinforcement learning agent with optimized strategies on edge computing nodes to perform online adaptive control of blasting parameters and adjustable vibration damping hole devices based on real-time monitoring data.

[0015] S400, based on blockchain technology, constructs a complete and reliable data chain from the sensor data and decision commands of the adaptive control adjustable vibration damping hole device to the effect evaluation, ensuring the auditability and traceability of the optimization process.

[0016] Furthermore, the digital twin construction method in step S100 of the present invention includes:

[0017] A smart wave theory model is constructed by integrating Geographic Information System (GIS), Building Information Modeling (BIM), and Internet of Things (IoT) sensor data. This model adopts a hybrid architecture that combines classical physics theory with data-driven correction.

[0018]

[0019] For the physical prediction term based on the angular spectrum diffraction theory, x is the change in spatial position and ω is the angular frequency; To trainable weights The parameterized neural network correction term takes historical wavefields as its input. Rock mass parameters Temperature gradient and seepage velocity .

[0020] Furthermore, the two-stage optimization strategy structure in step S200 of the present invention includes:

[0021] The first stage uses a multi-objective evolutionary algorithm to optimize the macroscopic layout of the vibration reduction hole group, and the second stage uses a topology optimization method to automatically generate a microscopic vibration reduction structure with irregular shape and composite filling material.

[0022] Among them, macro layout optimization: taking the number of rows, columns and spacing of vibration damping holes as macro layout parameters as design variables, the NSGA-III multi-objective genetic algorithm is used to conduct a global Pareto front search with the goals of vibration isolation rate, engineering cost and environmental impact, and to determine several high-performance macro layout schemes.

[0023] Microstructure topology optimization: Topology optimization is introduced to integrate strain energy and kinetic energy dissipation, simultaneously optimizing the stiffness and damping characteristics of the structure. The specific function is as follows:

[0024]

[0025]

[0026]

[0027] in, Due to volume constraints, This is the strain energy term; maximizing this term makes the structure more deformable by reflecting and scattering wave energy. This is the kinetic energy dissipation term. Maximizing this term allows the structure to more effectively convert wave energy into heat energy. For material density, Here, v is the damping coefficient related to the pseudo-density, and v is the vibration velocity field. This is a weighting factor.

[0028] Furthermore, the reinforcement learning agent in step S300 of the present invention includes:

[0029] The state value function incorporating the attention mechanism is as follows:

[0030]

[0031]

[0032] Among them, S t Let θ represent the system state at time t, and let θ be the set of learnable parameters in the state value function. These are state feature vectors from different sensors. The query vector is a learnable vector representing the current control objective. For attention weights, It is a trainable linear transformation matrix.

[0033] Furthermore, the method for constructing the trusted data link in step S300 of the present invention includes:

[0034] The real-time sensing data of the adjustable vibration damping hole device, the model prediction data of the digital twin, the optimized vibration damping structure parameters, the decision instructions generated by the reinforcement learning agent, and the evaluation results of the blasting vibration effect are used as key operational data.

[0035] Let the first The key operation data set for the batch is:

[0036]

[0037] in Represents the i-th type of data;

[0038] Perform a hash operation on each data item:

[0039]

[0040] in, For cryptographic hash functions;

[0041] Merkle trees are constructed for the key operation data in time batches, and their root hash values ​​are calculated;

[0042] Hash values ​​from the same batch As leaf nodes, construct a Merkle tree;

[0043] If the number of leaf nodes is odd, then copy the last leaf node to form an even number of nodes;

[0044] The hash value of a non-leaf node is calculated by concatenating the hash values ​​of its child nodes:

[0045]

[0046] The Merkle root hash of the batch of data is obtained by recursively calculating to the root node. ;

[0047] root hash With timestamp Package other metadata into transaction data:

[0048]

[0049] After signing the transaction, it is submitted to the blockchain network:

[0050]

[0051] After a transaction is verified through consensus, it is written into a block, forming an immutable data chain.

[0052] Based on an immutable data chain, it supports data auditing, effect traceability, and responsibility determination for the entire vibration reduction control process.

[0053] On the other hand, a digital twin-based intelligent vibration reduction system for foundation pit blasting includes: a digital twin engine module, an intelligent design brain module, an edge control node module, and a blockchain evidence storage platform module.

[0054] Each functional module communicates through an API gateway and is deployed on a hybrid cloud platform; among them, the intelligent design brain module integrates a macro optimizer and a micro topology optimizer; the edge control node module has a built-in reinforcement learning agent.

[0055] As can be seen from the above technical solutions, in summary, the method of the present invention integrates physical priors and data-driven approaches through an intelligent wave theory model, endowing the model with online self-calibration and continuous evolution capabilities, fundamentally solving the problem of high distortion of traditional models under complex geological conditions, and improving prediction accuracy by more than 30%.

[0056] By integrating strain energy and kinetic energy dissipation into topology optimization, a qualitative leap has been achieved in vibration damping hole design, moving from manual experience-based drawing to AI-generated optimal composite structures. The irregular rigid-damped composite structure discovered by this method achieves a 25%-40% improvement in vibration damping efficiency compared to traditional regular voids.

[0057] By employing reinforcement learning based on attention mechanisms, a cloud-edge collaborative architecture of "digital twin-edge intelligence" was constructed, establishing a self-regulating closed loop with millisecond-level response. The decision-making efficiency of the intelligent agent was improved by more than 40%, effectively responding to unexpected situations.

[0058] By introducing blockchain technology, a trusted data chain was established throughout the entire process, providing tamper-proof evidence to support engineering quality management and safety responsibility definition, thereby improving the transparency and trust level of the entire industry. Attached Figure Description

[0059] Figure 1 This is a flowchart of the intelligent vibration reduction method for foundation pit blasting based on digital twins according to the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0061] like Figure 1 As shown in this embodiment, the intelligent vibration reduction method for foundation pit blasting based on digital twins includes:

[0062] S100, constructing a digital twin of the foundation pit blasting site;

[0063] S200, based on a digital twin, runs a two-stage optimization strategy;

[0064] S300 deploys a lightweight reinforcement learning agent with optimized strategies on edge computing nodes to perform online adaptive control of blasting parameters and adjustable vibration damping hole devices based on real-time monitoring data.

[0065] S400, based on blockchain technology, constructs a complete and reliable data chain from the sensor data and decision commands of the adaptive control adjustable vibration damping hole device to the effect evaluation, ensuring the auditability and traceability of the optimization process.

[0066] The following provides a detailed explanation of each step:

[0067] S100, constructing a digital twin of the foundation pit blasting site;

[0068] The digital twin integrates geographic information system (GIS), building information model (BIM) and Internet of Things (IoT) sensor data to construct a digital twin that is completely mapped to the physical foundation pit. The core of this digital twin is the intelligent wave theory model.

[0069] This intelligent wave theory model breaks through traditional idealized assumptions and adopts a hybrid architecture that combines classical physics theory with data-driven correction:

[0070]

[0071] In the formula, It is a physical prediction term based on the theory of angular spectrum diffraction, which ensures the theoretical basis of the model. x is the change in spatial position and ω is the angular frequency. It is a neural network that can be trained online, based on Tron at a previous time step. Rock mass parameters Temperature gradient and seepage velocity The input is used to dynamically output model correction terms.

[0072] This model can learn online and compensate for prediction errors caused by medium inhomogeneity, nonlinearity and multi-physics coupling, realizing a leap from "static theory" to "dynamic self-evolution" and significantly improving prediction accuracy under complex geological conditions.

[0073] S200. Based on the digital twin, run a two-stage optimization strategy;

[0074] The optimization strategy employs a hierarchical design approach, coordinating optimization at both macro and micro scales.

[0075] The first stage uses a multi-objective evolutionary algorithm to optimize the macroscopic layout of the vibration damping hole group; the second stage uses a topology optimization method to automatically generate microscopic vibration damping structures with irregular shapes and composite filling materials for key areas.

[0076] Phase 1 (Macro Layout Optimization): Using macro layout parameters such as the number of rows, columns, and spacing of vibration damping holes as design variables, the NSGA-III multi-objective genetic algorithm is used to conduct a global Pareto front search with the goals of vibration isolation rate, engineering cost, and environmental impact (such as carbon emissions) to determine several high-performance macro layout schemes.

[0077] Phase Two (Microstructure Topology Optimization): Topology optimization techniques are introduced for the key regions selected in Phase One;

[0078] Topology optimization that integrates strain energy and kinetic energy dissipation

[0079] A novel objective function is proposed to simultaneously optimize the stiffness and damping characteristics of a structure:

[0080]

[0081]

[0082]

[0083] in, Due to volume constraints, This is the strain energy term. Maximizing this term makes the structure more deformable by reflecting and scattering wave energy. This is the kinetic energy dissipation term; maximizing this term allows the structure to more effectively convert wave energy into heat energy. For material density, Here, v is the damping coefficient related to the pseudo-density, and v is the vibration velocity field. This is a balancing factor used to weigh the contributions of stiffness and damping.

[0084] This function can automatically generate irregular multiphase composite microstructures containing cavities, high-damping polymers, and metal skeletons, achieving an integrated optimal design of "macro array-micro structure" with wave energy dissipation efficiency far exceeding that of traditional regular cavities.

[0085] S300 deploys a lightweight reinforcement learning agent with optimized strategies on edge computing nodes to perform online adaptive control of blasting parameters and adjustable vibration damping hole devices based on real-time monitoring data.

[0086] The optimized solution is deployed to on-site edge computing nodes. The reinforcement learning (RL) agent running within the node autonomously makes decisions and adjusts the adjustable vibration damping orifice device based on real-time vibration monitoring data.

[0087] To address the decision-making efficiency issue under multi-sensor data, an attention mechanism is introduced into the agent's state-value function:

[0088]

[0089] Among them, S t The system state at time t is the core variable describing the environment and state of the agent / system at a certain point in time.

[0090] θ is the set of learnable parameters in the state value function;

[0091] These are state characteristics from different sensors.

[0092] This is a learnable query vector that represents the current core control objective.

[0093] It automatically learns and assigns higher importance to key sensor data by assigning attention weights.

[0094] It is a trainable linear transformation matrix that enables the agent to dynamically and adaptively weigh the importance of different sensor data.

[0095] This enables intelligent agents to filter and focus information, significantly improving decision-making speed and control accuracy in complex and noisy environments, and achieving millisecond-level adaptive closed-loop control.

[0096] S400, based on blockchain technology, constructs a complete and reliable data chain from the sensor data and decision commands of the adaptive control adjustable vibration damping hole device to the effect evaluation, ensuring the auditability and traceability of the optimization process.

[0097] By using blockchain technology, key information such as sensor data, model versions, optimization schemes, and control commands are generated into hash values ​​and stored on the chain, forming an immutable and trustworthy data chain, providing a technical foundation for engineering quality traceability, safety assessment, and smart contract settlement.

[0098] Hashing of critical operation data

[0099] Let the first The key operation data set for the batch is:

[0100]

[0101] in This represents the i-th type of data (such as sensor data, decision instructions, etc.).

[0102] Perform a hash operation on each data item:

[0103]

[0104] in, For cryptographic hash functions (such as SHA-256).

[0105] The aforementioned adjustable vibration damping hole device is a mechanical-hydraulic composite structure that can receive instructions from edge intelligence agents and adjust its effective hole diameter, depth, and internal damping body density in real time.

[0106] The method for constructing a trusted data chain is as follows: generate a Merkle tree from key data and periodically write its root hash value into a public blockchain to achieve data tamper-proofing and end-to-end operation evidence storage;

[0107] Merkle tree construction and root hash calculation

[0108] Hash values ​​from the same batch Construct a Merkle tree using leaf nodes.

[0109] If the number of leaf nodes is odd, then copy the last leaf node to form an even number of nodes.

[0110] The hash value of a non-leaf node is calculated by concatenating the hash values ​​of its child nodes:

[0111]

[0112] The Merkle root hash of the batch of data is obtained by recursively calculating to the root node. (That is, the hash value of the root node, which does not need to be hashed again).

[0113] 3. Blockchain Evidence Storage Process

[0114] root hash With timestamp Package other metadata into transaction data:

[0115]

[0116] After signing the transaction, it is submitted to the blockchain network:

[0117]

[0118] After a transaction is verified through consensus, it is written into a block, forming an immutable data chain.

[0119] Based on an immutable data chain, it supports data auditing, effect traceability, and responsibility determination for the entire vibration reduction control process.

[0120] To support the above method, the present invention provides an intelligent vibration reduction system for foundation pit blasting to implement the above method, comprising: a digital twin engine module, an intelligent design brain module, an edge control node module, and a blockchain evidence storage platform module;

[0121] The system adopts a microservice architecture, with each functional module encapsulated as an independent service, communicating through an API gateway, and deployed on a hybrid cloud platform.

[0122] Digital Twin Engine Module: As the central hub of the system, it is responsible for the bidirectional synchronization and real-time interaction between the physical entity and the virtual model.

[0123] The intelligent design brain module integrates a macro-optimizer and a micro-topology optimizer, serving as the "thinking center" for generating innovative vibration reduction solutions.

[0124] Edge control node module: Deployed on-site, with built-in reinforcement learning agent, it is the "nerve ending" for performing precise control.

[0125] Blockchain-based evidence storage platform module: As the "cornerstone of trust" of the system, it ensures the authenticity and traceability of data throughout the entire process.

[0126] The following examples illustrate this:

[0127] Example 1: Vibration control of a deep foundation pit project for a subway adjacent to an operating tunnel

[0128] 1. Project Background and Challenges

[0129] A subway deep foundation pit project, with an excavation depth of 28 meters and a minimum clearance of only 15 meters from an existing operating tunnel, was under contractually tight budget. The contract required that the blasting vibration velocity be controlled below 2.0 cm / s.

[0130] 2. Implementation process of the invention

[0131] (1) Construction of digital twin

[0132] Accurate site parameters were obtained through geological surveys and BIM modeling: sandstone, longitudinal wave velocity. =2200 m / s, density =2450kg / m3.

[0133] Twenty-four vibration sensors were deployed, and real-time data was transmitted to the digital twin engine via a 5G network.

[0134] Initialize the intelligent fluctuation theory model, neural network correction term It adopts a 3-layer fully connected structure with a hidden layer dimension of 64 and the activation function is ReLU.

[0135] (2) Two-stage optimization design

[0136] Macro-layout optimization: NSGA-III algorithm is used, with a population size of 200 and 500 generations. Design variable: Number of rows. Hole spacing aperture Optimization goals: Vibration isolation rate > 70%, cost < budget, and lowest carbon emissions.

[0137] Micro-topology optimization: Three optimal macro-solutions are selected from the Pareto front, and micro-optimization is performed on key areas adjacent to the tunnel. Objective function weights are set. =0.6, volume constraint =0.4. After 150 iterations, a multi-level honeycomb composite structure was generated, and it is recommended that the interior be filled with a high-damping polyurethane material (damping coefficient = 0.4). =0.15).

[0138] (3) Edge intelligent control

[0139] Four edge computing nodes were deployed around the foundation pit.

[0140] Each node runs an agent based on the DQN algorithm with an attention mechanism. State characteristics. Includes triaxial vibration velocity and dominant frequency at 8 measuring points. Query vector. Initialized to [0.8, 0.1, 0.1], representing the priorities of "vibration control", "energy consumption" and "equipment lifespan" respectively.

[0141] (4) Blockchain-based evidence storage

[0142] All key data (sensor readings, model predictions, control commands) are packaged every 10 seconds to generate a Merkle root hash and written to the Ethereum testchain.

[0143] 3. Implementation effect verification

[0144] Safety indicators: During the entire blasting operation, the peak vibration velocity of the operating tunnel wall was successfully controlled below 1.85 cm / s, meeting the safety requirements.

[0145] Performance Comparison: Compared with another area on the same construction site that uses traditional experience-based design, the solution of this invention:

[0146] Vibration reduction effect improved by 35.1%.

[0147] Material usage reduced: 24.5%

[0148] Engineering costs decreased by 19.7%.

[0149] Innovative formula verification:

[0150] The correlation coefficient between the intelligent fluctuation model prediction and the measured data reached 0.91 (compared to 0.72 for the traditional model).

[0151] The composite structure generated by topology optimization has an average attenuation rate that is 28.7% higher than that of the regular aperture in the 30-80Hz main frequency band.

[0152] Attention mechanisms accelerated the training convergence speed of reinforcement learning agents by 45%.

[0153] Credibility: The blockchain records all 16,542 key operation logs, providing a complete and tamper-proof data evidence chain for project acceptance and auditing.

[0154] In summary, the method of this invention integrates physical priors and data-driven approaches through an intelligent wave theory model, endowing the model with online self-calibration and continuous evolution capabilities. This fundamentally solves the problem of high distortion in traditional models under complex geological conditions, improving prediction accuracy by more than 30%.

[0155] By integrating topology optimization with strain energy and kinetic energy dissipation, a qualitative leap has been achieved in vibration damping hole design, moving from "manual experience-based drawing" to "AI-generated optimal composite structure." The irregular rigid-damped composite structure discovered by this method has a vibration damping efficiency that is 25%-40% higher than that of traditional regular holes.

[0156] By employing reinforcement learning based on attention mechanisms, a cloud-edge collaborative architecture of "digital twin-edge intelligence" was constructed, establishing a self-regulating closed loop with millisecond-level response. The decision-making efficiency of the intelligent agent was improved by more than 40%, effectively responding to unexpected situations.

[0157] By introducing blockchain technology, a trusted data chain was established throughout the entire process, providing tamper-proof evidence to support engineering quality management and safety responsibility definition, thereby improving the transparency and trust level of the entire industry.

[0158] Practice has shown that, compared with traditional empirical methods, the method of this invention can improve the vibration reduction effect by more than 30% at a lower cost, while reducing material waste by more than 20% through precise topology optimization, thus achieving a win-win situation for multiple objectives of safety, economy and environmental protection.

[0159] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0160] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0161] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the digital twin-based intelligent vibration reduction methods for foundation pit blasting described in the above embodiments.

[0162] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0163] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0164] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0165] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0166] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A smart vibration reduction method for foundation pit blasting based on digital twins, characterized in that, Includes the following steps: S100, constructing a digital twin of the foundation pit blasting site; S200, based on a digital twin, runs a two-stage optimization strategy; S300 deploys a lightweight reinforcement learning agent with optimized strategies on edge computing nodes to perform online adaptive control of blasting parameters and adjustable vibration damping hole devices based on real-time monitoring data. S400, based on blockchain technology, constructs a complete and reliable data chain from the sensor data and decision-making instructions of the adaptive and adjustable vibration damping hole device to the effect evaluation, ensuring the auditability and traceability of the optimization process; The two-stage optimization strategy structure in step S200 includes: The first stage uses a multi-objective evolutionary algorithm to optimize the macroscopic layout of the vibration reduction hole group, and the second stage uses a topology optimization method to automatically generate a microscopic vibration reduction structure with irregular shape and composite filling material. Among them, macro layout optimization: taking the number of rows, columns and spacing of vibration damping holes as macro layout parameters as design variables, the NSGA-III multi-objective genetic algorithm is used to conduct a global Pareto front search with the goals of vibration isolation rate, engineering cost and environmental impact, and to determine several high-performance macro layout schemes. Microstructure topology optimization: Topology optimization is introduced to integrate strain energy and kinetic energy dissipation, simultaneously optimizing the stiffness and damping characteristics of the structure. The specific function is as follows: in, Due to volume constraints, This is the strain energy term; maximizing this term makes the structure more deformable by reflecting and scattering wave energy. This is the kinetic energy dissipation term. Maximizing this term allows the structure to more effectively convert wave energy into heat energy. For material density, Here, v is the damping coefficient related to the pseudo-density, and v is the vibration velocity field. For the weighting factor; The reinforcement learning agent in step S300 includes: The state value function incorporating the attention mechanism is as follows: Among them, S t Let θ represent the system state at time t, and let θ be the set of learnable parameters in the state value function. These are state feature vectors from different sensors. The query vector is a learnable vector representing the current control objective. For attention weights, It is a trainable linear transformation matrix.

2. The intelligent vibration reduction method for foundation pit blasting based on digital twins according to claim 1, characterized in that: The digital twin construction method in step S100 includes: A smart wave theory model is constructed by integrating Geographic Information System (GIS), Building Information Modeling (BIM), and Internet of Things (IoT) sensor data. This model adopts a hybrid architecture that combines classical physics theory with data-driven correction. For the physical prediction term based on the angular spectrum diffraction theory, x is the change in spatial position and ω is the angular frequency; To trainable weights The parameterized neural network correction term takes historical wavefields as its input. Rock mass parameters Temperature gradient and seepage velocity .

3. The intelligent vibration reduction method for foundation pit blasting based on digital twins according to claim 1, characterized in that, The method for constructing the trusted data link in step S300 includes: The real-time sensing data of the adjustable vibration damping hole device, the model prediction data of the digital twin, the optimized vibration damping structure parameters, the decision instructions generated by the reinforcement learning agent, and the evaluation results of the blasting vibration effect are used as key operational data. Let the first The key operation data set for the batch is: in Represents the i-th type of data; Perform a hash operation on each data item: in, For cryptographic hash functions; Merkle trees are constructed for the key operation data in time batches, and their root hash values ​​are calculated; Hash values ​​from the same batch As leaf nodes, construct a Merkle tree; If the number of leaf nodes is odd, then copy the last leaf node to form an even number of nodes; The hash value of a non-leaf node is calculated by concatenating the hash values ​​of its child nodes: The Merkle root hash of the t-th batch of data is obtained by recursively calculating up to the root node. ; root hash With timestamp Package other metadata into transaction data: After signing the transaction, it is submitted to the blockchain network: After a transaction is verified through consensus, it is written into a block, forming an immutable data chain. Based on an immutable data chain, it supports data auditing, effect traceability, and responsibility determination for the entire vibration reduction control process.

4. A digital twin-based intelligent vibration reduction system for foundation pit blasting, used to execute the method described in any one of claims 1-3, characterized in that, The intelligent vibration reduction system for blasting includes: a digital twin engine module, an intelligent design brain module, an edge control node module, and a blockchain evidence storage platform module; Each functional module communicates through an API gateway and is deployed on a hybrid cloud platform; among them, the intelligent design brain module integrates a macro optimizer and a micro topology optimizer; the edge control node module has a built-in reinforcement learning agent.

5. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the processor performs the method as described in any one of claims 1 to 3.

Images