Overwater pile foundation construction safety risk early warning system

Abstract

The invention relates to the field of data processing, and discloses an overwater pile foundation construction safety risk early warning system which comprises a data model updating module used for fusing real-time data to synchronize a digital data model; the risk prediction data generation module is used for performing simulation based on the synchronized model to generate a future risk index curve; and a management decision optimization module. When the risk index exceeds a preset threshold value, the module can automatically generate a candidate management instruction set, and the efficiency of each instruction is quantitatively evaluated through simulation deduction, so that a unique optimal management instruction is determined, and a negative management instruction is identified. According to the method, continuous quantitative prediction is carried out on future risks, and automatic deduction and optimization are carried out on the efficiency of multiple treatment measures, so that the problems that risk assessment lags behind, decision-making depends on artificial experience and optimality cannot be ensured in the prior art are solved, and automatic and prospective closed-loop management and control of the risks are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of household kitchen, in particular to a water pile foundation construction safety risk early warning system. BACKGROUND

[0002] Water pile foundation construction is a basic link in marine engineering, bridge construction and offshore wind farm construction. The operation process usually involves large construction platforms, heavy piling equipment and large-size piles to be installed, and is carried out in open waters with continuous changes in wind, wave, current and other environmental loads. The combination of these factors makes the dynamic nature and uncertainty of the construction operation process significant, and is accompanied by high operation risk.

[0003] The existing safety guarantee measures mainly rely on static operation procedures and safety plans formulated before construction. These procedures and plans are based on statistical analysis of historical environmental data and simplified physical model calculations to define a fixed safety operating window for the construction process. In the actual operation process, specific parameters (such as platform attitude angle, mooring cable tension, etc.) are monitored by on-site sensors, and the on-site management personnel compares them with the pre-set, isolated threshold values in the procedures to determine whether the current state is safe.

[0004] This approach lacks a mechanism for continuously comparing and evaluating real-time dynamic information during the construction process with pre-set safety boundaries, and cannot infer the state evolution trend of the system in the future based on the current state and future environmental forecasts. Therefore, risk assessment is not a continuous and forward-looking process, but a discrete and static check of the current state.

[0005] When the monitored parameters approach or exceed the safety threshold, management intervention measures need to be taken. In the prior art, the decision-making process of choosing which intervention measures (such as adjusting the water volume of a specific ballast tank, changing the tension of part of the mooring cable, or reducing the impact energy of the pile hammer) lacks the step of quantitatively predicting and comparing the system dynamic response triggered by the execution of different measures. Management decisions often rely on the qualitative judgment of on-site personnel rather than quantitative assessment of the future effectiveness of different intervention measures.

[0006] This makes the identification and disposal of risks lagging behind and unable to ensure that the management decisions taken are optimal in quantitative terms, making it difficult to achieve a precise balance between operation efficiency and safety guarantee. Therefore, there is a need in the field to provide a technical solution that can dynamically and forwardly quantify construction risks and automatically optimize management decisions. SUMMARY

[0007] In view of the deficiencies of the prior art, the water pile foundation construction safety risk early warning system is provided, which solves the problems that the risk assessment in the prior art depends on static threshold judgment, lacks foresight, and cannot quantitatively compare and automatically optimize the future efficiency of various management intervention measures when making risk disposal decisions.

[0008] To achieve the above object, the following technical scheme is adopted: a water pile foundation construction safety risk early warning system, comprising: a data model maintenance module, a real-time data receiving module, a data model updating module, a risk prediction data generation module, a management decision optimization module and a management instruction issuing module.

[0009] The data model maintenance module is used to maintain a digital data model representing a water pile foundation construction site. The digital data model is a computer processable data set, which includes: 1) geometric structure data representing physical objects such as construction platforms, piling equipment and pile foundations in the construction site; 2) physical property data associated with the geometric structure data, including mass, moment of inertia, material elastic modulus, yield strength, used to define the response characteristics of the geometric structure data under the driving of physical laws in subsequent calculations.

[0010] The real-time data receiving module is used to continuously receive real-time sensor data collected from various sensors deployed in the water pile foundation construction site. The real-time sensor data includes platform motion posture data, hydrological environment data, structure health data and equipment working condition data.

[0011] The data model updating module is used to continuously update the digital data model based on the real-time sensor data. Specifically, this module uses a data fusion algorithm, such as the extended Kalman filter (EKF) algorithm, to fuse the real-time sensor data with the system state prediction value of the digital data model, generate an optimal posterior state estimation value, and update the digital data model with the posterior state estimation value, so that the data state of the digital data model is synchronized with the true state of the physical construction site.

[0012] The risk prediction data generation module is used to perform forward-looking calculation based on the updated digital data model. This module takes the synchronized digital data model state as the initial condition, drives the internal physics engine solver to calculate the system behavior in the future at a faster rate than the real time elapses, generates a future state sequence representing the future system behavior. Based on the future state sequence, further generate risk index data representing future construction risks.

[0013] The risk index data is specifically dynamic composite risk index data. The dynamic composite risk index data The calculation process is defined by the following formula: ; in, For some point in the future; The total number of preset risk components; Index for risk components; This refers to the current management stage of underwater pile foundation construction; For the first Each risk component in the management phase Dynamic weighting coefficients; For the first Each risk component in the future The quantized value of a given moment.

[0014] The management decision optimization module is configured to be triggered when the peak value of the risk index data exceeds a preset risk threshold. Upon triggering, the module executes the following steps: Generate a candidate management instruction set containing multiple candidate management instructions. The candidate management instruction set includes specific management instructions for the current risk type extracted from a preset management plan library, as well as no-operation instructions for benchmark comparison. .

[0015] For the aforementioned candidate management instruction set Each candidate management instruction in Independent performance simulation calculations are performed. These calculations are based on a digital data model of the current state and simulate the execution of instructions. The subsequent system behavior is analyzed, and its corresponding risk index data evolution curve is generated. .

[0016] Based on step 2 For each A decision effectiveness evaluation value is calculated. The decision effectiveness evaluation value... The calculation process is defined by the following formula: ; in, Candidate management instructions The corresponding decision-making effectiveness evaluation value; The current moment; The preset evaluation time domain; To execute instructions The evolution curve of the risk index data afterward.

[0017] Based on the decision performance evaluation value, the optimal management instruction is determined from the candidate management instruction set. The method of determination is to find the... The candidate management instruction with the smallest value: .

[0018] At the same time, the decision-making effectiveness evaluation value Greater than the decision effectiveness evaluation value corresponding to no operation instruction Candidate management instructions are identified as negative management instructions.

[0019] The management instruction issuing module is used to generate and issue risk response decision information containing the optimal management instruction and the negative effect management instruction. This module also includes a data interface, which, when the risk index data exceeds a preset first-level hazard threshold, is configured to send the management instruction data of the optimal management instruction to the control system of the construction equipment to execute the instruction.

[0020] This invention provides a safety risk early warning system for underwater pile foundation construction. It has the following beneficial effects: 1. This invention uses a risk prediction data generation module to perform forward-looking extrapolation calculations based on an updated digital data model to generate risk index data representing a future period of time. This enables forward-looking prediction of construction risks, overcomes the problem of delayed early warning caused by relying on fixed thresholds in existing technologies, and reserves effective handling time for risk response.

[0021] 2. This invention performs independent performance calculations on multiple candidate management instructions through a management decision optimization module, and determines the optimal management instruction based on the decision performance evaluation value. It provides verified and quantitative optimal decision support, avoiding the uncertainty and risk that may arise from relying solely on the operator's experience to make decisions in emergency situations, and reducing the possibility of accidents caused by human error.

[0022] 3. By adopting dynamic composite risk index data and adaptively adjusting the dynamic weight coefficients of different risk components according to different management stages of underwater pile foundation construction, this invention improves the accuracy and pertinence of risk assessment, enabling the risk assessment results to more effectively reflect the core risk points under the current working conditions and enhance the overall effectiveness of the early warning system. Attached Figure Description

[0023] Figure 1 This is a system functional module architecture diagram of the present invention; Figure 2 This is a flowchart illustrating the overall system workflow of the present invention. Figure 3 This is a schematic diagram of the digital data model composition of the present invention; Figure 4 This is a flowchart illustrating the management decision optimization process of this invention. Detailed Implementation

[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Please see the appendix Figure 1 - Appendix Figure 4 This invention provides a safety risk early warning system for underwater pile foundation construction, including a data model maintenance module 110, a real-time data receiving module 120, a data model updating module 130, a risk prediction data generation module 140, a management decision optimization module 150, and a management instruction issuing module 160.

[0026] See attached document Figure 1 , Figure 1 This is a functional module architecture diagram of a data processing system for risk management of underwater pile foundation construction according to an embodiment of the present invention. The present invention provides a data processing system for risk management of underwater pile foundation construction, which may include: a data model maintenance module 110, a real-time data receiving module 120, a data model updating module 130, a risk prediction data generation module 140, a management decision optimization module 150, and a management instruction issuing module 160.

[0027] See attached document Figure 2 , Figure 2 This is a flowchart of a data processing system for risk management in offshore pile foundation construction according to an embodiment of the present invention. The specific workflow of the system is described below.

[0028] The real-time data receiving module 120 is configured to continuously receive real-time sensor data collected from multiple sensors deployed at the physical underwater pile foundation construction site via a data communication network. The real-time sensor data is structured into a data stream and sent to the data model update module 130.

[0029] Upon receiving real-time sensor data, the data model update module 130 invokes the digital data model stored in the data model maintenance module 110 and executes a data fusion algorithm. This algorithm fuses the real-time sensor data with the internal state predictions of the digital data model to generate a corrected posterior state estimate. This posterior state estimate is then used to update the data state of the digital data model. This process is repeated cyclically to ensure that the data state of the digital data model remains synchronized with the actual state of the physical construction site.

[0030] The risk prediction data generation module 140 periodically retrieves the latest version of the state-synchronized digital data model from the data model maintenance module 110. Based on this model, the risk prediction data generation module 140 drives an internal physics engine solver to perform forward-looking extrapolation calculations at a rate faster than real time, generating a future state sequence characterizing the behavior of the construction system over a future period, and calculating risk index data based on this sequence.

[0031] The generated risk index data is continuously monitored. When the value in the risk index data exceeds a preset risk threshold, the risk prediction data generation module 140 sends a trigger signal to the management decision optimization module 150 and transmits the current digital data model status along with the risk index data.

[0032] The management decision optimization module 150 is activated upon receiving a trigger signal. This module first generates a candidate management instruction set containing multiple candidate management instructions. Subsequently, for each candidate management instruction, the module performs independent performance simulation calculations in parallel, calculating a quantified decision performance evaluation value for each candidate management instruction.

[0033] After evaluating all candidate management instructions, the management decision optimization module 150 determines an optimal management instruction and identifies all negative management instructions by comparing all decision effectiveness evaluation values. This module then sends the calculated optimal and negative management instruction data to the management instruction issuing module 160.

[0034] After receiving the optimal management instruction and the negative effect management instruction data, the management instruction issuing module 160 integrates them into a structured risk response decision-making message. This message can be issued through a human-machine interface for management personnel to reference. Simultaneously, this module also includes a data interface that, under specific conditions, such as when the risk index data exceeds the first-level hazard threshold, can convert the optimal management instruction into executable management instruction data and send it to the construction equipment's control system.

[0035] See attached document Figure 3 , Figure 3 This is a schematic diagram illustrating the composition of a digital data model according to an embodiment of the present invention. During the initialization phase of the data processing system, the static construction of the digital data model must first be completed in the data model maintenance module 110. The digital data model is a computer-processable, structured collection of data, and its construction process includes the generation of geometric structure data and the configuration of physical attribute data.

[0036] In one specific embodiment, the geometric structure data portion of the digital data model is generated by fusing multi-source data. This process includes the following steps: First, importing Building Information Model (BIM) data containing 3D information of entities such as the offshore construction platform, piling equipment, and pile foundations; second, importing Geographic Information System (GIS) data containing underwater topography and geomorphological features of the construction area; finally, registering and fusing the BIM data and GIS data in a unified global coordinate system to generate a unified geometric structure data set that accurately represents all key entities and their spatial relationships at the offshore pile foundation construction site.

[0037] After generating the geometric structure data, it is necessary to configure the corresponding physical attribute data for each geometric component in the dataset. This process associates specific physical parameters with the corresponding geometric components and stores this data in the data model maintenance module 110. The physical attribute data specifically includes: Inertial properties include the mass of each component, the position of its center of mass, and the rotational inertia tensor that describes its rotational characteristics.

[0038] Material mechanical properties: including the elastic modulus, Poisson's ratio, and yield strength of the materials that make up the various structural components (such as piles and platform structures).

[0039] Equipment operating attributes: including the working parameters of power equipment such as pile hammers, such as hammer mass, maximum impact stroke and maximum impact energy.

[0040] Fluid dynamic properties: including fluid dynamic coefficients of the parts of the construction platform, pile body and other parts in contact with water, such as the added mass coefficient, damping coefficient and flow force coefficient.

[0041] Through the above steps, a static digital data model containing precise geometric information and complete physical properties is constructed. This static model provides fundamental data support for subsequent dynamic state synchronization and forward-looking risk simulation.

[0042] After the digital data model completes its static construction, the system enters the operational phase. During this phase, the data model update module 130 continuously executes dynamic synchronization tasks to ensure that the data state of the digital data model remains consistent with the actual state of the physical construction site. This dynamic synchronization process is implemented through a data fusion algorithm; in one specific embodiment, this algorithm is the Extended Kalman Filter (EKF) algorithm.

[0043] The specific steps for data model update module 130 to perform dynamic synchronization are as follows: First, define the system's state vector. This vector contains all the key variables describing the dynamic characteristics of the digital data model at a given moment. For example, the state vector. This can include the six degrees of freedom position and velocity of the construction platform, the stress and strain of key structural nodes, and the motion state of the pile driver.

[0044] The module's operation within a time step includes a prediction step and an update step.

[0045] Prediction steps: This step is based on the previous moment. The optimal state estimate is used to predict the current time. The system state. This process does not depend on the current sensor measurements, but is accomplished by driving the physical law solver within the digital data model.

[0046] The calculation process for state prediction is defined by the following formula: ; in, For at any time The prior state estimate, i.e. the predicted state; For at any time The posterior state estimate is the optimal estimate from the previous time step. For at any time The control input vector, such as the control command of the pile driver or the control command of the ballast system; This is a nonlinear state transition function, which in this embodiment is represented by the physics engine solver of the digital data model.

[0047] Simultaneously, the covariance matrix of the state estimate is predicted, and its calculation process is defined by the following formula: ; in, For at any time The prior estimate of the covariance matrix; For at any time The posterior estimated covariance matrix; For the state transition function in The Jacobian matrix calculated at point; The process noise covariance matrix represents the uncertainty of the physical model itself.

[0048] Update steps: This step uses the data provided by the real-time data receiving module 120 at the current moment. Real-time sensor data collected The prior state estimate generated in the prediction step Perform corrections to obtain the optimal state estimate at the current time. .

[0049] First, calculate the Kalman gain. Its calculation process is defined by the following formula: ; in, For observation function exist The Jacobian matrix calculated at the location, the observation function Used to map the system's state vector to the sensor's measurement space; The noise covariance matrix is ​​observed, which characterizes the uncertainty of sensor measurement data.

[0050] Then, based on the Kalman gain and the actual measured value, the current time is calculated. The posterior state estimate, i.e. the optimal state estimate, is calculated using the following formula: ; in, For at any time The actual sensor measurement vector.

[0051] Finally, the covariance matrix of the updated state estimate is used to prepare for the next round of prediction. The calculation process is defined by the following formula: ; in, It is an identity matrix.

[0052] The optimal state estimate obtained by performing the above update steps It is used to update the digital data model stored in the data model maintenance module 110. This process is executed repeatedly at each time step, thereby achieving continuous and dynamic state synchronization between the digital data model and the physical construction site.

[0053] In one specific embodiment, after the digital data model completes a state synchronization, the risk prediction data generation module 140 immediately executes a forward-looking extrapolation process. The purpose of this process is to calculate and generate a future state sequence characterizing the system behavior over a future period of time, based on the current, synchronized system state.

[0054] Specifically, the risk prediction data generation module 140 first obtains the current time from the data model update module 130. Optimal posterior state estimation This serves as the initial condition for the simulation calculations. Simultaneously, the module also obtains environmental condition forecasts for a future period from external data sources, such as wind, wave, and current forecasts, as external stimulus inputs for the simulation process.

[0055] The risk prediction data generation module 140 integrates a physics engine solver. This solver numerically solves for the multiphysics coupled dynamics of the system based on the geometric and physical property data defined in the digital data model. These physical fields include, but are not limited to: the rigid body dynamics of the construction platform, the finite element structural mechanics of the pile structure, and the fluid dynamics around the platform and pile.

[0056] The calculation process of forward-looking extrapolation takes place within a pre-defined prediction time domain. In discrete time steps The process is iterative. Within each time step, the physics engine solver performs one iterative calculation, the process of which is defined by the following general state equation: ; in, This represents the current moment in the deduction process, and its value range is... ; In order to simulate the moment The system state vector; It is a nonlinear state transition operator, represented by the physics engine solver, used to solve for the state of the system at the next time step. In order to simulate the moment The external excitation force vector applied to the system, which is calculated from environmental condition prediction data.

[0057] The iterative calculation starts from the initial conditions. Start, repeat execution until the deduced time. Reaching the end of the prediction time domain The execution time of the entire inference and calculation process was designed to be significantly shorter than the prediction time domain. The length.

[0058] After the extrapolation calculations are completed, all the system state vectors calculated within the prediction time domain are combined into a time series data set, i.e., the future state sequence. Its mathematical expression is: ; This future state sequence It contains deterministic data descriptions of the system's dynamic behavior over a future period and is used as the basis for subsequent calculations of dynamic composite risk index data.

[0059] The risk prediction data generation module 140 generates a future state sequence. The module then proceeds to the next step, which is to calculate the dynamic composite risk index data based on the sequence. This risk index data is a time series, representing the risk index over the prediction time domain. Within, the overall risk level of the construction system evolves over time.

[0060] Dynamic composite risk index data The calculation process is defined by the following formula: ; in, To predict a specific moment in the time domain, its value range is: ; The total number of preset risk components; Index for risk components; This is an identifier representing the current management stage of the underwater pile foundation construction; For the first Each risk component in the management phase Dynamic weighting coefficients; For the first Each risk component in the future The normalized quantized value at time t is within the range.

[0061] Risk Components The value is based on the future state sequence The risk component set is calculated from the specific data contained therein. In a specific embodiment, the risk component set may include, but is not limited to, the following two items: Platform collapse risk weight This component aims to quantify the risk of attitude instability of the construction platform at future moments. The calculation process for this component is defined by the following formula: ; in, and In the future The platform's roll and pitch angles are directly derived from the state vectors in the future state sequence. Extract from; The preset platform safety attitude angle limit.

[0062] Stress risk components of pile structure This component aims to quantify the risk of material failure in pile foundation structures at future moments. The calculation process for this component is defined by the following formula: ; in, For the future moment The maximum equivalent stress (e.g., von Mises stress) on the critical section of the pile is determined by the state vector. The results were obtained from finite element analysis. This represents the yield strength limit of the pile material.

[0063] Dynamic weighting coefficients The value is based on the current management stage of the construction. Adaptive configuration is performed during the management phase. It is a predefined identifier; for example, S1 represents the pile lifting stage, and S2 represents the pile driving stage. The system focuses on different risks at different management stages. For example: During management phase S1 (pile-lifting phase), platform attitude stability is the primary risk source. Therefore, the weighting coefficients are configured to place greater emphasis on platform capsizing risk, and their weight vector can be set as follows: .

[0064] In management phase S2 (pile driving phase), the structural bearing capacity of the pile itself is the primary source of risk. Therefore, the weighting coefficients are reconfigured to place greater emphasis on the stress risk of the pile structure, and its weight vector can be set as follows: .

[0065] Based on the above calculations, the risk prediction data generation module 140 ultimately generates a prediction time domain. Dynamic composite risk index data curve within The data curve is transmitted to the management decision optimization module 150 for subsequent risk assessment and optimal decision-making.

[0066] See attached document Figure 4 , Figure 4 This is a management decision optimization flowchart according to an embodiment of the present invention. When the dynamic composite risk index data generated by the risk prediction data generation module 140... When the peak value exceeds a preset risk threshold, the management decision optimization module 150 is activated and begins to execute the automated optimization process.

[0067] The first step in this process is to generate a candidate management instruction set C containing multiple candidate management instructions. The management decision optimization module 150 is internally connected to a pre-set management plan library. This plan library stores structured management instructions for different risk types.

[0068] The management decision optimization module 150 first analyzes the risk index data that triggered this optimization process. The composition of the risk components is analyzed to identify the risk component that contributes the most, thereby determining the current dominant risk type. For example, if the platform collapse risk component... The risk with the highest weight and value in the total risk is identified as platform instability.

[0069] Subsequently, based on the identified dominant risk type, the module retrieves and extracts all management instructions associated with that risk type from the pre-set management plan library. Each management instruction is a machine-readable data unit that defines specific operations and quantitative parameters.

[0070] For example, in the case where the dominant risk type is platform attitude instability, candidate management instructions extracted from the contingency plan library may include: instruction : Inject 50 tons of ballast water into ballast tank numbered A01.

[0071] instruction Adjust the mooring system and increase the tension of mooring cable number M03 by 10%.

[0072] For example, when the dominant risk type is excessive stress in the pile structure, candidate management instructions extracted from the contingency plan database may include: instruction The impact energy setting of the pile driver is reduced by 20%.

[0073] instruction : Suspend piling operations and keep the piling hammer at rest.

[0074] After extracting all relevant management instructions, the management decision optimization module 150 will also add a special instruction, namely the no-operation instruction, to the instruction set. This no-action instruction means that no proactive intervention measures will be taken for a period of time in the future. Its purpose is to provide a benchmark for subsequent effectiveness evaluation.

[0075] Ultimately, this module generates a complete set of candidate management instructions. ,in This represents the total number of instructions extracted from the contingency plan library. This instruction set is used as input for the next step, namely parallel simulation and performance evaluation.

[0076] The management decision optimization module 150 generated a set of candidate management instructions. Then, the module proceeds to process each candidate management instruction in the instruction set. (in From 0 to In a specific embodiment, to improve computational efficiency, the performance simulation calculations for different candidate management instructions are executed simultaneously in multiple parallel computation processes.

[0077] For a single candidate management instruction The specific steps for its performance deduction and calculation are as follows: First, the management decision optimization module 150 uses the current moment... The digital data model that has been synchronized with the state is used as the initial condition for this simulation.

[0078] Then, the candidate management instruction As a control input, it is applied to the digital data model in its initial state. For example, if the instruction... To inject 50 tons of water into a designated ballast tank, the mass and centroid parameters of the corresponding ballast tank in the digital data model will be updated according to this instruction at the beginning of the simulation calculation.

[0079] Next, the module drives the internal physics engine solver within a preset evaluation time domain. Inside, instructions were issued. The system's dynamic behavior is then simulated. This simulation process is consistent with the aforementioned risk prediction and deduction process, but its input includes system changes triggered by specific management instructions. After the simulation, a data set corresponding to that instruction is generated. The unique corresponding risk index data evolution curve within the assessment time domain. .

[0080] In obtaining the evolution curve of risk index data Subsequently, the management decision optimization module 150 continues to process the candidate management instruction. Calculate a quantitative evaluation value for decision effectiveness. This evaluation value characterizes the performance if the instruction is executed. The system will assess the cumulative risk over the future time domain.

[0081] Decision effectiveness evaluation value The calculation process is defined by the following formula: ; in, Candidate management instructions The corresponding decision-making effectiveness evaluation value; The current moment in time at which this calculation is performed; The preset evaluation time domain; To simulate the execution of instructions The evolution curve of the risk index data generated afterwards.

[0082] By analyzing the candidate management instruction set Each instruction in the process repeats the above calculation process, and the management decision optimization module 150 ultimately generates a corresponding decision effectiveness evaluation value for each candidate management instruction. These evaluation values ​​provide a quantitative basis for determining the optimal management instruction in the subsequent process.

[0083] After the management decision optimization module 150 completes the performance evaluation calculation of all instructions in the candidate management instruction set, the module obtains a dataset containing all candidate management instructions and their corresponding decision performance evaluation values. Based on this dataset, the module continues to execute the final instruction determination step.

[0084] First, the module evaluates all calculated decision effectiveness values. A comparison is made to determine the management instruction that minimizes the cumulative risk over the future assessment timeframe. This optimal management instruction... The determination process is defined by the following formula: ; in, The final, optimal management instruction; It is a collection that contains all candidate management instructions; Candidate management instructions The corresponding decision-making effectiveness evaluation value. Through this calculation, a unique, quantitatively optimal management instruction is determined. It has been confirmed.

[0085] At the same time, the module will also evaluate the decision-making effectiveness of each candidate management instruction. With no operation instruction The corresponding decision effectiveness evaluation value The comparison is then performed. This step aims to identify management instructions that, when executed, would increase system risk.

[0086] If a candidate management instruction The evaluation value meets This condition, then the instruction It has been identified and marked as a negative management instruction. This indicates that the cumulative risk resulting from executing this instruction is higher than the cumulative risk of not taking any action.

[0087] After completing the above determination and identification process, the management decision optimization module 150 will determine the optimal management instruction. The data, along with that of all identified negative management instructions, are combined to form a decision outcome dataset. This decision outcome dataset is then sent to the management instruction issuing module 160 for subsequent issuance and execution.

[0088] After receiving the decision result dataset sent by the management decision optimization module 150, the management instruction issuing module 160 executes the decision information issuance and closed-loop control process. This module's functions include issuing information to management personnel and issuing automatic control instructions to equipment.

[0089] In one specific embodiment, the management instruction issuing module 160 first formats the received decision result dataset to generate a structured risk response decision information. This information is then sent to a human-machine interface deployed at the construction site or remote monitoring center for display. The risk response decision information specifically includes: the determined optimal management instruction. The text description, a list of all identified negative management instructions, and the risk index data evolution curve after simulating the execution of the optimal management instructions. Evolution curve of risk index data under no-operation instruction The comparison chart.

[0090] In addition, the management instruction issuing module 160 also includes a data interface for communication with the equipment control system at the construction site, thereby forming a closed-loop control circuit. The function of this data interface is activated under specific conditions.

[0091] The activation condition is: dynamic composite risk index data initially generated by the risk prediction data generation module 140 without any intervention. The peak value exceeded a preset level 1 hazard threshold, which is higher than the usual risk threshold.

[0092] When this condition is met, the data interface is activated. Once activated, the interface will transmit the optimal management instructions determined by the management decision optimization module 150. Its internal logical representation is converted into a management instruction data packet that conforms to a specific industrial communication protocol (such as Modbus TCP or OPCUA protocol).

[0093] The management instruction data packet is then sent via wired or wireless network to the underlying control system of the physical device associated with the instruction, such as the ballast water pump control system of a construction platform, the mooring winch control system, or the PLC (Programmable Logic Controller) of a piling machine. After receiving and parsing the data packet, the device's control system automatically executes the instructions contained therein, thereby achieving automated, closed-loop handling of impending risks.

[0094] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0095] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A safety risk early warning system for underwater pile foundation construction, characterized in that, include: The data model maintenance module is used to maintain a digital data model that represents the construction site of underwater pile foundations. A real-time data receiving module is used to receive real-time sensor data collected from the construction site of the underwater pile foundation; The data model update module is used to continuously update the digital data model based on the real-time sensor data to reflect the current status of the underwater pile foundation construction site. The risk prediction data generation module is used to perform extrapolation and calculation based on the updated digital data model to generate risk index data that characterizes construction risks in the future period. The management decision optimization module is configured to generate a candidate management instruction set containing multiple candidate management instructions when the risk index data exceeds a preset risk threshold. Furthermore, the management decision optimization module is also used to perform independent performance deduction calculations for each candidate management instruction in the candidate management instruction set to generate a corresponding decision performance evaluation value, and to determine the optimal management instruction from the candidate management instruction set based on the decision performance evaluation value. The management instruction issuing module is used to issue risk response decision information containing the optimal management instructions.

2. The safety risk early warning system for underwater pile foundation construction according to claim 1, characterized in that, The digital data model includes: Geometric data characterizing the construction platform, piling equipment, and pile foundation at the construction site of the underwater pile foundation; Physical attribute data associated with the geometric structure data, the physical attribute data being used to define the response characteristics of the geometric structure data under the drive of physical laws.

3. The safety risk early warning system for underwater pile foundation construction according to claim 1, characterized in that, The data model update module is specifically used for: A data fusion algorithm is used to fuse the real-time sensor data with the predicted state of the digital data model in order to update the digital data model.

4. The safety risk early warning system for underwater pile foundation construction according to claim 1, characterized in that, The risk index data is dynamic composite risk index data, which is generated by weighted summation of multiple risk component data representing different risk types.

5. The safety risk early warning system for underwater pile foundation construction according to claim 4, characterized in that, The weights of the multiple risk component data are dynamic weight coefficients, which are adaptively adjusted according to the current management stage of the underwater pile foundation construction.

6. The safety risk early warning system for underwater pile foundation construction according to claim 1, characterized in that, The candidate management instruction set includes management instructions extracted from a preset management plan library, as well as no-operation instructions for benchmark comparison.

7. The safety risk early warning system for underwater pile foundation construction according to claim 1, characterized in that, The decision-making effectiveness evaluation value is obtained by performing an integral calculation on the risk index data corresponding to the candidate management instructions within a preset evaluation time domain.

8. The safety risk early warning system for underwater pile foundation construction according to claim 7, characterized in that, The management decision optimization module determines the optimal management instruction by finding the candidate management instruction that minimizes the decision effectiveness evaluation value.

9. A safety risk early warning system for underwater pile foundation construction according to claim 8, characterized in that, The management decision optimization module is further used to identify candidate management instructions whose decision effectiveness evaluation value is greater than the decision effectiveness evaluation value corresponding to the no-operation instruction as negative management instructions, and the management instruction issuing module is further used to issue the negative management instructions.

10. A safety risk early warning system for underwater pile foundation construction according to claim 1, characterized in that, The management instruction issuing module also includes a data interface, which is configured to send the management instruction data of the optimal management instruction to the control system of the construction equipment when the risk index data exceeds the first-level danger threshold.

Images