A digital and intelligent technology-based reef bombing ship intelligent platform

Abstract

This invention relates to the field of marine engineering technology, and in particular to a digital and intelligent platform for blasting vessels based on digital and intelligent technologies. The platform includes a data processing module for collecting and processing status and environmental data of the blasting vessel; a positioning module for acquiring data from the data processing module, locating and calibrating the position of the blasting vessel and its drilling points; a monitoring module for acquiring data from the data processing module and analyzing it to obtain safety risk levels; and a scheduling module for acquiring data from the data processing module, the monitoring module, and the construction tasks, and analyzing it to obtain operation scheduling strategies. This invention provides precise control over the entire blasting operation process, overcoming the technical limitations of traditional operations that rely on manual experience, shortening the blasting operation period, reducing costs, significantly improving the economic benefits of the project, and effectively reducing ecological damage and navigation safety hazards.

Description

Technical Field

[0001] This invention relates to the field of marine engineering technology, and in particular to a digital and intelligent platform for blasting vessels based on digital and intelligent technologies. Background Technology

[0002] In infrastructure projects such as port and shipping engineering and inland waterway improvement, blasting operations are a core component for dredging waterways and improving their navigation quality. Their efficiency directly determines the overall traffic capacity and operational safety of the waterway. Therefore, promoting the digital and intelligent transformation of this process has become an inevitable trend in smart port and shipping construction. While some digital exploration has been undertaken in blasting operations—for example, using a dual positioning system of BeiDou and GPS to accurately locate reefs and lay out blast holes, and utilizing intelligent monitoring systems for real-time control of blasting vibrations—traditional operational methods still have many core shortcomings that are difficult to overcome.

[0003] Traditional blasting vessel operations rely excessively on manual experience and judgment throughout the drilling, charging, and blasting processes. The subjectivity and limitations of manual operation lead to low efficiency and extended project durations. Furthermore, the lack of real-time monitoring and intelligent early warning mechanisms makes it difficult to identify safety hazards such as misfires and residual reefs, posing a dual safety threat to both construction personnel and passing vessels. Simultaneously, the traditional model relies primarily on manual data recording, resulting in fragmented and scattered data that hinders coordinated scheduling and optimized decision-making across multiple devices and stages. This makes the vessel significantly less adaptable and capable of adjusting to complex geological conditions, variable marine hydrological environments, or fluctuating water levels during flood season. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a digital and intelligent platform for blasting vessels based on digital and intelligent technologies. This platform enables precise control over the entire blasting operation process, overcoming the technical limitations of traditional operations that rely on manual experience. It shortens the blasting operation period, reduces costs, significantly improves the economic benefits of the project, and effectively reduces ecological damage and navigation safety hazards.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A digital and intelligent platform for blasting vessels based on digital and intelligent technologies includes:

[0007] The data processing module is used for collecting and processing data on the status of the blasting vessel and environmental data to obtain standard data, and the data processing module uploads the standard data to the server.

[0008] The positioning module is used for data acquisition by the data processing module. The positioning module locates the position of the blasting vessel and the drilling point of the blasting vessel, and the positioning module calibrates the positioning data.

[0009] The monitoring module is used to acquire data from the data processing module. The monitoring module sets different warning values ​​for depth, temperature, humidity and gas concentration according to geological conditions. The monitoring module compares the real-time data of the data processing module with the corresponding warning values ​​to obtain the safety risk level.

[0010] The scheduling module is used for data acquisition from the data processing module, the monitoring module, and construction tasks. The scheduling module optimizes the scheduling based on equipment utilization, energy consumption, safety risks, and construction progress to obtain a work scheduling strategy.

[0011] Furthermore, the data processing module includes a data acquisition submodule, a data processing submodule, and a communication submodule.

[0012] The data acquisition submodule is used to acquire the status data of the blasting vessel and the environmental data. The status data of the blasting vessel includes sensor data and shipborne positioning data of the blasting vessel, and the environmental data includes depth sensor data, temperature and humidity sensor data, and gas sensor data.

[0013] The data processing submodule is used to acquire data from the data acquisition submodule. Based on the status data of the blasting vessel, the data processing submodule acquires the position, heading, and attitude data of the blasting vessel, and also acquires the operating data of the drilling rig and the explosives loading system. The data processing submodule acquires operational environment parameters based on the environmental data. The data processing submodule connects the data from the data acquisition submodule to the data acquisition terminal via an industrial bus, and obtains standard data after format conversion.

[0014] The communication submodule is used to construct wired and wireless redundant link data interaction, so that the data of the data processing submodule can be uploaded to the server via industrial Ethernet in the near field, and the data of the data processing submodule can be uploaded to the server via 5G network and Beidou satellite in offshore operations.

[0015] 3. The digital and intelligent platform for blasting vessels based on digital and intelligent technologies according to claim 1, characterized in that: the positioning module includes a positioning output submodule, a positioning correction submodule, and a positioning calibration submodule.

[0016] The positioning output submodule is used for the initial configuration of basic information of the blasting vessel and geographical information of the construction area. The positioning output submodule constructs a state equation including position, velocity and attitude and a Beidou observation equation through an extended Kalman filter algorithm. After prediction, it obtains low-frequency high-precision Beidou data and high-frequency high-dynamic inertial navigation system data fused by the updated closed-loop process to obtain positioning data of the blasting vessel's position and the drilling point position of the blasting vessel.

[0017] The positioning correction submodule is used to perform real-time correction of the inertial sensor during the zero-speed phase of the low-speed operation of the reef blasting vessel.

[0018] The positioning calibration submodule performs dynamic error compensation during the blasting vessel's operation.

[0019] Furthermore, the positioning correction submodule is used to trigger the ZUPT mechanism during the zero-speed phase of the low-speed operation of the reef blasting vessel, so as to embed the zero speed of the reef blasting vessel as a virtual observation value into the filtering process, so as to correct the zero bias of the inertial sensor in real time.

[0020] Furthermore, after collecting manually calibrated point observation data with known WGS84 coordinates at preset intervals, the positioning calibration submodule constructs an error model to dynamically compensate for errors in the positioning data based on the manually calibrated point observation data. The error model is as follows:

[0021] Formula (1)

[0022] in, The data are observations of manually calibrated points in WGS84 coordinates. Scale factor; For location data; This is the position offset; This is the random error term;

[0023] The scale factor and the position offset are solved using the least squares method, and the optimal values ​​of the scale factor and the position offset are calculated with the objective of minimizing the sum of squared errors.

[0024] Formula (2)

[0025] in, The first The coordinates of the first location data, the first Coordinates of observation data from individual manual calibration points; The number of manual calibration points;

[0026] The partial derivatives of the scale factor and the position offset are calculated to minimize the sum of squared errors, and then set to zero to obtain the optimal values ​​of the scale factor and the position offset. These optimal values ​​are then used to compensate for the real-time positioning data. The method for calculating the optimal values ​​of the scale factor and the position offset is as follows:

[0027] Formula (3)

[0028] in, This is the optimal value of the scale factor; This is the optimal value for the position offset.

[0029] Furthermore, the positioning module also includes a scenario-based optimization submodule. This submodule preprocesses the data from the inertial sensor using a moving average filter with a preset fixed time window to suppress high-frequency vibration noise through mean calculation within the window. The filtering method is as follows:

[0030] Formula (4)

[0031] in, For the first Filtering at any time; To pass through a preset fixed time window; For the first Raw sampled values ​​from the inertial sensor at any given time, window coverage to Continuous sampling points at any given time.

[0032] Furthermore, the monitoring module includes a risk model construction submodule and a comparison submodule.

[0033] The risk model construction submodule sequentially performs anomaly removal, noise filtering, and normalization on the data of depth, temperature, humidity, and gas concentration to obtain effective data. Based on safety standards, a basic probability allocation function is constructed on the effective data, and multi-source information is fused through Dempster combination rules to calculate the comprehensive confidence level of different risk levels. Different warning values ​​are set for depth, temperature, humidity, and gas concentration according to geological conditions, and a risk model is constructed.

[0034] The comparison submodule is used to input the real-time data from the data processing module into the risk model, so that the risk model compares the real-time data with the corresponding warning value to obtain the safety risk level.

[0035] Furthermore, the risk model is equipped with an LSTM model, which acquires data from the data processing module at the most recent preset time to predict future time data, thereby obtaining predicted data on depth, temperature, humidity, and gas concentration. The risk model analyzes the predicted data to obtain the predicted safety risk level.

[0036] Furthermore, the scheduling module includes a scheduling scheme generation submodule and a scheduling scheme optimization submodule.

[0037] The scheduling scheme generation submodule constructs a fitness function based on equipment utilization, operational energy consumption, safety risks, and construction progress. It also sets corresponding weights for equipment utilization, operational energy consumption, safety risks, and construction progress based on data from the data processing module, the monitoring module, and construction tasks. Furthermore, it avoids premature convergence by using an adaptive crossover mutation rate, and iteratively outputs the optimal scheduling scheme.

[0038] The scheduling scheme optimization submodule uses the GM (1,1) model to predict the construction progress deviation at future times to determine the construction progress lag. When the construction progress lag exceeds the planned threshold, the scheduling scheme optimization submodule adjusts the weights to output an optimized scheduling scheme.

[0039] Furthermore, the construction progress deviation prediction method of the scheduling scheme optimization submodule includes the following steps:

[0040] S1. Perform first-order accumulation on historical construction progress data:

[0041] Formula (5)

[0042] in, This represents the historical construction progress data after first-order accumulation. This is historical construction progress data, and ;

[0043] S2. Construct a whitening differential equation for the first-order accumulated historical construction progress data:

[0044] Formula (6)

[0045] in, The development coefficient; The gray action quantity is used; the parameter vector is solved by the least squares method. get:

[0046] Formula (7)

[0047] in, It is a cumulative matrix; A vector of constant terms; The matrix is ​​the transpose; the accumulated matrix and the constant term vector are represented as follows:

[0048]

[0049] S3. Generate the predicted value of the cumulative sequence based on the solved parameters:

[0050] Formula (8)

[0051] in, This is the cumulative sequence prediction value for the next moment in the construction progress;

[0052] S4. Perform cumulative subtraction and restoration on the accumulated sequence predicted values ​​to obtain the predicted values ​​of the original construction progress data:

[0053] Formula (9)

[0054] in, This represents the predicted value for the next moment in the original construction schedule.

[0055] The beneficial effects of this invention are:

[0056] The data processing module collects multi-source data, including positioning, vibration, hydrology, and equipment operation data, from various deployed intelligent sensors and transmits and aggregates this data in real time. Hydrological parameters collected by underwater environment sensors provide a basis for dynamic adjustments to the blasting plan, while equipment operation sensors capture real-time equipment health status data. Simultaneously, the data processing module constructs a dynamic support system adapted to blasting operation scenarios, achieving full-scenario communication coverage for both near-field high-speed and far-sea bottom-protection operations. The positioning module employs a data source quality assurance scheme that combines initialization configuration and dynamic calibration, achieving accurate and stable positioning data in complex sea areas through multi-algorithm fusion and scenario-based optimization. The monitoring module implements proactive risk prevention based on a threshold trigger mechanism. Operators can set a three-level threshold system for depth, temperature and humidity, and gas concentration through a threshold preset interface based on geological conditions and safety regulations. This is achieved by constructing a system that includes data preprocessing, DS... This invention employs an algorithm system that integrates evidence theory and lightweight time-series prediction to achieve low-latency, hierarchical, and effective early warning capabilities. The scheduling module acquires data on equipment status, construction progress, weather, and sea conditions through a multi-source data access module. It then generates an optimized work scheduling plan using a fitness function. After operator approval via a confirmation interface, the instruction push program sends scheduling commands to the shipborne terminal, dynamically optimizing the work process to improve efficiency. This invention can be widely applied to port and waterway engineering scenarios such as inland waterway improvement, coastal port construction, and blasting reefs for cross-sea channels. It can shorten the blasting operation period and reduce costs, significantly improving the economic benefits of the project while effectively reducing ecological damage and navigation safety hazards. It can also promote the transformation of blasting operations from "experience-driven" to "data-driven," providing key technical support for smart port and waterway construction. Furthermore, the core concept can be extended to similar water engineering fields such as underwater excavation and offshore blasting, possessing broad prospects for industrial application. Attached Figure Description

[0057] Figure 1 This is a structural block diagram of a digital and intelligent platform for blasting vessels based on digital and intelligent technologies, according to a preferred embodiment of the present invention.

[0058] Figure 2This is a pyramid framework diagram of a preferred embodiment of the present invention, which is a digital and intelligent platform for blasting vessels based on digital and intelligent technologies.

[0059] Figure 3 This is a technical architecture diagram of a reef-blasting vessel digital intelligence platform based on digital and intelligent technologies, according to a preferred embodiment of the present invention.

[0060] Figure 4 This is a hardware block diagram of a digital and intelligent platform for blasting vessels based on digital and intelligent technologies, according to a preferred embodiment of the present invention.

[0061] In the diagram, 1-Data Processing Module, 11-Data Acquisition Submodule, 12-Data Processing Submodule, 13-Communication Submodule, 2-Positioning Module, 21-Positioning Output Submodule, 22-Positioning Correction Submodule, 23-Positioning Calibration Submodule, 24-Scenario-Based Optimization Submodule, 3-Monitoring Module, 31-Risk Model Construction Submodule, 32-Comparison Submodule, 4-Scheduling Module, 41-Scheduling Scheme Generation Submodule, 42-Scheduling Scheme Optimization Submodule. Detailed Implementation

[0062] 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 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.

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0064] Please see Figures 1 to 4 A preferred embodiment of the present invention provides a digital and intelligent platform for blasting vessels based on digital and intelligent technologies, comprising:

[0065] Data processing module 1 is used for data collection and processing of the status data and environmental data of the blasting vessel to obtain standard data, and data processing module 1 uploads the standard data to the server.

[0066] The data processing module 1 includes a data acquisition submodule 11, a data processing submodule 12, and a communication submodule 13.

[0067] The data acquisition submodule 11 is used to acquire data on the status of the blasting vessel and environmental data. The status data of the blasting vessel includes sensor data and shipborne positioning data of the blasting vessel, and the environmental data includes depth sensor data, temperature and humidity sensor data, and gas sensor data.

[0068] In this embodiment, high-precision real-time monitoring of the environment and equipment status is achieved by selecting various types of sensors. For depth monitoring, an ultrasonic underwater depth sensor is used to ensure continuous and stable operation in seawater immersion and drastic temperature fluctuation environments. For drill pipe stress monitoring, a patch-type strain sensor is deployed to effectively filter high-frequency noise generated by drilling rig vibration, with a signal-to-noise ratio of ≥95%. For equipment vibration monitoring, a triaxial accelerometer is used to capture abnormal vibration frequencies of the equipment in real time. All sensors are calibrated on-site to ensure reduced data errors.

[0069] The data processing submodule 12 is used to acquire data from the data acquisition submodule 11. Based on the status data of the blasting vessel, the data processing submodule 12 acquires the position, heading, and attitude data of the blasting vessel, as well as the operating data of the drilling rig and the charging system. The data processing submodule 12 also acquires operational environment parameters based on environmental data. The data processing submodule 12 connects the data from the data acquisition submodule 11 to the data acquisition terminal via an industrial bus, and obtains standard data after format conversion. In this embodiment, the data is processed by the format conversion program to form standardized data frames awaiting uploading.

[0070] The communication submodule 13 is used to construct wired and wireless redundant link data interaction, so as to upload the data of the data processing submodule 12 to the server via industrial Ethernet in the near field, and to upload the data of the data processing submodule 12 to the server via 5G network and Beidou satellite in offshore operations.

[0071] In this embodiment, to address the challenges of complex electromagnetic environments and long-range maritime communication, the transmission layer employs a dual-mode redundancy architecture of a 5G industrial gateway and a BeiDou satellite communication terminal. The 5G private network utilizes network slicing technology to ensure end-to-end latency, supporting real-time transmission of device status and positioning data. BeiDou satellite communication serves as a backup link in scenarios without public network coverage at sea, transmitting critical data such as fault warnings and emergency commands, eliminating communication blind spots in the maritime area. Key control commands are encrypted using national cryptographic algorithms and include a breakpoint resume module, automatically caching data during network interruptions and prioritizing retransmission upon connection restoration, ensuring the stability and security of data interaction.

[0072] Positioning module 2 is used for data acquisition by data processing module 1. Positioning module 2 locates the position of the blasting vessel and the drilling point of the blasting vessel, and calibrates the positioning data.

[0073] The positioning module 2 includes a positioning output submodule 21, a positioning correction submodule 22, a positioning calibration submodule 23, and a scenario-based optimization submodule 24.

[0074] The positioning output submodule 21 is used for the initial configuration of basic information of the blasting vessel and geographical information of the construction area. The positioning output submodule 21 constructs state equations including position, velocity and attitude and Beidou observation equations through extended Kalman filter algorithm. After prediction, it obtains low-frequency high-precision Beidou data and high-frequency high-dynamic inertial navigation system data fused by the updated closed-loop process to obtain positioning data of the blasting vessel's position and the drilling point position of the blasting vessel.

[0075] To construct a highly reliable spatiotemporal reference for the intelligent scheduling platform system for reef blasting vessels, a data source quality assurance scheme with a dual mechanism of initialization configuration and dynamic calibration is proposed. Through multi-algorithm fusion and scenario-based optimization, the accuracy and stability of positioning data in complex sea areas are achieved. Before operation, operators complete the initial configuration of basic vessel parameters and geographic information of the construction area through the platform's parameter input interface. Then, through a prediction-to-update closed-loop process, low-frequency high-precision BeiDou data and high-frequency high-dynamic inertial navigation system data are fused to output centimeter-level positioning results.

[0076] The positioning correction submodule 22 is used to perform real-time correction of the inertial sensor during the zero-speed phase of the low-speed operation of the blasting vessel.

[0077] The positioning correction submodule 22 is used to trigger the ZUPT mechanism during the zero-speed phase of the low-speed operation of the blasting vessel, so as to embed the zero speed of the blasting vessel as a virtual observation into the filtering process, so as to correct the zero bias of the inertial sensor in real time.

[0078] This embodiment embeds zero velocity as a virtual observation value into the filtering process to achieve real-time correction of the inertial sensor's zero bias and effectively suppress short-term drift errors.

[0079] The positioning calibration submodule 23 performs dynamic error compensation during the blasting vessel operation.

[0080] After collecting observation data from manually calibrated points with known WGS84 coordinates at preset intervals, the positioning calibration submodule 23 constructs an error model to dynamically compensate for errors in the positioning data based on the manually calibrated point observation data. This embodiment dynamically compensates for system errors, designing a dynamic calibration process every 30 minutes, collecting observation data from at least 3 manually calibrated points with known WGS84 coordinates to ensure the degrees of freedom in solving the linear error model. The error model is:

[0081] Formula (1)

[0082] in, The data are observations of manually calibrated points in WGS84 coordinates. Scale factor; For location data; This is the position offset; This is the random error term;

[0083] The scale factor and position offset are solved using the least squares method, with minimizing the sum of squared errors as the objective to calculate the optimal values ​​of the scale factor and position offset.

[0084] Formula (2)

[0085] in, The first The coordinates of the first location data, the first Coordinates of observation data from individual manual calibration points; The number of manual calibration points;

[0086] To minimize the sum of squared errors, calculate the partial derivatives of the scale factor and position offset separately, and make the partial derivatives of the scale factor and position offset equal to 0 to obtain the optimal values ​​of the scale factor and position offset. Then, compensate the real-time positioning data with the optimal values ​​of the scale factor and position offset. The method for calculating the optimal values ​​of the scale factor and position offset is as follows:

[0087] Formula (3)

[0088] in, This represents the optimal value of the scaling factor. This is the optimal value for the position offset.

[0089] In this embodiment, the optimal scaling factor obtained by solving will be... and position offset Reverse compensation is applied to the real-time positioning results to achieve dynamic correction of system errors and ensure that borehole position deviations meet construction accuracy requirements.

[0090] The scenario-based optimization submodule 24 preprocesses the inertial sensor data using a moving average filter with a preset fixed time window. This filter suppresses high-frequency vibration noise by calculating the mean value within the window. The filtering method is as follows:

[0091] Formula (4)

[0092] in, For the first Filtering at any time; In order to pass through a preset fixed time window, this embodiment ; For the first Raw sampled values ​​from the inertial sensor at any given time, window coverage to Continuous sampling points at any given time.

[0093] This embodiment addresses the unique environmental constraints of offshore operations by employing a scenario-based optimization submodule to preprocess inertial sensor data using a 5-second fixed-time window of moving average filtering. The window covers… to Five consecutive sampling points at each time point are smoothed by mean in the time domain, significantly reducing the interference of high-frequency noise caused by hull vibration on state estimation. Through a loose-tight combination switching logic of the BeiDou signal-to-noise ratio (SNR), when SNR < 15dB, it switches to loose combination mode to prioritize data stability; when SNR ≥ 15dB, it switches to tight combination mode to restore real-time fusion. This dynamic switching mechanism balances positioning accuracy and data stability.

[0094] The monitoring module 3 is used to acquire data from the data processing module 1. The monitoring module 3 sets different warning values ​​for depth, temperature, humidity and gas concentration according to geological conditions. The monitoring module 3 compares the real-time data of the data processing module 1 with the corresponding warning values ​​to obtain the safety risk level.

[0095] Monitoring module 3 includes risk model construction submodule 31 and comparison submodule 32.

[0096] The risk model construction submodule 31 sequentially performs anomaly removal, noise filtering, and normalization on the data of depth, temperature, humidity, and gas concentration to obtain valid data. Based on safety standards, a basic probability allocation function is constructed on the valid data, and multi-source information is integrated through Dempster combination rules to calculate the comprehensive confidence level of different risk levels. Different warning values ​​are set for depth, temperature, humidity, and gas concentration according to geological conditions, and a risk model is constructed.

[0097] In this embodiment, firstly, through The principle is to eliminate out-of-the-box errors in normally distributed data such as gas concentration, and then calculate the mean of the data series. and standard deviation It will exceed Instantaneous jump values ​​within a range are identified as anomalies and removed. For non-normally distributed state data such as equipment vibration, an outlier is identified using the isolated forest algorithm. A forest model containing 100 isolated trees is constructed, and by calculating the anomaly score of the samples, noise interference from ship vibration is effectively filtered out, improving the efficiency of the preprocessed data. Min-max normalization is used to unify the dimensions, ensuring effective data and further improving data efficiency.

[0098] Subsequently, using depth, temperature and humidity, and gas concentration as independent evidence, a basic probability allocation function was constructed based on safety standards. Multi-source information was fused through Dempster combination rules to calculate the comprehensive trust level for different risk levels. Based on preset thresholds, Level 1 (yellow light), Level 2 (audio-visual and shore-based push), and Level 3 (automatic protection) early warning responses were triggered.

[0099] The comparison submodule 32 is used to input the real-time data from the data processing module 1 into the risk model, so that the risk model can compare the real-time data with the corresponding warning value to obtain the safety risk level.

[0100] The risk model is equipped with an LSTM model. The LSTM model obtains data from the most recent preset time from the data processing module 1 to predict data for future time, thereby obtaining predicted data on depth, temperature, humidity, and gas concentration. The risk model analyzes the predicted data to obtain the predicted safety risk level.

[0101] This embodiment uses nearly 10 minutes of time-series data as input to predict parameter changes over the next 5 minutes, allowing time for intervention. The algorithm dynamically adjusts thresholds using fuzzy logic to adapt to different geological scenarios and is deployed on a shipborne edge terminal to ensure low-latency response, achieving closed-loop prevention and control of multi-dimensional risks.

[0102] Monitoring module 3 uses lightweight algorithms to preprocess local data, compares it against safety thresholds in real time, and triggers primary warnings, reducing remote transmission bandwidth usage. It also relies on historical fault correlation analysis from the fault database to provide real-time and source-tracing evidence for decision-making. The emergency command and control system acts as the central hub, linking a 3D monitoring interface to display visualized information about the blasting vessel's location, underwater blasting area, and equipment status. It also triggers a three-tiered warning response: Level 1 warnings only push local audio and visual signals, Level 2 warnings are simultaneously pushed to the application and mobile terminals, and Level 3 warnings directly trigger emergency commands such as equipment emergency stop and surrounding support dispatch. The multi-terminal linkage system ensures information synchronization and two-way command interaction between operators and commanders. Through data-driven decision transformation and tiered collaborative handling, it effectively reduces the safety risks of special operations on blasting vessels while significantly improving cross-role collaboration efficiency.

[0103] The scheduling module 4 is used to acquire data from the data processing module 1, the monitoring module 3, and the construction task. The scheduling module 4 optimizes the scheduling based on equipment utilization, energy consumption, safety risks, and construction progress to obtain the operation scheduling strategy.

[0104] The scheduling module 4 includes a scheduling scheme generation submodule 41 and a scheduling scheme optimization submodule 42.

[0105] The scheduling scheme generation submodule 41 constructs a fitness function based on equipment utilization, operational energy consumption, safety risks, and construction progress. It also sets corresponding weights for equipment utilization, operational energy consumption, safety risks, and construction progress based on data from data processing module 1, monitoring module 3, and construction tasks. Furthermore, it avoids premature convergence by using an adaptive crossover mutation rate and iteratively outputs the optimal scheduling scheme.

[0106] This embodiment first constructs a three-level hierarchical structure. The target layer aims for globally optimal operational efficiency; the criterion layer covers four core indicators: equipment utilization, operational energy consumption, safety risk, and construction progress; and the solution layer comprises different operational scheduling strategies. When constructing the criterion layer, pairwise comparison judgment matrices are used, and the maximum eigenvalue of the matrix is ​​solved using eigenvalue decomposition. The corresponding feature vectors are normalized to obtain the weight vector w=[0.4,0.1,0.2,0.3], where equipment utilization and construction progress are the core priority indicators. The improved GA uses integer encoding to represent the work sequence, constructs a fitness function that integrates equipment utilization, energy consumption, risk, and progress, avoids premature convergence through adaptive crossover and mutation rates, and iteratively outputs the optimal scheduling scheme.

[0107] The scheduling scheme optimization submodule 42 predicts the construction progress deviation at future times using a GM (1,1) model to determine the construction progress lag. When the construction progress lag exceeds the planned threshold, the scheduling scheme optimization submodule 42 adjusts the weights to output an optimized scheduling scheme. The GM (1,1) model predicts the construction progress deviation for the next 4 hours. This model does not rely on a large amount of sample data; it can achieve trend prediction using only a small amount of historical data, which aligns with the actual constraints of offshore operation data collection.

[0108] The steps of the construction progress deviation prediction method in the scheduling scheme optimization submodule 42 include:

[0109] S1. Perform first-order accumulation on historical construction progress data:

[0110] Formula (5)

[0111] in, The historical construction progress data is accumulated in the first order. By using the historical construction progress data accumulated in the first order, the random fluctuation of the original data can be weakened and the smoothness of the sequence can be enhanced. This is historical construction progress data, and ;

[0112] S2. Construct a whitening differential equation for the first-order accumulated historical construction progress data:

[0113] Formula (6)

[0114] in, The development coefficient; The gray action quantity is used; the parameter vector is solved by the least squares method. get:

[0115] Formula (7)

[0116] in, It is a cumulative matrix; A vector of constant terms; The matrix is ​​the transpose; the cumulative matrix and the vector of constant terms are represented as:

[0117]

[0118] S3. Generate the predicted value of the cumulative sequence based on the solved parameters:

[0119] Formula (8)

[0120] in, This is the cumulative sequence prediction value for the next moment in the construction progress;

[0121] S4. Perform cumulative subtraction to restore the cumulative sequence prediction values ​​to obtain the original construction progress data prediction values:

[0122] Formula (9)

[0123] in, This represents the predicted value for the next moment in the original construction schedule.

[0124] In this embodiment, the scheduling scheme optimization submodule 42 predicts the construction progress for the next 4 hours. If the predicted progress lags behind the plan by ≥10%, the dynamic adjustment mechanism of the scheduling scheme is automatically triggered. Through real-time equipment status data, resources are prioritized for allocation to drilling rigs and key operation nodes with optimal status, while simultaneously reducing the resource occupation time of non-critical operations. This forms a closed-loop management system for progress control, from prediction to adjustment and then execution, ensuring the overall project schedule is met. The algorithm employs offline pre-calculation and online fine-tuning modes to improve real-time performance. It also outputs three suboptimal schemes for manual selection, allowing operators to make decisions based on unforeseen on-site scenarios, achieving collaborative management from algorithm-driven to manual support. Finally, after confirmation by the operator, the optimized scheme is encrypted and pushed to the shipborne execution terminal via a standardized interface, ensuring the security of instruction transmission and the reliability of execution.

[0125] like Figure 2 and Figure 3As shown, the technical architecture of the intelligent platform for blasting reef vessels in this embodiment is divided into an intelligent perception and edge pre-control layer, a heterogeneous network fusion transmission layer, a multi-source data intelligent governance layer, and a human-machine collaborative application interaction layer.

[0126] Intelligent Sensing and Edge Pre-Control Layer: This layer serves as the system's physical sensing and near-end decision-making front end. Its modules include a high-precision positioning module, multi-parameter underwater environment sensors, equipment operation status sensors, and an edge computing early warning module equipped with a lightweight neural network. On one hand, the centimeter-level positioning module establishes a spatiotemporal reference for blasting operations; the hydrological parameters collected by the underwater environment sensors provide a basis for dynamic adjustments to the blasting plan; and the equipment operation sensors capture real-time equipment health status data. On the other hand, the edge computing early warning module, relying on the lightweight computing power of the edge gateway, performs real-time feature extraction and preliminary anomaly judgment on the sensed data, simultaneously triggering local audible and visual early warnings, achieving near-end risk interception without requiring full data backhaul. The core value of this layer lies in providing the entire system with a high-spatiotemporal accuracy and multi-dimensional correlation of raw data, while also compressing risk response latency to the second level through edge-level pre-control, becoming the first line of defense for the safety of special operations on blasting vessels.

[0127] Heterogeneous Network Convergence Transmission Layer: This layer serves as the cross-domain transmission hub connecting front-end sensing and back-end governance. Its modules are composed of a heterogeneous architecture based on 5G private networks and BeiDou satellite communication, complemented by edge computing nodes, lightweight data compression protocols, and algorithm-encrypted transmission security mechanisms. With coverage, efficiency, and security as the core guiding principles for transmission assurance, a dynamic assurance system adapted to blasting operations is constructed through multi-dimensional technological collaboration and process optimization. The 5G private network uses network slicing technology to ensure end-to-end latency, supporting real-time transmission of equipment status and location data. BeiDou satellite communication serves as a backup link in scenarios without public network coverage at sea, transmitting critical data such as fault warnings and emergency commands, eliminating communication blind spots in the maritime area. Edge computing nodes perform lightweight preprocessing on front-end data, using compression protocols to increase compressed data volume and reduce channel bandwidth usage. The encryption mechanism ensures the integrity and confidentiality of data on the heterogeneous link, preventing data tampering during transmission. The core value of this layer is to achieve full-scenario communication coverage for near-field high-speed and far-sea backup, providing low-latency and highly reliable multi-source data flow support for upper-layer applications such as 72-hour equipment failure prediction and blasting effect assessment.

[0128] Multi-Source Data Intelligent Governance Layer: This layer is the system's multi-source data standardization and high-availability management unit. It integrates the entire governance chain of raw data loading, cleaning, anomaly removal, and synchronization; a sliding window-based average filtering module; and a heterogeneous data storage architecture using MySQL, InfluxDB, and tiered storage. Standardization, adaptability, and high availability are the core governance objectives. The data governance process involves... The criteria for removing outliers from the perceived data and optimizing the stability of time-series data through moving average filtering employ a hybrid strategy of structured and time-series data storage. A MySQL database stores structured information such as job tasks and personnel scheduling, while an InfluxDB database carries time-series data collected every second, such as equipment vibration and location data. Simultaneously, a layered strategy of hot data memory caching and cold data file storage achieves a balance between data retrieval efficiency and storage cost. A cloud-based fusion mechanism further ensures cross-node data consistency across multiple databases. Through this design, multi-source heterogeneous front-end data is transformed into a high-quality data source with low redundancy, high correlation, and rapid retrieval, providing data quality support for intelligent decision-making in upper-layer human-machine collaborative applications.

[0129] Human-Machine Collaboration Application Interaction Layer: This layer serves as the system's intelligent control and human-machine interaction terminal. It integrates data support units such as real-time and fault databases, as well as core modules like the emergency command and linkage system, full-process 3D visualization monitoring, three-level early warning response, and multi-terminal linkage system, constructing a closed loop of human-machine collaboration encompassing data, decision-making, and execution. Specifically, the data support unit achieves millisecond-level retrieval of operational status through the real-time database and provides real-time and traceability-based dual-dimensional evidence for decision-making based on historical fault correlation analysis from the fault database. The emergency command and linkage system acts as the central hub, linking the 3D monitoring interface to display visualized information such as the location of the blasting vessel, the underwater blasting area, and equipment status. It also triggers a three-level early warning response: Level 1 warnings only push local audio and visual signals, Level 2 warnings are simultaneously pushed to the application and mobile terminals, and Level 3 warnings directly trigger emergency commands such as equipment emergency stop and surrounding support dispatch. The multi-terminal linkage system ensures information synchronization and two-way command interaction between operators and commanders. Through data-driven decision transformation and hierarchical collaborative handling, it effectively reduces the safety risks of special operations on blasting vessels while significantly improving cross-role collaboration efficiency.

[0130] like Figure 4 The diagram shows the hardware framework of the digital and intelligent platform for the blasting vessel based on digital and intelligent technologies in this embodiment.

[0131] In the perception layer, various types of sensors are selected for high-precision real-time monitoring of the environment and equipment status. Depth monitoring utilizes ultrasonic underwater depth sensors to ensure continuous and stable operation in seawater immersion and environments with drastic temperature fluctuations. Drill pipe stress monitoring employs patch-type strain sensors, effectively filtering high-frequency noise generated by drilling rig vibration, achieving a signal-to-noise ratio ≥95%. Equipment vibration monitoring uses triaxial accelerometers to capture abnormal vibration frequencies in real time. All sensors are calibrated on-site to minimize data errors. To address the challenges of complex electromagnetic environments and long-range communication, the transmission layer adopts a dual-mode redundant architecture of a 5G industrial gateway and a BeiDou satellite communication terminal, primarily carrying emergency warning information and critical control commands. End-to-end encryption is applied to transmitted data, and local caching and breakpoint resumption mechanisms are designed. Data is automatically stored during network interruptions and retransmitted according to timestamp priority after connection restoration, ensuring data integrity.

[0132] The computing and storage layer adopts a layered architecture of shipborne edge real-time processing and shore-based centralized storage backup. The shipborne end is equipped with industrial-grade edge computing terminals, which can directly connect to sensors and transmission modules. It performs local data preprocessing through lightweight algorithms, compares safety thresholds in real time, and triggers primary early warnings, reducing the bandwidth consumption of remote transmission. The shore-based end builds an industrial-grade server cluster to improve the computing efficiency of AI models such as geological feature recognition and blasting effect evaluation. At the same time, incremental synchronization is automatically completed every morning, and the historical data storage capacity supports TB-level data retention.

[0133] The safety layer addresses unforeseen risks through hardware redundancy and environmental adaptability design. The power system combines a shipboard main power supply with a 10kWh lithium iron phosphate backup power supply. In the event of a main power outage, an automatic switching module seamlessly takes over, ensuring continuous operation of core equipment such as sensors, edge terminals, and communication modules. Key hardware adopts a dual-machine hot standby architecture, with backup equipment taking over within 10 seconds in the event of a main equipment failure. Distributed data nodes are deployed, with one real-time data copy maintained on both the shipboard and shore-based sides, and data consistency is ensured through 1-minute incremental synchronization. All hardware installations utilize shock-resistant brackets, and the outer shell is coated with a marine-grade anti-corrosion coating, conforming to marine electrical equipment specifications. This ensures long-term stable operation under harsh conditions during reef blasting operations, providing reliable hardware support for the platform's software functionality.

Claims

1. A digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies, characterized in that: include: The data processing module (1) is used for data collection and processing of the status data and environmental data of the blasting vessel to obtain standard data, and the data processing module (1) uploads the standard data to the server. The positioning module (2) is used for data acquisition by the data processing module (1). The positioning module (2) locates the position of the blasting vessel and the drilling point of the blasting vessel, and calibrates the positioning data. The monitoring module (3) is used to acquire data from the data processing module (1), and the monitoring module (3) sets different warning values ​​for depth, temperature and humidity and gas concentration according to geological conditions. The monitoring module (3) compares the real-time data of the data processing module (1) with the corresponding warning values ​​to obtain the safety risk level. The scheduling module (4) is used to acquire data from the data processing module (1), the monitoring module (3), and the construction task. The scheduling module (4) optimizes the scheduling based on equipment utilization, operation energy consumption, safety risks, and construction progress to obtain the operation scheduling strategy.

2. The digital and intelligent platform for blasting vessels based on digital and intelligent technologies according to claim 1, characterized in that: The data processing module (1) includes a data acquisition submodule (11), a data processing submodule (12), and a communication submodule (13). The data acquisition submodule (11) is used for data acquisition of the status data of the blasting vessel and the environmental data. The status data of the blasting vessel includes sensor data and shipborne positioning data of the blasting vessel. The environmental data includes depth sensor data, temperature and humidity sensor data, and gas sensor data. The data processing submodule (12) is used to acquire data from the data acquisition submodule (11), and the data processing submodule (12) acquires the position data, heading data, and attitude data of the blasting vessel based on the status data of the blasting vessel, and acquires the operating data of the drilling rig and the charging system; the data processing submodule (12) acquires the operating environment parameters based on the environmental data; the data processing submodule (12) connects the data from the data acquisition submodule (11) to the data acquisition terminal through the industrial bus, and obtains standard data after format conversion processing; The communication submodule (13) is used to construct wired and wireless redundant link data interaction, so as to upload the data of the data processing submodule (12) to the server via industrial Ethernet in the near field, and upload the data of the data processing submodule (12) to the server via 5G network and Beidou satellite in offshore operations.

3. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 1, characterized in that: The positioning module (2) includes a positioning output submodule (21), a positioning correction submodule (22), and a positioning calibration submodule (23). The positioning output submodule (21) is used for the initial configuration of basic information of the blasting vessel and geographical information of the construction area. The positioning output submodule (21) constructs a state equation including position, velocity and attitude and a Beidou observation equation through an extended Kalman filter algorithm. After prediction, it obtains low-frequency high-precision Beidou data and high-frequency high-dynamic inertial navigation system data fused by the updated closed-loop process to obtain positioning data of the blasting vessel's position and the drilling point of the blasting vessel. The positioning correction submodule (22) is used to perform real-time correction of the inertial sensor during the zero-speed phase of the low-speed operation of the blasting vessel. The positioning calibration submodule (23) performs dynamic error compensation during the blasting vessel operation.

4. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 3, characterized in that: The positioning correction submodule (22) is used to trigger the ZUPT mechanism during the zero-speed phase of the low-speed operation of the blasting vessel, so as to embed the zero speed of the blasting vessel as a virtual observation value into the filtering process, so as to correct the zero bias of the inertial sensor in real time.

5. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 3, characterized in that: After collecting manually calibrated point observation data with known WGS84 coordinates at preset intervals, the positioning calibration submodule (23) constructs an error model to dynamically compensate for errors in the positioning data based on the manually calibrated point observation data. The error model is... Formula (1) in, The data are observations of manually calibrated points in WGS84 coordinates. Scale factor; For location data; This is the position offset; This is the random error term; The scale factor and the position offset are solved using the least squares method, and the optimal values ​​of the scale factor and the position offset are calculated with the objective of minimizing the sum of squared errors. Official (2) in, The first The coordinates of the first location data, the first Coordinates of observation data from individual manual calibration points; The number of manual calibration points; The partial derivatives of the scale factor and the position offset are calculated to minimize the sum of squared errors, and then set to zero to obtain the optimal values ​​of the scale factor and the position offset. These optimal values ​​are then used to compensate for the real-time positioning data. The method for calculating the optimal values ​​of the scale factor and the position offset is as follows: Official (3) in, This is the optimal value of the scale factor; This is the optimal value for the position offset.

6. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 3, characterized in that: The positioning module (2) further includes a scenario-based optimization submodule (24), which preprocesses the data from the inertial sensor by using a moving average filter with a preset fixed time window to suppress high-frequency vibration noise by calculating the mean within the window. The filtering method is as follows: Official (4) in, For the first Filtering at any time; To pass through a preset fixed time window; For the first Raw sampled values ​​from the inertial sensor at any given time, window coverage to Continuous sampling points at any given time.

7. The digital and intelligent platform for blasting vessels based on digital and intelligent technologies according to claim 1, characterized in that: The monitoring module (3) includes a risk model construction submodule (31) and a comparison submodule (32). The risk model construction submodule (31) sequentially performs anomaly removal, noise filtering, and normalization on the data of depth, temperature, humidity, and gas concentration to obtain effective data. Based on safety specifications, a basic probability allocation function is constructed on the effective data, and multi-source information is fused through Dempster combination rules to calculate the comprehensive trust level of different risk levels. Different warning values ​​are set for depth, temperature, humidity, and gas concentration according to geological conditions, and a risk model is constructed. The comparison submodule (32) is used to input the real-time data of the data processing module (1) into the risk model, so that the risk model compares the real-time data with the corresponding warning value to obtain the safety risk level.

8. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 7, characterized in that: The risk model is equipped with an LSTM model. The LSTM model obtains the data from the data processing module (1) at the most recent preset time to predict the data at future time, so as to obtain the predicted data of depth, temperature, humidity and gas concentration. The risk model analyzes the predicted data to obtain the predicted safety risk level.

9. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 1, characterized in that: The scheduling module (4) includes a scheduling scheme generation submodule (41) and a scheduling scheme optimization submodule (42). The scheduling scheme generation submodule (41) constructs a fitness function based on equipment utilization, operation energy consumption, safety risk and construction progress, and sets corresponding weights for equipment utilization, operation energy consumption, safety risk and construction progress based on the data processing module (1), the monitoring module (3) and the data of the construction operation task, and avoids premature convergence by adaptive cross mutation rate, so as to iteratively output the optimal scheduling scheme. The scheduling scheme optimization submodule (42) predicts the construction progress deviation at future times using the GM (1,1) model to determine the construction progress lag. When the construction progress lag exceeds the planned threshold, the scheduling scheme optimization submodule (42) adjusts the weights to output an optimized scheduling scheme.

10. The digital and intelligent platform for blasting reef vessels based on digital and intelligent technologies according to claim 9, characterized in that: The construction progress deviation prediction method of the scheduling scheme optimization submodule (42) includes the following steps: S1. Perform first-order accumulation on historical construction progress data: Official (5) in, This represents the historical construction progress data after first-order accumulation. This is historical construction progress data, and ; S2. Construct a whitening differential equation for the first-order accumulated historical construction progress data: Official (6) in, The development coefficient; The gray action quantity is used; the parameter vector is solved by the least squares method. get: Official (7) in, It is a cumulative matrix; A vector of constant terms; The matrix is ​​the transpose; the accumulated matrix and the constant term vector are represented as follows: S3. Generate the predicted value of the cumulative sequence based on the solved parameters: Official (8) in, This is the cumulative sequence prediction value for the next moment in the construction progress; S4. Perform cumulative subtraction and restoration on the accumulated sequence predicted values ​​to obtain the predicted values ​​of the original construction progress data: Official (9) in, This represents the predicted value for the next moment in the original construction schedule.

Images