A proppant delivery risk assessment method and system incorporating a controlled environment

Abstract

The application provides a fracturing ship powder conveying risk assessment method and system combined with a controllable environment, relates to the technical field of ship equipment safety analysis, and specifically comprises the following steps: based on a plurality of single controllable environment factors, at least one fused controllable environment factor is constructed; the fused controllable environment factor is used for representing a systematic safety situation generated under the coupling action of a plurality of single controllable environment factors; a control structure model is established, the control structure model comprises a controller, a controlled process, a control instruction and a feedback loop of the powder conveying system, and the single controllable environment factor and the fused controllable environment factor are taken as components of the feedback loop; based on the control structure model, in the context defined by the single controllable environment factor and the fused controllable environment factor, unsafe control behaviors leading to system hazards are identified, and causation scenarios leading to the occurrence of the unsafe control behaviors are determined to complete risk assessment.

Description

Technical Field

[0001] This invention relates to the field of ship equipment safety analysis technology, and more specifically, to a method and system for risk assessment of powder transportation on fracturing vessels in a controlled environment. Background Technology

[0002] With the deepening development of offshore oil and gas resources, large fracturing vessels with high integration and high operational efficiency have become important marine engineering equipment. One of the core operations of a fracturing vessel is hydraulic fracturing, the key step of which is mixing powdered particulate materials used as proppant with liquid to produce high-pressure fracturing fluid. The powder conveying system is the core subsystem responsible for this crucial task, precisely and stably conveying dry powder materials stored in the ship's large silos to the mixing unit via pneumatic or mechanical means. This system typically consists of storage tanks, conveying pipelines, valves, fans, dust collectors, and a series of automated controllers, making it a complex system with a high degree of mechatronics integration and close interaction between equipment.

[0003] Compared to traditional land-based fracturing operations, applying powder delivery systems to fracturing vessels presents more severe and unique safety challenges. First, the operating environment on ships is far harsher than on land. The system must withstand the periodic rolling and tilting of the hull caused by wind and waves, which continuously affects the stable flow of powder and the structure of equipment. Simultaneously, the inherent high-salt spray environment of the marine atmosphere causes continuous and accelerated corrosion to the metal structures, electrical components, and precision sensors in the system, easily leading to material failure or control signal distortion. Second, as an independent platform integrating production and living areas, fracturing vessels have a compact spatial layout. The powder delivery system is often very close to personnel activity areas and other high-risk operating areas (such as power and manifold areas), requiring operators and crew to coexist in the same confined space. This makes the consequences of dust leaks, fires, or explosions far more severe than on land, potentially directly threatening the lives of all personnel on board and the structural safety of the vessel.

[0004] Traditional safety analysis methods, such as Hazard and Operability Analysis (HAZOP), have significant limitations in risk assessment for such complex industrial systems. These methods typically focus on the physical failure modes of individual devices or components within the system, such as valve jamming or pipeline rupture, and deduce their potential impact by analyzing failure event chains. However, for highly integrated and automated powder conveying systems, many significant risks do not stem from simple failures of single components, but rather from unsafe interactions between multiple normally operating components under specific time sequences and environments, as well as complex human-machine-environment couplings. Traditional methods are insufficient to identify such systemic risks. Furthermore, the effectiveness of these traditional methods largely depends on historical accident data and failure rate statistics, which are obviously unavailable in the early design stages for systems first applied to large fracturing vessels, making assessment difficult.

[0005] To overcome the shortcomings of traditional methods, fields such as aerospace have developed System Theory-Based Process Analysis (STPA) methods. STPA treats safety as a dynamic control problem, identifying risks by constructing functional control structure models and analyzing unsafe control behaviors (UCAs) between controllers in the system. It can effectively reveal systemic risks caused by complex interactions, software defects, and human factors. However, the standard STPA method has a key limitation in its theoretical foundation when applied to specific scenarios such as ships. Standard STPA explicitly requires the separation of the system from its environment when defining the analysis boundary, assuming that the environment is an external condition beyond the control of system designers and engineers. For example, in aviation, engineers cannot control the weather along flight routes; in manufacturing, engineers cannot control urban planning outside the factory area. This fundamental assumption of treating the environment as an uncontrollable external disturbance, while valid in many fields, limits its accuracy on highly self-sufficient and manageable platforms such as fracturing ships. Therefore, a safety analysis method more suited to the unique environment of ships is urgently needed to guide their safety design. Summary of the Invention

[0006] This specification proposes a risk assessment method for fracturing vessel powder transportation in a controlled environment, the method comprising:

[0007] Multiple individual controllable environmental factors were collected from the powder conveying system of the fracturing vessel.

[0008] Based on the multiple individually controllable environmental factors, at least one integrated controllable environmental factor is constructed; the integrated controllable environmental factor is used to characterize the systemic security situation generated under the coupled effect of the multiple individually controllable environmental factors;

[0009] A control structure model is established, which includes the controller, controlled process, control command, and feedback loop of the powder conveying system, and the individual controllable environmental factors and the integrated controllable environmental factors are included as components of the feedback loop.

[0010] Based on the control structure model, in the context defined by the individual controllable environmental factors and the integrated controllable environmental factors, unsafe control behaviors that lead to system hazards are identified, and the causal scenarios that cause the unsafe control behaviors are determined to complete the risk assessment.

[0011] The collection of multiple individually controllable environmental factors specifically includes: cameras outside the hull collecting overall video of the hull, and calculating the ship's tilt angle and hull vibration based on the overall video of the hull; cameras inside the cabin collecting video of operations inside the cabin, and establishing a three-dimensional model of the equipment layout and identifying personnel positions and trajectories based on the video of operations inside the cabin; and measuring dust concentration through sensors deployed inside the cabin.

[0012] The constructed integrated controllable environmental factors include personnel-space safety posture. The construction method is as follows: based on the three-dimensional model of the device layout, multiple geometric volume regions are predefined. And associate a security level with each area. The first one identified based on the video of the operation inside the cabin. Real-time 3D position of an individual The personnel-space safety posture is determined by comparing it with the geometric volume region and by the following rules:

[0013] .

[0014] The constructed integrated controllable environmental factors include a dynamic dust cloud explosion risk index. The construction method includes: calculating the intensity of hull rolling based on the ship's tilt angle. :

[0015] in, for The intensity of the ship's rolling at any given moment. and They are respectively The ship's roll and pitch angular velocities at any given time;

[0016] Based on the aforementioned ship sway intensity and the dust concentration The dynamic dust cloud explosion risk index is calculated through preset fuzzy logic reasoning. .

[0017] The constructed integrated controllable environmental factors include the equipment corrosion-fatigue comprehensive health index. The construction method includes: calculating the real-time equivalent damage rate based on the ship's vibration, salt spray corrosion environment, and the cumulative operating time of the equipment. ; by measuring the equivalent damage rate over the cumulative running time Integrate within the range to obtain the total cumulative damage. The corrosion-fatigue comprehensive health index of the equipment was calculated. :

[0018] in, for The comprehensive health index of equipment corrosion-fatigue over time. This is a preset total damage threshold representing the failure of a component.

[0019] The constructed integrated controllable environmental factors include the work permit environmental safety window. The construction method is as follows: the job permit environment security window is determined by the logical AND operation result of the following four conditions. Status:

[0020] in, The status is either on or off; The attitude stability condition is determined based on the ship's tilt angle and a preset attitude threshold. Based on the dust concentration and preset concentration threshold The conditions for determining dust concentration; Normal ventilation conditions are determined based on the status of the ventilation system. The calculated personnel-space security situation The safety conditions of the personnel being assessed.

[0021] The ship's tilt angle is obtained by processing an image sequence of video footage of the entire ship. The calculation method further includes: calculating the instantaneous roll angle for each frame of the image sequence. The instantaneous roll angle is optimally estimated using a Kalman filter, and its state update formula is as follows:

[0022] in, for The optimal state estimate at time t. for The predicted state value at time 10:00. For Kalman gain, for The instantaneous roll angle observation at time t. This is the observation matrix.

[0023] The hull vibration is obtained by processing an image sequence of video of the entire hull, and the calculation method further includes: calculating a time-series signal representing the vertical vibration displacement of the hull based on the image sequence. The power spectrum is obtained by performing spectral analysis on the time series signal. The dominant vibration frequency is determined based on the power spectrum. and dominant vibration amplitude :

[0024] in, The dominant vibration frequency, Power spectrum, For video capture frame rate, This represents the number of signal points involved in the transformation.

[0025] The control structure model is established as follows: a hierarchical control structure including a safety monitoring and decision controller is established; the safety monitoring and decision controller is used to receive the individual controllable environmental factors as feedback and generate the integrated controllable environmental factors; the safety monitoring and decision controller is also used to generate a work permit signal based on the integrated controllable environmental factors and send the signal as a mandatory safety constraint to the powder conveying main controller.

[0026] This specification also proposes a risk assessment system for fracturing vessel powder transportation in conjunction with a controlled environment, the system comprising:

[0027] Data acquisition module: Collects multiple individual controllable environmental factors of the fracturing vessel powder conveying system;

[0028] Fusion Module: Based on the multiple individually controllable environmental factors, at least one fused controllable environmental factor is constructed; the fused controllable environmental factor is used to characterize the systemic security situation generated under the coupled effect of the multiple individually controllable environmental factors;

[0029] Control structure model construction module: The control structure model includes the controller, controlled process, control command and feedback loop of the powder conveying system, and the individual controllable environmental factors and the integrated controllable environmental factors are included as components of the feedback loop;

[0030] Identification module: Based on the control structure model, in the context defined by the individual controllable environmental factors and the integrated controllable environmental factors, identify unsafe control behaviors that lead to system hazards, determine the causal scenarios that cause the unsafe control behaviors to occur, and complete the risk assessment.

[0031] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the aforementioned risk assessment method for fracturing vessel powder transportation in conjunction with a controlled environment.

[0032] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned risk assessment method for fracturing vessel powder transportation in a controlled environment.

[0033] Compared with existing technologies, the present invention provides a risk assessment method and system for fracturing vessel powder transportation in a controlled environment, which has the following advantages:

[0034] This invention improves the application scope of System Theoretical Process Analysis (STPA) by introducing and systematizing the concept of a controllable environment, fundamentally solving the limitations of existing safety analysis methods when applied to specialized marine engineering equipment such as fracturing vessels. Traditional STPA methods treat the environment as an uncontrollable external factor, which, on a highly self-sufficient and manageable platform like a ship, overlooks numerous critical risks arising from the interaction between the environment and the system. This invention formally incorporates factors such as ship tilt and personnel activities—originally considered external disturbances—into internal variables of risk analysis. This transforms the entire assessment process from theoretical modeling detached from reality into a more precise and profound systemic hazard identification closely integrated with the application scenario, significantly enhancing the effectiveness and practical guidance of risk assessment.

[0035] This invention proposes and implements a higher-dimensional risk perception method based on multi-source information fusion. Existing technologies often monitor individual environmental parameters in isolation, failing to reveal emergent risks arising from the coupling of multiple factors. This invention, by constructing a dynamic dust cloud explosion risk index and an equipment corrosion-fatigue comprehensive health index, integrates controllable environmental factors to quantify and predict complex hazards with nonlinear characteristics that only manifest under the synergistic effect of multiple individual factors. For example, this invention can identify situations where the explosion risk increases sharply due to violent hull rolling at moderate dust concentrations. This predictive risk perception capability is unparalleled by traditional methods, transforming safety control from a passive, reactive response to a proactive, pre-emptive warning, significantly improving the inherent safety level of the system.

[0036] This invention utilizes video footage of the entire ship's hull captured by cameras deployed externally to the ship. Through time-series analysis of the sea-line, the real-time tilt angle of the ship is calculated. Furthermore, optical flow combined with spectral analysis is used to extract the vibration characteristics of the ship's structure caused by waves. Compared to traditional contact-based inertial sensors or accelerometers, this visual measurement method not only avoids problems such as sensor corrosion, damage, and accuracy drift in harsh marine environments like strong salt spray, high humidity, and strong vibration, significantly improving the long-term reliability and stability of the sensing, but also achieves a high degree of hardware integration. A single visual sensing unit can simultaneously perform multiple tasks such as attitude, vibration, and personnel monitoring, greatly simplifying the system hardware configuration and reducing the complexity and cost of installation and maintenance.

[0037] In summary, this invention deeply couples the perception and fusion of controllable environments with system safety analysis methods, constructing a complete, closed-loop risk assessment system from data acquisition and risk quantification to analytical decision-making. It not only identifies systemic and emergent risks that traditional methods cannot detect, but also ensures the reliability of assessment inputs through visual perception technology. The final risk assessment results (including unsafe control behaviors and causal scenarios) are closely linked to specific, measurable environmental contexts, making the formulation of safety countermeasures more targeted and effective, thereby comprehensively improving the overall safety level of fracturing vessel powder conveying systems in complex marine environments. Attached Figure Description

[0038] Figure 1 This is a flowchart of the risk assessment for powder transportation on fracturing vessels in a controlled environment, based on the present invention.

[0039] Specific implementation methods

[0040] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0041] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0042] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this application, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0043] Additionally, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that practice can be carried out without these specific details.

[0044] This specification presents an embodiment of a risk assessment method for fracturing vessel powder transportation in a controlled environment, the method comprising:

[0045] Multiple individual controllable environmental factors were collected from the powder conveying system of the fracturing vessel.

[0046] Based on the multiple individually controllable environmental factors, at least one integrated controllable environmental factor is constructed; the integrated controllable environmental factor is used to characterize the systemic security situation generated under the coupled effect of the multiple individually controllable environmental factors;

[0047] A control structure model is established, which includes the controller, controlled process, control command, and feedback loop of the powder conveying system, and the individual controllable environmental factors and the integrated controllable environmental factors are included as components of the feedback loop.

[0048] Based on the control structure model, in the context defined by the individual controllable environmental factors and the integrated controllable environmental factors, unsafe control behaviors that lead to system hazards are identified, and the causal scenarios that cause the unsafe control behaviors are determined to complete the risk assessment.

[0049] This embodiment details the acquisition of individually controllable environmental factors in the method of the present invention. In this embodiment, the core of data acquisition is the construction of an integrated visual perception unit, which is used to acquire raw video data for subsequent risk analysis.

[0050] The method of this invention is applied to a powder conveying system installed on a large marine engineering fracturing vessel. The operating environment of this system is significantly unique and complex. Fracturing vessels operate in open seas, enduring the combined effects of wind, waves, and currents, resulting in continuous, multi-degree-of-freedom rolling motion of the hull, i.e., real-time changes in the ship's heel angle, and low-frequency periodic vibrations in the hull structure. The powder conveying system is installed in a closed or semi-closed compartment inside the hull. This compartment has limited space, a compact equipment layout, and houses operating and maintenance personnel. The system inherently carries the risk of dust leakage during conveying, and the high salt spray from the marine atmosphere accelerates equipment corrosion. The visual perception unit described in this embodiment is designed for this complex and dynamic application scenario, providing accurate and reliable environmental condition data for subsequent risk assessment.

[0051] The preparatory work before the risk assessment involves deploying and running an integrated visual perception unit, which consists of two sets of cameras with different functions, used to collect video data of the exterior and interior of the ship respectively.

[0052] The overall hull video is specifically designed for the analysis and quantification of two dynamic and controllable environmental factors: ship tilt and hull vibration. To ensure long-term stable operation in harsh marine environments, two industrial-grade network cameras with an IP67 protection rating or higher were selected. The cameras must possess a wide dynamic range to cope with strong sunlight reflection and shadow variations on the sea surface; CMOS sensors with global shutters were chosen to avoid image distortion during rapid hull movement. To obtain a wide field of view, the cameras are equipped with 2.8mm wide-angle lenses. The two cameras were mounted on either side of the superstructure above the main deck of the fracturing vessel, on structurally stable surfaces with unobstructed views, ensuring their lenses face outwards. This deployment ensures that the cameras can capture a clear and continuous sea-sky boundary without obstruction under any navigational attitude, which is fundamental for subsequent visual analysis of ship tilt. The mounting points are located on the main load-bearing structure of the hull, effectively transmitting hull structural vibrations caused by waves and providing a reliable signal source for subsequent visual analysis of hull vibrations. Deploying cameras on both sides provides redundancy, ensuring that if one side's field of view is limited due to surges or equipment obstruction, the other side can still provide valid data. Both cameras are configured to continuously capture video at a frame rate of 50 frames per second (50fps) and a resolution of 1920x1080 pixels. The acquired raw video streams are compressed in H.265 video encoding format via an industrial-grade Ethernet switch and transmitted in real time to a video data server located in the ship's central control room. The server labels the two received video streams as port-side hull-wide video and starboard-side hull-wide video, respectively, and adds a precise timestamp provided by the ship's network time protocol server before storing them, forming a raw video database for subsequent analysis.

[0053] The video footage of operations inside the compartment is used for subsequent analysis and modeling of two controllable environmental factors: personnel location and trajectory, and equipment layout. Four wide-angle dome network cameras with infrared night vision and explosion-proof certification were selected. Wide-angle lenses were chosen to maximize monitoring coverage within the limited compartment space; infrared night vision ensures continuous monitoring in situations of insufficient lighting or emergency power outages; and explosion-proof certification addresses the potential dust explosion risk of the powder conveying system, ensuring the cameras themselves do not become ignition sources. The four cameras were installed on the ceiling of the compartment containing the powder conveying system, evenly distributed across the four quadrants, with the lenses pointing vertically downwards. This deployment aims to create comprehensive, overlapping coverage of the entire work area without blind spots. The following areas received special attention during deployment: equipment areas requiring close operation or observation, such as screw feeders and rotary valves; areas with dense flange connections on the main conveying pipeline and near maintenance manholes, as these are potential leak points and areas requiring personnel maintenance; and control panels and main personnel access routes.

[0054] This layout ensures that no matter where personnel are moving within the cabin, they can be clearly captured by at least one camera, providing data support for subsequent personnel location and trajectory analysis, while also fully recording the static layout information of all equipment within the cabin.

[0055] All four cameras are configured to continuously capture video at a frame rate of 30 frames per second (30fps) and a resolution of 1920x1080 pixels. The video streams are also transmitted to a video data server via industrial Ethernet. The server tags each of the four received video streams with its physical installation location (e.g., cabin-Area A-video, cabin-Area B-video, etc.) and adds a precise timestamp before storing them. This video data will serve as the basis for building a 3D digital model of the equipment layout and as the input source for real-time analysis of personnel dynamic positions and trajectories.

[0056] The following embodiment details the acquisition of other controllable environmental factors besides visual information. The data from these factors, together with the acquired video data, constitute the complete input to the risk assessment method of this invention.

[0057] Dust concentration is the most direct indicator for assessing the explosion risk inside a powder conveying system compartment. Real-time and accurate acquisition of dust concentration data at key locations inside the compartment is the foundation for subsequent dynamic dust cloud explosion risk index fusion analysis and the development of immediate safety control strategies.

[0058] To achieve high sensitivity and rapid response, this embodiment uses four online dust concentration sensors based on laser scattering. These sensors can accurately measure suspended dust concentrations as low as 1 mg / m³, with a response time of less than 1 second. Furthermore, their probes and circuitry are designed to be explosion-proof in accordance with ATEX or IECEx standards, ensuring safe operation in potentially explosive environments.

[0059] One unit is installed approximately 1.5 meters directly above the powder inlet of the screw feeder or rotary valve to monitor the source dust dispersion caused by material agitation. One unit is installed in the center of the area with the highest density of flanges, valves, and flexible connections on the main conveying pipeline to monitor minor leaks caused by seal aging or failure. One unit is installed near the maintenance platform of the dust collector to monitor potential leaks in the dust collection system itself or instantaneous high concentrations of dust generated during backflushing cleaning. One unit is installed at the front end of the return air vent of the main ventilation system in the compartment to measure the background dust concentration after mixing throughout the entire compartment space and assess the overall environmental safety level.

[0060] Four sensors continuously sample at a frequency of 1 Hz, outputting the measured dust concentration value (unit: mg / m³) as a standard current signal. This signal is connected to a data acquisition module (DAQ), and after analog-to-digital conversion, it is transmitted in digital form to a process data server located in the central control room via the Modbus TCP / IP industrial Ethernet protocol. The server adds a precise timestamp and sensor location tag to each concentration data point before storing it, forming a historical dust concentration database.

[0061] Salt spray in marine environments is a key accelerator of equipment structural corrosion and control system aging. Quantifying the real-time severity of salt spray corrosion is a necessary prerequisite for subsequent integrated analysis of equipment corrosion-fatigue health indices, enabling predictive maintenance and risk warning.

[0062] Two all-in-one corrosion environment sensors were selected. These sensors integrate a temperature and humidity sensor and a surface conductivity sensor based on an interdigital electrode array. They can simultaneously measure ambient temperature, relative humidity, and the conductivity of the liquid film formed after atmospheric salt and moisture deposits on the electrode surface, providing a key indicator that directly reflects the potential for real-time corrosion rates.

[0063] One unit is tightly mounted on the outer wall of a representative section of the main conveying pipeline exposed to the cabin air using a special clamp, for directly measuring the microenvironmental corrosion intensity faced by this core component. Another unit is installed next to the housing of the main PLC control cabinet of the powder conveying system to assess the corrosion risk of the environment in which the control system and key electronic components are located.

[0064] The sensors collect temperature, humidity, and surface conductivity data every minute and send the data packets to the process data server via industrial Ethernet. The server parses the data, adds timestamps and location tags, and stores it in a database, providing continuous input for the health index model.

[0065] The ventilation system is a primary engineering measure for controlling dust concentration within the cabin and preventing the formation of explosive atmospheres. To ensure the reliability of monitoring and the accuracy of diagnosis, a dual-sensor redundancy design is adopted: a thermal anemometer is used to directly measure the air velocity within the ventilation ducts. The motor current sensor uses a clamp-on current transformer based on the Hall effect principle to non-contactly measure the operating current of the motor driving the ventilation fan.

[0066] Install the probe of the thermal anemometer inside a straight section of the main exhaust duct of the cabin to ensure a stable average wind speed is measured. Install the clamp of the motor current sensor onto any one phase of the power supply cable of the three-phase asynchronous motor driving the exhaust fan.

[0067] Wind speed and current sensors perform continuous monitoring. The measured signals are also connected to the data acquisition module (DAQ). The data undergoes preliminary processing locally. For example, when the current value is within the rated range and the wind speed value is greater than the preset minimum safe wind speed, the system determines that the ventilation system is in normal condition; otherwise, it is determined to be faulty or shut down. This processed status information (normal / faulty / shutdown), along with the original current and wind speed values, is sent to the process data server, timestamped, and recorded.

[0068] The following embodiment details the processing and analysis of the overall ship video collected and stored on the video data server to calculate the ship's tilt angle, a key, individually controllable environmental factor. In this invention, the ship's tilt angle is defined by two core parameters: the roll angle, which describes the ship's lateral sway, and the pitch angle, which describes its longitudinal undulation. To overcome the randomness introduced by single-frame image analysis, this embodiment employs a time-window-based sequential image analysis method. Through data cleaning and optimal state estimation, it outputs stable and reliable tilt angle values.

[0069] S3.1 Batch processing and instantaneous tilt angle calculation of video frame sequences: Image data within a time window is obtained from a continuous video stream, and the instantaneous, unprocessed tilt angle value is calculated for each frame.

[0070] S3.1.1 Set a sliding time window In this embodiment, The value is 1 second. Based on the overall ship video capture frame rate (50fps), this time window contains the number of image frames. It is 50 frames.

[0071] S3.1.2 Continuously read the most recent data from the video data server. Frame images constitute an image sequence. .

[0072] S3.1.3 For each frame of the image in the sequence Independently executed video frame preprocessing, sea-line edge detection, and sea-line straight-line fitting. Canny edge detection and Hough transform are used for each frame image. Calculate an instantaneous sea-line parameter pair .

[0073] Based on this parameter pair, the instantaneous roll angle for each frame is calculated using the following formula. and instantaneous pitch angle :

[0074]

[0075] This step yields a raw, instantaneous tilt angle data sequence containing 50 samples within a time window: .

[0076] S3.2 Instantaneous tilt angle data cleaning and outlier removal

[0077] Identify and remove outliers (outliers) in the original tilt angle data sequence caused by transient disturbances (such as splashing waves obscuring the view, seabirds flying by, strong sunlight reflection, etc.) to ensure the reliability of the dataset used for the final calculation.

[0078] For the roll angle sequence respectively and pitch angle sequence Perform data cleaning.

[0079] S3.2.1 For the sequence Sort the data points and calculate their first quartile. and the third quartile .

[0080] S3.2.2 Calculate the interquartile range .

[0081] S3.2.3 Set the upper and lower limits of the valid data range: the upper limit is... The lower limit is .

[0082] S3.2.4 Traverse each data point in the sequence. If its value exceeds the range defined by the upper and lower limits above, it is identified as an outlier and removed.

[0083] After outlier removal, a sample size less than or equal to [number missing] is obtained. A clean, instantaneous tilt angle data sequence.

[0084] S3.3 Based on Kalman filtering, the optimal estimation of the tilt state is obtained, resulting in a smooth, stable, and optimal representation of the current time window. The inclination angle values ​​of the inner hull's true attitude are processed using the cleaned instantaneous inclination angle data sequence to estimate the state of the dynamic system.

[0085] Independent Kalman filters are established for the roll and pitch angles respectively. For example:

[0086] Define system state for The roll angle and its angular velocity at time:

[0087]

[0088] in, k The roll angular velocity at time k.

[0089] Assuming an extremely short frame interval Within 0.02 seconds (in this example), the hull angular acceleration is zero-mean Gaussian white noise. Prediction The state at any given moment:

[0090]

[0091]

[0092] in, for The predicted state value at any given time. The optimal state estimate at time t. Here is the state transition matrix. . Let be the covariance matrix of the prediction step. Let be the process noise covariance matrix, representing the uncertainty of the model prediction.

[0093] The instantaneous roll angle obtained after data cleaning As Observations at time Calculate the Kalman gain. : Update the optimal state estimate : Update the covariance matrix :

[0094] in, For the observation matrix, . The observation noise covariance matrix represents the uncertainty of a single frame measurement. It is an identity matrix.

[0095] After traversing the time window After collecting all clean data points within the timeframe, take the last time point. State estimation vector The first element in the equation serves as the roll angle for the final output of that time window. For pitch angle Perform the exact same filtering process to obtain the final pitch angle. .

[0096] This embodiment employs a computer vision-based algorithm to achieve non-contact measurement of ship tilt angle by identifying and analyzing the sea-sky boundary line (hereinafter referred to as the "sea-sky line") in video frames. The algorithm mainly includes four core steps: video frame preprocessing, sea-sky line edge detection, straight line fitting, and tilt angle parameter calculation.

[0097] The video frame preprocessing aims to eliminate the impact of marine environmental noise on image quality, enhance the sea-line features, and prepare for subsequent accurate identification.

[0098] Read a single frame of original color image with a resolution of 1920x1080 pixels from the video data server in timestamp order. .

[0099] Color image Convert to a single-channel grayscale image For grayscale images A Gaussian filter is applied to smooth out random noise caused by the camera sensor or sea fog. In this embodiment, a filter with a size of [missing value] is selected. The Gaussian kernel can effectively suppress noise while preserving the sharpness of the critical edge of the sea horizon to the maximum extent.

[0100] The sea-line edge detection extracts a set of pixels representing the sea-line, specifically:

[0101] The Canny edge detection algorithm is used to process the denoised grayscale image.

[0102] Calculate the gradient magnitude and direction for each pixel in the image. Apply non-maximum suppression to thin wide edges into single-pixel width edges. Use a double thresholding method (high threshold...) and low threshold Distinguish between strong and weak edges. In this embodiment, Set to 100. Set to 50. Through hysteresis connection, weak edges connected to strong edges are preserved, ultimately outputting a binarized edge image. This includes only candidate pixels identified as the sea-line area.

[0103] The sea-line fitting will adjust the edge image. The discrete edge pixels are fitted into a straight line that can describe the position and angle of the sea surface.

[0104] Hough transform is used for line detection. For edge images... Each edge point in In Hough parameter space In this process, the parameters of all possible lines passing through the point are calculated, and the corresponding accumulator units are voted on. The polar equation of the line is expressed as: in, This represents the perpendicular distance from the origin of the image coordinate system to the line. This represents the angle between the vertical line and the x-axis. After traversing all edge points, find the peak point with the highest number of votes in the Hough accumulator. The coordinates of this peak point are... That is, the parameter of the straight line that best represents the horizon.

[0105] Next, this embodiment details the processing and analysis of the overall ship hull video collected and stored on the video data server to calculate the key, individually controllable environmental factor of hull vibration. In the application scenario of this invention, hull vibration specifically refers to the low-to-medium frequency, periodic displacement of the hull structure caused by wave impact and ship navigation.

[0106] This embodiment employs a visual measurement method based on dense optical flow and time-frequency domain analysis. By calculating the motion patterns (i.e., optical flow fields) of global pixels between consecutive video frames, and after compensating for the macroscopic motion of the field of view caused by the ship's tilt angle, the periodic displacement signal caused by hull vibration is separated, and finally, its main vibration characteristics are extracted through spectral analysis. The hull vibration calculation mainly includes three steps: dense optical flow field calculation, global motion vector extraction and tilt angle effect compensation, and time-frequency domain analysis of the vibration signal.

[0107] S4.1 Calculation of Dense Optical Flow Field

[0108] S4.1.1 uses the same sliding time window (1 second, including) (Frame image).

[0109] S4.1.2 For each pair of consecutive, pre-processed (grayscale, Gaussian filtering) image frames within the time window. The Gunnar Farnebäck dense optical flow algorithm was used for calculation.

[0110] The output of S4.1.3 is Each optical flow field. For each frame (From 2 to ), and its corresponding optical flow field It is a two-dimensional vector matrix, where each vector Represents pixels From the Frame to the Horizontal and vertical displacement of the frame.

[0111] S4.2 Global motion vector extraction and tilt effect compensation: Since the original optical flow contains the superposition effect of three motions: ship translation, ship tilt angle change and hull vibration, in order to separate the net displacement signal caused by the hull structure vibration from the complex panoramic optical flow, the first two must be compensated.

[0112] S4.2.1 Global Motion Vector Extraction: To eliminate interference from local irrelevant motions (such as ocean wave fluctuations), global motion vector extraction is performed on each optical flow field. Calculate its global average motion vector This serves as a representation of the overall motion of the ship's hull within that frame.

[0113] S4.2.2 Heel Effect Compensation: The roll and pitch motions of the hull are the main contributors to the overall motion. Using the calculated high-precision heel angle time series... Calculate the change in tilt angle and The theoretical pixel displacement caused by this. In this embodiment, we mainly focus on the vertical displacement caused by pitch (bow tilt). The analysis focuses on the vertical vibration (heave motion) caused by the roll, and the rotational effect caused by the roll. Therefore, compensation is applied to the global vertical motion vector.

[0114] in, For the first The net pixel displacement between frames caused by the vertical vibration (heave) of the ship's hull. For the first Global average vertical displacement measured between frames. To adapt to changes in tilt angle The theoretical global average vertical displacement caused by panning, calculated from the camera model.

[0115] After global motion vector extraction and tilt effect compensation, a time series signal representing the vertical vibration displacement of the hull within a time window is obtained. .

[0116] S4.3 Time-frequency domain analysis of vibration signals, from vibration displacement time series signals containing noise. In the process, the characteristic parameters of hull vibration are extracted: dominant vibration frequency and dominant vibration amplitude.

[0117] The Fast Fourier Transform (FFT) is used to perform spectral analysis on the signal.

[0118] S4.3.1 For time series signals The Hanning window function is applied to reduce the energy leakage effect in spectral analysis and improve frequency resolution.

[0119] S4.3.2 Perform an FFT on the windowed signal to transform it from the time domain to the frequency domain, obtaining a complex spectrum sequence. :

[0120] in The first frequency domain signal One portion, The number of signal points participating in the transformation ( ). It is a time-domain vibration displacement signal. These are the coefficients of the Hanning window function. This is the index for the frequency domain.

[0121] S4.3.3 Calculate the square of the modulus of the complex spectrum to obtain the power spectral density (PSD). It represents the energy distribution of a signal at different frequencies.

[0122] S4.3.4 Considering that the ship's vibration is mainly in the low to medium frequency range, the power spectrum is searched within the preset frequency range (preferably 0.1Hz to 2Hz). Peak value:

[0123] Dominant vibration frequency Calculation:

[0124] Dominant vibration amplitude Calculation:

[0125] in, The dominant frequency of the ship's hull vibration is measured in Hertz (Hz). The dominant amplitude of the ship's vibration is expressed in millimeters (mm). This is the video capture frame rate, i.e., the sampling frequency (50Hz in this example). The calibration coefficient is used to convert the amplitude units of the power spectrum from pixel space to the actual physical displacement (millimeters). This coefficient is obtained through a one-time calibration.

[0126] Next, this embodiment details the processing and analysis of the cabin operation videos collected and stored on the video data server to establish a static three-dimensional model of the equipment layout, and based on this model, to track the position and trajectory of all personnel working inside the cabin in real time and accurately.

[0127] This embodiment is divided into two phases. The first phase is a one-time equipment layout modeling phase, which generates a digital twin model of the cabin containing semantic safety information through multi-view 3D reconstruction technology. The second phase is a continuous real-time personnel tracking phase, which uses deep learning target detection, attitude estimation, and multi-view fusion technology to obtain the precise position of personnel in the 3D model in real time and generate their movement trajectory.

[0128] S5.1 Equipment Layout Modeling: This invention creates a static three-dimensional digital model that not only describes the physical location of each piece of equipment in the powder conveying system within the compartment, but more importantly, it pre-divides and marks different levels of safety zones in the model.

[0129] S5.1.1 After the initial deployment of the system or a major change in the equipment layout, a one-time joint calibration of the four installed in-cabin operation cameras is performed. The intrinsic parameter matrix of each camera is calculated using Zhang Zhengyou's calibration method. (Including focal length, principal point coordinates) and extrinsic parameter matrix (Describe the position and attitude of the camera in a unified ship coordinate system).

[0130] S5.1.2 Using calibrated multiple cameras, a three-dimensional reconstruction of the entire cabin is performed using Structure for Motion Restoration (SfM). By automatically extracting and matching static feature points (such as equipment corners and ground markings) from multi-view images, and combining them with camera extrinsic parameters, the three-dimensional spatial coordinates of these feature points are calculated using triangulation, ultimately generating a high-density three-dimensional point cloud model of the cabin.

[0131] S5.1.3 Based on the generated 3D point cloud model and combined with the equipment's CAD design drawings, semantic segmentation of the space is performed to define geometric volume regions with clear safety attributes. Each area They are all associated with a security level. .

[0132] For example, the area below the main delivery pipeline that is prone to dust leakage is defined as... Its security level is The cylindrical space within a 1-meter radius around the rotating components of the screw feeder is defined as... Its security level is The main passageway marked with yellow lines on the ground is defined as... Its security level is .

[0133] S5.1.4 Store the complete 3D model containing all safety zone definitions in the process data server.

[0134] S5.2 Real-time tracking of personnel position and trajectory: In a continuously running video stream, all personnel are detected in real time, and their three-dimensional coordinates in the equipment layout model are accurately calculated to form a continuous motion trajectory.

[0135] S5.2.1 Multi-view human body detection:

[0136] For each frame of image from four cameras, the data is input in parallel into a pre-trained deep learning object detection model (in this embodiment, the YOLOv8 model is used). The model identifies all human targets in the image and outputs the bounding box of each target's location. ,in For personnel indexing, Index the camera.

[0137] S5.2.2 Human Body Key Point Location:

[0138] To improve positioning accuracy, for each detected bounding box The image within the camera feed is input into a human pose estimation model (preferably Mediapipe Pose). The output is the two-dimensional pixel coordinates of each joint point on the human body. To stably represent the person's position on the ground, this embodiment selects the midpoint of the left and right ankle key points as the position of the person in front of the camera. 2D position anchor point in the view .

[0139] S5.2.3 Multi-view data fusion and 3D positioning:

[0140] Because of the same person It may be captured by multiple cameras simultaneously, and the system needs to use feature matching to separate the images from different camera views. Associated with the same physical person. For individuals successfully associated with at least two cameras. Using its two-dimensional anchor point coordinates in different views Using pre-calibrated camera intrinsic and extrinsic parameters, the optimal position of the personnel's anchor point in the ship's three-dimensional coordinate system is calculated using the least squares method. .

[0141] S5.2.4 Trajectory Generation and Smoothing Filtering: To suppress measurement noise and transient target loss that may occur during visual positioning, for each tracked person... Maintain an independent Kalman filter.

[0142] Define the status of personnel Includes its three-dimensional position and velocity: The latest calculated 3D position The Kalman filter for this person is updated based on the observed values. The filter outputs a smoothed position estimate. And velocity estimates. The continuous smoothed position points When connected together, they constitute the personnel. Real-time motion trajectory .

[0143] Traditional risk assessments often view individual environmental factors in isolation, for example, setting separate alarm thresholds for dust concentration and ship tilt angle. The limitation of this approach is its inability to identify complex risks arising from the interaction and synergistic effects of multiple factors. This invention elevates safety analysis from single-variable threshold judgments to a multi-dimensional, systemic situational awareness level by constructing four integrated controllable environmental factors.

[0144] Integration Factor 1: Personnel-Space Security Situation Within the enclosed compartment housing the powder conveying system, the primary risks to personnel are accidental contact with operating equipment or exposure to high-concentration dust leaks. Therefore, the risk level of personnel is not determined by their absolute coordinates, but rather by their position relative to a predefined hazard source (i.e., equipment layout). This fusion factor aims to transform the real-time location of personnel into a discrete state with a clearly defined safety level, providing a direct and clear basis for subsequent control decisions.

[0145] Personnel-space security situation The input is the smoothed 3D position estimate of the person. and equipment layout model containing multiple geometric volume regions with labeled safety levels and their corresponding security levels

[0146] .

[0147] Geometric attribution logic is used. For the first [unclear] within the cabin... The real-time personnel-space security situation of each identified person. Determined by the following rules:

[0148]

[0149] in: For the first Individuals The security situation at any given moment is categorized as safe, alert, or dangerous. Represents geometric judgment, i.e., personnel Location Does it fall into a predefined spatial region? Within.

[0150] If a person is not within any defined warning or danger zone, their default state is safe. This determination is performed in parallel and in real time for all identified personnel.

[0151] Factor Two: Dynamic Dust Cloud Explosion Risk Index Dust explosion risk is highly nonlinear, depending not only on the static concentration of dust but also closely related to environmental disturbances. The violent rolling of a fracturing vessel is the primary source of disturbance within the compartments, capable of re-erecting dust deposited on equipment and pipe surfaces, causing a sudden surge in localized concentration. The input for fusion factor two is the maximum real-time reading from multiple dust concentration sensors. and roll angle and pitch angle and its angular velocity .

[0152] Fuzzy logic reasoning is used for modeling.

[0153] dust concentration The membership is fuzzified into three membership functions: {low, medium, high}. The intensity of the ship's roll is calculated and characterized by the magnitude of the total angular velocity. It is then fuzzified into three membership functions: {stationary, moderate, drastic}.

[0154] Establish expert rules that reflect synergistic effects:

[0155] Rule 1: IF (dust concentration is low) AND (swaying intensity is stable) THEN (risk index is very low)

[0156] Rule 2: IF (dust concentration is high) THEN (risk index is very high) (Regardless of the swaying, high concentration) The degree itself is high-risk.

[0157] Rule 3: IF (dust concentration is medium) AND (swaying intensity is severe) THEN (risk index is very high) (Core synergy rule: Under drastic fluctuations, the risk level of moderate concentrations is amplified to the highest level)

[0158] Rule 4: IF (dust concentration is low) AND (swaying intensity is severe) THEN (risk level is moderate) (Even with low background concentration, vigorous shaking can cause localized dust accumulation to be stirred up, creating a moderate risk.)

[0159] By using Mamdani fuzzy inference and the centroid method for defuzzification, a dynamic dust cloud explosion risk index in the range [0, 100] is output. This index can be further mapped to discrete risk levels of {very low, low, medium, high, very high}.

[0160] Factor Three: Equipment Corrosion-Fatigue Comprehensive Health Index The pipes, flanges, and support structures of the powder conveying system are exposed to a high salt spray environment for extended periods, while also being subjected to periodic stresses from ship vibrations. These two factors produce a synergistic destructive effect of "corrosion fatigue," where vibration stress accelerates the cracking of salt spray corrosion points, and corrosion further exacerbates stress concentration, thereby significantly shortening the equipment's lifespan.

[0161] The input for the equipment corrosion-fatigue integrated health index is the corrosive environment level. (Assessed comprehensively by temperature, humidity, and electrical conductivity, quantified as a value in the range [0,1]) and the output dominant amplitude of hull vibration. and dominant frequency and the cumulative operating time of components .

[0162] Define a real-time equivalent damage rate :

[0163]

[0164] in, These are the weighting coefficients for salt spray, vibration, and their interaction, respectively.

[0165] Total cumulative damage The integral of damage rate over runtime:

[0166]

[0167] Equipment Corrosion-Fatigue Comprehensive Health Index Defined as:

[0168]

[0169] in, It is a preset total damage threshold representing the failure of a component. Starting from 100%, it gradually decreases due to operational and environmental factors.

[0170] Factor Four: Work Permit Environmental Safety Window Before carrying out high-risk powder conveying operations, it is essential to ensure that all critical environmental prerequisites are met. This fusion factor integrates the most critical real-time environmental conditions to form a veto-based work permit decision, providing an absolute safety interlock signal for system start-up and shutdown control.

[0171] The input for the work permit environmental safety window is the pan angle. and pitch angle Readings from all dust concentration sensors Ventilation system status (Ventilation_Status(t)) and calculated personnel-space safety status for all personnel. .

[0172] Boolean logic AND gates are used for the judgment. Work Permit Environmental Safety Window The state is determined by the result of the logical AND operation of the following four conditions:

[0173]

[0174] in, The output is {On, Off}, with the following four conditions:

[0175] (Attitude stability conditions):

[0176] (Dust concentration conditions):

[0177] (Under normal ventilation conditions):

[0178] (Personnel safety conditions):

[0179] Only when When all four conditions are true, The device is considered "on" only if it is in a certain state; otherwise, it is immediately turned off. All of these are preset safety thresholds.

[0180] Next, this embodiment details the establishment of a control structure for STPA analysis that can reflect the influence of individual and integrated controllable environmental factors. The fundamental difference between this control structure model and the traditional STPA model is that it no longer treats the environment as an external, uncontrollable disturbance, but rather regards the controllable environment as an internal, core source of feedback and the basis for control decisions, thereby enabling a more profound analysis of the risks arising from the interaction between the environment and the system.

[0181] In the application scenario of this invention, the safe operation of the powder conveying system depends not only on the state of its internal components (motors, valves, controllers), but also to a large extent on the external dynamic environment. Traditional control structures can only describe the instructions and feedback between the operator, controller, and actuator, but cannot reflect the systemic risks induced by the environment, such as a normally safe control instruction becoming dangerous when the ship is violently rocking.

[0182] Therefore, this embodiment constructs a novel hierarchical control structure containing a safety monitoring loop. All environmental perception elements are used as basic feedback and centrally input into a dedicated safety decision-making unit. The fusion algorithm executed by the safety decision-making unit extracts intuitive, high-level safety situation information from the raw, multi-dimensional environmental data. As a safety supervisor, the safety decision-making unit outputs safety situation information as higher-priority control constraints or decision-making basis, simultaneously providing it to the top-level operator and the bottom-level powder conveying main controller, forming a closed-loop safety monitoring and control system.

[0183] The control structure constructed in this embodiment mainly includes the following three levels of controllers and their interaction relationships:

[0184] Top-level controller: The operator is the ultimate decision-maker for the operation, responsible for issuing high-level commands such as start, stop, and parameter adjustment. They send intent commands, such as requesting system startup or setting the conveying rate, to the main powder conveying controller and the safety monitoring and decision controller. Etc. The understanding of the system state upon which operators rely to make safety decisions.

[0185] In the system of this invention, four integrated and controllable environmental factors, provided by the fusion analysis and decision-making layer and visually displayed on the HMI, are incorporated:

[0186] Work Permit Environmental Safety Window Status: {On, Off}

[0187] Dynamic dust cloud explosion risk index Level: {Very low, ..., Very high}

[0188] Equipment Corrosion-Fatigue Comprehensive Health Index The percentage value.

[0189] Real-time personnel-space security situation for all personnel .

[0190] The system receives integrated environmental information and alarm information from the safety monitoring and decision controller, as well as equipment status information from the powder conveying main controller, through a human-machine interface (HMI).

[0191] Mid-level controller: Security monitoring and decision controller, acting as an independent security oversight unit. It does not directly control physical devices, but rather monitors and adjudicates the operational permissions and security status of the entire system by processing environmental information.

[0192] Its control commands are meta-commands or status broadcasts directed to other controllers:

[0193] Send a high-priority, Boolean-type work permission signal to the powder conveying main controller. This signal is directly from the work permit environmental safety window. The state determines it.

[0194] The real-time status values ​​of all four fusion factors are broadcast to the operator (via HMI) and the process data server (for logging).

[0195] The process model of this controller is a collection of the entire system environment states, and its state vector is... It can be represented as:

[0196]

[0197] The middle-level controller receives real-time data streams from all individually controllable environmental factors from the environmental perception layer as input for its fusion analysis.

[0198] The low-level controller: The main controller for powder conveying is a traditional PLC or DCS system, responsible for executing the specific physical equipment control logic. It outputs low-level instructions to the physical equipment, such as opening valves. Setting the motor Rotation speed, etc. Its internal control logic incorporates mandatory safety preconditions.

[0199] Its startup procedure logic must follow: IF (Operator_Command == 'Start') AND (G_permit(t) == 'Permitted') THEN { / / Execute the startup sequence}

[0200] This means that the work permit signal comes from the safety monitoring and decision control system. It has become a necessary but not sufficient condition for all dangerous operations, and has the right to veto operator instructions.

[0201] The underlying controller receives sensor signals from physical devices (such as motor current and valve position switches) and operation permission signals from the safety monitoring and decision controller. .

[0202] Physical layer: Controlled processes and environmental perception layer

[0203] The controlled process is the physical system for conveying powder materials, including all hardware equipment such as pipes, valves, motors, and fans.

[0204] The environmental perception layer, as the foundation of the entire control structure, includes all deployed sensors (integrated visual perception units, dust sensors, corrosion sensors, etc.), providing continuous and comprehensive environmental status feedback for safety monitoring and decision-making controllers.

[0205] Next, this embodiment details the identification of unsafe control behaviors (UCAs) based on the constructed control structure that deeply integrates controllable environmental factors, and further traces the specific causal scenarios that lead to these UCAs. This process is what distinguishes this invention from traditional STPA, enabling it to reveal the emergent risks arising from the interaction between the environment and the system.

[0206] Unsafe control actions (UCA) in a controlled environment are analyzed by focusing on critical control instructions within the control structure. Each control instruction is examined in context, incorporating controllable environmental factors, to determine its potential unsafety based on four criteria: providing a hazardous control action, failing to provide a hazardous control action, providing a control action at an inappropriate time or in the wrong sequence, or a control action that is too long or too short, leading to a hazard.

[0207] In this embodiment, two of the most representative control commands at different levels are selected for analysis. Analysis object 1: The operator's request to start the system command.

[0208] Context: Job Permit Environment Safety Window The state.

[0209] UCA-1 [Provides control behavior that leads to danger]: The operator, in the context of the job permission environment safety window W(t) being closed, provides the instruction "Request to start the system".

[0210] This UCA directly violates the highest level of safety interlock set by the system of this invention. W(t) being closed means that there are at least one or more clearly defined hazardous conditions (such as personnel entering a danger zone, ship instability, excessive dust concentration, or ventilation system failure). Activating the system under these conditions will directly lead to serious accidents such as personal injury, equipment damage, or dust explosion.

[0211] Analysis Object 2: Work permit signals issued by the safety monitoring and decision controller.

[0212] Context: Job license environment security window The internal calculation state.

[0213] UCA-2 [No control behavior provided that leads to danger]: The safety monitoring and decision controller does not provide a work permission signal with the value Permitted to the powder conveying main controller in the context that the internal state of the work permission environment safety window W(t) is open.

[0214] While this UCA does not directly pose a physical hazard, it constitutes a safety function failure. It prevents the system from starting normally even under safe environmental conditions, potentially causing production interruptions. More dangerously, operators might misinterpret this as a system malfunction and attempt to operate the system manually, bypassing safety systems, thus introducing new and unknown risks.

[0215] UCA-3 [Provides control behaviors that lead to hazards]: Safety monitoring and decision controller, provides a work permission signal with a value of Permitted in the context that the internal state of the work permission environment safety window W(t) is closed.

[0216] This UCA is a fatal failure at the core of the safety monitoring system of this invention. It means that the safety brain has made an incorrect judgment and wrongly authorized the system to operate in a clearly hazardous environment. The consequences are the same as UCA-1, which will directly lead to a serious accident.

[0217] Identify the causative scenarios within a controllable environment, and for each UCA identified in the previous step, delve into its root cause. The analysis path will cover the entire control structure, focusing on potential defects in the controller (including its internal algorithm), process model (state perception), and feedback loop (including environmental perception and data transmission).

[0218] Example of causal scenario analysis: UCA-1 and UCA-3 are selected for in-depth causal scenario analysis.

[0219] For UCA-1: The operator requests to start the system even when W(t) is off.

[0220] Cause Scenario 1.1 [Operator Process Model Defect - HMI Display Error]: Internal State of Safety Monitoring and Decision Controller The system is indeed off, but an error occurred during communication when the data transmitted to the HMI, or the HMI software itself has a defect, causing the interface to still display a green "system ready, start-up allowed" status. Based on this erroneous feedback, the operator believes the system is safe and therefore provides the start command. The root cause is that the operator's process model is inconsistent with the actual system state, and this inconsistency is caused by the erroneous feedback.

[0221] Cause Scenario 1.2 [Operator Control Algorithm Defect - Violation of Operation]: The HMI correctly displays W(t) as off and issues an audible and visual alarm. However, due to production schedule pressure, the operator, based on a contingency plan or personal judgment, uses the system's emergency maintenance / forced start mode. This mode's design bypasses the safety monitoring and decision-making controller. Signal interlocking. The root cause lies in the existence of a higher-priority control path in the system design that can bypass the core security logic, and the failure to effectively constrain its use cases in operating procedures or management.

[0222] Regarding UCA-3: The safety controller incorrectly issued a Permitted signal when W(t) was off.

[0223] Cause Scenario 3.1 [Feedback Loop Defect - Sensor Data Error]: The actual physical environment is already in a dangerous state (e.g., a person has entered a dangerous area). Therefore, the condition for W(t) The actual value is false; W(t) should be off. However, the integrated visual perception unit failed to detect the person because the camera was obstructed by dust, thus sending the erroneous message ∀i, Si(t) ≠ danger to the safety monitoring and decision controller. The root cause lies in the incorrect feedback information from the environmental perception layer, leading to a disconnect between its upstream fusion decision and physical reality.

[0224] Cause Scenario 3.2 [Controller Process Model Defect - Data Delay]: The ship encounters a rapidly developing swell with a roll angle of... It rapidly exceeded the safety threshold within 0.5 seconds. The integrated visual perception unit correctly captured and calculated the change at 50fps. However, due to momentary overload of computing resources in the fusion analysis and decision layer or congestion of the internal data bus, its computational resources for calculating W(t) were limited. The value is outdated data from one second ago, still within a safe range. Based on this stale process model, the controller incorrectly determined that W(t) is enabled. The root cause is that the controller's process model is not updated in a timely manner and fails to synchronize with the rapidly changing physical environment.

[0225] Cause scenario 3.3 [Controller control algorithm defect - fusion logic error]: All environmental perception data are accurate, and the actual situation is a ventilation system malfunction. (If false), W(t) should be closed. However, in the work permit environment safety window... During the programming implementation of the fusion algorithm, a code error caused the omission of condition C3, resulting in the actual executed logic being W(t) = C1 ∧ C2 ∧ C4. Therefore, even though the ventilation system reported a fault state, the incorrect algorithm logic still calculated W(t) as "on". The root cause lies in the design or implementation flaws of the fusion algorithm itself, which is the core of the safety decision-making process.

[0226] Through the above steps, the present invention can systematically identify deep-seated risks caused by defects in environmental perception, data fusion, human-computer interaction, and other aspects that are easily overlooked by traditional methods, providing extremely specific and profound guidance for the safe design and verification of the system.

[0227] This specification also proposes an embodiment of a risk assessment system for fracturing vessel powder transportation combined with a controllable environment, the system comprising:

[0228] Data acquisition module: Collects multiple individual controllable environmental factors of the fracturing vessel powder conveying system;

[0229] Fusion Module: Based on the multiple individually controllable environmental factors, at least one fused controllable environmental factor is constructed; the fused controllable environmental factor is used to characterize the systemic security situation generated under the coupled effect of the multiple individually controllable environmental factors;

[0230] Control structure model construction module: The control structure model includes the controller, controlled process, control command and feedback loop of the powder conveying system, and the individual controllable environmental factors and the integrated controllable environmental factors are included as components of the feedback loop;

[0231] Identification module: Based on the control structure model, in the context defined by the individual controllable environmental factors and the integrated controllable environmental factors, identify unsafe control behaviors that lead to system hazards, determine the causal scenarios that cause the unsafe control behaviors to occur, and complete the risk assessment.

[0232] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the aforementioned risk assessment method for fracturing vessel powder transportation in conjunction with a controlled environment.

[0233] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned risk assessment method for fracturing vessel powder transportation in a controlled environment.

[0234] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0235] In this specification, the same or similar parts between the various embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the descriptions of the embodiments described later are relatively simple, and relevant parts can be referred to the descriptions of the foregoing embodiments.

[0236] The above description is merely a specific embodiment of this application, but the scope of protection of this application 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 this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A risk assessment method for fracturing vessel powder transportation combined with a controlled environment, characterized in that: Multiple individual controllable environmental factors were collected from the powder conveying system of the fracturing vessel. Based on the multiple individually controllable environmental factors, at least one integrated controllable environmental factor is constructed; the integrated controllable environmental factor is used to characterize the systemic security situation generated under the coupled effect of the multiple individually controllable environmental factors; A control structure model is established, which includes the controller, controlled process, control command, and feedback loop of the powder conveying system, and the individual controllable environmental factors and the integrated controllable environmental factors are included as components of the feedback loop. Based on the control structure model, in the context defined by the individual controllable environmental factors and the integrated controllable environmental factors, unsafe control behaviors that lead to system hazards are identified, and the causal scenarios that cause the unsafe control behaviors are determined to complete the risk assessment. The collection of multiple individually controllable environmental factors specifically includes: cameras outside the hull collecting overall video of the hull, and calculating the ship's tilt angle and hull vibration based on the overall video of the hull; Cameras inside the cabin capture video of operations inside the cabin, and based on the video, a 3D model of the equipment layout is built, and the positions and trajectories of personnel are identified; dust concentration is measured by sensors deployed inside the cabin. The constructed integrated controllable environmental factors include personnel-space safety posture. The construction method is as follows: based on the three-dimensional model of the device layout, multiple geometric volume regions are predefined. And associate a security level with each area. The first one identified based on the video of the operation inside the cabin. Real-time 3D position of an individual The personnel-space safety posture is determined by comparing it with the geometric volume region and by the following rules: 。 2. The method for risk assessment of fracturing vessel powder transportation in a controlled environment according to claim 1, characterized in that, The constructed integrated controllable environmental factors include a dynamic dust cloud explosion risk index. The construction method includes: calculating the intensity of hull rolling based on the ship's tilt angle. : in, for The intensity of the ship's rolling at any given moment. and They are respectively The ship's roll and pitch angular velocities at any given time; Based on the aforementioned ship swaying intensity and the dust concentration The dynamic dust cloud explosion risk index is calculated through preset fuzzy logic reasoning. .

3. The method for risk assessment of fracturing vessel powder transportation combined with a controlled environment as described in claim 2, characterized in that, The constructed integrated controllable environmental factors include the equipment corrosion-fatigue comprehensive health index. The construction method includes: calculating the real-time equivalent damage rate based on the ship's vibration, salt spray corrosion environment, and the cumulative operating time of the equipment. ; by measuring the equivalent damage rate over the cumulative running time Integrate within the range to obtain the total cumulative damage. The corrosion-fatigue comprehensive health index of the equipment was calculated. : in, for The comprehensive health index of equipment corrosion-fatigue over time. This is a preset total damage threshold representing the failure of a component.

4. The risk assessment method for fracturing vessel powder transportation combined with a controlled environment as described in claim 3, characterized in that, The constructed integrated controllable environmental factors include the work permit environmental safety window. The construction method is as follows: the job permit environment security window is determined by the logical AND operation result of the following four conditions. Status: in, The status is either on or off; The attitude stability condition is determined based on the ship's tilt angle and a preset attitude threshold. Based on the dust concentration and preset concentration threshold The conditions for determining dust concentration; Normal ventilation conditions are determined based on the status of the ventilation system. The calculated personnel-space security situation The safety conditions of the personnel being assessed.

5. The risk assessment method for fracturing vessel powder transportation combined with a controlled environment as described in claim 4, characterized in that, The ship's tilt angle is obtained by processing an image sequence of video footage of the entire ship. The calculation method further includes: calculating the instantaneous roll angle for each frame of the image sequence. The instantaneous roll angle is optimally estimated using a Kalman filter, and its state update formula is as follows: in, for The optimal state estimate at time t. for The predicted state value at time 10:

00. For Kalman gain, for The instantaneous roll angle observation at time t. This is the observation matrix.

6. The risk assessment method for fracturing vessel powder transportation combined with a controlled environment as described in claim 5, characterized in that, The hull vibration is obtained by processing an image sequence of video of the entire hull, and the calculation method further includes: calculating a time-series signal representing the vertical vibration displacement of the hull based on the image sequence. The power spectrum is obtained by performing spectral analysis on the time series signal. The dominant vibration frequency is determined based on the power spectrum. and dominant vibration amplitude : in, The dominant vibration frequency, Power spectrum, For video capture frame rate, This represents the number of signal points involved in the transformation.

7. The method for risk assessment of fracturing vessel powder transportation in conjunction with a controlled environment as described in claim 6, characterized in that, The control structure model is established as follows: a hierarchical control structure including a safety monitoring and decision controller is established; the safety monitoring and decision controller is used to receive the individual controllable environmental factors as feedback and generate the integrated controllable environmental factors; the safety monitoring and decision controller is also used to generate a work permit signal based on the integrated controllable environmental factors and send the signal as a mandatory safety constraint to the powder conveying main controller.

8. A risk assessment system for fracturing vessel powder transportation in conjunction with a controlled environment, used to execute the risk assessment method for fracturing vessel powder transportation in conjunction with a controlled environment as described in claim 1, characterized in that, The system includes: Data acquisition module: Collects multiple individual controllable environmental factors of the fracturing vessel powder conveying system; Fusion Module: Based on the multiple individually controllable environmental factors, at least one fused controllable environmental factor is constructed; the fused controllable environmental factor is used to characterize the systemic security situation generated under the coupled effect of the multiple individually controllable environmental factors; Control structure model construction module: The control structure model includes the controller, controlled process, control command and feedback loop of the powder conveying system, and the individual controllable environmental factors and the integrated controllable environmental factors are included as components of the feedback loop; Identification module: Based on the control structure model, in the context defined by the individual controllable environmental factors and the integrated controllable environmental factors, identify unsafe control behaviors that lead to system hazards, determine the causal scenarios that cause the unsafe control behaviors to occur, and complete the risk assessment.

Images