Intelligent ship bulwark drainage management system combined with multi-sensor data monitoring

Abstract

The invention discloses a ship bulwark drainage intelligent management system combined with multi-sensor data monitoring, and relates to the technical field of ship intelligent control and automation, and the ship bulwark drainage intelligent management system comprises a sensor group, a dynamic trust chain fusion decision module, a multi-modal cross validation module, a fault-tolerant execution and reverse verification module and a central controller. According to the ship bulwark drainage intelligent management system combined with multi-sensor data monitoring, by constructing a dynamic trust chain fusion decision model and a multi-modal cross validation mechanism, the accuracy and timeliness of multi-source sensor data fusion under complex sea conditions are improved; sensor misjudgment caused by mechanical vibration, bubble interference and the like is effectively suppressed, the false triggering rate of a drainage instruction is reduced and response delay is shortened in combination with digital twin rehearsal and a reverse verification closed loop, and cooperative fault tolerance of a sensor cluster is realized through two-stage screening and a majority voting mechanism. And continuous and stable operation of the system under single-point fault or abnormal interference is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of ship intelligent control and automation, and in particular to a ship bulwark drainage intelligent management system combined with multi-sensor data monitoring. Background Art

[0002] In the technological development of ship bulwark drainage systems, traditional management solutions generally rely on fixed thresholds or single sensor signals to trigger drainage. Due to the complexity of a ship's navigational environment, particularly when encountering wind and waves, sudden turns, or sudden load changes, the drastic changes in the ship's hull attitude can lead to highly dynamic water accumulation in the bulwark area. For example, a liquid level sensor may generate a false high-water-level signal due to the ship's tilt, while a flow rate sensor may experience data jumps due to eddy current interference in the drain pipe.

[0003] Although existing technologies attempt to improve monitoring comprehensiveness by increasing the number of sensors, the processing of multi-source heterogeneous data still uses simple weighted averaging or static logical judgment, and lacks a dynamic evaluation mechanism for the credibility of sensor data. When multiple sensors generate conflicting data due to environmental interference, the system cannot adjust the data fusion strategy according to real-time working conditions. For example, when the hull is rolling, the weight of the sidewall liquid level sensor that is susceptible to mechanical vibration is not reduced, or the instantaneous abnormal value of the drain pipe pressure sensor due to bubbles is ignored, which ultimately leads to delayed drainage instructions, false triggering or redundant execution. Existing improvement directions are mostly focused on optimizing hardware deployment or actuator response speed, but have not solved the problem of adaptability of multi-sensor collaborative decision-making, making it difficult for the system's robustness and real-time performance in complex sea conditions to meet the management needs of modern intelligent ships. Summary of the Invention

[0004] (1) Technical problems solved

[0005] In response to the shortcomings of the existing technology, the present invention provides an intelligent management system for ship bulwark drainage combined with multi-sensor data monitoring, which solves the problem of how to achieve dynamic fusion and priority decision-making of multi-source sensor data to improve the accuracy and timeliness of drainage strategies under complex sea conditions.

[0006] (2) Technical solution

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent management system for ship bulwark drainage combined with multi-sensor data monitoring, comprising:

[0008] Sensor group, dynamic trust chain fusion decision module, multimodal cross-validation module, fault-tolerant execution and reverse verification module and central controller;

[0009] The sensor group includes a liquid level sensor installed on the inside of the ship's bulwark, a flow rate sensor and pressure sensor in the drainage pipe, a ship's deck attitude sensor, and an external environmental monitoring sensor. Each sensor is connected to the dynamic trust chain fusion decision module via a data bus. During implementation, the liquid level sensor is installed vertically on the bulkhead surface on the inside of the bulwark near the drainage outlet. Its probe is encapsulated with corrosion-resistant materials and collects real-time water depth data in the bulwark area. It is connected to the data bus via a waterproof cable. The flow rate sensor and pressure sensor in the drainage pipe are embedded in the upstream and bend areas of the pipe's inner wall, respectively. The flow rate sensor measures water velocity based on the ultrasonic Doppler effect, while the pressure sensor detects fluid pressure fluctuations in the pipe using a pressure-sensitive element. Both signals are processed through an anti-electromagnetic interference shielding layer and synchronously transmitted to the dynamic trust chain fusion decision module. The deck attitude sensor is fixed to a deck structure near the ship's center of gravity. It integrates a three-axis accelerometer and gyroscope, and outputs roll angle, pitch angle, and angular acceleration data at a sampling frequency of 100Hz. The external environmental monitoring sensor is located at the top of the ship's mast and includes a wind speed and direction meter and a wave height radar to continuously collect external wind and wave parameters.

[0010] The dynamic trust chain fusion decision module receives real-time data from the sensor group and dynamically adjusts the confidence weight of each sensor based on the ship's motion attitude parameters and environmental parameters. The dynamic trust chain fusion decision module obtains real-time roll angular velocity and pitch angle data from the ship's deck attitude sensor, and simultaneously receives surge period and wind speed information collected by external environmental monitoring sensors to establish an association model between the ship's motion attitude and environmental interference.

[0011] The multimodal cross-validation module performs a two-level screening of the data output by the dynamic trust chain fusion decision module through a physical constraint model and a historical behavior database. During the two-level screening process, the physical constraint model prioritizes verifying the rationality of the data, while the historical behavior analysis performs a secondary filtering on data that has passed physical verification but has potential anomalies. Finally, a fusion data packet with a verification mark is generated and transmitted to the central controller. At the same time, the conflicting data features during the screening process are recorded for subsequent sensor health assessment.

[0012] The fault-tolerant execution and reverse verification module uses a digital twin model to replay the effects of drainage instruction execution and reverse-verify them against actual drainage results. During this reverse verification process, the module synchronously updates the digital twin model's parameter calibration coefficients, dynamically adjusting the fluid resistance coefficient and pump efficiency curve based on the actual drainage results, ensuring that the model's prediction accuracy is continuously optimized as the system operates. All verification results and operation records are transmitted to the central controller's decision log via an encrypted link, providing data support for subsequent strategy optimization.

[0013] The central controller generates drainage instructions based on the output signals from the dynamic trust chain fusion decision module, the multimodal cross-validation module, and the fault-tolerant execution and reverse verification module, and controls the actions of the actuators. The resulting drainage instructions are transmitted to the actuators via the CAN bus, and an operation log is simultaneously sent to the ship's central monitoring system, recording the instruction timestamp, weight distribution parameters, and reverse verification error values, forming a closed-loop control evidence chain.

[0014] Preferably, the dynamic trust chain fusion decision module includes:

[0015] The environmental sensing unit is connected to the ship's attitude sensor and environmental monitoring sensor to obtain the ship's roll angular velocity, pitch angle and surge period in real time;

[0016] The trust chain reconstruction unit is connected to the liquid level sensor, flow rate sensor and pressure sensor, and is used to perform time domain decomposition and frequency domain filtering on the sensor signals;

[0017] The weight dynamic allocation engine receives the output signal of the environment perception unit and the filtered data of the trust chain reconstruction unit, and generates the dynamic confidence weight of each sensor.

[0018] Preferably, the environment sensing unit generates a dynamic attenuation coefficient of the liquid level sensor according to the correlation between the ship's roll angular velocity and the surge period, specifically:

[0019] When the ship's rolling angular velocity exceeds the preset threshold, the dynamic attenuation coefficient of the liquid level sensor decreases according to a linear function;

[0020] When the surge period coincides with the ship's inherent rocking frequency, the credibility threshold of the flow velocity sensor is increased to a preset safety value.

[0021] Preferably, the trust chain reconstruction unit includes:

[0022] The liquid level signal processing subunit uses the wavelet packet decomposition algorithm to separate the mechanical vibration interference component in the liquid level sensor signal;

[0023] The pressure signal processing subunit uses a sliding window entropy calculation method to identify transient abnormal pulses in the pressure sensor data, where the transient abnormal pulses are caused by the bursting of bubbles in the drainage pipe.

[0024] During the implementation of the trust chain reconstruction unit, the liquid level signal processing subunit performs multi-scale wavelet packet decomposition on the raw signal of the liquid level sensor installed on the inner side of the bulwark. First, the acquired time domain signal is divided into eight sub-bands. The interference components overlapping with the ship's mechanical vibration frequency are identified through energy spectrum analysis. The signals of the third to fifth sub-bands are extracted as the vibration noise floor. The corresponding frequency band components are reversely eliminated from the original data, retaining the low-frequency trend signal reflecting the true liquid level. For the pressure sensor data processing subunit, a sliding time window is used to intercept the pressure data sequence of 50 consecutive sampling points. The Shannon entropy value of the data in each window is calculated. This includes: when the window entropy value suddenly increases by more than twice the standard deviation of the historical mean, it is determined to be a transient abnormal pulse caused by bubble burst. The data in that window is automatically eliminated and replaced by linear interpolation of the adjacent windows.

[0025] Preferably, the dynamic weight allocation engine performs the following operations under the high-frequency hull shaking condition:

[0026] Adjust the dynamic confidence weight of the sidewall liquid level sensor to the range of 0.2-0.3;

[0027] Increase the dynamic confidence weight of the drain pipe pressure sensor to the range of 0.7-0.8;

[0028] Based on the majority voting results of the majority voting mechanism of at least three sensors in the same monitoring area, the final output value of the dynamic confidence weight is corrected.

[0029] When the dynamic weight distribution engine is implemented under high-frequency hull shaking conditions, it first monitors the roll angular velocity and pitch angle change rate in real time through the ship attitude sensor. When the average roll angular velocity exceeds 8 degrees / second for 5 consecutive seconds and the fluctuation amplitude is greater than 3 degrees, it is judged to be a high-frequency shaking state.

[0030] The engine then starts a dynamic weight adjustment program, including: for the liquid level sensor installed on the side wall, the centrifugal force interference coefficient caused by the sway is calculated according to the relative distance between its installation position and the center of gravity axis of the ship, and combined with the current surge direction, the confidence weight is gradually reduced from the initial value of 1.0 at a rate of 0.12 for every 0.1g acceleration, and finally limited to the range of 0.2-0.3; at the same time, based on the data stability index of the drainage pipe pressure sensor in the past 3 minutes, its weight is increased from the basic value of 0.5 to the range of 0.7-0.8, and the data consistency verification results of at least two pressure sensors in adjacent monitoring areas are linked. If the verification passes, an additional 0.05 weight gain is added.

[0031] After the weight distribution is completed, the engine calls the historical data of the liquid level sensor, pressure sensor and flow rate sensor cluster in the same monitoring area, and calculates the deviation of the current output value of each sensor from the cluster mean. If the deviation of the liquid level sensor data exceeds 25% and the consistency of the pressure sensor cluster data reaches more than 85%, the majority voting mechanism is triggered, including: selecting the two with the most concentrated data among three or more sensors of the same type as valid inputs, forcibly overwriting the output value of the abnormal sensor, and renormalizing the corrected dynamic confidence weight.

[0032] Preferably, the multimodal cross-validation module includes:

[0033] The first-level screening unit verifies the logical consistency between the liquid level sensor data and the ship's draft and surge direction based on the fluid dynamics model;

[0034] The second-level screening unit uses a long-short-term memory network to analyze the historical behavior patterns of sensor data and mark condition-sensitive sensors with frequent abnormalities;

[0035] When the liquid level sensor reports a sudden increase in water level and the flow rate sensor data remains unchanged, the first-level screening unit triggers the delayed response protection mechanism, suspends the output of the drainage command and starts the data review process for three consecutive sampling cycles.

[0036] Preferably, after detecting the working condition sensitive sensor, the second-level screening unit performs the following operations:

[0037] Call the historical data of other sensors in the same monitoring area to calculate the current true value;

[0038] Overwriting abnormal output data of the operating condition-sensitive sensor with an estimated value;

[0039] Generate sensor fault diagnosis signals and upload them to the central controller.

[0040] Preferably, the fault-tolerant execution and reverse verification module includes:

[0041] The virtual sandbox unit simulates the ship's posture changes and the water level fluctuation curve in the cabin within 10 seconds after the drainage command is executed through a digital twin model;

[0042] The reverse comparison unit collects the actual flow data of the drainage pipe pressure sensor in real time and calculates the difference with the predicted value of the virtual sandbox unit;

[0043] If the difference exceeds 5% for three consecutive times, the actuator is judged to be abnormal, the backup drainage pipeline is switched and an alarm signal is triggered.

[0044] Preferably, the central controller executes the following logic before generating the drainage instruction:

[0045] Receive the dynamic confidence weight allocation result of the dynamic trust chain fusion decision module;

[0046] Receive a two-stage screening pass mark from the multimodal cross-validation module;

[0047] receiving a preview deviation detection result from the virtual sandbox unit;

[0048] When the above three inputs all meet the preset conditions, a drainage pump start command and a valve opening control signal are generated.

[0049] Preferably, the triggering conditions of the majority voting mechanism include:

[0050] The sum of the dynamic confidence weights of sensors in the same monitoring area is lower than the preset reliability threshold;

[0051] The difference between the data reported by at least two sensors exceeds the maximum allowable error range;

[0052] The system automatically selects the majority voting result as the final decision basis and marks the low-weight sensors for offline calibration.

[0053] (3) Beneficial effects

[0054] The present invention provides an intelligent management system for ship bulwark drainage that combines multi-sensor data monitoring. It has the following beneficial effects:

[0055] (1) The ship bulwark drainage intelligent management system, which combines multi-sensor data monitoring, improves the accuracy and timeliness of multi-source sensor data fusion under complex sea conditions by building a dynamic trust chain fusion decision model and a multimodal cross-validation mechanism. The dynamic weight distribution algorithm based on ship attitude and environmental parameters effectively suppresses sensor misjudgments caused by mechanical vibration, bubble interference, etc., and combines digital twin rehearsal and reverse verification closed loop to reduce the false trigger rate of drainage commands and shorten response delays. At the same time, through two-level screening and majority voting mechanisms, the collaborative fault tolerance of the sensor cluster is achieved, ensuring the continuous and stable operation of the system under single point failures or abnormal interference.

[0056] (2) The ship bulwark drainage intelligent management system, which combines multi-sensor data monitoring, solves the adaptability defects of traditional drainage management solutions under dynamic sea conditions, improves the navigation safety and energy efficiency management level of ships, and its adaptive learning ability can optimize sensor evaluation models and decision thresholds through historical data, reducing the frequency of manual maintenance; the modular design is compatible with a variety of ship types and drainage system architectures, providing technical support for the automation upgrade of smart ships. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a schematic diagram of the overall framework of the present invention;

[0058] Figure 2 This is a control logic timing diagram of the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] See also Figure 1 and Figure 2 The present invention provides a technical solution: an intelligent management system for ship bulwark drainage combined with multi-sensor data monitoring, comprising:

[0061] Sensor group, dynamic trust chain fusion decision module, multimodal cross-validation module, fault-tolerant execution and reverse verification module and central controller;

[0062] The sensor group includes a liquid level sensor installed on the inside of the ship's bulwark, a flow rate sensor and pressure sensor in the drainage pipe, a deck attitude sensor, and an external environmental monitoring sensor. Each sensor is connected to the dynamic trust chain fusion decision module via a data bus. In the sensor group's implementation, the liquid level sensor is installed vertically on the inner bulkhead surface near the drainage outlet. Its probe is encapsulated with corrosion-resistant materials and collects real-time water depth data in the bulwark area. It is connected to the data bus via a waterproof cable. The flow rate sensor and pressure sensor in the drainage pipe are embedded in the upstream and bend of the pipe wall, respectively. The flow rate sensor measures water velocity based on the ultrasonic Doppler effect, while the pressure sensor detects pressure fluctuations in the pipe using a pressure-sensitive element. Both signals are processed through an anti-electromagnetic interference shielding layer and synchronously transmitted to the dynamic trust chain fusion decision module. The deck attitude sensor is fixed to the deck structure near the ship's center of gravity. It integrates a three-axis accelerometer and gyroscope, and outputs roll angle, pitch angle, and angular acceleration data at a sampling frequency of 100Hz. The external environmental monitoring sensor is located atop the ship's mast and includes a wind speed and direction meter and a wave height radar to continuously collect external wind and wave parameters.

[0063] Sensor data is transmitted via an industrial-grade CAN bus. The bus controller incorporates built-in data verification and conflict arbitration mechanisms to ensure real-time synchronization of heterogeneous multi-source data. Upon receiving the data, the dynamic trust chain fusion decision module first applies a sliding mean filter to the liquid level sensor signal to eliminate transient spike noise caused by ship turbulence. It also performs baseline drift correction on the pressure sensor data. The module then dynamically calculates the initial weight coefficients for each sensor, combining the angular acceleration output by the deck attitude sensor, to form a raw data preprocessing chain.

[0064] The dynamic trust chain fusion decision module receives real-time data from the sensor group and dynamically adjusts the confidence weight of each sensor based on the ship's motion attitude parameters and environmental parameters. The module acquires real-time roll angular velocity and pitch angle data from the ship's deck attitude sensor. It also receives surge period and wind speed information from external environmental monitoring sensors to establish a correlation model between the ship's motion attitude and environmental interference. When the roll angular velocity exceeds a preset threshold, the module automatically triggers the liquid level sensor's credibility attenuation mechanism. This includes calculating the vibration interference intensity based on the angular acceleration value and performing a time-domain sliding window analysis on the raw liquid level sensor data. If the signal fluctuation amplitude within three consecutive windows is positively correlated with the ship's rolling frequency, the signal is identified as a false water level signal caused by mechanical vibration, and its confidence weight is linearly reduced from an initial value of 1.0 to 0.3. Furthermore, based on the surge period and the baseline pressure characteristics of the drainage pipe pressure sensor, if the current surge frequency is close to the inherent resonant frequency of the ship's drainage system, the decision priority of the pressure sensor data is increased, raising its confidence weight to above 0.7.

[0065] For velocity sensor data, the module uses frequency-domain bandpass filtering to eliminate pipeline eddy noise caused by ship sway. It also dynamically adjusts the confidence weight coefficient by performing a spatial consistency check on velocity data from adjacent monitoring points. When conflicting sensor data occurs, the module draws on the historical behavior database of sensor clusters in the same area, analyzing the probability of anomalies for each sensor under the current operating conditions. Data from sensors with high confidence weights is prioritized, and a majority voting mechanism is triggered. This includes overwriting the anomalous output values of lower-confidence sensors if the data from at least two sensors of the same type agree within a preset error range, ultimately generating a fused drainage status assessment result.

[0066] The multimodal cross-validation module uses a physical constraint model and a historical behavior database to perform a two-level screening of the data output by the dynamic trust chain fusion decision module. During implementation, the multimodal cross-validation module first uses a physical constraint model to perform a first-level screening of the data output by the dynamic trust chain fusion decision module. This includes: constructing a displacement-hull attitude balance equation based on the ship's current draft, drainage pipe diameter, and surge direction parameters, and verifying the logical correlation between the liquid level sensor data and the flow rate sensor data in real time. When the liquid level sensor reports a sudden increase in water level but the flow rate sensor shows that the drainage flow has not changed, the theoretical water volume in the bulwark area is calculated in combination with the ship's real-time roll angle. If the water volume corresponding to the sudden increase in water level exceeds the hull stability limit, it is determined to be transient interference data caused by waves on the deck, triggering the delayed response protection mechanism, suspending the generation of drainage instructions, and starting a data review process for three consecutive sampling cycles.

[0067] The second-level screening is based on the historical behavior database to trace the source of anomalies, including: analyzing the data fluctuation characteristics of each sensor under similar working conditions in the past 30 days through the long-short term memory network, and establishing a correlation map between sensor data and ship speed and steering angle. If the current flow rate sensor continuously experiences data jumps that exceed three times the standard deviation of the historical mean at a specific steering angle, it will be marked as a working condition-sensitive sensor, and the historical data mean of the pressure sensor and adjacent liquid level sensor in the same monitoring area will be automatically called for alternative compensation, and the abnormal data source will be marked as a low-priority input.

[0068] During the two-level screening process, the physical constraint model prioritizes verifying the rationality of the data, while historical behavior analysis performs secondary filtering on data that has passed physical verification but has potential anomalies. Finally, a fused data packet with a verification mark is generated and transmitted to the central controller. At the same time, the conflicting data characteristics during the screening process are recorded for subsequent sensor health assessment.

[0069] The fault-tolerant execution and reverse verification module previews the execution effect of the drainage instruction through the digital twin model, and reversely verifies it with the actual drainage results; during the implementation of the fault-tolerant execution and reverse verification module, a digital twin model is first constructed based on the ship's three-dimensional model, drainage system structural parameters and real-time sensor data. The model integrates the hull attitude dynamics equations and the fluid mechanics simulation engine.

[0070] When the central controller generates a drainage pump start-up command, the module inputs the current liquid level, pipeline pressure, valve opening and ship motion parameters into the digital twin model to simulate the water level drop curve in the bulwark area, the drainage pipe flow change and the center of gravity offset of the ship in the next 10 seconds.

[0071] If the simulation results show that the water level drop rate is lower than 95% of the expected value, or the increase in the hull roll angle exceeds the safety threshold, it is determined to be a potential abnormal operating condition, triggering dynamic strategy backtracking, including: pausing the current instruction execution, re-calling the dynamic trust chain fusion decision module to perform secondary fusion calculations on the sensor data, and generating a revised drainage strategy.

[0072] During the actual drainage process, the module collects actual pipeline flow data at a frequency of 20 times per second through a high-precision pressure sensor, compares it with the predicted value of the digital twin model in real time, and calculates the root mean square error between the two within the sliding time window.

[0073] When the error value of three consecutive sampling windows exceeds 5%, it is automatically determined to be an actuator failure or sensor failure, and the current drainage pipeline control signal is immediately cut off, switching to the backup drainage pipeline and activating the sound and light alarm device. At the same time, the abnormal data packet is marked as a fault case and stored in the historical database.

[0074] During the reverse verification process, the module synchronously updates the digital twin model's parameter calibration coefficients, dynamically adjusting the fluid resistance coefficient and pump efficiency curve based on actual drainage performance, ensuring that the model's prediction accuracy continuously improves as the system operates. All verification results and operation records are transmitted via an encrypted link to the central controller's decision log, providing data support for subsequent strategy optimization.

[0075] The central controller generates drainage instructions and controls the actuators based on the output signals from the dynamic trust chain fusion decision module, the multimodal cross-validation module, and the fault-tolerant execution and reverse verification module. During implementation, the central controller first receives the dynamic confidence weight distribution table for each sensor from the dynamic trust chain fusion decision module. It then simultaneously obtains the two-stage screening pass flag from the multimodal cross-validation module and the virtual sandbox preview deviation value from the fault-tolerant execution and reverse verification module.

[0076] When the weight of the liquid level sensor in the dynamic confidence weight table is higher than 0.5 and the weight of the pressure sensor is lower than 0.3, the controller calls the drainage pipe historical flow database and calculates the theoretical displacement threshold based on the current ship draft. If the physical constraint screening of the multimodal cross-validation module shows that the liquid level data meets the hull stability requirements and the virtual sandbox preview deviation value is less than 3%, the drainage pump start-up command and the corresponding valve opening control signal are generated.

[0077] During the command execution phase, the controller monitors the actuator's feedback current and valve displacement sensor data in real time. If it detects an abnormal increase in the drainage pump motor current exceeding 15% of the rated value, it immediately triggers a gradient load reduction strategy: the pump speed is reduced by 5% per second, while the valve opening is gradually reduced to prevent pipeline overpressure. If the fault-tolerant execution module detects three consecutive deviations between the actual drainage flow rate and the predicted value through reverse comparison, the controller interrupts the current control loop, switches to the backup drainage pipeline, and activates emergency drainage mode. In this mode, pressure sensor data is prioritized for decision-making, and fault information is encrypted and stored in the black box system.

[0078] Throughout the decision-making process, the controller uses a multi-threaded architecture to process input signals from each module in parallel, ensuring that the entire process, from data reception to command issuance, is completed within 200 milliseconds. A real-time drainage strategy decision tree diagram is displayed on the human-computer interface, with the weight distribution of each sensor and the pass status of each verification step noted. The resulting drainage command is transmitted to the actuator via the CAN bus, and an operation log is simultaneously sent to the ship's central monitoring system, recording the command timestamp, weight distribution parameters, and reverse verification error value, forming a closed-loop control chain of evidence.

[0079] The dynamic trust chain fusion decision module includes:

[0080] The environmental sensing unit is connected to the ship's attitude sensor and environmental monitoring sensor to obtain the ship's roll angular velocity, pitch angle and surge period in real time;

[0081] The trust chain reconstruction unit is connected to the liquid level sensor, flow rate sensor and pressure sensor, and is used to perform time domain decomposition and frequency domain filtering on the sensor signals;

[0082] The weight dynamic allocation engine receives the output signal of the environmental perception unit and the filtered data of the trust chain reconstruction unit, and generates the dynamic confidence weight of each sensor.

[0083] It should be further explained that in the specific implementation process, when implementing the dynamic trust chain fusion decision module, the environmental perception unit collects roll angular velocity and pitch angle data in real time through the ship deck attitude sensor, and at the same time connects to the external environmental monitoring sensor to obtain surge cycle and wind speed parameters, and establishes a mapping relationship library between the ship's motion state and environmental interference.

[0084] When the roll angular velocity exceeds the preset threshold, the environmental perception unit generates the vibration interference coefficient of the liquid level sensor, which is positively correlated with the angular acceleration value, and transmits it to the weight dynamic allocation engine in real time; the trust chain reconstruction unit performs time domain decomposition processing on the liquid level sensor signal, uses a sliding time window to intercept 10 seconds of continuous raw data, and uses fast Fourier transform to separate the vibration noise component consistent with the ship's rocking frequency, retaining the true liquid level change trend signal.

[0085] For pressure sensor data, the unit sets a frequency-domain bandpass filter to remove baseline drift below 0.1Hz and mechanical impact interference above 50Hz, extracting effective pressure fluctuation characteristics. The dynamic weight allocation engine receives the interference coefficient output by the environmental perception unit and the feature data processed by the trust chain reconstruction unit. It uses a fuzzy logic algorithm to dynamically calculate the confidence weight of each sensor. This includes: when the difference between the surge period and the natural frequency of the ship's drainage system is less than 0.5Hz, the flow rate sensor weight is reduced by 30%. At the same time, based on the consistency verification results of the pressure sensor data and the liquid level sensors in the adjacent area, the pressure sensor weight is increased to above the preset safety threshold.

[0086] Under severe ship sway conditions, the engine activates a majority voting mechanism, comparing the data dispersion of at least three sensors within the same monitoring area. If the variance of the level sensor data exceeds the allowable range and the consistency of the pressure sensor cluster data exceeds 90%, the pressure sensor data is mandatory for decision-making, and the abnormal level sensor is marked as a low-trust device. All weight allocation results and correction decisions are transmitted in real time to the central controller via a high-speed data channel and simultaneously updated to the dynamic trust chain database for subsequent sensor health assessment and self-learning model optimization.

[0087] The environmental perception unit generates the dynamic attenuation coefficient of the liquid level sensor based on the correlation between the ship's roll angular velocity and the surge period. Specifically, it is:

[0088] When the ship's rolling angular velocity exceeds the preset threshold, the dynamic attenuation coefficient of the liquid level sensor decreases according to a linear function;

[0089] When the surge period coincides with the ship's inherent rocking frequency, the credibility threshold of the flow velocity sensor is increased to a preset safety value.

[0090] It should be further explained that, during the specific implementation process, the environmental perception unit monitors the roll angular velocity data in real time through the ship deck attitude sensor. When the angular velocity value is detected to exceed the preset threshold of 5 degrees / second, the unit automatically activates the dynamic attenuation mechanism of the liquid level sensor.

[0091] This mechanism generates an attenuation coefficient that is positively correlated with the vibration intensity based on the integral calculation results of the roll angular acceleration, including: if the angular acceleration continues to increase within 3 seconds, the attenuation coefficient decreases linearly at a rate of 0.15 for every 0.1g acceleration, down to a minimum of 0.4; at the same time, the surge cycle data collected by the environmental monitoring sensor is matched with the ship's inherent swing frequency library. When the difference between the two is less than 0.5 seconds, the unit will increase the credibility threshold of the flow rate sensor from the default 0.6 to a safe value of 0.8, and associate it with the baseline stability index of the drainage pipe pressure sensor. If the standard deviation of the pressure data is lower than 10kPa for five consecutive sampling cycles, the flow rate sensor threshold will be further increased to 1.2 times the preset value.

[0092] During the application of the dynamic attenuation coefficient, the unit performs sliding mean filtering on the raw data of the liquid level sensor, and the window width is dynamically adjusted with the attenuation coefficient: when the coefficient drops below 0.5, the window is expanded to 15 seconds to suppress high-frequency noise. At the same time, the filtered data is compared with the equivalent liquid level value calculated by the pressure sensor in the same area. If the difference exceeds 10 cm, a secondary correction is triggered, and the attenuation coefficient is reduced by an additional 0.1.

[0093] All adjustment parameters are written into the ship motion-sensor error correlation matrix in real time and synchronized to the central controller's decision database via a data bus, providing a historical reference for subsequent weight allocation. When the ship enters a stable navigation state, the unit automatically restores the liquid level sensor attenuation coefficient to its initial value of 1.0 and updates the adaptive learning model for the flow velocity sensor's credibility threshold based on the surge cycle distribution characteristics over the past 30 minutes.

[0094] The trust chain reconstruction unit includes:

[0095] The liquid level signal processing subunit uses the wavelet packet decomposition algorithm to separate the mechanical vibration interference component in the liquid level sensor signal;

[0096] The pressure signal processing subunit uses a sliding window entropy calculation method to identify transient abnormal pulses in the pressure sensor data. Transient abnormal pulses are caused by the bursting of bubbles in the drainage pipe.

[0097] It should be further explained that during the specific implementation process, the trust chain reconstruction unit is implemented. The liquid level signal processing subunit performs multi-scale wavelet packet decomposition on the original signal of the liquid level sensor installed on the inner side of the bulwark. First, the collected time domain signal is divided into 8 sub-bands. The interference components overlapping with the ship's mechanical vibration frequency are identified through energy spectrum analysis. The signals of the third to fifth sub-bands are extracted as the vibration noise floor. The corresponding frequency band components are reversely eliminated from the original data, retaining the low-frequency trend signal reflecting the true liquid level. The ship's mechanical vibration frequency range is 2-15Hz. For the pressure sensor data processing subunit, a sliding time window is used to intercept the pressure data sequence of 50 consecutive sampling points. The Shannon entropy value of the data in each window is calculated. This includes: when the window entropy value suddenly increases by more than twice the standard deviation of the historical mean, it is determined to be a transient abnormal pulse caused by bubble burst. The data in that window is automatically eliminated and replaced by linear interpolation of the adjacent windows.

[0098] During the processing process, if the difference between the filtered level sensor data and the estimated level value from the pressure sensor exceeds 15 cm for more than 10 seconds, the sensor collaborative calibration mechanism is triggered. This involves inferring the theoretical level change using the flow integration data from the flow velocity sensor in the same area. If the inferred result matches the filtered level sensor value by more than 90%, the pressure sensor's current window data is marked as valid. Otherwise, the pressure sensor output is frozen and the backup sensor data replacement process is initiated. The processed level and pressure signals are then aligned and compensated for transmission delays using a time-domain alignment algorithm. This ultimately generates a synchronized feature dataset that is transmitted to the dynamic weight allocation engine. The frequency and duration of abnormal pulses are also recorded to update the adaptive parameters of the sensor health assessment model.

[0099] The dynamic weight distribution engine performs the following operations under high-frequency hull sway conditions:

[0100] Adjust the dynamic confidence weight of the sidewall liquid level sensor to the range of 0.2-0.3;

[0101] Increase the dynamic confidence weight of the drain pipe pressure sensor to the range of 0.7-0.8;

[0102] Based on the majority voting results of the majority voting mechanism of at least three sensors in the same monitoring area, the final output value of the dynamic confidence weight is corrected.

[0103] It should be further explained that, in the specific implementation process, when the weight dynamic allocation engine is implemented under the condition of high-frequency hull shaking, the ship attitude sensor is first used to monitor the roll angular velocity and pitch angle change rate in real time. When the average roll angular velocity exceeds 8 degrees / second for 5 consecutive seconds and the fluctuation amplitude is greater than 3 degrees, it is judged to be a high-frequency shaking state.

[0104] The engine then starts the dynamic weight adjustment program: for the liquid level sensor installed on the side wall, the centrifugal force interference coefficient caused by the shaking is calculated according to the relative distance between its installation position and the center of gravity axis of the ship. Combined with the current surge direction, the confidence weight is gradually reduced from the initial value of 1.0 at a rate of 0.12 for every 0.1g acceleration, and finally limited to the range of 0.2-0.3; at the same time, based on the data stability index of the drainage pipe pressure sensor in the past 3 minutes (standard deviation is less than 15kPa), its weight is increased from the basic value of 0.5 to the range of 0.7-0.8, and the data consistency verification results of at least two pressure sensors in adjacent monitoring areas are linked. If the verification passes, an additional 0.05 weight gain is added.

[0105] After the weight distribution is completed, the engine calls the historical data of the liquid level sensor, pressure sensor and flow rate sensor cluster in the same monitoring area, and calculates the deviation of the current output value of each sensor from the cluster mean. If the deviation of the liquid level sensor data exceeds 25% and the consistency of the pressure sensor cluster data reaches more than 85%, the majority voting mechanism is triggered, including: selecting the two with the most concentrated data among three or more sensors of the same type as valid inputs, forcibly overwriting the output value of the abnormal sensor, and renormalizing the corrected dynamic confidence weight.

[0106] During the correction process, the engine compares the calculated drainage demand before and after the adjustment in real time. If the difference exceeds a preset threshold of 10%, a secondary arbitration process is initiated. Weights are fine-tuned based on the screening results of the multimodal cross-validation module, and a fused data packet with a credibility indicator is generated and transmitted to the central controller. All weight adjustment parameters and voting records are synchronously updated to the dynamic trust chain database for subsequent sensor performance degradation analysis and self-learning model training, ensuring that the system gradually optimizes the weight allocation strategy during ongoing operation.

[0107] The multimodal cross-validation module includes:

[0108] The first-level screening unit verifies the logical consistency between the liquid level sensor data and the ship's draft and surge direction based on the fluid dynamics model;

[0109] The second-level screening unit uses a long-short-term memory network to analyze the historical behavior patterns of sensor data and mark condition-sensitive sensors with frequent abnormalities;

[0110] When the liquid level sensor reports a sudden increase in water level and the flow rate sensor data remains unchanged, the first-level screening unit triggers the delayed response protection mechanism, suspends the output of the drainage command and starts the data review process for three consecutive sampling cycles.

[0111] It should be further explained that during the specific implementation process, when the multimodal cross-validation module is being implemented, the first-level physical constraint screening unit constructs a dynamic fluid dynamics model based on the ship's real-time draft and surge direction parameters. When the liquid level sensor reports a sudden increase in water level, the theoretical drainage efficiency is calculated in combination with the design flux of the drainage pipe. This includes: if the current flow rate sensor data shows that the drainage flow rate has not increased synchronously, and the ship's roll angle causes the actual volume of the bulwark area to be less than the theoretical water accumulation corresponding to the sudden increase in water level, it is determined to be a transient interference caused by an external surge impact, triggering the delayed response protection mechanism, freezing the generation of drainage instructions, and starting the data review process for three consecutive sampling cycles. During this period, the module synchronously retrieves data from the pressure sensor cluster in the adjacent area. If at least two pressure sensors show that the pressure fluctuations in the pipeline match the normal drainage pattern, the frozen state is released and a revised drainage instruction is generated.

[0112] The second-level screening unit analyzes the historical behavior patterns of sensors through a long-short-term memory network to establish a data fluctuation baseline library under different combinations of ship speeds, steering angles, and surge intensities. When the flow rate sensor continuously experiences data jumps that exceed three times the standard deviation of the historical mean at a specific steering angle (such as 35 degrees full right rudder), it is automatically marked as a working condition-sensitive sensor, and the historical data mean of the liquid level sensor and pressure sensor in the same monitoring area is called for alternative compensation, and the data weight of the abnormal sensor is forced to zero.

[0113] During the two-level screening process, if the liquid level sensor data passes the physical constraint verification but historical behavior analysis shows that its false alarm rate exceeds 15% under similar working conditions in the past 24 hours, the weight redistribution process of the dynamic trust chain fusion decision module is triggered, the sensor confidence weight is reduced, and the backup sensor data redundancy check is initiated.

[0114] All screening results and compensation operations are recorded in the verification log and synchronized to the decision tree database of the central controller in real time, ensuring the dual verification results of comprehensive physical laws and historical behavior characteristics when generating drainage strategies.

[0115] After the second-level screening unit detects the condition-sensitive sensor, it performs the following operations:

[0116] Call the historical data of other sensors in the same monitoring area to calculate the current true value;

[0117] Use estimated values to cover abnormal output data of condition-sensitive sensors;

[0118] Generate sensor fault diagnosis signals and upload them to the central controller.

[0119] It should be further explained that, in the specific implementation process, after detecting the working condition-sensitive sensor, the second-level screening unit first calls the historical data of the liquid level sensor, pressure sensor and flow rate sensor in the same monitoring area for the past 30 days, matches similar working condition segments according to the current ship speed, steering angle and surge intensity, and extracts the mean and standard deviation of each sensor in the corresponding time period as a benchmark reference.

[0120] If the current output value of the marked flow rate sensor exceeds the historical mean ±3 times the standard deviation range, the data replacement compensation algorithm will be activated, including: inferring the theoretical value of the pipeline flow based on the pressure sensor data, calculating the equivalent flow rate based on the water level change rate of the adjacent liquid level sensor, and generating a compensation value to cover the original output of the abnormal sensor. During the covering process, a sliding time window is used for smooth transition to avoid data jumps.

[0121] At the same time, the system automatically triggers a sensor redundancy check process, which includes activating a redundant pressure sensor group in the backup drainage pipeline and comparing its data with the consistency of the primary sensor cluster. If the redundant group data deviates from the compensation value by less than 5%, the replacement is confirmed to be effective and a sensor fault diagnosis signal is generated. This signal includes the time of the anomaly, operating parameters, and a comparison curve of the data before and after compensation. The signal is then uploaded to the central controller's maintenance terminal via an encrypted protocol. If the redundancy check fails, the digital twin model's simulation data is further called up, combined with the virtual sandbox pre-run results to generate a dynamic compensation coefficient. The sensor is then marked as a high-risk device, forcing it to participate in decision-making only as an auxiliary data source for the next 10 minutes.

[0122] All replacement operations are recorded in the sensor health profile for subsequent self-learning model updates. The abnormal sensor location and compensation status are highlighted on the human-machine interface, prompting the crew to perform manual review or maintenance. Once a fault diagnosis signal is generated, the system automatically reduces the initial weight of that sensor in the dynamic trust chain fusion decision module to below 0.2 and increases the priority of backup sensors in adjacent areas to ensure the continued and stable execution of the drainage strategy.

[0123] The fault-tolerant execution and reverse verification modules include:

[0124] The virtual sandbox unit simulates the ship's posture changes and the water level fluctuation curve in the cabin within 10 seconds after the drainage command is executed through a digital twin model;

[0125] The reverse comparison unit collects the actual flow data of the drainage pipe pressure sensor in real time and calculates the difference with the predicted value of the virtual sandbox unit;

[0126] If the difference exceeds 5% for three consecutive times, the actuator is judged to be abnormal, the backup drainage pipeline is switched and an alarm signal is triggered.

[0127] It should be further explained that in the specific implementation process, during the implementation of the fault-tolerant execution and reverse verification module, when the central controller generates a drainage pump start-up instruction, the module first constructs a dynamic simulation scenario based on the three-dimensional digital model of the ship and real-time sensor data, simulates the gradual adjustment process of the drainage valve opening from 10% to the target value, and predicts the water level drop rate in the bulwark area, pipeline pressure fluctuations and hull center of gravity offset in the next 10 seconds.

[0128] If the simulation results show that the water level drop rate is lower than 95% of the theoretical value or the increase in the hull roll angle exceeds the safety threshold of 2 degrees, the module immediately triggers dynamic strategy backtracking, including: pausing the current instruction execution, sending a data re-verification request to the dynamic trust chain fusion decision module, and calling the multimodal cross-validation module to perform a secondary screening of the liquid level and pressure sensor data.

[0129] During the actual drainage stage, the module obtains the actual flow value through the pressure sensor in the drainage pipe at a sampling frequency of 20 times per second, performs a sliding window comparison with the digital twin model prediction curve, and calculates the average absolute error within each 3-second time window.

[0130] When the error of three consecutive windows exceeds 5%, the module determines that the actuator is abnormal, cuts off the current main drainage pipeline control signal, activates the backup pipeline solenoid valve and increases the redundant pump power to 120% of the rated value, and at the same time issues a third-level alarm signal through the deck sound and light alarm.

[0131] During the reverse verification process, the module synchronously analyzes the long-term deviation trend between the actual drainage data and the predicted value. If it is found that the cumulative deviation of the fluid resistance coefficient exceeds 8% of the initial parameter, the pipeline friction coefficient calibration value of the digital twin model is automatically updated, and the pump efficiency curve fitting model is retrained.

[0132] All operation records and verification results are transmitted to the central controller's secure storage unit via AES-256 encryption, generating a complete chain of evidence including timestamps, simulation parameters, actual error values, and fault codes for subsequent accident tracing and system optimization. After switching to the backup pipeline, the module continuously monitors the pressure stability of the new pipeline. If the error alarm is not triggered again within 30 seconds, a recovery readiness signal is sent to the central controller. Otherwise, the full system safety lock mechanism is activated and an emergency shutdown request is transmitted to the ship's control center.

[0133] The central controller executes the following logic before generating a drain command:

[0134] Receive the dynamic confidence weight allocation result of the dynamic trust chain fusion decision module;

[0135] Receive a two-stage screening pass mark from the multimodal cross-validation module;

[0136] receiving a preview deviation detection result from the virtual sandbox unit;

[0137] When the above three inputs all meet the preset conditions, a drainage pump start instruction and a valve opening control signal are generated.

[0138] It should be further explained that, in the specific implementation process, when the central controller receives the dynamic confidence weight allocation table of the dynamic trust chain fusion decision module, it first verifies whether the weight of the liquid level sensor is higher than 0.5 and the weight of the pressure sensor is lower than 0.3. If the conditions are met, the draft and displacement relationship curve in the ship stability database is called, and the theoretical displacement flow threshold is calculated in combination with the current actual draft value.

[0139] The two-level screening pass mark of the multimodal cross-validation module is read synchronously, including: physical constraint screening needs to verify the logical consistency of the liquid level surge data and the surge direction and hull roll angle; historical behavior screening needs to confirm that there are no high-frequency false alarm records under similar working conditions.

[0140] The virtual sandbox rehearsal results of the fault-tolerant execution module are simultaneously obtained. If the deviation is less than 3% and the predicted water level drop rate reaches or exceeds 90% of the rated value, the drainage instruction generation process is initiated. The controller uses a weighted voting mechanism to integrate three types of input signals: dynamic confidence weight, which accounts for 50% of the decision weight, two-stage screening results, which account for 30%, and virtual sandbox deviation, which accounts for 20%. When the combined score exceeds the preset threshold of 0.8, a hierarchical control instruction is generated: first, the electric valve of the main drainage line is opened to 60% and the centrifugal pump is started at 80% of the rated speed.

[0141] During the execution of the instruction, the feedback signal of the valve displacement sensor and the pump current value are collected in real time. If it is detected that the actual valve opening deviates from the instruction value by more than 5% for 5 seconds, or the pump current fluctuation amplitude exceeds 10% of the rated value, the gradient adjustment strategy is triggered, including: gradually increasing the valve opening by 2% per second, and at the same time increasing the pump speed to 85% until the actual drainage flow matches the theoretical threshold.

[0142] When the fault-tolerant execution module reverse verification finds that the flow is continuously abnormal, the controller interrupts the current control signal, switches to the backup pipeline and activates the emergency mode. In this mode, the dynamic weight distribution function is turned off and the average value of the pressure sensor and redundant liquid level sensor is directly used to drive the decision.

[0143] All operation data is encrypted with a timestamp to generate a decision log, which includes weight distribution parameters, verification identification status and actuator response curve. It is synchronously transmitted to the ship's central monitoring system through a dual-channel redundant bus, forming a traceable closed-loop control evidence chain. The drainage strategy decision tree is dynamically displayed on the human-machine interface, highlighting the currently effective sensor data source and the verification link pass status.

[0144] The triggering conditions for the majority voting mechanism include:

[0145] The sum of the dynamic confidence weights of sensors in the same monitoring area is lower than the preset reliability threshold;

[0146] The difference between the data reported by at least two sensors exceeds the maximum allowable error range;

[0147] The system automatically selects the majority voting result as the final decision basis and marks the low-weight sensors for offline calibration.

[0148] It should be further explained that, during implementation, the majority voting mechanism of the ship bulwark drainage intelligent management system automatically triggers a majority voting process when the sum of the dynamic confidence weights of sensors within the same monitoring area falls below a preset reliability threshold of 0.7. This process involves first selecting at least three valid sensor data points from the level, pressure, and flow rate sensor clusters within the area, calculating their mean and dispersion, and then using these two sensor data points as valid input sets for a final decision. For example, if the difference between the level difference and the valid input set exceeds 150% of the maximum allowable error, the system marks the sensor as low-weighted, freezes its data output, and generates an offline calibration instruction. This instruction contains the abnormal time point, operating parameters, and deviation history, and is transmitted to the maintenance terminal via an encrypted link. During calibration, the system automatically increases the initial weight of the backup sensors in adjacent areas to 1.2 times the original value, and calls the digital twin model simulation data to fill in the missing values to ensure decision continuity.

[0149] After the vote is complete, the system reverse-validates the revised data against the original weight distribution. If the revised value causes the displacement strategy to adjust by more than 20% from the initial decision, the dynamic trust chain fusion decision module's weight redistribution process is triggered, reassessing the health status of each sensor and updating the attenuation coefficient. All voting records and calibration operations are synchronously written to the sensor health profile, which is used to optimize the threshold settings of subsequent self-learning models. The ship's human-machine interface also displays the sensor credibility distribution in each area as a heat map, noting the location of low-weighted sensors and the calibration countdown.

[0150] When the ship enters a stable sailing state, the system automatically unfreezes the low-weight sensors and dynamically restores their initial weights to 80% of the preset value based on the voting results in the past 2 hours. If no abnormalities are triggered after three consecutive votes, their normal right to participate in decision-making will be fully restored.

[0151] A method for intelligent management of ship bulwark drainage combined with multi-sensor data monitoring, comprising the following steps:

[0152] Step S1: Data is collected in real time through the bulwark liquid level sensor, drainage pipe flow rate / pressure sensor, ship attitude sensor and environmental monitoring sensor, and transmitted to the dynamic trust chain fusion decision module via the anti-interference data bus to complete the sliding mean filtering and baseline drift correction of the original data.

[0153] Step S2: Generate the dynamic attenuation coefficient of the sensor based on the ship's roll angular velocity, surge period and environmental parameters, use wavelet packet decomposition and entropy calculation to separate the signal interference component, dynamically assign confidence weights to each sensor, and trigger the majority voting mechanism to eliminate conflicting data.

[0154] Step S3: Verify the physical rationality of the liquid level / flow rate data through the fluid mechanics model, analyze the sensor abnormal pattern in combination with the historical behavior database, and start the data replacement compensation algorithm for the sensor that has passed the physical verification but has high-frequency false alarms.

[0155] Step S4: Use the digital twin model to simulate the changes in the hull posture and water level after the displacement instruction is executed. If the prediction deviation exceeds 5% or the hull stability risk increases, the dynamic strategy backtracking is triggered and the instruction is regenerated.

[0156] Step S5: The central controller generates a graded drainage instruction based on the dynamic weight, verification mark and preview result, controls the valve opening and pump speed, and monitors the actuator current and displacement feedback signal in real time.

[0157] Step S6: The actual drainage flow is collected through the pressure sensor and compared with the digital twin prediction value in a sliding window. If the error exceeds the limit for three consecutive times, it is determined that the execution is abnormal and the backup pipeline is switched.

[0158] Step S7: Mark the faulty sensor as low-weighted and start redundant data compensation. After switching to the backup pipeline, continuously monitor the pressure stability. If there is no secondary alarm within 30 seconds, restore the system to normal mode.

[0159] Step S8: Encrypt and record the weight distribution parameters, verification results, and execution response curves, update the friction coefficient and pump efficiency curve of the digital twin model, and simultaneously optimize the threshold setting of the self-learning algorithm.

[0160] Step S9: Based on the majority voting results and historical calibration records, a sensor health heat map is generated, the location of low-weight devices and the offline calibration countdown are marked, and a manual maintenance prompt is triggered.

[0161] Step S10: Use actual drainage data to reversely train the dynamic trust chain fusion model, optimize the weight distribution strategy and majority voting trigger conditions, and improve the system's adaptability under complex sea conditions.

[0162] By constructing a dynamic trust chain fusion decision model and a multimodal cross-validation mechanism, the accuracy and timeliness of multi-source sensor data fusion under complex sea conditions are improved; the dynamic weight distribution algorithm based on ship posture and environmental parameters effectively suppresses sensor misjudgments caused by mechanical vibration, bubble interference, etc., and combines digital twin rehearsal and reverse verification closed loop to reduce the false trigger rate of drainage commands and shorten response delays. At the same time, through two-level screening and majority voting mechanisms, the collaborative fault tolerance of the sensor cluster is achieved, ensuring the system's continued stable operation under single-point failures or abnormal interference.

[0163] It solves the adaptability defects of traditional drainage management solutions under dynamic sea conditions, improves the navigation safety and energy efficiency management level of ships, and its adaptive learning ability can optimize sensor evaluation models and decision thresholds through historical data, reducing the frequency of manual maintenance; the modular design is compatible with a variety of ship types and drainage system architectures, providing technical support for the automated upgrade of smart ships.

[0164] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0165] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent management system for ship bulwark drainage combined with multi-sensor data monitoring, characterized in that: include: Sensor group, dynamic trust chain fusion decision module, multimodal cross-validation module, fault-tolerant execution and reverse verification module and central controller; The sensor group includes a liquid level sensor installed on the inside of the ship's bulwark, a flow rate sensor and a pressure sensor in the drainage pipe, a ship's deck posture sensor, and an external environmental monitoring sensor. Each sensor is connected to the dynamic trust chain fusion decision module via a data bus; The dynamic trust chain fusion decision module receives real-time data from the sensor group and dynamically adjusts the confidence weight of each sensor based on the ship's motion posture parameters and environmental parameters; The multimodal cross-validation module performs two-level screening on the data output by the dynamic trust chain fusion decision module through the physical constraint model and the historical behavior database; The fault-tolerant execution and reverse verification module previews the drainage instruction execution effect through the digital twin model and performs reverse verification with the actual drainage result; The central controller generates drainage instructions according to the output signals of the dynamic trust chain fusion decision module, the multimodal cross-validation module and the fault-tolerant execution and reverse verification module, and controls the action of the actuator.

2. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 1 is characterized by: The dynamic trust chain fusion decision module includes: The environmental sensing unit is connected to the ship's attitude sensor and environmental monitoring sensor to obtain the ship's roll angular velocity, pitch angle and surge period in real time; The trust chain reconstruction unit is connected to the liquid level sensor, flow rate sensor and pressure sensor, and is used to perform time domain decomposition and frequency domain filtering on the sensor signals; The weight dynamic allocation engine receives the output signal of the environmental perception unit and the filtered data of the trust chain reconstruction unit, and generates the dynamic confidence weight of each sensor.

3. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 2 is characterized by: The environment sensing unit generates a dynamic attenuation coefficient of the liquid level sensor based on the correlation between the ship's roll angular velocity and the surge period, specifically: When the ship's rolling angular velocity exceeds the preset threshold, the dynamic attenuation coefficient of the liquid level sensor decreases according to a linear function; When the surge period coincides with the ship's natural rocking frequency, the credibility threshold of the flow velocity sensor is increased to a preset safety value.

4. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 2 is characterized by: The trust chain reconstruction unit includes: The liquid level signal processing subunit uses the wavelet packet decomposition algorithm to separate the mechanical vibration interference component in the liquid level sensor signal; The pressure signal processing subunit uses a sliding window entropy calculation method to identify transient abnormal pulses in the pressure sensor data, where the transient abnormal pulses are caused by the bursting of bubbles in the drainage pipe.

5. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 2 is characterized by: The dynamic weight distribution engine performs the following operations under the high-frequency hull swaying condition: Adjust the dynamic confidence weight of the sidewall liquid level sensor to the range of 0.2-0.3; Increase the dynamic confidence weight of the drain pipe pressure sensor to the range of 0.7-0.8; Based on the majority voting results of the majority voting mechanism of at least three sensors in the same monitoring area, the final output value of the dynamic confidence weight is corrected.

6. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 1, characterized in that: The multimodal cross-validation module includes: The first-level screening unit verifies the logical consistency between the liquid level sensor data and the ship's draft and surge direction based on the fluid dynamics model; The second-level screening unit uses a long-short-term memory network to analyze the historical behavior patterns of sensor data and mark condition-sensitive sensors with frequent abnormalities; When the liquid level sensor reports a sudden increase in water level and the flow rate sensor data remains unchanged, the first-level screening unit triggers the delayed response protection mechanism, suspends the output of the drainage command and starts the data review process for three consecutive sampling cycles.

7. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 6 is characterized by: After detecting the condition-sensitive sensor, the second-level screening unit performs the following operations: Call the historical data of other sensors in the same monitoring area to calculate the current true value; Overwriting abnormal output data of the operating condition-sensitive sensor with an estimated value; Generate sensor fault diagnosis signals and upload them to the central controller.

8. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 1 is characterized by: The fault-tolerant execution and reverse verification module includes: The virtual sandbox unit simulates the ship's posture changes and the water level fluctuation curve in the cabin within 10 seconds after the drainage command is executed through a digital twin model; The reverse comparison unit collects the actual flow data of the drainage pipe pressure sensor in real time and calculates the difference with the predicted value of the virtual sandbox unit; If the difference exceeds 5% for three consecutive times, the actuator is judged to be abnormal, the backup drainage pipeline is switched and an alarm signal is triggered.

9. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 1, characterized in that: The central controller executes the following logic before generating a drainage instruction: Receive the dynamic confidence weight allocation result of the dynamic trust chain fusion decision module; Receive a two-stage screening pass mark from the multimodal cross-validation module; receiving a preview deviation detection result from the virtual sandbox unit; When the above three inputs all meet the preset conditions, a drainage pump start command and a valve opening control signal are generated.

10. The ship bulwark drainage intelligent management system combined with multi-sensor data monitoring according to claim 5, characterized in that: The triggering conditions of the majority voting mechanism include: The sum of the dynamic confidence weights of sensors in the same monitoring area is lower than the preset reliability threshold; The difference between the data reported by at least two sensors exceeds the maximum allowable error range; The system automatically selects the majority voting result as the final decision basis and marks the low-weight sensors for offline calibration.

Images