Method for evaluating sealing state of underwater robot

Abstract

The invention provides an underwater robot sealing state evaluation method, and relates to the technical field of information, and the method comprises the steps: collecting depth change data and water pressure gradient fluctuation information during the operation of an underwater robot, carrying out the high-frequency sampling of the position of a sealing joint, obtaining the change trend of environment parameters around a sealing surface, and calculating the sealing state of the underwater robot; obtaining a preliminary environment dynamic distribution result; performing filtering and feature extraction on the water pressure gradient fluctuation data according to the preliminary environment dynamic distribution result to obtain dynamic response data of the stress state of the sealing surface, and analyzing the water pressure fluctuation frequency and fluctuation amplitude to determine periodic features of gap deformation of the sealing surface; generating a dynamic strategy adjustment instruction according to the abnormal state distribution of the sealing joint, and obtaining operation parameter configuration after the adjustment instruction is executed; and synchronously transmitting the adjusted operation parameter configuration and the sealing joint state monitoring data to an underwater robot control system, and updating the operation load distribution of the sealing joint in real time to obtain optimized equipment operation state data.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for evaluating the sealing status of an underwater robot. Background Technology

[0002] Underwater robots, as crucial tools for deep-sea exploration and operations, possess irreplaceable value in fields such as marine resource development, environmental monitoring, and national defense. Their performance stability and reliability directly impact mission success, especially in the complex and ever-changing deep-sea environment, where the condition of the hull's sealing joints becomes a critical factor in ensuring safe operation. These sealing joints must withstand extreme water pressure and adapt to environmental changes caused by frequent depth adjustments, placing extremely high demands on technological research. However, current methods for assessing the condition of underwater robot sealing joints still have significant shortcomings. Many approaches often neglect the comprehensive impact of the dynamic environment during depth changes, focusing more on static performance under single operating conditions, while failing to fully consider the continuous effect of water pressure gradient changes on the sealing structure. This neglect leads to discrepancies between assessment results and actual operating conditions, particularly when the robot frequently rises and falls, making it difficult to accurately capture real-time changes in the sealing structure, thus affecting the equipment's safety and lifespan. Against this backdrop, the core challenges facing research are becoming increasingly apparent. Fluctuations in water pressure gradients caused by depth changes exert periodic effects on the sealing structure, leading to repeated deformation of the sealing surface gap. This deformation further triggers the continuous migration of stress concentration points within the sealing material, significantly increasing the risk of material fatigue and damage. Due to the lack of precise monitoring and analysis of this dynamic process, real-time assessment of the sealing condition is often delayed, failing to reflect potential problems in a timely manner and hindering effective maintenance decisions. Therefore, how to construct a real-time assessment system and dynamically adjust its strategies to address the dynamic characteristics of periodic deformation of the sealing surface gap and stress migration within the material in the complex environment of frequent depth adjustments by underwater robots has become a key issue in ensuring sealing performance and extending service life. Summary of the Invention

[0003] To address the above problems, this invention provides a method for evaluating the sealing status of an underwater robot, mainly comprising: Data on depth changes and water pressure gradient fluctuations during the operation of the underwater robot are collected. Sampling is performed at the sealing joint location to obtain the changing trends of environmental parameters around the sealing surface, resulting in dynamic environmental distribution data. Based on this dynamic distribution data, signal processing is performed on the water pressure gradient fluctuation information to extract dynamic response data of the sealing surface's stress state. The frequency and amplitude of water pressure fluctuations are analyzed to determine the periodic characteristics of the sealing surface gap deformation. These periodic characteristics are input into a pre-established simulation model to calculate the sealing surface gap deformation. The range of internal stress migration in the material is determined based on stress distribution and material properties. Based on this stress migration range, combined with historical operating data and material fatigue characteristic data, the stress concentration factor of key parts of the sealing joint is calculated. Risk classification is performed based on the stress concentration factor and fatigue life curve to obtain the degree of damage accumulation of the sealing joint under the current environment. High-risk sealing areas are identified based on the degree of damage accumulation, and the abnormal state distribution of the sealing joint is determined. Dynamic adjustment instructions are generated based on the abnormal state distribution to obtain the adjusted operating parameter configuration. The operating parameter configuration and the sealing joint status monitoring data are transmitted to the control system to update the operating load allocation, resulting in optimized equipment operating status data.

[0004] Furthermore, the process involves collecting depth change data and water pressure gradient fluctuation information during the underwater robot's operation, sampling at the sealing joint location, obtaining the environmental parameter change trends around the sealing surface, and acquiring dynamic environmental distribution data, including: The underwater robot continuously collects depth change data using a sensor array. The pressure difference between adjacent sampling points is calculated by dividing the time interval to obtain the pressure change rate. Based on the correspondence between the pressure change rate and depth change, the pressure distribution data at the sealing joint is determined. High-frequency sampling is performed on the sensors at the sealing joint location to acquire pressure fluctuation data around the sealing surface. A sliding window is used to process the pressure distribution data, calculating the standard deviation and dominant frequency of the pressure values ​​within the window to obtain spatiotemporal pressure change characteristics. Based on these characteristics, a three-dimensional pressure field is constructed around the sealing joint. The difference between pressure fields at adjacent times is calculated, and the pressure gradient vector field is extracted. Regions with modulus values ​​exceeding a threshold are marked as abnormal regions. Combining depth change data and the distribution of abnormal regions, dynamic environmental distribution data is obtained.

[0005] Furthermore, the step of processing the water pressure gradient fluctuation information based on the environmental dynamic distribution data to extract dynamic response data of the stress state of the sealing surface includes: The water pressure gradient fluctuation information in the environmental dynamic distribution data is filtered to remove high-frequency noise and obtain a smooth pressure time series signal. The time-frequency characteristics of the pressure time series signal are extracted to obtain the main frequency component and its corresponding amplitude. Based on the main frequency component and amplitude, the frequency change rate is calculated and converted to obtain the dynamic response data of the sealing surface.

[0006] Furthermore, the step of inputting the periodic characteristics of the sealing surface gap deformation into a pre-established simulation model, calculating the sealing surface gap deformation, and determining the range of internal stress migration in the material based on stress distribution and material properties includes: The periodic characteristics of the sealing surface gap deformation are input into the simulation model, material parameters are set, nodal displacements are calculated, and a simulated deformation sequence is obtained. The simulated deformation sequence is compared with the actual measured data, the mapping relationship between the deformation and water pressure fluctuation is fitted, the strain energy density is calculated, and the stress migration range is determined.

[0007] Furthermore, based on the stress migration range, combined with historical operating data and material fatigue characteristic data, the stress concentration factor of key parts of the sealing joint is calculated. Risk classification is then performed based on the stress concentration factor and fatigue life curve to obtain the degree of damage accumulation of the sealing joint under the current environment, including: Based on the stress migration range, nodes with stress values ​​exceeding a preset proportion are identified as critical parts. The maximum stress value and cycle number of the critical parts are extracted from historical operating data, and the stress concentration factor is calculated. The stress amplitude is corrected based on the stress concentration factor, the fatigue life curve is queried, the allowable number of cycles is determined, and the fatigue damage ratio is calculated. The risk level is determined by comparing the fatigue damage ratio with a threshold, and the sum of the products of the risk level assignment of each critical part and the fatigue damage ratio is calculated to obtain the degree of damage accumulation of the sealing joint in the current environment.

[0008] Furthermore, the step of identifying high-risk sealing areas based on the degree of damage accumulation and determining the abnormal distribution of the sealing joint's condition includes: Based on the degree of damage accumulation, nodes exceeding the threshold are marked as high-risk points. Adjacent high-risk points are connected to form high-risk sealing areas, and a risk distribution map is constructed. The real-time pressure measurement values ​​of the high-risk areas in the risk distribution map are compared with the standard pressure values ​​to calculate the pressure deviation rate. The coordinates and data of deviation rates exceeding the standard are recorded to determine the abnormal distribution of the status.

[0009] Furthermore, the step of generating dynamic adjustment instructions based on the abnormal state distribution and obtaining the adjusted operating parameter configuration includes: Based on the abnormal distribution of the status, identify the area with the highest degree of abnormality and its pressure value, calculate the depth adjustment value, speed adjustment value and load distribution ratio; determine the target working depth based on the depth adjustment value, determine the maximum allowable speed in combination with the speed adjustment value, redistribute power, and form the operating parameter configuration.

[0010] Furthermore, the step of transmitting the configured operating parameters and the sealing joint status monitoring data to the control system, updating the operating load allocation, and obtaining optimized equipment operating status data includes: The operating parameter configuration and sealing joint status monitoring data are encoded into a data packet and transmitted to the control system to verify data integrity. The operating parameter configuration is parsed, the control module is updated, the pressure difference between the old and new is calculated, the load distribution is adjusted, the adjusted pressure and vibration data are collected, the uniformity index and performance index are calculated, and the optimized equipment operating status data is obtained.

[0011] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a method for assessing the sealing status of an underwater robot. It collects depth change and water pressure gradient fluctuation data using a sensor array, analyzes environmental parameters around the sealing surface, and obtains dynamic response data of the stress state of the sealing surface. Combined with a finite element simulation model, a mapping relationship between deformation and water pressure fluctuation is established, and stress concentration risk is assessed based on material fatigue characteristics to identify high-risk sealing areas. Based on the abnormal distribution of the sealing joint's state, this invention generates dynamic strategy adjustment instructions, optimizes operating parameter configuration, and updates the operating load distribution of the sealing joint in real time. This method can effectively monitor the sealing joint status of underwater robots, adjust operating strategies in a timely manner, improve equipment reliability and operational efficiency, extend the service life of the sealing joint, and provide a guarantee for the safe and stable operation of underwater robots. Attached Figure Description

[0012] Figure 1 This is a flowchart of a method for evaluating the sealing status of an underwater robot according to the present invention.

[0013] Figure 2 This is a schematic diagram of a method for evaluating the sealing status of an underwater robot according to the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0015] like Figure 1-2 This embodiment of a method for evaluating the sealing status of an underwater robot may specifically include: S101. Collect depth change data and water pressure gradient fluctuation information of the underwater robot during operation, perform high-frequency sampling at the sealing joint location, obtain the environmental parameter change trend around the sealing surface, and obtain preliminary environmental dynamic distribution results.

[0016] The sensor array continuously collects depth change data of the underwater robot at a preset sampling frequency, while simultaneously recording the water pressure gradient values ​​at each sensor node location. The pressure change rate is obtained by subtracting the pressure values ​​from adjacent sampling points and dividing by the time interval. Based on the correspondence between the pressure change rate and depth change, the pressure distribution data at the sealing joint at each moment is determined. High-density sensors deployed at the sealing joint location perform high-frequency sampling, exceeding a preset frequency threshold, to acquire minute pressure fluctuation data around the sealing surface. A fixed-duration sliding window is used to segment the pressure distribution data, calculating the standard deviation and dominant frequency of the pressure values ​​within each window to obtain the spatiotemporal pressure change characteristics around the sealing joint. Based on the spatial coordinates and corresponding pressure values ​​in the spatiotemporal pressure change characteristics, a three-dimensional pressure field around the sealing joint is constructed using Kriging interpolation. By calculating the difference in pressure fields at adjacent moments, a pressure gradient vector field is extracted. If the magnitude of the gradient vector exceeds a preset threshold, the area is marked as an abnormal region. Combining the depth change data and the distribution of marked abnormal regions, a preliminary dynamic environmental distribution result is obtained, including the pressure field distribution, gradient field distribution, and abnormal region markings.

[0017] Specifically, the arrangement of sensor arrays in underwater robot sealing monitoring directly affects the accuracy of data acquisition.

[0018] Specifically, the sensor array is typically arranged in a ring, forming a multi-layered monitoring network around the sealed joint. The inner sensors, closer to the sealing surface, are primarily responsible for capturing minute pressure changes; the outer sensors monitor the overall environmental pressure trend. This layered arrangement allows the system to simultaneously acquire both local details and overall trend information.

[0019] In one possible implementation, the calculation of the pressure change rate involves processing the pressure difference between adjacent sampling points. When the sensor collects data at a frequency of 100Hz, the pressure values ​​at two adjacent moments are P1 and P2, with a time interval of 0.01 seconds. The pressure change rate is then (P2-P1) / 0.01. This change rate reflects the dynamic load intensity borne by the sealing joint. A sharp increase in the change rate often indicates a potential problem with the sealing performance.

[0020] It should be noted that high-frequency sampling is of particular significance in monitoring sealing joints. While conventional monitoring points may use a sampling frequency of 10Hz, sensors at sealing joint locations increase the frequency to 50Hz or higher. This differentiated sampling strategy is based on the characteristic that sealing failure often occurs instantaneously; high-frequency sampling can capture the complete process of pressure surges, providing ample data support for subsequent analysis. The sliding window method exhibits unique advantages in processing pressure fluctuation data.

[0021] For example, a window duration of 0.5 seconds is set, containing 50 sampling points, with the window sliding forward by 0.1 seconds each time. Within each window, the standard deviation of the pressure values ​​is calculated to characterize the fluctuation intensity, and the dominant frequency component is extracted using a Fast Fourier Transform. When the standard deviation suddenly increases within a window, and the dominant frequency shifts, it indicates that the sealing surface may have been subjected to abnormal disturbance during that period. Kriging interpolation plays a crucial role in constructing the three-dimensional pressure field. This method considers spatial correlation, extrapolating the pressure distribution around the entire sealing joint based on the pressure values ​​at known sensor locations. During interpolation, data from nearby sensors are given greater weight, ensuring the continuity and smoothness of the pressure field. The pressure field constructed in this way can visually display areas of concentrated pressure, providing a basis for identifying potential leak points. The extraction process of the pressure gradient vector field reveals the directional characteristics of pressure changes. In the three-dimensional pressure field, the gradient vector points in the direction of the fastest pressure increase, and its magnitude represents the rate of change. When a small gap appears on the sealing surface, the internal and external pressure difference will form a significant pressure gradient anomaly near the gap. By setting a gradient threshold, the system can automatically mark these abnormal areas, achieving precise location of weak points in the seal. Compared to simple pressure monitoring, this gradient analysis-based method can detect signs of sealing performance degradation earlier.

[0022] S102. Based on the preliminary environmental dynamic distribution results, the water pressure gradient fluctuation data is filtered and feature extracted to obtain the dynamic response data of the stress state of the sealing surface. The frequency and amplitude of the water pressure fluctuation are analyzed to determine the periodic characteristics of the sealing surface gap deformation.

[0023] Based on the preliminary environmental dynamic distribution results, a Butterworth filter was used to filter the water pressure gradient fluctuation data. A cutoff frequency was set as a preset threshold to remove high-frequency noise interference, obtaining a smooth pressure time-series signal. The time-frequency characteristics of the signal were extracted using a short-time Fourier transform to obtain the main frequency components and corresponding amplitudes of the sealing surface at different times. For the main frequency components and corresponding amplitudes, the rate of change of frequency within adjacent time periods was calculated. According to a pre-established amplitude-deformation mapping table, the amplitude data was multiplied by the mapping coefficient and a reference value was added to convert it into dynamic response data of the sealing surface. This dynamic response data characterizes the real-time deformation of the sealing surface under pressure. Autocorrelation was performed on the dynamic response data to calculate the correlation coefficient under different time delays. When a periodic peak appeared in the correlation coefficient, the time interval between adjacent peaks was determined as the time period of the sealing surface deformation. The maximum value of the dynamic response data within each period was found as the peak displacement. Based on the peak displacement values ​​at each monitoring point on the sealing surface, displacement boundary conditions are constructed, and the elasticity equation is solved using the finite element method to obtain the stress distribution pattern inside the sealing surface. The deformation period, peak displacement, and stress distribution pattern are then output as periodic characteristics of the sealing surface gap deformation.

[0024] Specifically, the Butterworth filter exhibits a smooth frequency response when processing water pressure gradient fluctuation data.

[0025] In one possible implementation, a fourth-order Butterworth low-pass filter is chosen, which has the flattest amplitude response within its passband. When the underwater robot operates at different depths, the water pressure signal is mixed with high-frequency turbulence noise and low-frequency actual pressure changes. By setting the cutoff frequency to 5Hz, disturbance signals above this frequency are effectively attenuated, while the actual pressure change signal experienced by the sealing surface is preserved. The application of short-time Fourier transform makes time-frequency feature extraction possible.

[0026] Specifically, 256 sampling points were selected as the window length, with a window overlap rate of 50%. Within each time window, a Fourier transform was performed on the pressure signal to obtain the spectral distribution for that period. By tracking the dominant frequency components at different times, the dynamic changes in the stress on the sealing surface could be observed. When the dominant frequency abruptly changes from 2Hz to 8Hz, it often indicates that the sealing surface is experiencing a severe pressure shock.

[0027] It should be noted that the amplitude-deformation mapping table is established based on the principles of materials mechanics and experimental calibration. The mapping coefficients reflect the elastic modulus characteristics of the sealing material, while the reference values ​​correspond to the initial pre-tightening state of the sealing surface.

[0028] For example, when the pressure amplitude is 100 kPa, the deformation calculated through the mapping relationship may be 0.2 mm. This linear mapping has high accuracy within the elastic deformation range of the sealing material. Autocorrelation calculations play a crucial role in identifying periodic patterns.

[0029] In one embodiment, autocorrelation calculation is performed on dynamic response data with a length of 1000 sampling points, with the delay time gradually increasing from 0 to 500 sampling points. The first correlation peak appears when the delay is 100 sampling points, indicating that the signal has a periodicity of 100 sampling points. Combined with the sampling frequency, the actual time period of the sealing surface deformation can be determined. The spatial distribution of the peak displacement reveals weak areas on the sealing surface.

[0030] For example, a peak displacement of 0.3 mm was detected at the 12 o'clock position of the sealing ring, while it was only 0.1 mm at the 6 o'clock position. This uneven distribution indicates that the upper sealing surface is subjected to a larger dynamic load. The finite element method requires reasonable mesh generation and boundary condition settings when calculating stress distribution. The sealing surface is discretized into triangular elements, and the element size is adaptively adjusted according to the displacement gradient. In areas with drastic displacement changes, the mesh is automatically refined to improve calculation accuracy. In solving the elasticity equations, the measured peak displacement is used as the displacement boundary condition, and the pressure difference between the inside and outside of the sealing surface is used as the load condition. Through iterative solution, the stress tensor of each node is obtained, and thus the principal stress distribution is derived. When the equivalent stress in a certain region approaches the yield strength of the material, that region becomes a potential failure initiation point. This method of inverting stress based on measured displacement reflects the true stress state of the sealing surface better than traditional pure theoretical calculations.

[0031] S103. Based on the periodic characteristics of the sealing surface gap deformation, input the deformation period and peak displacement into the pre-established finite element simulation model to simulate and calculate the sealing surface gap deformation. By comparing the simulation results with the actual measurement data, obtain the mapping relationship between the deformation and water pressure fluctuation. Based on the stress distribution pattern and the material elastic modulus, determine the possible stress migration range inside the material.

[0032] To address the periodic characteristics of the sealing surface gap deformation, the deformation period and peak displacement are input into a pre-established finite element simulation model. The elastic modulus and lateral deformation coefficient of the sealing material are set as material parameters. A transient dynamics solver is used to calculate the nodal displacements at each moment, obtaining a simulated deformation sequence of the sealing surface gap over time. Based on this simulated deformation sequence, the difference between its maximum and minimum values ​​is extracted as the simulated deformation amplitude. This amplitude is compared with the actual measured sealing surface deformation data, and the correlation coefficient between the two is calculated. If the correlation coefficient is greater than a preset threshold, the relationship between the simulated deformation and the corresponding water pressure fluctuation value is fitted using the least squares method to obtain the mapping relationship between the deformation and the water pressure fluctuation. Using this mapping relationship and the stress tensor of each node in the finite element simulation results, the strain energy density is calculated based on each component of the stress tensor and the elastic modulus. The strain energy density is equal to half the product of stress and strain. Regions where the difference in strain energy density between adjacent nodes exceeds a preset threshold are identified. Based on the principal stress direction in the stress distribution pattern, the path of stress concentration in this region propagating to the surrounding area is determined as the stress migration range within the material.

[0033] Specifically, the application of finite element simulation models in the deformation analysis of sealing surfaces requires precise parameter settings.

[0034] Specifically, the elastic modulus reflects a material's ability to resist elastic deformation; for rubber sealing materials, its value is typically in the range of 1-10 MPa. The lateral deformation coefficient, or Poisson's ratio, characterizes the degree of lateral expansion of a material under uniaxial compression. The Poisson's ratio of rubber materials is close to 0.5, indicating that they are approximately incompressible. These two parameters together determine the stress-strain relationship of the material. The transient dynamics solver works based on a discretized form of Newton's second law.

[0035] In one possible implementation, the solver discretizes the continuous time domain into multiple time steps, solving for the acceleration, velocity, and displacement of the nodes within each time step. When the input deformation period is 0.5 seconds and the peak displacement is 0.3 mm, the solver iterates at a time step of 0.001 seconds, ultimately outputting nodal displacement data for 500 time points, forming a complete sequence of simulated deformation.

[0036] It should be noted that the extraction process of the simulated deformation amplitude involves statistical analysis of the time-series data. The maximum value of 0.35 mm and the minimum value of 0.05 mm were identified from the deformation sequence; the difference of 0.3 mm is the simulated deformation amplitude. This amplitude reflects the maximum deformation range of the sealing surface within a complete cycle. Simultaneously, actual measurement data comes from displacement sensors installed near the sealing surface, which record the actual deformation of the sealing surface at the same sampling frequency. The correlation coefficient was calculated using the Pearson correlation coefficient method, with values ​​ranging from -1 to 1. When the correlation coefficient reaches 0.85 or higher, it indicates a strong correlation between the simulation results and the measured data, verifying the accuracy of the simulation model. Based on this strong correlation, the least squares method can establish a reliable mapping relationship.

[0037] For example, when the water pressure fluctuation is 50 kPa, the deformation calculated through the mapping relationship is 0.25 mm. This quantitative relationship provides a theoretical basis for subsequent real-time monitoring. The physical meaning of strain energy density is the elastic energy stored per unit volume.

[0038] In one embodiment, when the normal stress at a node is 2 MPa and the shear stress is 0.5 MPa, combined with the elastic modulus of 5 MPa, the strain energy density at that point can be calculated. The spatial distribution of strain energy density is non-uniform, and high strain energy regions are often locations of stress concentration. The identification of stress migration paths is based on the principle of minimum energy. When stress concentration occurs at a certain point on the sealing surface, the stress will be transmitted to the surrounding areas along the path of least resistance. By tracing the principal stress directions, the stress transmission path can be predicted.

[0039] For example, when a concentrated stress of 2.5 MPa occurs on the upper part of the sealing surface, the stress will be transmitted to both sides along a direction at a 45-degree angle to the sealing surface, forming a butterfly-shaped stress migration zone. This stress migration phenomenon can cause the stress level in the originally low-stress area to increase, potentially leading to new failure risks. By identifying these stress migration ranges in advance, targeted reinforcement measures can be taken during the design phase.

[0040] S104. Based on the stress migration range obtained from the simulation, combined with the cyclic load response data from the historical operating data and the material fatigue characteristic database, the stress concentration factor of the key parts of the sealing joint is obtained. The stress concentration risk is graded and assessed based on the stress concentration factor and the fatigue life curve, and the potential damage accumulation degree of the sealing joint under the current dynamic environment is obtained.

[0041] Based on the simulated stress migration range, nodes within this range where the stress value exceeds a preset proportion of the material's yield strength are identified as critical locations. The maximum stress value and stress cycle count for these locations over past operating cycles are extracted from historical operating data. The maximum stress value is divided by the average stress value far from the stress concentration area to obtain the stress concentration factor for each critical location. The actual stress amplitude is corrected using this stress concentration factor. The fatigue life curve for the corresponding material is retrieved from the material fatigue characteristic database. The allowable number of cycles is determined on the curve based on the corrected stress amplitude. The actual number of stress cycles is divided by the allowable number of cycles to obtain the fatigue damage ratio for each critical location. The fatigue damage ratio is compared with a preset risk level threshold. If the fatigue damage ratio is less than the first threshold, it is rated as low risk and assigned a value of 1; if it is between the first and second thresholds, it is rated as medium risk and assigned a value of 2; if it exceeds the second threshold, it is rated as high risk and assigned a value of 3. The product of the risk level assignment for each critical location and the corresponding fatigue damage ratio is accumulated to obtain the potential damage accumulation degree of the sealing joint under the current dynamic environment.

[0042] Specifically, the process of identifying the range of stress migration involves in-depth analysis of the simulation results.

[0043] In one possible implementation, when a concentrated load is applied to a point on the sealing surface, the stress propagates outwards in a wave-like manner, forming a gradually decaying stress field. By tracing the trajectory of the stress value decreasing outwards from the peak position, the region within the boundary where the stress drops to 30% of the peak value is the stress migration range. The material within this range experiences significant alternating stress and is a high-risk area for fatigue damage. The calculation of the stress concentration factor requires a clearly defined reference benchmark.

[0044] Specifically, the average stress value far from the stress concentration area is usually taken from a region on the sealing surface where the stress distribution is relatively uniform, and these regions are at least three times the feature size away from the geometric discontinuity.

[0045] For example, when the maximum stress at the corner of the sealing groove reaches 3 MPa, while the far-field average stress is 1 MPa, the stress concentration factor at that location is 3. This factor directly reflects the degree to which the geometry affects the amplification of local stress.

[0046] It should be noted that the value of historical operational data lies in providing accurate information about the load history. When underwater robots operate at different depths, the pressure cycles experienced by the sealing joints exhibit clear patterns. Statistical analysis reveals that the pressure cycle amplitude is smaller but the frequency is higher in shallow water, while the opposite is true in deep water. This difference in load characteristics directly affects the rate of fatigue damage accumulation. The fatigue life curve is an important characterization of the fatigue properties of materials.

[0047] In one embodiment, the curve is presented as a descending straight line in a logarithmic coordinate system, with the horizontal axis representing the number of cycles and the vertical axis representing the stress amplitude. When the stress amplitude is 2 MPa, the curve indicates an allowable number of cycles of 100,000; however, when the stress amplitude increases to 3 MPa, the allowable number of cycles drops sharply to 20,000. This nonlinear relationship illustrates the significant impact of stress level on fatigue life. The fatigue damage ratio is physically the ratio of actual damage to critical damage. According to the linear cumulative damage theory, fatigue failure occurs when this ratio reaches 1.

[0048] For example, if a critical component has undergone 80,000 stress cycles, while the permissible number of cycles under the current stress level is 100,000, then the fatigue damage ratio is 0.8, indicating that this component has consumed 80% of its fatigue life. Quantifying the risk level transforms qualitative assessments into calculable numerical values.

[0049] Specifically, a low-risk level of 1 indicates that the area still has sufficient safety margin; a medium-risk level of 2 suggests the need for enhanced monitoring; and a high-risk level of 3 warns of possible near-term failure. By multiplying the risk level by the fatigue damage ratio, both the absolute degree of damage and the relative severity of the risk are considered.

[0050] S105. Identify high-risk sealing areas based on the potential damage accumulation of the sealing joint under the current dynamic environment. Obtain the real-time status of the sealing joint by comparing the standard operating parameters with the current measurement values. Determine the abnormal distribution of the sealing joint status by analyzing the damage differences of each sealing area.

[0051] Based on the potential damage accumulation of the sealing joint under the current dynamic environment, sealing surface nodes with damage accumulation exceeding a preset threshold are marked as high-risk points. The spatial coordinates and corresponding damage values ​​of these high-risk points are obtained. Adjacent high-risk points are connected according to their spatial location to form high-risk sealing regions, constructing a risk distribution map that includes region boundaries and damage levels. For each high-risk sealing region marked in the risk distribution map, the real-time pressure measurement values ​​of the sensors within that region are read and compared with the standard pressure values ​​under design conditions to calculate the pressure deviation rate. If the deviation rate exceeds a preset allowable range, the coordinates and deviation rate of the measurement point are recorded as a real-time status dataset. Based on the spatial coordinates of each measurement point in the real-time status dataset, the difference in damage accumulation between any two measurement points is calculated and divided by the distance between the two points to obtain the damage change rate. Regions with damage change rates exceeding a preset threshold are spatially superimposed with regions with excessive pressure deviation rates to form an abnormal distribution of the sealing joint status, including uneven damage distribution and real-time pressure anomalies.

[0052] Specifically, the process of identifying high-risk sealing areas is based on the spatial distribution characteristics of the degree of damage accumulation.

[0053] In one possible implementation, when the cumulative damage level of a sealing surface node reaches 0.7, and the preset threshold is 0.6, that node is marked as a high-risk point. These discrete high-risk points are often not isolated; there is an inherent connection between adjacent high-risk points in terms of stress transfer and damage propagation. By connecting high-risk points with a distance of less than 5 mm, a continuous high-risk region is formed. This connection is not a simple geometric line but is based on the principle of continuity of the damage field. The construction of the risk distribution map needs to comprehensively consider both spatial location and damage degree dimensions.

[0054] Specifically, a color-coding method is used to map the degree of damage to different color depths, with 0.7 corresponding to dark red, 0.6 to orange, and so on, decreasing in intensity. Area boundaries are defined by iso-damage lines, similar to contour lines on a topographic map. This visual representation allows maintenance personnel to intuitively identify weak points in the sealing joints.

[0055] It should be noted that comparing real-time pressure measurements with standard pressure values ​​is of significant diagnostic value. Standard pressure values ​​are derived from design conditions and represent the pressure the sealing joint should withstand under ideal conditions.

[0056] For example, at a water depth of 100 meters, the standard pressure value is 1.1 MPa. When the real-time measured value is 1.25 MPa, the pressure deviation rate is calculated as (1.25 - 1.1) / 1.1 × 100% = 13.6%. This deviation may originate from factors such as water flow disturbance, sealing surface deformation, or material aging. The allowable range for the pressure deviation rate needs to take into account the elastic limit and safety factor of the sealing material.

[0057] In one embodiment, the allowable range is set to ±10%, exceeding this range indicates that the sealing joint is under abnormal load. The real-time status dataset not only records the location of the exceedance points but also includes the direction and magnitude of the deviation; positive deviation indicates excessive pressure, while negative deviation indicates insufficient pressure, and the two have different mechanisms of influence on sealing performance. The calculation of the damage change rate reveals the spatial gradient distribution of damage.

[0058] For example, if the cumulative damage at point A is 0.8 and at point B it is 0.6, and the two points are 10 mm apart, then the damage change rate is (0.8-0.6) / 10 = 0.02 / mm. Areas with high damage change rates are often the forefront of stress concentration and damage propagation. Although these areas may not currently have the highest damage level, their damage development speed is the fastest. The spatial superposition process achieves the fusion of multi-source information. When areas with excessive damage change rates and areas with excessive pressure deviations spatially overlap, it indicates that the area has both structural damage and is subjected to abnormal loads, making it the most likely location for failure. This superposition is not a simple Boolean operation but considers the severity and mutual influence of the two anomalies. The resulting state anomaly distribution map comprehensively reflects the structural damage and load anomalies of the sealing joint, providing a scientific basis for precise maintenance.

[0059] S106. Generate dynamic strategy adjustment instructions based on the abnormal distribution of the sealing joint status, and obtain the operating parameter configuration after executing the adjustment instructions.

[0060] Based on the abnormal distribution of the sealing joint's condition, the area with the highest degree of abnormality and its corresponding pressure value are identified. If this pressure value exceeds a preset proportion of the design pressure, the difference between the current depth and the safe depth is calculated, generating an adjustment value to reduce the operating depth. Simultaneously, an adjustment value for the operating speed and a load redistribution ratio are generated based on the spatial offset of the abnormal distribution. Using the adjusted operating depth value, the current depth is subtracted to obtain the adjusted depth. The maximum permissible speed at this depth is determined according to the depth-speed correspondence. The adjusted operating speed value is added to the current speed and compared with the maximum permissible speed; the smaller of the two is taken as the adjusted operating speed. Using the load redistribution ratio, the total power of the underwater robot is redistributed to each actuator according to this ratio, forming an operating parameter configuration that includes the adjusted depth, adjusted operating speed, and power allocation values. The adjusted depth serves as the target operating depth, the adjusted operating speed as the maximum permissible speed, and the power allocation values ​​constitute the power allocation scheme for each system.

[0061] Specifically, there is a direct mapping relationship between the distribution of abnormal states and the generation of adjustment instructions.

[0062] In one possible implementation, a concentrated high-damage anomaly in the upper region of the sealing joint indicates that the area is under excessive pressure load. If the highest pressure value identified at this point reaches 120% of the design pressure, it means that the water pressure at the current operating depth has exceeded the safe range of the seal design. By calculating the difference between the current depth of 150 meters and the corresponding safe depth of 125 meters at 120% pressure, an adjustment value of 25 meters is obtained. The concept of spatial offset reflects the asymmetric nature of the anomaly distribution.

[0063] Specifically, if the abnormal area is concentrated on one side of the sealing ring, forming an eccentric distribution, this offset can be quantified by calculating the distance between the center of gravity of the abnormal area and the geometric center of the sealing ring. When the offset reaches 5mm, it indicates that the load distribution is severely uneven, and the operating posture needs to be adjusted to rebalance the load. The adjustment value of the operating speed is proportional to the offset; the larger the offset, the greater the speed reduction is required to reduce the impact of dynamic loads.

[0064] It should be noted that the depth-velocity relationship is based on fluid dynamics principles. As the operating depth increases, the density and viscosity of water change, correspondingly increasing the resistance to robot movement. At a depth of 100 meters, the maximum permissible speed may be 2 m / s, while at a depth of 150 meters, it drops to 1.5 m / s. This relationship was determined through a combination of experimental testing and theoretical calculations, ensuring the robot's safe and stable operation at different depths. The core of load redistribution lies in changing the power distribution ratio of each actuator.

[0065] For example, when an abnormality is detected in the left-side seal, the originally evenly distributed propulsion power needs to be adjusted. The power of the right-side thruster increases to 60% of the total power, while the power of the left-side thruster decreases to 40%. This asymmetrical propulsion generates torque to counteract the effects of the abnormal load. This dynamic adjustment is not fixed but changes according to the real-time variations in the abnormal distribution. The formation of the power distribution scheme requires consideration of multiple constraints.

[0066] In one embodiment, while keeping the total power constant, the power adjustment of each actuator must meet the maneuverability requirements. If the power of a certain thruster drops too low, it will affect the robot's maneuverability; if it rises too high, it may exceed the equipment's rated power. An optimization algorithm finds the optimal allocation scheme under these constraints, mitigating sealing anomalies while maintaining basic mobility. The final operating parameter configuration is a comprehensive solution. The target operating depth not only considers the current anomaly but also reserves a safety margin; the maximum permissible speed ensures that excessive dynamic pressure is not generated at the adjusted depth; and the power allocation scheme reduces pressure in the anomaly area by changing the load distribution. These three parameters are interrelated and work together to form a dynamically adaptive control strategy, enabling the underwater robot to continue performing its mission while protecting the sealing joint.

[0067] S107. The adjusted operating parameter configuration and the sealing joint status monitoring data are synchronously transmitted to the underwater robot control system to update the operating load distribution of the sealing joint in real time and obtain the optimized equipment operating status data.

[0068] The adjusted operating parameters and sealing joint status monitoring data are encoded into data packets and transmitted to the underwater robot control system via a communication link. A checksum is used to verify data integrity, and a successful transmission identifier is obtained upon successful verification. Based on the successful transmission identifier, the control system parses the target operating depth, maximum permissible speed, and power distribution scheme from the data packet, updates these parameter values ​​to the corresponding control modules, and simultaneously reads the real-time pressure values ​​at each monitoring point of the sealing joint, calculating the difference between the old and new pressure values. The load adjustment amount is determined based on this pressure difference, and the output power of each propulsion device is adjusted according to the power distribution scheme to redistribute the load on the sealing joint. Pressure and vibration data are collected within a preset time period after the power adjustment, and the collected data is compared and analyzed with the data before adjustment. The standard deviation of the pressure distribution is calculated as a uniformity index, and the ratio of the peak pressure to the design value is used as a performance index, resulting in optimized equipment operating status data that includes both uniformity and performance indicators.

[0069] Specifically, the packet encoding process involves key steps in uniformly formatting heterogeneous data.

[0070] In one possible implementation, the operating parameter configuration includes numerical data such as the target operating depth of 125 meters, the maximum permissible speed of 1.5 m / s, and the power distribution ratio of 60:40, while the sealing joint status monitoring data includes pressure time-series data from each monitoring point. During encoding, the numerical parameters are first converted into fixed-length binary representations, then the time-series data is compressed and encoded, and finally, a data type identifier header and a timestamp are added to form a structured data packet. The checksum is used to ensure data integrity during transmission.

[0071] Specifically, the CRC-16 checksum algorithm is used to perform a polynomial operation on the entire data packet, generating a 16-bit checksum that is appended to the end of the packet. The receiving end performs the same operation and compares the result with the received checksum. If they match, the data is considered complete and error-free. This mechanism is particularly important because the underwater communication environment is complex, and signal attenuation and noise interference can lead to data errors.

[0072] It's important to note that updating the control module's parameters is not a simple numerical replacement. When a new target operating depth is reached, the depth control module calculates the difference between the current depth and the target depth, generating control commands for ascent or descent. The speed control module sets a speed limit based on the maximum permissible speed to prevent overspeeding. The power distribution module allocates the total power to the left and right thruster groups in a 60:40 ratio, controlling the actual output of each thruster through pulse-width modulation signals. The calculation of the pressure difference reveals the actual effect of the parameter adjustments.

[0073] For example, the pressure at a monitoring point was 1.8 MPa before adjustment and dropped to 1.5 MPa after adjustment, a difference of 0.3 MPa. This difference not only reflects the change at a single point, but more importantly, by analyzing the distribution of differences across multiple points, it can be determined whether the load redistribution has achieved the expected result. If the pressure at all monitoring points decreases, it indicates a reduction in the overall load; if the pressure decreases at some points while increasing at others, it indicates a load shift. The standard deviation of pressure distribution serves as a uniformity indicator with a clear physical meaning.

[0074] In one embodiment, assuming there are 8 monitoring points on the sealing ring, the pressure values ​​before adjustment are 1.8, 2.1, 1.6, 2.2, 1.7, 2.0, 1.9, and 2.3 MPa, with a calculated standard deviation of 0.23 MPa. After adjustment, the pressure values ​​become 1.7, 1.8, 1.6, 1.8, 1.7, 1.8, 1.7, and 1.9 MPa, and the standard deviation decreases to 0.09 MPa. The significant reduction in standard deviation indicates a more uniform pressure distribution, which is beneficial for extending the seal life. The performance indicators are calculated based on the comparison between the peak pressure and the design value. If the seal design pressure is 2.0 MPa, the peak pressure before adjustment is 2.3 MPa, with a ratio of 1.15, indicating a 15% overload; after adjustment, the peak pressure drops to 1.9 MPa, with a ratio of 0.95, which is within the safe range. This quantitative evaluation method allows the system to objectively judge the optimization effect, providing data support for subsequent maintenance decisions. By continuously monitoring the changing trends of these indicators, signs of seal performance degradation can be detected in a timely manner.

[0075] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for evaluating the sealing status of an underwater robot, characterized in that, The method includes: Data on depth changes and water pressure gradient fluctuations during the operation of the underwater robot are collected. Sampling is performed at the sealing joint location to obtain the changing trends of environmental parameters around the sealing surface, resulting in dynamic environmental distribution data. Based on this dynamic distribution data, signal processing is performed on the water pressure gradient fluctuation information to extract dynamic response data of the sealing surface's stress state. The frequency and amplitude of water pressure fluctuations are analyzed to determine the periodic characteristics of the sealing surface gap deformation. These periodic characteristics are input into a pre-established simulation model to calculate the sealing surface gap deformation. The range of internal stress migration in the material is determined based on stress distribution and material properties. Based on this stress migration range, combined with historical operating data and material fatigue characteristic data, the stress concentration factor of key parts of the sealing joint is calculated. Risk classification is performed based on the stress concentration factor and fatigue life curve to obtain the degree of damage accumulation of the sealing joint under the current environment. High-risk sealing areas are identified based on the degree of damage accumulation, and the abnormal state distribution of the sealing joint is determined. Dynamic adjustment instructions are generated based on the abnormal state distribution to obtain the adjusted operating parameter configuration. The operating parameter configuration and the sealing joint status monitoring data are transmitted to the control system to update the operating load allocation, resulting in optimized equipment operating status data.

2. The underwater robot sealing condition assessment method according to claim 1, characterized in that, The process involves collecting depth change data and water pressure gradient fluctuation information during the underwater robot's operation, sampling at the sealing joint location, obtaining the environmental parameter change trends around the sealing surface, and acquiring dynamic environmental distribution data, including: The underwater robot continuously collects depth change data using a sensor array. The pressure difference between adjacent sampling points is calculated by dividing the time interval to obtain the pressure change rate. Based on the correspondence between the pressure change rate and depth change, the pressure distribution data at the sealing joint is determined. High-frequency sampling is performed on the sensors at the sealing joint location to acquire pressure fluctuation data around the sealing surface. A sliding window is used to process the pressure distribution data, calculating the standard deviation and dominant frequency of the pressure values ​​within the window to obtain spatiotemporal pressure change characteristics. Based on these characteristics, a three-dimensional pressure field is constructed around the sealing joint. The difference between pressure fields at adjacent times is calculated, and the pressure gradient vector field is extracted. Regions with modulus values ​​exceeding a threshold are marked as abnormal regions. Combining depth change data and the distribution of abnormal regions, dynamic environmental distribution data is obtained.

3. The underwater robot sealing condition assessment method according to claim 1, characterized in that, The step of processing the water pressure gradient fluctuation information based on the environmental dynamic distribution data and extracting the dynamic response data of the stress state of the sealing surface includes: The water pressure gradient fluctuation information in the environmental dynamic distribution data is filtered to remove high-frequency noise and obtain a smooth pressure time series signal. The time-frequency characteristics of the pressure time series signal are extracted to obtain the main frequency component and its corresponding amplitude. Based on the main frequency component and amplitude, the frequency change rate is calculated and converted to obtain the dynamic response data of the sealing surface.

4. The underwater robot sealing condition assessment method according to claim 1, characterized in that, The process of inputting the periodic characteristics of the sealing surface gap deformation into a pre-established simulation model, calculating the sealing surface gap deformation, and determining the range of internal stress migration in the material based on stress distribution and material properties includes: The periodic characteristics of the sealing surface gap deformation are input into the simulation model, material parameters are set, nodal displacements are calculated, and a simulated deformation sequence is obtained. The simulated deformation sequence is compared with the actual measured data, the mapping relationship between the deformation and water pressure fluctuation is fitted, the strain energy density is calculated, and the stress migration range is determined.

5. The underwater robot sealing condition assessment method according to claim 1, characterized in that, Based on the stress migration range, combined with historical operating data and material fatigue characteristic data, the stress concentration factor of key parts of the sealing joint is calculated. Risk classification is then performed based on the stress concentration factor and fatigue life curve to obtain the degree of damage accumulation of the sealing joint under the current environment, including: Based on the stress migration range, nodes with stress values ​​exceeding a preset proportion are identified as critical parts. The maximum stress value and cycle number of the critical parts are extracted from historical operating data, and the stress concentration factor is calculated. The stress amplitude is corrected based on the stress concentration factor, the fatigue life curve is queried, the allowable number of cycles is determined, and the fatigue damage ratio is calculated. The risk level is determined by comparing the fatigue damage ratio with a threshold, and the sum of the products of the risk level assignment of each critical part and the fatigue damage ratio is calculated to obtain the degree of damage accumulation of the sealing joint in the current environment.

6. The underwater robot sealing condition assessment method according to claim 1, characterized in that, The step of identifying high-risk sealing areas based on the degree of damage accumulation and determining the abnormal distribution of the sealing joint's condition includes: Based on the degree of damage accumulation, nodes exceeding the threshold are marked as high-risk points. Adjacent high-risk points are connected to form high-risk sealing areas, and a risk distribution map is constructed. The real-time pressure measurement values ​​of the high-risk areas in the risk distribution map are compared with the standard pressure values ​​to calculate the pressure deviation rate. The coordinates and data of deviation rates exceeding the standard are recorded to determine the abnormal distribution of the status.

7. The underwater robot sealing condition assessment method according to claim 1, characterized in that, The step of generating dynamic adjustment instructions based on the abnormal state distribution and obtaining the adjusted operating parameter configuration includes: Based on the abnormal distribution of the status, identify the area with the highest degree of abnormality and its pressure value, calculate the depth adjustment value, speed adjustment value and load distribution ratio; determine the target working depth based on the depth adjustment value, determine the maximum allowable speed in combination with the speed adjustment value, redistribute power, and form the operating parameter configuration.

8. The underwater robot sealing condition assessment method according to claim 1, characterized in that, The process of transmitting the configured operating parameters and the sealing joint status monitoring data to the control system, updating the operating load allocation, and obtaining optimized equipment operating status data includes: The operating parameter configuration and sealing joint status monitoring data are encoded into a data packet and transmitted to the control system to verify data integrity. The operating parameter configuration is parsed, the control module is updated, the pressure difference between the old and new is calculated, the load distribution is adjusted, the adjusted pressure and vibration data are collected, the uniformity index and performance index are calculated, and the optimized equipment operating status data is obtained.

Images