Method and device for detecting a net cage structure breakage

Abstract

The present application relates to the technical field of ocean engineering, and particularly relates to a net cage structure damage detection method and device. The method comprises the following steps: acquiring six-degree-of-freedom data, netting displacement data and environmental data of the net cage; preprocessing the six-degree-of-freedom data, the netting displacement data and the environmental data to obtain preprocessed data; extracting features from the preprocessed data to obtain anchor chain features and netting features; inputting the anchor chain features and the netting features into a net cage structure damage detection model to obtain a first net cage structure damage result; wherein the first net cage structure damage result comprises an anchor chain state result and a netting state result; and performing joint verification based on the first net cage structure damage result to obtain a second net cage structure damage result. In this way, the present application can improve the detection accuracy of net cage structure damage.

Description

Technical Field

[0001] This invention relates to the field of marine engineering technology, and in particular to a method and apparatus for detecting damage to net cage structures. Background Technology

[0002] With the rapid development of the deep-sea aquaculture industry, truss cages have become core equipment for offshore aquaculture due to their strong resistance to wind and waves and large aquaculture capacity. However, in complex marine environments, their anchor chains and nets face the risk of damage: when sea states reach level 4-6 (corresponding to wave heights of 2-6m and current velocities of 1-1.5m / s), the alternating stress on the anchor chains can reach 2-3 times that of the static load, and the probability of fatigue failure of anchor chains in single-point mooring systems is more than 30% higher than that in multi-point mooring systems; the nets are subject to long-term wave impact, attachment of marine organisms (such as shellfish attachment, which can increase net resistance by 40%), and collisions with aquaculture objects, and are more prone to tearing under severe weather conditions such as typhoons.

[0003] Currently, the marine environment is highly corrosive and subject to large instantaneous loads, resulting in low reliability of traditional sensors. For example, underwater strain gauges have a 30%-50% failure rate within 3 months in salt spray environments. Secondly, the scale of net cages is large (the perimeter of a single truss net cage often exceeds 50m), making single-point monitoring coverage incomplete. Manual inspections require the suspension of aquaculture operations and cannot be carried out in winds and waves of level 6 or higher. Thirdly, net cages are affected by wave currents and exhibit 6 degrees of freedom motion, including swaying and transverse swaying. The noise signals generated can easily mask structural damage characteristics, leading to a high false alarm rate for traditional monitoring methods.

[0004] Based on this, the present invention proposes a method and device for detecting damage to cage structures to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention describes a method and apparatus for detecting damage to cage structures, which can improve the detection accuracy of cage structure damage.

[0006] According to a first aspect, the present invention provides a method for detecting damage to a cage structure, comprising:

[0007] Acquire six degrees of freedom data of the net cage, netting displacement data, and environmental data;

[0008] The six-degree-of-freedom data, the mesh displacement data, and the environmental data are preprocessed to obtain preprocessed data.

[0009] The preprocessed data is subjected to feature extraction to obtain anchor chain features and netting features;

[0010] The anchor chain features and the netting features are input into the cage structure damage detection model to obtain the first cage structure damage result; wherein, the first cage structure damage result includes the anchor chain status result and the netting status result;

[0011] Based on the joint verification of the first cage structure damage result, the second cage structure damage result is obtained.

[0012] According to a second aspect, the present invention provides a detection device for damage to a cage structure, comprising:

[0013] The acquisition unit is configured to acquire six degrees of freedom data of the net cage, net displacement data, and environmental data;

[0014] The first data processing unit is configured to preprocess the six-degree-of-freedom data, the mesh displacement data, and the environmental data to obtain preprocessed data.

[0015] The second data processing unit is configured to extract features from the preprocessed data to obtain anchor chain features and netting features.

[0016] The third data processing unit is configured to input the anchor chain features and the netting features into the cage structure damage detection model to obtain the first cage structure damage result; wherein, the first cage structure damage result includes the anchor chain status result and the netting status result;

[0017] The fourth data processing unit is configured to perform joint verification based on the first cage structure damage result to obtain the second cage structure damage result.

[0018] Thirdly, embodiments of this specification also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the method described in any embodiment of this specification.

[0019] Fourthly, embodiments of this specification also provide a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the methods described in any embodiment of this specification.

[0020] According to the method and apparatus for detecting damage to a wire mesh cage structure provided by the present invention, sensors installed at preset positions on the wire mesh cage first acquire six-degree-of-freedom (6DOF) data, netting displacement data, and environmental data. The 6DOF data reflects the overall dynamic attitude of the wire mesh cage (covering sway, roll, heave, pitch, and head-on). The netting displacement data captures the displacement of local deformations of the netting (including real-time positional deviations and deformation amplitudes at each monitoring point). The environmental data characterizes external influencing factors. After data acquisition, three types of raw data need to be preprocessed: for the 6DOF data and netting displacement data, outlier removal (filtering instantaneous interference signals from sensors based on the 3σ criterion), data smoothing (suppressing high-frequency noise through moving averages), and spatiotemporal alignment (unifying the timestamps and spatial coordinates of multi-sensor data) are performed. For the environmental data, normalization is used to eliminate differences in the magnitude of different parameters, ultimately obtaining standardized, highly reliable preprocessed data. Subsequently, targeted feature extraction is performed based on the preprocessed data: dynamic features related to anchor chain stress (such as peak alternating stress, stress change frequency, and attitude change amplitude) are extracted from the six-degree-of-freedom data; deformation features related to the mesh structure (such as maximum tensile strength, deformation uniformity, and local wrinkle frequency) are extracted from the mesh displacement data. Simultaneously, feature thresholds are corrected by combining environmental data, forming anchor chain and mesh features that can be directly used for detection. The anchor chain and mesh features are input into the mesh cage structure damage detection model to obtain the first mesh cage structure damage result; the first mesh cage structure damage result includes anchor chain status results (such as normal, slight fatigue, severe damage) and mesh status results (such as normal, local tear, large-area damage). Considering the possible occasional errors in the output of a single model, joint verification is performed based on the first mesh cage structure damage result to obtain the second mesh cage structure damage result, ultimately yielding a second mesh cage structure damage result with higher accuracy. Thus, this invention can improve the detection accuracy of mesh cage structure damage. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating a method for detecting damage to a wire mesh cage structure according to one embodiment is shown.

[0023] Figure 2 A schematic block diagram of a detection device for cage structure damage according to one embodiment is shown;

[0024] Figure 3 A schematic diagram of the division of a single piece of mesh fabric area according to one embodiment is shown. Detailed Implementation

[0025] The solution provided by the present invention will now be described with reference to the accompanying drawings.

[0026] Figure 1 This diagram illustrates a flowchart of a method for detecting damage to a wire mesh cage structure according to one embodiment. It is understood that this method can be executed by any device, equipment, platform, or cluster of devices with computing and processing capabilities. Figure 1 As shown, the method includes:

[0027] Step 100: Obtain the six degrees of freedom data of the net cage, the netting displacement data, and the environmental data;

[0028] Step 102: Preprocess the six-degree-of-freedom data, netting displacement data, and environmental data to obtain preprocessed data;

[0029] Step 104: Extract features from the preprocessed data to obtain anchor chain features and netting features;

[0030] Step 106: Input the anchor chain features and netting features into the cage structure damage detection model to obtain the first cage structure damage result; wherein, the first cage structure damage result includes the anchor chain status result and the netting status result;

[0031] Step 108: Perform joint verification based on the damage results of the first cage structure to obtain the damage results of the second cage structure.

[0032] In this embodiment, sensors positioned at preset locations on the net cage first acquire six degrees of freedom (DOF) data, netting displacement data, and environmental data. The six DEF data reflects the overall dynamic attitude of the net cage (including sway, roll, heave, pitch, and head-up). The netting displacement data captures local deformation of the netting (including real-time positional deviations and deformation amplitudes at each monitoring point). The environmental data characterizes external influencing factors. After data acquisition, three types of raw data undergo preprocessing: for the six DEF and netting displacement data, outlier removal (filtering instantaneous interference signals from sensors based on the 3σ criterion), data smoothing (suppressing high-frequency noise through moving averages), and spatiotemporal alignment (unifying timestamps and spatial coordinates of multi-sensor data) are performed. For the environmental data, normalization eliminates differences in the magnitude of different parameters, ultimately yielding standardized, highly reliable preprocessed data. Subsequently, targeted feature extraction is performed based on the preprocessed data: dynamic features related to anchor chain stress (such as peak alternating stress, stress change frequency, and attitude change amplitude) are extracted from the six-degree-of-freedom data; deformation features related to the mesh structure (such as maximum tensile strength, deformation uniformity, and local wrinkle frequency) are extracted from the mesh displacement data. Simultaneously, feature thresholds are corrected by combining environmental data, forming anchor chain and mesh features that can be directly used for detection. The anchor chain and mesh features are input into the mesh cage structure damage detection model to obtain the first mesh cage structure damage result; the first mesh cage structure damage result includes anchor chain status results (such as normal, slight fatigue, severe damage) and mesh status results (such as normal, local tear, large-area damage). Considering the possible occasional errors in the output of a single model, joint verification is performed based on the first mesh cage structure damage result to obtain the second mesh cage structure damage result, ultimately yielding a second mesh cage structure damage result with higher accuracy. Thus, this invention can improve the detection accuracy of mesh cage structure damage.

[0033] In one embodiment of the present invention, the anchor chain features include the mean displacement offset rate and the coefficient of variation of the rotation period; the netting features include the peak displacement amplitude ratio and the neighborhood correlation coefficient; the environmental data include wind speed, wind direction, wave height, period, flow velocity, and flow direction; and the preprocessing includes standard time-series processing, data cleaning processing, and noise reduction processing.

[0034] In this embodiment, anchor chain features are used to characterize the dynamic mechanical state, specifically including the mean displacement deviation rate (reflecting the positional displacement trend of the anchor chain under long-term stress, which can help determine whether there is cumulative deformation) and the coefficient of variation of the rotation period (reflecting the stability of the anchor chain's swing pattern; abnormal fluctuations are often associated with fatigue damage). Netting features target local deformation characteristics, including the peak displacement amplitude ratio (comparing the maximum deformation differences in different areas of the netting to identify the risk of local tearing) and the neighborhood correlation coefficient (measuring the displacement coordination of adjacent monitoring points on the netting; low correlation may indicate deformation discontinuity caused by netting damage). Environmental data is further refined into wind speed, wind direction, wave height, period, current velocity, and current direction parameters to comprehensively capture the impact of the external marine environment on the net cage structure. The preprocessing stage involves standard time-series processing (unifying the time dimension of multi-source data to ensure temporal consistency), data cleaning (removing outliers caused by sensor malfunctions or interference), and noise reduction (suppressing interference from environmental noise and equipment noise on the effective signal).

[0035] In one embodiment of the present invention, the mean displacement offset rate and the coefficient of variation of the rotation period are determined by the following formula:

[0036]

[0037] In the formula, , , The displacement mean offset rate, The mean offset rate of cage sway. The mean offset rate of cage sway. The mean offset rate of cage heave. This represents the average displacement over the current 10 seconds. This represents the average displacement when the anchor chain is intact. , , The coefficient of variation of the rotation period. The coefficient of variation of the cage's sway period. This is the coefficient of variation of the cage's pitching period. The coefficient of variation of the bow rocking period of the net cage. The standard deviation of the corresponding rotation period, This is the average value of the corresponding rotation period.

[0038] In one embodiment of the present invention, the peak shift ratio and the neighborhood correlation coefficient are determined by the following formula:

[0039]

[0040] In the formula, The peak shift amplitude ratio, This represents the current peak displacement. This represents the peak displacement under normal sea conditions. The neighborhood correlation coefficient Let be the covariance of the two displacement parameters. These are the displacement parameters of two adjacent mesh sensors. The first standard deviation, This is the second standard deviation.

[0041] In one embodiment of the present invention, a joint verification is performed based on the damage result of the first cage structure to obtain the damage result of the second cage structure, including:

[0042] When both the anchor chain status result and the netting status result are damaged, the weight of the neighborhood correlation coefficient in the netting feature is increased to obtain the updated netting feature.

[0043] The updated mesh and anchor chain features are input into the cage structure damage detection model to obtain the second cage structure damage result.

[0044] In this embodiment, considering the mechanical relationship between the netting and the anchor chain (anchor chain damage easily leads to local stress concentration in the netting), the weight of the neighborhood correlation coefficient in the netting features is increased (this coefficient accurately reflects the deformation synergy of adjacent areas of the netting; increasing the weight allows for more sensitive capture of subtle features of netting damage), resulting in updated netting features. Subsequently, the updated netting features and the original anchor chain features (retaining the dynamic features of anchor chain damage to ensure the consistency of anchor chain status judgment) are input into the gabion structure damage detection model (when the neighborhood correlation coefficient > 0.8, the netting is considered not damaged). Thus, this invention effectively avoids the bias of single-feature judgment, ultimately outputting a second gabion structure damage result with higher accuracy and reliability, further reducing the false alarm rate and false negative rate of damage detection.

[0045] In one embodiment of the present invention, the mesh displacement data is obtained through the following steps:

[0046] The mesh of the cage is divided into regions to obtain multiple sub-regions of the mesh.

[0047] Obtain the set of displacements of the center points and boundary points of multiple sub-regions, and record it as the mesh displacement data;

[0048] Displacement sensors are installed at the center and boundary points of multiple sub-regions, and each displacement sensor is assigned a unique address.

[0049] In this embodiment, the netting displacement data is acquired through a data collection scheme, with the following steps: First, based on the overall size and structural characteristics of the netting, the area is divided according to the principle of "uniform coverage and emphasis on key areas." Using the center of the netting as the origin, equal-sized sub-regions are divided radially and circumferentially. The scale of these sub-regions is reduced for vulnerable areas such as edges and corners, forming multiple sub-regions with full coverage and denser coverage of key areas, ensuring comprehensive monitoring. Next, displacement monitoring points are deployed in each sub-region: one sensor is installed at the geometric center point to collect the overall average displacement of the sub-region; one sensor is installed at each of the four boundary vertices to capture local deformation displacement. All sensors are high-precision underwater type and are pre-assigned unique digital addresses containing sub-region numbers and point types, ensuring data traceability to the corresponding location.

[0050] like Figure 3 As shown, in this embodiment, the area is divided into 9 regions and 3 categories: main region, sub-region, and central region. Each region consists of 5 monitoring points, for a total of 25 monitoring points. The damage area is determined based on the X, Y, and Z direction displacement data of the region's center point. Then, the precise location of the damage is determined based on the displacement data of the four boundary points of that region. Finally, the netting damage judgment result is verified by comparing the displacement data of the monitoring points in adjacent sub-monitoring regions and the central monitoring region with their correlation coefficients. Innovation: The concept of dividing monitoring regions for netting damage localization is introduced, improving the model's ability to locate damage. Furthermore, the correlation coefficients of data from adjacent regions are introduced to verify the damage situation, further improving the accuracy of netting damage monitoring.

[0051] In one embodiment of the present invention, the cage structure damage detection model is a neural network model.

[0052] In this embodiment, the cage structure damage detection model adopts a neural network model adapted to complex deep-sea scenarios. This model is not a general-purpose network architecture, but rather specifically optimized based on the mechanical properties and damage patterns of truss-type cages. By introducing an attention mechanism, it prioritizes indicators such as the alternating stress characteristics of the anchor chain and the correlation coefficient of the netting's neighborhood, effectively avoiding noise interference generated by the six degrees of freedom motion of the cage. During the model training phase, based on a massive amount of labeled "environmental data - six degrees of freedom data - netting displacement data - damage state" samples, it learns the characteristic patterns of anchor chain fatigue and netting tearing under different sea conditions (such as 4-6 level winds and waves, typhoons). During the detection phase, it can efficiently receive pre-processed anchor chain and netting feature inputs, and through the collaborative calculation of multiple layers of neurons, accurately output the state results of the anchor chain and netting, providing a highly reliable initial judgment basis for subsequent joint verification, significantly outperforming the detection accuracy and anti-interference capabilities of traditional monitoring methods.

[0053] The foregoing has described specific embodiments of the invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0054] According to another embodiment, the present invention provides a detection device for damage to a cage structure. Figure 2 A schematic block diagram of a detection device for cage structure damage according to one embodiment is shown. It will be understood that this device can be implemented by any device, apparatus, platform, or cluster of devices with computing and processing capabilities. Figure 2 As shown, the device includes: an acquisition unit 200, a first data processing unit 202, a second data processing unit 204, a third data processing unit 206, and a fourth data processing unit 208. The main functions of each component are as follows:

[0055] Acquisition unit 200 is configured to acquire six degrees of freedom data of the net cage, net cage displacement data, and environmental data;

[0056] The first data processing unit 202 is configured to preprocess the six-degree-of-freedom data, the netting displacement data, and the environmental data to obtain preprocessed data.

[0057] The second data processing unit 204 is configured to extract features from the preprocessed data to obtain anchor chain features and netting features.

[0058] The third data processing unit 206 is configured to input the anchor chain features and the netting features into the cage structure damage detection model to obtain a first cage structure damage result; wherein, the first cage structure damage result includes anchor chain status result and netting status result;

[0059] The fourth data processing unit 208 is configured to perform joint verification based on the first cage structure damage result to obtain the second cage structure damage result.

[0060] In one embodiment of the present invention, the anchor chain features include the mean displacement offset rate and the rotation period variation coefficient; the netting features include the peak displacement amplitude ratio and the neighborhood correlation coefficient; the environmental data include wind speed, wind direction, wave height, period, flow velocity, and flow direction; and the preprocessing includes standard time-series processing, data cleaning processing, and noise reduction processing.

[0061] In one embodiment of the present invention, the mean displacement offset rate and the coefficient of variation of the rotation period are determined by the following formula:

[0062]

[0063] In the formula, , , The displacement mean offset rate, The mean offset rate of cage sway. The mean offset rate of cage sway. The mean offset rate of cage heave. This represents the average displacement over the current 10 seconds. This represents the average displacement when the anchor chain is intact. , , The coefficient of variation of the rotation period. The coefficient of variation of the cage's sway period. This is the coefficient of variation of the cage's pitching period. The coefficient of variation of the bow rocking period of the net cage. The standard deviation of the corresponding rotation period, This is the average value of the corresponding rotation period.

[0064] In one embodiment of the present invention, the peak shift ratio and the neighborhood correlation coefficient are determined by the following formula:

[0065]

[0066] In the formula, The peak shift amplitude ratio, This represents the current peak displacement. This represents the peak displacement under normal sea conditions. The neighborhood correlation coefficient is... Let be the covariance of the two displacement parameters. These are the displacement parameters of two adjacent mesh sensors. The first standard deviation, This is the second standard deviation.

[0067] In one embodiment of the present invention, the fourth data processing unit 208 is configured to perform the following operations:

[0068] When both the anchor chain status result and the netting status result are damaged, the weight of the neighborhood correlation coefficient in the netting feature is increased to obtain the updated netting feature;

[0069] The updated mesh features and anchor chain features are input into the cage structure damage detection model to obtain the second cage structure damage result.

[0070] In one embodiment of the present invention, the apparatus further includes a fifth data processing unit, the fifth data processing unit being configured to perform the following operations:

[0071] The mesh of the cage is divided into regions to obtain multiple sub-regions of the mesh.

[0072] Obtain the set of displacements of the center points and boundary points of multiple sub-regions, and record it as the mesh displacement data;

[0073] Displacement sensors are installed at the center and boundary points of multiple sub-regions, and each displacement sensor is assigned a unique address.

[0074] In one embodiment of the present invention, the cage structure damage detection model is a neural network model.

[0075] According to another embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed in a computer, causes the computer to perform a combination Figure 1 The method described.

[0076] According to another embodiment, an electronic device is also provided, including a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, it implements a combination... Figure 1 The method described.

[0077] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the apparatus embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0078] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in this invention can be implemented using hardware, software, firmware, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium.

[0079] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solution of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting damage to a wire mesh cage structure, characterized in that, include: Acquire six degrees of freedom data of the net cage, netting displacement data, and environmental data; The six-degree-of-freedom data, the mesh displacement data, and the environmental data are preprocessed to obtain preprocessed data. The preprocessed data is subjected to feature extraction to obtain anchor chain features and netting features; wherein, the anchor chain features include the mean displacement offset rate and the rotation period variation coefficient; the netting features include the displacement peak amplitude ratio and the neighborhood correlation coefficient. The anchor chain features and the netting features are input into the cage structure damage detection model to obtain the first cage structure damage result; wherein, the first cage structure damage result includes the anchor chain status result and the netting status result; Based on the joint verification of the first cage structure damage result, the second cage structure damage result is obtained; The mean displacement offset rate and the coefficient of variation of the rotation period are determined by the following formula: In the formula, R surge , R sway , R heave The displacement mean offset rate, R surge The mean offset rate of cage sway. R sway The mean offset rate of cage sway. R heave The mean offset rate of cage heave. This represents the average displacement over the current 10 seconds. This represents the average displacement when the anchor chain is intact. C roll , C pitch , C yaw The coefficient of variation of the rotation period. C roll The coefficient of variation of the cage's sway period. C pitch This is the coefficient of variation of the cage's pitching period. C yaw The coefficient of variation of the bow rocking period of the net cage. The standard deviation of the corresponding rotation period, This is the average of the corresponding rotation periods; The peak displacement amplitude ratio and the neighborhood correlation coefficient are determined by the following formula: In the formula, R peak The ratio of the peak displacement amplitude, X current This represents the current peak displacement. X normal This represents the peak displacement under normal sea conditions. ρ i,j The neighborhood correlation coefficient is... cov(x i ,x j ) Let be the covariance of the two displacement parameters. x i ,x j These are the displacement parameters of two adjacent mesh sensors. The first standard deviation, This is the second standard deviation.

2. The method according to claim 1, characterized in that, The environmental data includes wind speed, wind direction, wave height, period, flow velocity, and flow direction; the preprocessing includes standard time-series processing, data cleaning processing, and noise reduction processing.

3. The method according to claim 2, characterized in that, The joint verification based on the damage results of the first cage structure to obtain the damage results of the second cage structure includes: When both the anchor chain status result and the netting status result are damaged, the weight of the neighborhood correlation coefficient in the netting feature is increased to obtain the updated netting feature; The updated mesh features and anchor chain features are input into the cage structure damage detection model to obtain the second cage structure damage result.

4. The method according to claim 1, characterized in that, The mesh displacement data is obtained through the following steps: The mesh of the cage is divided into regions to obtain multiple sub-regions of the mesh. Obtain the set of displacements of the center points and boundary points of multiple sub-regions, and record it as the mesh displacement data; Displacement sensors are installed at the center and boundary points of multiple sub-regions, and each displacement sensor is assigned a unique address.

5. The method according to claim 1, characterized in that, The damage detection model for the cage structure is a neural network model.

6. A device for detecting damage to a wire mesh cage structure, characterized in that, include: The acquisition unit is configured to acquire six degrees of freedom data of the net cage, net displacement data, and environmental data; The first data processing unit is configured to preprocess the six-degree-of-freedom data, the mesh displacement data, and the environmental data to obtain preprocessed data. The second data processing unit is configured to extract features from the preprocessed data to obtain anchor chain features and netting features; wherein, the anchor chain features include the mean displacement offset rate and the rotation period variation coefficient; the netting features include the peak displacement amplitude ratio and the neighborhood correlation coefficient. The third data processing unit is configured to input the anchor chain features and the netting features into the cage structure damage detection model to obtain the first cage structure damage result; wherein, the first cage structure damage result includes the anchor chain status result and the netting status result; The fourth data processing unit is configured to perform joint verification based on the first cage structure damage result to obtain the second cage structure damage result; The mean displacement offset rate and the coefficient of variation of the rotation period are determined by the following formula: In the formula, R surge , R sway , R heave The displacement mean offset rate, R surge The mean offset rate of cage sway. R sway The mean offset rate of cage sway. R heave The mean offset rate of cage heave. This represents the average displacement over the current 10 seconds. This represents the average displacement when the anchor chain is intact. C roll , C pitch , C yaw The coefficient of variation of the rotation period. C roll The coefficient of variation of the cage's sway period. C pitch This is the coefficient of variation of the cage's pitching period. C yaw The coefficient of variation of the bow rocking period of the net cage. The standard deviation of the corresponding rotation period, This is the average of the corresponding rotation periods; The peak displacement amplitude ratio and the neighborhood correlation coefficient are determined by the following formula: In the formula, R peak The ratio of the peak displacement amplitude, X current This represents the current peak displacement. X normal This represents the peak displacement under normal sea conditions. ρ i,j The neighborhood correlation coefficient is... cov(x i ,x j ) Let be the covariance of the two displacement parameters. x i ,x j These are the displacement parameters of two adjacent mesh sensors. The first standard deviation, This is the second standard deviation.

7. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed in a computer, causes the computer to perform the method described in any one of claims 1-5.

Images