Buoy reliability analysis method, terminal equipment and storage medium

Abstract

The invention relates to a buoy reliability analysis method, terminal equipment and a storage medium, and the method comprises the steps: generating a large amount of wave data, and representing the wave height distribution of waves in a time domain and a frequency domain through a wave map; then calculating a surpass probability curve, and further dividing buoy motion response types into multiple types; then, an FNO model is used for establishing the mapping relation between the sub-atlas of the wave atlas and the buoy motion response type, and the mapping relation is embedded into a variational auto-encoder to guide the variational auto-encoder to learn the mapping rule; and finally, calculating an override probability curve of the buoy motion response through a small number of samples simulated by the deep learning model. According to the method, the transcendental probability curve for representing the reliability of the buoy is drawn through a small number of samples obtained through deep learning, and rapid and accurate analysis of the reliability of the buoy is achieved.

Description

Technical Field

[0001] The present invention relates to the field of ocean monitoring technology, and in particular to a buoy reliability analysis method, terminal equipment and storage medium. Background Art

[0002] The ocean covers about 70 percent of the earth's surface. It not only has abundant water, a wide variety of marine life, minerals and other explicit resources, but also plays an implicit role in regulating climate and purifying the environment. Therefore, it is called the cradle of life.

[0003] The buoy system in the ocean plays an important role in the development of marine resources because it can monitor marine environmental parameters such as wave height, temperature and salinity in real time. However, the effectiveness of these buoys is greatly affected by waves. The height, frequency and period of waves directly determine the motion state and stability of the buoy, which in turn affects the accuracy of its data collection. Therefore, in order to fully understand the motion characteristics of the buoy under different sea conditions, it is particularly important to analyze the exceedance probability curve of the buoy motion response under the action of waves. It not only reveals the distribution of the motion behavior of the buoy under wave conditions, but also provides a solid scientific basis for evaluating the reliability of the buoy. Through in-depth analysis of these data, the buoy design can be optimized, its stability in harsh environments can be improved, and the safety and effectiveness of marine resource development can be ensured. However, obtaining an accurate exceedance probability curve of the buoy motion response requires a large amount of wave height time series input, which consumes a lot of computing resources and takes a long time. Summary of the invention

[0004] In order to solve the above problems, the present invention proposes a buoy reliability analysis method, a terminal device and a storage medium.

[0005] The specific plan is as follows: A buoy reliability analysis method comprises the following steps: S1: randomly generate a large amount of wave data, and represent each wave data through a wave spectrum that can characterize the distribution of wave height in the time domain and frequency domain; S2: Establish multiple buoy models corresponding to different rolling periods, and use the generated wave data to calculate the buoy motion response corresponding to each buoy model; S3: based on the calculated motion responses of all buoys, draw a exceedance probability curve of the buoy motion response; S4: Divide the exceedance probability from 0 to 100% into multiple intervals, and set the buoy motion response category corresponding to each interval; S5: extract the sub-atlas corresponding to each buoy model from the wave atlas according to the corresponding roll period of each buoy model; construct a training set based on all wave atlases and their corresponding sub-atlases, and annotate the roll period and buoy motion response category corresponding to each sub-atlas; S6: Train the FNO model using the training set. The input of the model is the sub-map, and the output is the buoy motion response category. S7: embed the trained FNO model into the variational autoencoder; the embedded model first reconstructs a new wave spectrum according to the input wave spectrum through the variational autoencoder, then extracts the sub-spectrum corresponding to the roll period from the generated wave spectrum according to the input roll period, and finally predicts the buoy motion response category of the sub-spectrum through the trained FNO model; S8: Divide the training set according to different rolling periods and different buoy motion response categories, so that each sub-training set obtained by the division corresponds to only one rolling period and one buoy motion response category; train the embedded model through each sub-training set respectively, and use the variational autoencoder in the trained model as the wave spectrum generation model of the rolling period and buoy motion response category corresponding to the sub-training set used for training; S9: according to the wave spectrum corresponding to the current wave data in the area where the buoy to be analyzed is located and the current rolling period of the buoy to be analyzed, a wave spectrum corresponding to different buoy motion response categories of the buoy to be analyzed is generated by a wave spectrum generation model corresponding to different buoy motion response categories with the same rolling period; S10: After reconstructing the wave spectrum into wave data, the exceedance probability curve of the buoy motion response is calculated based on the wave data of all buoy motion response categories corresponding to the buoy to be analyzed, and then the reliability analysis of the buoy to be analyzed is performed according to the exceedance probability curve.

[0006] Furthermore, the wave data are generated by numerical simulation techniques of multi-directional irregular waves.

[0007] Furthermore, the buoy model was established using ANSYS software.

[0008] Furthermore, the number of intervals divided by the exceedance probability is four, namely [0, 25%), [25%, 50%), [50%, 75%), and [75%, 100%], and the corresponding buoy motion response categories are four.

[0009] Furthermore, according to the corresponding rolling period of each buoy model, the method of extracting the sub-atlas corresponding to each buoy model from the wave atlas is as follows: the rolling period corresponding to the buoy model is set to T, and the sub-atlas intercepts the part with a period of [a*T, b*T] from the wave atlas, where the value range of the two coefficients a and b satisfies 0 <a<1,b> 1.

[0010] A buoy reliability analysis terminal device comprises a processor, a memory and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method in an embodiment of the present invention when executing the computer program.

[0011] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described above in an embodiment of the present invention are implemented.

[0012] The present invention adopts the above technical solution, and uses a small number of samples obtained through deep learning to draw a exceedance probability curve for characterizing the reliability of the buoy, which can realize a rapid and accurate analysis of the reliability of the buoy. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 Shown is a flow chart of a method according to a first embodiment of the present invention.

[0014] Figure 2 Shown is a schematic diagram of a curve showing the variation of the wave height of random waves over time in this embodiment.

[0015] Figure 3 Shown is a schematic diagram of the wave pattern in this embodiment.

[0016] Figure 4 FIG. 4 is a schematic diagram of the exceedance probability curve in this embodiment. DETAILED DESCRIPTION

[0017] To further illustrate various embodiments, the present invention provides drawings. These drawings are part of the disclosure of the present invention, which are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementations and advantages of the present invention.

[0018] The present invention will now be further described with reference to the accompanying drawings and specific implementation methods.

[0019] Embodiment 1: The embodiment of the present invention provides a buoy reliability analysis method, such as Figure 1 As shown, the method comprises the following steps: S1: A large amount of wave data is randomly generated, and each wave data is represented by a wave spectrum that can characterize the wave height distribution in the time domain and the frequency domain.

[0020] In this embodiment, wave data is generated by numerical simulation technology of multi-directional irregular waves. The numerical simulation technology of multi-directional irregular waves is an important tool for studying wave propagation and its interaction with engineering structures, and is widely used in the fields of marine engineering, ports and coastal engineering. In this embodiment, a large amount of wave data is generated by formulas (1)-(6).

[0021]

[0022] Where S(ω,θ) is the directional spectral density function of multi-directional irregular waves, ω is the component wave frequency, and θ is the directional angle; S(ω) is the spectrum of the waves, and the specific calculation formulas are shown in Formulas (2) and (3); G(ω,θ) is the directional distribution function of the waves, and the specific calculation formulas are shown in Formulas (4), (5) and (6).

[0023]

[0024] In the formula, is the peak enhancement factor, is the peak shape parameter, is the effective wave height, is the effective period, and are the peak frequency and peak period, respectively.

[0025]

[0026]

[0027] In the formula, [ , ] is the direction distribution range of the waves; parameter The value of is:

[0028] Since the wave energy of multi-directional irregular waves is distributed in a certain frequency domain and direction range at the same time, the frequency domain is divided into M parts and the direction is divided into I parts. There are a total of M×I component units. The component wave of each unit is regarded as a simple harmonic wave. The multi-directional irregular waves formed by the superposition of the component waves are:

[0029] The parameters in the formula are calculated as follows:

[0030] Where x, y are position coordinates; is the wave number; d is the water depth; For [0, 2 ] is a random variable uniformly distributed within ; , is the serial number of the frequency component; , is the serial number of the direction; are uniformly distributed, uncorrelated random variables on the interval [0, 1]; is the maximum value of the component wave frequency; is the minimum value of the wave frequency; g is the gravitational acceleration; is the frequency value of the mth frequency component in the i-th direction.

[0031] The curve of wave height changing with time in the generated wave data is as follows: Figure 2 As shown, the converted wave spectrum is as follows Figure 3 As shown, the three coordinate systems in the wave spectrum are time, frequency and wave height.

[0032] S2: Establish multiple buoy models corresponding to different rolling periods, and use the generated wave data to calculate the buoy motion response corresponding to each buoy model.

[0033] In this embodiment, the buoy model is established by ANSYS software.

[0034] The motion response of the buoy in the waves is mostly represented by the heave height and the roll angle, and in this embodiment it is represented by the roll angle.

[0035] The motion response of the buoy in the waves can be described by the buoy dynamics equations. These equations are based on Newton's second law and the principles of ocean dynamics, and take into account the force balance and motion of the buoy. In order to calculate the motion response of the buoy, the wave data of the buoy can be input into the buoy dynamics equations, and the motion response of the buoy can be obtained by solving these equations. The motion response of the buoy is mostly represented by the heave height and the roll angle, and in this embodiment, it is represented by the roll angle.

[0036] Each wave data has a different buoy motion response for different buoy models. If there are n wave data and m buoy models, there are n*m ​​calculated buoy motion responses.

[0037] S3: Based on the calculated motion responses of all buoys, a exceedance probability curve of the buoy motion response is plotted.

[0038] Step S3 includes the following steps: S301: Calculate the mean and standard deviation of all buoy motion responses and approximate them to a normal distribution.

[0039]

[0040] In the formula, f(x) is the approximate normal distribution of the buoy motion response; x represents the buoy motion response, which is the roll angle in this embodiment; is the standard deviation of all buoy motion responses; is the mean of all buoy motion responses. S302: Calculate the cumulative distribution of the approximate normal distribution of the buoy motion response, that is, the exceedance probability curve of the buoy motion response.

[0041]

[0042] In the formula, is the exceedance probability of the buoy motion response.

[0043] S4: Divide the exceedance probability from 0 to 100% into multiple intervals, and set the buoy motion response category corresponding to each interval.

[0044] In this embodiment, the exceedance probability is preferably divided into four intervals from 0 to 100%, and the four intervals are [0, 25%), [25%, 50%), [50%, 75%), and [75%, 100%). The corresponding buoy motion response categories are four categories, namely, category I, category II, category III, and category IV, as follows:

[0045] That is, when the exceedance probability of the buoy motion response is When corresponding to a certain range in the above formula (11), the corresponding buoy motion response belongs to one of the categories of Class I, Class II, Class III and Class IV.

[0046] S5: According to the roll period corresponding to each buoy model, the sub-atlas corresponding to each buoy model is extracted from the wave atlas; the first training set is constructed based on the sub-atlases corresponding to all wave atlases, and the roll period and buoy motion response category corresponding to each sub-atlas are marked.

[0047] In this embodiment, according to the corresponding rolling period of each buoy model, the method of extracting the sub-spectrum corresponding to each buoy model from the wave spectrum is as follows: the rolling period corresponding to the buoy model is set to T, and the sub-spectrum intercepts the part with a period of [a*T, b*T] from the wave spectrum, where the value range of the two coefficients a and b satisfies 0 <a<1,b> 1. In this embodiment, it is preferred to set a=0.1 and b=2. In other embodiments, those skilled in the art can make settings according to actual needs, which are not limited here.

[0048] The roll period corresponding to the sub-map is the roll period of the buoy model corresponding to the sub-map.

[0049] The buoy motion response category corresponding to the sub-atlas is the buoy motion response category corresponding to the buoy motion response calculated by the wave data corresponding to the wave atlas to which the sub-atlas belongs according to the buoy model corresponding to the sub-atlas.

[0050] S6: The FNO (Fourier Neural Operator) model is trained using the training set to obtain the mapping relationship between each sub-spectrum and the buoy motion response category.

[0051] FNO is a new type of deep learning architecture that is specifically designed to learn mappings from infinite-dimensional function spaces to function spaces, which makes it excellent at solving partial differential equations. Unlike traditional finite or finite difference-based methods, FNO can operate directly in function space without relying on grid discretization or explicit numerical solutions. The core idea of ​​FNO is to map function signals to the Fourier frequency domain, operate on these signals in the frequency domain, and then return them to the original space through the inverse Fourier transform. Compared with traditional neural network methods, FNO has stronger generalization capabilities.

[0052] The input of the model is a subgraph , the output is the corresponding buoy motion response category. After sufficient training, the mapping relationship between the sub-atlas and the buoy response category can be established.

[0053] S7: Embed the trained FNO model into a variational autoencoder (VAE).

[0054] Variational autoencoders are generative deep learning models that are designed to learn the underlying probability distribution of high-dimensional data. Unlike traditional autoencoders, VAEs can model the distribution of data in latent variable space and generate new data samples from it by introducing the ideas of probabilistic inference and generative modeling. Its core goal is to maximize the marginal log-likelihood function of the data in order to learn the mapping from latent variables to data space.

[0055] The input of the embedded model includes wave spectrum and roll period, and the output is the buoy motion response category (that is, the buoy motion response category corresponding to the buoy motion response calculated by the buoy model corresponding to the wave data corresponding to the input roll period according to the input wave spectrum). The specific implementation method is: the embedded model first reconstructs a new wave spectrum according to the input wave spectrum through the variational autoencoder, and then extracts the sub-spectrum corresponding to the roll period from the generated wave spectrum according to the input roll period, and finally predicts the buoy motion response category of the sub-spectrum through the trained FNO model.

[0056] The role of the embedded FNO is to predict the corresponding buoy motion response category based on the input sub-atlas, and to determine whether the predicted buoy motion response category is consistent with the actual buoy motion response category corresponding to the training data, and to feed back the judgment result to the variational autoencoder to guide its training optimization. The trained variational autoencoder can simulate and generate a typical wave atlas that satisfies the corresponding rolling period and buoy motion response category.

[0057] S8: Divide the training set according to different roll periods and different buoy motion response categories, so that each sub-training set obtained by the division corresponds to only one roll period and one buoy motion response category; train the embedded model through each sub-training set respectively, and use the variational autoencoder in the trained model as the wave spectrum generation model of the roll period and buoy motion response category corresponding to the sub-training set used for training.

[0058] The wave atlas generation model trained by the sub-training set can generate the wave atlas of the rolling period and buoy motion response category corresponding to the sub-training set according to the input wave atlas.

[0059] S9: Based on the wave spectrum corresponding to the current wave data in the area where the buoy to be analyzed is located and the current roll period of the buoy to be analyzed, a wave spectrum generation model corresponding to different buoy motion response categories with the same roll period is used to generate a wave spectrum corresponding to different buoy motion response categories for the buoy to be analyzed.

[0060] Since the buoy motion response categories in this embodiment include four categories, in step S9, four wave spectrum generation models corresponding to four different buoy motion response categories with the same rolling period are used to generate wave spectrums corresponding to the four buoy motion response categories of the buoy to be analyzed. Each wave spectrum generation model may generate one or more wave spectrums.

[0061] S10: After reconstructing the wave spectrum into wave data, the exceedance probability curve of the buoy motion response is calculated based on the wave data of all buoy motion response categories corresponding to the buoy to be analyzed, and then the reliability analysis of the buoy to be analyzed is performed according to the exceedance probability curve.

[0062] The calculation method of the exceedance probability curve of the buoy motion response is the same as step S3.

[0063] In this embodiment, the reliability of the buoy is determined based on the shape of the exceedance probability curve, such as Figure 4 As shown in the figure, the roll angle exceedance probability distribution of buoys with different roll periods (the black line and the yellow line correspond to buoys with two different roll periods) under the same wave energy spectrum distribution conditions is compared and analyzed. The larger the roll angle of the buoy, the easier it is for the buoy to capsize. Figure 4It can be found that under the same exceedance probability, the maximum roll angle of the buoy corresponding to the black line is significantly smaller than the maximum roll angle of the buoy corresponding to the yellow line. The results show that under the same ocean conditions, the probability of the buoy corresponding to the yellow line encountering a large roll motion is significantly higher than that of the buoy corresponding to the black line, and it has a higher risk of capsizing under extreme sea conditions. In practical applications, the reliability of the buoy can be judged based on whether the actual roll angle under a fixed exceedance probability is less than the set roll angle threshold (less than reliable, otherwise unreliable). The roll angle threshold can be set based on experimental or empirical data.

[0064] The embodiment of the present invention first uses the numerical simulation technology of multi-directional irregular waves to generate a large amount of wave data, and uses the wave spectrum to characterize the distribution of wave heights in the time domain and frequency domain. Then, the Fourier neural operator is used to establish the mapping relationship between the sub-spectrum of the wave spectrum and the buoy motion response type, and is embedded in the variational autoencoder to guide it to learn this mapping law. Finally, a small number of samples simulated by the deep learning model are used to calculate the exceedance probability curve of the buoy motion response, thereby realizing a rapid and refined analysis of the buoy reliability.

[0065] Embodiment 2: The present invention also provides a buoy reliability analysis terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in the above-mentioned method embodiment of embodiment 1 of the present invention when executing the computer program.

[0066] Further, as an executable solution, the buoy reliability analysis terminal device can be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The buoy reliability analysis terminal device can include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the composition structure of the above-mentioned buoy reliability analysis terminal device is only an example of the buoy reliability analysis terminal device, and does not constitute a limitation on the buoy reliability analysis terminal device. It can include more or less components than the above, or combine certain components, or different components. For example, the buoy reliability analysis terminal device can also include input and output devices, network access devices, buses, etc., and the embodiments of the present invention do not limit this.

[0067] Further, as an executable solution, the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the buoy reliability analysis terminal device, and uses various interfaces and lines to connect various parts of the entire buoy reliability analysis terminal device.

[0068] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the buoy reliability analysis terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required for a function; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0069] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method in the embodiment of the present invention are implemented.

[0070] If the module / unit integrated in the buoy reliability analysis terminal device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory) and software distribution medium, etc.

[0071] Although the present invention has been specifically shown and described in conjunction with the preferred embodiments, it should be understood by those skilled in the art that various changes may be made to the present invention in form and details without departing from the spirit and scope of the present invention as defined by the appended claims, all of which are within the scope of protection of the present invention.

Claims

1. A buoy reliability analysis method, characterized in that: The following steps are involved: S1: randomly generate a large amount of wave data, and represent each wave data through a wave spectrum that can characterize the distribution of wave height in the time domain and frequency domain; S2: Establish multiple buoy models corresponding to different rolling periods, and use the generated wave data to calculate the buoy motion response corresponding to each buoy model; S3: based on the calculated motion responses of all buoys, draw a exceedance probability curve of the buoy motion response; S4: Divide the exceedance probability from 0 to 100% into multiple intervals, and set the buoy motion response category corresponding to each interval; S5: extract the sub-atlas corresponding to each buoy model from the wave atlas according to the corresponding roll period of each buoy model; construct a training set based on all wave atlases and their corresponding sub-atlases, and annotate the roll period and buoy motion response category corresponding to each sub-atlas; S6: Train the FNO model using the training set. The input of the model is the sub-map, and the output is the buoy motion response category. S7: embed the trained FNO model into the variational autoencoder; the embedded model first reconstructs a new wave spectrum according to the input wave spectrum through the variational autoencoder, then extracts the sub-spectrum corresponding to the roll period from the generated wave spectrum according to the input roll period, and finally predicts the buoy motion response category of the sub-spectrum through the trained FNO model; S8: Divide the training set according to different rolling periods and different buoy motion response categories, so that each sub-training set obtained by the division corresponds to only one rolling period and one buoy motion response category; train the embedded model through each sub-training set respectively, and use the variational autoencoder in the trained model as the wave spectrum generation model of the rolling period and buoy motion response category corresponding to the sub-training set used for training; S9: according to the wave spectrum corresponding to the current wave data in the area where the buoy to be analyzed is located and the current rolling period of the buoy to be analyzed, a wave spectrum corresponding to different buoy motion response categories of the buoy to be analyzed is generated by a wave spectrum generation model corresponding to different buoy motion response categories with the same rolling period; S10: After reconstructing the wave spectrum into wave data, the exceedance probability curve of the buoy motion response is calculated based on the wave data of all buoy motion response categories corresponding to the buoy to be analyzed, and then the reliability analysis of the buoy to be analyzed is performed according to the exceedance probability curve.

2. The buoy reliability analysis method according to claim 1, characterized in that: The wave data are generated by numerical simulation techniques of multi-directional irregular waves.

3. The buoy reliability analysis method according to claim 1, characterized in that: The buoy model was established using ANSYS software.

4. The buoy reliability analysis method according to claim 1, characterized in that: There are four intervals for the exceedance probability division, namely [0, 25%), [25%, 50%), [50%, 75%), and [75%, 100%], and the corresponding buoy motion response categories are four categories.

5. The buoy reliability analysis method according to claim 1, characterized in that: According to the corresponding rolling period of each buoy model, the method of extracting the sub-atlas corresponding to each buoy model from the wave atlas is as follows: set the rolling period corresponding to the buoy model as T, then the sub-atlas intercepts the part with a period of [a*T, b*T] from the wave atlas, where the value range of the two coefficients a and b satisfies 0 <a<1,b> 1.

6. A buoy reliability analysis terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 5 when executing the computer program.

Images