Ship data sampling method
Through independent update and multi-sample generation of ship data parent samples, the accuracy problem of the evolutionary sampling algorithm under the differences in complex probability density functions and variable ranges is solved, and the accuracy and data volume of ship stability prediction are improved.
Patent Information
- Application Number
- CN202411098350.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-12
AI Technical Summary
The existing evolutionary sampling algorithms are difficult to cope with the complex multi-dimensional probability density function and variable range differences, resulting in a decrease in the sampling accuracy of ship data and cannot meet the high-precision requirements of ship stability prediction.
By independently updating each parent sample feature component of the ship data parent sample, using multiple different proposal distributions and weight functions to generate candidate sample feature components, perform rejection sampling operations, select samples with higher acceptance rates, and gradually update the sample set. , until the preset iteration number is satisfied.
It effectively increases the diversity and reliability of samples, improves the accuracy of ship stability prediction, adapts to the sampling range differences of different components, and amplifies the reliable data volume.
Smart Images

Figure CN119025857A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a ship data sampling method. Background Art
[0002] In the process of ship production and design, a large amount of ship-related characteristic data is often needed to train the model to predict the stability of the ship, including ship length, ship width, draft, initial trim, initial stability height, waterline length, roll period, center of gravity height, square coefficient, bilge keel area, water inlet angle, wind pressure moment, measured projection area, measured projection surface center height and displacement, etc.
[0003] However, in the actual production and design of ships, there are few channels for obtaining ship characteristic data and it is difficult to obtain it. Therefore, it is necessary to conduct supplementary collection based on the existing ship characteristic data.
[0004] Existing supplementary acquisition methods often use the evolutionary sampling algorithm (ES algorithm), but the current evolutionary sampling algorithm is difficult to cope with the sampling accuracy requirements brought about by complex multi-dimensional probability density functions and variable range differences, as follows:
[0005] (1) As the complexity of the probability density function increases, it not only involves multidimensional variables, but also often contains nonlinear and multi-peak characteristics, which makes the sampling process extremely difficult. At the same time, with the continuous improvement of sample quality requirements in the field of ship sampling, especially in applications such as machine learning and statistical inference, the requirements for sample representativeness and diversity are becoming more and more stringent, and the sampling accuracy of the existing evolutionary sampling algorithm has been difficult to meet the requirements;
[0006] (2) There are large differences in the sampling ranges of different variables in ship data. For example, for a specific ship type, it may be necessary to sample the depth within a small range, while for the draft, it is necessary to collect data from a larger range. However, the existing evolutionary sampling algorithm uses the same transfer density function for all components to supplement the sample collection, which is difficult to deal with the situation where the variable distribution range is unequal, and thus will also lead to a decrease in sampling accuracy.
[0007] The above defects of the algorithm make the effective ship characteristic data that can be obtained in practice still very limited, which further leads to inaccurate ship stability prediction.
[0008] In summary, how to provide a ship data sampling method that can efficiently expand the number of effective samples for predicting ship stability has become a technical problem that needs to be urgently solved by technical personnel in this field. Summary of the invention
[0009] The present invention provides a ship data sampling method to solve the problems existing in the related art that it is difficult to obtain a large amount of effective ship type characteristic data, resulting in low accuracy in ship stability prediction, etc.
[0010] As a first aspect of the present invention, a ship data sampling method is provided, comprising:
[0011] Obtaining a parent sample of ship data for predicting ship stability, the parent sample of ship data comprising a plurality of parent sample feature components, each parent sample feature component corresponding to a ship stability feature;
[0012] Performing data augmentation processing on each parent sample feature component to obtain at least one candidate sample feature component corresponding to each parent sample feature component;
[0013] Obtain at least one candidate sample of the parent sample of the ship data according to at least one candidate sample feature component corresponding to each parent sample feature component;
[0014] Perform a rejection sampling operation on each candidate sample to obtain at least one candidate result sample of the ship data parent sample, wherein the number of candidate result samples of the ship data parent sample, the number of candidate samples of the ship data parent sample, and the number of candidate sample feature components corresponding to each parent sample feature component are all equal;
[0015] If the number of candidate result samples is 1, the candidate result sample is determined to be a subsample of the parent sample of ship data;
[0016] If the number of candidate result samples is at least 2, one sub-sample of the parent sample of ship data is determined from at least two candidate result samples according to the degree of dispersion of each candidate result sample;
[0017] Update the parent sample of ship data according to the sub-sample, repeat the above steps until the preset number of iterations is met, and obtain the final sample.
[0018] Further, performing data augmentation processing on each parent sample feature component to obtain at least one candidate sample feature component corresponding to each parent sample feature component includes:
[0019] Generate at least two candidate feature components corresponding to each parent sample feature component according to at least two different proposal distributions;
[0020] Obtain at least one candidate sample feature component corresponding to each parent sample feature component according to at least one weight function and at least two candidate feature components corresponding to each parent sample feature component;
[0021] The number of candidate sample feature components corresponding to each parent sample feature component is equal to the number of weight functions.
[0022] Further, obtaining at least one candidate sample feature component corresponding to each parent sample feature component according to at least one weight function and at least two candidate feature components corresponding to each parent sample feature component includes:
[0023] Selecting at least two candidate feature components corresponding to each parent sample feature component in turn according to at least two different weight functions to obtain at least two temporary feature components corresponding to each parent sample feature component, wherein the number of temporary feature components corresponding to each parent sample feature component is equal to the number of weight functions;
[0024] A rejection sampling operation is performed on each temporary feature component to obtain at least two candidate sample feature components corresponding to each parent sample feature component.
[0025] Further, selecting in turn from at least two candidate feature components corresponding to each parent sample feature component according to at least two different weight functions to obtain at least two temporary feature components corresponding to each parent sample feature component includes:
[0026] At least two different weight functions are obtained according to the adjustment result of the weight parameter of the weight function, wherein the weight parameter represents the bias of the control weight function;
[0027] Calculate the weight value of each candidate feature component corresponding to each parent sample feature component under each weight function;
[0028] For each parent sample feature component, select the corresponding candidate feature components calculated by the same weight function as a group, and select a temporary feature component of the parent sample feature component from the corresponding candidate feature components by random selection according to the proportion of each group of weight values;
[0029] After traversing each candidate feature component corresponding to each parent sample feature component, at least two temporary feature components corresponding to each parent sample feature component are obtained.
[0030] Furthermore, the at least two different weight functions include: a weight function with a weight parameter of 0 and a weight function with a weight parameter of 2.5.
[0031] Further, a rejection sampling operation is performed on each temporary feature component to obtain at least two candidate sample feature components corresponding to each parent sample feature component, including:
[0032] Calculate the acceptance rate of each temporary feature component separately;
[0033] Generate random numbers in sequence. If the generated random numbers are less than the acceptance rate of the corresponding temporary feature component, then accept the temporary feature component as a candidate sample feature component of the corresponding parent sample feature component.
[0034] If the generated random number is greater than or equal to the acceptance rate of the corresponding temporary feature component, then the candidate sample feature component of the temporary feature component corresponding to the parent sample feature component is rejected, and the parent sample feature component is determined as the candidate sample feature component;
[0035] After traversing each temporary feature component corresponding to each parent sample feature component, at least two candidate sample feature components corresponding to each parent sample feature component are obtained.
[0036] Further, obtaining at least one candidate sample of the parent sample of the ship data according to at least one candidate sample feature component corresponding to each parent sample feature component includes:
[0037] Form candidate samples of the parent sample of the ship data according to the candidate sample feature components selected based on the same weight function of each parent sample feature component;
[0038] After traversing each candidate sample feature component corresponding to each parent sample feature component, at least two candidate samples of the ship data parent sample are obtained.
[0039] Further, a rejection sampling operation is performed on each candidate sample to obtain at least one candidate result sample of the parent sample of the ship data, including:
[0040] Calculate the acceptance rate of each candidate sample separately;
[0041] Generate random numbers in sequence. If the generated random numbers are less than the acceptance rate of the corresponding candidate samples, the candidate samples are accepted as the candidate result samples of the parent samples of the ship data.
[0042] If the generated random number is greater than or equal to the acceptance rate of the corresponding candidate sample, the candidate sample is rejected as the candidate result sample of the parent sample of the ship data, and the parent sample of the ship data is determined as the candidate result sample;
[0043] After traversing each candidate sample of the parent sample of the ship data, at least one candidate result sample of the parent sample of the ship data is obtained.
[0044] Furthermore, the acceptance rate of each candidate sample is calculated separately, including:
[0045] Calculate the target bandwidth matrix of the multivariate kernel density estimation function based on the AMISE minimization bandwidth calculation method and the Kendall correlation coefficient;
[0046] The acceptance rate of each candidate sample is calculated according to the target bandwidth matrix.
[0047] Further, if the number of the candidate result samples is at least 2, determining a sub-sample of the parent sample of the ship data from at least two candidate result samples according to the degree of dispersion of each candidate result sample includes:
[0048] If the number of candidate result samples is at least 2, calculate the standard deviation value of each group of candidate result samples respectively;
[0049] The candidate result sample with the smallest standard deviation value is determined as the subsample of the parent sample of ship data.
[0050] Compared with updating the entire sample as a whole, the present invention independently updates the characteristic component of each parent sample of the parent sample of ship data. This not only increases the diversity of samples, but also effectively addresses the problem of decreased sampling accuracy caused by large differences in the sampling ranges of different components of ship data. The final sample obtained effectively expands the amount of reliable ship data for ship stability prediction, thereby improving the accuracy of ship stability prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention, but do not constitute a limitation of the present invention.
[0052] Figure 1 A brief flow chart of the ship data sampling method provided by the present invention.
[0053] Figure 2 A brief flow chart of the data augmentation process provided by the present invention.
[0054] Figure 3 A brief flow chart of generating candidate sample feature components provided by the present invention.
[0055] Figure 4 A brief flow chart of generating temporary feature components provided by the present invention.
[0056] Figure 5 A specific flow chart of generating candidate sample feature components provided by the present invention.
[0057] Figure 6 A brief flowchart of generating candidate samples provided by the present invention.
[0058] Figure 7 A brief flow chart of generating a sample of candidate results provided by the present invention.
[0059] Figure 8 A simplified flow chart for calculating the acceptance rate provided by the present invention.
[0060] Fig. 9 A brief flowchart of obtaining subsamples provided by the present invention.
[0061] Fig.10 The test function π in Table 2 provided by the present invention 1Convergence plot of the standard deviation of .
[0062] Fig.11 The test function π in Table 2 provided by the present invention 1 Convergence plot of the normalization factor.
[0063] Fig.12 The test function π in Table 2 provided by the present invention 2 Convergence plot of the standard deviation of .
[0064] Fig.13 The test function π in Table 2 provided by the present invention 2 Convergence plot of the normalization factor.
[0065] Fig.14 The test function π in Table 2 provided by the present invention 4 Convergence plot of the standard deviation of .
[0066] Fig.15 The test function π in Table 2 provided by the present invention 4 Convergence plot of the normalization factor.
[0067] Fig.16 The test function π in Table 2 provided by the present invention 6 Convergence plot of the standard deviation of .
[0068] Fig.17 The test function π in Table 2 provided by the present invention 6 Convergence plot of the normalization factor.
[0069] Fig.18 The test function π in Table 2 provided by the present invention 8 Convergence plot of the standard deviation of .
[0070] Fig.19 The test function π in Table 2 provided by the present invention 8 Convergence plot of the normalization factor.
[0071] Fig. 20 The test function π in Table 2 provided by the present invention 10 Convergence plot of the standard deviation of .
[0072] Fig.21 The test function π in Table 2 provided by the present invention 10 Convergence plot of the normalization factor.
[0073] Fig. 22 The test function π in Table 2 provided by the present invention 11 Convergence plot of the standard deviation of .
[0074] Fig.23 The test function π in Table 2 provided by the present invention 11 Convergence plot of the normalization factor.
[0075] Fig.24 The test function π in Table 2 provided by the present invention 12 Convergence plot of the standard deviation of .
[0076] Fig.25 The test function π in Table 2 provided by the present invention 12 Convergence plot of the normalization factor.
[0077] Fig.26 The test function π in Table 2 provided by the present invention 13 Convergence plot of the standard deviation of .
[0078] Fig. 27 The test function π in Table 2 provided by the present invention 13 Convergence plot of the normalization factor. DETAILED DESCRIPTION
[0079] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0080] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.
[0081] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0082] The embodiment of the present invention provides a ship data sampling method, such as Figure 1 As shown, including:
[0083] S100, obtaining a parent sample of ship data for predicting ship stability, the parent sample of ship data comprising a plurality of parent sample feature components, each parent sample feature component corresponding to a ship stability feature;
[0084] When the embodiment of the present invention is actually implemented, a ship data sample set is obtained, which includes several ship data parent samples for predicting ship stability. The ship data parent samples include several parent sample feature components, and each parent sample feature component corresponds to a ship stability feature, such as ship length, ship width, draft, initial trim, initial stability height, waterline length, roll period, center of gravity height, square coefficient, bilge keel area, water inlet angle, wind pressure moment, measured projection area, measured projection surface center height and displacement, etc.
[0085] Specifically, the number of parent samples of ship data is preset to N, that is, the number of parent samples of ship data is N, and the number of iterations of parent samples of ship data Iter is preset to initialize the parent samples of ship data. The initialized parent sample set of ship data is specifically expressed as It means that the i-th and t-th iteration parent sample of the ship data in the parent sample set of the ship data has D parent sample characteristic components, where i = 1, 2, ..., N, It represents the k-th parent sample feature component in the parent sample of the ship data and undergoes the t-th iteration, k=1, 2, ..., D. The following steps 200 to 700 are performed for each parent sample of the ship data in the parent sample set of the ship data.
[0086] S200, performing data augmentation processing on each parent sample feature component to obtain at least one candidate sample feature component corresponding to each parent sample feature component;
[0087] Specifically, the embodiment of the present invention can use evolutionary algorithms, such as random sampling, Latin hypercube sampling, Gibbs sampling, etc. to select the parent sample of ship data. Each parent sample characteristic component Perform data augmentation processing to obtain the characteristic components of each parent sample At least one candidate sample feature component S≥1.
[0088] S300, obtaining at least one candidate sample of the parent sample of the ship data according to at least one candidate sample feature component corresponding to each parent sample feature component;
[0089] Specifically, according to each parent sample feature component Corresponding to at least one candidate sample feature component Get the parent sample of ship data At least one candidate sample
[0090] S400, performing a rejection sampling operation on each candidate sample to obtain at least one candidate result sample of the ship data parent sample, wherein the number of candidate result samples of the ship data parent sample, the number of candidate samples of the ship data parent sample, and the number of candidate sample feature components corresponding to each parent sample feature component are all equal;
[0091] Specifically, this embodiment calculates the candidate samples The acceptance rate of the parent sample of the ship data Each candidate sample One by one with the parent sample of the ship data Compare and select, and the selected one is the candidate result sample of the parent sample of the ship data, so as to obtain the candidate result sample of the parent sample of the ship data s=1,2,……,S, the parent sample of the ship data Sample of candidate results The number of parent samples of the ship data Candidate samples of The number and parent sample of the ship data Each parent sample characteristic component Corresponding candidate sample feature components The quantities are equal.
[0092] S500, if the number of the result samples to be selected is 1, determine that the result sample to be selected is a subsample of the parent sample of the ship data;
[0093] Specifically, when the parent sample of the ship data There is only one sample of results to be selected When the result sample to be selected is determined to be the parent sample of the ship data Subsample of
[0094] S600, if the number of candidate result samples is at least 2, determine a sub-sample of the parent sample of ship data from at least two candidate result samples according to the discrete degree of each candidate result sample;
[0095] When the parent sample of the ship data There are more than two candidate result samples When , it is necessary to select several samples of the results to be selected Specifically, in the embodiment of the present invention, among the multiple result samples to be selected, Select the one with the smallest degree of dispersion as the parent sample of the ship data Subsample of
[0096] S700, updating the parent sample of ship data according to the sub-sample, repeating the above steps until the preset number of iterations is met, and obtaining the final sample;
[0097] Specifically, the embodiment of the present invention updates the parent sample of the ship data according to the subsample: The above steps 200 to 700 are iteratively performed on the parent sample of the ship data until the preset number of iterations Iter is met to obtain the final sample, and after traversing each parent sample of the ship data in the parent sample set of the ship data, the final sample set is obtained.
[0098] In summary, the embodiment of the present invention independently updates the characteristic component of each parent sample of the parent sample of ship data. Compared with updating the entire sample as a whole, it can not only increase the diversity of samples, but also effectively deal with the problem of decreased sampling accuracy caused by large differences in the sampling ranges of different components of ship data. The final sample obtained effectively expands the amount of reliable ship data for ship stability prediction, thereby improving the accuracy of ship stability prediction.
[0099] Furthermore, if Figure 2 As shown, data augmentation processing is performed on each parent sample feature component to obtain at least one candidate sample feature component corresponding to each parent sample feature component, including:
[0100] S210, generating at least two candidate feature components corresponding to each parent sample feature component according to at least two different proposed distributions;
[0101] Proposal Distribution, also known as suggestion distribution, is an importance sampling method for obtaining weighted sub-sampling of posterior probability distribution. Different proposal distributions can affect the search range of evolutionary sampling.
[0102] In the embodiment of the present invention, the number of proposed distributions is set to M, M ≥ 2, for the parent sample of ship data The parent sample characteristic component Through M different proposal distributions Generate M candidate feature components z 1 ,z 2 ,…,z M .
[0103] Specifically, since the Lévy distribution has the characteristics of infinite second-order moment and adjustable control parameters, the embodiment of the present invention uses the Lévy distribution to obtain M different proposed distributions by adjusting the control parameters of the Lévy distribution. The control parameter α of the Lévy distribution can control the search range of the proposed distribution, so that the probability distribution presents different shapes, especially the tail area. The smaller the parameter α, the longer the tail, 0<α<2.
[0104] Then, each parent sample feature component is taken as the center of the proposal distribution, and the candidate feature components of the parent sample feature component are generated by using M proposal distributions as the transition probability function.
[0105] Specifically, according to the M proposal distribution As a transition probability function, a set of candidate feature components is generated for each parent sample feature component of the corresponding parent sample of ship data Including M candidate feature components z 1 ,z 2 ,…,z M , where j represents the jth proposal distribution, j = 1, 2, …, M.
[0106] The present invention generates multiple candidate feature components for each parent sample feature component of the parent sample of the ship data through multiple different proposed distributions. After obtaining M candidate feature components of each parent sample feature component, each parent sample feature component is independently updated, and through a subsequent evolution process, relatively excellent candidate sample feature components are selected to form candidate samples. This can not only increase sample diversity and explore the parameter space more widely, but also effectively deal with the problem of decreased sampling accuracy caused by large differences in the sampling ranges of different components.
[0107] Specifically, since the analytical expression of the candidate feature component that satisfies this integral is unknown for the general control parameter α, the analytical expression of the candidate feature component can be obtained in some special cases. In particular, for α=1, this integral can be reduced to a Cauchy distribution. When α→2, that is, α infinitely approaches 2, the distribution is no longer a Lévy distribution, but becomes a Gaussian distribution. The candidate feature components generated by the proposed distributions obtained in these two cases can often obtain a stable evolution process. Therefore, when the embodiment of the present invention is implemented, the control parameter α=1 is adjusted to obtain a proposed distribution, namely, the Cauchy distribution, and the control parameter α→2 is adjusted to obtain another proposed distribution, namely, the Gaussian distribution.
[0108] S220, obtaining at least one candidate sample feature component corresponding to each parent sample feature component according to at least one weight function and at least two candidate feature components corresponding to each parent sample feature component;
[0109] The number of candidate sample feature components corresponding to each parent sample feature component is equal to the number of weight functions.
[0110] Specifically, in the embodiment of the present invention, the number of weight functions is set to S, S ≥ 1, that is, S different weight functions ω are set 1 ,ω 2 , ..., ω s , for the parent sample of ship data The parent sample characteristic component Through the weight function ω 1 ,ω 2 , ..., ω s (s=1,2,…,S) in the parent sample feature component The corresponding candidate feature component z 1 ,z 2 ,…,z M Selective evolution is performed in the process to obtain the characteristic components of each parent sample Corresponding to at least one candidate sample feature component
[0111] Furthermore, if Figure 3 As shown, according to at least one weight function and at least two candidate feature components corresponding to each parent sample feature component, obtaining at least one candidate sample feature component corresponding to each parent sample feature component includes:
[0112] S221, selecting in sequence from at least two candidate feature components corresponding to each parent sample feature component according to at least two different weight functions, to obtain at least two temporary feature components corresponding to each parent sample feature component, wherein the number of temporary feature components corresponding to each parent sample feature component is equal to the number of weight functions;
[0113] In the specific implementation, for the parent sample of ship data The parent sample characteristic component Through the weight function ω 1 ,ω 2 , ..., ω s (s=1,2,…,S) in the parent sample feature component The corresponding candidate feature component z 1 ,z 2 ,…,Z M Select from among them, and assign each selected feature component to the temporary feature component The proposed distribution corresponding to the selected feature component is recorded as j * , traverse the parent sample of the ship data Each parent sample characteristic component Then obtain the characteristic component of each parent sample S temporary characteristic components of
[0114] S222: Perform a rejection sampling operation on each temporary feature component to obtain at least two candidate sample feature components corresponding to each parent sample feature component.
[0115] Specifically, the embodiment of the present invention needs to perform a rejection sampling operation on each parent sample feature component. The corresponding temporary characteristic components , so as to obtain the characteristic components of each parent sample Corresponding to at least two candidate sample feature components It should be understood that the number of candidate sample feature components corresponding to each parent sample feature component is determined by the number of temporary feature components, that is, the number of weight functions.
[0116] Since each parent sample characteristic component of the parent sample of the ship data is updated independently, on this basis, in order to take into account the efficiency of evolutionary sampling, the embodiment of the present invention introduces at least two weight functions, that is, the number of weight functions S≥2, and obtains candidate samples according to the characteristic components of each candidate sample obtained by the same weight function, and selects the optimal sub-sample using the standard deviation value of the candidate samples generated under different weight functions, that is, the weight function is adaptively selected by the evolutionary process, and the search strategy is dynamically adjusted according to the actual situation of the search process, balancing the global search and local search in the evolutionary process, further improving the efficiency of evolutionary sampling, expanding the adaptability of the sampling method, and taking into account the sampling efficiency while optimizing the sampling accuracy.
[0117] Furthermore, if Figure 4 As shown, selecting at least two candidate feature components corresponding to each parent sample feature component in turn according to at least two different weight functions to obtain at least two temporary feature components corresponding to each parent sample feature component, including:
[0118] S221a, obtaining at least two different weight functions according to the adjustment result of the weight parameters of the weight function, wherein the weight parameters represent the bias of the weight function;
[0119] The weight parameter β of the weight function can balance the distance of the candidate feature component from the current state and the probability under the target proposal distribution. Setting a suitable weight function with the weight parameter β can improve the efficiency of the sampling chain, help the evolutionary sampling algorithm move more effectively in the search space, and find better solutions.
[0120] Specifically, the expression of the weight function is shown in the following formula (1):
[0121]
[0122] in, express Except for the k-th parent sample feature component quilt Replace, and keep everything else the same, i.e.
[0123] for λ j,k(x,y) is a non-negative symmetric function. It determines the value of the weight function, and there are many options, such as: {T j,k (x,y)T j,k (y,x)} -β and 1.
[0124] In the embodiment of the present invention, j,k (x, y) takes into account both the transition probability and the moving distance, which is specifically expressed as shown in formula (2):
[0125] λ j,k (x,y)=T j,k (x,y) -1 ||yx|| β (2)
[0126] Among them, β represents the weight parameter that controls the bias of the weight function, which can balance the distance of the candidate feature component from the current state and the probability under the target proposal distribution, thereby affecting the evolution process. It needs to be set according to the actual problem.
[0127] S221b, respectively calculating the weight values of each candidate feature component corresponding to each parent sample feature component under each weight function;
[0128] The embodiment of the present invention sequentially calculates the weight values of the M candidate feature components corresponding to each parent sample feature component under S different weight functions, such as:
[0129] in, Indicates the parent sample of the ship data The k-th parent sample feature component The corresponding Mth feature component to be selected z M In the Sth weight function ω s The weight value below.
[0130] S221c, for each parent sample feature component, select the corresponding candidate feature components calculated by the same weight function as a group, and select a temporary feature component of the parent sample feature component from the corresponding candidate feature components by a random selection method according to the proportion of each group of weight values;
[0131] Specifically, for each parent sample feature component, the weight values of the corresponding M candidate feature components calculated by the same weight function are selected as a group, such as: As a group, a temporary feature component of the parent sample feature component is selected from the corresponding candidate feature components by random selection method according to the proportion of each group weight value The proposed distribution corresponding to the selected temporary feature component is recorded as j * .
[0132] S221d. After traversing each candidate feature component corresponding to each parent sample feature component, obtain at least two temporary feature components corresponding to each parent sample feature component.
[0133] Since the embodiment of the present invention adopts S different weight functions, for each parent sample feature component It is necessary to make S selections among the corresponding M candidate feature components to obtain each parent sample feature component S temporary characteristic components of
[0134] Furthermore, the at least two different weight functions include: a weight function with a weight parameter of 0 and a weight function with a weight parameter of 2.5.
[0135] When the weight parameter β is set to 2.5, the evolution process tends to explore in a wide search space, taking into account the target probability in the current state while maintaining sufficient exploratory power to discover possible better solutions. This setting helps the evolution process to jump out of the local optimal solution and find potential better solutions in the global scope. When the weight parameter β is set to 0, the evolution process focuses more on the candidate feature components with higher target probability in the current state, making the evolution process enter the local search stage, thereby increasing the probability of accepting high-quality solutions and accelerating the convergence speed of the algorithm.
[0136] Therefore, the embodiment of the present invention uses the standard deviation value of the candidate samples generated under different weight functions to select sub-samples in the subsequent evolution process, that is, the weight function is adaptively selected by the evolution process, which can dynamically adjust the search strategy according to the actual situation of the search process, so as to satisfy the algorithm to maintain global exploration while also being able to perform local search when gradually converging to the optimal solution in the subsequent process, thereby improving the adaptability and efficiency of the evolutionary method.
[0137] Furthermore, if Figure 5 As shown, a rejection sampling operation is performed on each temporary feature component to obtain at least two candidate sample feature components corresponding to each parent sample feature component, including:
[0138] S222a, respectively calculating the acceptance rate of each temporary feature component;
[0139] Specifically, the embodiment of the present invention takes two proposal distributions and two weight functions as an example, and the specific calculation formula of the acceptance rate of each temporary feature component is shown in formula (3):
[0140]
[0141] in, Represents the parent sample of ship data The k-th parent sample feature component The corresponding first candidate feature component z 1 In the first weight function ω 1 The weight value under Represents the parent sample of ship data The k-th parent sample feature component The corresponding second candidate feature component z 2 In the second weight function ω 2 The weight value under Represents temporary feature components In the corresponding proposal distribution j * The weight value under Represents temporary feature components The weight value under another proposed distribution.
[0142] S222b, generating random numbers in sequence, and if the generated random numbers are less than the acceptance rate of the corresponding temporary feature component, accepting the temporary feature component as a candidate sample feature component of the corresponding parent sample feature component;
[0143] Specifically, a random number is generated on [0,1] in sequence. If the generated random number is smaller than the corresponding temporary feature component The acceptance rate A, accepting the temporary feature component is the corresponding parent sample feature component The candidate sample feature components Right now
[0144] S222c, if the generated random number is greater than or equal to the acceptance rate of the corresponding temporary feature component, then rejecting the candidate sample feature component of the temporary feature component as the corresponding parent sample feature component, and determining the parent sample feature component as the candidate sample feature component;
[0145] If the generated random number is greater than or equal to the corresponding temporary characteristic component The acceptance rate A is higher, then the temporary feature components are rejected. is the corresponding parent sample feature component The candidate sample feature components Determine the characteristic component of the parent sample is the candidate sample feature component Right now
[0146] S222d. After traversing each temporary feature component corresponding to each parent sample feature component, obtain at least two candidate sample feature components corresponding to each parent sample feature component.
[0147] Based on the above steps, traverse each parent sample feature component Each temporary characteristic component of After that, the characteristic components of each parent sample are obtained The S candidate sample feature components
[0148] Furthermore, if Figure 6 As shown, obtaining at least one candidate sample of the parent sample of the ship data according to at least one candidate sample feature component corresponding to each parent sample feature component includes:
[0149] S310, forming candidate samples of the parent sample of the ship data according to the candidate sample feature components selected based on the same weight function of each parent sample feature component;
[0150] Since each parent sample feature component The S candidate sample feature components They are obtained through S weight functions respectively, and each parent sample feature component Corresponding candidate sample feature components selected based on the same weight function Form the parent sample of the ship data candidate samples.
[0151] S320, after traversing each candidate sample feature component corresponding to each parent sample feature component, obtain at least two candidate samples of the ship data parent sample.
[0152] After traversing each candidate sample feature component corresponding to each parent sample feature component, the parent sample of ship data is obtained S candidate samples
[0153] Furthermore, if Figure 7 As shown, a rejection sampling operation is performed on each candidate sample to obtain at least one candidate result sample of the parent sample of ship data, including:
[0154] S410, respectively calculating the acceptance rate of each candidate sample;
[0155] Specifically, the calculation formula of the acceptance rate of candidate samples is as follows:
[0156]
[0157] Among them, Q represents the target proposal distribution and p represents the multivariate kernel density estimation function.
[0158] S420, generating random numbers in sequence, and if the generated random numbers are less than the acceptance rate of the corresponding candidate samples, accepting the candidate samples as the candidate result samples of the parent samples of the ship data;
[0159] Specifically, a random number is generated on [0,1] in sequence. If the generated random number is smaller than the corresponding candidate sample Acceptance rate Then accept the candidate sample The parent sample of ship data Sample of candidate results That is
[0160] S430, if the generated random number is greater than or equal to the acceptance rate of the corresponding candidate sample, then reject the candidate sample as the candidate result sample of the parent sample of the ship data, and determine the parent sample of the ship data as the candidate result sample;
[0161] Specifically, if the generated random number is greater than or equal to the acceptance rate of the corresponding candidate sample Then reject the candidate sample The parent sample of ship data Sample of candidate results Determine the parent sample of the ship data Itself is a sample of the selected results, that is,
[0162] S440, after traversing each candidate sample of the parent sample of the ship data, obtaining at least one candidate result sample of the parent sample of the ship data.
[0163] Traverse the parent sample of ship data based on the above steps Each candidate sample Y t+1 After that, get the parent sample of ship data S samples of candidate results
[0164] Furthermore, if Figure 8 As shown, the acceptance rate of each candidate sample is calculated separately, including:
[0165] S001. Calculate the target bandwidth matrix of the multivariate kernel density estimation function according to the AMISE minimization bandwidth calculation method and the Kendall correlation coefficient;
[0166] Specifically, it is used to calculate candidate samples The expression of the multivariate kernel density estimation function of the acceptance rate is shown in the following formula (5):
[0167]
[0168] in, H represents a d×d symmetric positive definite bandwidth matrix, and |·| represents the computational determinant. It should be understood that That is, p in the above formula (4), K[·] represents the Gaussian kernel function, and the specific expression of the Gaussian kernel function is shown in formula (6):
[0169]
[0170] Among them, π represents pi, d represents dimension, and T represents matrix transpose.
[0171] In order to optimize the calculation process, the embodiment of the present invention calculates the target bandwidth matrix of the multivariate kernel density estimation function by using the AMISE minimization bandwidth calculation method to update the symmetric positive definite bandwidth matrix H.
[0172] Specifically, based on the optimal diagonal bandwidth empirical method of the AMISE minimization bandwidth calculation method, the expression of the multivariate kernel density estimation AMISE can be derived according to the derivation formula of the univariate kernel density estimation AMISE as shown in formula (7):
[0173]
[0174] Among them, det means to find the determinant, d means the dimension, tr{·} means the trace of the matrix, and f(Y) means to calculate the candidate sample The multivariate kernel density estimation function of , f"(Y) represents the second-order derivative of f(Y), that is, the Hessian Matrix. Assuming that f(Y) obeys the multivariate normal distribution, the target bandwidth matrix that minimizes AMISE can be derived, as shown in formula (8):
[0175]
[0176] Among them, σ i represents the standard deviation of the i-th dimension of the sample. This method provides a simple and effective way to select an appropriate bandwidth matrix, especially for data sets with normal distribution characteristics.
[0177] The specific expression of the target bandwidth matrix is as follows:
[0178]
[0179] Among them, ρ is the Kendall correlation coefficient (Kendall correlation coefficient), and the Kendall correlation coefficient has the advantages of being more stable and reliable in evaluating the correlation between variables and being more sensitive to nonlinear relationships. Furthermore, considering that in a specific candidate sample, there may be correlation between the characteristic components of each candidate sample, and this correlation may affect the accuracy of the kernel density estimation function, the embodiment of the present invention reflects the correlation of variables in the target bandwidth matrix, and at the same time considers the different effects of strong correlation and weak correlation on the kernel density estimation, and sets the elements on the non-diagonal line to the square value of the correlation coefficient between variables, thereby amplifying the influence between strongly correlated variables and reducing the influence of weak correlation values. In addition, by retaining the sign of the correlation coefficient, it reflects whether the variables are positively correlated or negatively correlated, so as to more accurately adjust the kernel density estimation to reflect the true distribution of the data.
[0180] Since the embodiment of the present invention has a known function when calculating the acceptance rate of the temporary feature component, there is no need to use the multivariate kernel density estimation function. If there is no known function, the multivariate kernel density estimation function can also be used to calculate the temporary feature component. The acceptance rate is as shown in the following formula (11):
[0181]
[0182] Where s = 1, 2, ..., S, it should be understood that It refers to the temporary characteristic component under the same weight function, H represents the d×d symmetric positive definite bandwidth matrix, |· represents the calculation determinant. K[·] represents the Gaussian kernel function, and the specific representation form of the target bandwidth matrix H and the form of the correlation coefficient are the same as above, which will not be repeated here.
[0183] S002. Calculate the acceptance rate of each candidate sample or the acceptance rate of each temporary feature component according to the target bandwidth matrix.
[0184] The embodiment of the present invention generates a target bandwidth matrix of a multivariate kernel density estimation function based on the AMISE minimization bandwidth calculation method and the Kendall correlation coefficient through the sample information collected each time, and uses the multivariate kernel density estimation function to calculate the acceptance rate of temporary feature components and the acceptance rate of candidate samples, thereby effectively avoiding the subjectivity and lack of adaptability caused by manual setting of bandwidth parameters, thereby reducing the risk of decreased accuracy of the acceptance rate calculated by the multivariate kernel density estimation function due to improper bandwidth parameter setting, and effectively ensuring the performance of the ship data sampling method.
[0185] Furthermore, if Fig. 9As shown, if the number of candidate result samples is at least 2, a sub-sample of the parent sample of ship data is determined from at least two candidate result samples according to the degree of dispersion of each candidate result sample, including:
[0186] S610, if the number of the candidate result samples is at least 2, respectively calculate the standard deviation value of each group of candidate result samples;
[0187] Specifically, if the number of candidate result samples is at least 2, calculate the parent sample of ship data respectively Sample of candidate results The standard deviation of
[0188] S620. Determine the candidate result sample with the smallest standard deviation value as the subsample of the parent sample of the ship data.
[0189] In the ship data parent sample S samples of candidate results In the above example, the candidate result sample with the smallest standard deviation value is selected as the parent sample of ship data. The subsample of the ship data is updated
[0190] The simulation experiment data provided by the embodiment of the present invention are as follows:
[0191] Table 1 shows the test function set required for the simulation experiment, which consists of 15 test functions, divided into three single-peak functions π 1 ~π 3 , 6 slightly local optimal functions π 4 ~π 9 and 6 obvious local optimal functions π 10 ~π 15 , for some functions, the variable correlation is set to include only positive correlation between variables, only negative correlation between variables, and both positive and negative correlation between variables, while the other functions keep the variables independent of each other.
[0192] Table 1 Test functions
[0193]
[0194]
[0195] Table 2 shows the 9 test functions π randomly selected from Table 1. 1 , π 2 , π 4 , π 6 , π 8 , π 10 , π 11 , π 12 and π13 , and randomly set different variance sizes for each variable.
[0196] Table 2 Test functions (different distribution ranges between variables)
[0197]
[0198]
[0199] The performance of the evolutionary sampling algorithm is usually evaluated using standard deviation and convergence. The convergence of the algorithm is observed by calculating the standardization factor using the following formula (12). When λ converges to a stable value, we determine that the algorithm converges and obtains an optimal stable solution. Therefore, when the difference between λ of two adjacent iterations is less than a very small value, the algorithm can be determined to terminate.
[0200]
[0201] According to the following formula (13), the standard deviation J can be used to indicate the approximation between the distribution of the support sample set and the distribution of the target sample set. The smaller J is, the closer the distribution of the support sample set is to the distribution of the target sample set, and the higher the accuracy of the sample.
[0202]
[0203] Where p(x) = λ * p * (x), p(x) represents the standardized output, and π(x) represents the target distribution. Set the data points uniformly distributed in the target definition domain to calculate the standard deviation.
[0204] The specific experimental settings are as follows: 300 random samples that meet the range of each dimension are generated from a uniform distribution as the initial sample set, that is, N = 300; the number of iterations Iter = 300. In order for the evaluation index to show the difference between the algorithms of different transfer density functions, we set a corresponding number of test points uniformly distributed in the definition domain to calculate the values of λ and J, and repeat the operation 30 times for each test function. The influence of different values of the weight function parameter α in the single-component multiple-attempt evolutionary sampling (LCMTES) algorithm based on Lévy distribution on the algorithm performance is verified, and the performance is compared with that of the adaptive single-component evolutionary sampling (LACES) algorithm based on Lévy distribution when processing functions with the same distribution range and different distribution ranges between variables. Among them, the LACES algorithm is an evolutionary sampling algorithm based on adaptive Lévy distribution that uses a component-by-component update strategy but does not use the multiple-attempt strategy in the ship data sampling method provided in the embodiment of the present invention.
[0205] Experiment 1: Select all test functions in Table 1, set the weight function parameter α = 0, α = 2.5 and α = [0, 2.5], verify the effect of different values of the weight function parameter α in the LCMTES algorithm on the algorithm performance and compare the performance with the LACES algorithm in the same distribution range of processing variables. The algorithms involved are abbreviated as follows
[0206] LCMTES-1: parameter α of the weight function = 0;
[0207] LCMTES-2: weight function parameter α = 2.5;
[0208] According to Table 3, the standard deviations of LCMTES-1, LCMTES-2 and LCMTES algorithms are compared. It can be seen that on unimodal functions, the performance of the LCMTES-1 algorithm can be similar to or even slightly better than that of the LCMTES-2 algorithm. Therefore, the method of only considering the target probability value in the current state for the weight value is suitable for unimodal functions. However, on functions with local optimal values, it can be clearly seen that the performance of the LCMTES-2 algorithm is better than that of the LCMTES-1 algorithm, and the standard deviation is reduced by 3.9% or more on most functions. Therefore, for functions with local optimal values, compared with functions with only considering the target probability value in the current state, the weight function setting that balances the jump distance from the current state and the probability under the target distribution π is more suitable. Since the LCMTES algorithm considers both α=0 and α=2.5, it can be clearly seen from Table 3 that the performance of the LCMTES algorithm is better than that of LCMTES-1 and LCMTES-2.
[0209] Compared with the LACES algorithm, it can be seen from Table 3 that in the case of the single-peak function π 1 and π 3 On the function with slight local optimum, for the function π 4 and π 5 , the standard deviation obtained by the LACES algorithm is slightly smaller than that of the LCMTES algorithm. 6 , π 7 and π 8 , the standard deviation of the LCMTES algorithm is slightly smaller. On functions with obvious local optima, the performance of the two algorithms is similar, and the standard deviation does not differ by more than 2%. In summary, when the differences between variables are the same, the performance of the LCMTES algorithm and the LACES algorithm is similar.
[0210] Table 3 Standard deviation and normalization factor of the last generation of samples
[0211]
[0212]
[0213] Experiment 2: Select the 9 test functions π in Table 1 1 , π 2 , π 4 , π 6 , π 8 , π 10 , π 11 , π 12 and π 13 , randomly set different variance sizes for each variable, as shown in Table 2, comparing the performance of LACES and LCMTES algorithms when processing functions with different distribution ranges between variables.
[0214] According to Table 4, it can be seen that in the single-peak function π 1 and π 2 In general, the standard deviation of the LACES algorithm is similar to that of the LCMTES algorithm, and the difference between the two does not exceed 2%. Fig.14 , Fig.16 , Fig.18 , Fig. 20 , Fig. 22 , Fig.24 and Fig.26 It can be clearly seen that the LCMTES algorithm not only shows a higher descent rate, but also has a lower standard deviation, and its performance is significantly better than the LACES algorithm. Fig.15 , Fig.17 and Fig.19 It can be seen that compared with the LACES algorithm, the LCMTES algorithm converges faster and its standard deviation is reduced by 7.5% or more. For functions with obvious local optimality, Fig.21 , Fig.23 , Fig.25 and Fig. 27 It can be seen that the speed at which the LCMTES algorithm converges to a stable solution is slightly slower than that of the LACES algorithm, but combined with Fig. 20 , Fig. 22 , Fig.24 and Fig.26 , it can be seen that the standard deviation of the LCMTES algorithm is obviously lower than that of the LACES algorithm. Compared with the LACES algorithm, the standard deviation of the LCMTES algorithm is reduced by more than 20%. In summary, when the target distribution is relatively simple, despite the difference in the variable range, the samples collected by the LACES algorithm can also have a good approximation effect on the target function. When the target function is more complex, the samples collected by the LCMTES algorithm can better approximate the target distribution than the LACES algorithm and have higher accuracy.
[0215] Table 4 Standard deviation and normalization factor of the last generation of samples
[0216]
[0217] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.
Claims
1. A ship data sampling method, characterized in that: include: Acquire a parent sample of ship data for predicting ship stability, wherein the parent sample of ship data includes a plurality of parent sample feature components, and each parent sample feature component corresponds to a ship stability feature; Performing data augmentation processing on each of the parent sample feature components to obtain at least one candidate sample feature component corresponding to each of the parent sample feature components; Obtain at least one candidate sample of the parent sample of the ship data according to at least one candidate sample feature component corresponding to each of the parent sample feature components; Perform a rejection sampling operation on each of the candidate samples to obtain at least one candidate result sample of the ship data parent sample, wherein the number of the candidate result samples of the ship data parent sample, the number of the candidate samples of the ship data parent sample, and the number of the candidate sample feature components corresponding to each of the parent sample feature components are all equal; If the number of the candidate result samples is 1, determining that the candidate result sample is a subsample of the parent sample of the ship data; If the number of the candidate result samples is at least 2, determining one of the candidate result samples as a sub-sample of the parent sample of the ship data from at least two candidate result samples according to the degree of dispersion of each candidate result sample; The parent sample of the ship data is updated according to the sub-sample, and the above steps are repeated until a preset number of iterations is met to obtain a final sample.
2. The method according to claim 1, characterized in that The step of performing data augmentation processing on each of the parent sample feature components to obtain at least one candidate sample feature component corresponding to each of the parent sample feature components includes: Generate at least two candidate feature components corresponding to each of the parent sample feature components according to at least two different proposal distributions; Obtaining at least one candidate sample feature component corresponding to each of the parent sample feature components according to at least one weight function and at least two candidate feature components corresponding to each of the parent sample feature components; The number of candidate sample feature components corresponding to each parent sample feature component is equal to the number of weight functions.
3. The method according to claim 2, characterized in that The step of obtaining at least one candidate sample feature component corresponding to each of the parent sample feature components according to at least one weight function and at least two candidate feature components corresponding to each of the parent sample feature components comprises: Selecting in turn at least two candidate feature components corresponding to each of the parent sample feature components according to at least two different weight functions to obtain at least two temporary feature components corresponding to each of the parent sample feature components, wherein the number of temporary feature components corresponding to each of the parent sample feature components is equal to the number of the weight functions; A rejection sampling operation is performed on each of the temporary feature components to obtain at least two candidate sample feature components corresponding to each of the parent sample feature components.
4. The method according to claim 3, characterized in that The step of selecting in sequence from the at least two candidate feature components corresponding to each of the parent sample feature components according to at least two different weight functions to obtain at least two temporary feature components corresponding to each of the parent sample feature components includes: Obtaining at least two different weight functions according to a result of adjusting a weight parameter of the weight function, wherein the weight parameter represents a bias of controlling the weight function; Respectively calculating the weight values of each candidate feature component corresponding to each of the parent sample feature components under each of the weight functions; For each of the parent sample feature components, select the corresponding candidate feature components calculated by the same weight function as a group, and select a temporary feature component of the parent sample feature component from the corresponding candidate feature components by a random selection method according to the proportion of each group of weight values; After traversing each candidate feature component corresponding to each of the parent sample feature components, at least two temporary feature components corresponding to each of the parent sample feature components are obtained.
5. The method according to claim 3, characterized in that: The at least two different weight functions include: A weight function with a weight parameter of 0 and a weight function with a weight parameter of 2.
5.
6. The method according to claim 3, characterized in that: The step of performing a rejection sampling operation on each of the temporary feature components to obtain at least two candidate sample feature components corresponding to each of the parent sample feature components includes: Calculating the acceptance rate of each temporary feature component respectively; Generate random numbers in sequence, and if the generated random numbers are less than the acceptance rate of the corresponding temporary feature component, accept the temporary feature component as a candidate sample feature component of the corresponding parent sample feature component; If the generated random number is greater than or equal to the acceptance rate of the corresponding temporary feature component, then the temporary feature component is rejected as a candidate sample feature component of the corresponding parent sample feature component, and the parent sample feature component is determined as a candidate sample feature component; After traversing each temporary feature component corresponding to each parent sample feature component, at least two candidate sample feature components corresponding to each parent sample feature component are obtained.
7. The method according to claim 3, characterized in that The step of obtaining at least one candidate sample of the parent sample of the ship data according to at least one candidate sample feature component corresponding to each parent sample feature component comprises: Forming candidate samples of the parent sample of the ship data according to the candidate sample feature components selected based on the same weight function of each of the parent sample feature components; After traversing each candidate sample feature component corresponding to each parent sample feature component, at least two candidate samples of the parent sample of the ship data are obtained.
8. The method according to claim 1, characterized in that The step of performing a rejection sampling operation on each candidate sample to obtain at least one candidate result sample of the parent sample of the ship data comprises: Calculate the acceptance rate of each candidate sample respectively; Generate random numbers in sequence, and if the generated random numbers are less than the acceptance rate of the corresponding candidate samples, then accept the candidate samples as the candidate result samples of the parent samples of the ship data; If the generated random number is greater than or equal to the acceptance rate of the corresponding candidate sample, the candidate sample is rejected as the candidate result sample of the parent sample of the ship data, and the parent sample of the ship data is determined as the candidate result sample; After traversing each candidate sample of the parent sample of the ship data, at least one to-be-selected result sample of the parent sample of the ship data is obtained.
9. The method according to claim 8, characterized in that The respectively calculating the acceptance rate of each candidate sample comprises: Calculate the target bandwidth matrix of the multivariate kernel density estimation function based on the AMISE minimization bandwidth calculation method and the Kendall correlation coefficient; The acceptance rate of each candidate sample is calculated according to the target bandwidth matrix.
10. The method according to claim 1, characterized in that If the number of the candidate result samples is at least 2, determining a sub-sample of the parent sample of the ship data from at least two candidate result samples according to the discrete degree of each candidate result sample, comprising: If the number of the candidate result samples is at least 2, respectively calculate the standard deviation value of each group of the candidate result samples; The candidate result sample with the smallest standard deviation value is determined as the subsample of the parent sample of the ship data.
Citation Information
Patent Citations
Ship rolling integrated prediction method and system based on deep learning
CN114528756A
Ship stability forecasting method and system based on GD weighted fusion RBFNN and random forest
CN116424508A
Ship data evolution sampling method based on Levy distribution
CN116720443A
Method for data imputation and classification and system for data imputation and classification
US20200193220A1
Method for enhancing SAR image data for ship target detection
WO2023050746A1
