Wave system space tracking method based on parameter continuity

Abstract

The invention belongs to the technical field of wave numerical simulation, and provides a wave system space tracking method based on parameter continuity, which comprises the following steps: S1, acquiring parameters of a wave system of all grid points in a research area through a numerical model; s2, detecting and combining the pseudo wave system and the homologous wave system of each grid point, and calculating parameters of a combined new wave system; and S3, determining a search starting point, and clustering the wave systems which are homologous with the wave system at the starting point in the computational domain by using a spatial search algorithm so as to obtain wave fields of the wave systems with different sources. According to the method, the problem of discontinuous tracking results or too many tracking results caused by interference of a pseudo-wave system or a homologous wave system in an existing method can be solved, so that a wave field which better conforms to actual conditions can be obtained.

Description

Technical Field

[0001] The invention belongs to the technical field of ocean wave numerical simulation, and in particular relates to a wave system spatial tracking method based on parameter continuity. Background Art

[0002] Ocean waves usually refer to the fluctuations on the sea surface caused by wind. They have an important impact on many fields such as ship navigation, marine engineering construction, and marine resource development. They are one of the hydrological elements that focus on marine scientific research and engineering applications. Under the joint action of different meteorological activities in the ocean, wave components of different origins overlap to form mixed waves. For wave components generated by the same meteorological activity, their period and wave direction characteristics are similar, which is called a wave system. The wave system is generated by specific meteorological activities. Clarifying its origin and related meteorological activities is of great significance for understanding the formation, characteristics and wave evolution of mixed waves.

[0003] The two-dimensional spectrum gives the detailed distribution of wave energy in frequency and direction. Using the two-dimensional spectrum segmentation technology, the wave systems that make up the waves can be segmented and identified, and each wave system can be parameterized (such as wave height, period, and wave direction, etc.). Based on the characteristic that the parameters of the same source wave system maintain continuity at adjacent spatial points, the wave system of each grid point can be clustered according to its origin using a spatial search algorithm, and the wave field of a wave system of a specific origin can be tracked.

[0004] However, existing spatial tracking methods usually directly use the two-dimensional spectrum segmentation results given by the numerical model as input, which has two serious problems:

[0005] (1) Since two-dimensional spectrum segmentation is essentially an image segmentation algorithm, there will be a problem of over-segmentation, which will lead to the appearance of unreasonable pseudo-wave systems in the two-dimensional spectrum segmentation results, thereby significantly reducing the accuracy and stability of subsequent spatial tracking.

[0006] (2) In terms of the identification of homologous wave systems, existing methods generally use the continuity criteria of wave height, period and wave direction for evaluation. When the study area is large, wave systems formed at different times will coexist. According to the linear wave dispersion relationship, wave systems with large periods propagate faster than wave systems with small periods. The large-period wave systems formed later will catch up with the small-period wave systems formed earlier, resulting in the simultaneous existence of multiple homologous wave systems at the same spatial point or the periods of homologous wave systems having mutations at adjacent grid points. Since the periods of these homologous wave systems vary greatly, existing methods use the period as one of the main evaluation parameters. When encountering the above situation during spatial tracking, interruptions often occur, resulting in missing or excessive wave field data for spatial tracking, which causes the spatial tracking results to be inconsistent with the actual situation.

[0007] The patent with publication number CN111222531A discloses a method and system for extracting information on the spatiotemporal evolution of a wave field. Although the patent proposes a method for finding homologous wave systems, it does not take into account the phenomenon that there are multiple homologous wave systems at the same spatial point or that the period of a homologous wave system has mutations at adjacent grid points. The patent still directly uses the segmentation result of the two-dimensional spectrum as input, and the wave field finally obtained is quite different from the actual situation. Summary of the invention

[0008] In order to overcome the shortcomings of the prior art, the present invention provides a wave system spatial tracking method based on parameter continuity. The present invention can solve the problem of intermittent or excessive tracking results caused by interference from pseudo wave systems or homologous wave systems in the prior art, thereby obtaining a wave field that is more in line with the actual situation.

[0009] The above object of the present invention is achieved through the following technical solutions:

[0010] A wave system spatial tracking method based on parameter continuity comprises the following steps:

[0011] S1. Obtain the parameters of the wave system at all grid points in the study area through the numerical model;

[0012] S2, detecting and merging the pseudo wave system and the homologous wave system of each grid point, and calculating the parameters of the merged new wave system;

[0013] S3. Determine the search starting point, and use the spatial search algorithm to cluster the wave systems in the calculation domain that have the same origin as the wave system at the starting point, so as to obtain the wave fields of wave systems with different origins.

[0014] Furthermore, the parameters of the wave system in S1 are the peak period T of each wave system. p , effective wave height H s , average wave direction θ m and the average directional distribution width σ m .

[0015] Furthermore, the specific method for detecting and merging the pseudo wave system and the homologous wave system in S2 is as follows:

[0016] The average wave direction θ m The difference between the two is used to judge the proximity of the wave system in the spectral domain, and the average directional distribution width σ m To evaluate whether the development state of the wave system is similar; when the average wave direction θ of any two wave systems in the spectrum domain m and the average directional distribution width σ mThe relationship of formula (1) is satisfied, indicating that the two wave systems are extremely close in the spectrum domain and are at a similar stage of development, that is, they correspond to the same origin, so they are merged into a new wave system;

[0017]

[0018] In the above formula, superscripts i and j represent wave system i and wave system j in the current spectrum domain, respectively. represents the average wave direction of wave system i, represents the average wave direction of wave system j, represents the average directional distribution width of wave system i, represents the average directional distribution width of wave system j, θ x , σ x Represent the set coefficients respectively.

[0019] Furthermore, the parameter calculation formula of the new wave system after merging in S2 is as follows:

[0020]

[0021] In the above formula, represents the peak period of wave system i, represents the significant wave height of wave system i, represents the average wave direction of wave system i, represents the average directional distribution width of wave system i;

[0022] represents the peak period of wave system j, represents the significant wave height of wave system j, represents the average wave direction of wave system j, represents the average directional distribution width of wave system j;

[0023] Represents the peak period of the new wave system after merging, represents the effective wave height of the new wave system after merging, Represents the average wave direction of the new wave system after merging. Represents the average directional distribution width of the new wave system after the merger.

[0024] Furthermore, the specific steps of S3 are:

[0025] T1. Set the starting point of spatial tracking and mark all grid points as untracked;

[0026] T2, put a wave system at the starting point into the search sequence and set it as the benchmark wave system, delete the information of the benchmark wave system in the original data set, and mark the starting point as tracked;

[0027] T3, sequentially calculating the similarity between each wave system at all untracked adjacent grid points and the benchmark wave system;

[0028] T4. Find the wave system in the neighboring grid points that is most similar to the benchmark wave system. If the similarity between the wave system and the benchmark wave system is less than the threshold S th , then the wave system is considered to be of the same origin as the benchmark wave system, the wave system is added to the search sequence, and the grid point where the wave system is located is marked as tracked;

[0029] T5. If no homologous wave system can be found in the wave systems of adjacent grid points, the reference wave system is placed in a result sequence and deleted from the search sequence, and the next wave system in the search sequence is set as the reference wave system;

[0030] T6, repeating steps T3-T5 until there is no wave system in the search sequence, indicating that the tracking of the wave system with the same origin as the wave system selected at the starting point in the calculation domain is completed, and the current result sequence is saved and output;

[0031] T7, reset the tracking status of each grid point to untracked, put another wave system at the starting point into the search sequence, set the wave system as the reference wave system, delete the information of the reference wave system in the original data set, and mark the starting point as tracked;

[0032] T8. Repeat steps T3-T7 until all wave systems at the starting point have been searched, completing the spatial tracking of all wave systems at the starting point.

[0033] Furthermore, the calculation formula of the similarity in T3 is:

[0034]

[0035] In the above formula, represents the significant wave height of the reference wave system, Represents the directional distribution width of the benchmark wave system, Represents the average wave direction of the benchmark wave system;

[0036] represents the effective wave height of the wave system at the adjacent point, represents the directional distribution width of the wave system at the neighboring point, Represents the average wave direction of the wave system at the adjacent point;

[0037] S represents the similarity, d represents the geographical distance between two grid points; H a , σ b ,θ c Represent weight factors respectively.

[0038] Furthermore, the adjacent grid points are 8 grid points connected to the grid point where the reference wave system is located, namely, the top, bottom, left, right, upper left, upper right, lower left, and lower right points of the grid point where the reference wave system is located.

[0039] The beneficial effects of the present invention compared with the prior art solutions are:

[0040] 1. After the two-dimensional spectrum is segmented, the present invention will detect and merge the pseudo-wave system and the homologous wave system, thereby improving the quality of the benchmark data set and eliminating the influence of the existence of multiple homologous wave systems at the same spatial point and the mutation of the period of the homologous wave system at adjacent grid points on subsequent tracking.

[0041] 2. The present invention uses the average wave direction θ when detecting and merging the pseudo wave system and the homologous wave system. m and the average directional distribution width σ m To determine whether the wave system can be merged. The present invention adopts the average wave direction θ m The difference between the two is used to judge the proximity of the wave system in the spectral domain, and the average directional distribution width σ m To evaluate whether the development status of the wave system is similar. m A small difference means that the distance between the two wave systems in the spectral domain is close. If the directional distribution width m is also similar, it indicates that their development states are also similar. Therefore, these two parameters can fully ensure that the merged wave system is of the same origin and effectively prevent the merging of wave systems from different sources.

[0042] 3. The present invention proposes a method based on the effective wave height H s , average wave direction θ m and the average directional distribution width σ m The continuity criterion of the wave system is used to evaluate the similarity of the wave system, which solves the problem of the existing technology that relies on the peak period T of the wave system. p However, the problem of difficulty in accurately tracking homologous wave systems generated at different times is that the peak period is discontinuous. The tracking method of the present invention is more stable, and the tracking results are more consistent with the actual situation, which is of great significance for the research work of analyzing the origin and evolution of wave systems using the spatial characteristics of wave fields.

[0043] 4. In steps T3-T5, the present invention searches for whether there is a homologous wave system at all grid points by searching the neighboring grid points of each reference wave system. Since the parameters of the homologous wave system maintain continuity at adjacent spatial points, when a homologous wave system exists at adjacent grid points, the wave system of the adjacent grid points can be used as a new reference wave system to continue searching for untracked grid points. The existing search method is to find a homologous wave system by directly searching all grid points, while the search method of the present invention can effectively reduce the search for some grid points where it is impossible to have a homologous wave system compared to the existing method, greatly improving the search speed and efficiency and avoiding the waste of computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0045] Figure 2 It is a schematic diagram of the two-dimensional spectrum at a certain grid point, the wave system composition and the wave parameters of each wave system obtained by using the numerical model WAVEWATCHIII in Example 1 of the present invention;

[0046] Figure 3 In Example 1 of the present invention, Figure 2 The original spectrum segmentation result is shown as a schematic diagram of the wave system composition and corresponding wave parameters after the wave system is merged;

[0047] Figure 4 is a schematic diagram of the spatial tracking result of the target point wave system using the method of the present invention in Example 1 of the present invention;

[0048] Figure 5 The present invention does not use the wave system merging method in Comparative Example 1, and uses the existing parameters (based on the effective wave height H s , peak period T p and the average wave direction θ m ) is a schematic diagram of the tracking results obtained. DETAILED DESCRIPTION

[0049] The present invention is described in detail below through specific embodiments, but the protection scope of the present invention is not limited thereto.

[0050] Example 1

[0051] Figure 1 The figure shows the overall flow chart of the wave system spatial tracking method based on wave parameter continuity of the present invention. In order to more clearly demonstrate and explain the technical scheme and principle of the present invention, the following is a step-by-step explanation in combination with a specific implementation example in a certain sea area and the accompanying drawings. The steps are as follows:

[0052] S1: Use the numerical model to obtain the wave system parameters at each grid point in the study area;

[0053] Firstly, the numerical model WAVEWATCHIII (v6.07) was used to obtain the two-dimensional spectrum and wave system segmentation results at each grid point in a certain sea area. Figure 2 The two-dimensional spectrum and wave system composition of a grid point position (19°25′29″S, 85°04′39″W) are given. From the two-dimensional spectrum, we can see that there are four energy clusters (peaks), that is, four wave systems, which are recorded as W1, W2, W3 and W4 respectively. The wave parameters of each wave system, that is, the effective wave height H s , peak period T p , average wave direction θ m and the average directional distribution width σ m All are certain.

[0054] S2: Merge the pseudo wave system and the homologous wave system at each grid point, and calculate the wave parameters of the new wave system after the wave system is merged;

[0055] The two-dimensional spectrum segmentation method is an image segmentation method that relies on the morphological characteristics of the spectrum to identify and separate wave systems. It is inevitable that there is a problem of over-segmentation, which leads to the generation of pseudo-wave systems. In addition, there are often multiple homologous wave systems at the same grid point. Due to the lack of consideration of the actual situation, the two-dimensional spectrum segmentation cannot accurately determine whether the segmented wave systems are homologous. In order to ensure the stability and accuracy of subsequent wave system tracking, the present invention proposes to eliminate the pseudo-wave system of each grid point in advance and merge the homologous wave systems. This approach greatly improves the quality of tracking benchmark data. When the wave parameters of any two wave systems at the same spatial point satisfy formula (7), it is determined that the two wave systems correspond to the same origin and are merged:

[0056]

[0057] In the above formula, superscripts i and j represent wave system i and wave system j at the current grid point, respectively. represents the average wave direction of wave system i, represents the average wave direction of wave system j, represents the average directional distribution width of wave system i, represents the average directional distribution width of wave system j.

[0058] Calculate the wave parameters of the new wave system after merging as follows:

[0059]

[0060] In the above formula, represents the peak period of wave system i, represents the significant wave height of wave system i, represents the average wave direction of wave system i, represents the average directional distribution width of wave system i;

[0061] represents the peak period of wave system j, represents the significant wave height of wave system j, represents the average wave direction of wave system j, represents the average directional distribution width of wave system j;

[0062] Represents the peak period of the new wave system after merging, represents the effective wave height of the new wave system after merging, Represents the average wave direction of the new wave system after merging. Represents the average directional distribution width of the new wave system after the merger.

[0063] Figure 3 Given the Figure 2 Merged results for the medium wave system. Figure 2 It can be seen that in Figure 3 In the above equation, the number of wave systems is reduced to 3. According to the combined criterion of formula (7), Figure 2 The θ of wave systems W3 and W4 in m and σ m The values ​​are extremely close, which means that the two wave systems are close in distance in the spectrum domain and have similar development states, indicating that they are generated by the same meteorological activity. Therefore, they are merged and the parameters of the merged wave system are recalculated.

[0064] S3: Determine the search starting point, use the spatial search algorithm to search for the wave systems of all grid points, cluster the wave systems in the computational domain that are homologous to the wave system at the starting point, and thus obtain the wave fields of wave systems with different origins. The process includes:

[0065] (1) Select the starting point of spatial tracking and mark all grid points as untracked;

[0066] (2) Select a wave system at the starting point as the reference wave system, mark the starting point as tracked, and put the reference wave system into a search sequence that follows the first-in-first-out rule. At the same time, delete the information of the reference wave system in the original data set;

[0067] (3) The similarity between each wave system and the benchmark wave system at the untracked grid points adjacent to the benchmark wave system is calculated in turn. The similarity is evaluated based on the continuity criterion of wave height, direction distribution and wave direction:

[0068]

[0069] In the above formula, represents the significant wave height of the reference wave system, Represents the directional distribution width of the benchmark wave system, Represents the average wave direction of the benchmark wave system; represents the effective wave height of the wave system at the adjacent point, represents the directional distribution width of the wave system at the neighboring point, represents the average wave direction of the wave system at the neighboring points; S represents the similarity, d represents the geographical distance between the two grid points; H a , σ b ,θ c They represent weight factors respectively and can be adjusted according to actual needs. The recommended weight factor is H a , σ b ,θ c Set to 2m, 5° and 5° respectively.

[0070] The adjacent grid points are 8 grid points connected to the grid point where the reference wave system is located, namely, the top, bottom, left, right, upper left, upper right, lower left, and lower right points of the grid point where the reference wave system is located.

[0071] (4) Find the wave system in each adjacent grid point that is most similar to the benchmark wave system. If the similarity of the wave system is less than the set threshold S th , the wave system is considered to be of the same origin as the reference wave system, the wave system is stored in the search sequence, the wave system information is deleted from the original data set, and the adjacent grid point is marked as tracked. Preferably, the threshold in is the significant wave height value of the reference wave system;

[0072] (5) If the same wave system is still not found after searching all adjacent points, the current reference wave system is placed in a result sequence and deleted from the search sequence, and then the next wave system in the search sequence is set as the reference wave system, and the spatial search continues;

[0073] (6) Repeat steps (3) to (5) until the search sequence is empty, indicating that the tracking of the wave system with the same origin as the wave system selected at the starting point in the spectrum domain is completed, and the corresponding tracking result sequence is saved and output;

[0074] (7) Reset the tracking status of each grid point to untracked, put another wave system at the starting point into the search sequence, set the wave system as the reference wave system, delete the information of the reference wave system in the original data set, and mark the starting point as tracked;

[0075] (8) Repeat steps (3)-(7) until the wave system at the starting point is empty, indicating that the spatial tracking of all wave systems at the starting point is completed.

[0076] Figure 4 For a certain research area, the starting point Figure 3 Spatial tracking results of the three wave systems in .

[0077] Comparative Example 1:

[0078] In order to verify the effect of the tracking method of the present invention, according to the technical scheme of the existing method, the pre-merging link of the homologous wave system in step S2 of embodiment 1 is removed, and the calculation parameter (formula (12)) of the wave system similarity in step S3 is adjusted to the effective wave height H s , peak period T p and the average wave direction θ m , the remaining steps remain unchanged, and the tracing starting point is the same as in Example 1.

[0079] Figure 5 To obtain the wave system tracking result, since the wave system merging operation is not performed on the spectrum segmentation result before tracking, the comparative embodiment is based on Figure 2 The original four wave systems in the space tracking are used to obtain the tracking results of the four wave fields. Figure 5 (c) and Figure 5 From the wave field distribution range of (d), it can be seen that the two have a large overlap, which can be inferred that the two wave systems are of the same origin but generated at different times. Because in the process of spatial propagation, according to the wave dispersion relationship, wave systems formed at different times will meet, resulting in the existence of multiple homologous wave systems at the same grid point. The existing method directly uses the spectrum segmentation data as input, and the error of the spectrum segmentation result has a great impact on the tracking result. In addition, when there are multiple homologous wave systems at the same spatial point, the existing method is difficult to accurately identify, and thus the wave field of two homologous wave systems is tracked incorrectly. In addition, the existing method uses the peak period T p As one of the important parameters for evaluating the similarity of wave systems, T will appear when wave systems generated at different times meet. p Sudden change, when the difference of wave parameters exceeds the set threshold, it will cause tracking interruption, which is also Figure 5 (b) and Figure 5 The wave field shown in (c) is incomplete, and the area is significantly smaller than the result of Example 1 (see Figure 4 ) reasons.

[0080] The method of the present invention applies a technical solution of merging wave systems before wave system tracking. On the one hand, the pseudo wave system generated by over-segmentation in the spectral segmentation algorithm is effectively eliminated to reduce the interference of input data set errors on wave system tracking; on the other hand, wave system merging can merge multiple homologous wave systems coexisting on the same grid point into a whole, avoiding the omission of wave system components during subsequent spatial tracking. In addition, the present invention also proposes a new wave parameter combination for evaluating the similarity of wave systems, which combines the average directional distribution width σ m It is introduced, so that the homologous wave systems formed at different times can be completely presented in the same wave field, which solves the problems of large number of tracking results and missing data in existing methods.

[0081] The above-described embodiments are only preferred embodiments of the present invention, but not all feasible embodiments of the present invention. For those skilled in the art, any obvious changes made thereto without departing from the principles and spirit of the present invention should be considered to be included in the scope of protection of the claims of the present invention.

Claims

1. A wave system spatial tracking method based on parameter continuity, characterized in that: The following steps are involved: S1. Obtain the parameters of the wave system at all grid points in the study area through the numerical model; S2, detecting and merging the pseudo wave system and the homologous wave system of each grid point, and calculating the parameters of the merged new wave system; S3. Determine the search starting point, and use the spatial search algorithm to cluster the wave systems in the calculation domain that have the same origin as the wave system at the starting point, so as to obtain the wave fields of wave systems with different origins.

2. A wave system spatial tracking method based on parameter continuity according to claim 1, characterized in that: The parameters of the wave system in S1 are the peak period T of each wave system. p , effective wave height H s , average wave direction θ m and the average directional distribution width σ m .

3. A wave system spatial tracking method based on parameter continuity according to claim 2, characterized in that: The specific method for detecting and merging pseudo wave systems and homologous wave systems in S2 is as follows: The average wave direction θ m The difference between the two is used to judge the proximity of the wave system in the spectral domain, and the average directional distribution width σ m To evaluate whether the development state of the wave system is similar; when the average wave direction θ of any two wave systems in the spectrum domain m and the average directional distribution width σ m The relationship of formula (1) is satisfied, indicating that the two wave systems are extremely close in the spectrum domain and are at a similar stage of development, that is, they correspond to the same origin, so they are merged into a new wave system; In the above formula, superscripts i and j represent wave system i and wave system j in the current spectrum domain, respectively. represents the average wave direction of wave system i, represents the average wave direction of wave system j, represents the average directional distribution width of wave system i, represents the average directional distribution width of wave system j, θ x , σ x Represent the set coefficients respectively.

4. The wave system spatial tracking method based on parameter continuity according to claim 3, characterized in that: The parameter calculation formula of the new wave system after merging in S2 is as follows: In the above formula, represents the peak period of wave system i, represents the significant wave height of wave system i, represents the average wave direction of wave system i, represents the average directional distribution width of wave system i; represents the peak period of wave system j, represents the significant wave height of wave system j, represents the average wave direction of wave system j, represents the average directional distribution width of wave system j; Represents the peak period of the new wave system after merging, represents the effective wave height of the new wave system after merging, Represents the average wave direction of the new wave system after merging. Represents the average directional distribution width of the new wave system after the merger.

5. A wave system spatial tracking method based on parameter continuity according to claim 4, characterized in that: The specific steps of S3 are: T1. Set the starting point of spatial tracking and mark all grid points as untracked; T2, put a wave system at the starting point into the search sequence and set it as the benchmark wave system, delete the information of the benchmark wave system in the original data set, and mark the starting point as tracked; T3, sequentially calculating the similarity between each wave system at all untracked adjacent grid points and the benchmark wave system; T4. Find the wave system in the neighboring grid points that is most similar to the benchmark wave system. If the similarity between the wave system and the benchmark wave system is less than the threshold S th , then the wave system is considered to be of the same origin as the benchmark wave system, the wave system is added to the search sequence, and the grid point where the wave system is located is marked as tracked; T5. If no homologous wave system can be found in the wave systems of adjacent grid points, the reference wave system is placed in a result sequence and deleted from the search sequence, and the next wave system in the search sequence is set as the reference wave system; T6, repeating steps T3-T5 until there is no wave system in the search sequence, indicating that the tracking of the wave system with the same origin as the wave system selected at the starting point in the calculation domain is completed, and the current result sequence is saved and output; T7, reset the tracking status of each grid point to untracked, put another wave system at the starting point into the search sequence, set the wave system as the reference wave system, delete the information of the reference wave system in the original data set, and mark the starting point as tracked; T8. Repeat steps T3-T7 until all wave systems at the starting point have been searched, completing the spatial tracking of all wave systems at the starting point.

6. A wave system spatial tracking method based on parameter continuity according to claim 5, characterized in that: The calculation formula of the similarity in T3 is: In the above formula, represents the significant wave height of the reference wave system, Represents the directional distribution width of the benchmark wave system, Represents the average wave direction of the benchmark wave system; represents the effective wave height of the wave system at the adjacent point, represents the directional distribution width of the wave system at the neighboring point, Represents the average wave direction of the wave system at the adjacent point; S represents the similarity, d represents the geographical distance between two grid points; H a , σ b ,θ c Represent weight factors respectively.

7. A wave system spatial tracking method based on parameter continuity according to claim 6, characterized in that: The adjacent grid points are 8 grid points connected to the grid point where the reference wave system is located, namely, the top, bottom, left, right, upper left, upper right, lower left, and lower right points of the grid point where the reference wave system is located.

Images