Array performance quantification analysis method and system for sea wave directional spectrum estimation

By using wavenumber spectral window calculation and multi-dimensional index evaluation, the quantitative problems of sensor array shape selection and performance evaluation are solved, and the accurate quantification and optimization of array performance are realized. This method is applicable to the extreme performance prediction and optimization of irregular arrays.

CN122384753APending Publication Date: 2026-07-14TIANJIN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610431358.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-02
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

In the existing technology, there is a lack of effective quantitative means for the selection of sensor array shape and its performance evaluation, which leads to problems such as estimation error and insufficient array observation capability. Traditional methods cannot accurately evaluate the extreme performance of irregular arrays.

Method used

A quantitative analysis method for array performance estimation of wave direction spectrum is proposed. The upper limit of resolution is determined by wavenumber spectrum window calculation, the equivalent number of measurement points is compared with the actual number of measurement points, the wavenumber sidelobe level and normalized average nearest neighbor distance are calculated, the main lobe concentration is evaluated, and a quantitative evaluation is carried out by combining five dimensions.

Benefits of technology

It enables accurate quantitative evaluation of sensor array performance, can identify the causes of estimated distortion of the array under different wave characteristics, is applicable to arrays with any number and shape of measuring points, and provides more accurate extreme performance prediction and optimization guidance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122384753A_ABST
    Figure CN122384753A_ABST
Patent Text Reader

Abstract

The application provides an array performance quantification analysis method and system for sea wave direction spectrum estimation, and relates to the technical field of array performance quantification analysis. The method comprises the following steps: determining the analytical upper limit of the array through wave number spectrum window calculation; determining the information utilization rate of the array by comparing the equivalent measurement point number and the real measurement point number of the array; obtaining the wave number sidelobe level by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wave number spectrum window, and then determining the energy leakage level of the array; determining the observation ability of the array by calculating the normalized average nearest neighbor distance between each measurement point of the array; obtaining the main lobe concentration degree by calculating the angle deviation between the direction of the vector and the direction of the center line of the correlation matrix of the array, and then determining the advantages and disadvantages of the main lobe shape of the array; and comprehensively evaluating the quantitative evaluation results of the above five dimensions to form the quantitative evaluation of the comprehensive performance of the array. The application is suitable for arrays with any measurement point number, and the performance pre-evaluation can be completed before the array is deployed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of array performance quantification analysis technology, and in particular to an array performance quantification analysis method and system for wave direction spectrum estimation. Background Technology

[0002] Wave direction spectrum estimation is a wave analysis technique that uses a planar sensor array deployed in the ocean to simultaneously collect wave signals (such as free surface elevation, near-bottom pressure, and bidirectional current velocity) at different spatial locations. Various estimation algorithms are then used to reveal the correlations between the signals within the array, thereby reconstructing the wave energy distribution. This energy distribution is ultimately manifested as a two-dimensional wave spectrum (frequency-direction, e.g., ...). Figure 1 (as shown) and three-dimensional wave spectrum (frequency-wave number-direction).

[0003] In practical sensor array installation, various constraints are often encountered, such as limited installation space, a limited number of available sensors, and the need to adjust the position of some measurement points due to terrain and other factors. This leads to significant differences in the planar distribution, number of sensors, and structural complexity of sensor arrays used in actual observations. Examples of array configurations currently in use or under research and analysis include... Figure 2 As shown. In existing research on ocean wave observation, scholars have mostly focused on optimizing the wave direction spectrum estimation algorithm itself, and are committed to developing estimation methods with better performance, such as iterative maximum likelihood estimation and Bayesian direction spectrum estimation. They hope to improve the estimation accuracy by enhancing the analytical capability of the algorithm and fully mining the wave energy direction distribution information hidden in the signals from various sensors.

[0004] However, the selection of sensor array morphology and its performance evaluation, which provide the raw data foundation, still mainly rely on subjective experience or calculation methods based on the simplest observation upper limit. This neglect of the array's own observation capabilities leads to many estimation errors stemming from array construction being wrongly attributed to insufficient algorithm performance. It also makes it difficult for researchers to identify fundamental estimation biases caused by insufficient array observation capabilities, resulting in an over-reliance on algorithm optimization while ignoring the quality of the data source. Regarding array morphology performance evaluation, existing methods mainly include two types: one is the "no redundancy" criterion proposed by Aida (2010), which states that the spatial vector formed by any two measurement points within the array should avoid repetition as much as possible; the other is the co-array criterion proposed by Davis and Young, which requires that the distribution of the co-array in different directions should be as extensive and uniform as possible. However, the "no redundancy" standard can only guarantee the independence of the measurement point information, that is, the utilization rate of the information, and does not mean that a high information utilization rate means better observation performance. Therefore, it is only a necessary but not sufficient condition for a high-performance observation array. The co-array standard essentially projects the array spatial measurement points to a single direction. Its evaluation results are only applicable to unidirectional wave situations and cannot effectively evaluate the observation performance of the array in real multidirectional wave environments. In addition, the judgment of "extension" and "uniform distribution" in this standard is still subjective qualitative and lacks rigorous quantitative basis.

[0005] In calculating the analytical upper limit of array configurations, existing methods still simply follow the Nyquist sampling theorem under uniform sampling conditions in time-frequency signal analysis. Within this theoretical framework, array observation is considered a spatial domain sampling, and the highest wavenumber observable by the array ( k max Limited by the minimum measurement point spacing (d) in the array x min ),Right now k max =π / d x While spatial observation and temporal signal analysis of an array are mathematically equivalent, real-world sensor arrays are typically spatially irregularly sampled, with uneven spacing and irregular distribution of measurement points. This leads to discrepancies between the predictions of traditional methods and the actual performance of the array. Traditional predictions are often too conservative, resulting in the rejection of a large amount of valid wave information during post-processing; conversely, overly aggressive performance predictions may lead to distorted estimates being misinterpreted as true wave characteristics. Therefore, traditional upper limit calculation methods suitable for regular sampling are insufficient to accurately predict the performance limits of irregularly sampled sensor arrays in the field. Summary of the Invention

[0006] To solve or alleviate one or more of the above-mentioned technical problems, the present invention provides a method and system for quantitative analysis of array performance of wave direction spectrum estimation.

[0007] According to one aspect of the present invention, a method for quantitative analysis of array performance in wave direction spectrum estimation is proposed, the method comprising: The upper limit of the array's resolution is determined by wavenumber spectral window calculation; The information utilization rate of the array is determined by comparing the equivalent number of measurement points with the actual number of measurement points. The wavenumber sidelobe level is obtained by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectral window, and then the energy leakage level of the array is determined. The observation capability of the array is determined by calculating the normalized average nearest neighbor distance between each measuring point in the array. By calculating the angular deviation between the direction of the vector at each measuring point in the array and the direction of the co-array centerline, the main lobe concentration is obtained, and the quality of the main lobe morphology of the array is determined. The overall performance of the array is quantitatively evaluated by combining the evaluation results of five dimensions: the resolution limit, the information utilization rate, the energy leakage level, the observation capability, and the quality of the main lobe morphology.

[0008] Furthermore, the determination of the upper limit of array resolution through wavenumber spectral window calculation includes: Project the array onto a certain direction to obtain a linear projection array in that direction; remove redundant measurement points from the linear projection array to obtain an independent simplified linear projection array; calculate the interval of all measurement points in the simplified linear projection array to obtain a set of measurement point intervals; determine the upper limit of resolution for the current direction based on the relationship between the defined linear array basis and the set of measurement point intervals; for all directions, after obtaining the upper limit of resolution for each direction according to the above process, determine the upper limit of resolution of the array as the lowest value of the upper limit of resolution for all directions.

[0009] Furthermore, determining the analytical upper limit for the current direction based on the defined relationship between the linear array basis and the measurement point interval set includes: The relationship between the linear array basis and the measurement point interval set satisfies: ; In the formula, Represents the set of measurement point intervals; represents the basis of a linear array, which satisfies two constraints: all values ​​are positive integers and none of the positive integers have a common common divisor; s represents the scaling factor; This indicates the fitting deviation or actual installation error. Based on the constraints of the linear array basis, we search for scaling factors and linear array basis such that the relationship between the linear array basis and the measurement point interval set holds. Based on the characteristic that the wavenumber spectral window response period corresponding to the linear array basis is always 2π, and combined with the scaling factor s, the upper limit of the array's resolution in the current direction is determined. .

[0010] Furthermore, determining the information utilization rate of the array by comparing the equivalent number of measurement points with the actual number of measurement points includes: if the equivalent number of measurement points of the array... Less than its actual number of measurement points If the array's information utilization is insufficient, then the number of equivalent measurement points is insufficient. The calculation formula is: ; In the formula, This represents the number of independent array measurement point pairs vectors.

[0011] Furthermore, the formula for calculating the wavenumber sidelobe level is as follows: ; In the formula, Indicates the wavenumber sidelobe level; max indicates taking the maximum value; This represents the sidelobe peak value in the wavenumber spectral window; This represents the peak value of the main lobe in the wavenumber spectral window. Represents all wavenumber vectors in wavenumber space. Represents the wavenumber vector to be analyzed; A preset energy leakage threshold is set, if the wavenumber sidelobe level If the energy leakage level exceeds the energy leakage threshold, the array is considered to have an excessively high energy leakage level; the energy leakage threshold is a negative value.

[0012] Furthermore, the formula for calculating the normalized average nearest neighbor distance is: ; In the formula, Indicates the average nearest neighbor distance. , R represents the actual number of measurement points; R is the radius of the circumcircle of the array, which is the radius of the smallest circle that can contain all measurement points. A preset distance threshold is set. If the normalized average nearest neighbor distance is less than the distance threshold, the array is determined to have high observation capability.

[0013] Furthermore, the formula for calculating the main lobe concentration is: ; In the formula, Indicates the concentration of the main lobe; For a vector of measurement point pairs in the measurement point interval set; θ mn The direction angle of the vector relative to the measuring point; θ p The angle of the co-array centerline; This indicates that the main lobe of the array has the most severe morphological distortion; This indicates that the main lobe shape of the array is optimal.

[0014] According to another aspect of the present invention, a system for quantitative analysis of array performance based on wave direction spectrum estimation is proposed. The system is implemented based on the aforementioned method for quantitative analysis of array performance based on wave direction spectrum estimation. The system includes: The upper limit of resolution calculation module is configured to determine the upper limit of resolution of the array through wavenumber spectral window calculation; The information utilization rate quantification module is configured to determine the information utilization rate of the array by comparing the equivalent number of measurement points of the array with the actual number of measurement points; The energy leakage level quantification module is configured to obtain the wavenumber sidelobe level by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectrum window, and then determine the energy leakage level of the array. The observation capability quantification module is configured to determine the observation capability of the array by calculating the normalized average nearest neighbor distance between each measuring point in the array. The main lobe morphology quality quantification module is configured to obtain the main lobe concentration by calculating the angle deviation between the direction of the vector of each measuring point in the array and the direction of the co-array centerline, thereby determining the quality of the main lobe morphology of the array. The comprehensive performance evaluation module is configured to integrate the evaluation results of five dimensions: the resolution limit, the information utilization rate, the energy leakage level, the observation capability, and the main lobe morphology, to form a quantitative evaluation of the array's overall performance.

[0015] The embodiments of the present invention have the following technical effects: This invention aims to address two key analytical gaps in the quantitative analysis of array performance in existing technologies: firstly, the lack of effective quantification methods for performance, and secondly, the difficulty in accurately defining the upper limit of performance. To address these gaps, this invention proposes a complete and systematic method for quantitative analysis of array morphology performance. This method includes an upper limit calculation method for irregular spatial sampling arrays, and four independent array performance indicators. The new upper limit calculation method can more accurately assess the array's ultimate observation capabilities; the four array performance indicators characterize four aspects that may affect the estimation results during observation, explaining the causes of estimation distortion under different wave characteristics. This invention is applicable to array morphologies with any number of measurement points. Each indicator has been normalized during the definition phase, ensuring comparability of performance between different array morphologies. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the frequency-direction wave spectrum; Figure 2 Examples of multiple array configurations ranging from four to eight measurement points; Figure 3 This is a flowchart of an array performance quantization analysis method for wave direction spectrum estimation as described in this invention; Figure 4 This is a schematic diagram of the array projected in a single direction. The red dots in the diagram represent the linear array after projection. Figure 5 Examples of simplified window calculation results for regular arrays (corresponding to the left figure) and irregular arrays (corresponding to the right figure); Figure 6 This is a schematic diagram illustrating the calculation process for the general array analytical upper limit based on a linear array basis. Figure 7 This is a diagram illustrating the prediction results of the general array analytical upper bound calculation based on the linear array basis; Figure 8 Examples of the calculation results of equivalent number of measurement points and perfect ratio for a six-point array (corresponding to Figure (a)) and a seven-point array (corresponding to Figure (b)); Figure 9 Example of the calculation effect of simplified wavenumber spectrum window (corresponding to Figure (a)) and wavenumber sidelobe level (corresponding to Figure (b)) for a six-point array; Figure 10 Examples of multi-directional wave observation effects using the array's average nearest neighbor distance index; Figures (a) to (d) represent the most unfavorable observations to the best observations, respectively. Figure 11 A schematic diagram showing the relationship between the main lobe morphology and the measurement point pair vector; Figure 12 This is a schematic diagram illustrating the effect of the main lobe morphology description on the main lobe concentration index. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are part of this invention.

[0019] The present invention aims to provide a systematic evaluation method for quantifying the observation performance of sensor array morphology in wave observation and directional spectrum estimation.

[0020] This invention proposes a method for quantitative analysis of array performance in wave direction spectrum estimation, such as... Figure 3 As shown, the method includes: S1. Determine the upper limit of the array's resolution by calculating using the wavenumber spectral window; S2. Determine the information utilization rate of the array by comparing the equivalent number of measurement points with the actual number of measurement points; S3. By calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectral window, the wavenumber sidelobe level is obtained, and the energy leakage level of the array is determined. S4. Determine the observation capability of the array by calculating the normalized average nearest neighbor distance between each measuring point in the array; S5. By calculating the angular deviation between the direction of the vector at each measuring point in the array and the direction of the co-array centerline, the main lobe concentration is obtained, and the quality of the main lobe morphology of the array is determined. S6. Based on the evaluation results of the five dimensions of the resolution upper limit, information utilization rate, energy leakage level, observation capability, and main lobe morphology, a quantitative assessment of the overall performance of the array is formed.

[0021] First, in S1, the upper limit of the array's resolution is determined by wavenumber spectral window calculation.

[0022] According to an embodiment of the present invention, for any array configuration (including regular and irregular arrays), its limiting performance should be the highest number of resolvable waves without aliasing in a unidirectional wave. In the wave spectrum estimation stage of a unidirectional wave, each measuring point at different locations can be projected onto the main wave direction without considering information distortion caused by projection. Therefore, at this time, all measuring point information of the array is used for solving in a single direction. The highest wave value that can be solved while ensuring estimation quality is the limiting performance of the array. A schematic diagram of the projection of a spatial array in a certain direction is shown below. Figure 4 As shown.

[0023] Traditional methods only consider the minimum interval and ignore the actual spatial distribution and measurement points of the array, thus failing to accurately reflect the array's true response to wavenumber analysis. To fully consider the common response of the projected linear array to the wavenumber to be analyzed, a simplified wavenumber window is used to calculate the response of a linear array. The simplified linear wavenumber window is calculated as follows.

[0024] (1) In the formula, k is the wavenumber to be analyzed, in rad / m; m and n are two measurement points of a linear array along a certain projection direction; the vector between the two measurement points is the measurement point pair vector. , For measuring points m The relative spatial position.

[0025] Despite its numerous simplifications, this simplified linear wavenumber window retains the most basic wavenumber aliasing identification capability. At the same time, it reflects the common response of all measurement points and is more suitable for real-world spatial arrangements of irregular arrays compared to traditional methods.

[0026] The wavenumber response and aliasing prediction performance of this simplified linear wavenumber window are demonstrated using the following two examples: 1) Uniform linear array: total length 15 units, with one spatial sampling per unit, for a total of 16 spatial measurement points. 2) Irregular linear array: total length 15 units, with sampling only performed at the spatial locations corresponding to [0, 2, 5, 9, 15], for a total of 5 spatial measurement points.

[0027] According to traditional calculation methods: The highest wavenumber observable by the array Limited by the minimum measurement point spacing in the array The minimum interval of a uniform linear array is 1 unit, therefore its wavenumber resolution upper limit should be 1 / 2. In the spectral window, it appears as The response period; however, for irregular linear arrays, since the minimum interval is 2, the traditional prediction is an upper limit. Spectral window representation The response period. The spectral windowing of the two arrays is performed using equation (1) as follows. Figure 5 As shown.

[0028] Depend on Figure 5 It can be seen that regular linear arrays, due to their sampling method being consistent with the theoretical framework of traditional methods, exhibit good prediction results, and the response spectrum window demonstrates... The response cycle confirms that the upper limit of parsing at this time is... However, although the spectral window of the irregular array exhibits slight distortion due to its low sampling rate, its response period is consistent with that of the regular array spectral window, confirming that the true upper limit of resolution of this array is also [missing information]. This contradicts the predictions made by traditional methods. Therefore, it is clear that traditional methods, which ignore the spatial structure of the array and rely solely on the minimum interval calculation model, are unreliable in predicting the ultimate performance of irregular arrays. A more comprehensive upper limit calculation method applicable to irregular sampling is urgently needed.

[0029] Therefore, this invention proposes the following array upper limit calculation method with the aid of simplified spectral window sensing and calculation, which fully considers the effect of irregular spatial arrangement of actual measurement points in the array.

[0030] Let the set of all measurement point intervals of the real linear array be . Introducing assumptions, It can be viewed as a set of points with an equal number of measurement points that satisfy a certain standard. The scaling result, the scaling factor is denoted as Meanwhile, taking into account minute practical errors, and The transformation relationship can be expressed as shown in equation (2).

[0031] (2) In the formula, Indicates fitting deviation or actual installation error; set Defined as a linear array basis (hereinafter abbreviated as linear basis), a linear array basis satisfies two core constraints: first, all values ​​in the set are positive integers; second, the positive integers have no common common divisor. Example: [2, 4, 5] is a linear array basis, while [2, 4, 6] does not satisfy the definition of a linear array basis because it has a common common divisor of 2.

[0032] Due to its two strong constraints, the wavenumber spectral window response period of a set satisfying the definition of a linear basis will always be zero. That is, the upper limit of the maximum wavenumber analysis is always . Therefore, the set The approximate analytic upper bound can be obtained by scaling the upper bound of the linear basis. The maximum analytic upper bound satisfying this scaling assumption is... It can be calculated from equation (3).

[0033] (3) Figure 6 The complete calculation process of this upper limit calculation method is demonstrated. Specifically, it includes: projecting the array onto a certain direction to obtain a linear projection array in that direction; removing redundant measurement points from the linear projection array to obtain an independent simplified linear projection array; calculating the intervals of all measurement points in the simplified linear projection array to obtain a set of measurement point intervals; determining the analytical upper limit for the current direction based on the defined relationship between the linear array basis and the set of measurement point intervals, including finding a scaling factor and a linear array basis based on the constraints of the defined linear array basis, such that the relationship between the linear array basis and the set of measurement point intervals holds; and determining the analytical upper limit of the array in the current direction based on the characteristic that the wavenumber spectral window response period corresponding to the linear array basis is always 2π, combined with the scaling factor s. For all directions, after obtaining the upper limit of resolution for each direction according to the above process, the upper limit of resolution of the array is determined to be the lowest value of the upper limit of resolution for all directions.

[0034] Using a classic six-point regular pentagonal array as a calculation example, the basic calculation process is as follows: 1) Project the array to be analyzed into any direction to obtain a linear projection array in that direction, i.e. For example, projecting a regular pentagonal array onto direction θ0 yields [x n The value is [0, 7.27, 19.02, 19.02, 30.78, 38.04], where 19.02 appears twice, indicating that the projected positions of two measuring points in this direction coincide. Figure 6 As shown in (a).

[0035] 2) Remove linear projection array By eliminating redundant measurement points with extremely close spatial distances, an independent simplified linear projection array is obtained. Specifically, remove projections with a spacing less than 3% of the total span of the array in that direction. In this embodiment, the threshold for projection spacing can be determined as 3% × 38.04 = 1.14m.

[0036] 3) Calculate the simplified linear projection array All measurement point intervals are used to obtain the measurement point interval set. Taking the above projection as an example, = [7.26, 7.27, 11.75, 11.76, 19.02, 19.02, 23.51, 30.77, 30.78, 38.04]), find a certain scaling factor. and its corresponding linear array basis This invention is implemented using second-order spatial difference and iterative refinement, but other numerical methods can also be used to solve it. For example, s = 2.376 can be determined. = [3, 3, 5, 5, 8, 8, 10, 13, 13, 16], where All elements are positive integers and their common divisor is 1. In this case, s× It can be used as an approximation Equation (2) holds true, such as Figure 6 As shown in (b).

[0037] 4) Based on the linear array basis Spectral window features and scaling factor Inversely calculate the approximate response period of the real array. Therefore, the upper limit of array resolution in the current direction is determined to be... ,like Figure 6 As shown in (c).

[0038] 5) Repeat steps 1) to 4) to calculate the array upper limit results in all directions, obtaining the complete distribution curve of the array upper limit along the direction. At this point, the overall performance limit of the array is defined as the lowest upper limit in each direction, such as... Figure 6 As shown in (d).

[0039] Figure 7 The results of the array's analytical performance test are shown, with the horizontal axis representing the operating range. It is evident that the novel upper limit calculation method (red line) in this invention successfully calculates the array's performance in unidirectional waves (directional diffusion). The array limit at its minimum (the smallest) has a more accurate prediction effect compared to the traditional method (black line).

[0040] However, Figure 7 This also shows that the reliable resolution upper limit of the array (orange line) is lower than the calculated results of the two upper limits mentioned above, which means that the array performance is poor and there are other mechanisms causing performance degradation. However, there is currently no reliable quantitative method in the industry to control or quantify this array performance degradation. This invention proposes four subsequent important array indicators to control array performance and ensure that its true performance (orange line) can approach its limit performance (red line) as closely as possible.

[0041] The four subsequent important array indicators focus on exploring whether the current array arrangement makes full use of the information from all its measuring points. In short, whether the information from each measuring point is independent and whether the array has reached its maximum amount of information collected.

[0042] Subsequently, in S2, the information utilization rate of the array is determined by comparing the equivalent number of measurement points of the array with the actual number of measurement points.

[0043] According to an embodiment of the present invention, there is already a quantitative indicator for calculating the information utilization rate of an array, called the array's perfect ratio. The calculation method of this indicator is as follows: (4) In the formula, The number of independent array measurement point pairs vectors. This represents the actual number of measurement points. Although this indicator is a general dimensionless indicator and its mathematical concept is clear and easy to understand, it can only reflect whether the array itself has reached its own performance limit, and cannot be used for comparison between different array configurations.

[0044] To balance the need for information utilization assessment with consideration for different array schemes, this invention proposes a first important array indicator—the equivalent number of measurement points—to replace the original method for calculating the array perfect ratio. Equivalent number of measurement points It can be calculated using the following formula: (5) Unlike the perfect ratio of an array, the physical meaning of the equivalent number of measurement points is a... An array of measuring points, the amount of spatial sampling information of its measuring points is equivalent to An array of measurement points. Therefore, the method for determining the information utilization rate of the equivalent number of measurement points is: if the equivalent number of measurement points of an array... Less than its actual number of measurement points If so, the information utilization rate of the array is insufficient.

[0045] at the same time, Its advantage lies in allowing comparison between different arrays, which can be achieved by... Figure 8 This can be seen from the text. Figure 8 In Figure (a), the perfect ratio of the six-point array is 1, meaning the array utilization reaches its theoretical maximum. However, the perfect ratio of the seven-point array in Figure (b) is only 0.71, indicating duplicate data collection at the measurement points. Based solely on the perfect ratio, the six-point array seems to be the better option. However, the equivalent number of measurement points shows that both arrays have the equivalent information content of six measurement points, meaning their information content is completely identical. Furthermore, the seven-point array still reflects that it did not fully utilize its acquisition performance, resulting in wasted resources due to duplicate data collection. At the same time, the equivalent number of measurement points also confirms that the number of measurement points in both array schemes is equivalent. Therefore, the practical application of the equivalent number of measurement points is more convenient than the array perfect ratio.

[0046] Subsequently, in S3, the wavenumber sidelobe level is obtained by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectral window, thereby determining the energy leakage level of the array.

[0047] According to embodiments of the present invention, a second important array metric—wavenumber sidelobe level—is proposed. This metric is used to perform basic quantization of the array's analytical response in the wavenumber domain using a simplified two-dimensional wavenumber spectral window, and is used to quantify the energy leakage level of the array when resolving a specific wavenumber. The expression for the wavenumber spectral window is shown below: (6) In the formula, For all wavenumber vectors in wavenumber space, Here, the wavenumber vector to be analyzed is... and Both are two-dimensional wavenumber vectors, containing information on wavenumber magnitude and direction, unlike the scalar k in equation (1) which only represents the wavenumber magnitude; Intervals between each measuring point The spectral window response weighting coefficients. Based on this, the simplified two-dimensional wavenumber spectral window is defined as follows: (7) Similar to the side-lobe level (SLL) in classical time-frequency analysis, a similar wavenumber side-lobe level can be defined for this two-dimensional simplified wavenumber spectral window as shown in the following equation: (8) In this formula, the numerator represents the maximum sidelobe peak in the wavenumber window, and the denominator represents the main lobe peak. The subscript sidelobe indicates the response peak in the region outside the main lobe of the wavenumber window, the subscript mainlobe indicates the response peak at the main lobe, and max(·) indicates taking the maximum value.

[0048] The wavenumber sidelobe level index is negative; a larger absolute value indicates weaker sidelobes, less energy leakage, and higher resolution. Because the directional distribution of wave energy is continuous and differs from the high-precision resolution requirements of signal analysis, in wave directional spectrum estimation, the array... (The smaller this value, the weaker the sidelobes and the more ideal the resolution.) The requirements can be appropriately lowered. Therefore, a preset energy leakage threshold is set. If the wavenumber sidelobes are low... If the energy leakage exceeds the energy leakage threshold, the array is considered to have an excessively high energy leakage level; the energy leakage threshold is a negative value. After testing, the energy leakage threshold is equal to... It can effectively suppress energy leakage.

[0049] It is worth noting that the upper limit of equation (8) adopts the new analytical upper limit in S1 of this invention. . Continue The unit is decibel (dB), with a theoretical range of (-20, 0], corresponding to the ideal state where the side lobes are much smaller than the main lobe, and the worst state where the side lobes are at the same level as the main lobe.

[0050] Two example arrays are selected and their wavenumber spectral windows are plotted according to Equation (7). The corresponding values ​​are calculated using Equation (8). Numerical values, results as follows Figure 9 As shown. By Figure 9 As can be seen, the classical array in Figure (a) actually has a very obvious sidelobe level, which means that the array is accompanied by continuous spectral window energy leakage during the solution process, and its The value was only -2.76 dB, confirming severe leakage in this array; conversely, the array in Figure (b) had a significantly lower sidelobe level and weaker leakage. The value of -6.63dB confirms that the array has better resolution.

[0051] Subsequently, in S4, the observation capability of the array is determined by calculating the normalized average nearest neighbor distance between each measurement point in the array.

[0052] According to embodiments of the present invention, a third important array metric—mean nearest neighbor distance—is proposed. This metric is used to evaluate the array's ability to observe waves whose phase is spatially unstable or easily distorted. The metric is defined as follows: (9) In the formula, This represents the actual number of measurement points; min indicates taking the minimum value. This describes the average distance between each measuring point and its nearest neighbor, excluding itself, in a given array. Physically, it describes the density of the array's measuring points in space. The normalized calculation of this index is shown in Equation (10), obtained by calculating its own... The numerical value is achieved by comparing it with the ratio of a circular array of equal-scale measurement points.

[0053] (10) In the formula, R represents the normalized average nearest neighbor distance; R is the radius of the array's circumcircle, which is the radius of the smallest circle that can contain all measurement points and is used to characterize the spatial scale of the array; the denominator is the theoretical average nearest neighbor distance of a circular array with the same number of measurement points and the same scale, which serves as the normalization benchmark.

[0054] The mean nearest neighbor distance index has a good and stable quantitative effect on the array's ability to observe complex waves. The smaller the index value, the more significant the array's ability to capture wave details. Therefore, a distance threshold is preset. If the normalized mean nearest neighbor distance is less than this threshold, the array is considered to have high observation capability. This distance threshold can be obtained through experiments or experience.

[0055] Figure 10 Examples of multi-directional wave observation effects using the array's average nearest neighbor distance index; where Figures (a) to (d) represent the worst-case observation to the best-case observation, respectively.

[0056] Subsequently, in S5, the main lobe concentration is obtained by calculating the angular deviation between the direction of the vector at each measuring point in the array and the direction of the co-array centerline, thereby determining the quality of the main lobe morphology of the array.

[0057] According to embodiments of the present invention, a fourth important array indicator—main lobe concentration—is proposed. This indicator describes the array's ability to distinguish multi-directional wave crest regions. Different arrays produce main lobes with varying morphology and concentration, and the morphology of these main lobes determines the objectivity of the array's response to the true wave number distribution of multi-directional waves. Arrays with distorted main lobes will misinterpret and misjudge the true wave number characteristics of observed waves, thus interfering with subsequent analysis results. Figure 11 This is a schematic diagram showing the relationship between the main lobe morphology and the measurement point pair vector. Figure 11In the figure, multiple similar figures (b), (c), and (d) are superimposed to form figure (a).

[0058] Main lobe concentration The definition is shown in Equation (11). The quality of the main lobe morphology is quantified by calculating the deviation of the vector at each measuring point of the array from the center line of its co-array. That is, the main lobe concentration is obtained by calculating the angular deviation between the direction of the vector at each measuring point of the array and the direction of the center line of the co-array. (11) In the formula, θ mn = arg(x mn () represents the direction angle of the vector from the measuring point. θ p The angle of the co-array centerline of the array.

[0059] Equation (11) shows that the specific value of the main lobe concentration index of the array is related to the spatial orientation of the current array. Therefore, its final value is defined as the angle that causes the most adverse effect among all array orientations. θ p The corresponding indicator value. This is the most unfavorable angle. θ p The analytical solution is shown in equation (12).

[0060] (12) Since the definition of equation (11) is essentially the ratio of the square of the Y-axis extension component of the coarray to its total extension, the main lobe concentration is a non-negative dimensionless index. The most unfavorable spatial extension corresponds to the U-shaped main lobe morphology with the most severe distortion in the array. This indicates the optimal main lobe morphology, closest to the optimal main lobe of the impulse function. In this case, the array can most objectively reflect the wave number pattern. The main lobe concentration index's descriptive effect on the main lobe morphology is as follows: Figure 12 As shown. Figure 12 Figures (a), (b), (c), and (d) correspond to respectively =0.03、 =0.34、 =0.67、 =1.

[0061] Subsequently, in S6, the evaluation results of the five dimensions—the upper limit of resolution, the information utilization rate, the energy leakage level, the observation capability, and the quality of the main lobe morphology—are combined to form a quantitative assessment of the overall performance of the array.

[0062] This invention also proposes a system for quantitative analysis of array performance based on wave direction spectrum estimation. The system is implemented based on the wave direction spectrum estimation method for quantitative analysis of array performance described in the above embodiments. The system includes: The upper limit of resolution calculation module is configured to determine the upper limit of resolution of the array through wavenumber spectral window calculation; The information utilization rate quantification module is configured to determine the information utilization rate of the array by comparing the equivalent number of measurement points of the array with the actual number of measurement points; The energy leakage level quantification module is configured to obtain the wavenumber sidelobe level by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectrum window, and then determine the energy leakage level of the array. The observation capability quantification module is configured to determine the observation capability of the array by calculating the normalized average nearest neighbor distance between each measuring point in the array. The main lobe morphology quality quantification module is configured to obtain the main lobe concentration by calculating the angle deviation between the direction of the vector of each measuring point in the array and the direction of the co-array centerline, thereby determining the quality of the main lobe morphology of the array. The comprehensive performance evaluation module is configured to integrate the evaluation results of five dimensions: the resolution limit, the information utilization rate, the energy leakage level, the observation capability, and the main lobe morphology, to form a quantitative evaluation of the array's overall performance.

[0063] For the undescribed parts of the array performance quantification analysis system for wave direction spectrum estimation according to an embodiment of the present invention, please refer to the detailed description of the method embodiment above.

[0064] This invention abstracts and unifies the practical engineering problem of wave sensor array optimization into a mathematical and signal analysis problem of "optimizing the shape and response characteristics of spatial signal sampling points." It establishes a quantitative relationship between the geometric shape of the sensor array and the wavenumber domain response quality as the theoretical basis for array optimization. Drawing on the idea of ​​the influence of spectral window functions on the quality of observed signals in signal analysis, it extends this concept to array systems, proposing the concept of an array wavenumber window to characterize the array's observation characteristics in the wavenumber domain. Its evaluation indicators mainly include the concentration of the main lobe and the level of the side lobes. Simultaneously, considering the physical characteristics of accelerated dissipation of wave phase with propagation direction under complex multidirectional wave conditions, an average nearest neighbor distance index is introduced to enhance the array's ability to resolve and robust to spatial phase changes. Furthermore, considering the limited and costly deployment of measuring points in actual marine field construction, this invention introduces an equivalent number of measuring points index to quantify the differences in information utilization efficiency under different array configurations. This invention achieves systematic and quantifiable optimization of array configuration and its observation performance by functionally analogizing the spatial sampling behavior of the array with the window function design concept in classical signal analysis.

[0065] Existing array design methods lack a common theoretical foundation, with each criterion existing in isolation. This invention unifies array performance issues within the framework of spatial sampling and wavenumber domain spectral window analysis, making the relationship between array geometry and estimation error interpretable and calculable. It also clarifies the dominant sources of error under different conditions: wavenumber aliasing is the primary cause of narrow-direction diffusion, while insufficient cross-spectral sampling is the primary cause of wide-direction diffusion, thus providing a theoretical basis for targeted optimization.

[0066] Traditional Nyquist methods rely solely on the minimum measurement point spacing, resulting in systematic biases for irregular arrays. This can lead to either overestimation (causing artifacts to be misidentified as genuine wave features) or underestimation (discarding valuable information). This invention, compared to traditional methods, better meets the analytical needs of irregular arrays. In the analytical upper bound calculation, the linear array basis comprehensively utilizes the spatial arrangement information of all measurement points, providing a more accurate prediction of the limiting performance of irregular arrays. Furthermore, this invention supports horizontal comparisons. Existing methods (without redundancy standards and co-array standards) are all qualitative descriptions, unable to quantify or compare different arrays. The four indicators proposed in this invention, each targeting an independent dimension of performance, are all normalized and can be directly used for quantitative comparison and optimization ranking between arrays with different numbers of measurement points and different geometric shapes.

[0067] In terms of construction, this invention only requires the coordinates of the measurement points as input, resulting in minimal computational load and allowing for performance pre-evaluation before array deployment. The improved array designed based on this invention has an effective working range of up to 3.45, a significant improvement compared to existing classic arrays (generally below 2.0).

[0068] It should be noted that in the calculation of the wavenumber spectral window, the embodiments of the present invention adopt a simplified form that only retains the real part. Using the complete complex form or introducing amplitude weighting are different approximations of the same physical quantity, and the array performance they describe is essentially the same. Regarding the selection of normalization benchmarks for various indicators, the embodiments of the present invention use the theoretically optimal layout with the same number of measurement points and scale as the benchmark. Using other reasonable benchmarks does not change the physical meaning of the indicators, but only affects the reference point of the values. Furthermore, the present invention focuses on two-dimensional planar arrays. If the projection direction is extended from the planar angle to the three-dimensional spherical solid angle, the same calculation framework can be directly applied to the performance analysis of three-dimensional volumetric arrays.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for quantitative analysis of array performance in wave direction spectrum estimation, characterized in that, include: The upper limit of the array's resolution is determined by wavenumber spectral window calculation; The information utilization rate of the array is determined by comparing the equivalent number of measurement points with the actual number of measurement points. The wavenumber sidelobe level is obtained by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectral window, and then the energy leakage level of the array is determined. The observation capability of the array is determined by calculating the normalized average nearest neighbor distance between each measuring point in the array. By calculating the angular deviation between the direction of the vector at each measuring point in the array and the direction of the co-array centerline, the main lobe concentration is obtained, and the quality of the main lobe morphology of the array is determined. The overall performance of the array is quantitatively evaluated by combining the evaluation results of five dimensions: the resolution limit, the information utilization rate, the energy leakage level, the observation capability, and the quality of the main lobe morphology.

2. The array performance quantification analysis method for wave direction spectrum estimation according to claim 1, characterized in that, The determination of the upper limit of array resolution through wavenumber spectral window calculation includes: Project the array onto a certain direction to obtain a linear projection array in that direction; remove redundant measurement points from the linear projection array to obtain an independent simplified linear projection array; calculate the interval of all measurement points in the simplified linear projection array to obtain a set of measurement point intervals; determine the upper limit of resolution for the current direction based on the relationship between the defined linear array basis and the set of measurement point intervals; for all directions, after obtaining the upper limit of resolution for each direction according to the above process, determine the upper limit of resolution of the array as the lowest value of the upper limit of resolution for all directions.

3. The array performance quantification analysis method for wave direction spectrum estimation according to claim 2, characterized in that, The determination of the analytical upper limit for the current direction based on the defined relationship between the linear array basis and the measurement point interval set includes: The relationship between the linear array basis and the measurement point interval set satisfies: ; In the formula, Represents the set of measurement point intervals; represents the basis of a linear array, which satisfies two constraints: all values ​​are positive integers and none of the positive integers have a common common divisor; s represents the scaling factor; This indicates the fitting deviation or actual installation error. Based on the constraints of the linear array basis, we search for scaling factors and linear array basis such that the relationship between the linear array basis and the measurement point interval set holds. Based on the characteristic that the wavenumber spectral window response period corresponding to the linear array basis is always 2π, and combined with the scaling factor s, the upper limit of the array's resolution in the current direction is determined. .

4. The array performance quantification analysis method for wave direction spectrum estimation according to claim 1, characterized in that, The process of determining the information utilization rate of the array by comparing the equivalent number of measurement points with the actual number of measurement points includes: if the equivalent number of measurement points of the array... Less than its actual number of measurement points If the array's information utilization is insufficient, then the number of equivalent measurement points is insufficient. The calculation formula is: ; In the formula, This represents the number of independent array measurement point pairs vectors.

5. The array performance quantification analysis method for wave direction spectrum estimation according to claim 1, characterized in that, The formula for calculating the wavenumber sidelobe level is: ; In the formula, Indicates the wavenumber sidelobe level; max indicates taking the maximum value; This represents the sidelobe peak value in the wavenumber spectral window; This represents the peak value of the main lobe in the wavenumber spectral window. Represents all wavenumber vectors in wavenumber space. Represents the wavenumber vector to be analyzed; A preset energy leakage threshold is set, if the wavenumber sidelobe level If the energy leakage exceeds the energy leakage threshold, the energy leakage level of the array is determined to be too high. The energy leakage threshold is a negative value.

6. The array performance quantification analysis method for wave direction spectrum estimation according to claim 1, characterized in that, The formula for calculating the normalized average nearest neighbor distance is as follows: ; In the formula, Indicates the average nearest neighbor distance. , R represents the actual number of measurement points; R is the radius of the circumcircle of the array, which is the radius of the smallest circle that can contain all measurement points. A preset distance threshold is set. If the normalized average nearest neighbor distance is less than the distance threshold, the array is determined to have high observation capability.

7. The method for quantitative analysis of array performance for wave direction spectrum estimation according to claim 1, characterized in that, The formula for calculating the main lobe concentration is: ; In the formula, Indicates the concentration of the main lobe; For a vector of measurement point pairs in the measurement point interval set; θ mn The direction angle of the vector relative to the measuring point; θ p The angle of the co-array centerline; This indicates that the main lobe of the array has the most severe morphological distortion; This indicates that the main lobe shape of the array is optimal.

8. A system for quantitative analysis of array performance in wave direction spectrum estimation, characterized in that, The system is implemented based on the array performance quantization analysis method for wave direction spectrum estimation according to any one of claims 1-7; the system includes: The upper limit of resolution calculation module is configured to determine the upper limit of resolution of the array through wavenumber spectral window calculation; The information utilization rate quantification module is configured to determine the information utilization rate of the array by comparing the equivalent number of measurement points of the array with the actual number of measurement points; The energy leakage level quantification module is configured to obtain the wavenumber sidelobe level by calculating the ratio of the maximum sidelobe peak value to the main lobe peak value in the wavenumber spectrum window, and then determine the energy leakage level of the array. The observation capability quantification module is configured to determine the observation capability of the array by calculating the normalized average nearest neighbor distance between each measuring point in the array. The main lobe morphology quality quantification module is configured to obtain the main lobe concentration by calculating the angle deviation between the direction of the vector of each measuring point in the array and the direction of the co-array centerline, thereby determining the quality of the main lobe morphology of the array. The comprehensive performance evaluation module is configured to integrate the evaluation results of five dimensions: the resolution limit, the information utilization rate, the energy leakage level, the observation capability, and the main lobe morphology, to form a quantitative evaluation of the array's overall performance.