A filter performance automatic evaluation method and system

Abstract

The application is suitable for the technical field of filter data optimization, and provides a filter performance automatic evaluation method and system.The method comprises the following steps: generating a time delay shaping strategy according to the deviation direction and deviation amount between the current equivalent multipath delay spread parameter and the historical optimal interval; and performing time delay shaping processing on the echo signal of an input target filter based on the time delay shaping strategy, so that the equivalent multipath delay spread of the shaped echo signal falls into the historical optimal interval when entering the target filter, so that the echo peak value resolution is in an optimal state.The application does not depend on changing the structure of the filter itself, but realizes performance optimization by regulating the input signal characteristics, can further tap the processing potential of the existing filter, and improves the stability of the filtering effect and the accuracy of subsequent target identification.

Description

Technical Field

[0001] This invention belongs to the field of filter data optimization technology, and in particular relates to an automatic filter performance evaluation method and system. Background Technology

[0002] In underwater sonar systems, filters, as crucial units in echo signal preprocessing, are primarily used to suppress environmental noise, enhance effective signals, and improve signal quality, thereby providing fundamental data support for subsequent target detection, feature extraction, and recognition. Existing technologies typically employ bandpass filtering, adaptive filtering, and multi-scale filtering to address noise interference from the complex underwater propagation environment, aiming to improve the signal-to-noise ratio and ensure overall signal stability. Furthermore, to address multipath effects caused by sea surface reflection, seabed reflection, and changes in the water medium, some technologies optimize filter parameters or introduce dynamic adjustment mechanisms to reduce the impact of multipath interference on the filtering results, ensuring that the echo signal meets pre-defined processing requirements at the overall level.

[0003] However, in the aforementioned existing technologies, the evaluation and optimization of filter performance typically focus on overall metrics, such as signal-to-noise ratio, energy retention, or bandwidth characteristics, while paying less attention to the finer-grained structural features of the filtered output signal, especially lacking a systematic analysis of echo peak resolution. In fact, echo peak resolution, as a crucial factor affecting target separation and recognition accuracy, directly relates to the performance of subsequent processing stages. However, because it is a result metric after filtering, existing technologies generally do not establish a correlation between it and the multipath structural characteristics of the signal before or during filtering, resulting in this metric not being effectively utilized in filter design and evaluation.

[0004] Furthermore, in practical applications, even when multipath interference is controlled within acceptable limits using existing filtering techniques, the equivalent multipath delay spread parameter of the echo signal will still fluctuate within a certain range, causing corresponding changes in echo peak resolution. Existing technologies typically treat these fluctuations as random disturbances and ignore them, failing to identify potential patterns of change or establish a correspondence between the equivalent multipath delay spread parameter and echo peak resolution. Therefore, there is a lack of technical means to model and utilize these variations. Thus, how to uncover and utilize the inherent correlation between these parameters without altering the filter structure to further improve echo peak resolution and optimize the overall filter performance has become a pressing technical problem in this field. Summary of the Invention

[0005] The purpose of this invention is to provide an automatic filter performance evaluation method and system, which aims to solve the problems mentioned in the background art.

[0006] This invention is implemented as follows: an automatic filter performance evaluation method, the method comprising:

[0007] S1. Obtain the historical sample dataset, which includes multiple historical echo samples and their filtering results that match the working background of the current target filter.

[0008] S2. Extract the equivalent multipath delay spread parameters of each historical echo sample, and extract the echo peak resolution based on the filtering results;

[0009] S3. Based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, establish a variation relationship model, and determine the historical optimal interval corresponding to the optimal echo peak resolution according to the variation relationship model.

[0010] S4. Obtain the current echo signal, extract the equivalent multipath delay spread parameter corresponding to the current echo signal, and determine whether it falls within the historical optimal range.

[0011] S5. If the signal does not fall within the historical optimal range, a time delay shaping strategy is generated based on the deviation direction and amount between the current equivalent multipath delay spread parameter and the historical optimal range. The time delay shaping strategy is then used to perform time delay shaping on the echo signal input to the target filter so that the equivalent multipath delay spread of the shaped echo signal falls within the historical optimal range when it enters the target filter, thereby ensuring that the echo peak resolution is in a better state.

[0012] As a further limitation of the technical solution of the present invention, the matching with the working background of the current target filter specifically refers to the following: the historical echo samples and the current echo signal meet the preset similarity conditions in terms of sonar working parameters and environmental parameters. The sonar working parameters include the type of transmitted signal, the working frequency band and the transmitted power, and the environmental parameters include water depth, seabed reflection characteristics and noise level.

[0013] As a further limitation of the technical solution of this embodiment of the invention, the specific calculation process of the equivalent multipath delay spread parameter includes:

[0014] The echo signal is decomposed by time delay to obtain the time delay parameters and corresponding energy parameters of several multipath components;

[0015] The average delay is obtained by weighting the delay parameters of each multipath component based on the energy parameters of each multipath component.

[0016] The energy-weighted root mean square delay spread is calculated based on the deviation between the delay parameters of each multipath component and the average delay.

[0017] The energy-weighted root mean square delay spread is determined as the equivalent multipath delay spread parameter.

[0018] As a further limitation of the technical solution of this embodiment of the invention, the process of extracting the echo peak resolution specifically includes:

[0019] Peak detection is performed on the echo signal after filtering to determine the peak position and peak amplitude of different echo peaks in the echo signal;

[0020] For adjacent echo peaks, calculate the difference between their peak positions as the time interval, and calculate the ratio of their peak amplitudes;

[0021] The corresponding echo peak resolution is obtained by weighting and combining the ratio of the time interval to the peak amplitude.

[0022] As a further limitation of the technical solution of this embodiment of the invention, S3 specifically includes:

[0023] S31. Sort several historical echo samples in ascending order according to the equivalent multipath delay spread parameter to obtain a sample sequence;

[0024] S32. Obtain the echo peak resolution corresponding to the sample sequence, and perform curve fitting based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution to obtain the change relationship model.

[0025] S33. Analyze the change relationship model to determine whether it exhibits a non-monotonic change trend of first rising and then falling.

[0026] S34. When the change relationship model shows a non-monotonic change trend of first rising and then falling, determine the peak position of the change relationship model, and extract the parameter interval within a preset range as the historical optimal interval, centered on the equivalent multipath delay expansion parameter corresponding to the peak position.

[0027] As a further limitation of the technical solution of this embodiment of the invention, the generation process of the time delay shaping strategy includes:

[0028] When the current equivalent multipath delay spread parameter is less than the historical optimal interval, a delay spread strategy is generated to increase the dispersion of the delay distribution of multipath components, so as to increase the delay interval between each multipath component.

[0029] When the current equivalent multipath delay spread parameter is greater than the historical best interval, a delay compression strategy is generated to reduce the dispersion of the delay distribution of multipath components, so as to reduce the delay interval between each multipath component.

[0030] As a further limitation of the technical solution of this embodiment of the invention, the increase or decrease of the dispersion of the multipath component delay distribution specifically means:

[0031] Multipath component decomposition is performed on the echo signal to obtain the time delay parameters and energy parameters of each multipath component;

[0032] When it is necessary to increase the dispersion of the delay distribution of multipath components, additional delays are applied to at least some of the multipath components to increase the delay interval of each multipath component before reconstruction.

[0033] When it is necessary to reduce the dispersion of the delay distribution of multipath components, the delay of at least some multipath components is aligned, and the delay interval of each multipath component is reduced before reconstruction.

[0034] As a further limitation of the technical solution of the embodiments of the present invention, the method further includes:

[0035] After time delay shaping and reconstruction of the echo signal, the equivalent multipath delay spread parameter of the shaped echo signal is extracted again.

[0036] Determine whether the equivalent multipath delay spread parameter falls within the historical optimal range;

[0037] If it falls into the historical optimal range, the corresponding time delay shaping strategy is used to process the echo signal of the subsequent input target filter;

[0038] If the delay does not fall within the historical optimal range, the delay shaping strategy is adjusted and the delay shaping and reconstruction process is repeated until the equivalent multipath delay extension parameter falls within the historical optimal range.

[0039] An automatic filter performance evaluation system, the system comprising:

[0040] The historical sample acquisition module is used to acquire a historical sample dataset, which includes multiple historical echo samples and their filtering results that match the working background of the current target filter.

[0041] The parameter extraction module is used to extract the equivalent multipath delay spread parameters of each historical echo sample and extract the echo peak resolution based on the filtering results.

[0042] The model building module is used to establish a variation relationship model based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, and to determine the historical optimal interval corresponding to the optimal echo peak resolution based on the variation relationship model.

[0043] The current signal analysis module is used to acquire the current echo signal, extract the equivalent multipath delay spread parameter corresponding to the current echo signal, and determine whether it falls within the historical optimal range.

[0044] The time delay shaping module is used to generate a time delay shaping strategy based on the deviation direction and amount between the current equivalent multipath delay spread parameter and the historical optimal range if the signal does not fall into the historical optimal range. The time delay shaping strategy is then used to perform time delay shaping on the echo signal input to the target filter so that the equivalent multipath delay spread of the shaped echo signal falls into the historical optimal range when it enters the target filter, thereby ensuring that the echo peak resolution is in a better state.

[0045] As a further limitation of the technical solution of the present invention, the matching with the working background of the current target filter specifically refers to the following: the historical echo samples and the current echo signal meet the preset similarity conditions in terms of sonar working parameters and environmental parameters. The sonar working parameters include the type of transmitted signal, the working frequency band and the transmitted power, and the environmental parameters include water depth, seabed reflection characteristics and noise level.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] This invention establishes a correspondence model between equivalent multipath delay spread parameters and echo peak resolution based on historical sample data, and identifies the historical optimal interval that achieves a better echo peak resolution, transforming small variations that were originally considered random disturbances into usable regular information. Based on this, by judging the deviation between the equivalent multipath delay spread parameters corresponding to the current echo signal and the historical optimal interval, a corresponding time delay shaping strategy is generated to structurally adjust the echo signal of the input target filter, making its equivalent multipath time delay distribution tend towards the optimal state, thereby improving echo peak resolution. This invention does not rely on changing the filter's structure itself, but achieves performance optimization by adjusting the characteristics of the input signal, further exploring its processing potential under existing filter conditions, and improving the stability of the filtering effect and the accuracy of subsequent target recognition. Attached Figure Description

[0048] Figure 1 A flowchart of the method provided in the embodiments of the present invention;

[0049] Figure 2 This is a flowchart illustrating the process of determining the historical optimal interval in the method provided in this embodiment of the invention;

[0050] Figure 3 The application architecture diagram of the system provided in the embodiments of the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0052] Figure 1 A flowchart of the method provided by an embodiment of the present invention is shown.

[0053] Specifically, an automatic filter performance evaluation method includes the following steps:

[0054] Step S1: Obtain the historical sample dataset, which includes multiple historical echo samples and their filtering results that match the working background of the current target filter.

[0055] The matching with the working background of the current target filter specifically means that the historical echo samples and the current echo signal meet the preset similarity conditions in terms of sonar operating parameters and environmental parameters. The sonar operating parameters include the type of transmitted signal, the operating frequency band and the transmitted power, and the environmental parameters include water depth, seabed reflection characteristics and noise level.

[0056] In this embodiment of the invention, the method is applicable to filter performance evaluation and optimization in underwater sonar data processing scenarios. Underwater sonar systems detect and identify targets by emitting sound waves and receiving target echo signals. However, the echo signals are affected by various factors during propagation, such as seawater medium characteristics, sea surface conditions, and seabed reflections, resulting in a complex signal structure with multipath characteristics. For such signals, preprocessing with filters is typically required to suppress noise and enhance effective signal components. Therefore, filters are critical processing units in sonar systems, and their performance directly affects the accuracy of subsequent target detection and identification.

[0057] For filters used in sonar data processing, the signal is inevitably affected by changes in water surface conditions during the reception and filtering process. These include dynamic changes in the reflective interface caused by wave undulations, enhanced scattering due to bubble formation, and instability in the reflection path caused by variations in water surface roughness. These factors alter the propagation paths and energy distribution of multipath components in the echo signal, resulting in changes in the time-domain distribution of these components, manifested as fluctuations in the equivalent multipath delay spread parameter. Existing filter technologies typically suppress and compensate for these interferences through bandpass design, adaptive filtering, or noise suppression algorithms. This keeps the overall impact of such interference on the filtered output within a preset standard range, generally preventing significant degradation of the filtering effect.

[0058] However, through extensive practical observation and data analysis, those skilled in the art have discovered that even when the aforementioned interference is within acceptable limits, the equivalent multipath delay spread parameter of the echo signal still undergoes a certain degree of change, further leading to a corresponding change in the echo peak resolution in the filtered output. Further analysis shows that this change is not entirely random, but exhibits a certain regularity under specific operating conditions and environmental backgrounds. Specifically, as the equivalent multipath delay spread parameter changes, the echo peak resolution shows a non-monotonic trend of first increasing and then decreasing, or first decreasing and then increasing.

[0059] It should be noted that the echo peak resolution is typically used in subsequent target detection, target separation, or feature extraction stages to determine whether adjacent echo peaks can be effectively distinguished, thus affecting the accuracy of target recognition. However, in existing technologies, filter design and optimization mainly focus on indicators such as noise suppression, signal fidelity, and overall signal-to-noise ratio improvement, while echo peak resolution, as a performance characterization parameter for subsequent processing stages, is usually not pre-correlated with the multipath structure characteristics of the filter's input signal.

[0060] Therefore, although existing technologies have controlled the fluctuations in the equivalent multipath delay spread parameter caused by environmental factors to a small range through filter optimization, they still fail to further analyze the impact of these small fluctuations on the echo peak resolution, and even more so, they fail to establish a correspondence between the equivalent multipath delay spread parameter and the echo peak resolution. In other words, existing technologies have not linked the small changes in the equivalent multipath delay spread parameter that still exist before and after filtering to the subtle changes in the echo peak resolution in subsequent stages. Consequently, they have failed to identify the potential non-monotonic variation patterns and have failed to utilize these patterns to further optimize or control the filter performance.

[0061] Based on the above findings, this invention introduces a historical sample dataset to model the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, and further determines the historical optimal interval corresponding to the optimal echo peak resolution, thereby providing a basis for subsequent time delay shaping strategies.

[0062] The historical sample dataset originates from an echo data record database accumulated during the long-term operation of the sonar system. This database can be acquired by an actual sea trial data acquisition system or generated by a simulation platform under different combinations of environmental parameters. This database typically contains raw echo signal data, corresponding filtering results, and relevant operating and environmental parameter information. It represents a conventionally available data source in this field and is well-supported by existing technologies. In practical implementation, data relevant to the current target filter's operating background can be extracted from this database as the historical sample dataset.

[0063] In step S1, the purpose of screening historical samples is to ensure that the established change relationship model is targeted and effective. Since the relationship between the equivalent multipath delay spread parameter and the echo peak resolution is influenced by both sonar operating parameters and environmental parameters, directly using all historical data for modeling can easily introduce interference from differences between operating conditions, thereby reducing the model's accuracy. Therefore, it is necessary to screen historical echo samples that match the current target filter's operating background to ensure that the constructed change relationship can truly reflect the characteristic change patterns under the current application scenario.

[0064] In this embodiment, the screening conditions are relatively strict in order to minimize the influence of non-target factors on the modeling results, making the equivalent multipath delay spread parameter the dominant variable, thereby more accurately characterizing its relationship with echo peak resolution. The screening conditions can include not only sonar operating parameters such as transmitted signal type, operating frequency band, and transmitted power, as well as environmental parameters such as water depth, seabed reflection characteristics, and noise level, but can also be further extended to factors such as sound velocity profile characteristics, water surface state parameters (e.g., sea state), platform motion state, and target type.

[0065] It should be noted that "meeting the preset similarity conditions" does not require that historical echo samples and current echo signals be completely identical in all parameters, but rather allows for differences within a certain tolerance range. This is because in the actual marine environment, various parameters are difficult to be completely identical. If a strict and consistent screening standard is adopted, it will lead to an insufficient number of usable samples, affecting the stability of the model. Therefore, in this embodiment, by setting a parameter similarity threshold or tolerance range, the historical samples are kept consistent with the current working context in terms of overall characteristics, while retaining a certain degree of data diversity, thereby improving the robustness and generalization ability of the changing relationship model.

[0066] Furthermore, the automatic filter performance evaluation method also includes the following steps:

[0067] S2. Extract the equivalent multipath delay spread parameters of each historical echo sample, and extract the echo peak resolution based on the filtering results.

[0068] The specific calculation process of the equivalent multipath delay spread parameter includes: performing time delay decomposition on the echo signal to obtain time delay parameters and corresponding energy parameters of several multipath components; performing a weighted average of the time delay parameters of each multipath component based on the energy parameters of each multipath component to obtain the average time delay; calculating the energy-weighted root mean square time delay spread based on the deviation between the time delay parameters of each multipath component and the average time delay; and determining the energy-weighted root mean square time delay spread as the equivalent multipath delay spread parameter.

[0069] The process of extracting the echo peak resolution specifically includes: performing peak detection on the echo signal after filtering to determine the peak position and peak amplitude of different echo peaks in the echo signal; for adjacent echo peaks, calculating the difference between their peak positions as the time interval, and calculating the ratio of their peak amplitudes; and weighting the time interval and the ratio of peak amplitudes to obtain the corresponding echo peak resolution.

[0070] In this embodiment of the invention, step S2 is used to extract two core parameters from historical echo samples and their corresponding filtering results: the equivalent multipath delay spread parameter and the echo peak resolution. The equivalent multipath delay spread parameter characterizes the time-domain dispersion of several multipath components in the echo signal, while the echo peak resolution characterizes the ease with which adjacent echo peaks are distinguished in the filtered echo signal. Both parameters can be obtained from actual acquired echo signal data through deterministic signal processing and numerical calculation, possessing clear data sources, processing paths, and result meanings, thus enabling practical implementation.

[0071] To extract the equivalent multipath delay spread parameters, the echo signals corresponding to historical echo samples are first decomposed into time delay parameters and corresponding energy parameters of several multipath components. Time delay decomposition refers to splitting the received mixed echo signal into multiple multipath components with different arrival times and energy levels. This process can be implemented using conventional multipath estimation methods in the field, such as matched filtering combined with peak search, sparse decomposition, maximum likelihood estimation, subspace decomposition, or methods based on channel impulse response estimation. Any method that can identify multiple multipath components from the echo signal and provide the time delay and energy parameters of each multipath component can be used in this invention.

[0072] After obtaining the time delay and energy parameters of each multipath component, a weighted average is first calculated based on the energy parameters of each multipath component to obtain the average time delay. Specifically, the weighted average means that multipath components with higher energy have a larger weight in the average time delay calculation, while multipath components with lower energy have a smaller weight, thus making the obtained average time delay more consistent with the overall time delay distribution characteristics of the echo signal. Then, based on the deviation of the time delay parameters of each multipath component from the average time delay, weighted statistics are performed in conjunction with the corresponding energy parameters, and the root mean square (RMS) result is taken after normalization to obtain the energy-weighted RMS time delay spread. This energy-weighted RMS time delay spread is determined as the equivalent multipath delay spread parameter. In other words, this parameter essentially reflects the degree of dispersion of each multipath component around the average time delay; the greater the dispersion, the larger the parameter; the smaller the dispersion, the smaller the parameter. The energy-weighted RMS time delay spread can be calculated using conventional RMS statistical methods in this field.

[0073] In practical implementation, to ensure the stability and repeatability of the calculated equivalent multipath delay spread parameters, the echo signal can be preprocessed before time delay decomposition. This preprocessing may include removing DC components, normalizing amplitude, truncating the effective echo time window, and removing noise peaks below a preset threshold. For multipath component identification, minimum energy threshold, minimum peak value threshold, or minimum component interval can be set to avoid misidentifying noise disturbances as effective multipath components. Therefore, the calculation of the equivalent multipath delay spread parameters is not an abstract description, but rather a result obtained through step-by-step processing of measurable multipath component parameters in the echo signal, possessing a clear implementation basis.

[0074] The extraction of echo peak resolution is performed based on the filtered echo signal. Specifically, peak detection is first performed on the filtered echo signal to determine the peak position and amplitude of different echo peaks. The peak position can be represented by time position or sampling point position, and the peak amplitude is the signal amplitude at the corresponding peak position. Peak detection can be implemented using conventional local maximum detection methods, or it can be combined with minimum peak height, minimum peak spacing, and peak width constraints for filtering to eliminate the influence of noise spikes on the detection results.

[0075] After identifying multiple echo peaks, for adjacent echo peaks, the difference in their peak positions is calculated as a time interval, and the ratio of their peak amplitudes is also calculated. The time interval characterizes the separation between two adjacent echo peaks on the time axis; a larger time interval indicates that the two adjacent echo peaks are more easily distinguishable in the time domain. The ratio of peak amplitudes characterizes the balance in amplitude between two adjacent echo peaks. Preferably, the smaller peak amplitude can be divided by the larger peak amplitude, resulting in a value between zero and one. Thus, the closer the peak amplitudes of two adjacent echo peaks are, the closer the ratio of peak amplitudes is to one; the closer the peak amplitudes of two adjacent echo peaks are to zero, the closer the ratio of peak amplitudes is to zero. This method provides a more intuitive reflection of the distinguishing characteristics of adjacent echo peaks at the amplitude level.

[0076] After obtaining the ratio of time interval to peak amplitude, the two are weighted and combined to obtain the corresponding echo peak resolution. This weighted combination can be a linear weighted combination. Specifically, the weights corresponding to the time interval and the peak amplitude ratio can be preset, and the time interval is normalized before being combined with the peak amplitude ratio. The purpose of the normalization process is to eliminate the influence of the inconsistency between the time dimension and the amplitude ratio dimension, allowing them to participate in the combination within a unified numerical range. The result obtained can then be used as the echo peak resolution corresponding to a set of adjacent echo peaks. If multiple sets of adjacent echo peaks exist in the same echo signal, the echo peak resolution corresponding to each set of adjacent echo peaks can be further averaged, weighted averaged, or optimized to obtain the overall echo peak resolution corresponding to the echo signal.

[0077] In a preferred implementation, the weight of the time interval and the weight of the peak amplitude ratio can be set according to the actual application scenario. For example, when more attention is paid to the temporal separation effect of adjacent echo peaks, the weight of the time interval can be appropriately increased; when more attention is paid to the balance of amplitude between adjacent echo peaks, the weight of the peak amplitude ratio can be appropriately increased. This weight setting method allows those skilled in the art to flexibly adjust it according to different application requirements, while not affecting the basic idea of ​​the present invention for extracting the resolvability of echo peaks.

[0078] It should be noted that the above-mentioned process for extracting echo peak resolution also has a clear implementation basis. The data it relies on all comes from the echo signal itself after filtering, and the peak detection, time interval calculation, amplitude ratio calculation, and weighted combination methods used are all conventional and feasible data processing methods in this field. Therefore, this invention does not merely propose conceptual parameters, but rather provides a directly executable implementation path.

[0079] Furthermore, the automatic filter performance evaluation method also includes the following steps:

[0080] Step S3: Based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, establish a variation relationship model, and determine the historical optimal interval corresponding to the optimal echo peak resolution according to the variation relationship model.

[0081] Specifically, Figure 2 A flowchart for determining the historical best interval is shown.

[0082] The process of establishing a variation relationship model based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, and determining the historical optimal interval corresponding to the optimal echo peak resolution based on the variation relationship model, specifically includes the following steps:

[0083] Step S31: Sort several historical echo samples in ascending order according to the equivalent multipath delay spread parameter to obtain a sample sequence;

[0084] Step S32: Obtain the echo peak resolution corresponding to the sample sequence, and perform curve fitting based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution to obtain the change relationship model;

[0085] Step S33: Analyze the change relationship model to determine whether it exhibits a non-monotonic change trend of first rising and then falling;

[0086] Step S34: When the change relationship model exhibits a non-monotonic trend of first rising and then falling, determine the peak position of the change relationship model, and extract a parameter interval within a preset range as the historical optimal interval, centered on the equivalent multipath delay spread parameter corresponding to the peak position. In other embodiments, the non-monotonic change trend can also be expressed as first falling and then rising, as long as the parameter interval corresponding to the better echo peak resolution can be determined.

[0087] In this embodiment of the invention, step S3 is used to mine the intrinsic correlation between the equivalent multipath delay spread parameter and the echo peak resolution based on historical sample data, thereby determining the parameter range that optimizes the echo peak resolution. This step directly echoes the research objective described in step S1, which is to eliminate interference from non-target factors and highlight the impact of multipath delay distribution on the filtering results by screening historical echo samples that match the current target filter's operating background. Based on this, by modeling the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution in historical samples, small fluctuations that were originally considered random disturbances can be transformed into regular trends, thus providing a basis for subsequent delay shaping strategies.

[0088] Specifically, in step S31, historical echo samples are sorted from smallest to largest according to the equivalent multipath delay spread parameter, transforming discrete samples into an ordered sample sequence. This makes the parameter change process continuous at the data level, facilitating subsequent trend analysis. In step S32, based on the ordered sample sequence, the corresponding echo peak resolution is obtained, and the correspondence between the two is smoothed using conventional curve fitting methods (such as polynomial fitting, moving average, or local regression methods), thereby obtaining a change relationship model and avoiding the influence of discrete data fluctuations on trend judgment. In step S33, the change relationship model is analyzed, for example, based on derivative sign changes, local extremum detection, or trend segment division methods, to determine whether the echo peak resolution exhibits a non-monotonic trend of first increasing and then decreasing with the change of the equivalent multipath delay spread parameter. In step S34, after confirming the existence of a non-monotonic change trend, the peak position of the change relationship model is further determined using an extremum detection method. The corresponding parameter interval is extracted using the equivalent multipath delay spread parameter corresponding to this peak position as the center, combined with a preset window range, as the historical optimal interval.

[0089] Determining the historical optimal range involves extracting the optimal matching range between the equivalent multipath delay spread parameter and the echo peak resolution from historical samples. This transforms the originally scattered and difficult-to-use performance variation patterns into quantifiable range constraints, thus providing a basis for subsequent judgment on whether the current echo signal state is in a favorable range. At the same time, when the current equivalent multipath delay spread parameter deviates from this range, a targeted time delay shaping strategy can be generated to make the echo signal input to the target filter approach the optimal range in an equivalent manner, thereby stabilizing and improving the echo peak resolution, and ultimately improving the overall filtering effect.

[0090] Furthermore, the automatic filter performance evaluation method also includes the following steps:

[0091] S4. Obtain the current echo signal, extract the equivalent multipath delay spread parameter corresponding to the current echo signal, and determine whether it falls within the historical optimal range.

[0092] S5. If the signal does not fall within the historical optimal range, a time delay shaping strategy is generated based on the deviation direction and amount between the current equivalent multipath delay spread parameter and the historical optimal range. The time delay shaping strategy is then used to perform time delay shaping on the echo signal input to the target filter so that the equivalent multipath delay spread of the shaped echo signal falls within the historical optimal range when it enters the target filter, thereby ensuring that the echo peak resolution is in a better state.

[0093] The generation process of the delay shaping strategy includes: when the current equivalent multipath delay expansion parameter is less than the historical optimal interval, generating a delay expansion strategy to increase the dispersion of the delay distribution of multipath components, so as to increase the delay interval between each multipath component; when the current equivalent multipath delay expansion parameter is greater than the historical optimal interval, generating a delay compression strategy to reduce the dispersion of the delay distribution of multipath components, so as to reduce the delay interval between each multipath component.

[0094] The process of increasing or decreasing the dispersion of multipath component delay distribution specifically involves: performing multipath component decomposition on the echo signal to obtain the delay parameters and energy parameters of each multipath component; when it is necessary to increase the dispersion of multipath component delay distribution, applying additional delays to at least some multipath components to increase the delay interval of each multipath component before reconstruction; when it is necessary to decrease the dispersion of multipath component delay distribution, aligning the delays of at least some multipath components to decrease the delay interval of each multipath component before reconstruction.

[0095] After time delay shaping and reconstruction of the echo signal, the equivalent multipath delay spread parameter of the shaped echo signal is extracted again. It is then determined whether the equivalent multipath delay spread parameter falls within the historical optimal range. If it falls within the historical optimal range, the corresponding time delay shaping strategy is used to process the echo signal of the subsequent input target filter. If it does not fall within the historical optimal range, the time delay shaping strategy is adjusted and the time delay shaping and reconstruction process is repeated until the equivalent multipath delay spread parameter falls within the historical optimal range.

[0096] In this embodiment of the invention, steps S4 and S5 constitute a closed-loop processing mechanism for dynamically optimizing and continuously monitoring the filter performance evaluation results. The current echo signal obtained in step S4 can be a short signal segment from a continuous sonar data stream, used for real-time or near-real-time evaluation of the filter performance under the current operating state. By extracting the equivalent multipath delay spread parameter corresponding to the current echo signal and comparing it with the historical best interval, it is possible to quickly determine whether the multipath delay distribution of the current input signal is in a favorable state.

[0097] When the judgment result indicates that the current equivalent multipath delay spread parameter does not fall within the historical optimal range, step S5 initiates the generation and execution process of the delay shaping strategy. Specifically, firstly, based on the positional relationship of the current equivalent multipath delay spread parameter relative to the historical optimal range, its deviation direction and deviation amount are determined. The deviation direction is used to determine whether the current delay distribution is too concentrated or too dispersed, and the deviation amount is used to quantify the degree of difference between it and the optimal range. Based on this, a corresponding delay shaping strategy is generated.

[0098] When the current equivalent multipath delay spread parameter is less than the historical best range, it indicates that the multipath components in the echo signal are too concentrated in the time domain, easily leading to overlap between echo peaks and reducing the resolvability of the echo peaks. In this case, a delay spread strategy is generated by applying additional delays to some multipath components to appropriately increase their spacing on the time axis. Specifically, the echo signal can be first decomposed into multipath components to obtain the delay and energy parameters of each component; then, multipath components with higher energy or greater impact on the overall structure are selected, and their delay values ​​are increased according to preset rules or proportionally; finally, the adjusted multipath components are re-superimposed to obtain a new echo signal. This processing increases the delay interval between the multipath components, thereby improving the separation between echo peaks.

[0099] When the current equivalent multipath delay spread parameter is greater than the historical best range, it indicates that the multipath components in the echo signal are too dispersed in the time domain, which may lead to an overly discrete energy distribution and affect the concentration of the main echo peaks. In this case, a time delay compression strategy is generated by aligning the time delays of some multipath components, bringing them closer to the main path or the center delay position. Specifically, this involves first decomposing the multipath components and identifying the time delay and energy parameters of each component; then selecting some multipath components and shifting their delays towards the average time delay or the main path delay to reduce the time interval between different paths; finally, signal reconstruction is performed. This method makes the multipath components more concentrated on the time axis, thereby enhancing the clustering characteristics of the echo peaks.

[0100] After completing one round of time delay shaping and signal reconstruction, the equivalent multipath delay spread parameter of the shaped echo signal is extracted again, and it is determined whether it falls into the historical optimal range. If it has, the current shaping strategy is effective and can be applied to the echo signal processing of the subsequent input target filter. If it still has not fallen into the range, the time delay shaping strategy is adjusted according to the new deviation, such as changing the magnitude of the additional time delay, adjusting the number of multipath components involved in shaping, or modifying the alignment reference position. The above shaping and reconstruction process is repeated until the equivalent multipath delay spread parameter enters the historical optimal range. This forms a closed-loop optimization process driven by the evaluation results, ensuring that the echo signal is always in a relatively optimal multipath time delay distribution state before entering the target filter.

[0101] Through the above steps, this invention achieves a closed-loop transition from "evaluation" to "control," not only identifying whether the current echo signal is in a favorable state but also proactively adjusting the signal structure when it deviates from the optimal range, thereby improving the resolvability of the echo peak. Compared with existing technologies, this invention is not limited to optimizing the filter structure itself but rather achieves better processing results under the existing structure by controlling the equivalent multipath characteristics of the input signal, thus further exploring the performance potential of the filter.

[0102] From an overall beneficial perspective, this invention addresses a key issue raised in the aforementioned research: the relationship between the equivalent multipath delay spread parameter, which exhibits regular variations despite interference remaining within acceptable limits, and the resolvability of the echo peak value. By determining the historical optimal interval and introducing a closed-loop delay shaping strategy, this invention effectively utilizes this regularity. This method can further enhance the resolvability of the echo peak value without altering the filter hardware structure or algorithm framework, thereby improving the performance of subsequent target detection and recognition.

[0103] In terms of application prospects, this invention can be widely applied to data processing scenarios of various underwater sonar systems, including but not limited to underwater target detection, seabed topography mapping, underwater communication, and underwater acoustic monitoring. Especially in complex sea conditions or environments with significant multipath effects, this invention can improve the stability and robustness of filtering processing through adaptive control of the echo signal structure, demonstrating good engineering application value and promising prospects for widespread adoption.

[0104] Furthermore, Figure 3 An application architecture diagram of the system provided in an embodiment of the present invention is shown.

[0105] In another preferred embodiment of the present invention, an automatic filter performance evaluation system includes:

[0106] The historical sample acquisition module 100 is used to acquire a historical sample dataset, which includes multiple historical echo samples and their filtering results that match the operating background of the current target filter. Specifically, matching the operating background of the current target filter means that the historical echo samples and the current echo signal meet preset similarity conditions in terms of sonar operating parameters and environmental parameters. The sonar operating parameters include the transmitted signal type, operating frequency band, and transmitted power, while the environmental parameters include water depth, seabed reflection characteristics, and noise level.

[0107] Furthermore, the automatic filter performance evaluation system also includes:

[0108] The parameter extraction module 200 is used to extract the equivalent multipath delay spread parameters of each historical echo sample and extract the echo peak resolution based on the filtering results.

[0109] Furthermore, the automatic filter performance evaluation system also includes:

[0110] The model building module 300 is used to establish a variation relationship model based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, and to determine the historical optimal interval corresponding to the optimal echo peak resolution based on the variation relationship model.

[0111] Furthermore, the automatic filter performance evaluation system also includes:

[0112] The current signal analysis module 400 is used to acquire the current echo signal, extract the equivalent multipath delay spread parameter corresponding to the current echo signal, and determine whether it falls within the historical optimal range.

[0113] Furthermore, the automatic filter performance evaluation system also includes:

[0114] The time delay shaping module 500 is used to generate a time delay shaping strategy based on the deviation direction and amount between the current equivalent multipath delay spread parameter and the historical optimal range if the signal does not fall into the historical optimal range. The time delay shaping strategy is then used to perform time delay shaping on the echo signal input to the target filter so that the equivalent multipath delay spread of the shaped echo signal falls into the historical optimal range when it enters the target filter, thereby making the echo peak resolution better.

[0115] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0116] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0117] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0118] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0119] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An automatic filter performance evaluation method, characterized in that, The method includes: S1. Obtain the historical sample dataset, which includes multiple historical echo samples and their filtering results that match the working background of the current target filter. S2. Extract the equivalent multipath delay spread parameters of each historical echo sample, and extract the echo peak resolution based on the filtering results; S3. Based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, establish a variation relationship model, and determine the historical optimal interval corresponding to the optimal echo peak resolution according to the variation relationship model. S4. Obtain the current echo signal, extract the equivalent multipath delay spread parameter corresponding to the current echo signal, and determine whether it falls within the historical optimal range. S5. If the signal does not fall within the historical optimal range, a time delay shaping strategy is generated based on the deviation direction and amount between the current equivalent multipath delay spread parameter and the historical optimal range. The time delay shaping strategy is then used to perform time delay shaping on the echo signal input to the target filter so that the equivalent multipath delay spread of the shaped echo signal falls within the historical optimal range when it enters the target filter, thereby ensuring that the echo peak resolution is in a better state.

2. The automatic filter performance evaluation method according to claim 1, characterized in that, The matching with the working background of the current target filter specifically means that the historical echo samples and the current echo signal meet the preset similarity conditions in terms of sonar operating parameters and environmental parameters. The sonar operating parameters include the type of transmitted signal, the operating frequency band and the transmitted power, and the environmental parameters include water depth, seabed reflection characteristics and noise level.

3. The automatic filter performance evaluation method according to claim 1, characterized in that, The specific calculation process for the equivalent multipath delay spread parameter includes: The echo signal is decomposed by time delay to obtain the time delay parameters and corresponding energy parameters of several multipath components; The average delay is obtained by weighting the delay parameters of each multipath component based on the energy parameters of each multipath component. The energy-weighted root mean square delay spread is calculated based on the deviation between the delay parameters of each multipath component and the average delay. The energy-weighted root mean square delay spread is determined as the equivalent multipath delay spread parameter.

4. The automatic filter performance evaluation method according to claim 1, characterized in that, The process of extracting the echo peak resolution specifically includes: Peak detection is performed on the echo signal after filtering to determine the peak position and peak amplitude of different echo peaks in the echo signal; For adjacent echo peaks, calculate the difference between their peak positions as the time interval, and calculate the ratio of their peak amplitudes; The corresponding echo peak resolution is obtained by weighting and combining the ratio of the time interval to the peak amplitude.

5. The automatic filter performance evaluation method according to claim 1, characterized in that, S3 specifically includes: S31. Sort several historical echo samples in ascending order according to the equivalent multipath delay spread parameter to obtain a sample sequence; S32. Obtain the echo peak resolution corresponding to the sample sequence, and perform curve fitting based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution to obtain the change relationship model. S33. Analyze the change relationship model to determine whether it exhibits a non-monotonic change trend of first rising and then falling. S34. When the change relationship model shows a non-monotonic change trend of first rising and then falling, determine the peak position of the change relationship model, and extract the parameter interval within a preset range as the historical optimal interval, centered on the equivalent multipath delay expansion parameter corresponding to the peak position.

6. The automatic filter performance evaluation method according to claim 1, characterized in that, The generation process of the time delay shaping strategy includes: When the current equivalent multipath delay spread parameter is less than the historical optimal interval, a delay spread strategy is generated to increase the dispersion of the delay distribution of multipath components, so as to increase the delay interval between each multipath component. When the current equivalent multipath delay spread parameter is greater than the historical best interval, a delay compression strategy is generated to reduce the dispersion of the delay distribution of multipath components, so as to reduce the delay interval between each multipath component.

7. The automatic filter performance evaluation method according to claim 6, characterized in that, The increase or decrease in the dispersion of multipath component delay distribution specifically refers to: Multipath component decomposition is performed on the echo signal to obtain the time delay parameters and energy parameters of each multipath component; When it is necessary to increase the dispersion of the delay distribution of multipath components, additional delays are applied to at least some of the multipath components to increase the delay interval of each multipath component before reconstruction. When it is necessary to reduce the dispersion of the delay distribution of multipath components, the delay of at least some multipath components is aligned, and the delay interval of each multipath component is reduced before reconstruction.

8. The automatic filter performance evaluation method according to claim 7, characterized in that, The method further includes: After time delay shaping and reconstruction of the echo signal, the equivalent multipath delay spread parameter of the shaped echo signal is extracted again. Determine whether the equivalent multipath delay spread parameter falls within the historical optimal range; If it falls into the historical optimal range, the corresponding time delay shaping strategy is used to process the echo signal of the subsequent input target filter; If the delay does not fall within the historical optimal range, the delay shaping strategy is adjusted and the delay shaping and reconstruction process is repeated until the equivalent multipath delay extension parameter falls within the historical optimal range.

9. An automatic filter performance evaluation system, characterized in that, The system includes: The historical sample acquisition module is used to acquire a historical sample dataset, which includes multiple historical echo samples and their filtering results that match the working background of the current target filter. The parameter extraction module is used to extract the equivalent multipath delay spread parameters of each historical echo sample and extract the echo peak resolution based on the filtering results. The model building module is used to establish a variation relationship model based on the correspondence between the equivalent multipath delay spread parameter and the echo peak resolution, and to determine the historical optimal interval corresponding to the optimal echo peak resolution based on the variation relationship model. The current signal analysis module is used to acquire the current echo signal, extract the equivalent multipath delay spread parameter corresponding to the current echo signal, and determine whether it falls within the historical optimal range. The time delay shaping module is used to generate a time delay shaping strategy based on the deviation direction and amount between the current equivalent multipath delay spread parameter and the historical optimal range if the signal does not fall into the historical optimal range. The time delay shaping strategy is then used to perform time delay shaping on the echo signal input to the target filter so that the equivalent multipath delay spread of the shaped echo signal falls into the historical optimal range when it enters the target filter, thereby ensuring that the echo peak resolution is in a better state.

10. The automatic filter performance evaluation system according to claim 9, characterized in that, The matching with the working background of the current target filter specifically means that the historical echo samples and the current echo signal meet the preset similarity conditions in terms of sonar operating parameters and environmental parameters. The sonar operating parameters include the type of transmitted signal, the operating frequency band and the transmitted power, and the environmental parameters include water depth, seabed reflection characteristics and noise level.

Images