X-band phased array weather radar electromagnetic interference suppression method and device

Abstract

The application provides an X-band phased array weather radar electromagnetic interference suppression method and device, comprising: identifying abnormal interference pulses from pulse domain original signals collected by an X-band phased array weather radar, and performing interpolation correction on the abnormal interference pulses; processing the pulse domain original signals after interpolation correction to obtain reflectivity factor data as radar echo data; extracting multi-dimensional feature parameters from the radar echo data; performing fuzzy processing on each feature parameter using a fuzzy logic discrimination algorithm, and in the case where it is determined according to the fuzzy processing result that the radar echo data is radial ray interference echo, the radar echo data is removed. The application realizes accurate identification and suppression of various electromagnetic interferences.

Description

Technical Field

[0001] This invention relates to the field of weather monitoring technology, and in particular to a method and apparatus for suppressing electromagnetic interference of X-band phased array weather radar. Background Technology

[0002] X-band phased array weather radar, as a new generation of meteorological detection equipment, plays a crucial role in monitoring and refining severe convective weather due to its high spatiotemporal resolution and flexible beam scanning capabilities. However, with the rapid development of wireless communication technology and the increasing scarcity of spectrum resources, radar data often contains electromagnetic interference echoes caused by electromagnetic radiation from external devices operating on the same or adjacent frequencies. These electromagnetic interference echoes manifest morphologically as radial ray interference echoes, spiral band interference echoes, or strip-shaped speckled interference echoes. Radial ray interference echoes are characterized by rays extending from the radar center to the furthest point of detection, with a relatively uniform texture. Spiral band interference echoes appear as spiral bands on the radar's planar position display, exhibiting a distinct spiral structure. Strip-shaped speckled interference echoes appear as strip-shaped specks extending over long distances, with a relatively coarse texture.

[0003] Electromagnetic interference echoes, as non-meteorological echoes, severely affect the normal observation of radar base data and further reduce the accuracy of radar products. For example, when using reflectivity factor data for precipitation estimation, electromagnetic interference echoes may cause the algorithm to incorrectly estimate precipitation in areas without precipitation, thus affecting the accuracy of precipitation estimation and causing missed or false alarms for short-term heavy precipitation. Therefore, identifying and removing electromagnetic interference echoes from radar echo data is crucial for improving radar data quality and the accuracy of subsequent precipitation estimation.

[0004] Current methods for electromagnetic interference (EMI) suppression include filtering, power analysis, fuzzy logic, and neural networks. However, these methods can only identify a limited number of EMI echoes, processing only a single type. Their accuracy in identifying EMI echoes is low, and they suffer from significant false cancellation issues for meteorological echoes. Therefore, developing an efficient, comprehensive, and accurate EMI suppression method for X-band phased array weather radar is of significant practical importance. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method and apparatus for suppressing electromagnetic interference in X-band phased array weather radar.

[0006] This invention provides a method for suppressing electromagnetic interference in X-band phased array weather radar, comprising:

[0007] Abnormal interference pulses are identified from the raw pulse domain signals acquired by the X-band phased array weather radar, and the abnormal interference pulses are corrected by interpolation.

[0008] The original pulse domain signal after interpolation correction is processed to obtain reflectivity factor data as radar echo data;

[0009] Extract multidimensional feature parameters from the radar echo data;

[0010] The fuzzy logic discrimination algorithm is used to fuzzify each feature parameter. If the radar echo data is determined to be radial ray interference echo based on the fuzzification result, the radar echo data is discarded.

[0011] The present invention also provides an electromagnetic interference suppression device for X-band phased array weather radar, comprising:

[0012] The correction module is used to identify abnormal interference pulses from the raw pulse domain signals acquired by the X-band phased array weather radar and to perform interpolation correction on the abnormal interference pulses.

[0013] The processing module is used to process the interpolated pulse domain original signal to obtain reflectivity factor data as radar echo data.

[0014] The extraction module is used to extract multi-dimensional feature parameters from the radar echo data;

[0015] The elimination module is used to perform fuzzification processing on each feature parameter using a fuzzy logic discrimination algorithm. If the radar echo data is determined to be a radial ray interference echo based on the fuzzification processing result, the radar echo data is eliminated.

[0016] The electromagnetic interference suppression method and apparatus for X-band phased array weather radar provided by this invention first processes helical band interference and strip-shaped speckle interference at the signal processing layer using pulse interpolation; then, at the echo identification layer, it extracts the power difference between near and far distances, the proportion of strong echoes, the inter-pulse correlation coefficient, the extended average power, the radial gradient, and the inter-column correlation coefficient, and uses fuzzy logic to identify and eliminate radial ray interference echoes, thereby achieving accurate identification and suppression of various electromagnetic interferences.

[0017] The core innovation and beneficial effects are as follows:

[0018] (1) Architecture innovation: For the first time, a layered progressive processing architecture of “signal processing layer (pulse-level interference source suppression) + echo identification layer (accurate removal of echo domain structural interference)” is proposed. The two algorithms perform their respective functions and complement each other, achieving full coverage of spiral band interference, strip-shaped pockmark interference and radial ray interference.

[0019] (2) Algorithm innovation: In the signal processing layer, a two-level pulse-level interference suppression method was designed. The burst pulse is quickly corrected by comparing the amplitude of adjacent periods, and the discrete pulse is selected by adaptive window selection and threshold filtering based on "multi-scale clutter feature saliency". It can simultaneously handle isolated spike interference and continuous abnormal pulses.

[0020] (3) Protecting meteorological echoes: Through pulse interpolation (instead of direct zeroing) and multi-feature screening mechanism, while effectively suppressing electromagnetic interference, the continuity and power spectrum characteristics of meteorological echoes are preserved to the greatest extent, significantly reducing the false cancellation rate of meteorological echoes in precipitation areas.

[0021] (4) Demonstration significance: This method deeply integrates radar meteorological business logic with image recognition and statistical feature extraction technology, providing an engineering-featured technical path for the quality control of electromagnetic interference of phased array weather radar. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the electromagnetic interference suppression method for X-band phased array weather radar provided by the present invention.

[0024] Figure 2 This is another schematic diagram of the electromagnetic interference suppression method for X-band phased array weather radar provided by the present invention;

[0025] Figure 3 These are reflectivity factor diagrams before and after processing by the suppression algorithm in the X-band phased array weather radar electromagnetic interference suppression method provided by this invention. Part (a) is the original reflectivity factor diagram; part (b) is the reflectivity factor diagram after processing by the radial fuzzy logic algorithm; part (c) is the reflectivity factor diagram after processing by the Vaisala interference filter algorithm; and part (d) is the reflectivity factor diagram after suppression by the algorithm of this invention.

[0026] Figure 4 This is a schematic diagram of the structure of the electromagnetic interference suppression device for X-band phased array weather radar provided by the present invention. Detailed Implementation

[0027] 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 with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0028] The following is combined Figure 1 The present invention describes an electromagnetic interference suppression method for X-band phased array weather radar, comprising:

[0029] Step 101: Identify abnormal interference pulses from the original pulse domain signal acquired by the X-band phased array weather radar, and perform interpolation correction on the abnormal interference pulses.

[0030] Step 102: Process the interpolated pulse domain original signal to obtain reflectivity factor data as radar echo data;

[0031] Step 103: Extract multi-dimensional feature parameters from the radar echo data;

[0032] Step 104: Use a fuzzy logic discrimination algorithm to fuzzify each feature parameter. If the radar echo data is determined to be a radial ray interference echo based on the fuzzification result, the radar echo data is discarded.

[0033] Figure 2 This is another flowchart illustrating the electromagnetic interference suppression method for X-band phased array weather radar of the present invention. The suppression method mainly includes a comprehensive progressive processing of two layers: a signal processing layer and an echo identification layer.

[0034] The signal processing layer is the pre-processing layer for the raw radar signal, operating during the I / Q data acquisition stage and preceding the generation of reflectivity factor data. This layer focuses on optimizing the raw signal in the pulse domain. By identifying abnormal interference pulses and performing interpolation corrections, it suppresses raw pulse-type interference such as helical band interference and stripe-shaped speckle interference at their source, thus purifying the raw radar echo signal from the ground up. The radar data suppressed by the signal processing layer is then passed to the echo identification layer for further electromagnetic interference suppression.

[0035] The echo identification layer is a post-processing layer for meteorological echo data, operating on the mature data stage of reflectivity factor generated after signal processing. This layer relies on the echo statistical and spatial distribution characteristics of the reflectivity factor field, using multi-dimensional feature extraction combined with fuzzy logic discrimination algorithms to accurately identify and remove radial ray-type electromagnetic interference echoes, thus eliminating residual interference.

[0036] This embodiment proposes an electromagnetic interference suppression method for X-band phased array weather radar based on hierarchical progressive processing. First, at the signal processing layer, helical band interference and strip-shaped speckle interference are processed by pulse interpolation. Then, at the echo identification layer, the power difference between near and far distances, inter-pulse correlation, radial structure characteristics, proportion of strong echoes, extended average power, and inter-column correlation coefficient are extracted. Fuzzy logic is used to identify and eliminate radial ray interference echoes, thereby achieving accurate identification and suppression of various electromagnetic interferences.

[0037] Based on the above embodiments, the abnormal interference pulse in this embodiment includes a sudden interference pulse. The abnormal interference pulse is identified from the raw pulse domain signal acquired by the X-band phased array weather radar, and interpolation correction is performed on the abnormal interference pulse, including:

[0038] Obtain pulse groups from the same distance library from the original pulse domain signal;

[0039] Calculate the ratio between the amplitude of each pulse in the pulse group acquired in the current scan cycle and the amplitude of the normal pulse in the pulse group acquired in the previous scan cycle.

[0040] If the ratio is greater than a first preset threshold, the pulse is determined to be a sudden interference pulse; otherwise, the pulse is determined to be a normal pulse.

[0041] The amplitude of the sudden interference pulse is replaced and completed using the amplitude of the normal pulse in the pulse group acquired in the previous scan cycle.

[0042] First, the raw pulse domain signal (raw I / Q data) of the X-band phased array weather radar is acquired and pulse compression is performed. The raw I / Q data is organized in the form of "azimuth – range library – pulse group", where each range library corresponds to a group of continuous pulse data S, denoted as:

[0043] (1)

[0044] Where s1 represents the complex echo signal of the first pulse, s2 represents the complex echo signal of the second pulse, and s N This represents the complex echo signal of the Nth pulse, where N is the total number of pulses in a pulse group. The complex echo signal s of the nth pulse... n The representation is as follows, where I n For the real part, Q n The imaginary part is j, which is the square root of -1.

[0045] (2)

[0046] Sudden pulse interference refers to short-duration abnormal pulse interference that occurs randomly within a radar pulse group. This type of interference is typically characterized by a low overall duty cycle, short duration, and discontinuous spatial distribution. Within the same pulse group, it usually manifests as only a single or a few abnormal pulses, rather than forming continuous, large areas of interference. It is a direct manifestation of helical band interference at the underlying pulse signal level.

[0047] To address this type of interference, pulse-level anomaly detection and replacement processing are performed. Specifically, this includes:

[0048] 1) Traverse each pulse in the pulse group under the same distance library, and let the amplitude of the pulse signal in the current scan cycle be A. n The normal pulse amplitude of the distance library corresponding to the previous scan cycle is A. n-1 Calculate the amplitude of the nth pulse and the amplitude ratio R between the current pulse and the pulse of the previous cycle. n The calculation formula is shown below, R n Used to quantify the degree of amplitude change of the current pulse:

[0049] (3)

[0050] (4)

[0051] 2) Based on the prior characteristics of actual radar noise and electromagnetic interference, set an amplitude ratio threshold (first preset threshold) T. R The value is 2, when R is satisfied. n >T R When this occurs, the pulse is determined to be a sudden interference pulse. Such pulses are often characterized by a sharp increase in local power, isolated distribution, and adjacent pulses being stable and normal echoes. R The parameters can be obtained from typical electromagnetic interference data of X-band phased array weather radar through statistical analysis and manual calibration.

[0052] 3) Abnormal interference pulses detected during the current scan period T The normal pulses of the same distance library from the previous scan cycle T-1 are directly used. The amplitude is replaced and completed, and while preserving the continuity of the real meteorological echo signal, sudden electromagnetic interference pulses, such as spiral band interference echoes, are quickly filtered out, completing the first level of interference pre-suppression in the signal processing layer.

[0053] (5)

[0054] Based on the above embodiments, the abnormal interference pulse in this embodiment also includes discrete interference pulses. Abnormal interference pulses are identified from the original pulse domain signal acquired by the X-band phased array weather radar, and interpolation correction is performed on the abnormal interference pulses, including:

[0055] Calculate the clutter feature saliency of each distance library under different candidate window lengths;

[0056] The optimal window length for each distance library is determined based on the clutter feature salience of each distance library under different candidate window lengths.

[0057] Construct radial windows centered on each distance library and the optimal window length, and count the proportion of sudden interference pulses within the radial windows;

[0058] When the proportion is greater than or equal to the second preset threshold, the relative power of each pulse in the radial window is averaged and compressed, and then compared with the adaptive judgment threshold. Based on the comparison result, it is determined whether each pulse is a discrete interference pulse.

[0059] The power of the discrete interference pulses is fitted and corrected using a quadratic interpolation algorithm.

[0060] Discrete pulse interference refers to the high duty cycle, continuously distributed, patchy abnormal pulses that remain in the radar signal after the first stage of burst-type pulse interference suppression. This type of interference exhibits multi-pulse continuous distortion characteristics, with abnormal pulses preceding and following the interference sample, and is usually not effectively identified and corrected using the single-cycle amplitude ratio method.

[0061] For interference signals that are not completely filtered out by the above-mentioned sudden pulses, such as strip-shaped pockmarks, a pulse suppression algorithm combining basis noise estimation and two-dimensional window adaptive threshold is adopted. The local basis noise is estimated in the pulse-distance two-dimensional plane, and an adaptive threshold is constructed based on the statistical characteristics in the two-dimensional window to further identify and suppress residual abnormal pulses.

[0062] To avoid false suppression of continuous strong meteorological echoes, before performing discrete pulse interference suppression, the proportion of pulses marked as sudden interference within the current radial window is first counted. .like (A second preset threshold can be used) If the current signal is not detected, skip the current suppression stage and output the current signal directly; otherwise, proceed to the subsequent adaptive threshold detection process.

[0063] Based on the above embodiments, the steps for obtaining the relative power of each pulse in this embodiment include:

[0064] Calculate the power of each pulse in the pulse group of each distance library;

[0065] Select a preset number of pulses with the lowest power from the pulse group, and calculate the average power of the selected pulses as the local base noise;

[0066] The ratio between the power of each pulse and the local base noise is used as the relative power.

[0067] For each pulse group S in the distance library, calculate the pulse s of each pulse. n Power P n :

[0068] (6)

[0069] Select the M pulses with the lowest power from the pulse group, where M can be 3, and calculate the average value as the local base noise P. base The M-parameter represents the current background noise floor level of the unit. The M-parameter can be set based on statistical analysis and manual calibration using typical electromagnetic interference data from an X-band phased array weather radar.

[0070] (7)

[0071] Among them, P min , k This represents the power of the k-th pulse among the M pulses with the lowest power after sorting within the pulse group.

[0072] Calculate each pulse s n relative power P rel (n):

[0073] (8)

[0074] Based on the radial echo variation characteristics, the average gradient G in the radial direction of each pulse is calculated in the distance dimension. r :

[0075] (9)

[0076] Where R represents the total number of distances in a single radial path. and These represent the power of each pulse in the pulse array corresponding to the i-th and i+1-th distance libraries, respectively.

[0077] Based on the above embodiments, this embodiment calculates the clutter feature saliency of each distance library under different candidate window lengths, including:

[0078] Construct a current radial window centered on each distance library based on each distance library and each candidate window length;

[0079] Calculate the mean and standard deviation of the relative power of each pulse within the current radial window;

[0080] Calculate the mean and standard deviation of the relative power of each pulse within the neighborhood background window of the current radial window;

[0081] Calculate the average radial variation characteristics within the current radial window and the neighboring background window of the current radial window, respectively;

[0082] The clutter feature significance is calculated based on the mean and standard deviation of the relative power of each pulse in the current radial window, the mean and standard deviation of the relative power of each pulse in the neighboring background window, and the average radial variation characteristics in the current radial window and the neighboring background window.

[0083] To avoid selecting the window length solely based on the average radial gradient and a fixed threshold, a multi-scale statistical feature significance method is used to adaptively determine the radial window length. Let the set of candidate radial window lengths L be:

[0084] (10)

[0085] Let the relative power of the nth pulse in the pulse group at the r-th distance be . Given the current distance i and the length of the m-th candidate window. Construct a radial window centered at i:

[0086] (11)

[0087] And construct its neighborhood background window:

[0088] (12)

[0089] in, This represents the number of pulses within the pulse group. Used to characterize the background echo level around the current distance from the reservoir.

[0090] At each candidate window scale, the mean power, standard deviation, and radial variation characteristics of the current window and the background window are calculated respectively. Their mean... and standard deviation They are represented as follows:

[0091] (13)

[0092] (14)

[0093] Here, A can represent the current window. or background window ,Right now This represents the average relative power within the current window. This represents the average relative power within the background window; This represents the number of valid samples within the corresponding window.

[0094] Optionally, the average radial variation characteristics within the current radial window and its neighboring background window are calculated, including:

[0095] Calculate the absolute value of the difference between the relative power of each pulse in the current radial window at the r-th and (r+1)-th distances, and average the absolute values ​​of all pulses in the current radial window to obtain the average radial variation characteristic in the current radial window.

[0096] Calculate the absolute value of the difference between the relative power of each pulse in the neighborhood background window of the current radial window at the r-th and r+1-th distances. Take the average of the absolute values ​​corresponding to all pulses in the neighborhood background window as the average radial variation feature in the neighborhood background window of the current radial window.

[0097] Radial variation characteristics are defined as follows:

[0098] (15)

[0099] Where G(n,r) represents the radial variation characteristic of the nth pulse at the rth distance, and P rel (n,r) and P rel (n, r+1) represent the relative power of the nth pulse at distances r and r+1, respectively.

[0100] Calculate the average radial variation characteristics of all pulses within the current window. and the average radial variation characteristics of all pulses within the neighborhood background window .

[0101] The clutter feature significance can be obtained by weighted summing the differences (or ratios) between the mean, standard deviation, and average radial variation characteristics of the relative power of each distance library within the current radial window and the corresponding characteristics within the neighborhood background window.

[0102] We can also construct the following distance library with the length of the m-th candidate window: Clutter feature salience:

[0103] (16)

[0104] in, The length of the i-th distance library in the m-th candidate window The significance of clutter characteristics is shown below. All three fractions are dimensionless ratios, and background statistics are introduced into the denominators. Normalization is performed to make the three fractions comparable in numerical range, allowing for linear weighted fusion. , These are the mean and standard deviation of the relative power within the current window, respectively. , These represent the mean and standard deviation of the relative power within the background window, respectively. , These represent the average radial variation characteristics within the current window and the background window, respectively. To prevent extremely small positive numbers with a denominator of zero; , and Let the weighting coefficients satisfy:

[0105] (17)

[0106] Preferably, it can be taken The weighting parameters can be set based on statistical analysis and manual calibration of typical electromagnetic interference data from X-band phased array weather radar.

[0107] The sample availability factor is used to reduce the significance when there are many missing or invalid samples. It is defined as follows:

[0108] (18)

[0109] in, and These represent the number of valid samples in the current window and the background window, respectively. and These represent the theoretical number of samples within the corresponding window.

[0110] Finally, based on the clutter feature saliency under different candidate window lengths, the optimal radial window length for the current range library is determined:

[0111] (19)

[0112] in, For the first The optimal radial window length corresponds to each distance library. In other words, within the candidate window length... In this process, the scale with the highest clutter feature significance is selected as the radial processing window of the current range library.

[0113] Selected optimal window length The corresponding maximum significance (Recommended) When the distance is 1.2, it indicates that the difference between the current distance library and the surrounding background is not significant, and the default window length is used in this case. This is to avoid frequent changes in window length due to noise fluctuations.

[0114] Candidate set of window lengths in this embodiment It can be preset according to the actual radar pulse repetition frequency and range resolution. In actual engineering implementation, the computational complexity can be controlled to the O(N) order of magnitude by updating statistical quantities such as mean and standard deviation through sliding cumulative updates, which meets the real-time processing requirements of phased array radar.

[0115] Using the above method, the window length is no longer directly determined by a fixed gradient threshold, but rather by a comprehensive consideration of multi-scale statistical characteristics such as local power abrupt changes, radial variation intensity, and the degree of dispersion within the window. When interference or clutter exhibits local abrupt changes, a smaller window is easier to select; when interference or clutter exhibits continuous distribution characteristics, a larger window can obtain more stable statistical results, thereby improving the reliability of subsequent adaptive threshold detection and impulse suppression.

[0116] Based on the above embodiments, the step of obtaining the adaptive judgment threshold in this embodiment includes:

[0117] The relative power of each pulse within the radial window is averaged and compressed to obtain a relative power sequence;

[0118] Calculate the mean and standard deviation of the relative power series;

[0119] The adaptive judgment threshold is determined based on the mean and standard deviation.

[0120] Along the radial distance, L centered on the current distance library i * (i) Within the window, the relative power of each pulse is averaged and compressed to obtain a one-dimensional relative power sequence P. r :

[0121] (20)

[0122] Based on a one-dimensional relative power sequence P r Statistical series mean and standard deviation :

[0123] (twenty one)

[0124] (twenty two)

[0125] Construct an adaptive discrimination threshold T Q :

[0126] (twenty three)

[0127] in, For coefficients, the preferred option is... =1.5, The parameters were obtained through statistical analysis and manual calibration based on typical electromagnetic interference data from X-band phased array weather radar.

[0128] When the result P of the relative power mean compression of the nth pulse is satisfied r (n) > T Q When the nth pulse is determined to be a discrete interference pulse, the abnormal pulse power is finally fitted and corrected by a quadratic interpolation algorithm to achieve the second level of fine interference suppression in the signal processing layer.

[0129] After two stages of interference suppression, the I / Q data undergoes conventional radar signal processing, including autocorrelation calculations, to obtain reflectivity factor data for suppressing pulse interference signals. This reflectivity factor data is then processed by an echo identification layer algorithm to further eliminate radial ray interference echoes.

[0130] Based on the above embodiments, the multidimensional characteristic parameters in this embodiment include multiple parameters such as near-far power difference, strong echo ratio, inter-pulse correlation coefficient, extended average power, radial gradient, and inter-column correlation coefficient.

[0131] To distinguish between radial ray interference echoes and normal meteorological echoes, the following characteristic parameters are used: near-far power difference (PD), strong echo proportion (SEP), inter-pulse correlation coefficient (PCC), extended average power (EAP), radial gradient (RG), and inter-column correlation coefficient (RCC), defined as follows:

[0132] First, the 0th-order delay autocorrelation function R(0) and the 1st-order delay autocorrelation function R(1) are obtained by performing autocorrelation calculation on the radar pulse group S:

[0133] (twenty four)

[0134] (25)

[0135] Where N is the number of pulses in the pulse group, s i and s i+1 These are the complex echo signals of the i-th and (i+1)-th pulses in pulse group S, respectively.

[0136] The signal power Z of a distance library can be further calculated, and its logarithmic form is as follows:

[0137] (26)

[0138] The power difference PD(j) between the near and far distances of the j-th distance reservoir is the power value Z(j) of the j-th distance reservoir and the average power of the last 10% of the distance reservoirs in its radial direction. The difference, in dB, is defined as follows:

[0139] (27)

[0140] (28)

[0141] Where R is the total number of distances along a single radial path.

[0142] Strong Echo Percentage (SEP) is a statistic used to measure the amount of effective signal power in the radial direction of a given range bank. It is calculated as follows: using 200 as the threshold for effective signal power, the ratio of the number of range banks with signal power exceeding this threshold in a single radial direction to the total number of range banks is calculated. The definition is as follows:

[0143] (29)

[0144] (30)

[0145] (31)

[0146] Where Val is the effective power threshold, Z(j) is the power value of the j-th distance in a single radial direction, and N e This is a flag indicating the effective power.

[0147] For the inter-pulse correlation coefficient PCC, first obtain the 0th-order delay autocorrelation function R(0) and the 1st-order delay autocorrelation function R(1), and calculate the ratio of the absolute value of R(1) to R(0) to obtain the PCC.

[0148] (32)

[0149] For the extended average power EAP, calculate the current distance library Z. r The radar detects echo points extending five distances into the distance and calculates their average power.

[0150] (33)

[0151] Where r is the index of the current distance database, Z r+j Let be the power of the (r+j)th distance library.

[0152] The radial gradient RG is used to calculate the power difference with respect to the radial direction of the distance library, defined with a radial step size of Δr. This reflects the radial variation trend.

[0153] (34)

[0154] Where r is the current distance library number, R is the total number of distance libraries in a single radial direction, Z is the signal power of each distance library, and the subscript of Z is the distance library number.

[0155] Introduce the Coogaus weighted coefficient w, which represents the distance between two points on the left and right. k Construct the smoothed local gradient:

[0156] (35)

[0157] In the formula: the weighting coefficients satisfy w -2 =w2=0.15, w -1 =w1=0.35, w0=0.5, strengthen the center distance library gradient weights. The above parameters are obtained through statistical analysis and manual calibration based on typical electromagnetic interference data from X-band phased array weather radar.

[0158] Take the absolute value of all smooth gradients along the entire radial direction ( ) and root mean square (RMS) average, eliminating the problem of positive and negative gradient cancellation, serve as global radial gradient features:

[0159] (36)

[0160] The inter-column correlation coefficient RCC, for the current echo's radial range, is the radial power Z of the (R-1) range columns prior to the radar center. x The radial power Z of the next (R-1) distances after the distance of one distance is... y Calculate the inter-liquid correlation coefficient along the entire radial direction:

[0161] (37)

[0162] Where Cov represents the covariance operation. For Z x standard deviation For Z y The standard deviation.

[0163] Based on collected field and simulated radar data, a database containing radial ray interference and normal meteorological echoes was established through expert judgment and labeling. Based on the extraction of near-far power difference (PD), strong echo proportion (SEP), inter-pulse correlation coefficient (PCC), extended average power (EAP), radial gradient (RG), and inter-column correlation coefficient (RCC), a database of characteristic parameters for radial ray interference and normal meteorological echoes was further obtained. Then, probability distribution maps of each characteristic parameter in positive and negative samples were obtained. Based on the probability distributions of radial ray interference echoes and meteorological echoes, the triangular membership functions of each characteristic parameter were determined.

[0164] For radar echo data used for testing, the extracted feature parameters are fuzzified using a triangular membership function to obtain 0-1 value criteria for each feature parameter for different types of echoes. The criteria values ​​are then weighted and accumulated, and if they exceed a preset threshold, they are judged as radial ray interference echoes and discarded.

[0165] The electromagnetic interference suppression algorithm proposed in this embodiment adopts a two-level progressive architecture of signal processing layer and echo identification layer to achieve layered governance and step-by-step filtering of different types of radar electromagnetic interference. At the bottom signal processing level, hierarchical suppression of sudden interference pulses and discrete interference pulses is completed successively. Sudden spike interference is quickly corrected by comparing the amplitude of adjacent periods. Then, combined with the radial gradient calculation of pulse groups and the adaptive window strategy, fine correction of high duty cycle continuous interference pulses is achieved, weakening pulse-level electromagnetic interference clutter from the source of the original I / Q signal. After completing the above steps, the echo identification layer completes the accurate identification and elimination of radial interference echoes based on the multi-dimensional interference characteristics of reflectivity factor and combined with fuzzy logic discrimination method. The two-layer algorithm performs its own function and complements each other. It can effectively suppress local abnormal interference at the pulse scale in the time domain and accurately eliminate structural electromagnetic interference clutter in the two-dimensional echo domain. It is fully adaptable to typical electromagnetic interference forms such as spiral band interference, strip-shaped pockmark interference and radial ray interference, significantly improving the quality and effectiveness of weather radar original detection signals and echo data.

[0166] Electromagnetic interference suppression experiments were conducted on phased array weather radar data using the method proposed in this invention. Figure 3 Part (a) of the diagram shows a PPI (Plan Position Indicator, Reflectivity Factor) map of electromagnetic interference echoes according to an embodiment of the present invention. The echo distribution detected by radar indicates the presence of various types of electromagnetic interference echoes around the station, manifested as large-amplitude radial ray interference, a small amount of strip-shaped speckle interference, and spiral band interference echoes. To verify the effectiveness of the present invention, two existing methods are compared: Method 1 is the interference filter in the Vaisala radar signal processor from Finland. Its basic principle is to compare the power (dB value) of three consecutive pulses in the pulse domain. When the power reaches the corresponding determination formula, it is determined to be an interference pulse and replaced by the previous pulse. This method is effective for isolated pulse-type interference, but its ability to handle continuous strip-shaped speckle interference and radial ray interference dependent on spatial structure features is limited. Method 2 is a radial fuzzy logic algorithm. This type of method extracts features such as radial gradients and combines them with fuzzy logic for interference discrimination. It is mainly designed for radial ray interference, but its suppression of pulse interference in the pulse domain is insufficient.

[0167] Figure 3 Parts (b) and (c) are the reflectivity factor PPI diagrams after suppression by the radial fuzzy logic algorithm and the interference filter algorithm in the Finnish Vaisala radar signal processor, respectively. It can be seen that both methods have electromagnetic interference residual problems to varying degrees.

[0168] After being processed by the layered progressive processing method of the present invention, as follows Figure 3In part (d), large-area electromagnetic interference is effectively removed, while the original meteorological echoes are effectively preserved. This verifies the superiority of the layered and progressive architecture of this invention, which combines "signal processing layer suppressing pulse-level interference + echo identification layer removing structural interference," compared to single-layer methods.

[0169] The electromagnetic interference suppression device for X-band phased array weather radar provided by the present invention is described below. The electromagnetic interference suppression device for X-band phased array weather radar described below can be referred to in correspondence with the electromagnetic interference suppression method for X-band phased array weather radar described above.

[0170] like Figure 4 As shown, the device includes a correction module 401, a processing module 402, an extraction module 403, and a rejection module 404, wherein:

[0171] The correction module 401 is used to identify abnormal interference pulses from the original pulse domain signal acquired by the X-band phased array weather radar and to perform interpolation correction on the abnormal interference pulses.

[0172] Processing module 402 is used to process the interpolated pulse domain original signal to obtain reflectivity factor data as radar echo data;

[0173] Extraction module 403 is used to extract multidimensional feature parameters from the radar echo data;

[0174] The elimination module 404 is used to perform fuzzification processing on each feature parameter using a fuzzy logic discrimination algorithm. If the radar echo data is determined to be a radial ray interference echo based on the fuzzification processing result, the radar echo data is eliminated.

[0175] This embodiment proposes an electromagnetic interference suppression method for X-band phased array weather radar based on hierarchical progressive processing. First, at the signal processing layer, helical band interference and strip-shaped speckle interference are processed by pulse interpolation. Then, at the echo identification layer, the power difference between near and far distances, the proportion of strong echoes, the inter-pulse correlation coefficient, the extended average power, the radial gradient, and the inter-column correlation coefficient are extracted. Fuzzy logic is used to identify and eliminate radial ray interference echoes, thereby achieving accurate identification and suppression of various electromagnetic interferences.

[0176] 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 of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for suppressing electromagnetic interference in an X-band phased array weather radar, characterized in that, include: Abnormal interference pulses are identified from the raw pulse domain signals acquired by the X-band phased array weather radar, and the abnormal interference pulses are corrected by interpolation. The original pulse domain signal after interpolation correction is processed to obtain reflectivity factor data as radar echo data; Extract multidimensional feature parameters from the radar echo data; The fuzzy logic discrimination algorithm is used to fuzzify each feature parameter. If the radar echo data is determined to be radial ray interference echo based on the fuzzification result, the radar echo data is discarded. The abnormal interference pulses also include discrete interference pulses. Abnormal interference pulses are identified from the raw pulse domain signal acquired by the X-band phased array weather radar, and interpolation corrections are performed on the abnormal interference pulses, including: Calculate the clutter feature saliency of each distance library under different candidate window lengths; The optimal window length for each distance library is determined based on the clutter feature salience of each distance library under different candidate window lengths. Construct radial windows centered on each distance library and the optimal window length, and count the proportion of sudden interference pulses within the radial windows; When the proportion is greater than or equal to the second preset threshold, the relative power of each pulse in the radial window is averaged and compressed, and then compared with the adaptive judgment threshold. Based on the comparison result, it is determined whether each pulse is a discrete interference pulse. The power of the discrete interference pulses is fitted and corrected using a quadratic interpolation algorithm; Calculate the clutter feature saliency of each distance library under different candidate window lengths, including: Construct a current radial window centered on each distance library based on each distance library and each candidate window length; Calculate the mean and standard deviation of the relative power of each pulse within the current radial window; Calculate the mean and standard deviation of the relative power of each pulse within the neighborhood background window of the current radial window; Calculate the average radial variation characteristics within the current radial window and the neighboring background window of the current radial window, respectively; The clutter feature significance is calculated based on the mean and standard deviation of the relative power of each pulse in the current radial window, the mean and standard deviation of the relative power of each pulse in the neighboring background window, and the average radial variation characteristics in the current radial window and the neighboring background window. Calculate the average radial variation features within the current radial window and its neighboring background window, including: Calculate the absolute value of the difference between the relative power of each pulse in the current radial window at the r-th and (r+1)-th distances, and average the absolute values ​​of all pulses in the current radial window to obtain the average radial variation characteristic in the current radial window. Calculate the absolute value of the difference between the relative power of each pulse in the neighborhood background window of the current radial window at the r-th and r+1-th distances. Take the average of the absolute values ​​corresponding to all pulses in the neighborhood background window as the average radial variation feature in the neighborhood background window of the current radial window.

2. The electromagnetic interference suppression method for X-band phased array weather radar according to claim 1, characterized in that, The abnormal interference pulses include sudden interference pulses. Abnormal interference pulses are identified from the raw pulse domain signal acquired by the X-band phased array weather radar, and interpolation correction is performed on the abnormal interference pulses, including: Obtain pulse groups from the same distance library from the original pulse domain signal; Calculate the ratio between the amplitude of each pulse in the pulse group acquired in the current scan cycle and the amplitude of the normal pulse in the pulse group acquired in the previous scan cycle. If the ratio is greater than a first preset threshold, the pulse is determined to be a sudden interference pulse; otherwise, the pulse is determined to be a normal pulse. The amplitude of the sudden interference pulse is replaced and completed using the amplitude of the normal pulse in the pulse group acquired in the previous scan cycle.

3. The electromagnetic interference suppression method for X-band phased array weather radar according to claim 1, characterized in that, The steps for obtaining the relative power of each pulse include: Calculate the power of each pulse in the pulse group of each distance library; Select a preset number of pulses with the lowest power from the pulse group, and calculate the average power of the selected pulses as the local base noise; The ratio between the power of each pulse and the local base noise is used as the relative power.

4. The electromagnetic interference suppression method for X-band phased array weather radar according to claim 1, characterized in that, The steps for obtaining the adaptive judgment threshold include: The relative power of each pulse within the radial window is averaged and compressed to obtain a relative power sequence; Calculate the mean and standard deviation of the relative power series; The adaptive judgment threshold is determined based on the mean and standard deviation.

5. The electromagnetic interference suppression method for X-band phased array weather radar according to claim 1, characterized in that, The multidimensional characteristic parameters include multiple parameters such as near-far power difference, strong echo ratio, inter-pulse correlation coefficient, extended average power, radial gradient, and inter-column correlation coefficient.

6. An electromagnetic interference suppression device for X-band phased array weather radar, characterized in that, The electromagnetic interference suppression method for X-band phased array weather radar according to any one of claims 1-5 includes: The correction module is used to identify abnormal interference pulses from the raw pulse domain signals acquired by the X-band phased array weather radar and to perform interpolation correction on the abnormal interference pulses. The processing module is used to process the interpolated pulse domain original signal to obtain reflectivity factor data as radar echo data. The extraction module is used to extract multi-dimensional feature parameters from the radar echo data; The elimination module is used to perform fuzzification processing on each feature parameter using a fuzzy logic discrimination algorithm. If the radar echo data is determined to be a radial ray interference echo based on the fuzzification processing result, the radar echo data is eliminated.

Images