Watson-watt direction finder amplitude-phase calibration method and device

Abstract

The application discloses a Watson-Watt direction-finder amplitude-phase calibration method and device, constructs a double-channel amplitude-phase coupling deviation reference spectrum through amplitude-phase deviation extraction and drift direction classification; establishes a channel deviation coupling relationship spectrum by jointly tracing the temperature drift and frequency dependence of each deviation factor, generates an overshoot risk mark by combining the compensation direction flip risk quantization; based on the channel deviation coupling relationship spectrum and the overshoot risk mark, carry out calibration residual convergence evaluation, determine the deviation level and identify the blocking factor set; according to the cross-frequency transfer path of the blocking factor set, complete the frequency range classification weight distribution, obtain the deviation value of each frequency range, and determine the deviation distribution type through polarity flip detection and asymmetry grading; finally, complete the cross-frequency range amplitude-phase linkage compensation amount solution and output the amplitude-phase calibration compensation parameters combined with the deviation level, realize the system identification and linkage compensation of the double-channel amplitude-phase deviation in the wide frequency range working scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of radio direction finding technology, in particular to a Watson-Watt direction finder amplitude-phase calibration method and device. BACKGROUND

[0002] The Watson-Watt direction finder relies on the amplitude-phase consistency of the sine channel and the cosine channel to realize azimuth angle calculation. In actual engineering, the amplitude-phase inconsistency error of the two channels is caused by device processing deviation, transmission path difference and working environment change. This error presents complex nonlinear distribution characteristics with the change of frequency and temperature, which directly affects the azimuth estimation accuracy of the direction finder. The existing calibration method usually performs static correction on the amplitude-phase deviation of the double channel with a fixed compensation value, which cannot accurately reflect the dynamic characteristics of the deviation with the change of frequency, and the calibration residual error is large in the wide frequency band working scene.

[0003] The formation mechanism of the double-channel amplitude-phase deviation involves the cross coupling of temperature drift and frequency dependence. Part of the deviation factors appears compensation overshoot in the iterative compensation process due to nonlinear response, resulting in secondary rebound of the residual error after the first convergence. The existing method lacks systematic identification and processing ability for the above coupling characteristics and convergence blocking phenomenon. The deviation between different frequency bands exists in the cross-band transmission path in the wide frequency band multi-working condition. The mutual interference of cross-band deviation components cannot be eliminated by single-band independent compensation, and the overall calibration accuracy is restricted. An intelligent calibration method is needed to solve at least one of the above problems. SUMMARY

[0004] The present application discloses a Watson-Watt direction finder amplitude-phase calibration method and device, which aims to identify the deviation factors of the cross coupling of temperature drift and frequency dependence by constructing a double-channel amplitude-phase coupling deviation reference spectrum, evaluate the overshoot risk and convergence blocking characteristics in the iterative compensation process, and then classify the blocking factor set by frequency band and identify the deviation polarity reversal distribution pattern. Finally, the cross-band amplitude-phase interactive compensation amount is calculated combined with the deviation distribution type and the deviation level, the amplitude-phase calibration compensation parameters facing the wide frequency band working scene are output, and the accuracy and reliability of the double-channel amplitude-phase calibration of the direction finder are improved.

[0005] The first aspect of the present application proposes a Watson-Watt direction finder amplitude-phase calibration method, comprising the following steps: Obtain known direction reference signal data and double-channel receiver response data, and construct a double-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the double-channel receiver response data; Perform deviation factor extraction on the double-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set, perform temperature drift and frequency dependence joint tracing on the candidate deviation factor set to establish a channel deviation coupling relationship spectrum, and perform single compensation overshoot risk evaluation on the candidate deviation factor set to generate an overshoot risk label; Based on the channel deviation coupling spectrum and the overshoot risk marker, a deviation classification parameter is generated by real-time calibration residual convergence evaluation. The deviation level is determined according to the deviation classification parameter. The calibration convergence blocking factors are identified from the deviation classification parameter to generate a blocking factor set. The set of blocking factors is divided into frequency bands and the deviation is classified to obtain the deviation value of each frequency band. The amplitude and phase deviation polarity reversal detection is performed on the deviation value of each frequency band to identify the deviation polarity reversal frequency point. The distribution pattern recognition is performed on the deviation polarity reversal frequency point and the deviation value of each frequency band to obtain the deviation distribution type. Based on the deviation distribution type and the deviation polarity reversal frequency point, the cross-band amplitude and phase linkage compensation amount is calculated to generate the frequency band deviation distribution characteristics. The frequency band deviation distribution characteristics and the deviation level are combined to output the amplitude and phase calibration compensation parameters.

[0006] A second aspect of the present invention provides a Watson-Watt direction finder amplitude and phase calibration device, comprising: The reference construction module is used to acquire known direction reference signal data and dual-channel receiver response data, and construct a dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the dual-channel receiver response data; The source tracing analysis module is used to extract deviation factors from the dual-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set, perform joint source tracing of temperature drift and frequency dependence on the candidate deviation factor set to establish a channel deviation coupling relationship spectrum, and perform single-compensation overshoot risk assessment on the candidate deviation factor set to generate an overshoot risk label. The convergence evaluation module is used to perform real-time calibration residual convergence evaluation based on the channel deviation coupling relationship spectrum and the overshoot risk marker to generate deviation classification parameters, determine the deviation level according to the deviation classification parameters, and identify calibration convergence blocking factors from the deviation classification parameters to generate a blocking factor set. The frequency band analysis module is used to classify the deviation of the blocking factor set by frequency band partition to obtain the deviation value of each frequency band, perform amplitude and phase deviation polarity reversal detection on the deviation value of each frequency band to identify the deviation polarity reversal frequency point, and perform distribution pattern recognition based on the deviation polarity reversal frequency point and the deviation value of each frequency band to obtain the deviation distribution type. The calibration output module is used to calculate the cross-band amplitude and phase linkage compensation amount based on the deviation distribution type and the deviation polarity reversal frequency point to generate the frequency band deviation distribution characteristics, and output the amplitude and phase calibration compensation parameters by combining the frequency band deviation distribution characteristics and the deviation level.

[0007] The beneficial effects of this invention are reflected in the following points: 1. By extracting the amplitude and phase deviation of the dual-channel receiver response data and the known direction reference signal data at frequency points, a dual-channel amplitude and phase coupling deviation benchmark spectrum containing common-mode and differential-mode drift classification information is constructed. On this basis, temperature drift sensitivity detection and frequency scanning response analysis are carried out on each deviation factor. The channel deviation coupling relationship spectrum is established through cross-correlation strength evaluation, and the nonlinear slope and compensation direction reversal risk of each factor in iterative compensation are quantitatively labeled, realizing the systematic tracing and risk prediction of the source of dual-channel amplitude and phase deviation. 2. Based on the network transmission structure of the channel deviation coupling relationship spectrum and the amplitude limiting constraint of the overshoot risk label, the convergence trajectory of the iterative compensation residual of each deviation factor is tracked in real time. By identifying the residual secondary recovery segment and combining it with the amplitude and phase convergence rate difference sequence to screen mismatch factors, the recovery ratio is calculated to complete the deviation severity ranking, thereby determining the overall deviation level and identifying the blocking factors that restrict calibration convergence, establishing the ability to quantitatively identify and grade convergence blocking phenomena. 3. By analyzing the inter-band amplitude-phase coupling transmission path of the blocking factor set, a transmission path diagram is generated. Based on the transmission chain length and effective transmission strength, cross-band classification weight allocation is completed. After obtaining the deviation values ​​of each frequency band, the deviation distribution type is determined by polarity reversal detection and asymmetry classification. Based on the deviation distribution type, a centralized compensation or uniform distribution strategy is selected to complete the calculation of cross-band amplitude-phase linkage compensation. Based on the deviation level, the compensation application rhythm output amplitude-phase calibration compensation parameters are determined, realizing the adaptive matching of cross-band deviation transmission characteristics and compensation strategy. Attached Figure Description

[0008] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0009] Figure 1 This is a schematic flowchart of a Watson-Watt direction finder amplitude and phase calibration method according to the present invention.

[0010] Figure 2 This is a structural block diagram of a Watson-Watt direction finder amplitude and phase calibration device according to the present invention. Detailed Implementation

[0011] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0012] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0013] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0014] The technical solutions of the embodiments of this application are described below.

[0015] like Figure 1 As shown, this embodiment of the invention provides a Watson-Watt direction finder amplitude and phase calibration method, including the following steps S110-S150: Step S110: Obtain known direction reference signal data and dual-channel receiver response data, and construct a dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and dual-channel receiver response data.

[0016] Specifically, known direction reference signal data and dual-channel receiver response data are acquired. The known direction reference signal data originates from a precisely calibrated signal source output at a specific azimuth angle. The azimuth calibration accuracy of the reference signal data directly determines the accuracy of subsequent amplitude and phase deviation benchmark calculations. Reference signal data with azimuth calibration errors exceeding the tolerance are discarded during the acquisition stage. Dual-channel receiver response data is acquired synchronously through the sine and cosine channels of a Watson-Watt direction finder. Both channels are acquired synchronously under the same reference signal excitation. Due to the different lengths of their respective RF links, the sine and cosine channels have a nanosecond-level acquisition delay difference. If alignment compensation is not performed during the acquisition stage, the dual-channel response data at the same frequency will correspond to signal states at different times, leading to spurious phase differences in subsequent amplitude and phase deviation calculations. The frequency coverage of the known direction reference signal data must match the operating frequency band of the dual-channel receiver response data. If the output power of the reference signal data is unstable at a certain frequency, causing abnormal amplitude in the dual-channel receiver response data at that frequency, the validity markers of both types of data at that frequency are updated synchronously during the acquisition stage. In wideband calibration scenarios, the known direction reference signal data is acquired in the low-frequency band after completing omnidirectional angular scanning. The dual-channel receiver response data is then refreshed segment by segment as the reference signal data's frequency band changes. The frequency band correspondence between the reference signal data and the dual-channel receiver response data is recorded frame-by-frame using frequency band identifiers during the acquisition phase to prevent misalignment of cross-frequency band entries in deviation calculations. Invalid markers for channel fault frames are written synchronously with the corresponding frequency point's reference signal data entry during the acquisition phase to prevent subsequent deviation calculations from mistakenly including fault frame response values.

[0017] In some embodiments, constructing a dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the dual-channel receiver response data includes: extracting theoretical amplitude-phase responses of known directions from the reference signal data to generate a theoretical response reference; calculating amplitude deviation and phase deviation on the dual-channel receiver response data based on the theoretical response reference to generate amplitude-phase deviation pair sequences; performing joint drift direction classification on the amplitude-phase deviation pair sequences to identify common-mode drift segments and differential-mode drift segments to generate a drift direction classification spectrum; and fusing the drift direction classification spectrum with the amplitude-phase deviation pair sequences to construct a dual-channel amplitude-phase coupling deviation reference spectrum.

[0018] The theoretical amplitude and phase responses in known directions are extracted from the reference signal data to generate a theoretical response benchmark. The signal parameters at each calibrated azimuth angle in the reference signal data are used to calculate the theoretical amplitude and phase responses of the sine and cosine channels based on the Watson-Watt direction finding principle. The theoretical amplitude response is calculated analytically using the sine and cosine values ​​of the calibrated azimuth angle, while the theoretical phase response is calculated based on the phase relationship of the two channels under ideal, unbiased conditions. Both types of theoretical responses are calculated azimuth-by-azimuth and frequency-by-frequency and then written into the theoretical response benchmark. The theoretical response benchmark stores the theoretical amplitude and phase response values ​​of the sine and cosine channels using azimuth and frequency identifiers as dual indexes. Entries with higher azimuth calibration accuracy in the reference signal data have stronger confidence in the corresponding theoretical amplitude and phase response values ​​in the theoretical response benchmark. During wideband calibration, the theoretical amplitude response of the reference signal data in the high-frequency band deviates from the analytical pattern of the low-frequency band due to the distortion of the sine and cosine channel pattern. A pattern correction term is introduced at the corresponding frequency point in the high-frequency band to compensate for the analytical deviation. The continuity of the theoretical response reference generated from the low-frequency band entries and the theoretical response reference generated from the high-frequency band entries in the reference signal data in the frequency dimension is ensured by the correction term. The introduction of the high-frequency band pattern correction term compensates for the analytical deviation caused by the distortion of the sine and cosine channel pattern, maintaining the continuity of the reference in the frequency dimension between the low-frequency and high-frequency bands.

[0019] Amplitude and phase deviation pairs are generated by calculating the amplitude and phase deviations of the dual-channel receiver response data based on the theoretical response benchmark. The measured amplitudes and phases of the sine and cosine channels in the dual-channel receiver response data are successively subtracted from the theoretical amplitude and phase response values ​​at the corresponding azimuth and frequency points in the theoretical response benchmark. The amplitude deviation is characterized by the difference between the measured amplitude of the dual-channel receiver response data and the corresponding theoretical amplitude in the theoretical response benchmark, and the phase deviation is characterized by the difference between the measured phase of the dual-channel receiver response data and the corresponding theoretical phase in the theoretical response benchmark. Both types of deviation values ​​are written into the amplitude and phase deviation pair sequences using a triple index of channel identifier, azimuth identifier, and frequency point identifier. The amplitude deviation is expressed in decibels, and the phase deviation is expressed in degrees. Amplitude and phase deviation pair sequence entries generated at the same frequency point under multiple calibrated azimuth angles are stored independently. Subsequent deviation factor extraction uses the statistical distribution of all azimuth angle entries as input rather than a single azimuth angle result, avoiding the influence of single-point measurement deviations on the representativeness of the deviation characteristics. When a direction finder was calibrated at a frequency of 1.2 GHz, the measured amplitude of the sinusoidal channel was 0.6 dB higher than the theoretical amplitude corresponding to the theoretical response benchmark. The amplitude-phase deviation recorded for the sequence at this frequency was +0.6 dB. The measured phase deviation of the cosine channel at the same frequency was 4.3° behind the theoretical value, and the amplitude-phase deviation recorded for the sequence was -4.3°. The channel identifier, azimuth identifier, and frequency identifier constitute a triple index. The amplitude is expressed in decibels, and the phase is expressed in degrees. Each deviation entry can be uniquely traced to the corresponding difference between the measured and theoretical responses.

[0020] A joint drift direction classification is performed on the amplitude-phase deviation pair sequence to identify common-mode drift segments and differential-mode drift segments, generating a drift direction classification spectrum. The joint drift direction classification uses the sign consistency of the sinusoidal and cosine channel amplitude deviation changes at adjacent frequency points in the amplitude-phase deviation pair sequence as the amplitude drift criterion, and the sign consistency of the sinusoidal and cosine channel phase deviation changes as the phase drift criterion. Frequency bands where both amplitude and phase drift criteria have the same sign are labeled as common-mode drift segments, and frequency bands with opposite signs for either criterion are labeled as differential-mode drift segments. The drift direction classification spectrum records the drift direction type label for each frequency point using the frequency point identifier as an index. When a direction finder is calibrated in the 400MHz to 600MHz frequency band, the amplitude and phase deviations of the sequence in this band show a monotonically increasing trend in both sine and cosine channel amplitude deviations, and a trend of simultaneous deflection in both sine and cosine channel phase deviations. The amplitude drift criterion and the phase drift criterion both have the same sign, and the drift direction classification spectrum labels this frequency band as a common-mode drift segment. In the 600MHz to 800MHz frequency band, the amplitude and phase deviations of the sequence show an increasing trend in sine channel amplitude deviation and a decreasing trend in cosine channel amplitude deviation. The amplitude drift criterion has the opposite sign, and the drift direction classification spectrum labels this frequency band as a differential-mode drift segment. Both amplitude and phase drift criteria must satisfy sign consistency to be classified into the common-mode segment; if either criterion has the opposite sign, it is classified into the differential-mode segment. This dual constraint reduces the probability of misclassification under a single criterion.

[0021] A dual-channel amplitude-phase coupling deviation reference spectrum is constructed by fusing the drift direction classification spectrum with the amplitude-phase deviation pair sequence. The drift direction type label of the drift direction classification spectrum provides a classification label for the physical interpretation of the deviations at each frequency point in the amplitude-phase deviation pair sequence. The fusion process of the dual-channel amplitude-phase coupling deviation reference spectrum uses the type label of each frequency point in the drift direction classification spectrum as the classification key. The amplitude-phase deviation values ​​of the sine and cosine channels at the corresponding frequency points of the amplitude-phase deviation pair sequence are categorized into common-mode and differential-mode labels and then merged into the corresponding partitions of the dual-channel amplitude-phase coupling deviation reference spectrum. The dual-channel amplitude-phase coupling deviation reference spectrum uses the frequency point identifier as the row index and consists of multiple columns of fields, including drift direction type, sine channel amplitude deviation, sine channel phase deviation, cosine channel amplitude deviation, and cosine channel phase deviation, to form a complete entry. The structural division into common-mode and differential-mode partitions allows for the extraction of deviation factors using different tracing paths for different drift types. Common-mode partition entries point to device or transmission path errors affecting dual-channel synchronization, while differential-mode partition entries point to the sources of dual-channel response asymmetry. The coexistence of these two types of partitions in the reference spectrum ensures that the deviation factor set covers both single and overlapping driving sources. After the full-band calibration of a certain direction finder, the sine and cosine channel amplitude deviations of the 400MHz to 600MHz common-mode partition in the dual-channel amplitude-phase coupling deviation reference spectrum are both positive and in the same direction. The sine channel amplitude deviation of the 600MHz to 800MHz differential-mode partition is positive while the cosine channel amplitude deviation is negative. The difference in the deviation sign distribution of the two types of partitions in the dual-channel amplitude-phase coupling deviation reference spectrum clearly distinguishes the frequency band contribution range of the common-mode and differential-mode error sources.

[0022] Step S120: Extract deviation factors from the dual-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set; perform joint tracing of temperature drift and frequency dependence on the candidate deviation factor set to establish a channel deviation coupling relationship spectrum; and perform a single-compensation overshoot risk assessment on the candidate deviation factor set to generate an overshoot risk marker.

[0023] Specifically, deviation factor extraction is performed on the dual-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set. Deviation factor extraction identifies driving characteristics from the deviation distribution pattern of each partition of the dual-channel amplitude-phase coupling deviation reference spectrum. Continuous frequency bands with consistent pattern characteristics are grouped into a class of candidate deviation factors—that is, the device or path error source driving the dual-channel amplitude-phase deviation in that frequency band—and written into the candidate deviation factor set. The candidate deviation factor set uses factor identifiers as row indexes and records the frequency band range, drift direction type, and corresponding amplitude-phase deviation amplitude characteristics of each candidate deviation factor. The candidate deviation factors in the common-mode partition and differential-mode partition are respectively labeled with the source partition type. After the full-band calibration of a direction finder was completed, the amplitude deviation of the sinusoidal channel in the common-mode region (300MHz to 500MHz) of the dual-channel amplitude-phase coupling deviation reference spectrum continuously increased with increasing frequency and the phase deflected in the same direction. Deviation factor extraction classified this frequency band as a monotonic drift-type candidate deviation factor, and its source was determined to be the inconsistency in the transmission path accumulated with increasing operating frequency. The amplitude deviation of the sinusoidal channel in the adjacent differential-mode region (500MHz to 700MHz) exhibited alternating positive and negative periodic fluctuations, and was classified as a periodic fluctuation-type candidate deviation factor. Its source was determined to be the amplitude-phase oscillation caused by the resonant characteristics of a certain device in this frequency band. The monotonic drift-type factor deviation accumulated unidirectionally with temperature, while the periodic fluctuation-type factor deviation exhibited frequency-periodic oscillations. The physical causes of these two types are different, and temperature drift tracing and frequency dependence analysis employed different identification paths for them.

[0024] In some embodiments, the step of establishing a channel deviation coupling spectrum by jointly tracing the temperature drift and frequency dependence of the candidate deviation factor set includes: performing temperature sensitivity detection on each factor in the candidate deviation factor set to generate a temperature drift sensitive factor set; performing frequency scan response analysis on each factor in the candidate deviation factor set to generate a frequency dependent factor set; performing cross-correlation strength evaluation on the temperature drift sensitive factor set and the frequency dependent factor set to generate coupling strength weights; and establishing a channel deviation coupling spectrum based on the coupling strength weights.

[0025] Temperature sensitivity testing is performed on each factor in the candidate deviation factor set to generate a temperature drift sensitive factor set. Temperature sensitivity testing collects the amplitude and phase deviation changes of each factor in the candidate deviation factor set under different ambient temperature conditions. The difference in deviation amplitude of the same candidate deviation factor under high and low temperature test conditions is compared to assess the factor's response strength to temperature changes. Temperature sensitivity testing performs temperature drift quantification on each factor in the candidate deviation factor set in both the amplitude and phase dimensions. The amplitude temperature drift coefficient is defined as the change in amplitude deviation caused by a unit temperature change (dB / °C), and the phase temperature drift coefficient is defined as the change in phase deviation caused by a unit temperature change (° / °C). Candidate deviation factors whose temperature drift coefficients exceed the preset sensitivity threshold are identified as temperature drift sensitive factors and added to the temperature drift sensitive factor set. In the candidate bias factor set, a monotonic drift factor exhibits an amplitude deviation change of 0.8 dB and a phase deviation change of 2.1° during a temperature rise from 25°C to 55°C. Both its amplitude and phase temperature drift coefficients exceed the sensitivity threshold. Temperature sensitivity testing identifies this factor as a temperature drift sensitive factor and adds it to the temperature drift sensitive factor set. Another periodic fluctuation factor, however, shows amplitude and phase deviation changes below the sensitivity threshold within the same temperature range. The difference in temperature drift sensitivity between these two types of factors will lead to a significant divergence in the coupling strength weighting during cross-correlation strength assessment. Both amplitude and phase temperature drift coefficients must exceed the sensitivity threshold simultaneously for inclusion in the set; factors exceeding the threshold in only one dimension are excluded to prevent the one-sidedness of temperature response from affecting the accuracy of cross-correlation strength assessment.

[0026] For example, the step of performing frequency scanning response analysis on each factor in the candidate deviation factor set to generate a frequency-dependent factor set includes: performing wideband amplitude and phase response simulation on each factor in the candidate deviation factor set to generate a simulation response spectrum; performing amplitude and phase abrupt change synchronization detection on each frequency point in the simulation response spectrum to generate a synchronization change frequency point set; performing synchronization change intensity evaluation on the synchronization change frequency point set to generate an amplitude and phase synchronization change intensity sequence; and selecting factors with high synchronization change intensity based on the amplitude and phase synchronization change intensity sequence to form a frequency-dependent factor set.

[0027] Wideband amplitude and phase response simulations are performed on each factor in the candidate deviation factor set to generate simulation response spectra. Based on the circuit topology parameters and device models of each factor in the candidate deviation factor set, the wideband amplitude and phase response simulations calculate the amplitude and phase responses of each factor to sinusoidal and cosine channels point-by-point across the full frequency range covering the operating frequency band of the dual-channel receiver. The simulation results are written into the simulation response spectrum factor-by-factor and frequency-by-frequency point. The simulation response spectrum records the amplitude and phase response values ​​of each factor at each frequency point using factor identifiers and frequency point identifiers as dual indices. Factors with more precise device parameters in the candidate deviation factor set show a higher degree of agreement between the simulation results and measured deviations in the simulation response spectrum. The frequency resolution of the simulation response spectrum is adaptively configured according to the frequency band range of each factor in the candidate deviation factor set. When the device resonant frequency of a factor in the candidate deviation factor set falls within the operating frequency band, the simulation response spectrum exhibits a sharp drop in amplitude and a rapid change in phase near the corresponding resonant frequency of that factor. The simulation response spectra of adjacent non-resonant factors show smooth amplitude and phase responses within the same frequency range. The difference in response patterns between the two types of factors in the simulation response spectrum provides direct frequency characteristic input for subsequent abrupt synchronization detection. The sudden drop in amplitude and abrupt change in phase at the device's resonant frequency make the simulated curve of the resonant factor significantly different from that of the smooth non-resonant region. As a result, the difference between the two types of response modes in the detection of amplitude and phase change synchronization is more obvious.

[0028] A synchronous mutation frequency set is generated by detecting the synchronicity of amplitude and phase abrupt changes at each frequency point in the simulated response spectrum. The amplitude and phase responses of each factor in the simulated response spectrum are differentially analyzed along the frequency axis to form amplitude and phase change sequences. Frequency points in the amplitude change sequence whose absolute values ​​exceed the amplitude mutation threshold are identified as amplitude mutation frequencies, and frequency points in the phase change sequence whose absolute values ​​exceed the phase mutation threshold are identified as phase mutation frequencies. Frequency points that simultaneously satisfy both amplitude and phase mutation conditions are identified as synchronous mutation frequencies and added to the synchronous mutation frequency set. The synchronous mutation frequency set records the synchronous mutation positions of each factor using both factor identifiers and frequency identifiers as dual indices. Factors with significant resonance characteristics in the simulated response spectrum have a higher entry density in the synchronous mutation frequency set than non-resonant factors. The entry density of the synchronous mutation frequency set directly reflects the concentration of frequency dependence characteristics of each factor in the simulated response spectrum. In the simulated response spectrum, the absolute value of the amplitude change sequence of a certain factor near the resonant frequency reaches 3.2 dB / MHz, and the absolute value of the phase change sequence simultaneously reaches 18° / MHz, both exceeding the corresponding abrupt change threshold. The set of synchronous abrupt change frequency points is written into the entry for that frequency point. The absolute values ​​of the amplitude and phase change sequences in adjacent smooth frequency bands are all below the threshold, thus clearly locating the concentrated distribution of the frequency dependence characteristics within the operating frequency band. The higher the entry density of a factor, the more concentrated its frequency dependence characteristics are within the operating frequency band. The difference in entry density directly reflects the strength of the sensitivity of each factor to frequency changes.

[0029] Synchronous mutation intensity assessment is performed on the synchronous mutation frequency point set to generate an amplitude-phase synchronous mutation intensity sequence. Synchronous mutation frequencies of each factor in the synchronous mutation frequency point set exhibit different mutation intensities in terms of amplitude and phase changes. The amplitude-phase synchronous mutation intensity assessment uses the normalized product of the absolute values ​​of the amplitude change sequence and the absolute values ​​of the phase change sequence corresponding to each entry in the synchronous mutation frequency point set as the intensity quantification index. A higher normalized product indicates a greater synergistic intensity of amplitude and phase mutations at that frequency point. The amplitude-phase synchronous mutation intensity sequence records the intensity quantification value of each synchronous mutation frequency point using both factor identifier and frequency point identifier as dual indexes. Multiple synchronous mutation frequencies of the same factor in the synchronous mutation frequency point set are recorded independently in the amplitude-phase synchronous mutation intensity sequence, with the frequency point with the highest intensity value designated as the primary mutation frequency point for that factor. The normalized value of the amplitude change of a certain factor at the resonant frequency in the synchronous mutation frequency set is 0.86, and the normalized value of the phase change is 0.91. The product of the two, 0.78, is written into the amplitude-phase synchronous mutation intensity sequence. The normalized product of the synchronous mutation frequency points of the same factor in the non-resonant frequency band is only 0.23. The comparison of the intensity values ​​of the amplitude-phase synchronous mutation intensity sequence at the two types of frequency points clearly distinguishes the intensity distribution of the main mutation frequency point and the weak mutation frequency point. The higher the normalized product, the stronger the coordination between the amplitude mutation and the phase mutation at that frequency point. The intensity value of the main mutation frequency point is the highest value among multiple synchronous mutation frequency points of the same factor and is marked separately.

[0030] A frequency-dependent factor set is formed by screening factors with high synchronous mutation intensity based on the amplitude-phase synchronous mutation intensity sequence. The intensity value of the principal mutation frequency point of each factor in the amplitude-phase synchronous mutation intensity sequence reflects the maximum ability of that factor to generate amplitude-phase linked mutations within the working frequency band. The screening criterion for high synchronous mutation intensity factors is that the intensity value of the principal mutation frequency point of each factor in the amplitude-phase synchronous mutation intensity sequence exceeds the global intensity mean plus one standard deviation. Factors that meet the criteria are identified as frequency-dependent factors and added to the frequency-dependent factor set. The frequency-dependent factor set is indexed by factor identifiers, and includes the principal mutation frequency point identifier and corresponding intensity value of each factor. Factors with higher intensity values ​​in the amplitude-phase synchronous mutation intensity sequence have more significant frequency dependence characteristics in the frequency-dependent factor set. The number of entries in the frequency-dependent factor set reflects the proportion of factors with frequency dependence characteristics in the candidate bias factor set. When the intensity of a primary mutation frequency point of a factor in the amplitude-phase synchronous mutation intensity sequence is 0.78, exceeding the screening threshold of the global mean plus one standard deviation, the factor is added to the frequency-dependent factor set along with a primary mutation frequency point identifier. Conversely, if the intensity of a primary mutation frequency point of another factor is 0.31, below the screening threshold, it is not included in the frequency-dependent factor set. The screening boundary based on intensity determines the factor coverage of the frequency-dependent driving path in the cross-association strength assessment. The admission condition is set to the primary mutation frequency point intensity exceeding the global mean plus one standard deviation. This threshold effectively excludes interference from low-intensity noise factors while identifying significant frequency-dependent characteristics.

[0031] A cross-correlation strength assessment is performed on the temperature drift sensitive factor set and the frequency dependent factor set to generate coupling strength weights. The cross-correlation strength assessment is performed on the factor identifiers in both sets. For factors appearing in both sets, the normalized product of their temperature drift coefficient and amplitude-phase synchronization abrupt change intensity is calculated as the coupling strength weight of that factor. The formula is W = α_norm × S_norm, where W is the coupling strength weight, α_norm is the normalized value of the temperature drift coefficient, and S_norm is the intensity value of the main abrupt change frequency point. Factors appearing only in the temperature drift sensitive factor set have W = α_norm as their coupling strength weight, and factors appearing only in the frequency dependent factor set have W = S_norm as their coupling strength weight. The coupling strength weight records the quantified coupling strength value of each candidate deviation factor using the factor identifier as an index. The coupling strength weight of factors from both sets, which combine the characteristics of both temperature drift and frequency dependence, is usually higher than the weight of factors from a single source. The level of the coupling strength weight directly represents the comprehensive driving capability of each candidate deviation factor on the dual-channel amplitude-phase deviation. A factor appearing in both the temperature drift sensitive factor set and the frequency dependent factor set has a normalized temperature drift coefficient of 0.74 and a main abrupt change frequency point intensity of 0.78. Its normalized product of 0.58 is written into the coupling strength weight. Another factor appearing only in the temperature drift sensitive factor set has a coupling strength weight of 0.38, and a factor appearing only in the frequency dependent factor set has a coupling strength weight of 0.42. The numerical distribution of the three sources in the coupling strength weight reflects the different driving path strengths of each factor on the dual-channel deviation. Factors appearing in both sets typically have a higher normalized product weight than single-source factors, and will receive a higher priority in the subsequent node weight assignment of the channel deviation coupling relationship spectrum.

[0032] A channel deviation coupling spectrum is established based on coupling strength weights. The weight values ​​of each candidate deviation factor in the coupling strength weights reflect the comprehensive ability of that factor to drive the amplitude and phase deviation of the two channels. The channel deviation coupling spectrum uses the coupling strength weights as node weights and the degree of overlap of the amplitude and phase response frequency bands between each factor as edge weights, organizing all factors in the candidate deviation factor set into a weighted correlation network structure. The nodes of the channel deviation coupling spectrum are each factor in the candidate deviation factor set, and the node weights are directly assigned by the coupling strength weights. The edge weights between two nodes are calculated based on the overlap rate of the frequency bands covered by the corresponding two factors in the dual-channel amplitude and phase coupling deviation reference spectrum. The higher the frequency band overlap rate, the larger the edge weight between the factor pairs, indicating a stronger frequency band coupling degree between the two factors in deviation driving. In a direction finding system calibration scenario, two adjacent resonant factors in the candidate deviation factor set exhibit a frequency overlap rate of 0.73 in the dual-channel amplitude-phase coupling deviation reference spectrum. The channel deviation coupling spectrum establishes a strong connection edge with a weight of 0.73 between the two nodes. Another wideband monotonically drifting factor has a frequency overlap rate of only 0.18 with the aforementioned resonant factor, and its channel deviation coupling spectrum establishes a low-weight weak connection edge between the corresponding nodes. The distribution density of strong and weak connection edges in the channel deviation coupling spectrum reflects the frequency coupling pattern of each factor in the candidate deviation factor set. Factors with higher frequency overlap rates have larger edge weights, indicating tighter frequency coupling between the two factors in deviation-driven bias. The network structure of the channel deviation coupling spectrum provides a weighted basis for the deviation propagation path between factors in subsequent propagation path analysis.

[0033] In some embodiments, the step of performing a single-compensation overshoot risk assessment on the candidate deviation factor set to generate an overshoot risk label includes: performing channel response nonlinear slope calculation on each factor in the candidate deviation factor set to generate a nonlinear slope sequence; identifying factors whose slopes exceed the applicable range of linear compensation based on the nonlinear slope sequence to form an overshoot candidate factor set; performing compensation direction reversal risk quantification on the overshoot candidate factor set to generate a reversal risk coefficient; and marking the candidate deviation factor set with overshoot risk according to the reversal risk coefficient to generate an overshoot risk label.

[0034] A nonlinear slope sequence is generated by calculating the channel response nonlinear slope for each factor in the candidate deviation factor set. The nonlinear slope of the channel response is calculated using the slope of the change in amplitude and phase deviation values ​​of each factor in the dual-channel amplitude-phase coupling deviation reference spectrum relative to the compensation amount. Under linear compensation conditions, the slope of the deviation value relative to the compensation amount should remain constant and equal to the unit slope. Factors whose slopes deviate from the unit value indicate that their channel response exhibits nonlinear characteristics. The nonlinear slope calculation progressively calculates the first-order slope of the residual with respect to the compensation amount for each factor in the candidate deviation factor set as the compensation amount increases from zero to full scale. The first-order slope value at each compensation step is recorded as the nonlinear slope value at that step and written into the nonlinear slope sequence. The greater the deviation of the value in the nonlinear slope sequence from the unit value, the stronger the nonlinearity of the response in that compensation interval. The nonlinear slope sequence records the nonlinear slope values ​​of each factor in each compensation interval using the factor identifier and the compensation step size identifier as dual indices. For factors in the candidate deviation factor set containing resonant characteristics, the absolute value of the nonlinear slope sequence near the resonant frequency is usually higher than that of non-resonant factors. In the candidate bias factor set, the nonlinear slope value of one factor surges from 0.12 to 0.89 after the compensation exceeds 60% of the full scale. This shows a significant jump in the nonlinear slope sequence within the compensation interval. In contrast, the nonlinear slope sequence of another factor remains below 0.15 throughout the entire compensation interval. The difference in slope distribution between these two types of factors provides a direct criterion for subsequent screening of overshoot candidate factors. The more significant the slope deviation from a unit value, the weaker the applicability of the linear compensation assumption within that interval. The slope jump amplitude of resonant factors near the resonant frequency is typically much higher than that of non-resonant factors. These two types of differences constitute the criteria for screening overshoot candidate factors.

[0035] Factors with slopes exceeding the applicable range of linear compensation are identified based on nonlinear slope sequences to form an overshoot candidate factor set. The maximum absolute value of the slope of each factor in the nonlinear slope sequence within the full compensation interval is taken as the peak nonlinear slope of that factor. Factors whose peak nonlinear slope exceeds the threshold of the applicable range of linear compensation indicate that they will deviate from the linear response assumption during the compensation process, posing a potential risk of a reverse increase in the deviation value after the compensation amount is applied. These factors are added to the overshoot candidate factor set. The overshoot candidate factor set is indexed by factor identifiers, along with the peak nonlinear slope value of each factor and the compensation step size identifier where the peak value is located. Factors with higher peak nonlinear slopes have a stronger potential overshoot risk in the overshoot candidate factor set. When a factor with a peak nonlinear slope of 0.89 exceeds the applicable range threshold of 0.5 in the nonlinear slope sequence, this factor is added to the overshoot candidate factor set along with the peak compensation step size identifier. Factors with a peak nonlinear slope of 0.15 are below the threshold and are not included in the overshoot candidate factor set. The screening boundary of the peak nonlinear slope directly determines the factor input range for flip risk quantification. Entering this set means that the peak nonlinear slope has exceeded the applicable range of linear compensation, and this factor has the potential risk of the deviation value increasing in the opposite direction in full-scale compensation.

[0036] For the overshoot candidate factor set, a risk quantification of the compensation direction reversal is performed to generate a reversal risk coefficient. The nonlinear slope of each factor in the overshoot candidate factor set may continue to increase after the peak compensation step and exceed the unit slope. Exceeding the unit slope means that the reverse change of the deviation caused by the unit compensation amount exceeds the forward compensation amount. The risk quantification of the compensation direction reversal takes the integral value of the slope of the nonlinear slope sequence of each factor in the overshoot candidate factor set after the peak compensation step as the quantification basis. The reversal risk coefficient R is calculated using the formula R=∫(k(x)-1)dx (x is integrated from the peak step to full scale, k(x) is the slope value of the nonlinear slope sequence at the compensation step x, the unit of x is consistent with the unit of the compensation amount, and the dimension of R is the same as the dimension of the compensation amount). The larger the R value, the higher the risk of the factor reversing its direction under a single full-scale compensation. The flip risk coefficient records the R-value of each overshoot candidate factor using the factor identifier as an index. Factors with an R-value greater than zero in the overshoot candidate factor set indicate that they have a directional flip risk range within the full-scale compensation range. Factors with a negative R-value, although exhibiting nonlinearity, do not have a slope exceeding the unit slope after the peak, and their flip risk is relatively controllable. If the nonlinear slope sequence of a certain factor in the overshoot candidate factor set has a slope exceeding 1.0 after the peak compensation step, the integral result of the flip risk coefficient R is recorded as positive. If the slope of another factor falls back below 1.0 after the peak, the R-value is recorded as negative. The R-value integration range extends from the peak step to the full scale. Factors with a slope exceeding the unit value after the peak have a positive R, while factors with a slope falling back after the peak have a negative R. The positive and negative results directly distinguish the degree of flip risk of the two types of factors.

[0037] Overshoot risk labels are generated by labeling the candidate deviation factor set based on the flip risk coefficient. Factors with a positive R-value in the flip risk coefficient indicate a risk range of directional flip during a single full-scale compensation. Overshoot risk labeling is primarily based on the sign of the R-value for each factor in the flip risk coefficient. Factors with a positive R-value are labeled with high overshoot risk on the corresponding entry in the candidate deviation factor set. Overshoot candidate factors with a negative R-value but whose peak nonlinear slope exceeds the threshold applicable to linear compensation are labeled with medium overshoot risk. Candidate deviation factors not included in the overshoot candidate factor set are labeled with low overshoot risk. The overshoot risk label records the overshoot risk level of all factors in the candidate deviation factor set using the factor identifier as an index. High overshoot risk factors are accompanied by the flip risk coefficient R-value in the overshoot risk label; medium overshoot risk factors are accompanied by the peak nonlinear slope value; and low overshoot risk factors are written in a low-risk state without any accompanying parameters. This three-level labeling covers all factors in the candidate deviation factor set, ensuring that no factor enters the subsequent processing stage in an undefined risk state. Factors with higher R-values ​​among those with high overshoot risk exhibit a more significant increase in deviation under full-scale compensation. Factors at the same high overshoot risk level are further quantified by their accompanying parameters based on their different R-values. A factor with a positive R-value in the overshoot candidate factor set is marked as having high overshoot risk and accompanied by its R-value in the corresponding entry of the candidate deviation factor set. Factors not included in the overshoot candidate factor set are batch-written as low-risk in the overshoot risk label. A positive R-value is marked as high risk and accompanied by its R-value; a negative R-value but with a peak slope exceeding a threshold is marked as medium risk and accompanied by its peak slope. The difference between these two types of accompanying parameters allows for a quantifiable distinction of the risk level of each factor.

[0038] Step S130: Real-time calibration residual convergence assessment is performed based on the channel deviation coupling relationship spectrum and overshoot risk marker to generate deviation classification parameters. The deviation level is determined according to the deviation classification parameters. Calibration convergence blocking factors are identified from the deviation classification parameters to generate a blocking factor set.

[0039] In some embodiments, the step of generating deviation grading parameters by real-time calibration residual convergence assessment based on the channel deviation coupling spectrum and the overshoot risk marker includes: extracting iterative compensation response curves of each deviation factor from the channel deviation coupling spectrum to generate a compensation response set; identifying secondary recovery segments after the first convergence of the residuals for each factor in the compensation response set to form a recovery segment set; calculating the ratio of the recovery amplitude to the first convergence amount for the recovery segment set in combination with the overshoot risk marker to generate a recovery ratio sequence; and sorting each deviation factor by severity according to the recovery ratio sequence to generate deviation grading parameters.

[0040] The iterative compensation response curves of each deviation factor are extracted from the channel deviation coupling spectrum to generate a compensation response set. The iterative compensation response curves are plotted with the absolute value of the residual after each round of compensation iterations on the vertical axis and the iteration round on the horizontal axis. The coupling strength weight of each node in the channel deviation coupling spectrum determines the amount of compensation resources allocated to the corresponding factor in each iteration. Nodes with higher weights receive more compensation resources in each iteration, and their iterative compensation response curves show a greater slope of residual decrease in the initial round. The compensation response set stores the iterative compensation response curves of each factor using factor identifiers as indexes. Factors with strong connections in the channel deviation coupling spectrum exhibit a correlation in the residual decrease rhythm of their iterative compensation response curves in the compensation response set; when the residual of one factor decreases, the iterative compensation response curve of the other factor may experience a temporary increase in residual. In the channel deviation coupling spectrum, a factor with a coupling strength weight of 0.78 exhibits a rapid decrease in residuals during the first three rounds of iterative compensation response curves in the compensation response set. Meanwhile, the iterative compensation response curve of another strongly connected factor shows a slight increase in residuals during the second round, followed by a return to a decrease. The morphological correlation between these two iterative compensation response curves in the compensation response set confirms the characterization of residual propagation by the edge weights of the channel deviation coupling spectrum. Nodes with higher coupling strength weights receive more compensation resources per round, resulting in a steeper slope for residual decrease in the initial round. The residual propagation correlation between strongly connected factors is directly reflected in the morphological fluctuations of their respective curves.

[0041] A set of rebound intervals is formed by identifying the secondary rebound intervals after the initial convergence of the residuals for each factor in the compensation response set. Under normal convergence conditions, the iterative compensation response curves of each factor in the compensation response set should exhibit a monotonically decreasing trend until the residuals fall below the convergence threshold. If the residuals rise again in subsequent iterations after initially falling below the convergence threshold, a secondary rebound interval is formed. The appearance of a secondary rebound interval indicates that the factor has experienced a residual rebound after the initial convergence due to residual propagation from adjacent strongly connected factors or its own nonlinear characteristics. Based on the iterative compensation response curves of each factor in the compensation response set, the iterative intervals where the residuals first fall below the convergence threshold and then exceed the convergence threshold again are identified and written into the set of rebound intervals as secondary rebound intervals. Factors whose iterative compensation response curves in the compensation response set are always monotonically decreasing are not included in the list of rebound intervals. In the compensation response set, the iterative compensation response curve of a certain factor falls below the convergence threshold for the first time in the fifth round, and then rises again in the seventh round, exceeding the convergence threshold. The recovery segment set identifies the interval from the seventh to the ninth round with this factor. Another factor's iterative compensation response curve monotonically decreases throughout the entire iteration round; the recovery segment set does not generate an entry for this factor. Factors whose iterative compensation response curves consistently decrease monotonically converge stably, while factors exhibiting a second rise indicate that their residuals are affected by external propagation or their own nonlinearity after the first convergence. The boundary between these two cases is the coverage area of ​​the recovery segment set. A second rise in residuals after the first convergence indicates that the factor is affected by residual propagation from adjacent strongly connected factors or by its own nonlinear characteristics. The initial iteration round of the second rise segment marks the position where convergence stability begins to be lost.

[0042] For example, the step of calculating the ratio of the recovery amplitude to the first convergence amount and generating a recovery ratio sequence for the recovery segment set in combination with the overshoot risk marker includes: extracting amplitude residual convergence curves and phase residual convergence curves for each factor in the recovery segment set to generate amplitude-phase convergence curve pairs; comparing the amplitude convergence rate and phase convergence rate of the amplitude-phase convergence curve pairs to generate a convergence rate difference sequence; identifying factors whose rate mismatch exceeds the tolerance based on the convergence rate difference sequence and the overshoot risk marker to form a mismatch factor set; and calculating the ratio of the recovery amplitude to the first convergence amount and generating a recovery ratio sequence based on the degree of mismatch in the mismatch factor set.

[0043] Amplitude and phase residual convergence curves are extracted from each factor in the recovery segment set to generate amplitude-phase convergence curve pairs. The iterative compensation response curves of each factor in the recovery segment set evolve independently in both the amplitude and phase residual dimensions. The extraction of amplitude-phase convergence curve pairs is based on reading the amplitude and phase residual sequences of each factor in the recovery segment set round by round throughout the entire iteration cycle. The two residual curves are stored separately in the amplitude-phase convergence curve pairs using factor identifiers as dual indices. Under normal dual-channel compensation conditions, the amplitude and phase residual convergence curves in the amplitude-phase convergence curve pairs should maintain synergy in their convergence rhythms. The smaller the difference in the iteration cycle at which the two curves first fall below their respective convergence thresholds, the stronger the synergy in amplitude-phase compensation; a larger difference in the iteration cycle indicates a significant decoupling of the amplitude-phase convergence rhythms. In the recovery phase set, the amplitude residual convergence curve for a certain factor converges for the first time in the fifth round, while the phase residual convergence curve converges for the first time in the eighth round. The difference in the number of convergence rounds for the amplitude and phase convergence curves for this factor is 3 rounds. For another factor, the amplitude and phase convergence curves for the two curves converge simultaneously for the first time in the sixth round, with a zero difference in the number of convergence rounds. The smaller the difference in the number of iteration rounds when the two curves first fall below their respective convergence thresholds, the stronger the linkage and synergy of amplitude and phase compensation; the larger the difference in the number of rounds, the more obvious the decoupling of the amplitude and phase convergence rhythm.

[0044] A convergence rate difference sequence is generated by comparing the amplitude convergence rate and the phase convergence rate of the amplitude-phase convergence curve pair. The amplitude convergence rate is calculated as the decrease in residual between adjacent iterations of the amplitude residual convergence curve in the amplitude-phase convergence curve pair, and the phase convergence rate is calculated as the decrease in residual between adjacent iterations of the phase residual convergence curve. The convergence rate difference sequence is calculated and written into the absolute value of the difference between the normalized amplitude convergence rate and the phase convergence rate, and the calculation formula is δ_i(t)=|Δa_i(t) / a_i(1)-Δφ_i(t) / φ_i(1)|, where δ_i(t) is the value of the amplitude-phase convergence rate. The rate difference values ​​of factors i in round t, Δa_i(t) is the residual decrease of the amplitude residual convergence curve of factor i in round t, a_i(1) is the absolute value of the amplitude residual of factor i in the first round, Δφ_i(t) is the residual decrease of the phase residual convergence curve of factor i in round t, and φ_i(1) is the absolute value of the phase residual of factor i in the first round. Normalization is based on the absolute value of the first round residual of each factor's corresponding dimension to convert the rate values ​​into dimensionless ratios, ensuring that the rate values ​​of amplitude and phase are comparable under the same dimensions. The convergence rate difference sequence records the rate difference values ​​of each factor in each iteration round using factor identifier and iteration round as dual indices. In the sixth round, the amplitude residual of a certain resonant factor decreased by 0.3 dB and the phase residual decreased by 0.08°. After normalization, the amplitude convergence rate was much higher than the phase convergence rate, and the difference in rate was large. This indicates that the amplitude compensation of this factor has almost converged while the phase compensation is still lagging. This asynchrony is usually caused by the large difference between the amplitude response slope and the phase response slope near the resonance point. After several rounds, it will drive the amplitude residual to rebound twice.

[0045] A mismatch factor set is formed by combining the convergence rate difference sequence with overshoot risk markers to identify factors whose rate mismatch exceeds the tolerance limit. The more iterations in which the rate difference value of each factor in the convergence rate difference sequence exceeds the rate mismatch tolerance threshold, the more persistent the amplitude-phase compensation progress of that factor is uncoordinated across multiple iterations. The mismatch factor set is selected based on the criterion that the proportion of iterations in which each factor in the convergence rate difference sequence exceeds the tolerance limit exceeds the mismatch proportion threshold. The impact of the overshoot risk marker on the selection of the mismatch factor set is reflected in the dynamic adjustment of the tolerance threshold. Factors with high overshoot risk naturally have lower amplitude-phase convergence rates due to compensation limitation, and their tolerance thresholds are correspondingly relaxed in the selection of the mismatch factor set. Factors with low overshoot risk are judged using the standard tolerance threshold, ensuring that the selection results of the mismatch factor set do not misjudge the rate difference caused by the limitation as amplitude-phase decoupling. When the percentage of iterations exceeding the standard tolerance threshold for a certain factor in the convergence rate difference sequence reaches 65%, exceeding the mismatch ratio threshold, an entry for that factor is added to the mismatch factor set. For another high overshoot risk factor, after relaxing the tolerance threshold, if the percentage of iterations exceeding the standard tolerance threshold drops to 30%, below the mismatch ratio threshold, an entry for that factor is not generated in the mismatch factor set. This differentiation between the two screening results effectively separates the amplitude-limiting side effect from the true amplitude-phase decoupling, preventing overshoot risk factors from being misjudged as sources of mismatch. The tolerance threshold for high overshoot risk factors is correspondingly relaxed during screening, effectively isolating the amplitude-limiting side effect from the true amplitude-phase decoupling in the mismatch judgment results, avoiding misjudging rate differences caused by amplitude limiting as amplitude-phase decoupling.

[0046] The recovery ratio sequence is generated by calculating the ratio of the recovery amplitude to the first convergence amount based on the degree of mismatch in the mismatch factor set. The degree of mismatch of each factor in the mismatch factor set is represented by the average rate difference of the convergence rate difference sequence exceeding the tolerance round. The higher the degree of mismatch, the greater the residual rebound amplitude in the secondary recovery segment corresponding to the recovery segment set. The calculation of the ratio of recovery amplitude to first convergence amount uses the degree of mismatch in the mismatch factor set as a weighting coefficient to correct the secondary recovery amplitude in the recovery segment set. The corrected recovery amplitude is divided by the first convergence amount of the corresponding factor to obtain the recovery ratio. The calculation formula is ρ=(ΔR×w) / C_first, where ρ is the recovery ratio, ΔR is the secondary recovery amplitude in the recovery segment set, w is the normalized weighting coefficient of the degree of mismatch, and C_first is the first convergence amount. The recovery ratio sequence records the recovery ratio of each mismatched factor, indexed by its factor identifier. A higher recovery ratio indicates a larger proportion of secondary residual rebound relative to the initial convergence, contributing a higher ranking weight in subsequent bias severity ranking. A factor with a high degree of mismatch in the mismatch factor set corresponds to a secondary recovery amplitude of 0.52 dB and an initial convergence of 0.85 dB in the recovery segment set. After mismatch degree weighting correction, its recovery ratio is 0.61 and entered into the recovery ratio sequence. Another factor with a lower degree of mismatch has a corrected recovery ratio of 0.29. The numerical difference between the two factors in the recovery ratio sequence directly characterizes the differentiated contribution of amplitude-phase convergence incoordination to the severity of secondary rebound. The degree of mismatch is used as a weighting coefficient to correct the secondary rebound amplitude; the corrected recovery ratio simultaneously reflects the combined severity of both amplitude-phase convergence incoordination and residual rebound amplitude.

[0047] Based on the rebound ratio sequence, the severity of each deviation factor is ranked to generate deviation grading parameters. The rebound ratio of each factor in the rebound ratio sequence reflects the relative severity of the residual's second rebound during the iterative compensation process. The severity ranking uses the rebound ratio of each factor in the rebound ratio sequence as the primary sorting key. Candidate deviation factors not included in the rebound ratio sequence participate in the ranking with a zero rebound ratio. The ranking results are written into the severity grading entries of the deviation grading parameters from high to low. The deviation grading parameters record the ranking results of all factors using a triplet of factor identifier, rebound ratio, and severity level. Factors with rebound ratios exceeding the high severity threshold are marked as high severity level in the deviation grading parameters, factors with rebound ratios between the medium and low thresholds are marked as medium severity level, and factors with a rebound ratio of zero are marked as low severity level. When a factor with a recovery ratio of 0.61 in the recovery ratio sequence exceeds the high severity threshold, the deviation grading parameter labels it as high severity; a factor with a recovery ratio of 0.29 is labeled as medium severity; and candidate deviation factors not included in the recovery ratio sequence are batched into the low severity level. The high, medium, and low levels are divided by the two recovery ratio thresholds. The level labeling of each factor, together with the recovery ratio, constitutes the quantitative input for deviation level determination and blocking factor identification.

[0048] The deviation level is determined based on the deviation grading parameters. The severity distribution of each factor in the deviation grading parameters reflects the overall status of the current dual-channel amplitude and phase deviation in terms of convergence stability. The deviation level is determined by two criteria: the proportion of high-severity factors and the highest recovery ratio. A high deviation level is defined as one where the proportion of high-severity factors exceeds a threshold or the highest recovery ratio exceeds an extreme threshold. A low deviation level is defined as one where both criteria are below the threshold, and a medium deviation level is defined as one where only one criterion exceeds the threshold. The deviation level is output as an overall level identifier, supported by the proportion of high-severity factors and the highest recovery ratio in the deviation grading parameters. A high deviation level is defined as one where the proportion of high-severity factors is 35%, exceeding the 20% threshold. A low deviation level is defined as one where the proportion of high-severity factors is only 8% and the highest recovery ratio is below the extreme threshold. The level of deviation affects two downstream stages simultaneously: firstly, it determines the tightness of the blocking ratio threshold to control the coverage of the blocking factor set; secondly, it determines the application strategy of the compensation amount when the final amplitude and phase calibration compensation parameter is output. High levels trigger step-by-step application to avoid overshoot, while low levels allow full application at once. Therefore, the deviation level becomes the core control parameter connecting the two stages of convergence evaluation and compensation execution.

[0049] A set of blocking factors is generated by identifying calibration convergence blocking factors from the deviation grading parameters. The identification of these blocking factors uses high-severity factors in the deviation grading parameters as candidate sources. The rebound ratio of high-severity factors indicates that they continuously generate residual rebounds during iterative compensation. When the residuals still cannot stably converge within the threshold after multiple iterations, this factor is determined to block overall calibration convergence. The blocking factor identification criterion is that the cumulative number of iterations in which the response curve of a high-severity factor in the deviation grading parameters exceeds the convergence threshold relative to the total number of iterations exceeds a blocking proportion threshold. Factors meeting this condition are added to the blocking factor set. The blocking proportion threshold is dynamically adjusted according to the deviation level; when the deviation level is high, the blocking proportion threshold is lowered to expand the coverage of the blocking factor set. In the deviation grading parameters, a certain high-severity factor exceeded the convergence threshold in 14 out of 20 iterations, representing 70% of the total. This exceeds the reduced blocking ratio threshold of 60% under high deviation levels, and this factor is added to the blocking factor set. Another high-severity factor, with an exceedance rate of 45% below the blocking ratio threshold, is not included in the blocking factor set. The dynamic tightening of the blocking ratio threshold ensures that no convergence-threatening factors are missed in high-deviation-level scenarios. The blocking ratio threshold is dynamically adjusted with the deviation level, decreasing at high deviation levels to broaden coverage and ensure that no key factors posing a substantial threat to calibration convergence are overlooked when overall deviation is severe.

[0050] Step S140: The blocking factor set is divided into frequency bands and the deviation is classified to obtain the deviation value of each frequency band. The amplitude and phase deviation polarity reversal detection is performed on the deviation value of each frequency band to identify the deviation polarity reversal frequency point. The distribution pattern is identified based on the deviation polarity reversal frequency point and the deviation value of each frequency band to obtain the deviation distribution type.

[0051] In some embodiments, the step of classifying the blocking factor set by frequency band partition to obtain the deviation value of each frequency band includes: performing inter-band amplitude-phase coupling transmission path analysis on each factor in the blocking factor set to generate a transmission path diagram; identifying cross-frequency band amplitude-phase transmission chains in the transmission path diagram to form a transmission chain length sequence; allocating frequency band classification weights to the blocking factor set based on the transmission chain length sequence to generate a frequency band classification weight table; and calculating and determining the deviation value of each frequency band by accumulating the values ​​according to each frequency band partition based on the frequency band classification weight table.

[0052] Inter-band amplitude-phase coupling propagation path analysis is performed on each factor in the blocking factor set to generate a propagation path graph. The analysis starts with the edge connections between nodes corresponding to the blocking factor set in the channel deviation coupling spectrum, tracing the path through which each blocking factor propagates amplitude-phase deviation components to adjacent frequency band factors via strong connections. The frequency band identifiers and edge weights of each node on the propagation path are combined to represent a directed propagation chain and written into the propagation path graph. The propagation path graph uses the blocking factor identifier as the root node and the identifiers of adjacent factors propagated hierarchically as child nodes. The weights of the connections between nodes inherit the corresponding edge weights from the channel deviation coupling spectrum. The hierarchical depth of the propagation path graph reflects the cross-band influence extension range of each blocking factor. In a direction-finding machine calibration scenario, a certain blocking factor in the channel deviation coupling spectrum is transmitted to an adjacent resonant factor through a strong connection edge with a weight of 0.73. The transmission path graph records a cross-frequency band transmission chain under the root node of this factor. This adjacent resonant factor is further connected to a third factor in a more distant frequency band through an edge with a weight of 0.41. The transmission path graph records the second-hop transmission path at this layer. Another blocking factor, however, has only a zero-weight isolated node connection in the channel deviation coupling spectrum. The transmission path graph for this factor is written as a single-node, non-outward transmission path. The multi-hop transmission path driven by strong connections and the non-outward transmission of isolated nodes create a significant hierarchical differentiation in the path graph. The path hierarchy depth directly reflects the cross-frequency band influence extension range of each blocking factor.

[0053] In the transmission path diagram, cross-band amplitude and phase transmission chains are identified to form a transmission chain length sequence. The number of cross-band node hops involved in the transmission path extending downwards from the root node of each blocking factor in the transmission path diagram constitutes the transmission chain length of that blocking factor. The more cross-band node hops a blocking factor has, the farther its amplitude and phase deviation affects the frequency range through the cascading transmission of the channel deviation coupling spectrum. The transmission chain length sequence records the longest cross-band transmission chain hops starting from the root node of each factor, indexed by the blocking factor identifier. When there are multiple transmission paths for the same blocking factor in the transmission path diagram, the transmission chain length sequence takes the longest path hop count. The product of the weights of each path edge is recorded as the effective transmission strength of that path in the transmission chain length sequence. In the transmission path diagram, a blocking factor starts from its root node and reaches a third-band factor after two hops. The product of the weights of the two edges is 0.73 × 0.41 = 0.30. The transmission chain length sequence for this factor records a transmission chain length of 2 hops and an effective transmission strength of 0.30. The transmission chain length sequence for a single-node blocking factor records zero hops. When there are multiple transmission paths for the same blocking factor, the longest hop count is taken. The product of the edge weights of each path is recorded as the effective transmission strength. The combination of hop count and strength quantifies the actual ability of each blocking factor to spread deviation to the far-end frequency band.

[0054] A frequency band classification weight table is generated by assigning frequency band classification weights to the set of blocking factors based on the transmission chain length sequence. The transmission chain length and effective transmission strength of each blocking factor in the transmission chain length sequence jointly determine its contribution weight to the calculation of the deviation value for each frequency band. Blocking factors with longer transmission chain lengths and higher effective transmission strengths have greater classification weights for distant frequency band partitions. The frequency band classification weight allocation uses the hop-by-hop effective transmission strength of each factor in the transmission chain length sequence as the weight source. Classification weights are calculated for each blocking factor and each frequency band partition and written into the frequency band classification weight table. The frequency band classification weight table records the classification weight of each blocking factor for each frequency band partition using both the blocking factor identifier and the frequency band partition identifier as dual indexes. The classification weight of the same blocking factor for its own frequency band partition is 1.0. The classification weight for frequency band partitions covered by the transmission chain is assigned based on the effective transmission strength of the corresponding hop count. The classification weight for frequency band partitions not covered by the transmission chain is zero. In the transmission chain length sequence, a blocking factor with a two-hop effective transmission strength of 0.30 is assigned a weight of 0.30 to the frequency band classification weight table of the third frequency band partition and a weight of 1.0 to the classification weight of the first frequency band partition. A zero-hop single-node blocking factor is assigned a weight of 1.0 only to its own frequency band partition. The weight of its own frequency band partition is fixed at 1.0, and the transmission chain coverage partition is assigned a value based on the hop-by-hop effective transmission strength. If a blocking factor has a weight of 0.73 at the first hop and its product decreases to 0.30 at the second hop, the far-end third frequency band partition will only include the bias contribution of this factor with a weight of 0.30, reflecting the actual attenuation of the bias energy after multiple cascaded transmissions.

[0055] The deviation value of each frequency band is determined by cumulative calculation according to the frequency band classification weight table and each frequency band partition. The classification weight of each blocking factor in the frequency band classification weight table provides a weighting coefficient for the cumulative deviation calculation. The cumulative calculation of the deviation value of each frequency band is multiplied by the classification weight of all blocking factors in the corresponding frequency band partition column of the frequency band classification weight table. The mean amplitude and phase deviation within the corresponding frequency band range of each factor in the blocking factor set is multiplied by the classification weight and then summed factor by factor. The calculation formula is V_k=Σ(w_ik×μ_i) (i traverses all blocking factors), where V_k is the frequency band deviation value of the k-th frequency band partition, w_ik is the classification weight of the i-th blocking factor in the k-th frequency band partition, and μ_i is the mean amplitude and phase deviation within the frequency band range of the i-th blocking factor. The summation result is written as the frequency band deviation value of that frequency band partition and output. The average amplitude and phase deviations of the sinusoidal and cosine channels for each frequency band deviation value, weighted and accumulated using the frequency band classification weight table, reflect the quantitative result of the contribution of the blocking factor set to the overall deviation of each frequency band through the transmission path. Frequency band partitions with a classification weight of zero are not included in the deviation contribution of the corresponding blocking factor in the accumulation calculation. When the classification weight of the third frequency band partition for two-hop blocking factors in the frequency band classification weight table is 0.30, the average amplitude and phase deviation within the frequency band range of this blocking factor is multiplied by 0.30 and accumulated to the deviation values ​​of each frequency band in the third frequency band partition. Zero-hop blocking factors have a classification weight of zero for the third frequency band partition and are not included in the accumulation. The difference in the contribution of the two types of factors to the accumulation of deviation values ​​in the third frequency band partition reflects the quantitative characterization of the far-end frequency band deviation diffusion by the transmission path map. Taking the third frequency band partition as an example, the independent statistics of a single frequency band only include the mean deviation of the partition's own blocking factors. The weighted summation adds the deviation contribution of the two-hop transmission factors with a weight of 0.30. The difference between the two is the additional deviation component introduced by the cross-frequency band transmission path in this partition. If this component is ignored, the compensation amount of the third frequency band partition will be systematically low.

[0056] Amplitude and phase deviation polarity reversal detection is performed on the deviation values ​​of each frequency band to identify the frequency points where the deviation polarity is reversed. The sign changes of the mean amplitude deviation and mean phase deviation of the sine and cosine channels in each frequency band deviation value between adjacent frequency band partitions reflect the polarity reversal characteristics of the deviation. The amplitude and phase deviation polarity reversal detection is based on the sign difference of the mean amplitude deviation and mean phase deviation between adjacent frequency band partitions of each frequency band deviation value. The boundary frequency points where the sign of the mean amplitude deviation of adjacent frequency band partitions changes from positive to negative or from negative to positive are marked as amplitude polarity reversal frequency points. The boundary frequency points where the sign of the mean phase deviation changes in the same way are marked as phase polarity reversal frequency points. The frequency points where both amplitude polarity reversal frequency points and phase polarity reversal frequency points appear are marked as amplitude and phase synchronous reversal frequency points and are preferentially written into the deviation polarity reversal frequency points. The deviation polarity reversal frequency point is recorded with dual attributes: a reversal type identifier and a frequency band boundary identifier. For each frequency band deviation value, the polarity reversal frequency points for the sine and cosine channels are respectively labeled with the channel source. When both channels undergo polarity reversal at the same boundary frequency point, they are merged and written using the amplitude-phase synchronous reversal type. During the calibration of a direction finder, the deviation values ​​for each frequency band show that the average amplitude deviation of the sine channel in the first frequency band partition is +0.4dB, while in the second frequency band partition it drops to -0.2dB. The amplitude-phase deviation polarity reversal detection identifies the boundary frequency points of the two partitions as the amplitude polarity reversal frequency points of the sine channel. The average phase deviation of the cosine channel synchronously undergoes sign reversal at the same boundary frequency point. The deviation polarity reversal frequency point is written to this boundary frequency point using the amplitude-phase synchronous reversal type. The sign of the frequency band deviation values ​​at the boundaries of the second to third frequency band partitions does not change, and the deviation polarity reversal detection does not generate a reversal frequency point at this boundary. The amplitude and phase synchronous flip type is preferentially merged and written, while the boundary frequency points of the two channels that flip independently are marked with the channel source respectively. The difference between the two recording methods preserves the linkage and independent information of the polarity change of the two channels.

[0057] In some embodiments, the step of obtaining the deviation distribution type by performing distribution pattern recognition based on the deviation polarity reversal frequency point and the deviation values ​​of each frequency band includes: performing segment extraction on both sides of the reversal point of the deviation values ​​of each frequency band based on the deviation polarity reversal frequency point to generate a set of deviation segments on both sides; performing asymmetry calculation on the deviation amplitude on both sides of each reversal point in the set of deviation segments on both sides to generate a reversal asymmetry sequence; performing asymmetry degree classification based on the reversal asymmetry sequence to generate an asymmetry classification identifier; and determining the deviation distribution type based on the asymmetry classification identifier.

[0058] Based on the deviation polarity reversal frequency point, the deviation values ​​of each frequency band are segmented on both sides of the reversal point to generate two deviation segment sets. The deviation polarity reversal frequency point divides the frequency axis of each frequency band deviation value into the frequency band to the left and right of the reversal point. The extraction of the two deviation segment sets uses each entry of the deviation polarity reversal frequency point as the segmentation node. For each reversal frequency point, the average amplitude and phase deviation values ​​of all frequency band partitions to the left of the reversal frequency point are collected into the left deviation segment, and all frequency band partitions to the right are collected into the right deviation segment. The left and right deviation segments are written side by side into the two deviation segment sets using the reversal frequency point identifier as the composite key. The two deviation segment sets use the reversal frequency point identifier as the row index, and the left and right deviation segment columns store the corresponding amplitude and phase deviation average value sequences. During the calibration of a direction finder, the deviation polarity reversal frequency point is at the 700MHz boundary. Among the deviation values ​​in each frequency band, 400MHz to 700MHz is the left deviation segment, and 700MHz to 900MHz is the right deviation segment. The sets of deviation segments on both sides are written into the amplitude and phase deviation mean sequences of the three left frequency band partitions and the two right frequency band partitions respectively at the reversal frequency point. The average amplitude of the left deviation segment is generally positive, while the average amplitude of the right deviation segment is generally negative. The left and right deviation segments are stored side-by-side using the reversal frequency point identifier as a key. The distribution characteristic of the left side's overall positive and the right side's negative amplitude mean will be quantified as a high asymmetry normalization ratio in the asymmetry calculation.

[0059] Asymmetry calculation is performed on the deviation amplitudes on both sides of each flip point in the deviation segment set to generate a flip asymmetry sequence. The asymmetry is calculated as the normalized ratio of the difference between the absolute value of the mean absolute value of the left deviation segment amplitude and the mean absolute value of the right deviation segment amplitude, divided by the sum of their means. The formula is D=|μ_L-μ_R| / (μ_L+μ_R), where D is the normalized ratio of asymmetry, μ_L is the mean absolute value of the left deviation segment amplitude, and μ_R is the mean absolute value of the right deviation segment amplitude. The value of D ranges from 0 to 1. The closer the value is to 1, the more significant the difference in deviation amplitudes on both sides. The closer the value is to 0, the more balanced the amplitude distribution on both sides. The flip asymmetry sequence records the normalized ratio of asymmetry for each flip frequency point, indexed by the flip frequency point identifier. For flip frequencies with unequal numbers of frequency band partitions on the left and right sides of the deviation segment concentration, the mean amplitude of each side is first weighted and normalized by the number of effective frequency band partitions before asymmetry calculation, ensuring that the asymmetry ratio of each flip frequency point remains comparable even when the number of partitions on both sides differs. When a direction finder operates in a wide frequency band, if the amplitude deviation of the frequency band on the left side of the 700MHz flip frequency point is significantly higher than that on the right side, and the difference in the mean absolute amplitude values ​​on both sides is large, the flip asymmetry sequence records a high D value at this frequency point, indicating that the deviation energy on both sides of the flip point is highly concentrated on one side. If the mean amplitude deviations on both sides of another flip frequency point are similar, and the D value is close to zero, it indicates that the deviation amplitudes of the frequency bands before and after the flip point are relatively balanced. The difference in the D value of the two types of flip frequencies in the flip asymmetry sequence directly reflects the concentrated or dispersed distribution characteristics of the deviation on both sides of the frequency band flip boundary.

[0060] Asymmetry classification labels are generated based on the flip asymmetry sequence. The normalized ratio of each flip frequency point in the flip asymmetry sequence reflects the strength of the imbalance in the magnitude of the deviations on both sides. The asymmetry classification is based on the global distribution of the flip asymmetry sequence. Flip frequencies with normalized ratios exceeding the high asymmetry threshold are labeled as high asymmetry level; those with normalized ratios between the medium and low thresholds are labeled as medium asymmetry level; and those with normalized ratios below the low threshold are labeled as low asymmetry level. These three level labels are written into the asymmetry classification label for each flip frequency point. The asymmetry classification label records the asymmetry level of each flip frequency point using the flip frequency label as an index. High asymmetry level entries include the normalized ratio, while low asymmetry level entries are written with the level label as the only content. In the flip asymmetry sequence, flip frequencies with a normalized ratio of 0.49 exceeding the high asymmetry threshold of 0.4 are labeled as high asymmetry and accompanied by the ratio. Flip frequencies with a normalized ratio of 0.03 are below the low asymmetry threshold of 0.1 and are labeled as low asymmetry. The high, medium, and low levels are divided by the normalized ratio thresholds. High asymmetry level entries include the ratio, while low asymmetry level entries are written solely with the level label. The difference in information density between the two types of entries reflects the granularity of the classification results.

[0061] The deviation distribution type is determined based on the asymmetric classification identifier. The asymmetric classification identifier reflects the amplitude balance of the deviation values ​​on both sides of the flip point. The determination of the deviation distribution type is mainly based on the proportion of high asymmetric classification flip points to the total number of flip points in the asymmetric classification identifier. When the proportion of high asymmetric classification exceeds the distribution type determination threshold, the deviation distribution type is determined to be asymmetric concentrated; when the proportion of high asymmetric classification is below the threshold, the deviation distribution type is determined to be symmetrical diffuse; when both types of flip points are equally distributed, it is determined to be a mixed distribution. The deviation distribution type is output as a type identifier, along with a list of identifiers for high asymmetric classification flip points in the asymmetric classification identifier. In the calibration of a certain direction finder, one of the two flip points in the asymmetric classification identifier is of high asymmetric classification, and the proportion of high asymmetric classification is 50%, exceeding the determination threshold of 40%, so the deviation distribution type is determined to be asymmetric concentrated. In the calibration of another direction finder, all flip points in the asymmetric classification identifier are of low asymmetric classification, so the deviation distribution type is determined to be symmetrical diffuse. If the proportion of high asymmetry levels exceeds the judgment threshold, it is determined to be an asymmetric concentration type; if both are below the threshold, it is determined to be a symmetric diffusion type; and if both are equal, it is determined to be a mixed distribution type. The three types correspond to different compensation allocation strategies.

[0062] Step S150: Based on the deviation distribution type and the deviation polarity reversal frequency point, calculate the cross-band amplitude and phase linkage compensation amount to generate the frequency band deviation distribution characteristics, and output the amplitude and phase calibration compensation parameters by combining the frequency band deviation distribution characteristics and the deviation level.

[0063] Specifically, cross-band amplitude and phase linkage compensation is calculated based on the deviation distribution type and the frequency point where the deviation polarity reverses, generating frequency band deviation distribution characteristics. The cross-band amplitude and phase linkage compensation calculation uses the deviation distribution type as the strategy selection basis. For asymmetric concentrated deviation distribution types, a concentrated compensation strategy is triggered, distributing the compensation amount centrally to the frequency band where the high asymmetry level reversal frequency point is located. For symmetric diffuse deviation distribution types, a uniform distribution strategy is triggered, distributing the compensation amount evenly to each frequency band partition according to the amplitude ratio of the deviation values ​​of each frequency band. The deviation polarity reversal frequency point serves as the boundary constraint node for the compensation amount. The calculation adjusts the polarity of the compensation amount according to the reversal boundary of the deviation polarity reversal frequency point. The polarity of the compensation amount in the frequency band partition to the left of the reversal frequency point is reversed according to the sign of the corresponding frequency band deviation values, and the frequency band partition to the right of the reversal frequency point also independently takes the reversed direction according to the sign of the corresponding frequency band deviation values ​​after the polarity reversal. When a direction finder is calibrated, its deviation distribution type is asymmetric concentrated. The calculation allocates 65% of the compensation to the left band of the high asymmetry level reversal frequency point and 35% to the right band. At the 700MHz boundary, the polarity of the compensation on both sides is constrained to be positive and negative respectively. The calculated compensation amplitude and polarity for each frequency band are written into the frequency band deviation distribution characteristics for each band. When another direction finder has a symmetrical diffusion type deviation distribution, the compensation amplitude distribution in each frequency band tends to be balanced. In the asymmetric concentrated type, the compensation is concentrated on the main side of the high asymmetry level reversal frequency point. In the symmetrical diffusion type, the compensation is evenly distributed according to the deviation amplitude ratio of each frequency band, and the polarity of the compensation is independently constrained on both sides of the reversal frequency point.

[0064] The amplitude and phase calibration compensation parameters are output based on the frequency band deviation distribution characteristics and deviation level. For high deviation levels, a step-by-step application strategy is used, dividing the compensation amount for each frequency band into multiple rounds for gradual application to avoid overshoot risk. For medium deviation levels, a two-step application strategy is used, applying half the compensation amount in each of two rounds. For low deviation levels, a one-time application strategy is used to directly output the full compensation amount. The amplitude and phase calibration compensation parameters consist of multiple columns of fields: frequency band partition identifier, sine channel amplitude compensation amount, sine channel phase compensation amount, cosine channel amplitude compensation amount, cosine channel phase compensation amount, and compensation application strategy type. The field structure of the compensation parameter entries allows the calibration execution module to directly read the compensation amount and application rhythm by frequency band partition without needing to backtrack the deviation analysis process. The four compensation amounts and strategy types for each frequency band partition have already been calculated and encapsulated in the parameter output stage. Taking a high-deviation-level direction finder in the 700MHz band as an example, the amplitude compensation of the sinusoidal channel is +0.6dB and the phase compensation is -3.2°, while the amplitude compensation of the cosine channel is -0.4dB and the phase compensation is +1.8°. The above compensation is divided into three rounds, each applying one-third and written with a step-by-step three-round identifier. The compensation of another low-deviation-level direction finder in the same band is directly written in full with an application identifier. The difference in the strategy columns of the two parameter records drives the calibration execution module to adopt different compensation writing rhythms. The polarity of the compensation of each band has been corrected in the cross-band linkage compensation calculation stage according to the deviation polarity flip frequency point. The sign of each band compensation in the output parameters strictly corresponds to the direction of amplitude and phase deviation.

[0065] To implement the Watson-Watt direction finder amplitude and phase calibration method corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects. See also Figure 2 , Figure 2 This application provides a structural block diagram of a Watson-Watt direction finder amplitude and phase calibration device 200, which includes: The reference construction module 201 is used to acquire known direction reference signal data and dual-channel receiver response data, and construct a dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the dual-channel receiver response data; Source tracing analysis module 202 is used to extract deviation factors from the dual-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set, perform joint source tracing of temperature drift and frequency dependence on the candidate deviation factor set to establish a channel deviation coupling relationship spectrum, and perform single-compensation overshoot risk assessment on the candidate deviation factor set to generate an overshoot risk label. The convergence evaluation module 203 is used to perform real-time calibration residual convergence evaluation based on the channel deviation coupling relationship spectrum and the overshoot risk mark to generate deviation classification parameters, determine the deviation level according to the deviation classification parameters, and identify calibration convergence blocking factors from the deviation classification parameters to generate a blocking factor set. The frequency band analysis module 204 is used to classify the deviation of the blocking factor set by frequency band partition to obtain the deviation value of each frequency band, perform amplitude and phase deviation polarity reversal detection on the deviation value of each frequency band to identify the deviation polarity reversal frequency point, and perform distribution pattern recognition based on the deviation polarity reversal frequency point and the deviation value of each frequency band to obtain the deviation distribution type. The calibration output module 205 is used to calculate the cross-band amplitude and phase linkage compensation amount based on the deviation distribution type and the deviation polarity reversal frequency point to generate the frequency band deviation distribution characteristics, and output the amplitude and phase calibration compensation parameters in combination with the frequency band deviation distribution characteristics and the deviation level.

[0066] The Watson-Watt direction finder amplitude and phase calibration device 200 described above can implement the Watson-Watt direction finder amplitude and phase calibration method of the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining contents of this application embodiment can be referred to the contents of the above method embodiments, and will not be repeated in this embodiment.

[0067] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

Claims

1. A Watson-Watt direction finder amplitude and phase calibration method, characterized in that, include: Acquire known direction reference signal data and dual-channel receiver response data, and construct a dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the dual-channel receiver response data; Deviation factors are extracted from the dual-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set. Temperature drift and frequency dependence are jointly traced to establish the channel deviation coupling relationship spectrum. A single-compensation overshoot risk assessment is performed on the candidate deviation factor set to generate an overshoot risk label. Based on the channel deviation coupling spectrum and the overshoot risk marker, a deviation classification parameter is generated by real-time calibration residual convergence evaluation. The deviation level is determined according to the deviation classification parameter. The calibration convergence blocking factors are identified from the deviation classification parameter to generate a blocking factor set. The set of blocking factors is divided into frequency bands and the deviation is classified to obtain the deviation value of each frequency band. The amplitude and phase deviation polarity reversal detection is performed on the deviation value of each frequency band to identify the deviation polarity reversal frequency point. The distribution pattern recognition is performed on the deviation polarity reversal frequency point and the deviation value of each frequency band to obtain the deviation distribution type. Based on the deviation distribution type and the deviation polarity reversal frequency point, the cross-band amplitude and phase linkage compensation amount is calculated to generate the frequency band deviation distribution characteristics. The frequency band deviation distribution characteristics and the deviation level are combined to output the amplitude and phase calibration compensation parameters.

2. The method according to claim 1, characterized in that, The construction of the dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the dual-channel receiver response data includes: The theoretical amplitude and phase responses in known directions are extracted from the reference signal data to generate a theoretical response reference; Based on the theoretical response benchmark, the amplitude deviation and phase deviation are calculated respectively for the dual-channel receiver response data to generate an amplitude-phase deviation pair sequence; The amplitude and phase deviation pairs are subjected to joint drift direction classification to identify common-mode drift segments and differential-mode drift segments, generating a drift direction classification spectrum; The drift direction classification spectrum is fused with the amplitude-phase deviation pair sequence to construct a dual-channel amplitude-phase coupling deviation reference spectrum.

3. The method according to claim 1, characterized in that, The step of establishing a channel deviation coupling relationship spectrum by jointly tracing the source of temperature drift and frequency dependence for the candidate deviation factor set includes: Temperature sensitivity detection is performed on each factor in the candidate deviation factor set to generate a temperature drift sensitive factor set; Frequency scan response analysis is performed on each factor in the candidate bias factor set to generate a frequency-dependent factor set; Perform a cross-correlation strength assessment on the set of temperature drift sensitive factors and the set of frequency dependent factors to generate coupling strength weights; A channel deviation coupling relationship spectrum is established based on the coupling strength weights.

4. The method according to claim 1, characterized in that, The step of performing a single-time compensated overshoot risk assessment on the candidate deviation factor set to generate an overshoot risk label includes: For each factor in the candidate deviation factor set, perform channel response nonlinear slope calculation to generate a nonlinear slope sequence; Based on the nonlinear slope sequence, factors whose slopes exceed the applicable range of linear compensation are identified to form an overshoot candidate factor set; For the overshoot candidate factor set, perform compensation direction reversal risk quantification to generate a reversal risk coefficient; Based on the flip risk coefficient, the candidate deviation factor set is labeled with overshoot risk to generate an overshoot risk label.

5. The method according to claim 1, characterized in that, The step of generating deviation classification parameters based on the channel deviation coupling spectrum and the overshoot risk marker for real-time calibration residual convergence evaluation includes: The iterative compensation response curves of each deviation factor are extracted from the channel deviation coupling relationship spectrum to generate a compensation response set; For each factor in the compensation response set, identify the secondary recovery segment after the first convergence of the residuals to form a recovery segment set; For the set of recovery segments, the ratio of the recovery magnitude to the first convergence amount is calculated in conjunction with the overshoot risk marker to generate a recovery ratio sequence; Based on the recovery ratio sequence, the severity of each deviation factor is ranked to generate deviation grading parameters.

6. The method according to claim 1, characterized in that, The step of classifying the blocking factor set by frequency band partition to obtain the deviation value of each frequency band includes: Perform inter-band amplitude-phase coupling transmission path analysis on each factor in the blockage factor set to generate a transmission path diagram; In the transmission path diagram, cross-frequency band amplitude-phase transmission chains are identified to form a transmission chain length sequence; Based on the transmission chain length sequence, frequency band classification weights are assigned to the blocking factor set to generate a frequency band classification weight table. The deviation value of each frequency band is determined by cumulative calculation according to the frequency band classification weight table.

7. The method according to claim 1, characterized in that, The step of obtaining the deviation distribution type by performing distribution pattern recognition based on the deviation polarity reversal frequency point and the deviation values ​​of each frequency band includes: Based on the deviation polarity reversal frequency point, the deviation values ​​of each frequency band are segmented on both sides of the reversal point to generate a set of deviation segments on both sides. For each flip point in the set of two-sided deviation segments, perform asymmetry calculation to generate a flip asymmetry sequence; Based on the flipped asymmetry sequence, an asymmetry degree classification is performed to generate an asymmetry classification identifier; The type of deviation distribution is determined based on the asymmetric grading identifier.

8. The method according to claim 3, characterized in that, The step of generating a frequency-dependent factor set by performing frequency scan response analysis on each factor in the candidate bias factor set includes: Perform wideband amplitude and phase response simulation on each factor in the candidate deviation factor set to generate a simulation response spectrum; Synchronization detection of amplitude and phase abrupt changes is performed on each frequency point in the simulated response spectrum to generate a set of synchronous abrupt change frequency points; For the set of synchronous mutation frequency points, perform synchronous mutation intensity assessment to generate an amplitude-phase synchronous mutation intensity sequence; Based on the amplitude-phase synchronous mutation intensity sequence, a frequency-dependent factor set is formed by screening factors with high synchronous mutation intensity.

9. The method according to claim 5, characterized in that, The step of calculating the ratio of the recovery amplitude to the first convergence amount and generating a recovery ratio sequence for the recovery segment set in combination with the overshoot risk marker includes: For each factor in the recovery segment, the amplitude residual convergence curve and the phase residual convergence curve are extracted to generate amplitude-phase convergence curve pairs; A convergence rate difference sequence is generated by comparing the amplitude convergence rate and the phase convergence rate of the amplitude and phase convergence curves. The convergence rate difference sequence is combined with the overshoot risk marker to identify factors that cause rate mismatch exceeding the tolerance, forming a mismatch factor set; Based on the degree of mismatch in the set of mismatched factors, the ratio of the recovery magnitude to the first convergence amount is calculated to generate a recovery ratio sequence.

10. A Watson-Watt direction finder amplitude and phase calibration device, characterized in that, include: The reference construction module is used to acquire known direction reference signal data and dual-channel receiver response data, and construct a dual-channel amplitude-phase coupling deviation reference spectrum based on the reference signal data and the dual-channel receiver response data; The source tracing analysis module is used to extract deviation factors from the dual-channel amplitude-phase coupling deviation reference spectrum to form a candidate deviation factor set, perform joint source tracing of temperature drift and frequency dependence on the candidate deviation factor set to establish a channel deviation coupling relationship spectrum, and perform single-compensation overshoot risk assessment on the candidate deviation factor set to generate an overshoot risk label. The convergence evaluation module is used to perform real-time calibration residual convergence evaluation based on the channel deviation coupling relationship spectrum and the overshoot risk marker to generate deviation classification parameters, determine the deviation level according to the deviation classification parameters, and identify calibration convergence blocking factors from the deviation classification parameters to generate a blocking factor set. The frequency band analysis module is used to classify the deviation of the blocking factor set by frequency band partition to obtain the deviation value of each frequency band, perform amplitude and phase deviation polarity reversal detection on the deviation value of each frequency band to identify the deviation polarity reversal frequency point, and perform distribution pattern recognition based on the deviation polarity reversal frequency point and the deviation value of each frequency band to obtain the deviation distribution type. The calibration output module is used to calculate the cross-band amplitude and phase linkage compensation amount based on the deviation distribution type and the deviation polarity reversal frequency point to generate the frequency band deviation distribution characteristics, and output the amplitude and phase calibration compensation parameters by combining the frequency band deviation distribution characteristics and the deviation level.

Images