A calibration method and system for a swept frequency signal generator

Abstract

The present invention discloses a calibration method and system for a swept frequency signal generator, which relates to the field of signal processing technology. The method can comprehensively obtain frequency response information at different locations by setting multiple collection points on a high-voltage long-distance transmission line and collecting power data in real time. In the calibration server, feature extraction is performed on the collected power data to obtain a frequency shift feature vector set, and a standardized digital feature set is obtained through preprocessing. Based on these standardized feature sets, the output frequency offset prediction value △f1 is further calculated, and a preliminary comparative evaluation is performed with the error critical threshold Eth. If the frequency offset error exceeds the threshold, the common-mode spectrum interference suppression mechanism will be triggered, and the output common-mode interference anti-aliasing suppression function △f2 will be calculated, which can significantly improve the frequency calibration accuracy of the swept frequency signal generator, especially in long-distance transmission lines and high-voltage environments, to ensure accurate calibration of the signal.

Description

Technical Field

[0001] The present invention relates to the field of signal processing technology, and in particular to a calibration method and system for a swept frequency signal generator. Background Art

[0002] Calibration methods for swept-frequency signal generators fall within the realm of signal processing technology, particularly frequency response calibration of transmission lines, a key subfield within power system monitoring and control. In modern power transmission systems, especially ultra-high voltage (UHV) transmission lines, frequency sweeps using a swept-frequency signal generator are a common method for assessing line status, monitoring equipment operating conditions, and locating partial discharge (PD). With the continuous expansion of power networks and the increasing complexity of equipment, especially in the context of long-distance, high-voltage transmission lines, the accuracy of frequency response is crucial to the safe operation and efficient management of power systems. To improve the measurement accuracy of these swept-frequency signal generators and their adaptability in complex environments, a suite of refined calibration methods has been gradually developed.

[0003] Currently, traditional swept-frequency signal generator calibration methods primarily focus on compensating for structural frequency offsets, specifically correcting for frequency offsets caused by factors such as uneven path impedance and device coupling. However, with the increasing complexity of power networks and the increase in long-distance transmission lines, these traditional methods face increasing challenges. In particular, in high-voltage, long-distance transmission lines, the asymmetric refraction effect at different line locations, caused by the line's uneven structure and local environmental variations, is often overlooked. Current calibration methods fail to effectively address the impact of these factors on signal frequency shifts, resulting in large frequency offset errors. In particular, the frequency response consistency at remote locations is poor, making it difficult to accurately locate discharge sources and other fault sources.

[0004] The shortcomings of this traditional calibration method, particularly when it neglects the high impedance refractive index (AIHI), significantly reduce the accuracy of frequency offset compensation. At the end of long transmission lines, asymmetric dispersion can cause severe signal distortion and frequency shift, especially at the far end of the frequency response. This can prevent the swept frequency signal from effectively reflecting the true line state, affecting the accuracy of PD source location and overall power system fault detection. This offset error can not only lead to misjudgment of equipment status but also impede subsequent maintenance and optimization efforts due to inaccurate measurements. In severe cases, it can even pose safety risks to the power system. Summary of the Invention

[0005] In view of the deficiencies in the prior art, the present invention provides a calibration method and system for a swept frequency signal generator, which solves the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: comprising the following steps:

[0007] S1. Set up collection points on high-voltage long-distance transmission lines and install power sensors at each collection point to collect power data in real time and transmit the power data to the calibration server;

[0008] S2. Extracting features from the power data in the calibration server to obtain a frequency shift feature vector set, and preprocessing the frequency shift feature vector set to obtain a standardized digital feature set;

[0009] S3. Based on the standardized digital feature set, calculate and output the frequency offset prediction value △f1, calculate and output the frequency offset error value Ef based on the frequency offset prediction value △f1, and set the error critical threshold Eth and the frequency offset error value Ef for preliminary comparative evaluation;

[0010] S4. Based on the preliminary comparative evaluation results, trigger the common-mode spectrum interference suppression mechanism, extract the interference data, and calculate and output the common-mode interference anti-aliasing suppression function △f2 based on the interference data;

[0011] S5. The frequency offset prediction value △f1 and the common-mode interference anti-aliasing suppression function △f2 are fused and calculated to output the frequency correction function fcorr. The difference between the frequency correction function fcorr and the actual frequency response value fobs is calculated. A secondary comparative evaluation is performed based on the difference result and the error critical threshold Eth to determine the frequency calibration status of the swept frequency signal generator.

[0012] Preferably, said S1 includes S11 and S12;

[0013] S11. Set a collection point every 13 km on the high-voltage long-distance transmission line. The collection point includes a near-end node N, a mid-section node M, and an end node F. At the same time, set a power sensor group in each collection point to collect power data in real time.

[0014] The proximal node N is arranged at the substation end of the high-voltage long-distance transmission line;

[0015] The middle section node M is arranged at the grounding pole of the middle section tower base;

[0016] The terminal node F is set at the terminal monitoring point before the load is connected;

[0017] The power sensor group includes GPS, PMU synchronized phasor measurement unit, ADC module and FFT spectrum analyzer;

[0018] The power data includes a synchronously measured voltage V(x) at the line position x, a synchronously measured current I(x) at the line position x, a phase angle Xw(x, f) at the line position x at a frequency f, and a background spectrum S(f) at a frequency f;

[0019] The route position x is obtained through GPS positioning;

[0020] The synchronous measurement voltage V(x) at the line position x and the synchronous measurement current I(x) at the line position x are injected with a swept frequency signal at the frequency f of the swept frequency signal at both ends of the high-voltage long-distance transmission line through a swept frequency signal generator, and the current I and voltage V at different line positions x are synchronously measured by a PMU synchronized phasor measurement unit;

[0021] The phase angle Xw(x, f) at the line position x at the frequency f is obtained by using an ADC module to extract the phase change of each mid-segment node M when injecting the frequency f of the swept frequency signal;

[0022] The background spectrum S(f) at the frequency f is obtained by performing a fast Fourier transform (FFT) on the frequency f signal within the acquisition period using an FFT spectrum analyzer at the terminal node F during the period when the sweep signal generator is turned off.

[0023] S12. Use industrial Ethernet and power sensor groups in the substation to aggregate the power data of all sampling points to the substation. Use 5G industrial routers in the substation to transmit the power data to the calibration server.

[0024] Preferably, said S2 includes S21 and S22;

[0025] S21. Extracting features from the power data in the calibration server to obtain a frequency shift feature vector set;

[0026] The frequency shift feature vector set includes the length impedance gradient Kz(x) of the line position x, the phase density △Xw(x,f) of the line position x at the frequency f, and the background power spectrum density Pc(f) generated by the common mode interference at the frequency f;

[0027] The length impedance gradient Kz(x) at the line position x is obtained by extracting the synchronously measured voltage V(x) and the synchronously measured current I(x) at all line positions x in the power data, performing ratio calculation to obtain the transient impedance Z(x) at the line position x, and then taking the difference between the transient impedance Z(x) at the line position x of adjacent collection points and dividing by the distance to extract the length impedance gradient Kz(x) at the line position x;

[0028] The phase density △Xw(x, f) at the line position x at the frequency f is calculated and extracted by extracting the phase angle Xw(x, f) at the line position x at the frequency f and fitting the rate of change of the phase with distance. The specific algorithm formula is: , where d represents the calculus function, and dx represents the calculus variable of the line position x;

[0029] The background power spectrum density Pc(f) generated by the common-mode interference at the frequency f is obtained by performing smooth curve fitting on the background spectrum S(f) at the frequency f using a quintic B-spline, and the fitting curve is set to the background power spectrum density Pc(f) generated by the common-mode interference at the frequency f;

[0030] S22, preprocessing the frequency shift feature vector set obtained based on feature extraction to obtain a standardized digital feature set;

[0031] The preprocessing uses the Min-Max normalization method to unify the value range of the frequency shift feature vector set and eliminate the dimension effect.

[0032] Preferably, said S3 includes S31;

[0033] S31. Construct a mathematical model to predict the frequency offset that the current frequency sweep signal will generate in the transmission path. Extract the length impedance gradient Kz(x) at line position x and the phase density △Xw(x, f) at line position x at frequency f from the standardized digital feature set and input them into the mathematical model. Calculate and output the predicted frequency offset △f1, which measures the ability of the frequency sweep signal to offset the structural offset error along the path by pre-adjusting the frequency at the transmitting end.

[0034] The frequency offset prediction value Δf1 is calculated and outputted by the following mathematical model;

[0035] ;

[0036] Where △f1(f) represents the predicted frequency offset at frequency f, L represents the total length of the high-voltage long-distance transmission line, sin represents the sine function, e represents the exponential function, dx represents the calculus variable at the line location x, and u represents the energy dissipation factor.

[0037] Preferably, said S3 further includes S32 and S33;

[0038] S32, performing dimensionless processing based on the actual observed response frequency fobs(f) extracted in real time at the frequency f and the ideal response frequency fideal(f) set by the user at the frequency f to eliminate the unit dimension, performing error calculation on the frequency offset prediction value △f1(f) at the frequency f, and obtaining a frequency shift error value Ef;

[0039] The frequency shift error value Ef is calculated and outputted by the following algorithm formula:

[0040] ;

[0041] Where, Ef(f) represents the frequency shift error value at frequency f;

[0042] S33. Based on the user setting of the critical error threshold Eth, a preliminary comparison and evaluation is performed between the frequency shift error value Ef(f) obtained in real time at the frequency f and the critical error threshold Eth to determine the frequency shift of the pre-adjusted frequency of the swept signal generator. The specific evaluation contents are as follows:

[0043] When the frequency shift error value Ef(f) at frequency f is ≤ the critical error threshold Eth, it means that the frequency shift of the pre-adjusted frequency of the sweep signal generator is normal and no intervention is required;

[0044] When the frequency shift error value Ef(f) at frequency f is greater than the critical error threshold Eth, it indicates that the frequency shift of the pre-adjusted frequency of the sweep signal generator is abnormal, and the common mode spectrum interference suppression mechanism is triggered.

[0045] Preferably, the S4 includes S41;

[0046] S41. Triggering a common-mode spectrum interference suppression mechanism through preliminary comparative evaluation. The common-mode spectrum interference suppression mechanism extracts the background power spectrum density Pc(f) generated by the common-mode interference at frequency f and calculates and outputs the convolution energy density X(x, f) at the line position x at frequency f.

[0047] The convolution energy density X(x, f) of the line position x at the frequency f is calculated and output by the following algorithm formula;

[0048] ;

[0049] Where exp represents the natural exponential function, xs(f) represents the estimated spatial position of the interference source at frequency f, It represents the spatial propagation attenuation factor of interference energy at frequency f, describing how quickly energy decays with distance along the path. The specific value is set by the user and is dimensionless.

[0050] Preferably, the S4 further includes S42;

[0051] S42, calculating and outputting a common-mode interference anti-aliasing suppression function △f2 based on the background power spectrum density Pc(f) generated by the common-mode interference at the frequency f and the convolution energy density X(x, f) at the line position x at the frequency f;

[0052] The common mode interference anti-aliasing suppression function Δf2 is calculated and outputted by the following algorithm formula:

[0053] ;

[0054] Where △f2(f) represents the common mode interference anti-aliasing suppression function at frequency f, Indicates the path energy attenuation rate.

[0055] Preferably, the S5 includes S51;

[0056] S51, performing a fusion calculation based on the common-mode interference anti-aliasing suppression function △f2(f) at frequency f and the frequency offset prediction value △f1(f) at frequency f, and outputting a frequency correction function fcorr as the actual control compensation term of the swept frequency signal generator, which directly acts on the control parameters of the swept frequency signal source;

[0057] The frequency correction function fcorr is calculated and outputted by the following algorithm formula:

[0058] ;

[0059] Where fcorr(f) represents the frequency correction function of frequency f.

[0060] Preferably, the S5 further includes S52;

[0061] S52. Calculate the difference between the frequency correction function fcorr(f) currently controlling the frequency f of the swept frequency signal generator and the actual observed response frequency fobs(f) at the frequency f, and perform a secondary comparative evaluation of the difference calculation result with the error critical threshold Eth to analyze the calibration of the swept frequency signal generator frequency f after correction and compensation. The specific evaluation content is as follows:

[0062] when When the error threshold Eth is less than or equal to the critical threshold, it indicates that the sweep signal generator has been calibrated successfully and no intervention is required.

[0063] when When the error threshold Eth is greater than the critical threshold, it indicates that the calibration of the sweep signal generator is abnormal, and the iterative adjustment mechanism is executed;

[0064] The iterative adjustment mechanism collects the background spectrum S(f, x) at the line position x at the frequency f after correction and compensation by the sweep signal generator, and reconstructs the background power spectrum density Pc(f) generated by the common mode interference at the frequency f. The specific reconstruction method is:

[0065] Pc(f)=Env(S(f,x)), where Env represents the envelope, which is a smooth depiction of the local amplitude changes of the signal in the time domain or frequency domain. In spectrum analysis, the envelope usually refers to the average value of each frequency component in the signal spectrum. The common-mode interference anti-aliasing suppression function △f2 and the frequency correction function fcorr are recalculated based on the background power spectral density Pc(f) generated by the common-mode interference at the reconstructed frequency f, and an iterative secondary comparative evaluation is performed until the calibration is successful and the iteration is stopped.

[0066] A calibration system for a frequency sweep signal generator includes a frequency shift data acquisition module, a frequency shift data processing module, a frequency shift error analysis module, an interference suppression module, and a final compensation module;

[0067] The frequency shift data acquisition module collects power data in real time by setting up acquisition points on the high-voltage long-distance transmission line and setting up power sensors in each acquisition point, and transmits the power data to the calibration server;

[0068] The frequency shift data processing module extracts features from the power data in the calibration server to obtain a frequency shift feature vector set, and pre-processes the frequency shift feature vector set to obtain a standardized digital feature set;

[0069] The frequency shift error analysis module calculates and outputs a frequency shift prediction value △f1 based on a standardized digital feature set, calculates and outputs a frequency shift error value Ef based on the frequency shift prediction value △f1, and sets an error critical threshold Eth to perform preliminary comparative evaluation with the frequency shift error value Ef;

[0070] The interference suppression module triggers a common-mode spectrum interference suppression mechanism based on a preliminary comparison and evaluation result, extracts interference data, and calculates and outputs a common-mode interference anti-aliasing suppression function Δf2 based on the interference data;

[0071] The final compensation module fuses the frequency offset prediction value △f1 and the common-mode interference anti-aliasing suppression function △f2 to output a frequency correction function fcorr, performs a difference calculation based on the frequency correction function fcorr and the actual frequency response value fobs, and performs a secondary comparative evaluation based on the difference result and the error critical threshold Eth to determine the frequency calibration status of the swept signal generator.

[0072] The present invention provides a method and system for calibrating a swept frequency signal generator, which has the following beneficial effects:

[0073] (1) This method can comprehensively obtain frequency response information at different locations by setting up multiple collection points on high-voltage long-distance transmission lines and collecting power data in real time. In the calibration server, the collected power data is feature extracted to obtain a frequency shift feature vector set, and a standardized digital feature set is obtained through preprocessing. Based on these standardized feature sets, the output frequency offset prediction value △f1 is further calculated and preliminarily compared and evaluated with the error critical threshold Eth. If the frequency offset error exceeds the threshold, the common-mode spectrum interference suppression mechanism will be triggered, and the output common-mode interference anti-aliasing suppression function △f2 will be calculated, which can significantly improve the frequency calibration accuracy of the swept signal generator, especially in long-distance transmission lines and high-voltage environments, ensuring accurate signal calibration.

[0074] (2) This method introduces an asymmetric high-impedance refractive index, which can dynamically compensate for the frequency offset caused by the asymmetric refraction effect. Especially in long-distance transmission lines, the impedance gradient, phase density, and common-mode interference on the path may cause the frequency response to shift. The present invention triggers the common-mode spectrum interference suppression mechanism after preliminary comparative evaluation, further extracts interference data, and calculates the common-mode interference anti-aliasing suppression function △f2 based on the interference data, thereby fine-tuning the frequency offset. Combined with the predicted frequency offset △f1, it can effectively handle interference fluctuations in the path, enhance the adaptability of the system in complex and high-interference environments, and ensure accurate frequency response.

[0075] (3) This method uses the frequency response of long-distance transmission lines to identify possible errors in the location of local discharge sources caused by interference and frequency shift. The method of the present invention first performs a preliminary frequency correction using the frequency correction function fcorr, and further fine-tunes it during the secondary comparative evaluation. The accuracy of the frequency response is ensured by performing a differential calculation between the frequency-corrected result and the actual frequency response value, and performing a secondary comparative evaluation with the error critical threshold Eth. When the frequency correction result does not meet the calibration requirements, an iterative adjustment mechanism is automatically executed to achieve accurate secondary correction by reconstructing the background spectrum and recalculating the common-mode interference anti-aliasing suppression function △f2. This optimized compensation can improve the location accuracy of the local discharge source, reduce fault detection time, and ensure the high reliability and safety of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 A schematic diagram of the steps of a calibration method for a swept frequency signal generator according to the present invention;

[0077] Figure 2 A schematic diagram of a calibration system flow for a swept frequency signal generator according to the present invention;

[0078] Figure 3 Schematic diagram of collection point settings. DETAILED DESCRIPTION

[0079] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0080] Example 1: Please refer to Figure 1 and Figure 3 The present invention provides a calibration method for a swept frequency signal generator. To achieve the above object, the present invention is implemented by the following technical solution: comprising the following steps:

[0081] S1. Set up collection points on high-voltage long-distance transmission lines and install power sensors at each collection point to collect power data in real time and transmit the power data to the calibration server;

[0082] S2. Extracting features from the power data in the calibration server to obtain a frequency shift feature vector set, and preprocessing the frequency shift feature vector set to obtain a standardized digital feature set;

[0083] S3. Based on the standardized digital feature set, calculate and output the frequency offset prediction value △f1, calculate and output the frequency offset error value Ef based on the frequency offset prediction value △f1, and set the error critical threshold Eth and the frequency offset error value Ef for preliminary comparative evaluation;

[0084] S4. Based on the preliminary comparative evaluation results, trigger the common-mode spectrum interference suppression mechanism, extract the interference data, and calculate and output the common-mode interference anti-aliasing suppression function △f2 based on the interference data;

[0085] S5. The frequency offset prediction value △f1 and the common-mode interference anti-aliasing suppression function △f2 are fused and calculated to output the frequency correction function fcorr. The difference between the frequency correction function fcorr and the actual frequency response value fobs is calculated. A secondary comparative evaluation is performed based on the difference result and the error critical threshold Eth to determine the frequency calibration status of the swept frequency signal generator.

[0086] In this embodiment, the method establishes multiple collection points along a high-voltage transmission line and equips it with power sensors to collect line data in real time, enabling comprehensive acquisition of frequency response information at different locations. After transmitting this data to a calibration server, feature extraction and normalization are performed to eliminate dimensionality effects and ensure data consistency. Based on this, a frequency offset prediction value Δf1 is further calculated using a frequency shift feature set and a standardized digital feature set. This is then compared and evaluated against a set critical error threshold, Eth, to assess the frequency response error. If a significant deviation is detected, the common-mode spectrum interference suppression mechanism is triggered, extracting interference data and calculating a common-mode interference anti-aliasing suppression function Δf2 to further correct the signal. Finally, the output frequency correction function fcorr is calculated by combining the frequency offset prediction value Δf1 and the common-mode interference anti-aliasing suppression function Δf2. The difference between this value and the actual frequency response value fobs is then compared and evaluated again against the critical error threshold, Eth, to ensure the signal generator's frequency has been accurately calibrated. This method enables multi-stage, dynamic calibration of the swept frequency signal generator, improving both frequency response accuracy and adaptability in high-interference environments. This method, particularly when dealing with the asymmetric dispersion effects of long-distance, high-voltage transmission lines, further optimizes signal frequency calibration by introducing an asymmetric high-impedance refractive index, effectively reducing frequency errors. This method significantly improves the reliability and accuracy of swept-frequency signal generators, ensuring the accuracy of localizing partial discharge sources, enhancing fault diagnosis capabilities, reducing equipment maintenance costs, and enhancing power safety and stability.

[0087] Example 2: Please refer to Figure 1 and Figure 3 ,Specifically: S1 includes S11 and S12;

[0088] S11. Set up a collection point every 13 km on the high-voltage long-distance transmission line. The collection points include the near-end node N, the mid-section node M, and the end node F. At the same time, set up a power sensor group at each collection point to collect power data in real time.

[0089] The near-end node N is set at the substation end of the high-voltage long-distance transmission line to record the original state of the frequency of the transmitting signal source;

[0090] The middle section node M is set at the grounding pole of the middle section tower base to obtain the phase and impedance fluctuation of the middle path;

[0091] The terminal node F records the final spectrum distortion by setting up a terminal monitoring point before the load is connected;

[0092] The power sensor group includes GPS, PMU synchronized phasor measurement unit, ADC module and FFT spectrum analyzer;

[0093] The power data includes a synchronously measured voltage V(x) at the line position x, a synchronously measured current I(x) at the line position x, a phase angle Xw(x, f) at the line position x at a frequency f, and a background spectrum S(f) at a frequency f;

[0094] The route position x is obtained through GPS positioning;

[0095] The synchronous measurement voltage V(x) at line position x and the synchronous measurement current I(x) at line position x are injected into both ends of the high-voltage long-distance transmission line through a swept frequency signal generator at a frequency f, and the current I and voltage V at different line positions x are synchronously measured through the PMU synchronized phasor measurement unit;

[0096] The phase angle Xw(x, f) at the line position x at the frequency f is obtained by using the ADC module to extract the phase change of each mid-segment node M when injecting the frequency f of the swept frequency signal;

[0097] The background spectrum S(f) at the frequency f is obtained by performing a fast Fourier transform (FFT) on the frequency f signal within the acquisition period using an FFT spectrum analyzer at the terminal node F when the sweep signal generator is turned off.

[0098] S12. Use industrial Ethernet and power sensor groups in the substation to aggregate the power data of all sampling points to the substation. Use 5G industrial routers in the substation to transmit the power data to the calibration server.

[0099] In this embodiment, the method monitors the changes in frequency response in real time by synchronously measuring voltage, current, phase angle and spectrum data through the power sensor group at each collection point, including the proximal node N, the mid-section node M and the end node F. Through the set GPS, PMU synchronous phasor measurement unit, ADC module and FFT spectrum analyzer, the voltage, current, phase change and background spectrum information of the line can be accurately obtained, and the collected data is transmitted to the calibration server for processing via industrial Ethernet and 5G industrial router. In a specific embodiment, the proximal node N is used to record the original state of the frequency of the transmitting signal source, the mid-section node M obtains the phase and impedance fluctuations of the intermediate path, and the end node F monitors the final spectrum distortion. This setting can fully cover the frequency response information from the signal source to the terminal, ensuring that the calibration can track and correct the frequency offset in real time. In the calibration server, by extracting and preprocessing the collected power data, a standardized digital feature set is obtained to provide data support for the subsequent frequency offset prediction calculation. This method can effectively improve the accuracy of the swept signal generator calibration. By precisely collecting and transmitting frequency response data, combined with data processing and calculations in a calibration server, frequency offset can be accurately predicted and dynamically compensated based on real-time data. This real-time calibration not only enhances frequency offset prediction capabilities but also mitigates external interference and signal distortion in long-distance, high-voltage transmission lines, improving the stability and accuracy of the frequency sweep signal. This method offers greater accuracy in applications such as localizing partial discharge sources and monitoring equipment, reducing equipment fault location time and enhancing the operational reliability and safety of power systems.

[0100] Example 3: Please refer to Figure 1 , specifically: S2 includes S21 and S22;

[0101] S21. Extracting features from the power data in the calibration server to obtain a frequency shift feature vector set;

[0102] The frequency shift feature vector set includes the length impedance gradient Kz(x) at the line position x, the phase density △Xw(x, f) at the line position x at the frequency f, and the background power spectrum density Pc(f) generated by the common mode interference at the frequency f;

[0103] The length impedance gradient Kz(x) of line position x is obtained by extracting the synchronously measured voltage V(x) and the synchronously measured current I(x) of all line positions x in the power data, performing ratio calculation to obtain the transient impedance Z(x) of line position x. The length impedance gradient Kz(x) of line position x is then extracted by subtracting the transient impedance Z(x) of the line position x at adjacent collection points and dividing by the distance.

[0104] The phase density △Xw(x, f) at the line position x at the frequency f is calculated by extracting the phase angle Xw(x, f) at the line position x at the frequency f and fitting the rate of change of the phase with distance. The specific algorithm formula is: , where d represents the calculus function, and dx represents the calculus variable of the line position x;

[0105] The background power spectrum density Pc(f) generated by the common-mode interference at frequency f is smoothed by fitting the background spectrum S(f) at frequency f using a quintic B-spline. The fitting curve is set to the background power spectrum density Pc(f) generated by the common-mode interference at frequency f, which expresses the background interference trend rather than instantaneous burst interference.

[0106] S22, preprocessing the frequency shift feature vector set obtained based on feature extraction to obtain a standardized digital feature set;

[0107] The preprocessing uses the Min-Max normalization method to unify the value range of the frequency shift feature vector set and eliminate the dimension effect.

[0108] In this embodiment, the method constructs a frequency shift feature vector set by extracting key information from the power data. The vector set includes the length impedance gradient Kz(x) at the line position x, the phase density △Xw(x,f) at the line position x at the frequency f, and the background power spectrum density Pc(f) generated by the common-mode interference at the frequency f. By synchronously measuring the voltage and current in the power data and combining the transient impedance Z(x) at the line position, the length impedance gradient Kz(x) at the line position x is calculated to further evaluate the structural disturbance of the line. In addition, by extracting the phase angle Xw(x,f) at the frequency f and fitting the rate of change of the phase with distance, the phase density △Xw(x,f) can be accurately calculated. In order to accurately describe the interference trend, the quintic B-spline B-spline is used to perform a smooth curve fitting on the background spectrum S(f) at the frequency f, thereby obtaining the background power spectrum density Pc(f) generated by the common-mode interference, avoiding the interference of instantaneous burst interference on the analysis results. After feature extraction is complete, step S22 preprocesses the frequency shift feature vector set and uses the Min-Max normalization method to unify the value range of the feature vector set, effectively eliminating dimensionality effects and ensuring data consistency and comparability. Through this processing method, the calibration server can accurately obtain a standardized digital feature set, providing reliable data support for subsequent frequency offset prediction, error calculation, and interference suppression. This method achieves the goal of high-precision calibration by comprehensively collecting line data, accurately extracting frequency shift features, and standardizing the features. By eliminating interference and errors, the frequency compensation and correction process is optimized, significantly improving the frequency offset prediction accuracy and interference suppression capability of the swept frequency signal generator in long-distance transmission lines. This not only improves the stability and reliability of equipment operation, but also provides more accurate data support for fault diagnosis and local discharge source location in high-voltage long-distance transmission systems, ultimately improving the safety and efficient operation of the power system.

[0109] Example 4: Please refer to Figure 1 , specifically: S3 includes S31;

[0110] S31. Construct a mathematical model to predict the frequency offset that the current frequency sweep signal will generate in the transmission path. Extract the length impedance gradient Kz(x) at line position x and the phase density △Xw(x, f) at line position x at frequency f from the standardized digital feature set and input them into the mathematical model. Calculate and output the predicted frequency offset △f1, which measures the ability of the frequency sweep signal to offset the structural offset error along the path by pre-adjusting the frequency at the transmitting end.

[0111] The frequency offset prediction value △f1 is calculated and output by the following mathematical model;

[0112] ;

[0113] Where △f1(f) represents the predicted frequency offset at frequency f, L represents the total length of the high-voltage long-distance transmission line, sin represents the sine function, e represents the exponential function, dx represents the calculus variable at the line position x, and u represents the energy dissipation factor. The specific value is set by the user and is dimensionless, simulating the energy attenuation of the swept frequency signal during line transmission.

[0114] The calculation logic and physical meaning of the formula divides the entire transmission line into countless small segments dx, calculates the local frequency offset caused by structural disturbances and phase fluctuations in each segment, and performs a weighted integral on these local offsets to obtain the overall frequency offset prediction △f1(f);

[0115] The frequency offset prediction △f1(f) at frequency f is based on factors such as line structural disturbances, phase density fluctuations, and path attenuation. It predicts the frequency deviation of the swept-frequency signal caused by the physical structure during transmission under ideal conditions without interference. In optical fibers, if the medium is unevenly distributed, the light wave will deviate from the original path. In this method, impedance fluctuations and phase changes in the line can cause the frequency center point to be structurally pulled or compressed. This deviation is determined by the physical structure and is not random.

[0116] That is, each small segment dx will introduce a small frequency deviation due to its own impedance change Kz and phase change ΔXw, using the exponential factor Simulate the path attenuation characteristics. The farther the segment, the smaller the impact. Finally, through integration, these effects are summed up to form the overall frequency offset prediction △f1(f);

[0117] The core structure of the formula, where the length impedance gradient Kz(x) at line position x represents the structural disturbance intensity of this line segment. If the impedance mutation is more severe, the signal will be more strongly refracted or reflected;

[0118] Describes the propagation nonlinearity of this segment. If the phase density is large and the bending is severe, the path will experience frequency compression or expansion, resulting in a center frequency offset.

[0119] Exponential Factor is the path distance attenuation weight. The closer to the signal source, the greater the impact. Although there is interference at farther points, the signal has been weakened and its offset contribution is reduced.

[0120] The logic of integration is essentially to construct a frequency offset response function. The output is the offset that naturally occurs in this line due to the path structure at the frequency f. This offset is not random but structural and physical, and must be actively considered during compensation.

[0121] S3 also includes S32 and S33;

[0122] S32, performing dimensionless processing based on the actual observed response frequency fobs(f) extracted in real time at the frequency f and the ideal response frequency fideal(f) set by the user at the frequency f to eliminate the unit dimension, performing error calculation on the frequency offset prediction value △f1(f) at the frequency f, and obtaining a frequency shift error value Ef;

[0123] The frequency shift error value Ef is calculated and output by the following algorithm formula;

[0124] ;

[0125] Where, Ef(f) represents the frequency shift error value at frequency f;

[0126] Among them, the ideal response frequency fideal(f) at frequency f is the goal of the entire control system;

[0127] The predicted frequency shift at frequency f, △f1(f), is how much the structural prediction tells us to shift;

[0128] is the frequency response that should be observed, i.e. the target after compensation;

[0129] The actual observed response frequency fobs(f) at frequency f is the frequency response we actually observe;

[0130] so, The difference between the two is the "error remaining after structural compensation". If the ideal response frequency fideal(f) at frequency f is not introduced, the error loses its meaning relative to the target frequency.

[0131] S33. Setting a critical error threshold Eth based on the user. The critical error threshold Eth is set based on how much the user wants the frequency response error to deviate from within an acceptable range. A preliminary comparison and evaluation is then performed between the frequency shift error value Ef(f) obtained in real time at the frequency f and the critical error threshold Eth to determine the frequency shift of the pre-adjusted frequency of the swept signal generator. The specific evaluation contents are as follows:

[0132] When the frequency shift error value Ef(f) at frequency f is ≤ the critical error threshold Eth, it means that the frequency shift of the pre-adjusted frequency of the sweep signal generator is normal and no intervention is required;

[0133] When the frequency shift error value Ef(f) at frequency f is greater than the critical error threshold Eth, it indicates that the frequency shift of the pre-adjusted frequency of the sweep signal generator is abnormal, and the common mode spectrum interference suppression mechanism is triggered.

[0134] In this embodiment, this method, based on the length impedance gradient Kz(x) and phase density ΔXw(x,f) in a standardized digital feature set, predicts the frequency offset Δf1 generated by the swept-frequency signal along the transmission path, effectively offsetting frequency offset errors caused by the path structure. A mathematical model is used to calculate the frequency offset impact of each small segment of the line path dx, and these local frequency offsets are accumulated to obtain the final frequency offset prediction. This process takes into account structural perturbations, propagation nonlinearities, and path distance attenuation, ensuring that the calibration process actively compensates for structural offset errors. Based on this, step S32 calculates the error between the observed frequency fobs and the ideal frequency fideal in real time to obtain a frequency offset error value Ef. This value is then compared and evaluated against a user-set critical error threshold Eth. If the frequency offset error value is less than or equal to the threshold, the frequency offset is normal and no system intervention is required. If the frequency offset error value is greater than the threshold, the system triggers the common-mode spectrum interference suppression mechanism, further extracting interference data and performing compensation to ensure that the final frequency correction result achieves the desired accuracy. This method enables precise frequency calibration of swept-frequency signal generators in high-voltage, long-distance transmission lines, addressing structural offsets and external interference. This calibration process not only dynamically adjusts frequency offset but also monitors and automatically corrects frequency errors in real time. Particularly in long-distance transmission lines, this method effectively reduces frequency offset errors caused by asymmetric dispersion and interference, significantly improving the calibration accuracy and stability of the swept-frequency signal. Ultimately, this provides more precise localization of local discharge sources and fault diagnosis, enhancing the reliability and safety of power systems.

[0135] Example 5: Please refer to Figure 1 , specifically: S4 includes S41;

[0136] S41. Trigger the common-mode spectrum interference suppression mechanism through preliminary comparative evaluation. The common-mode spectrum interference suppression mechanism extracts the background power spectrum density Pc(f) generated by the common-mode interference at frequency f, calculates and outputs the convolution energy density X(x, f) at the line position x at frequency f, and uses this to analyze whether the common-mode interference is concentrated at certain key points along the path.

[0137] The convolution energy density X(x, f) at the line position x at the frequency f is calculated and output by the following algorithm formula;

[0138] ;

[0139] Where exp represents the natural exponential function, xs(f) represents the estimated spatial position of the interference source at frequency f, The spatial propagation attenuation factor of the interference energy at frequency f, describing how quickly the energy decays with distance along the path. The specific value is set by the user and is dimensionless.

[0140] Represents the spatial propagation kernel function, which describes the distribution attenuation of interference in space;

[0141] The convolution energy density X(x, f) at the line position x at the frequency f is used to quantify the spatial diffusion trend of the interference energy at a certain frequency point in the path, that is, the actual intensity of the interference energy at the frequency f at the line position x.

[0142] S4 also includes S42;

[0143] S42. Calculate and output a common-mode interference anti-aliasing suppression function △f2 based on the background power spectrum density Pc(f) generated by the common-mode interference at frequency f and the convolution energy density X(x, f) at the line position x at frequency f. Estimate the interference suppression frequency shift at the acquisition point line position x on the entire line, which is used to actively perform frequency compression processing in frequency bands with large errors.

[0144] The common mode interference anti-aliasing suppression function △f2 is calculated and output by the following algorithm formula;

[0145] ;

[0146] Where △f2(f) represents the common mode interference anti-aliasing suppression function at frequency f, Represents the path energy attenuation rate, which is set by the user and dimensionless. It is used to simulate the impact of path loss on interference weight.

[0147] The core logic of the formula is that it is essentially an interference intensity harmonic suppression integral. For each line position x in the path, the suppression contribution to the frequency offset is calculated and then averaged to form the final frequency offset estimate.

[0148] Numerator: Provides the interference source strength at frequency f along the entire path, representing the original energy of the external interference on the signal. If the frequency is a harmonic of a device, this value will be large. At low interference frequencies, this value approaches 0.

[0149] Denominator: Downweights path segments with high interference density to prevent concentrated interference from overweighting compensation. A larger convolution energy density X(x, f) at line location x at frequency f indicates stronger interference at that location. A larger denominator and a smaller overall value indicate that the more interference there is, the more its contribution to frequency offset is suppressed.

[0150] Represents the distance attenuation factor, which simulates the natural attenuation trend of the interference signal propagating in the line space path. The farther the line position x is, the smaller the weight is to avoid the weak interference at a long distance dominating the compensation estimation;

[0151] Finally, the interference effect of the entire path on the current frequency point is weighted averaged to form the final frequency compensation value.

[0152] In this embodiment, the method triggers the common-mode spectrum interference suppression mechanism through preliminary comparative evaluation, extracts the common-mode interference background power spectral density Pc(f) at frequency f, and then calculates the convolution energy density X(x, f) at line position x at frequency f. This density quantifies the diffusion trend of interference along the transmission path, helping to identify whether the interference is concentrated at certain key points. By combining the natural exponential function and the spatial propagation kernel function, this method can effectively quantify the contribution of interference sources to frequency offset, providing an accurate basis for subsequent compensation. In S42, based on the calculated background power spectral density Pc(f) and convolution energy density X(x, f), the common-mode interference anti-aliasing suppression function Δf2 is further calculated. This function harmonizes the suppression integral based on the energy attenuation rate and interference intensity of the path, and forms the final frequency shift compensation value by weightedly averaging the interference contribution of each line position in the path. In this way, path segments with higher interference intensity are downweighted to avoid overcompensation of areas with higher interference density, thereby ensuring the accuracy of frequency correction. The implementation of this method significantly improves the calibration accuracy of the swept frequency signal generator. In the complex environments of high-voltage, long-distance transmission lines, especially when faced with external common-mode interference and path heterogeneity, this invention accurately calculates and suppresses interference, reducing frequency offset errors caused by interference. By actively suppressing high-interference frequency bands, the consistency of frequency response is effectively improved, the accuracy of localizing partial discharge sources is enhanced, and equipment fault detection and power system reliability are enhanced. This precise interference suppression and frequency correction improves overall system performance, reduces the impact of errors on calibration results, and enhances the operational stability and safety of the power system.

[0153] Example 6: Please refer to Figure 1 , specifically: S5 includes S51;

[0154] S51, performing a fusion calculation based on the common-mode interference anti-aliasing suppression function △f2(f) at frequency f and the frequency offset prediction value △f1(f) at frequency f, and outputting a frequency correction function fcorr as the actual control compensation term of the swept frequency signal generator, which directly acts on the control parameters of the swept frequency signal source;

[0155] The frequency correction function fcorr is calculated and outputted by the following algorithm formula;

[0156] ;

[0157] Where fcorr(f) represents the frequency correction function of frequency f.

[0158] S5 also includes S52;

[0159] S52. Calculate the difference between the frequency correction function fcorr(f) currently controlling the frequency f of the swept frequency signal generator and the actual observed response frequency fobs(f) at the frequency f, and perform a secondary comparative evaluation of the difference calculation result with the error critical threshold Eth to analyze the calibration of the swept frequency signal generator frequency f after correction and compensation. The specific evaluation content is as follows:

[0160] when When the error threshold Eth is less than or equal to the critical threshold, it indicates that the sweep signal generator has been calibrated successfully and no intervention is required.

[0161] when When the error threshold Eth is greater than the critical threshold, it indicates that the calibration of the sweep signal generator is abnormal, and the iterative adjustment mechanism is executed;

[0162] The iterative adjustment mechanism reconstructs the background power spectrum density Pc(f) generated by the common-mode interference at frequency f by collecting the background spectrum S(f, x) at the line position x at frequency f after correction and compensation by the swept signal generator. The specific reconstruction method is:

[0163] Pc(f)=Env(S(f,x)), where Env represents the envelope, which is a smooth depiction of the local amplitude changes of the signal in the time domain or frequency domain. In spectrum analysis, the envelope usually refers to the average value of each frequency component in the signal spectrum. The common-mode interference anti-aliasing suppression function △f2 and the frequency correction function fcorr are recalculated based on the background power spectral density Pc(f) generated by the common-mode interference at the reconstructed frequency f, and an iterative secondary comparative evaluation is performed until the calibration is successful and the iteration is stopped.

[0164] In this embodiment, the method fuses the common-mode interference anti-aliasing suppression function Δf2(f) with the frequency offset prediction Δf1(f) to obtain a frequency correction function fcorr. This correction function, serving as the actual control compensation term for the swept-frequency signal generator, directly acts on the control parameters of the signal source, effectively offsetting the frequency offset caused by the path structure and the impact of common-mode interference on the frequency response. In step S52, the calibration process is further refined. By calculating the difference between the frequency correction function fcorr at the current frequency f and the actual observed response frequency fobs, and comparing and evaluating it with the user-set error threshold Eth, the system can determine whether the swept-frequency signal generator has been successfully calibrated. When the error is less than or equal to the set threshold, calibration is completed; when the error exceeds the threshold, an iterative adjustment mechanism is triggered. This mechanism collects the frequency-corrected background spectrum S(f,x), reconstructs the background power spectral density Pc(f), and uses envelope smoothing Env to remove noise from the signal. It further optimizes the common-mode interference anti-aliasing suppression function Δf2 and the frequency correction function fcorr to ensure the system's calibration accuracy. Through this method, the swept-frequency signal generator can achieve precise frequency correction and dynamically adjust in the face of complex power network interference, ensuring the accuracy of the frequency response. The introduction of the iterative adjustment mechanism greatly enhances the system's adaptability, allowing frequency errors to be quickly corrected even in severe or complex interference environments, avoiding the risk of calibration failure. Ultimately, this method significantly improves the frequency response consistency and accuracy of the swept-frequency signal generator in high-voltage transmission lines, enhances the power system's fault location capabilities and equipment monitoring accuracy, and ensures the safe operation and reliability of the power system.

[0165] Example 7: Please refer to Figure 1 and Figure 2 , a calibration system for a swept frequency signal generator, comprising a frequency shift data acquisition module, a frequency shift data processing module, a frequency shift error analysis module, an interference suppression module and a final compensation module;

[0166] The frequency shift data acquisition module collects power data in real time by setting up collection points on high-voltage long-distance transmission lines and power sensors in each collection point, and transmits the power data to the calibration server;

[0167] The frequency shift data processing module extracts features from the power data in the calibration server to obtain a frequency shift feature vector set, and pre-processes the frequency shift feature vector set to obtain a standardized digital feature set;

[0168] The frequency shift error analysis module calculates and outputs the frequency shift prediction value △f1 based on the standardized digital feature set, and calculates and outputs the frequency shift error value Ef based on the frequency shift prediction value △f1. At the same time, the error critical threshold Eth is set to perform preliminary comparative evaluation with the frequency shift error value Ef;

[0169] The interference suppression module triggers the common-mode spectrum interference suppression mechanism based on the preliminary comparative evaluation results, extracts the interference data, and calculates and outputs the common-mode interference anti-aliasing suppression function △f2 based on the interference data;

[0170] The final compensation module fuses the frequency offset prediction value △f1 and the common-mode interference anti-aliasing suppression function △f2 to output the frequency correction function fcorr. The difference between the frequency correction function fcorr and the actual frequency response value fobs is calculated. A secondary comparative evaluation is performed based on the difference result and the error critical threshold Eth to determine the frequency calibration status of the swept signal generator.

[0171] While the embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions, and alterations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. A method for calibrating a swept frequency signal generator, characterized in that: The following steps are involved: S1. Set up collection points on high-voltage long-distance transmission lines and install power sensors at each collection point to collect power data in real time and transmit the power data to the calibration server; S2. Extracting features from the power data in the calibration server to obtain a frequency shift feature vector set, and preprocessing the frequency shift feature vector set to obtain a standardized digital feature set; S3. Based on the standardized digital feature set, calculate and output the frequency offset prediction value △f1, calculate and output the frequency offset error value Ef based on the frequency offset prediction value △f1, and set the error critical threshold Eth and the frequency offset error value Ef for preliminary comparative evaluation; S4. Based on the preliminary comparative evaluation results, trigger the common-mode spectrum interference suppression mechanism, extract the interference data, and calculate and output the common-mode interference anti-aliasing suppression function △f2 based on the interference data; Said S4 includes S41; S41. Triggering a common-mode spectrum interference suppression mechanism through preliminary comparative evaluation. The common-mode spectrum interference suppression mechanism extracts the background power spectrum density Pc(f) generated by the common-mode interference at frequency f and calculates and outputs the convolution energy density X(x, f) at the line position x at frequency f. The convolution energy density X(x, f) of the line position x at the frequency f is calculated and output by the following algorithm formula; ； Where exp represents the natural exponential function, xs(f) represents the estimated spatial position of the interference source at frequency f, Represents the spatial propagation attenuation factor of the interference energy at frequency f; Said S4 also includes S42; S42, calculating and outputting a common-mode interference anti-aliasing suppression function △f2 based on the background power spectrum density Pc(f) generated by the common-mode interference at the frequency f and the convolution energy density X(x, f) at the line position x at the frequency f; The common mode interference anti-aliasing suppression function Δf2 is calculated and outputted by the following algorithm formula: ； Where △f2(f) represents the common mode interference anti-aliasing suppression function at frequency f, represents the path energy attenuation rate, L represents the total length of the high-voltage long-distance transmission line; S5. The frequency offset prediction value △f1 and the common-mode interference anti-aliasing suppression function △f2 are fused and calculated to output the frequency correction function fcorr. The difference between the frequency correction function fcorr and the actual frequency response value fobs is calculated. A secondary comparative evaluation is performed based on the difference result and the error critical threshold Eth to determine the frequency calibration status of the swept frequency signal generator.

2. The method for calibrating a swept frequency signal generator according to claim 1, wherein: Said S1 includes S11 and S12; S11. Set a collection point every 13 km on the high-voltage long-distance transmission line. The collection point includes a near-end node N, a mid-section node M, and an end node F. At the same time, set a power sensor group in each collection point to collect power data in real time. The proximal node N is arranged at the substation end of the high-voltage long-distance transmission line; The middle section node M is arranged at the grounding pole of the middle section tower base; The terminal node F is set at the terminal monitoring point before the load is connected; The power sensor group includes GPS, PMU synchronized phasor measurement unit, ADC module and FFT spectrum analyzer; The power data includes a synchronously measured voltage V(x) at the line position x, a synchronously measured current I(x) at the line position x, a phase angle Xw(x, f) at the line position x at a frequency f, and a background spectrum S(f) at a frequency f; The route position x is obtained through GPS positioning; The synchronous measurement voltage V(x) at the line position x and the synchronous measurement current I(x) at the line position x are injected with a swept frequency signal at the frequency f of the swept frequency signal at both ends of the high-voltage long-distance transmission line through a swept frequency signal generator, and the current I and voltage V at different line positions x are synchronously measured by a PMU synchronized phasor measurement unit; The phase angle Xw(x, f) at the line position x at the frequency f is obtained by using an ADC module to extract the phase change of each mid-segment node M when injecting the frequency f of the swept frequency signal; The background spectrum S(f) at the frequency f is obtained by performing a fast Fourier transform (FFT) on the frequency f signal within the acquisition period using an FFT spectrum analyzer at the terminal node F during the period when the sweep signal generator is turned off. S12. Use industrial Ethernet and power sensor groups in the substation to aggregate the power data of all sampling points to the substation. Use 5G industrial routers in the substation to transmit the power data to the calibration server.

3. The method for calibrating a frequency sweep signal generator according to claim 2, wherein: Said S2 includes S21 and S22; S21. Extracting features from the power data in the calibration server to obtain a frequency shift feature vector set; The frequency shift feature vector set includes the length impedance gradient Kz(x) of the line position x, the phase density △Xw(x,f) of the line position x at the frequency f, and the background power spectrum density Pc(f) generated by the common mode interference at the frequency f; The length impedance gradient Kz(x) at the line position x is obtained by extracting the synchronously measured voltage V(x) and the synchronously measured current I(x) at all line positions x in the power data, performing ratio calculation to obtain the transient impedance Z(x) at the line position x, and then taking the difference between the transient impedance Z(x) at the line position x of adjacent collection points and dividing by the distance to extract the length impedance gradient Kz(x) at the line position x; The phase density △Xw(x, f) at the line position x at the frequency f is calculated and extracted by extracting the phase angle Xw(x, f) at the line position x at the frequency f and fitting the rate of change of the phase with distance. The specific algorithm formula is: , where d represents the calculus function, and dx represents the calculus variable of the line position x; The background power spectrum density Pc(f) generated by the common-mode interference at the frequency f is obtained by performing smooth curve fitting on the background spectrum S(f) at the frequency f using a quintic B-spline, and the fitting curve is set to the background power spectrum density Pc(f) generated by the common-mode interference at the frequency f; S22, preprocessing the frequency shift feature vector set obtained based on feature extraction to obtain a standardized digital feature set; The preprocessing uses the Min-Max normalization method to unify the value range of the frequency shift feature vector set and eliminate the dimension effect.

4. The method for calibrating a swept frequency signal generator according to claim 3, wherein: Said S3 includes S31; S31. Construct a mathematical model to predict the frequency offset that the current frequency sweep signal will generate in the transmission path. Extract the length impedance gradient Kz(x) at line position x and the phase density △Xw(x, f) at line position x at frequency f from the standardized digital feature set and input them into the mathematical model. Calculate and output the predicted frequency offset △f1, which measures the ability of the frequency sweep signal to offset the structural offset error along the path by pre-adjusting the frequency at the transmitting end. The frequency offset prediction value Δf1 is calculated and outputted by the following mathematical model; ； Where △f1(f) represents the predicted frequency offset at frequency f, L represents the total length of the high-voltage long-distance transmission line, sin represents the sine function, e represents the exponential function, dx represents the calculus variable at the line location x, and u represents the energy dissipation factor.

5. The method for calibrating a frequency sweep signal generator according to claim 4, wherein: Said S3 also includes S32 and S33; S32, performing dimensionless processing based on the actual observed response frequency fobs(f) extracted in real time at the frequency f and the ideal response frequency fideal(f) set by the user at the frequency f to eliminate the unit dimension, performing error calculation on the frequency offset prediction value △f1(f) at the frequency f, and obtaining a frequency shift error value Ef; The frequency shift error value Ef is calculated and outputted by the following algorithm formula: ； Where, Ef(f) represents the frequency shift error value at frequency f; S33. Based on the user setting of the critical error threshold Eth, a preliminary comparison and evaluation is performed between the frequency shift error value Ef(f) obtained in real time at the frequency f and the critical error threshold Eth to determine the frequency shift of the pre-adjusted frequency of the swept signal generator. The specific evaluation contents are as follows: When the frequency shift error value Ef(f) at frequency f is ≤ the critical error threshold Eth, it means that the frequency shift of the pre-adjusted frequency of the sweep signal generator is normal and no intervention is required; When the frequency shift error value Ef(f) at frequency f is greater than the critical error threshold Eth, it indicates that the frequency shift of the pre-adjusted frequency of the sweep signal generator is abnormal, and the common mode spectrum interference suppression mechanism is triggered.

6. The method for calibrating a frequency sweep signal generator according to claim 1, wherein: The S5 includes S51; S51, performing a fusion calculation based on the common-mode interference anti-aliasing suppression function △f2(f) at frequency f and the frequency offset prediction value △f1(f) at frequency f, and outputting a frequency correction function fcorr as the actual control compensation term of the swept frequency signal generator, which directly acts on the control parameters of the swept frequency signal source; The frequency correction function fcorr is calculated and outputted by the following algorithm formula: ； Where fcorr(f) represents the frequency correction function of frequency f, and fideal(f) represents the ideal response frequency at frequency f.

7. The method for calibrating a swept frequency signal generator according to claim 6, wherein: Said S5 also includes S52; S52. Calculate the difference between the frequency correction function fcorr(f) currently controlling the frequency f of the swept frequency signal generator and the actual observed response frequency fobs(f) at the frequency f, and perform a secondary comparative evaluation of the difference calculation result with the error critical threshold Eth to analyze the calibration of the swept frequency signal generator frequency f after correction and compensation. The specific evaluation content is as follows: when When the error threshold Eth is less than or equal to the critical threshold, it indicates that the sweep signal generator has been calibrated successfully and no intervention is required. when When the error threshold Eth is greater than the critical threshold, it indicates that the calibration of the sweep signal generator is abnormal, and the iterative adjustment mechanism is executed; The iterative adjustment mechanism collects the background spectrum S(f, x) at the line position x at the frequency f after correction and compensation by the sweep signal generator, and reconstructs the background power spectrum density Pc(f) generated by the common mode interference at the frequency f. The specific reconstruction method is: Pc(f)=Env(S(f,x)), where Env represents the envelope, which is a smooth depiction of the local amplitude changes of the signal in the time domain or frequency domain. In spectrum analysis, the envelope usually refers to the average value of each frequency component in the signal spectrum. The common-mode interference anti-aliasing suppression function △f2 and the frequency correction function fcorr are recalculated based on the background power spectral density Pc(f) generated by the common-mode interference at the reconstructed frequency f, and an iterative secondary comparative evaluation is performed until the calibration is successful and the iteration is stopped.

8. A calibration system for a swept frequency signal generator, applied to a calibration method for a swept frequency signal generator according to any one of claims 1 to 7, characterized in that: It includes frequency shift data acquisition module, frequency shift data processing module, frequency shift error analysis module, interference suppression module and final compensation module; The frequency shift data acquisition module collects power data in real time by setting up acquisition points on the high-voltage long-distance transmission line and setting up power sensors in each acquisition point, and transmits the power data to the calibration server; The frequency shift data processing module extracts features from the power data in the calibration server to obtain a frequency shift feature vector set, and pre-processes the frequency shift feature vector set to obtain a standardized digital feature set; The frequency shift error analysis module calculates and outputs a frequency shift prediction value △f1 based on a standardized digital feature set, calculates and outputs a frequency shift error value Ef based on the frequency shift prediction value △f1, and sets an error critical threshold Eth to perform preliminary comparative evaluation with the frequency shift error value Ef; The interference suppression module triggers a common-mode spectrum interference suppression mechanism based on a preliminary comparison and evaluation result, extracts interference data, and calculates and outputs a common-mode interference anti-aliasing suppression function Δf2 based on the interference data; The final compensation module fuses the frequency offset prediction value △f1 and the common-mode interference anti-aliasing suppression function △f2 to output a frequency correction function fcorr, performs a difference calculation based on the frequency correction function fcorr and the actual frequency response value fobs, and performs a secondary comparative evaluation based on the difference result and the error critical threshold Eth to determine the frequency calibration status of the swept signal generator.

Images