Rocket Electromagnetic Interference Source Localization and Suppression System and Method

Abstract

This invention discloses a rocket electromagnetic interference source localization and suppression system and method. The system comprises five modules: excitation source acquisition, electromagnetic response acquisition, localization calculation, evaluation, and suppression control. The excitation source acquisition module acquires artificial lightning current, return stroke coordinates, and time signals to form a reference signal packet. The electromagnetic response acquisition module synchronously acquires electromagnetic response signals through electromagnetic field sensors at multiple locations on the wind turbine. The localization calculation module deconvolves the compensated signals to calculate the spatial coordinates of the interference source. The evaluation module extracts discharge characteristics and compares them with a fingerprint database to generate a risk list. The suppression control module generates pitch, shielding, or yaw suppression commands based on the location and risk of the interference source. This invention solves the problems of wind turbine electromagnetic interference source localization being susceptible to structural multipath effects and the lack of targeted suppression.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic interference detection and control technology for wind turbine generator sets, specifically to a system and method for locating and suppressing electromagnetic interference sources in rockets. Background Technology

[0002] With the expansion of wind power installations, wind turbines are becoming increasingly larger and are widely installed in mountainous and coastal areas with frequent lightning activity. These turbines integrate converters, control cabinets, numerous power cables, and sensors. Operating under strong electric fields and mechanical vibrations for extended periods, the insulation materials are prone to defects such as partial discharge and corona discharge, creating sources of electromagnetic interference. The broadband electromagnetic noise emitted by these interference sources can couple to sensitive signal circuits through conduction or radiation, leading to measurement deviations, protection malfunctions, and even unplanned shutdowns, severely impacting turbine availability and power generation efficiency. To identify potential hazards, the industry uses partial discharge detectors or high-frequency antennas for passive monitoring inside the nacelle and tower, attempting to locate the source by capturing interference signals. However, wind turbines have complex structures such as tall metal towers, rotating blades, and nacelle shells. Electromagnetic waves propagate through these structures undergo multiple reflections and diffractions, resulting in severe multipath effects. The sensors receive signals that are a mixture of direct waves and multiple reflected waves, causing severe distortion of arrival time and angle. Traditional positioning algorithms based on time difference or spatial spectrum have significant errors and often yield incorrect locations. Furthermore, passive methods rely on the self-discharge of the interference source, whose occurrence time and intensity are random and uncontrollable, resulting in weak signals, low signal-to-noise ratios, and difficulty in capturing the panoramic data required for effective positioning. Regarding interference source suppression, existing systems often employ simple shutdown or power reduction strategies upon detecting anomalies, failing to distinguish the specific coordinates and type of the interference source, let alone implement precise suppression measures such as directional shielding or electric field control at specific locations. Frequent shutdowns lead to significant power generation losses. While artificial lightning induction technology has been used in lightning protection tests of power transmission and transformation equipment, capable of generating lightning with known current waveforms and precise spatiotemporal coordinates, it has not yet been used as an active excitation source for detecting interference sources inside wind turbine units. Therefore, there is an urgent need for an electromagnetic interference source mitigation solution that can actively excite, compensate for structural propagation distortion, accurately locate, and precisely suppress interference sources in a tiered manner. Summary of the Invention

[0003] In view of the shortcomings of the prior art, the purpose of this invention is to provide a rocket electromagnetic interference source localization and suppression system and method to solve the problem that the localization of electromagnetic interference sources in wind turbine units is easily affected by structural multipath and the suppression lacks specificity.

[0004] This invention generates a strong electromagnetic excitation with precise spatiotemporal parameters by launching a lightning-inducing rocket. Multi-channel sensors are deployed on the wind turbine tower, nacelle, and blades to synchronously collect electromagnetic response signals. The multipath impulse response function determined by the pre-built structural electromagnetic model is used to deconvolve and compensate for each response signal to remove structural reflection interference. After extracting the direct wave, the interference source is accurately located by combining the arrival time difference and angle spectrum. Then, the risk level is determined by comparing the discharge phase resolution characteristics with the defect fingerprint database. Finally, targeted suppression commands such as pitch control, partial shielding, or yaw are generated according to the location and risk, replacing the crude measure of blindly shutting down the turbine.

[0005] This invention provides a rocket electromagnetic interference source localization and suppression system, comprising: The excitation source acquisition module acquires the current waveform signal, return stroke position coordinate signal, and occurrence time signal of artificial lightning, forming an excitation source reference signal packet; The electromagnetic response acquisition module uses multiple electromagnetic field sensors arranged on the wind turbine tower, nacelle and blade root to synchronously acquire electromagnetic response signals under lightning excitation and form multi-channel response signals. The positioning and calculation module receives the excitation source reference signal packet and the multi-channel response signal, calls the multipath impulse response function determined by the pre-stored wind turbine structure electromagnetic simulation model to perform deconvolution compensation on the multi-channel response signal, and generates the spatial coordinate signal of the interference source based on the arrival time difference and arrival angle spectrum of the compensated signal. The evaluation module receives the spatial coordinate signal of the interference source, extracts the discharge pulse waveform of the interference source corresponding to the coordinate, constructs the discharge phase resolution feature and compares it with the preset defect electromagnetic fingerprint database to generate an interference source list signal with risk level. The suppression control module receives the interference source list signal and generates suppression control command signals to adjust the blade pitch angle, deploy local shields, or change the yaw direction based on the location and risk level of the interference sources.

[0006] In one embodiment of the present invention, the excitation source acquisition module includes a broadband Rogowski coil current sensor and a fast electric field change meter. The broadband Rogowski coil current sensor is installed at the base of the grounding wire of the artificial lightning-inducing rocket to acquire current waveform signals. The fast electric field change meter is arranged in an array on the ground to determine the return stroke position coordinate signal and the occurrence time signal by measuring the time difference and amplitude of the electric field change signals at different locations. The excitation source acquisition module also receives timing signals from the Global Navigation Satellite System, timestamps the current waveform signal, the return stroke position coordinate signal, and the occurrence time signal, and packages them with the current waveform characteristic parameters, three-dimensional spatial coordinates, and the trigger time to form an excitation source reference signal packet.

[0007] In one embodiment of the present invention, the electromagnetic response acquisition module includes multiple electromagnetic field sensors, including electric field time-varying rate probes and magnetic field time-varying rate probes. The electric field time-varying rate probes and magnetic field time-varying rate probes are installed in a ring array or a linear array at the flanges at the bottom, middle, and top of the wind turbine tower, inside the nacelle near the control cabinet and converter, and in the inner cavity at the blade root, forming a multi-channel synchronous acquisition network. The electromagnetic response acquisition module also includes a multi-channel signal conditioning and synchronous acquisition unit, which receives timing signals from the Global Navigation Satellite System and trigger signals from the excitation source acquisition module, and performs synchronous analog-to-digital conversion on the signals from each sensor to form a multi-channel response signal.

[0008] In one embodiment of the present invention, the multipath impulse response function of the pre-stored electromagnetic simulation model of the wind turbine structure in the positioning and calculation module is determined in the following manner: a full structural geometric model including each segment of the wind turbine tower, flange, platform, nacelle cover, generator housing, blade skin and grounding system is established in advance using the finite element method, and the conductivity, dielectric constant and magnetic permeability are set for each part; the return stroke position coordinate signal and current waveform signal in the excitation source reference signal packet are used as excitation input, and the propagation, reflection and diffraction process of electromagnetic waves in the full structural geometric model is calculated using the finite difference method in the time domain; virtual probes are set at the coordinate positions corresponding to the actual sensors to record the time domain field strength waveform; Wiener deconvolution is performed on the recorded time domain field strength waveform and the excitation source waveform to extract the multipath impulse response function at each sensor position. The multipath impulse response function is a set of pulse sequences with different time delays and amplitude attenuations, representing the structural multipath propagation characteristics suffered at that position.

[0009] In one embodiment of the present invention, the specific process by which the positioning calculation module performs deconvolution compensation on the multi-channel response signals and generates spatial coordinate signals of the interference source is as follows: For the multi-channel response signals of each sensor channel, a time-domain Wiener deconvolution operation is performed using the multipath impulse response function of the corresponding sensor position to remove the reflection and diffraction components caused by the metal structure of the wind turbine, thereby obtaining the compensated direct wave signal; the compensated direct wave signals of all sensor channels are cross-correlated pairwise to extract the arrival time difference between each channel pair; simultaneously, a multi-signal classification algorithm is applied to multiple compensated direct wave signals for spatial... Inter-spectral estimation is performed to obtain the arrival angle spectrum including azimuth and elevation angles. A cost function is constructed, consisting of arrival time difference residuals and angle residuals. The arrival time difference residuals are the sum of squares of the differences between the measured arrival time difference and the theoretical arrival time difference calculated from the candidate spatial coordinates, and the angle residuals are the weighted sum of squares of the differences between the angles corresponding to the peaks in the arrival angle spectrum and the theoretical angles calculated from the candidate spatial coordinates. The coordinates that minimize the cost function are searched in a preset 3D spatial grid, and local optimization is performed using gradient descent. The resulting coordinates are the spatial coordinate signals of the interference source. The formula for calculating the cost function is as follows:

[0010] Where J is the cost function, and x=(x,y,z) are the three-dimensional spatial coordinates of the interference source candidate. Let be the measured arrival time difference between the i-th and j-th sensor channels. , , Let be the three-dimensional coordinates of the i-th sensor, c be the speed of electromagnetic waves in air, and N be the total number of sensor channels. Here, M represents the weighting coefficients for the angle residual term, and M represents the number of peak values ​​in the arriving angle spectrum. The measured angle of arrival corresponds to the k-th peak. The candidate coordinates correspond to the theoretical angle of arrival.

[0011] In one embodiment of the present invention, the process by which the evaluation module extracts the discharge pulse waveform of the interference source corresponding to the spatial coordinates of the interference source and constructs the discharge phase-resolved feature is as follows: Based on the position indicated by the spatial coordinate signal of the interference source, the evaluation module applies a beamforming algorithm to the multi-channel response signal to enhance the signal in that position direction, extracts the time series of the interference source discharge event, and identifies the waveform of each discharge pulse; using the power frequency voltage phase at the grid connection of the wind turbine as the reference phase, the module calculates the discharge quantity corresponding to each discharge pulse and statistically analyzes the distribution of the discharge quantity with phase over multiple power frequency cycles to form a discharge quantity-phase distribution spectrum; simultaneously, it statistically analyzes the distribution of the discharge pulse repetition rate with pulse amplitude to form a pulse repetition rate-amplitude distribution spectrum; it calculates the skewness and kurtosis of the discharge quantity-phase distribution spectrum, as well as the peak value, half-width at half-maximum, and asymmetry of the pulse repetition rate-amplitude distribution spectrum, and combines these statistical parameters as the discharge phase-resolved feature of the interference source. The calculation formula for the discharge quantity-phase distribution spectrum is as follows:

[0012] Where S is the skewness of the discharge quantity-phase distribution spectrum, and K is the number of phase sampling points within the power frequency cycle. This represents the discharge quantity corresponding to the p-th phase sampling point. This represents the average discharge amount. This represents the standard deviation of the discharge quantity.

[0013] In one embodiment of the present invention, the process by which the evaluation module generates an interference source list signal with risk levels is as follows: A preset defect electromagnetic fingerprint database stores floating potential discharge fingerprints and sharp corona discharge fingerprints calibrated by applying high voltage to typical defects under laboratory conditions. Each fingerprint includes the skewness and kurtosis of the corresponding discharge quantity-phase distribution spectrum and the shape parameters of the pulse repetition rate-amplitude distribution spectrum. The evaluation module calculates the Euclidean distance between the discharge phase resolution characteristics of the interference source and each defect fingerprint in the defect electromagnetic fingerprint database, and selects the one with the smallest distance as the matching defect type. Simultaneously, based on the spatial coordinate signal of the interference source, the spatial distance from the interference source to the sensitive equipment in the wind turbine is calculated. The sensitive equipment includes control cabinets, converters, and sensor signal conditioning circuits. Based on the risk coefficient corresponding to the matching defect type and the spatial distance attenuation function, the interference risk index is calculated. The interference risk index is compared with preset low, medium, and high risk thresholds to determine the risk level, and an interference source list signal containing the spatial coordinates of the interference source, the defect type, and the risk level is generated. The calculation formula for the interference risk index is as follows:

[0014] Where R is the interference risk index. To match the hazard coefficient corresponding to the defect type, d is the spatial distance from the interference source to the sensitive device. Where S is the distance attenuation characteristic length, and S is the current discharge quantity-phase distribution skewness of the interference source. Let L be the kurtosis of the pulse repetition rate-amplitude distribution of the current interference source, and L be the total number of fingerprint types in the defect electromagnetic fingerprint database. For the skewness of the fingerprint of type n defects, Let be the kurtosis of the nth type of defect fingerprint.

[0015] In one embodiment of the present invention, the suppression control module also communicates with the wind turbine main control system to receive blade position signals and turbine operating status signals. When the interference source list signal indicates that the interference source is located in the blade tip or trailing edge region and the risk level is higher than a preset threshold, the suppression control module generates a pitch compensation angle command. This command includes the pitch direction and compensation angle value and is output to the pitch actuator. By adjusting the pitch angle of the corresponding blade, the spatial electric field distribution between the blade surface and the tower is changed, so that the local electric field intensity in the blade tip region is lower than the air initiation discharge intensity, thereby suppressing the blade tip discharge.

[0016] In one embodiment of the present invention, the suppression control module is connected to a relay matrix in the engine compartment. Each channel of the relay matrix corresponds to a local auxiliary shielding cover preset next to the converter cabinet and the generator terminal box. When the interference source list signal indicates that the interference source is located in a locatable device in the engine compartment and the risk level is higher than a preset threshold, the suppression control module generates a shielding activation command, drives the corresponding relay to close, and connects the local auxiliary shielding cover at that location to the grounding circuit, forming a local electromagnetic shielding surrounding the device and suppressing the leakage interference source generated by the device.

[0017] This invention also includes a method for locating and suppressing electromagnetic interference sources in rockets, comprising: S1: Collect the current waveform signal, return stroke position coordinate signal, and occurrence time signal of artificial lightning to form an excitation source reference signal packet; S2: Multiple electromagnetic field sensors are arranged on the wind turbine tower, nacelle and blade root to synchronously collect electromagnetic response signals under lightning excitation and form multi-channel response signals. S3: Receive the excitation source reference signal packet and multi-channel response signal, call the multipath impulse response function determined by the pre-stored wind turbine structure electromagnetic simulation model to perform deconvolution compensation on the multi-channel response signal, and generate the interference source spatial coordinate signal based on the arrival time difference and arrival angle spectrum of the compensated signal. S4: Receive the spatial coordinate signal of the interference source, extract the discharge pulse waveform of the interference source corresponding to the coordinate, construct the discharge phase resolution feature and compare it with the preset defect electromagnetic fingerprint database to generate an interference source list signal with risk level. S5: Receives the interference source list signal and generates suppression control command signals to adjust the blade pitch angle, deploy local shields, or change the yaw direction based on the location and risk level of the interference sources.

[0018] The rocket electromagnetic interference source localization and suppression system and method provided by this invention generates a strong electromagnetic excitation with precise spatiotemporal parameters by launching a lightning-inducing rocket. Multi-channel sensors are arranged on the wind turbine tower, nacelle and blades to synchronously collect electromagnetic response signals. The multipath impulse response function determined by the pre-built structural electromagnetic model is called to perform deconvolution compensation on each response signal to remove structural reflection interference. After extracting the direct wave, the interference source is accurately located by combining the arrival time difference and angle spectrum. Then, the risk level is determined by comparing the discharge phase resolution characteristics with the defect fingerprint database. Finally, targeted suppression commands such as pitch control, partial shielding or yaw are generated according to the location and risk, replacing the crude measures of blindly shutting down the turbine. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 System architecture diagram of a rocket electromagnetic interference source localization and suppression system; Figure 2 Here is a flowchart of the internal algorithm of the localization and solution module; Figure 3 Evaluation-inhibition control closed-loop decision-making flowchart; Figure 4 A flowchart illustrating the method for locating and suppressing electromagnetic interference sources in rockets. Detailed Implementation

[0021] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0022] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0023] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0024] Please see Figure 1-4The figure shows the rocket electromagnetic interference source localization and suppression system and method of the present invention. The rocket electromagnetic interference source localization and suppression system of the present invention includes an excitation source acquisition module, which acquires the current waveform signal, return stroke position coordinate signal, and occurrence time signal of artificial lightning, forming an excitation source reference signal packet; an electromagnetic response acquisition module, which synchronously acquires the electromagnetic response signal under lightning excitation through multiple electromagnetic field sensors arranged in the wind turbine tower, nacelle, and blade root, forming a multi-channel response signal; a localization calculation module, which receives the excitation source reference signal packet and the multi-channel response signal, calls the multipath impulse response function determined by the pre-stored wind turbine structure electromagnetic simulation model to perform deconvolution compensation on the multi-channel response signal, and generates the interference source spatial coordinate signal based on the arrival time difference and arrival angle spectrum of the compensated signal; an evaluation module, which receives the interference source spatial coordinate signal, extracts the interference source discharge pulse waveform corresponding to the coordinate, constructs the discharge phase resolution feature and compares it with the pre-set defect electromagnetic fingerprint database, generating an interference source list signal with risk level; and a suppression control module, which receives the interference source list signal and generates suppression control command signals to adjust the blade pitch angle, deploy local shielding, or change the yaw azimuth according to the location and risk level of the interference source.

[0025] like Figure 1As shown, the system presents the complete signal flow from signal acquisition, location calculation, risk assessment to suppression execution, including risk classification branches and closed-loop feedback links, clearly demonstrating the collaborative logic and operating mechanism of each functional unit. The excitation source acquisition module, as the system's signal source, plays a crucial role in accurately capturing key excitation information during artificial lightning strikes. It comprehensively collects the current waveform signal, return stroke coordinate signal, and occurrence time signal of the artificial lightning strike, integrating and packaging multi-dimensional information into an excitation source reference signal package. This provides a benchmark reference for subsequent electromagnetic response calibration and interference source location, ensuring the accuracy and consistency of signal tracing throughout the system. The electromagnetic response acquisition module is the system's sensing terminal. Relying on multiple electromagnetic field sensors deployed on the wind turbine tower, nacelle, and blade roots, it constructs a distributed synchronous acquisition network. It synchronously acquires electromagnetic response signals from key parts of the unit under artificial lightning excitation in real time, summarizing them into multi-channel response signals. This comprehensively captures the electromagnetic changes generated by the propagation and reflection of electromagnetic waves within the unit structure, providing raw data support for interference source location. The positioning and calculation module is the core positioning hub of the system. It synchronously receives the excitation source reference signal packet and multi-channel response signals, calls the multipath impulse response function determined by the pre-stored electromagnetic simulation model of the wind turbine structure, and performs deconvolution compensation processing on the multi-channel response signals to effectively eliminate signal distortion interference caused by the unit's metal structure. Based on the arrival time difference and arrival angle spectrum of the compensated signal, it accurately generates the spatial coordinate signal of the interference source, realizing the accurate conversion from electromagnetic response signal to interference source location. The evaluation module is responsible for defect identification and risk classification. It receives the spatial coordinate signal of the interference source output by the positioning and calculation module, extracts the discharge pulse waveform of the interference source at the corresponding coordinate position, constructs standardized discharge phase resolution features, and accurately identifies the type of interference source defect by comparing and matching with a pre-set defect electromagnetic fingerprint database, thereby generating a list signal of interference sources with risk levels. The module has a low-risk continuous monitoring branch to maintain real-time monitoring of low-risk interference sources and periodically review them to avoid unnecessary intervention and ensure system operating efficiency.

[0026] The suppression control module is the system's execution decision-making unit. It receives the risk level list signal output by the assessment module and, based on the specific location of the interference source and its corresponding risk level, intelligently matches and generates suppression control command signals to adjust the blade pitch angle, deploy local shielding, or change the yaw direction, achieving precise output of the suppression strategy. The wind turbine itself acts as the command execution carrier, receiving the suppression control command and completing the corresponding physical adjustments. Simultaneously, it collects the changes in electromagnetic response after adjustment in real time and transmits the electromagnetic response feedback signal back to the electromagnetic response acquisition module, forming a complete closed-loop feedback link. The module has a high-risk dynamic adjustment branch that continuously optimizes the adjustment strategy for high-risk interference sources, strengthening the suppression effect and ensuring the safety of the unit's core equipment. Overall, the modules are logically connected vertically, with smooth signal flow. Branch decisions adapt to different risk scenarios, and closed-loop feedback ensures dynamic system optimization, forming a complete closed-loop control system from excitation acquisition to interference suppression. This system meets the actual needs of wind turbine electromagnetic interference prevention and control in artificial lightning scenarios, featuring stable operation, timely response, and precise strategies, effectively solving the problem of electromagnetic interference caused by artificial lightning.

[0027] Specifically, the system first includes an excitation source acquisition module. The core function of this module is to acquire the current waveform signal, return stroke location coordinate signal, and occurrence time signal of the artificial lightning strike, and package these signals into an excitation source reference signal packet. Artificial lightning strikes are achieved by launching a lightning-inducing rocket carrying a grounding wire from a predetermined location around the wind turbine. After the rocket ascends, the wire is straightened and ultimately triggers lightning, which has precisely known spatiotemporal parameters. In practical implementation, the excitation source acquisition module includes a broadband Rogowski coil current sensor, which is installed at the base of the grounding wire of the artificial lightning-inducing rocket. Due to the Rogowski coil's characteristics of fast response, wide bandwidth, and unsaturation, it can accurately record the pulse current waveform from tens of amperes to tens of thousands of amperes during the lightning return stroke, forming a current waveform signal. Simultaneously, multiple fast electric field change meters are arranged in an array on the ground. These meters are located at known coordinates around the lightning induction point, and each measures the electric field change signal excited by the lightning return stroke. Because electromagnetic waves propagate at the speed of light, the timing of the electric field change front received by the fast electric field change instrument at different locations varies slightly. By using high-precision time measurement and the time difference of arrival (TDOA) positioning principle, the three-dimensional spatial coordinates of the bottom of the return stroke channel can be retrieved, i.e., the return stroke position coordinate signal. Simultaneously, the precise moment of the return stroke is recorded as the occurrence time signal. To ensure all signals have a unified and accurate time reference, the excitation source acquisition module also receives timing signals from the Global Navigation Satellite System (GNSS) and timestamps the current waveform signal, return stroke position coordinate signal, and occurrence time signal, achieving a timestamp accuracy on the nanosecond level. Finally, the module packages the timestamped current waveform characteristic parameters, three-dimensional spatial coordinates, and trigger time into a preset data frame format, forming an excitation source reference signal packet. This signal packet not only provides known spatiotemporal parameters of the excitation source for subsequent positioning calculations but also serves as a synchronous trigger reference for the multi-channel electromagnetic response acquisition module.

[0028] The system also includes an electromagnetic response acquisition module. This module synchronously acquires electromagnetic response signals under lightning excitation through multiple electromagnetic field sensors deployed on the wind turbine tower, nacelle, and blade roots, forming a multi-channel response signal. Specifically, the electromagnetic field sensors include electric field time-varying rate probes and magnetic field time-varying rate probes. These two types of probes sense the electric field and magnetic field change rates of spatial electromagnetic waves, respectively, and have the advantages of high sensitivity and flat frequency response. These sensors are arranged in a ring array or linear array, specifically according to the structural characteristics of the wind turbine: sensors are installed at the flanges at the bottom, middle, and top of the tower to capture the electromagnetic field radiated by the induced current propagating along the tower and the creeping waves on the tower surface; sensors are installed inside the nacelle near the control cabinet and converter to monitor the electromagnetic environment near these critical devices; sensors are installed inside the blade root cavity, as the blade root connects to the hub and is the necessary path for the electromagnetic signals from the internal discharge of the blade to be transmitted to the nacelle, and placing sensors here can effectively capture signals from blade-related interference sources. All sensors together form a multi-channel synchronous acquisition network. The electromagnetic response acquisition module also includes a multi-channel signal conditioning and synchronous acquisition unit. This unit receives timing signals from the Global Navigation Satellite System to obtain a unified time reference, and simultaneously receives trigger signals from the excitation source acquisition module. These trigger signals are correlated with the occurrence time signal in the excitation source reference signal packet, ensuring strict synchronization between the start time of electromagnetic response acquisition and the artificial lightning triggering event. After conditioning the signals from each sensor through impedance matching, filtering, and amplification, this unit performs synchronous sampling using a high-speed multi-channel analog-to-digital converter. The sampled digital signal sequences are arranged according to channel number and time order to form multi-channel response signals. Each channel's response signal is a time series, completely recording the change in the electromagnetic field at the sensor location over time from the start of lightning triggering.

[0029] like Figure 2As shown in the figure, this diagram illustrates the complete algorithm flow of the localization solution module, from data input, signal preprocessing, feature extraction to iterative optimization and result output. It includes key steps such as parallel data input, core parameter loading, signal compensation processing, multi-dimensional feature extraction, cost function construction, grid search, and gradient descent iterative loops, clearly presenting the algorithm logic for accurately calculating the spatial coordinates of the interference source. Receiving the excitation source reference package serves as the first input step in the algorithm flow. Its core function is to acquire the baseline excitation information for artificial lightning strikes, including key parameters such as the return stroke location coordinates, current waveform data, signal occurrence time, and time stamp. This provides a precise baseline for subsequent multipath pulse response function calls, theoretical time difference, and angle calculations, ensuring the accuracy of the algorithm's input excitation dimension and avoiding positioning errors caused by excitation baseline deviations. Receiving multi-channel response signals is the parallel input step. It receives multi-channel electromagnetic response data from the electromagnetic response acquisition module, covering sensor signals from key areas such as the tower bottom, middle, top flange, nacelle interior, and blade roots. This fully presents the electromagnetic response status of various parts of the unit under artificial lightning excitation, providing comprehensive raw data support for signal preprocessing and feature extraction. Loading the multipath impulse response function (MECT) is the parameter preparation step of the algorithm. It retrieves the MECT output from the pre-stored full-structure electromagnetic simulation model of the wind turbine. This function accurately characterizes the multipath propagation characteristics of electromagnetic waves at each sensor location after reflection and diffraction by the turbine's metal structure. It is a core parameter for eliminating signal distortion and purifying the direct wave signal, providing a crucial basis for subsequent deconvolution compensation. Time-domain Wiener deconvolution compensation is the core step in signal preprocessing. For the response signal of each sensor channel, a time-domain Wiener deconvolution operation is performed, matching the corresponding MECT at the corresponding location. This effectively filters out electromagnetic wave reflection and diffraction components caused by the wind turbine's metal structure, stripping away interference signals and obtaining a pure compensated direct wave signal, thus eliminating the impact of the turbine structure's multipath effect on positioning accuracy at its source. Extracting the arrival time difference between channels is the time-dimensional feature extraction step. Pairwise cross-correlation operations are performed on the compensated direct wave signals of all sensor channels to accurately calculate the arrival time difference between each pair of channels, constructing a time-dimensional positioning feature matrix. This provides a time-dimensional error basis for constructing the cost function. Multiple signal classification → Arrival angle spectrum is the spatial dimension feature extraction step. Multiple signal classification algorithm is applied to multiple compensated direct wave signals to perform spatial spectrum estimation, accurately obtain the azimuth and elevation angles of the interference source signal arriving at the sensor array, generate the arrival angle spectrum, supplement the spatial angle dimension positioning features, and make up for the limitations of single time difference positioning in the spatial dimension.

[0030] The construction of the cost function is the first step in establishing the algorithm's optimization objective. It integrates the arrival time difference residual and the angle residual. The arrival time difference residual quantifies the matching error between the measured and theoretical time differences, while the angle residual quantifies the matching error between the measured and theoretical angles, forming a multi-dimensional error fusion cost function to establish the target criterion for optimizing the coordinates of interference sources. The 3D mesh search for optimal coordinates is the initial optimization step. Within the preset 3D spatial mesh range of the wind turbine, all candidate spatial coordinates are traversed, and the cost function value corresponding to each coordinate is calculated one by one. This initially filters out the candidate coordinate range with smaller cost function values, narrowing the search interval for subsequent iterations and improving the algorithm's computational efficiency. Gradient descent iterative optimization is the core of precise optimization. Iterative gradient descent calculations are performed on the initially selected candidate coordinates, and the convergence of the optimization results is judged in real time. If convergence fails, the algorithm returns to the 3D mesh search step to re-filter candidate coordinates, continuously iterating to gradually reduce the cost function error. If convergence occurs, the positioning accuracy requirement is met, and the iteration process terminates. The final step in the algorithm process is to output the spatial coordinates of the interference source. This outputs precise 3D spatial coordinates after iterative optimization and convergence, completing the core function of interference source localization. This provides accurate spatial location data support for subsequent evaluation modules' defect identification and risk assessment. The overall algorithm process is tightly integrated vertically, with clear data flow logic. Parallel input ensures data integrity, iterative loops improve positioning accuracy, and all stages work together effectively to overcome electromagnetic signal distortion caused by the complex metal structure of wind turbines. This achieves high-precision 3D localization of electromagnetic interference sources in artificial lightning scenarios, laying a solid foundation for subsequent risk assessment and precise suppression of interference.

[0031] The core processing unit of the system is the location calculation module. This module receives the excitation source reference signal packet and the multi-channel response signal, calls the multipath impulse response function determined by the pre-stored electromagnetic simulation model of the wind turbine structure to perform deconvolution compensation on the multi-channel response signal, and generates the spatial coordinate signal of the interference source based on the arrival time difference and arrival angle spectrum of the compensated signal. Its working principle is based on the following understanding: When a strong electromagnetic pulse generated by artificial lightning strikes a wind turbine, various potential interference sources inside the wind turbine will be excited and generate secondary radiation. These secondary radiation signals are received by the sensor along with the direct signal from the excitation source. However, due to the presence of large structures such as the metal tower, nacelle shell, and blades of the wind turbine, electromagnetic waves will undergo multiple reflections and diffractions during propagation. The signal received by the sensor is actually a superposition of the direct wave and numerous reflected and diffracted waves, i.e., there is a serious multipath effect. In order to accurately extract the direct wave signal of the interference source from this aliased signal, the location calculation module first calls the multipath impulse response function determined by the pre-stored electromagnetic simulation model of the wind turbine structure. The process of determining the multipath impulse response function is as follows: A complete structural geometric model, including all sections of the wind turbine tower, flanges, platform, nacelle cover, generator casing, blade skin, and grounding system, is established beforehand using the finite element method. Corresponding electromagnetic material parameters such as conductivity, dielectric constant, and permeability are set for each part. Using the return stroke position coordinate signal and current waveform signal from the excitation source reference signal packet as excitation input, the entire process of electromagnetic wave propagation is calculated in the model using the finite-difference time-domain method. Virtual probes are set at coordinate points corresponding to the actual sensor installation locations to record the time-domain field strength waveform. The recorded time-domain field strength waveform is deconvolved with the original excitation source waveform by Wiener deconvolution to extract the multipath impulse response function at each sensor location. This function represents a sequence of pulses with different time delays and amplitude attenuations. Each pulse corresponds to a path propagating from the excitation source through the structure to the sensor, fully describing the structural multipath propagation characteristics experienced at that location. During actual positioning, the positioning solution module performs time-domain Wiener deconvolution on the multi-channel response signals of each sensor channel using the multipath impulse response function corresponding to the sensor location. This process extracts the multipath components introduced by structural reflection and diffraction from the aliased signal, yielding the compensated direct wave signal. After compensating each channel, the positioning solution module performs pairwise cross-correlation on the compensated direct wave signals of all sensor channels. By detecting the position of the peak value of the cross-correlation function, the arrival time difference between each channel pair is extracted. Simultaneously, the module also applies a multi-signal classification algorithm to perform spatial spectrum estimation on multiple compensated direct wave signals. Utilizing the orthogonality of the signal subspace and noise subspace, it scans in a two-dimensional angle space composed of azimuth and elevation angles to obtain the arrival angle spectrum containing azimuth and elevation information. The peak position of the angle spectrum indicates the direction of arrival of the interference source signal.After obtaining the time difference of arrival (TDOA) and angle of arrival (AHA) spectra, the localization module constructs a cost function consisting of TDOA residuals and angle residuals. The TDOA residual is the sum of squares of the differences between the measured TDOA and the theoretical TDOA calculated from the candidate spatial coordinates. The angle residual is the weighted sum of squares of the differences between the angles corresponding to the peak values ​​in the angle of arrival and the theoretical angles calculated from the candidate spatial coordinates. Weighting coefficients are used to adjust the proportion of TDOA and angle information in the optimization process. The module traverses all candidate coordinates in a pre-defined 3D spatial grid with a certain step size, calculating the cost function value for each candidate point to initially search for the candidate coordinates that minimize the cost function. Then, using these candidate coordinates as the initial point, gradient descent is used for local optimization iterations until the cost function converges to a preset accuracy. The resulting spatial coordinates are the spatial coordinate signals of the interference source, thus achieving accurate 3D localization of the interference source in a multipath environment.

[0032] like Figure 3As shown in the figure, this diagram presents a complete closed-loop decision-making process from interference source location input, discharge feature extraction, defect fingerprint comparison, risk quantification and classification to suppression strategy execution and effect feedback optimization. It includes key steps such as feature extraction, fingerprint matching, risk classification branching, differentiated suppression execution, and closed-loop feedback loop, clearly demonstrating the decision-making logic for accurate interference risk assessment and dynamic suppression. Receiving the spatial coordinates of the interference source is the initial step in the process. This obtains the precise three-dimensional spatial location information of the interference source output by the location calculation module, providing a spatial reference for subsequent targeted extraction of partial discharge signals. This ensures that signal extraction focuses on the area where the interference source is located, avoids irrelevant signal interference, and guarantees the accuracy of discharge feature extraction. Beamforming to extract the discharge waveform is the core step in signal purification. Based on the location indicated by the spatial coordinates of the interference source, a beamforming algorithm is applied to the multi-channel response signal. Spatial filtering technology is used to enhance the signal strength in the direction of the interference source, suppress irrelevant electromagnetic interference in other directions, accurately extract the complete time series of the interference source discharge event, and clearly identify the specific waveform of each discharge pulse, providing clean discharge signal data for subsequent feature construction. The quantitative characterization step involves constructing discharge phase-resolved features. Using the power frequency voltage phase at the wind turbine grid connection point as a unified reference, the discharge quantity for each discharge pulse is accurately calculated. The distribution pattern of discharge quantity with phase over multiple power frequency cycles is statistically analyzed, generating a discharge quantity-phase distribution spectrum. Simultaneously, the distribution characteristics of discharge pulse repetition rate with pulse amplitude are statistically analyzed, generating a pulse repetition rate-amplitude distribution spectrum. Furthermore, key statistical parameters such as skewness, kurtosis, peak value, and half-width at half-maximum (HWHM) of the two types of spectra are calculated. These multi-dimensional parameters are combined to form standardized and quantifiable discharge phase-resolved features, providing a unified comparison basis for defect fingerprint comparison. The defect identification step involves comparing the constructed discharge phase-resolved features with typical defect fingerprints such as floating potential discharge and tip corona discharge stored in a pre-set defect electromagnetic fingerprint database. Each defect fingerprint in the database contains standard feature parameters calibrated in the laboratory. Accurate identification of interference source defect types is achieved through feature similarity matching, avoiding subjective errors in manual identification. Matching defect types and calculating risk indices constitute the risk quantification step. After determining the defect type of the interference source, the preset hazard coefficient corresponding to that defect type is retrieved. Combined with the spatial coordinates of the interference source, the spatial distance from it to sensitive equipment such as the unit control cabinet, converter, and sensor signal conditioning circuit is calculated. A spatial distance attenuation function is introduced to quantify the attenuation characteristics of electromagnetic interference intensity with distance. By combining the hazard coefficient, distance attenuation factor, and feature matching similarity, a standardized interference risk index is calculated, thereby achieving accurate quantification of the degree of interference hazard.

[0033] Risk level determination is a core step in branch decision-making. Based on the calculated interference risk index, it is compared with the preset low, medium, and high risk thresholds one by one to clearly define the risk level corresponding to the interference source, forming a clear branch decision node. This provides a direct basis for subsequent selection of differentiated suppression strategies, ensuring that suppression measures are accurately matched with the risk level. Low-risk continuous monitoring + periodic review is the low-risk response branch. When the risk level is determined to be low, there is no need to trigger active suppression operations. Only continuous real-time monitoring of the interference source signal is performed, and the risk index changes are periodically reviewed to dynamically track the interference status. Under the premise of ensuring normal unit operation, signs of risk level escalation are promptly captured, balancing operational efficiency and risk controllability. Medium-risk deployment of local shielding is the medium-risk response branch. When the risk level is determined to be medium, the relay matrix in the nacelle is triggered, driving the corresponding channel relays to close. This connects the local auxiliary shielding covers preset next to the converter cabinet and generator outlet box with the grounding circuit, forming a local electromagnetic shielding structure surrounding sensitive equipment. This effectively suppresses leakage interference sources generated by equipment in the nacelle and blocks the propagation path of interference signals. The high-risk response branch involves adjusting the pitch angle / changing the yaw position. When a high-risk level is identified, a pitch compensation angle command or a yaw adjustment command is generated and output to the unit's pitch actuator and yaw system. By adjusting the blade pitch angle, the spatial electric field distribution between the blade and the tower is changed, reducing the electric field intensity at the blade tip to suppress tip discharge. Alternatively, the unit's yaw position is changed to avoid areas of strong interference, thus curbing the generation of high-risk interference sources at their source. Executing the suppression command is the physical execution stage. It receives suppression commands corresponding to different risk levels and drives the pitch mechanism, relay matrix, yaw system, and other actuators to complete the corresponding physical adjustment actions, translating the decision commands into actual suppression measures to achieve the interference control objectives. Feedback on suppression effect → updating response signal is the closed-loop feedback stage. It collects the unit's electromagnetic response signal after the suppression measures are implemented in real time, evaluates the suppression effect, and updates the response signal data. The updated data is then transmitted back to the interference source spatial coordinate receiving stage at the beginning of the process, forming a complete closed-loop feedback link. The low-risk monitoring branch also synchronously transmits data back in a loop, enabling dynamic tracking of the risk status. The overall process is logically coherent, progressing step by step from feature extraction to suppression execution. Risk classification branches adapt to different hazard levels, and closed-loop feedback enables dynamic optimization of the suppression effect. All links work together to build a closed-loop control system of "location-assessment-suppression-feedback", which effectively solves the problems of inaccurate risk classification and weak targeting of suppression measures for electromagnetic interference of wind turbines in artificial lightning scenarios. This ensures the safe and stable operation of the core equipment of wind turbines and achieves precision, intelligence and dynamism in electromagnetic interference prevention and control.

[0034] The system further includes an evaluation module. This module receives the spatial coordinate signal of the interference source, extracts the discharge pulse waveform corresponding to that coordinate, constructs discharge phase-resolved features, and compares them with a pre-set defect electromagnetic fingerprint database to generate an interference source list signal with risk levels. The evaluation module's work is divided into three stages: feature extraction, fingerprint comparison, and risk assessment. In the feature extraction stage, the module applies a beamforming algorithm to the multi-channel response signal based on the location indicated by the spatial coordinate signal of the interference source. The beamforming algorithm applies phase compensation and amplitude weighting corresponding to the target location to the signals of each sensor channel, forming a spatial filtering effect, enhancing the signal from that coordinate direction, while suppressing interference signals from other directions, thereby separating the radiation signal time series of the interference source from the multi-channel response signal. The module identifies the start and end boundaries of each discharge pulse on this time series and extracts the waveform of a single discharge pulse. Using the power frequency voltage phase at the wind turbine's grid connection point as the reference phase, the voltage signal is obtained from the wind turbine's outlet voltage transformer, and the fundamental phase is extracted. The occurrence phase and corresponding discharge quantity of each discharge pulse in this phase coordinate system are calculated, with the discharge quantity defined as the calibrated conversion value of the pulse waveform's integral area or peak value. The phase and discharge quantity of all discharge pulses within multiple power frequency cycles are statistically analyzed to construct a discharge quantity-phase distribution spectrum with phase as the abscissa and discharge quantity as the ordinate. Simultaneously, the variation of the discharge pulse repetition rate with the pulse amplitude is statistically analyzed, constructing a pulse repetition rate-amplitude distribution spectrum. Based on this, the skewness and kurtosis of the discharge quantity-phase distribution spectrum are calculated. Skewness reflects the symmetry of the spectrum shape relative to the central phase, and kurtosis reflects the sharpness of the spectrum shape. The amplitude, half-width at half-maximum (HWHM), and asymmetry corresponding to the peak values ​​of the pulse repetition rate-amplitude distribution spectrum are calculated. HWHM measures the width of the distribution peak, and asymmetry measures the degree of asymmetry between the two sides of the distribution peak. These statistical parameters are combined to form the discharge phase-resolved feature of the interference source, which characterizes the discharge properties of the interference source in vector form. In the fingerprint comparison stage, the evaluation module calls a pre-set defect electromagnetic fingerprint database. This database was calibrated beforehand under laboratory conditions by applying high voltage to typical defects such as floating potential discharge and corona discharge. Each fingerprint also uses skewness, kurtosis, and the shape parameters of the pulse repetition rate-amplitude spectrum as feature vectors. The evaluation module calculates the Euclidean distance between the discharge phase-resolved feature of the interference source and the feature vectors of each defect fingerprint in the database, selecting the defect type with the smallest distance as the matching defect type to complete the defect nature identification of the interference source. In the risk assessment stage, the evaluation module calculates the spatial distance from the interference source to each sensitive device within the wind turbine unit based on the spatial coordinate signal of the interference source. Sensitive devices include control cabinets, converters, and sensor signal conditioning circuits, etc. Disturbance to these devices will directly affect the normal operation of the unit. Based on the hazard coefficient corresponding to the matching defect type and the spatial attenuation function with distance as the variable, the interference risk index is calculated.The risk factor is pre-calibrated based on prior knowledge of the interference capability of defect types on sensitive equipment, and the spatial attenuation function considers the attenuation law of electromagnetic waves in free space and along the structure. The calculated interference risk index is compared step-by-step with preset low, medium, and high risk thresholds to determine the risk level of the current interference source. Finally, the evaluation module lists the spatial coordinates, defect type, and risk level of each interference source, generating an interference source list signal. This list signal contains comprehensive information on all located and evaluated interference sources within the wind turbine unit.

[0035] The system concludes with a suppression control module. This module receives an interference source list signal and, based on the location and risk level of the interference sources, generates suppression control command signals to adjust the blade pitch angle, deploy local shielding, or change the yaw direction. The suppression control module has a data communication interface with the wind turbine's main control system, receiving blade position signals and turbine operating status signals, thus considering the turbine's current operating conditions when generating suppression commands. When the interference source list signal indicates that an interference source is located at the blade tip or trailing edge region, and the risk level of this interference source is higher than a preset threshold, the suppression control module activates a pitch suppression strategy for blade tip partial discharge. This strategy is based on the physical mechanism of blade tip partial discharge: the electric field strength in the blade tip region is related to the blade-tower spacing and the charge accumulation on the blade surface, which in turn is related to environmental factors such as air humidity and ion concentration in the trajectory area traversed by the rotating blade. Adjusting the pitch angle can change the aerodynamic shape and electric field distribution of the blade surface relative to the tower and surrounding space. The suppression control module calculates the required pitch direction and compensation angle values, generates a pitch compensation angle command, and outputs it to the pitch actuator. By adjusting the pitch angle of the corresponding blades, it changes the spatial electric field distribution between the blade surface and the tower, making the local electric field intensity in the blade tip region lower than the air initiation discharge intensity, thereby suppressing the blade tip discharge at its source. When the interference source list signal indicates that the interference source is located in a locatable device in the nacelle, such as inside the converter cabinet or near the generator terminal box, and the risk level is higher than a preset threshold, the suppression control module activates the equipment leakage source shielding switching suppression strategy. This module is connected to a preset relay matrix in the nacelle. Each channel of the relay matrix corresponds to a local auxiliary shield pre-installed near equipment such as the converter cabinet and generator terminal box. These local auxiliary shields are in a retracted position or disconnected from the grounding circuit under normal conditions, and do not affect equipment heat dissipation and maintenance. When suppression is needed, the suppression control module generates a shielding activation command, driving the corresponding channel's relay to close. The relay's action electrically connects the local auxiliary shielding cover at that location to the grounding circuit, forming a local electromagnetic shielding layer surrounding the equipment. This confines the leakage electromagnetic interference generated by the equipment within the shielding cover, effectively reducing coupling interference to other sensitive equipment in the nacelle. When the interference source list signal indicates that the interference source is located on the tower grounding down conductor path, and the location of the interference source is associated with the yaw azimuth of the wind turbine, meaning the abnormal discharge degree of the interference source changes with the yaw angle, the suppression control module activates the yaw suppression strategy for the tower induction source. The suppression control module senses the current orientation of the wind turbine axis by receiving the yaw angle signal provided by the wind turbine yaw encoder. When suppression is needed, the module generates a yaw adjustment command, which includes the yaw direction, yaw angle increment, and yaw speed. The yaw direction is the direction to reduce local current concentration on the grounding down conductor, the yaw angle increment is the angle value that causes the wind turbine axis to deviate from its current position, and the yaw speed is set according to the mechanical capability of the turbine's yaw system.This command is output to the yaw drive motor, changing the relative azimuth angle between the wind turbine and the tower. This alters the distribution of the lightning-induced current along the tower's grounding down conductor, reducing current density concentration in localized areas and suppressing induced discharge interference sources on the tower's grounding down conductor path caused by excessive local current. By selecting and executing these three suppression methods, the system achieves graded suppression precisely corresponding to the location and risk level of the interference source, avoiding unnecessary shutdowns of the entire unit and maximizing the continuous power generation operation of the wind turbine.

[0036] like Figure 4 As shown, this invention also provides a method for locating and suppressing electromagnetic interference sources in rockets, including: S1: collecting the current waveform signal, return stroke position coordinate signal, and occurrence time signal of artificial lightning to form an excitation source reference signal packet; S2: synchronously collecting electromagnetic response signals under lightning excitation through multiple electromagnetic field sensors arranged on the wind turbine tower, nacelle, and blade root to form a multi-channel response signal; S3: receiving the excitation source reference signal packet and the multi-channel response signal, calling the multipath impulse response function determined by the pre-stored electromagnetic simulation model of the wind turbine structure to perform deconvolution compensation on the multi-channel response signal, and generating an interference source spatial coordinate signal based on the arrival time difference and arrival angle spectrum of the compensated signal; S4: receiving the interference source spatial coordinate signal, extracting the interference source discharge pulse waveform corresponding to the coordinate, constructing discharge phase resolution features and comparing them with a pre-set defect electromagnetic fingerprint database to generate an interference source list signal with risk level; S5: receiving the interference source list signal, and generating suppression control command signals to adjust the blade pitch angle, deploy a local shield, or change the yaw azimuth according to the location and risk level of the interference source.

[0037] This invention provides a system and method for locating and suppressing electromagnetic interference sources in rockets. By launching a lightning-inducing rocket to generate a strong electromagnetic excitation with precise spatiotemporal parameters, multi-channel sensors are deployed on the wind turbine tower, nacelle, and blades to synchronously collect electromagnetic response signals. The multipath impulse response function determined by a pre-built structural electromagnetic model is used to deconvolve and compensate for each response signal to remove structural reflection interference. After extracting the direct wave, the system combines the arrival time difference and angular spectrum to accurately locate the interference source. Then, the risk level is determined by comparing the discharge phase resolution characteristics with a defect fingerprint database. Finally, targeted suppression commands such as pitch control, partial shielding, or yaw are generated based on the location and risk, replacing the crude measure of blindly shutting down the turbine.

[0038] Therefore, the rocket electromagnetic interference source localization and suppression system and method of the present invention solves the problem that the localization of electromagnetic interference sources in wind turbine units is easily affected by structural multipath and the suppression lacks specificity.

[0039] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A system for locating and suppressing sources of electromagnetic interference from rockets, characterized in that, include: The excitation source acquisition module acquires the current waveform signal, return stroke position coordinate signal, and occurrence time signal of artificial lightning, forming an excitation source reference signal packet; The electromagnetic response acquisition module uses multiple electromagnetic field sensors arranged on the wind turbine tower, nacelle and blade root to synchronously acquire electromagnetic response signals under lightning excitation and form multi-channel response signals. The positioning and calculation module receives the excitation source reference signal packet and the multi-channel response signal, calls the multipath impulse response function determined by the pre-stored wind turbine structure electromagnetic simulation model to perform deconvolution compensation on the multi-channel response signal, and generates the spatial coordinate signal of the interference source based on the arrival time difference and arrival angle spectrum of the compensated signal. The evaluation module receives the spatial coordinate signal of the interference source, extracts the discharge pulse waveform of the interference source corresponding to the coordinate, constructs the discharge phase resolution feature and compares it with the preset defect electromagnetic fingerprint database, and generates an interference source list signal with risk level. The suppression control module receives the interference source list signal and generates suppression control command signals to adjust the blade pitch angle, deploy a local shield, or change the yaw direction based on the location and risk level of the interference sources.

2. The system of claim 1, wherein, The excitation source acquisition module includes a broadband Rogowski coil current sensor and a fast electric field change meter. The broadband Rogowski coil current sensor is installed at the grounding wire base of the artificial lightning-inducing rocket to acquire current waveform signals. The fast electric field change meter is arranged in an array on the ground to determine the return stroke position coordinate signal and the occurrence time signal by measuring the time difference and amplitude of the electric field change signals at different locations. The excitation source acquisition module also receives timing signals from the Global Navigation Satellite System, timestamps the current waveform signal, return stroke position coordinate signal, and occurrence time signal, and packages them with current waveform characteristic parameters, three-dimensional spatial coordinates, and trigger time to form the excitation source reference signal packet.

3. The system of claim 1, wherein, The electromagnetic response acquisition module includes multiple electromagnetic field sensors, including electric field time-varying rate probes and magnetic field time-varying rate probes. These probes are installed in a ring array or a linear array at the flanges at the bottom, middle, and top of the wind turbine tower, inside the nacelle near the control cabinet and converter, and in the inner cavity at the blade root, forming a multi-channel synchronous acquisition network. The electromagnetic response acquisition module also includes a multi-channel signal conditioning and synchronous acquisition unit. This unit receives timing signals from the Global Navigation Satellite System and trigger signals from the excitation source acquisition module, and performs synchronous analog-to-digital conversion on the signals from each sensor to form a multi-channel response signal.

4. The system of claim 1, wherein, The pre-stored electromagnetic simulation model of the wind turbine structure in the positioning and calculation module determines the multipath impulse response function in the following way: a full structural geometric model including each section of the wind turbine tower, flange, platform, nacelle cover, generator housing, blade skin and grounding system is established in advance using the finite element method, and the conductivity, dielectric constant and magnetic permeability are set for each part; Using the return stroke position coordinate signal and current waveform signal in the excitation source reference signal packet as excitation input, the propagation, reflection and diffraction process of electromagnetic waves in the full structure geometric model are calculated by the finite-difference time-domain method. Virtual probes are set at coordinate positions corresponding to actual sensors to record time-domain field strength waveforms. Wiener deconvolution is performed on the recorded time-domain field strength waveform and the excitation source waveform to extract the multipath impulse response function at each sensor location. The multipath impulse response function is a set of pulse sequences with different time delays and amplitude attenuations, representing the structural multipath propagation characteristics experienced at that location.

5. The system of claim 4, wherein, The specific process by which the positioning and calculation module performs deconvolution compensation on the multi-channel response signals and generates spatial coordinate signals of the interference sources is as follows: For the multi-channel response signals of each sensor channel, a time-domain Wiener deconvolution operation is performed using the multipath impulse response function of the corresponding sensor location to remove reflection and diffraction components caused by the metal structure of the wind turbine, resulting in a compensated direct wave signal; pairwise cross-correlation is performed on the compensated direct wave signals of all sensor channels to extract the arrival time difference between each channel pair; simultaneously, a multi-signal classification algorithm is applied to multiple compensated direct wave signals for spatial spectrum estimation. Obtain the arrival angle spectrum including azimuth and elevation angles; construct a cost function consisting of arrival time difference residuals and angle residuals, where the arrival time difference residuals are the sum of squares of the differences between the measured arrival time difference and the theoretical arrival time difference calculated from the candidate spatial coordinates, and the angle residuals are the weighted sum of squares of the differences between the angles corresponding to the peaks in the arrival angle spectrum and the theoretical angles calculated from the candidate spatial coordinates; search for the coordinates that minimize the cost function in a preset three-dimensional spatial grid, and perform local optimization using the gradient descent method. The obtained coordinates are the spatial coordinate signals of the interference source. The calculation formula of the cost function is as follows: ； Where J is the cost function, and x=(x,y,z) are the three-dimensional spatial coordinates of the interference source candidate. Let be the measured arrival time difference between the i-th and j-th sensor channels. , , Let be the three-dimensional coordinates of the i-th sensor, c be the speed of electromagnetic waves in air, and N be the total number of sensor channels. Here, M represents the weighting coefficients for the angle residual term, and M represents the number of peak values ​​in the arriving angle spectrum. The measured angle of arrival corresponds to the k-th peak. The candidate coordinates correspond to the theoretical angle of arrival.

6. The system according to claim 1, characterized in that, The process by which the evaluation module extracts the discharge pulse waveform of the interference source corresponding to the spatial coordinates of the interference source and constructs the discharge phase-resolved features is as follows: the evaluation module applies a beamforming algorithm to the multi-channel response signal based on the position indicated by the spatial coordinate signal of the interference source, enhances the signal in the direction of that position, extracts the time series of the interference source discharge event, and identifies the waveform of each discharge pulse. Using the power frequency voltage phase at the grid connection point of the wind turbine as the reference phase, the discharge quantity corresponding to each discharge pulse is calculated, and the distribution of discharge quantity with phase over multiple power frequency cycles is statistically analyzed to form a discharge quantity-phase distribution spectrum. Simultaneously, the distribution of discharge pulse repetition rate with pulse amplitude is statistically analyzed to form a pulse repetition rate-amplitude distribution spectrum. The skewness and kurtosis of the discharge quantity-phase distribution spectrum, as well as the peak value, half-width at half-maximum, and asymmetry of the pulse repetition rate-amplitude distribution spectrum, are calculated. These statistical parameters are combined as the discharge phase resolution characteristics of the interference source. The calculation formula for the discharge quantity-phase distribution spectrum is as follows: ； Where S is the skewness of the discharge quantity-phase distribution spectrum, and K is the number of phase sampling points within the power frequency cycle. This represents the discharge quantity corresponding to the p-th phase sampling point. This represents the average discharge amount. This represents the standard deviation of the discharge quantity.

7. The system according to claim 6, characterized in that, The process by which the evaluation module generates a list of interference sources with risk levels is as follows: The pre-set defect electromagnetic fingerprint database stores floating potential discharge fingerprints and tip corona discharge fingerprints calibrated by applying high voltage to typical defects under laboratory conditions. Each fingerprint includes the skewness and kurtosis of the corresponding discharge quantity-phase distribution spectrum and the shape parameters of the pulse repetition rate-amplitude distribution spectrum. The evaluation module calculates the Euclidean distance between the discharge phase resolution characteristics of the interference source and each defect fingerprint in the defect electromagnetic fingerprint database, and selects the one with the smallest distance as the matching defect type. At the same time, based on the spatial coordinate signal of the interference source, the spatial distance from the interference source to the sensitive equipment in the wind turbine is calculated. The sensitive equipment includes the control cabinet, converter, and sensor signal conditioning circuit. Based on the hazard coefficient corresponding to the matching defect type and the spatial distance attenuation function, the interference risk index is calculated; The interference risk index is compared with preset low, medium, and high risk thresholds to determine the risk level, and an interference source list signal containing the spatial coordinates of the interference source, defect type, and risk level is generated. The calculation formula for the interference risk index is as follows: ； Where R is the interference risk index. To match the hazard coefficient corresponding to the defect type, d is the spatial distance from the interference source to the sensitive device. Where S is the distance attenuation characteristic length, and S is the current discharge quantity-phase distribution skewness of the interference source. Let L be the kurtosis of the pulse repetition rate-amplitude distribution of the current interference source, and L be the total number of fingerprint types in the defect electromagnetic fingerprint database. For the skewness of the fingerprint of type n defects, Let be the kurtosis of the nth type of defect fingerprint.

8. The system according to claim 1, characterized in that, The suppression control module also communicates with the wind turbine main control system to receive blade position signals and turbine operating status signals. When the interference source list signal indicates that the interference source is located in the blade tip or trailing edge region and the risk level is higher than a preset threshold, the suppression control module generates a pitch compensation angle command. This command includes the pitch direction and compensation angle value and is output to the pitch actuator. By adjusting the pitch angle of the corresponding blade, the spatial electric field distribution between the blade surface and the tower is changed, so that the local electric field intensity in the blade tip region is lower than the air initiation discharge intensity, thereby suppressing the blade tip discharge.

9. The system according to claim 1, characterized in that, The suppression control module is connected to a relay matrix in the engine compartment. Each channel of the relay matrix corresponds to a local auxiliary shielding cover preset next to the converter cabinet and the generator output box. When the interference source list signal indicates that the interference source is located in a locatable device in the engine compartment and the risk level is higher than a preset threshold, the suppression control module generates a shielding activation command, drives the corresponding relay to close, and connects the local auxiliary shielding cover at that location to the grounding circuit, forming a local electromagnetic shielding around the device and suppressing the leakage interference source generated by the device.

10. A method for locating and suppressing rocket electromagnetic interference sources according to any one of claims 1-9, comprising: S1: Collect the current waveform signal, return stroke position coordinate signal, and occurrence time signal of artificial lightning to form an excitation source reference signal packet; S2: Multiple electromagnetic field sensors are arranged on the wind turbine tower, nacelle and blade root to synchronously collect electromagnetic response signals under lightning excitation and form multi-channel response signals. S3: Receive the excitation source reference signal packet and the multi-channel response signal, call the multipath impulse response function determined by the pre-stored wind turbine structure electromagnetic simulation model to perform deconvolution compensation on the multi-channel response signal, and generate the interference source spatial coordinate signal based on the arrival time difference and arrival angle spectrum of the compensated signal. S4: Receive the spatial coordinate signal of the interference source, extract the discharge pulse waveform of the interference source corresponding to the coordinate, construct the discharge phase resolution feature and compare it with the preset defect electromagnetic fingerprint database to generate an interference source list signal with risk level. S5: Receive the interference source list signal, and generate suppression control command signals to adjust the blade pitch angle, deploy a local shield, or change the yaw direction based on the location and risk level of the interference sources.

Images