Ferromagnetic resonance flexible device based on power electronic converter

Abstract

The invention belongs to the technical field of overvoltage protection of a power system, and particularly relates to a ferromagnetic resonance flexible device based on a power electronic converter, a power electronic converter module adopts a full-bridge uncontrolled rectification and double-switch BUCK chopper circuit topology, and dynamic simulation equivalent damping resistance is adjusted through the duty ratio of an MOSFET (Metal-Oxide-Semiconductor Field Effect Transistor) switch; the signal acquisition module is integrated with a GMR sensor, a T-type thermocouple and a voltage detection unit; the control module carries out ferromagnetic resonance characteristic identification based on an FFT algorithm, combines critical conduction mode control and a model reference adaptive PI control algorithm, dynamically generates a PWM signal to drive the converter to work, and realizes rapid adaptive dissipation of resonance energy. Through multi-parameter fusion detection and a composite control strategy, the response speed and the adjustment precision of ferromagnetic resonance suppression are effectively improved, the functions of state self-diagnosis, parameter visual configuration and remote data interaction are achieved, and the operation reliability of a power system is remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of overvoltage protection of power systems, and in particular relates to a ferromagnetic resonance flexible device based on a power electronic converter. Background Art

[0002] Ferroresistance is a persistent resonance phenomenon in power systems (typically 10kV to 35kV neutral-point non-effectively grounded systems) caused by the mismatch between the parameters of nonlinear inductive components (such as electromagnetic voltage transformers) and the system's ground capacitance. It is often accompanied by overvoltage and overcurrent, leading to serious faults such as equipment insulation damage, fuse blowing, lightning arrester explosion, and transformer burnout. Specifically, when a system disturbance occurs (such as a circuit breaker trip or a single-phase ground fault), the magnetizing inductance of the electromagnetic voltage transformer (PT) can easily enter saturation, forming a resonant circuit with the line-to-ground capacitance, posing a direct threat to power grid security.

[0003] Currently, the suppression measures for ferromagnetic resonance mainly include passive detuning circuits, active detuning circuits and superconducting technology, but all of them have significant limitations.

[0004] The typical solution for passive detuning circuits is to connect a fixed damping resistor in parallel with a saturated reactor (a saturated reactor consists of a reactor and a series resistor). It relies on the energy dissipation of the series resistor after the reactor is saturated, but the response speed is slow (it takes several cycles). The fixed resistance value is difficult to adapt to multi-modal resonances such as fundamental frequency, frequency division and chaos, and the suppression efficiency is sharply reduced in low-load or high-frequency resonance scenarios. Although the resonant RLC filter can suppress resonance by blocking the resonant frequency, it will degrade the system frequency response characteristics and affect the harmonic measurement accuracy. Metal oxide varistors (MOVs) are limited by their fixed conduction threshold and energy capacity, making it difficult to cope with complex and changing resonant modes.

[0005] Active detuning schemes based on switching thyristor damping resistors can shorten the suppression time to within two cycles. However, thyristor state switching can easily cause discontinuous changes in the equivalent resistance, posing the risk of re-igniting resonance, and the complex control circuit reduces reliability. Patent CN 119134245 A proposes controlling the detuning module by detecting the magnetizing inrush current and offset voltage. While this improves the targeted operation, it still suffers from the multi-stage switching issue. Furthermore, its control logic relies on steady-state parameter calibration, which can lead to misjudgments during transient processes or chaotic modes.

[0006] To improve resonance suppression, a series of hybrid solutions and emerging technologies have been proposed. Dynamic impedance adjustment techniques combining RLC tuned filters with harmonic current injection require precise matching of the resonant frequency and are susceptible to parameter drift at high frequencies. Intelligent algorithms such as fuzzy PID control still suffer from regulation lag in chaotic resonance modes. Some designs have attempted to combine MOVs with reactors or series-connect DC inductor fault current limiters. While this improves suppression, it increases system complexity, and parameter tuning relies on manual experience, lacking dynamic adaptability. Patent CN 118137420A proposes using a circuit breaker with a phase-selective tripping function to suppress resonance. This approach weakens transient excitation by controlling the tripping phase, but its suppression effectiveness is affected by the coupling between the circuit breaker's operating sequence and system operating conditions, making it unreliable in complex multimodal resonances. While superconducting fault current limiters can simultaneously suppress short-circuit currents and resonances, their cryogenic cooling systems drive up operation and maintenance costs, making them difficult to adapt to the scale-up requirements of medium-voltage distribution networks. Patent CN 117805518A detects fundamental frequency ferromagnetic resonance based on the second-order derivative of zero-sequence voltage. However, this method is sensitive to high-frequency noise and is easily affected by electromagnetic interference, which affects detection accuracy. It is difficult to effectively coordinate with dynamic suppression strategies, resulting in delayed action or overcompensation. Patent CN 117826059 B predicts resonant faults by establishing a model and performing parameter analysis. However, it lacks adaptability to dynamic changes in grid topology and relies on high-precision parameter input, which limits its practical application.

[0007] In summary, existing methods have technical bottlenecks in terms of speed, wide-band multi-modal adaptability, coordination between detection methods and control strategies, and reliability. There is an urgent need for a flexible suppression device that combines fast response, dynamic impedance matching, multi-modal adaptability and high reliability to meet the stringent requirements of distribution networks for ferromagnetic resonance suppression. Summary of the Invention

[0008] In response to the technical problems of the above-mentioned existing methods in terms of rapidity, wide-band multi-modal adaptability, coordination between detection methods and control strategies, and reliability, the present invention provides a ferromagnetic resonance flexible device based on a power electronic converter, which aims to achieve rapid detection, accurate identification and efficient energy dissipation of ferromagnetic resonance.

[0009] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0010] A ferromagnetic resonance flexible device based on a power electronic converter, comprising:

[0011] The power electronic converter module, connected to the secondary open delta of the voltage transformer, is configured to dynamically simulate an equivalent damping resistor by adjusting the duty cycle of the switching signal to achieve rapid dissipation of resonant energy;

[0012] The signal acquisition module is connected to the secondary side of the voltage transformer, the giant magnetoresistance effect GMR sensor and the T-type thermocouple to collect the secondary side voltage, zero-sequence voltage, primary side excitation current and transformer core temperature;

[0013] The control module connects the signal acquisition module and the power electronic converter module, performs ferromagnetic resonance recognition and identification based on the collected signals, generates a pulse width modulation (PWM) signal to drive the power electronic converter, and implements status indication, parameter configuration, and data interaction with the upper-level monitoring system;

[0014] The damping resistor is connected to the output side of the power electronic converter module and is used to dissipate the resonant energy.

[0015] The power electronic converter module adopts a composite topology structure of a full-bridge uncontrolled rectifier circuit and a dual-switch BUCK chopper circuit, wherein the dual-switch BUCK chopper circuit is connected to the output end of the full-bridge uncontrolled rectifier circuit, and a MOSFET is used as a power switching device, and its output end is connected to a damping resistor.

[0016] The control module includes:

[0017] The resonance feature extraction unit performs fast Fourier transform (FFT) on the excitation current to extract the amplitude and frequency characteristics of the fundamental frequency, frequency division and high-frequency components, and calculates the open triangle voltage amplitude and change rate, the excitation current amplitude and change rate, and the core temperature rise rate;

[0018] The multi-parameter fusion diagnosis unit, based on the feature weight fusion algorithm, integrates the open triangle voltage change rate, excitation current change rate, excitation current distortion rate and temperature rise rate to generate a comprehensive resonance probability value. When the comprehensive resonance probability value exceeds the threshold, it is determined that ferroresonance has occurred.

[0019] The composite control unit adopts an algorithm that combines the fixed on-time control of critical conduction mode (CrCM) with the model reference adaptive PI control based on the voltage harmonic amplitude. It dynamically adjusts the PWM duty cycle according to the resonant mode to achieve continuous adjustment of the equivalent damping resistance.

[0020] The control module also includes a protection and status feedback unit that monitors the core temperature and excitation current in real time. When the temperature exceeds 100°C or the excitation current exceeds 10 times the nominal value, it switches to maximum damping mode and triggers an alarm. The resonance suppression status, damping resistance value, and fault code are uploaded to the upper-level monitoring system via the Modbus / TCP protocol, and a health assessment report is generated.

[0021] The control module continuously monitors the open triangle voltage spectrum after resonance suppression. If no resonance characteristics are detected within 30 seconds, the damping resistance is gradually reduced to the standby value. The resonance mode prediction model is trained based on historical data, the threshold weights of multi-parameter fusion diagnosis are optimized, and the sensors and sampling channels are regularly self-calibrated to ensure that the measurement accuracy error is less than 1%.

[0022] A resonance suppression method for a ferromagnetic resonance flexible device based on a power electronic converter comprises the following steps:

[0023] S1. Multi-dimensional signal acquisition and preprocessing: GMR sensors are used to collect the primary side excitation current, high-precision voltage sensors are used to collect the secondary side three-phase voltage and zero-sequence voltage, T-type thermocouples are used to monitor the core temperature, and the signals are filtered and noise-reduced.

[0024] S2. Resonance Feature Extraction and Pattern Recognition: Perform FFT spectrum analysis on the excitation current to extract the fundamental frequency, frequency division, and high-frequency component features; calculate the zero-sequence voltage change rate, excitation current distortion rate, and core temperature rise rate to identify core nonlinear saturation and overheating risks;

[0025] S3. Multi-parameter fusion diagnosis and resonance determination: Based on the feature weight fusion algorithm, the zero-sequence voltage change rate, excitation current distortion rate and temperature rise rate are integrated to generate a comprehensive resonance probability value. When the probability value exceeds 80%, resonance is determined to have occurred. The fundamental frequency, partial frequency and high-frequency resonance modes are distinguished based on the excitation current spectrum characteristics.

[0026] S4. Dynamic damping adjustment and energy dissipation: The control strategy is selected based on the resonant mode. The fundamental frequency and divided frequency resonance adopt CrCM fixed on-time control. The high-frequency resonance is combined with model-referenced adaptive PI control. The equivalent damping resistor (0.1Ω to 10kΩ) is dynamically adjusted by adjusting the PWM duty cycle. The resonant energy is dissipated through the damping resistor through the full-bridge rectification and buck chopper circuit.

[0027] S5, protection and status feedback: Real-time monitoring of core temperature and excitation current. If they exceed the limit, the system switches to maximum damping mode and issues an alarm. Status data is uploaded via Modbus / TCP protocol and a health assessment report is generated.

[0028] S6. Continuous monitoring and adaptive optimization: After resonance suppression, the open triangle voltage spectrum is continuously monitored. If there is no resonance feature, the damping is reduced to the standby value; the diagnostic threshold weight is optimized based on historical data, and the sensor and sampling channel are self-calibrated regularly.

[0029] The method for multi-parameter fusion diagnosis and resonance determination in S3 is:

[0030] S3.1, based on the feature weight fusion algorithm, the zero-sequence voltage change rate Ku, the excitation current distortion rate I THDAnd the temperature rise rate β generates a comprehensive resonance probability value P:

[0031]

[0032] When P>80%, it is determined that ferromagnetic resonance occurs;

[0033] S3.2. After ferromagnetic resonance occurs, ferromagnetic resonance is identified based on the excitation current spectrum characteristics using the following conditions:

[0034] Fundamental frequency resonance: 50Hz component amplitude>5 times rated excitation current and total excitation current distortion>30%;

[0035] Frequency division resonance: 25Hz / 17Hz component amplitude accounts for >60% of the total excitation current and the total excitation current distortion rate is >20%;

[0036] High-frequency resonance: The excitation current amplitude is greater than 8 times the rated excitation current, the amplitude of the 150Hz~1kHz frequency band component accounts for more than 20%, and the excitation current distortion rate is greater than 60%.

[0037] The method of dynamic damping adjustment and energy dissipation in S4 is:

[0038] S4.1. Prioritize control strategies based on resonance modes

[0039] The control circuit adopts a control algorithm that combines fixed on-time control based on critical conduction mode (CrCM) with model reference adaptive PI control based on voltage harmonic amplitude, aiming to achieve high-precision and stable control of the output voltage while ensuring high system efficiency.

[0040] Fundamental frequency / divided frequency resonance: Using critical conduction mode (CrCM) fixed on-time control, the equivalent damping resistance is continuously adjusted within the range of 0.1Ω to 1kΩ by adjusting the PWM duty cycle;

[0041] High-frequency resonance: Combined with critical conduction mode (CrCM), model-referenced adaptive PI control is enabled to dynamically optimize PI parameters based on the resonant energy intensity, and the duty cycle is adjusted to achieve fast switching of high resistance values from 1kΩ to 10kΩ.

[0042] S4.2, the dual-switch buck chopper circuit updates the PWM signal with a 50us regulation cycle to ensure that the equivalent resistance changes smoothly without step changes;

[0043] S4.3. The resonant energy is converted into a DC component through the full-bridge uncontrolled rectifier circuit and the BUCK chopper circuit, and is quickly released in the form of heat dissipation through the damping resistor.

[0044] The method for selecting the control strategy according to the resonance mode priority in S4.1 is:

[0045] S4.1.1, Fixed on-time CrCM control mode

[0046] Using a fixed on-time T on The strategy is the CrCM control mode. In the CrCM working mode, the inductor current starts at zero and returns to zero in each switching cycle, which means that quasi-zero current switching (ZCS) is achieved, which greatly reduces the switching loss.

[0047] MOSFET devices Q1 and Q2 are connected through a logic NOT gate. In each cycle, Q1 and Q2 calculate the output T of the model reference adaptive PI control. on The MOSFET conduction control is determined by the inductor current returning to zero, that is, the Q1 leakage current is close to zero, and the sampling resistor R Q Can collect characteristic voltage u Q1 , when u Q1 >H: It means the current is not zero yet, Q2 continues to conduct. Q1 ≤H: The inductor current is zero, Q2 is turned off, and the control process requires a zero-crossing detection circuit ZCD and an RS trigger to ensure the CrCM condition and have a high control response speed. The triangle wave comparator provides a ramp signal, which is compared with the reference voltage to set the on-time;

[0048] S4.1.2 Model reference adaptive PI control based on voltage harmonic amplitude

[0049] The model reference adaptive control (MRAC) mechanism is introduced to perform adaptive parameter adjustment on the PI controller to enhance the robustness and tracking capability of the system.

[0050] The transfer function of the Buck chopper circuit is:

[0051]

[0052] The transfer function of the PI controller is:

[0053]

[0054] The transfer function of the closed-loop system is:

[0055]

[0056] Then we get:

[0057]

[0058] The above formula is the duty cycle-impedance mapping model of the Buck chopper circuit. Based on the model, a reference model with a structure similar to that with ideal output characteristics is set:

[0059]

[0060] The system obtains the output voltage u through the sampling circuit o , the error is

[0061] e=u ref ―u o

[0062] Where: u ref =3kI0λ I R, I0 is the primary side rated excitation current, λ I is the current multiple, k is the transformer ratio;

[0063] Introducing the Lilapunov performance function

[0064]

[0065] According to the MIT gradient rule, the time rate of change of θ is proportional to the negative gradient J, which is Where e is the model error, θ is the controller parameter vector, The component of is the sensitivity derivative of the error to θ, and the parameter γ is the adaptive gain;

[0066] Apply the MIT gradient rule to determine the value of the PI controller parameter K P and K I .

[0067]

[0068] In the formula

[0069]

[0070] The adaptive rule is obtained:

[0071]

[0072] Set a in the reference model m1 =RLC, a m2 =L、a m3 =(1+DK P )R、a m4 =DRK I , the adaptive gain γ is defined as follows: γ P =-0.0005,γ I =-0.0005;

[0073] The protection and status feedback method in S5 is:

[0074] S5.1, Real-time monitoring of core temperature and excitation current:

[0075] If the temperature is greater than 100°C or the excitation current is greater than 10 times the nominal value, it will immediately switch to the maximum damping mode and trigger an alarm;

[0076] S5.2. Upload the resonance suppression status, damping resistance value and fault code to the upper-level monitoring system via Modbus / TCP protocol;

[0077] S5.3. Generate a health assessment report every 5 minutes, including cumulative energy consumption, device temperature rise, and control parameter offset.

[0078] The method of continuous monitoring and adaptive optimization in S6 is:

[0079] S6.1. After resonance suppression, continuously monitor the open triangle voltage spectrum. If no resonance characteristics are detected within 30 seconds, gradually reduce the damping until the damping resistance reaches the standby value.

[0080] S6.2. Train the resonance mode prediction model based on historical data and optimize the threshold weights of multi-parameter fusion diagnosis;

[0081] S6.3. Regularly self-calibrate the GMR sensor and voltage sampling channel to ensure measurement accuracy error <1%.

[0082] Compared with the prior art, the present invention has the following beneficial effects:

[0083] 1. The present invention uses FFT spectrum analysis and multi-parameter threshold judgment to complete the identification of fundamental frequency, frequency division and high-frequency resonance within one power frequency cycle. The resonance detection time is shortened to less than 20ms, and the suppression action is fully controlled within three power frequency cycles (<60ms), which is more than 5 times faster than traditional passive solutions. It supports 10Hz to 2kHz wide-band resonance suppression, covering fundamental frequency, frequency division and high-frequency multi-modal scenarios.

[0084] 2. The present invention is based on the duty cycle-impedance mapping model of the BUCK chopper circuit. Through the composite algorithm of CrCM and adaptive PI control, it realizes continuous and stepless adjustment of the equivalent damping resistance according to the resonant energy intensity, adapts to the impedance requirements of different resonant modes, eliminates the risk of secondary resonance caused by thyristor switching, and avoids the underdamping or overdamping problems caused by traditional fixed resistors.

[0085] 3. When the core temperature exceeds the limit (threshold of 100°C) or the excitation current exceeds the threshold (10 times the nominal value), this invention activates temperature-current coordinated protection, automatically switching to maximum damping mode and triggering an alarm to prevent equipment overheating and damage, ensuring high reliability and compatibility. Furthermore, this invention uses GMR sensors to avoid measurement distortion caused by traditional CT saturation. This invention supports Modbus / TCP protocol interaction with higher-level monitoring systems, enabling remote parameter setting and status monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are merely exemplary, and those skilled in the art can, without inventive effort, derive other implementation drawings based on the provided drawings.

[0087] The structures, proportions, sizes, etc. illustrated in this specification are intended solely to complement the contents disclosed herein and to facilitate understanding and reading by persons skilled in the art. They are not intended to limit the conditions under which the present invention may be implemented and therefore have no substantive technical significance. Any structural modifications, changes in proportions, or adjustments in sizes, without affecting the efficacy and objectives of the present invention, shall remain within the scope of the technical contents disclosed herein.

[0088] Figure 1 This is a block diagram of the overall structure of the device of the present invention;

[0089] Figure 2 This is a control structure diagram of the power electronic converter of the present invention;

[0090] Figure 3 This is a flow chart of the control algorithm of the present invention;

[0091] Figure 4 This is a control block diagram of the model reference adaptive PI control based on voltage harmonic amplitude of the present invention.

[0092] Among them: 1 is the power electronic converter module, 2 is the signal acquisition module, 3 is the control module, and 4 is the damping resistor. DETAILED DESCRIPTION

[0093] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only part of the embodiments of this application, not all the embodiments. These descriptions are only to further illustrate the features and advantages of the present invention, rather than to limit the claims of the present invention. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0094] The following embodiments of the present invention are described in further detail with reference to the accompanying drawings and examples. The following embodiments are used to illustrate the present invention but are not intended to limit the scope of the present invention.

[0095] In the description of this application, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0096] This embodiment provides a ferromagnetic resonance flexible device based on a power electronic converter, such as Figure 1 Shown, including:

[0097] Power electronic converter module 1: This module utilizes a composite topology combining a full-bridge uncontrolled rectifier circuit with a dual-switch buck chopper circuit, with MOSFET power devices forming the dual-switch buck circuit. This module connects to the secondary open-delta winding of the voltage transformer, with its output connected to damping resistor 4. By adjusting the PWM duty cycle, the equivalent damping resistor value (ranging from 0.1Ω to 10kΩ) is dynamically adjusted to achieve rapid and flexible dissipation of resonant energy.

[0098] Signal acquisition module 2: A sensor based on the giant magnetoresistance effect (GMR) collects the primary-side excitation current of the three-phase ABC in real time, with a sensitivity of 0.1A. A high-precision voltage sensor collects the secondary-side three-phase voltage and open-delta zero-sequence voltage with a sampling frequency of 10kHz. A pre-buried T-type thermocouple monitors the transformer core temperature (resolution of ±0.5°C) to construct a multi-dimensional resonant characteristic.

[0099] Control module 3: Integrates ferromagnetic resonance identification, adaptive control, and communication functions, including:

[0100] Resonance feature extraction: Analyze the excitation current spectrum based on fast Fourier transform (FFT) to identify fundamental frequency (50Hz), crossover frequency (25Hz / 17Hz) and high frequency (150Hz~1kHz) resonance modes;

[0101] Multi-parameter fusion diagnosis: Combines the comprehensive criteria of zero-sequence voltage change rate (threshold > 10V / ms), excitation current distortion rate (threshold > 15%) and core temperature rise rate (threshold > 5°C / s) to improve the reliability of resonance detection;

[0102] Composite control algorithm: Combining critical conduction mode (CrCM) fixed on-time control with model reference adaptive PI control, the PWM duty cycle is dynamically adjusted according to the resonant frequency and energy intensity (adjustment period < 2ms) to achieve continuous and smooth adjustment of the damping resistance value.

[0103] In order to achieve the purpose of suppressing ferromagnetic resonance, the power electronic converter module 1, the signal acquisition module 2, the control module 3 and the damping resistor 4 in the present invention constitute a complete suppression device, and the power electronic converter control structure diagram is shown in FIG. Figure 2 shown.

[0104] The control structure of the power electronic converter consists of the power electronic conversion circuit and the control circuit. The key parameters in the power electronic conversion circuit - the damping resistor R and the freewheeling inductor L - are selected as follows:

[0105]

[0106] Set the switching frequency to f k =20kHz, when the duty cycle is 50%, the on-time of MOSFET switch tube Q2 u △ When the value is 200V and the duty cycle is 50%, u o For 100V, take R = 10Ω and from formula (2) we can get L = 25mH.

[0107] The control circuit completes the suppression of overvoltage, such as Figure 3 As shown, follow these steps to run:

[0108] Step 1: Real-time acquisition and preprocessing of multi-dimensional signals

[0109] The primary-side excitation current of the three-phase ABC is collected in real time using a giant magnetoresistance (GMR) sensor with a sensitivity of 0.1A and a sampling frequency of 10kHz. A high-precision voltage sensor is used to synchronously collect the three-phase voltage and open-delta zero-sequence voltage on the secondary side of the voltage transformer. The transformer core temperature is monitored using a pre-buried T-type thermocouple with a temperature resolution of ±0.5°C. The collected current, voltage, and temperature signals are filtered and denoised to eliminate high-frequency interference and baseline drift.

[0110] Step 2: Resonance feature extraction and pattern recognition

[0111] Perform fast Fourier transform (FFT) spectrum analysis on the excitation current to extract the amplitude and frequency characteristics of the fundamental frequency (50Hz), frequency division (25Hz / 17Hz), and high-frequency (150Hz-1kHz) components; calculate the zero-sequence voltage change rate Ku (ΔU / Δt); if it exceeds the threshold of 10V / ms, it is marked as a transient resonance risk; analyze the excitation current waveform distortion rate I THD When the distortion rate is greater than 15%, it is determined to be a core nonlinear saturation characteristic; the core temperature rise rate β is monitored, and if it exceeds 5°C / s, an overheating warning is triggered. When all three conditions are met, further comprehensive judgment is made.

[0112] Step 3: Multi-parameter fusion diagnosis and resonance determination

[0113] Step 3.1: Based on the feature weight fusion algorithm, the zero-sequence voltage change rate Ku and the excitation current distortion rate I are integrated. THD and the temperature rise rate β to generate the comprehensive resonance probability value P

[0114]

[0115] When P>80%, it is determined that ferromagnetic resonance occurs.

[0116] Step 3.2: After ferromagnetic resonance occurs, ferromagnetic resonance is identified based on the excitation current spectrum characteristics using the following conditions.

[0117] Fundamental frequency resonance: 50Hz component amplitude>5 times rated excitation current and total excitation current distortion>30%;

[0118] Frequency division resonance: 25Hz / 17Hz component amplitude accounts for >60% of the total excitation current and the total excitation current distortion rate is >20%;

[0119] High-frequency resonance: The excitation current amplitude is greater than 8 times the rated excitation current, the amplitude of the 150Hz~1kHz frequency band component accounts for more than 20%, and the excitation current distortion rate is greater than 60%.

[0120] Step 4: Dynamic Damping Adjustment and Energy Dissipation

[0121] Step 4.1: Prioritize the control strategy based on the resonance mode

[0122] The control circuit adopts a control algorithm that combines fixed on-time control based on critical conduction mode (CrCM) with model-referenced adaptive PI control based on the voltage harmonic amplitude. It aims to achieve high-precision and stable control of the output voltage while ensuring high system efficiency.

[0123] Fundamental frequency / divided frequency resonance: Critical conduction mode (CrCM) fixed on-time control is used, and the equivalent damping resistance is continuously adjusted in the range of 0.1Ω to 1kΩ by adjusting the PWM duty cycle (step size 0.1%);

[0124] High-frequency resonance: Combined with critical conduction mode (CrCM), model-referenced adaptive PI control is enabled to dynamically optimize PI parameters based on the resonant energy intensity and adjust the duty cycle to achieve fast switching of high resistance values from 1kΩ to 10kΩ.

[0125] Step 4.1.1, Fixed on-time CrCM control mode

[0126] In order to simplify the control circuit and reduce the cost and noise introduced by the current detection circuit, this patent adopts a fixed on-time (T on) strategy, also known as CrCM control mode. In CrCM operating mode, the inductor current starts at zero and returns to zero during each switching cycle, achieving quasi-zero current switching (ZCS), which greatly reduces switching losses. The specific working process is as follows:

[0127] MOSFET devices Q1 and Q2 are connected through a logic NOT gate. In each cycle, Q1 and Q2 calculate the output T of the model reference adaptive PI control. on The MOSFET conduction control is determined by the inductor current returning to zero, that is, the Q1 leakage current is close to zero, and the sampling resistor R Q Can collect characteristic voltage u Q1 , when u Q1 >H (H is set to 0.2V): This means the current is not zero yet, Q2 continues to conduct. Q1 ≤H: The inductor current is zero, and Q2 is off. The control process requires a zero-crossing detection circuit (ZCD) and an RS trigger to ensure CrCM conditions and high control response speed. A triangle wave comparator provides a ramp signal, which is compared with a reference voltage to set the on-time.

[0128] Step 4.1.2: Model reference adaptive PI control based on voltage harmonic amplitude

[0129] CrCM itself is prone to increase in output voltage ripple and power supply jitter under light load and input changes. Traditional PI controllers are difficult to maintain good dynamic response in the entire operating range. To this end, the model reference adaptive control (MRAC) mechanism is introduced, such as Figure 4 As shown in the figure, the PI controller is adaptively adjusted to enhance the robustness and tracking capability of the system.

[0130] The transfer function of the Buck chopper circuit is:

[0131]

[0132] The transfer function of the PI controller is:

[0133]

[0134] The transfer function of the closed-loop system is: )

[0135]

[0136] Then we can get:

[0137]

[0138] Formula (7) is the duty cycle-impedance mapping model of the Buck chopper circuit. According to Formula (7), a reference model with a structure similar to that with ideal output characteristics is set:

[0139]

[0140] The system obtains the output voltage u through the sampling circuit o , the error is

[0141] e=u ref ―u o (9)

[0142] Where: u ref =3kI0λ I R, I0 is the primary side rated excitation current, λ I is the current multiple, and k is the transformer ratio.

[0143] Introducing the Lilapunov performance function

[0144]

[0145] According to the MIT gradient rule, the time rate of change of θ is proportional to the negative gradient J, which is Where e is the model error, θ is the controller parameter vector, The component of is the sensitivity derivative of the error to θ, and the parameter γ is the adaptive gain.

[0146] Apply the MIT gradient rule to determine the value of the PI controller parameter K in equation (7) P and K I .

[0147]

[0148] In the formula

[0149]

[0150] From equations (11) to (14), we can get the adaptive rule:

[0151]

[0152] K P , K I By integrating equations (15) and (16), we can get: m1 =RLC, a m2 =L、a m3 =(1+DK P )R、a m4 =DRK I, the adaptive gain γ is defined as follows: γ P =-0.0005,γ I =-0.0005.

[0153] Step 4.2: The dual-switch buck chopper circuit updates the PWM signal with a 50µs regulation period to ensure that the equivalent resistance changes smoothly without step changes.

[0154] Step 4.3: The resonant energy is converted into a DC component through the full-bridge uncontrolled rectifier circuit and the BUCK chopper circuit, and is quickly released in the form of heat dissipation through the damping resistor 4.

[0155] Step 5: Protection and Status Feedback

[0156] Step 5.1: Real-time monitoring of core temperature and excitation current:

[0157] If the temperature is greater than 100°C or the excitation current is greater than 10 times the nominal value, it will immediately switch to the maximum damping mode (0.1Ω) and trigger an alarm;

[0158] Step 5.2: Upload the resonance suppression status, damping resistance value, and fault code to the upper-level monitoring system via the Modbus / TCP protocol.

[0159] Step 5.3: Generate a health assessment report every 5 minutes, including the accumulated energy consumption, device temperature rise, and control parameter offset.

[0160] Step 6: Continuous Monitoring and Adaptive Optimization

[0161] Step 6.1: After the resonance is suppressed, the open triangle voltage spectrum is continuously monitored. If no resonance feature is detected within 30 seconds, the damping is gradually reduced so that the damping resistor 4 is changed to the standby value (10 kΩ);

[0162] Step 6.2: Train the resonance mode prediction model based on historical data and optimize the threshold weight of multi-parameter fusion diagnosis;

[0163] Step 6.3: Regularly self-calibrate the GMR sensor and voltage sampling channel to ensure that the measurement accuracy error is <1%.

[0164] The above only describes in detail the preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Various changes can be made within the knowledge of ordinary technicians in this field without departing from the purpose of the present invention, and various changes should be included in the scope of protection of the present invention.

Claims

1. A ferromagnetic resonance flexible device based on a power electronic converter, characterized in that: include: A power electronic converter module (1), connected to the secondary open triangle of the voltage transformer, is configured to dynamically simulate an equivalent damping resistor by adjusting the duty cycle of a switching signal to achieve rapid dissipation of resonant energy; A signal acquisition module (2) is connected to the secondary side of the voltage transformer, a giant magnetoresistance effect GMR sensor and a T-type thermocouple, and is used to collect the secondary side voltage, zero sequence voltage, primary side excitation current and transformer core temperature; The control module (3) is connected to the signal acquisition module (2) and the power electronic converter module (1), performs ferromagnetic resonance recognition and identification based on the acquired signals, generates a pulse width modulation (PWM) signal to drive the power electronic converter, and realizes status indication, parameter configuration, and data interaction with the upper-level monitoring system; The damping resistor (4) is connected to the output side of the power electronic converter module (1) and is used to dissipate resonance energy.

2. The ferromagnetic resonance flexible device based on a power electronic converter according to claim 1, characterized in that: The power electronic converter module (1) adopts a composite topology structure of a full-bridge uncontrolled rectifier circuit and a dual-switch BUCK chopper circuit, wherein the dual-switch BUCK chopper circuit is connected to the output end of the full-bridge uncontrolled rectifier circuit, and a MOSFET is used as a power switch device, and its output end is connected to a damping resistor (4).

3. The ferromagnetic resonance flexible device based on a power electronic converter according to claim 1, characterized in that: The control module (3) comprises: The resonance feature extraction unit performs fast Fourier transform (FFT) on the excitation current to extract the amplitude and frequency characteristics of the fundamental frequency, frequency division and high-frequency components, and calculates the open triangle voltage amplitude and change rate, the excitation current amplitude and change rate, and the core temperature rise rate; The multi-parameter fusion diagnosis unit, based on the feature weight fusion algorithm, integrates the open triangle voltage change rate, excitation current change rate, excitation current distortion rate and temperature rise rate to generate a comprehensive resonance probability value. When the comprehensive resonance probability value exceeds the threshold, it is determined that ferroresonance has occurred. The composite control unit adopts an algorithm that combines the fixed on-time control of critical conduction mode (CrCM) with the model reference adaptive PI control based on the voltage harmonic amplitude. It dynamically adjusts the PWM duty cycle according to the resonant mode to achieve continuous adjustment of the equivalent damping resistance.

4. The ferromagnetic resonance flexible device based on a power electronic converter according to claim 1, characterized in that: The control module (3) also includes a protection and status feedback unit, which monitors the core temperature and excitation current in real time. When the temperature exceeds 100°C or the excitation current exceeds 10 times the nominal value, it switches to the maximum damping mode and triggers an alarm; uploads the resonance suppression status, damping resistance value and fault code to the upper monitoring system via the Modbus / TCP protocol, and generates a health assessment report; The control module (3) continuously monitors the open triangle voltage spectrum after resonance suppression, and gradually reduces the damping resistance (4) to a standby value if no resonance feature is detected within 30 seconds; trains a resonance mode prediction model based on historical data, optimizes the threshold weight of multi-parameter fusion diagnosis, and regularly self-calibrates sensors and sampling channels to ensure that the measurement accuracy error is less than 1%.

5. A resonance suppression method for a ferromagnetic resonance flexible device based on a power electronic converter according to any one of claims 1 to 4, characterized in that: The following steps are involved: S1. Multi-dimensional signal acquisition and preprocessing: GMR sensors are used to collect the primary side excitation current, high-precision voltage sensors are used to collect the secondary side three-phase voltage and zero-sequence voltage, T-type thermocouples are used to monitor the core temperature, and the signals are filtered and noise-reduced. S2. Resonance feature extraction and pattern recognition: Perform FFT spectrum analysis on the excitation current to extract the fundamental frequency, frequency division, and high-frequency component features; Calculate the zero-sequence voltage change rate, excitation current distortion rate, and core temperature rise rate to identify core nonlinear saturation and overheating risks; S3. Multi-parameter fusion diagnosis and resonance determination: Based on the feature weight fusion algorithm, the zero-sequence voltage change rate, excitation current distortion rate and temperature rise rate are integrated to generate a comprehensive resonance probability value. When the probability value exceeds 80%, resonance is determined to have occurred. The fundamental frequency, partial frequency and high-frequency resonance modes are distinguished based on the excitation current spectrum characteristics. S4, dynamic damping adjustment and energy dissipation: the control strategy is selected according to the resonance mode. The fundamental frequency / divided frequency resonance adopts CrCM fixed on-time control, and the high frequency resonance is combined with the model reference adaptive PI control. The equivalent damping voltage is dynamically adjusted by adjusting the PWM duty cycle. The resonance energy is dissipated through the damping resistor (4) through the full-bridge rectification and BUCK chopper circuit. S5, protection and status feedback: Real-time monitoring of core temperature and excitation current. If they exceed the limit, the system switches to maximum damping mode and issues an alarm. Status data is uploaded via Modbus / TCP protocol and a health assessment report is generated. S6. Continuous monitoring and adaptive optimization: After resonance suppression, the open triangle voltage spectrum is continuously monitored. If there is no resonance feature, the damping is reduced to the standby value; the diagnostic threshold weight is optimized based on historical data, and the sensor and sampling channel are self-calibrated regularly.

6. The resonance suppression method of a ferromagnetic resonance flexible device based on a power electronic converter according to claim 5, characterized in that: The method for multi-parameter fusion diagnosis and resonance determination in S3 is: S3.1, based on the feature weight fusion algorithm, the zero-sequence voltage change rate Ku, the excitation current distortion rate I THD And the temperature rise rate β generates a comprehensive resonance probability value P: When P>80%, it is determined that ferromagnetic resonance occurs; S3.

2. After ferromagnetic resonance occurs, ferromagnetic resonance is identified based on the excitation current spectrum characteristics using the following conditions: Fundamental frequency resonance: 50Hz component amplitude>5 times rated excitation current and total excitation current distortion>30%; Frequency division resonance: 25Hz / 17Hz component amplitude accounts for >60% of the total excitation current and the total excitation current distortion rate is >20%; High-frequency resonance: The excitation current amplitude is greater than 8 times the rated excitation current, the amplitude of the 150Hz~1kHz frequency band component accounts for more than 20%, and the excitation current distortion rate is greater than 60%.

7. The resonance suppression method of a ferromagnetic resonance flexible device based on a power electronic converter according to claim 5, characterized in that: The method of dynamic damping adjustment and energy dissipation in S4 is: S4.

1. Prioritize control strategies based on resonance modes The control circuit uses a control algorithm that combines fixed on-time control based on critical conduction mode (CrCM) with model-referenced adaptive PI control based on voltage harmonic amplitudes. This algorithm aims to achieve high-precision and stable control of the output voltage while ensuring high system efficiency. Fundamental frequency / divided frequency resonance: Using critical conduction mode (CrCM) fixed on-time control, the equivalent damping resistance is continuously adjusted within the range of 0.1Ω to 1kΩ by adjusting the PWM duty cycle; High-frequency resonance: Combined with critical conduction mode (CrCM), model-referenced adaptive PI control is enabled to dynamically optimize PI parameters based on the resonant energy intensity, and the duty cycle is adjusted to achieve fast switching of high resistance values from 1kΩ to 10kΩ. S4.2, the dual-switch buck chopper circuit updates the PWM signal with a 50us regulation cycle to ensure that the equivalent resistance changes smoothly without step changes; S4.

3. The resonant energy is converted into a DC component through the full-bridge uncontrolled rectifier circuit and the BUCK chopper circuit, and is quickly released in the form of heat dissipation through the damping resistor (4).

8. The resonance suppression method of a ferromagnetic resonance flexible device based on a power electronic converter according to claim 7, characterized in that: The method for preferentially selecting the control strategy according to the resonance mode in S4.1 is: S4.1.1, Fixed on-time CrCM control mode Using a fixed on-time T on The strategy is the CrCM control mode. In the CrCM working mode, the inductor current starts at zero and returns to zero in each switching cycle, which means that quasi-zero current switching (ZCS) is achieved, which greatly reduces the switching loss. MOSFET devices Q1 and Q2 are connected through a logic NOT gate. In each cycle, Q1 and Q2 calculate the output T of the model reference adaptive PI control. on On and off; the MOSFET conduction control is determined by the inductor current returning to zero, that is, the Q1 leakage current is close to zero, and the sampling resistor R Q Can collect characteristic voltage u Q1 , when u Q1 >H: It means the current is not zero yet, Q2 continues to conduct. Q1 ≤H: The inductor current is zero, Q2 is turned off, and the control process requires a zero-crossing detection circuit ZCD and an RS trigger to ensure the CrCM condition and have a high control response speed; the triangle wave comparator provides a ramp signal, which is compared with the reference voltage to set the on-time; S4.1.2 Model reference adaptive PI control based on voltage harmonic amplitude The model reference adaptive control (MRAC) mechanism is introduced to perform adaptive parameter adjustment on the PI controller to enhance the robustness and tracking capability of the system. The transfer function of the Buck chopper circuit is: The transfer function of the PI controller is: The transfer function of the closed-loop system is: Then we get: The above formula is the duty cycle-impedance mapping model of the Buck chopper circuit. Based on the model, a reference model with a structure similar to that with ideal output characteristics is set: The system obtains the output voltage u through the sampling circuit o , the error is e=u ref ―u o Where: u ref =3kI0λ I R, I0 is the primary side rated excitation current, λ I is the current multiple, k is the transformer ratio; Introducing the Lilapunov performance function According to the MIT gradient rule, the time rate of change of θ is proportional to the negative gradient J, which is Where e is the model error, θ is the controller parameter vector, The component of is the sensitivity derivative of the error to θ, and the parameter γ is the adaptive gain; Apply the MIT gradient rule to determine the value of the PI controller parameter K P and K I ; In the formula The adaptive rule is obtained: Set a in the reference model m1 =RLC, a m2 =L、a m3 =(1+DK P )R、a m4 =DRK I , the adaptive gain γ is defined as follows: γ P =-0.0005,γ I =-0.0005.

9. The resonance suppression method of a ferromagnetic resonance flexible device based on a power electronic converter according to claim 5, characterized in that: The protection and status feedback method in S5 is: S5.1, Real-time monitoring of core temperature and excitation current: If the temperature is greater than 100°C or the excitation current is greater than 10 times the nominal value, it will immediately switch to the maximum damping mode and trigger an alarm; S5.

2. Upload the resonance suppression status, damping resistance value and fault code to the upper-level monitoring system via Modbus / TCP protocol; S5.

3. Generate a health assessment report every 5 minutes, including cumulative energy consumption, device temperature rise, and control parameter offset.

10. The resonance suppression method of a ferromagnetic resonance flexible device based on a power electronic converter according to claim 5, characterized in that: The method of continuous monitoring and adaptive optimization in S6 is: S6.

1. After the resonance is suppressed, the open triangle voltage spectrum is continuously monitored. If no resonance feature is detected within 30 seconds, the damping is gradually reduced so that the damping resistor (4) is changed to the standby value; S6.

2. Train the resonance mode prediction model based on historical data and optimize the threshold weights of multi-parameter fusion diagnosis; S6.

3. Regularly self-calibrate the GMR sensor and voltage sampling channel to ensure measurement accuracy error <1%.

Images