Power drive compensation method and system for monitoring magnetostrictive load characteristics

Abstract

This invention belongs to the field of magnetic variable measurement technology, specifically relating to a power-driven compensation method and system for monitoring the characteristics of magnetostrictive loads. The method includes: obtaining a reference zero point for the evolution of magnetic performance under cold conditions based on the center frequency of the electromagnetic system; calculating the current dominant frequency impedance magnitude and phase difference using an electrical parameter characterization method, and obtaining a thermal drift index reflecting the degree of hysteresis loop distortion in conjunction with the reference zero point; obtaining correction factors of various orders based on the thermal drift index and the average output power at the current moment, and dynamically updating the coefficient vector of the inverse model; constructing a pre-distortion driving voltage signal that can cancel nonlinear interference using the new coefficients and an ideal guided wave voltage sequence, and inputting it to a power amplifier to drive the magnetostrictive transducer. This invention solves the problem of pre-distortion failure in existing technologies under variable temperature conditions, significantly improving the signal-to-noise ratio of high-power, long-distance guided wave detection.

Description

Technical Field

[0001] This invention relates to the field of magnetic variable measurement technology. More specifically, this invention relates to a power drive compensation method and system for monitoring the characteristics of magnetostrictive loads. Background Technology

[0002] Magnetostrictive guided wave testing technology is widely used in non-destructive testing of long-distance structures such as pipelines and cables due to its advantages such as non-contact operation and long detection distance. This technology uses the magnetostrictive effect to convert alternating magnetic fields into mechanical vibrations, thereby exciting ultrasonic guided waves in the components. In order to achieve long-distance detection, a high-power amplifier is usually required to drive the magnetostrictive transducer to obtain sufficient energy.

[0003] However, magnetostrictive materials have inherent nonlinear hysteresis characteristics; when the driving signal is a standard sine wave or a Hanning window modulated wave, the actual mechanical vibration waveform generated on the pipe will be distorted, producing high-order harmonics; these harmonic components are very likely to excite unexpected redundant modes in the waveguide, resulting in a decrease in the signal-to-noise ratio of the echo signal.

[0004] Existing technologies typically employ model-based predistortion techniques to compensate for this nonlinearity, which involves pre-generating a distorted electrical signal with characteristics opposite to the nonlinearity so that it cancels out the nonlinearity after passing through the transducer.

[0005] Existing predistortion model parameters are usually fixed values ​​obtained through offline calibration in an ideal laboratory environment. However, in actual engineering testing, especially when high-power continuous drive is used to pursue long-distance testing, the magnetostrictive transducer coil will generate significant Joule heating, leading to an increase in transducer temperature. The temperature change will directly change the permeability and hysteresis loop shape of the magnetostrictive material, resulting in thermal drift. At this time, the fixed predistortion parameters based on cold-state calibration will no longer match the current thermal nonlinear characteristics, causing waveform correction to fail. Summary of the Invention

[0006] To address the technical problem in the prior art where thermal drift of magnetostrictive materials under high-power continuous drive leads to the failure of fixed pre-distortion parameters based on cold-state calibration, resulting in waveform distortion and reduced detection signal-to-noise ratio, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a power drive compensation method for monitoring the characteristics of magnetostrictive loads, comprising: determining the center frequency of an electromagnetic system; obtaining a reference impedance magnitude and a reference phase difference under cold conditions of the electromagnetic system as a reference zero point for the evolution of magnetic performance; and constructing an inverse model to obtain an initial coefficient vector; acquiring voltage and current data at the output of a power amplifier in real time; calculating the current dominant frequency impedance magnitude and phase difference based on the voltage and current data; and obtaining a thermal drift index reflecting the degree of hysteresis loop distortion by combining the reference impedance magnitude and the reference phase difference; obtaining correction factors of various orders based on the thermal drift index and the average output power at the current moment; dynamically updating the coefficient vector of the inverse model using the correction factors of various orders to obtain new coefficients at the current moment; constructing a predistortion drive voltage signal that can cancel nonlinear interference using the new coefficients at the current moment and an ideal guided wave voltage sequence; and inputting the predistortion drive voltage signal to a power amplifier to drive a magnetostrictive transducer, thereby achieving accurate restoration of the electrical characteristic indicators of the magnetostrictive load under variable temperature conditions.

[0008] This invention establishes a reference impedance magnitude and a reference phase difference as physical reference zero points in a cold system environment, and collects voltage and current data in real time during operation to monitor changes in the dominant frequency impedance and phase, thereby constructing a thermal drift index that reflects the degree of hysteresis loop distortion. This thermal drift index is used to dynamically update the inverse model coefficients with the current output power, generating a pre-distortion drive voltage signal that matches the current thermal nonlinear characteristics in real time. Utilizing the mapping mechanism of electrical parameters to changes in the material's microscopic magnetic properties, real-time sensing and closed-loop compensation of thermal drift phenomena are achieved without the need for external temperature sensors. This ensures that the system can accurately cancel nonlinear interference under varying temperature conditions, effectively suppressing the excitation of higher harmonics and redundant modes, and significantly improving the purity and signal-to-noise ratio of long-distance guided wave detection signals.

[0009] Preferably, the ideal guided wave voltage sequence is obtained as follows: the number of cycles of the excitation signal is set, and a Hanning window function is constructed based on the center frequency of the electromagnetic system; the Hanning window function is multiplied with the carrier sine wave to obtain the time-domain expression of the ideal guided wave voltage sequence, and then the voltage data at each moment is calculated; the calculated voltage data are arranged in chronological order to form the ideal guided wave voltage sequence.

[0010] This invention generates an ideal guided wave voltage sequence by setting the number of cycles of the excitation signal and modulating the carrier sine wave using the Hanning window function. This time-domain windowing process effectively suppresses the sidelobe effect of the signal, avoids spectral energy leakage, and ensures that the waveform used as the control target has a high degree of concentration in the frequency domain. This excites an ultrasonic guided wave with a single mode and weak dispersion effect in the component under test, thereby improving the identification of defect echoes.

[0011] Preferably, the step of constructing the inverse model to obtain the initial coefficient vector includes: injecting a sweep frequency signal into the power amplifier through a signal generator to acquire the voltage data sequence at the output of the power amplifier; constructing an observation matrix containing the odd power features of the voltage data sequence; and using the least squares method to calculate the mapping relationship between the observation matrix and the ideal guided wave voltage sequence in combination with the ideal guided wave voltage sequence to obtain the initial coefficient vector.

[0012] Preferably, obtaining the reference impedance magnitude and reference phase difference in a cold-state environment of the electromagnetic system includes: in a cold-state environment of the electromagnetic system, controlling a signal generator to transmit a standard pulse signal with a carrier frequency equal to the center frequency of the electromagnetic system; acquiring time-domain voltage data and time-domain current data at the output of the power amplifier and performing a fast Fourier transform to extract the complex impedance at the center frequency of the electromagnetic system; calculating the magnitude of the complex impedance as the reference impedance magnitude, and calculating the arctangent of the real and imaginary parts of the complex impedance as the reference phase difference.

[0013] Preferably, the formula for calculating the thermal drift index is: In the formula, Thermal drift index; For natural logarithm operations; The phase difference detected at the current moment; This is the reference phase difference in the cold state; To take the absolute value; The phase tolerance factor is equal to a preset proportion of the reference phase difference; It is a natural constant; The measured magnitude of the dominant frequency impedance at the current moment; This is the reference impedance magnitude in the cold state.

[0014] This invention obtains the thermal drift index by constructing a calculation relationship that includes an exponential term comprising the ratio of the absolute value of the impedance deviation to the reference impedance magnitude and a logarithmic term comprising the ratio of the absolute value of the phase deviation to the phase tolerance factor. By utilizing the amplification characteristics of the exponential function, the numerical sensitivity of the thermal drift index can be significantly improved when the impedance magnitude, which characterizes the decay of permeability, changes significantly. This enables sensitive quantification of the degradation of the microscopic magnetic properties of magnetostrictive materials through electrical parameters in the absence of a temperature sensor.

[0015] Preferably, obtaining the correction factors of each order includes: In the formula, For the first Correction factor for order coefficients; is the order of the inverse model polynomial, with values ​​of 1, 3, and 5; The evolution sensitivity coefficient has a value range of [0.05, 0.2]. Thermal drift index; Thermal drift index Power; It is the hyperbolic tangent function; The average output power at the current moment; This is the power threshold.

[0016] This invention obtains correction factors of various orders by introducing a power gating term containing a hyperbolic tangent function and an order power term of the thermal drift exponent. This makes the correction mechanism power-sensitive and order-selective. That is, it remains silent at low power to avoid introducing noise, while at high power when the thermal effect is significant, the correction strength increases exponentially with the increase of order. This achieves key suppression of the high-order harmonic components most severely affected by thermal drift and optimizes the compensation efficiency.

[0017] Preferably, dynamically updating the coefficient vector of the inverse model using the correction factors of each order to obtain the new coefficients at the current time includes: adjusting the coefficient vector of the initial coefficient vector by the first... The initial coefficients of the order and the corresponding first order Multiplying the correction factors of the first-order coefficients by the first-order coefficients yields the first-order coefficient in the new coefficients at the current time. Order coefficient, is the order of the inverse model polynomial, with values ​​of 1, 3, and 5.

[0018] Preferably, the formula for calculating the predistortion drive voltage signal is: In the formula, For a moment The generated predistortion drive voltage signal; is the order of the inverse model polynomial, with values ​​of 1, 3, and 5; For the updated number Order coefficient; For time in an ideal guided wave voltage sequence Voltage data.

[0019] This invention constructs a predistorted driving voltage signal by weighted summing of the new coefficients of each order at the current moment with the power waveform data of the corresponding order of the ideal guided wave voltage sequence. This superposition synthesis method constructs a signal in the digital domain that has undergone specific distortion in the waveform shape in advance. The pre-set nonlinear component in this signal can physically cancel out the nonlinear characteristics of the material under the current thermal state when it passes through the transducer, thereby ensuring a high degree of restoration of the output waveform without changing the hardware structure.

[0020] Preferably, the real-time acquisition of voltage and current data at the output of the power amplifier includes: synchronously acquiring voltage and current data at the output of the power amplifier through a high-frequency sampling circuit; performing cross-correlation calculations on the voltage and current data to eliminate transmission delay; and performing discrete Fourier transform on the data after delay elimination to obtain frequency domain characteristics.

[0021] In a second aspect, the present invention provides a power drive compensation system for monitoring the characteristics of magnetostrictive loads, comprising a processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the aforementioned power drive compensation method for monitoring the characteristics of magnetostrictive loads.

[0022] By adopting the above technical solution, the power drive compensation method for monitoring magnetostrictive load characteristics is generated into a computer program and stored in a memory for loading and execution by a processor. This allows for the creation of a terminal device based on the memory and processor, facilitating its use.

[0023] The beneficial effects of this invention are as follows:

[0024] This invention establishes a reference impedance magnitude and a reference phase difference as physical reference zero points in a cold system environment, and collects voltage and current data in real time during operation to monitor changes in the dominant frequency impedance and phase, thereby constructing a thermal drift index that reflects the degree of hysteresis loop distortion. This thermal drift index is used to dynamically update the inverse model coefficients with the current output power, generating a pre-distortion drive voltage signal that matches the current thermal nonlinear characteristics in real time. Utilizing the mapping mechanism of electrical parameters to changes in the material's microscopic magnetic properties, real-time sensing and closed-loop compensation of thermal drift phenomena are achieved without the need for external temperature sensors. This ensures that the system can accurately cancel nonlinear interference under varying temperature conditions, effectively suppressing the excitation of higher harmonics and redundant modes, and significantly improving the purity and signal-to-noise ratio of long-distance guided wave detection signals. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the power drive compensation method for monitoring the characteristics of magnetostrictive loads in this invention.

[0026] Figure 2 This is a schematic diagram illustrating the time-domain comparison of the output waveform of a magnetostrictive transducer under thermal conditions;

[0027] Figure 3 This is a schematic diagram illustrating the comparison of the spectral characteristics of the output signal. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0030] This invention discloses a power drive compensation method for monitoring the characteristics of magnetostrictive loads, as described in the embodiments below. Figure 1 This includes steps S1-S4:

[0031] S1. Determine the center frequency of the electromagnetic system, obtain the reference impedance magnitude and reference phase difference under the cold state of the electromagnetic system as the reference zero point for the evolution of magnetic performance, and construct the inverse model to obtain the initial coefficient vector.

[0032] It should be noted that the nonlinear characteristics of magnetostrictive materials vary under different temperatures and stress states. In order to achieve accurate compensation for thermal drift, this invention first needs to obtain the basic physical characteristics of the system under cold conditions at the initial stage of device startup as a zero-point reference for subsequent judgment of the degree of drift. At the same time, in order to ensure the effectiveness of guided wave detection, the optimal operating frequency must be determined in advance.

[0033] Specifically, during the system deployment phase, the wall thickness, diameter, and material sound velocity parameters of the component under test are obtained; the dispersion curve of the guided wave is plotted using the above parameters, and the center frequency corresponding to a single mode region is selected as the center frequency of the electromagnetic system; in subsequent steps, the carrier frequency of all driving signals generated by the signal generator is locked to the center frequency of the electromagnetic system; when performing spectrum analysis on the signal, the frequency point where the center frequency of the electromagnetic system is located is the fundamental frequency.

[0034] Furthermore, in a cold-state environment of the electromagnetic system, a stepped-amplitude sweep signal is injected into the power amplifier via a signal generator, and voltage data at the output of the power amplifier is collected. and current data ; using the collected voltage data and current data An inverse model is constructed, which employs a structure of static nonlinear elements connected in series with dynamic linear elements; the initial coefficient vector of the inverse model is obtained in order to... Construct an observation matrix containing input and output data. .

[0035] Wherein, the observation matrix The formula for calculation is:

[0036]

[0037] In the formula, The observation matrix; to These are the values ​​of each sampling point in the acquired voltage data sequence at the output of the power amplifier; This represents the number of all sampling points in the voltage data sequence. This represents the cube value of the corresponding voltage data sampling point; This represents the fifth power value of the corresponding voltage data sampling point.

[0038] The calculation formula maps the physical characteristic of magnetostrictive materials, which mainly produce odd-order harmonic distortion, by constructing a matrix structure containing fundamental, cubic, and quintic terms. Odd powers are selected as characteristic terms, making the observation matrix... It can accurately capture voltage fluctuation characteristics that are highly correlated with hysteresis nonlinearity, providing a mathematical basis for subsequent solutions to the coefficients that can offset these specific order distortions.

[0039] It should be further noted that the ideal guided wave voltage sequence It is the target reference waveform for the system to perform predistortion control, and its setting directly determines the spectral quality of the final excited ultrasonic guided wave. In order to obtain a single-mode and low-frequency dispersion guided wave signal, an infinitely long sine wave cannot be used directly, and it must be windowed in the time domain.

[0040] Specifically, based on the determined center frequency of the electromagnetic system An ideal guided wave voltage sequence is generated by modulating a sinusoidal wave using the Hanning window. First, set the number of cycles for the excitation signal. Typically set to 5; construct time series Its sampling interval is determined by the sampling rate of the signal generator; based on time series Calculate the Hanning window function Furthermore, the Hanning window function Multiplying with a carrier sine wave yields an ideal guided wave voltage sequence. Time-domain expression: ; For time in an ideal guided wave voltage sequence Voltage data; For the Hanning window function; The center frequency of the electromagnetic system; For the number of cycles; finally, the calculated... Arranged in chronological order, they form an ideal guided wave voltage sequence in vector form. This is used as the objective for subsequent least squares operations.

[0041] Furthermore, construct the observation matrix. Then, combined with the ideal guided wave voltage sequence The initial coefficient vector is calculated using the least squares method. The formula for its calculation is:

[0042]

[0043] In the formula, This is the initial coefficient vector; The observation matrix; This is the transpose of the observation matrix; It is the inverse matrix of the matrix product; It is an ideal guided wave voltage sequence, a standard sine wave.

[0044] This calculation formula utilizes the optimization idea of ​​the least squares method in the observation matrix. and ideal guided wave voltage sequence Find an optimal solution among the given and target signals such that the sum of squared errors between the reconstructed signal and the target signal is minimized; obtain the initial coefficient vector. It includes initial weighting coefficients corresponding to the fundamental, third, and fifth harmonics, which precisely quantify the degree of predistortion required to compensate for the inherent nonlinearity of the system under cold-state reference.

[0045] It should be noted that this invention injects a swept frequency signal into the power amplifier and constructs an observation matrix containing odd power characteristics of voltage data. It then uses the least squares method to establish a mapping relationship between the matrix and the ideal guided wave voltage sequence. This setup uses odd power terms to accurately fit the hysteresis nonlinearity characteristics of magnetostrictive materials that mainly generate odd harmonics. This provides the system with a set of initial coefficient vectors that can accurately describe the nonlinear distortion law under the initial cold state, providing an accurate mathematical benchmark for subsequent dynamic adjustments.

[0046] Furthermore, in the cold state of the electromagnetic system, the control signal generator transmits a standard pulse signal to the power amplifier. The carrier frequency of this standard pulse signal is the previously determined center frequency of the electromagnetic system. Time-domain voltage data at the output of the power amplifier is synchronously acquired through a high-frequency sampling circuit. and time-domain current data and the collected voltage data and current data Perform a fast Fourier transform to extract the complex impedance at the fundamental frequency.

[0047] Among them, the complex impedance at the fundamental frequency The formula for calculation is:

[0048]

[0049] In the formula, The complex impedance at the fundamental frequency; For Fast Fourier Transform; For the collected voltage data; For the collected current data; This represents the real part of the complex impedance, i.e., the resistive component. The imaginary unit; This is the imaginary part of the complex impedance, i.e., the reactance component.

[0050] Furthermore, based on the extracted complex impedance Calculate and store the reference zero point for the evolution of magnetic properties in the cold state, i.e., the reference impedance magnitude. and reference phase difference Among them, the reference impedance magnitude Reference phase difference , The arctangent function is used to obtain the reference impedance magnitude. The sum of the inductance and resistance of the coil in the cold state was characterized, serving as a physical benchmark for subsequent measurement of changes in permeability; the obtained benchmark phase difference The hysteresis loss angle state of the magnetostrictive material in the cold state was characterized, serving as a physical benchmark for subsequent measurement of the degree of hysteresis loop deformation.

[0051] It should be noted that this invention transmits a standard pulse signal with a carrier frequency equal to the center frequency of the electromagnetic system in a cold environment, and uses Fast Fourier Transform to extract the complex impedance at the fundamental frequency to obtain the reference impedance magnitude and reference phase difference. This operation ensures that the measurement is performed in the most stable intrinsic state of the material and is locked at the operating frequency, thereby obtaining a highly reliable physical reference with a high signal-to-noise ratio, providing a necessary comparative basis for subsequent accurate quantification of the degree of thermal drift.

[0052] Furthermore, a phase tolerance factor is set. Its value is the reference phase difference. 5%.

[0053] S2. Real-time acquisition of voltage and current data at the output of the power amplifier, calculation of the current main frequency impedance magnitude and phase difference using the electrical parameter characterization method, and acquisition of the thermal drift index by combining the reference impedance magnitude and reference phase difference.

[0054] It should be noted that, in response to the drift in impedance magnitude caused by changes in coil inductance and the shift in voltage and current phase difference caused by changes in hysteresis loss, this invention quantifies this thermal drift state by real-time monitoring of electrical parameters and constructing specific mathematical relationships.

[0055] Specifically, during the testing process, voltage and current data at the output of the power amplifier are acquired in real time, and cross-correlation calculations are performed on the voltage and current data to eliminate the time difference caused by hardware transmission delay; then, discrete Fourier transform is performed on the aligned voltage and current data to calculate the current dominant frequency impedance magnitude. and phase difference .

[0056] It should be noted that the present invention synchronously collects voltage and current data through a high-frequency sampling circuit and performs cross-correlation calculations to eliminate transmission delay. Subsequently, discrete Fourier transform is performed to obtain frequency domain characteristics. This processing flow effectively eliminates time difference interference caused by hardware line transmission delay or sampling asynchrony, ensuring that the calculated phase difference only reflects the hysteresis loss characteristics of the magnetostrictive material itself, thereby guaranteeing the authenticity and accuracy of thermal drift monitoring data.

[0057] Furthermore, based on the current dominant frequency impedance magnitude... and phase difference Calculate the thermal drift index The formula for its calculation is:

[0058]

[0059] In the formula, The thermal drift index quantifies the impedance magnitude drift caused by permeability decay, and quantifies the voltage and current phase shift caused by changes in hysteresis loss. For natural logarithm operations; The phase difference detected at the current moment; This is the reference phase difference in the cold state; This represents the absolute value of the phase deviation. This is the phase tolerance factor; It is a natural constant; The measured magnitude of the dominant frequency impedance at the current moment; This is the reference impedance magnitude in the cold state; This represents the absolute value of the impedance deviation.

[0060] In this calculation formula This term reflects the relative rate of change of the impedance magnitude, and is placed in an exponential function. The exponential position of this term means that when the impedance changes significantly, its value increases exponentially. The physical meaning of this change is that a drastic change in the impedance magnitude usually indicates a significant decrease in the permeability of the magnetostrictive material, at which point the phase deviation... The contribution weight to waveform distortion should be amplified; The function is used to map the exponentially amplified value to a range with better linearity, preventing overflow in subsequent calculations; when the thermal drift exponent... The larger the value, the more severe the current thermal drift phenomenon of the system and the higher the degree of magnetic hysteresis nonlinearity.

[0061] S3. Based on the thermal drift index and the average output power at the current moment, obtain the correction factors of each order, and use the correction factors of each order to dynamically update the coefficient vector of the inverse model to obtain the new coefficients at the current moment.

[0062] It should be noted that the effect of thermal drift on the magnetostrictive waveform is not uniform across the entire frequency band, but mainly leads to a sharp increase in the energy of higher harmonic components. This means that the higher-order coefficients used to control higher harmonics in the inverse model need to be significantly modified, while the lower-order coefficients used to control the fundamental wave only need to be fine-tuned. In addition, this modification should only be introduced when the power is large enough and the thermal effect is significant, so as to avoid introducing unnecessary interference in low-power noise environments. Therefore, this invention designs a dynamic evolution mechanism that is sensitive to order and power.

[0063] Specifically, calculate the average output power at the current moment. and set a power threshold. In this embodiment, the power threshold is... Set to 80% of the rated power of the power amplifier; calculate the corresponding correction factor for each order coefficient in the inverse model. The formula for its calculation is:

[0064]

[0065] In the formula, For the first Correction factor for order coefficients; is the order of the inverse model polynomial, with values ​​of 1, 3, and 5; The evolution sensitivity coefficient is taken as an empirical value of 0.1 in this embodiment; Thermal drift index; Thermal drift index Power; It is the hyperbolic tangent function; The average output power at the current moment; This is the power threshold.

[0066] Wherein, the evolution sensitivity coefficient The evolution sensitivity coefficient is a key gain term controlling the update step size of the predistortion parameters, and its value directly affects the stability and response speed of the system. If the evolution sensitivity coefficient is set too small, the system's ability to track thermal drift is insufficient, causing the change in the correction factor to lag behind the actual physical changes, and failing to suppress rapidly rising high-order harmonics in time. If the evolution sensitivity coefficient is set too large, it will lead to excessively strong feedback gain, causing the correction factor to oscillate violently or even diverge during the iteration process, thus introducing additional waveform distortion. Therefore, the evolution sensitivity coefficient... The value range is [0.05, 0.2]; in this embodiment, the evolution sensitivity coefficient is... Setting it to 0.1 achieves the best balance between response speed and control stability, ensuring that the system effectively counteracts the nonlinear effects of thermal drift while ensuring a smooth transition. In other embodiments, implementers can set the evolution sensitivity coefficient in the range of [0.05, 0.2] according to the actual situation.

[0067] Wherein, in the calculation formula The item reflects order sensitivity, when As it increases, i.e. for higher-order terms, The value will change with the thermal drift index. The value increases rapidly with the increase of the value, which means that when thermal drift occurs, the algorithm will prioritize and significantly adjust higher-order coefficients to suppress higher-order harmonics caused by increased nonlinearity; the introduction of a power gating term of the hyperbolic tangent function makes the correction mechanism sensitive to electrical load: when the actual average output power Much smaller than the power threshold hour, When the function value approaches 0, the correction factor... Approaching 1, that is, keeping the initial coefficients unchanged, when the average output power Approaching or exceeding the power threshold At that time, the high-power thermal effect is significant. When the function value approaches 1, the focus is on suppressing the higher harmonic components caused by magnetic saturation.

[0068] Furthermore, using correction factors For the initial coefficient vector Update the elements of each order in the matrix to obtain the new coefficients at the current time. The formula for its calculation is:

[0069]

[0070] In the formula, For the updated number Order coefficient; Initial coefficient vector The corresponding number in Initial coefficients; For the first Correction factor for order coefficients.

[0071] It should be noted that this invention obtains the new coefficients at the current moment by multiplying the initial coefficients of each order in the initial coefficient vector with the corresponding correction factor. This avoids the need to retrain the complex inverse model or perform large-scale matrix operations during online operation, greatly reducing the computational complexity of the algorithm and ensuring the real-time response capability of the control system, enabling it to quickly track the transient thermal changes of the magnetostrictive transducer.

[0072] S4. Using the new coefficients and ideal guided wave voltage sequence at the current moment, construct a predistortion drive voltage signal that can cancel nonlinear interference, and input the predistortion drive voltage signal to the power amplifier to drive the magnetostrictive transducer.

[0073] It should be noted that, after the calculations in the above steps, the system has obtained control parameters that can offset the current thermal nonlinear characteristics; in order to achieve the final waveform correction, these parameters need to be applied to the ideal waveform to generate a drive signal containing inverse distortion, which is then amplified by the power amplifier.

[0074] Specifically, using the updated coefficients and ideal guided wave voltage sequence Voltage data in the time frame, constructing time Predistortion drive voltage signal The formula for its calculation is:

[0075]

[0076] In the formula, For a moment The generated predistortion drive voltage signal; is the order of the inverse model polynomial, with values ​​of 1, 3, and 5; For the updated number Order coefficient; For time in an ideal guided wave voltage sequence Voltage data; For time in an ideal guided wave voltage sequence voltage data Power of 1.

[0077] The calculation formula constructs a signal whose waveform shape is pre-distorted by superimposing weighted power functions of different orders. The signal contains a nonlinear component that is opposite to the current thermal hysteresis characteristics. The amplitude and phase of these components have been precisely calculated to counteract the distortion generated during the subsequent physical transformation.

[0078] Furthermore, the calculated predistortion drive voltage signal The signal is converted into an analog signal and input to a power amplifier; the power amplifier amplifies the analog signal and applies it to the magnetostrictive transducer; due to the predistortion of the drive voltage signal... A nonlinear component opposite to the current thermal hysteresis characteristics is pre-set in the transducer. When this component passes through the magnetomechanical conversion process of the magnetostrictive transducer, it will cancel out the nonlinear characteristics of the material itself, thereby exciting a pure ultrasonic guided wave in the tested component, and realizing the accurate restoration of the electrical characteristics of the magnetostrictive load under variable temperature conditions.

[0079] For example, Figure 2 This is a time-domain comparison diagram of the output waveform of a magnetostrictive transducer under hot conditions. The ideal target waveform preset by the system is used as the comparison benchmark. For the curve corresponding to the output signal of the existing algorithm without compensation measures, significant amplitude compression and clipping distortion occur at the peak, and the waveform width widens, indicating severe nonlinear distortion. For the curve corresponding to the output signal after applying the thermal drift compensation method of this invention, the system detects an increase in the thermal drift exponent, automatically updates the coefficient vector, and generates a driving signal with inverse distortion. When this signal passes through the hot transducer, the pre-distortion of the driving signal cancels out the thermal distortion of the transducer. Therefore, the curve corresponding to the output signal after applying the thermal drift compensation method of this invention is full and symmetrical about the time axis, highly coinciding with the ideal sine wave. Figure 3 The diagram shows a comparison of the spectral characteristics of the output signal. For the existing algorithm without compensation measures, the curve corresponding to the signal spectrum has a high energy peak at the odd harmonic frequency, indicating that thermal drift causes serious high-order harmonic interference. However, for the curve corresponding to the signal spectrum after applying the thermal drift compensation method of this invention, the curve has a significant peak at the fundamental frequency, while the amplitude at the third and fifth harmonic frequencies is greatly suppressed, resulting in a clean spectral background.

[0080] This invention also discloses a power drive compensation system for monitoring the characteristics of magnetostrictive loads, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the power drive compensation method for monitoring the characteristics of magnetostrictive loads according to the present invention.

[0081] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A power drive compensation method for monitoring the characteristics of magnetostrictive loads, characterized in that, include: The center frequency of the electromagnetic system is determined, and the reference impedance magnitude and reference phase difference are obtained under the cold state of the electromagnetic system as reference zero points for the evolution of magnetic performance. An inverse model is then constructed to obtain the initial coefficient vector. The system acquires real-time voltage and current data at the output of the power amplifier. Based on this data, it calculates the current main frequency impedance magnitude and phase difference. Combined with the reference impedance magnitude and phase difference, it obtains the thermal drift index, which reflects the degree of hysteresis loop distortion. , ; For natural logarithm operations; The phase difference detected at the current moment; This is the reference phase difference in the cold state; To take the absolute value; This is the phase tolerance factor, which is equal to a preset proportion of the reference phase difference; It is a natural constant; The measured magnitude of the dominant frequency impedance at the current moment; This is the reference impedance magnitude in the cold state; Correction factors of each order are obtained based on the thermal drift index and the average output power at the current moment. , ; is the order of the inverse model polynomial, with values ​​of 1, 3, and 5; The evolution sensitivity coefficient has a value range of [0.05, 0.2]. Thermal drift index Power of; It is the hyperbolic tangent function; The average output power at the current moment; Using the power threshold, the coefficient vector of the inverse model is dynamically updated using correction factors of various orders to obtain the new coefficients at the current time. By utilizing the new coefficients and ideal guided wave voltage sequence at the current moment, a predistortion driving voltage signal that can cancel nonlinear interference is constructed. The predistortion driving voltage signal is then input to a power amplifier to drive a magnetostrictive transducer, thereby achieving accurate reproduction of the electrical characteristics of the magnetostrictive load under variable temperature conditions.

2. The power drive compensation method for monitoring the characteristics of magnetostrictive loads according to claim 1, characterized in that, The ideal guided wave voltage sequence is obtained as follows: Set the number of cycles of the excitation signal, and construct a Hanning window function based on the center frequency of the electromagnetic system; The Hanning window function is multiplied by the carrier sine wave to obtain the time-domain expression of the ideal guided wave voltage sequence, and then the voltage data at each time point is calculated. The calculated voltage data are arranged in chronological order to form the ideal guided wave voltage sequence.

3. The power drive compensation method for monitoring the characteristics of magnetostrictive loads according to claim 2, characterized in that, The construction of the inverse model to obtain the initial coefficient vector includes: A sweep frequency signal is injected into the power amplifier through a signal generator, and the voltage data sequence at the output of the power amplifier is collected. Construct an observation matrix containing the odd-power features of the voltage data sequence; By combining the ideal guided wave voltage sequence, the mapping relationship between the observation matrix and the ideal guided wave voltage sequence is calculated using the least squares method to obtain the initial coefficient vector.

4. The power drive compensation method for monitoring the characteristics of magnetostrictive loads according to claim 1, characterized in that, The acquisition of the reference impedance magnitude and reference phase difference in a cold-state environment of an electromagnetic system includes: In a cold environment of the electromagnetic system, the control signal generator transmits a standard pulse signal with a carrier frequency equal to the center frequency of the electromagnetic system. The time-domain voltage and time-domain current data at the output of the power amplifier are collected and subjected to fast Fourier transform to extract the complex impedance at the center frequency of the electromagnetic system. The magnitude of the complex impedance is calculated as the reference impedance magnitude, and the arctangent of the real and imaginary parts of the complex impedance is calculated as the reference phase difference.

5. The power drive compensation method for monitoring the characteristics of magnetostrictive loads according to claim 1, characterized in that, The coefficient vector of the inverse model is dynamically updated using the aforementioned correction factors of each order to obtain the new coefficients at the current time, including: adjusting the coefficient vector of the initial coefficient vector by the first... The initial coefficients of the order and the corresponding first order Multiplying the correction factors of the first-order coefficients by the first-order coefficients yields the first-order coefficient in the new coefficients at the current time. Order coefficient, is the order of the inverse model polynomial, with values ​​of 1, 3, and 5.

6. The power drive compensation method for monitoring the characteristics of magnetostrictive loads according to claim 1, characterized in that, The formula for calculating the predistortion drive voltage signal is: ； In the formula, For a moment The generated predistortion drive voltage signal; is the order of the inverse model polynomial, with values ​​of 1, 3, and 5; For the updated number Order coefficient; For time in an ideal guided wave voltage sequence Voltage data.

7. The power drive compensation method for monitoring the characteristics of magnetostrictive loads according to claim 1, characterized in that, The real-time acquisition of voltage and current data at the output of the power amplifier includes: Voltage and current data at the output of the power amplifier are synchronously acquired by a high-frequency sampling circuit; cross-correlation is performed on the voltage and current data to eliminate transmission delay; and discrete Fourier transform is performed on the data after delay elimination to obtain frequency domain characteristics.

8. A power drive compensation system for monitoring the characteristics of magnetostrictive loads, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the power drive compensation method for monitoring magnetostrictive load characteristics according to any one of claims 1-7.

Images