A nonlinear ultrasonic long pulse very narrow band harmonic detection method and device thereof

Abstract

The application relates to a nonlinear ultrasonic long-pulse extremely narrow-band harmonic detection method and device, computer equipment and a storage medium. The method comprises the following steps: transmitting a millisecond-level long-pulse ultrasonic wave to excite a fundamental wave ultrasonic wave in a measured object; during fundamental wave detection, an ultrasonic echo signal is acquired, pre-filtered through a fundamental wave band-pass filter and a low-pass filter, and a fundamental wave pre-filtered signal is obtained; during harmonic detection, pre-filtering is performed through a harmonic band-pass filter and a fundamental wave notch filter, and a harmonic pre-filtered signal is obtained; the fundamental wave pre-filtered signal and the harmonic pre-filtered signal are sequentially subjected to low-noise amplification, frequency up-conversion, band-pass filtering, frequency down-conversion, band-pass filtering through an extremely narrow-band crystal filter and quadrature phase detection, and the amplitudes of the fundamental wave signal and the target harmonic signal are extracted; the relative nonlinear coefficient is calculated according to the amplitudes of the fundamental wave signal and the target harmonic signal, and the material damage state of the measured object is evaluated. The application can reduce the excitation amplitude and suppress system nonlinearity and thermal noise.

Description

Technical Field

[0001] This application relates to the field of nonlinear ultrasonic nondestructive testing technology, and in particular to a nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method, device, computer equipment, and storage medium. Background Technology

[0002] With the increasing demands for the safety of critical components in aerospace, nuclear industry, and other fields, nonlinear ultrasonic nondestructive testing technology has attracted widespread attention due to its sensitivity to early micro-damage in materials. Existing nonlinear ultrasonic testing systems generally employ short-pulse, high-amplitude excitation schemes. For example, the RAM5000Snap system from RETIC Corporation in the United States generates hundreds of volts of high-voltage pulses through a Class A power amplifier to enhance the nonlinear response of defects, and the receiving end uses a superheterodyne phase-sensitive receiver structure to extract harmonic signals.

[0003] However, this scheme has inherent flaws: on the one hand, Class A power amplifiers have extremely low efficiency (usually less than 50%), generate significant heat, and their maximum transmit pulse width is only about 200 microseconds, which prevents the receiver filter bandwidth from being narrowed to below 50kHz, resulting in a high thermal noise level (measured noise floor of about -124dBm), which limits the extraction of weak harmonic signals; on the other hand, high amplitude excitation significantly increases the nonlinear distortion of the power amplifier and transducer themselves. These system nonlinearities are superimposed on the real nonlinear signals generated by defects, which seriously interferes with the accuracy of the detection results.

[0004] To overcome the above problems, researchers have tried improved schemes such as the mixing method and the zero-frequency method, or used digital signal processing technology for post-processing. However, the mixing method is complex and requires high phase accuracy, the zero-frequency method has lower signal amplitude and is sensitive to environmental vibration, and digital signal processing technology has limited adaptability to noise types. All of these methods are difficult to fundamentally solve the problem of accurate extraction of weak defect signals. Summary of the Invention

[0005] Therefore, it is necessary to provide a nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method, device, computer equipment, and storage medium that can reduce the nonlinear interference and thermal noise of the system itself and improve the detection sensitivity of weak harmonic signals, in order to address the above-mentioned technical problems.

[0006] In a first aspect, this application provides a nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method. The method includes: Transmission steps: Transmit millisecond-level long-pulse ultrasonic waves to excite fundamental ultrasonic waves in the object under test; Fundamental wave detection step: When performing fundamental wave detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a fundamental wave bandpass filter and a low-pass filter to obtain the fundamental wave pre-filtered signal; Harmonic detection steps: When performing harmonic detection, the transmission step is executed to obtain the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a harmonic bandpass filter and a fundamental notch filter to obtain the harmonic pre-filtered signal. The calculation and evaluation steps are as follows: The fundamental pre-selected filtered signal is sequentially subjected to low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and quadrature phase-locked loop (QLL) detection to extract the amplitude of the fundamental signal; the harmonic pre-selected filtered signal is sequentially subjected to low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and QLL detection to extract the amplitude of the target harmonic signal; the relative nonlinear coefficient is calculated based on the amplitudes of the fundamental signal and the target harmonic signal to quantitatively assess the material damage state of the tested object; The ultra-narrowband crystal filter has a bandwidth of 100Hz~1000Hz and satisfies... ,in The pulse width of the millisecond-level long pulse excitation signal is given. The bandwidth of the ultra-narrowband crystal filter is given.

[0007] In one embodiment, the launching step includes: A high-efficiency power amplifier based on gallium nitride devices is used to drive the high-efficiency power amplifier with a rectangular wave of set duty cycle to generate the millisecond-level long pulse ultrasonic wave, which is then applied to the transmitting transducer after frequency selection filtering.

[0008] In one embodiment, the setting of the duty cycle includes: When the target harmonic is the second harmonic, the duty cycle of the rectangular wave is set to 50%. When the target harmonic is the third harmonic, the duty cycle of the rectangular wave is set to... .

[0009] In one embodiment, the launching step further includes: The actual duty cycle of the high-voltage pulse wave output by the high-efficiency power amplifier is compared with the set duty cycle. Based on the comparison result, the duty cycle of the rectangular wave is adjusted to lock the actual duty cycle at the set duty cycle.

[0010] In one embodiment, the bandpass filtering after upconversion is implemented using a crystal filter; after low-noise amplification and before upconversion, filtering is further performed using an LC low-pass filter.

[0011] In one embodiment, the method further includes: Before performing the fundamental frequency detection or the harmonic detection, the ultra-narrowband crystal filter is subjected to frequency sweep detection to determine the actual passband center frequency of the ultra-narrowband crystal filter; the local oscillator frequency of the down-conversion is adjusted so that the signal center frequency before bandpass filtering of the ultra-narrowband crystal filter is equal to the actual passband center frequency.

[0012] In one embodiment, the high-efficiency power amplifier is a Class D power amplifier or a Class E power amplifier, and the noise figure of the amplifier used for the low-noise amplification is less than 2.

[0013] Secondly, this application also provides a nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection device. The device includes: Transmitting module: Transmits millisecond-level long-pulse ultrasonic waves to excite fundamental ultrasonic waves in the object under test; Fundamental wave detection module: When performing fundamental wave detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a fundamental wave bandpass filter and a low-pass filter to obtain a fundamental wave pre-filtered signal; Harmonic detection module: When performing harmonic detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a harmonic bandpass filter and a fundamental notch filter to obtain a harmonic pre-filtered signal; The calculation and evaluation module performs the following steps sequentially on the fundamental pre-selected filtered signal: low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and quadrature phase-locked loop detection, to extract the amplitude of the fundamental signal. Similarly, it performs the same steps on the harmonic pre-selected filtered signal, extracting the amplitude of the target harmonic signal. Based on the amplitudes of the fundamental signal and the target harmonic signal, it calculates a relative nonlinear coefficient to quantitatively assess the material damage state of the tested object. The ultra-narrowband crystal filter has a bandwidth of 100Hz~1000Hz and satisfies... ,in The pulse width of the millisecond-level long pulse excitation signal is given. The bandwidth of the ultra-narrowband crystal filter is given.

[0014] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the above-described nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method.

[0015] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the above-described nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method.

[0016] The aforementioned nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method, device, computer equipment, and storage medium utilize millisecond-level long-pulse ultrasonic excitation and establish a quantitative relationship between the pulse width and the bandwidth of the ultra-narrowband crystal filter at the receiving end. This allows the receiving end to employ an ultra-narrowband crystal filter, reducing the thermal noise floor and improving the signal-to-noise ratio. By setting different pre-selected filtering configurations for fundamental and harmonic detection, strong fundamental or high-order harmonic components are effectively suppressed, avoiding saturation distortion in subsequent circuits. A superheterodyne architecture consisting of up-conversion, bandpass filtering, and down-conversion shifts the target harmonic signal to a frequency suitable for processing by the ultra-narrowband crystal filter. Combined with orthogonal phase-locked loop detection, accurate extraction of weak nonlinear harmonic signals under small-to-medium amplitude excitation is achieved. This reduces its own nonlinearity while improving the detection sensitivity and repeatability of early micro-damage to the tested object and lowering the requirements for transducer withstand voltage performance. Attached Figure Description

[0017] Figure 1 This is a diagram illustrating the application environment of a nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method in one embodiment. Figure 2 This is a flowchart illustrating a nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method in one embodiment; Figure 3 This is a block diagram of the overall structure of the detection system in one embodiment; Figure 4 This is a schematic diagram of a high-efficiency, high-linearity pulse generation circuit based on gallium nitride in one embodiment. Figure 5 This is a circuit block diagram of a nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method in one embodiment; Figure 6 This is a structural block diagram of a nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection device in one embodiment; Figure 7 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. First, to facilitate understanding of the technical solutions provided by the embodiments of this application, the background technology involved in the embodiments of this application will be described below.

[0019] In critical engineering fields such as aerospace, nuclear industry, and pressure vessels, metal components develop early micro-damage such as fatigue and creep during long-term service. If not detected and assessed in time, this can lead to catastrophic accidents. Nonlinear ultrasonic nondestructive testing technology has received widespread attention in recent years due to its sensitivity to changes in the microstructure of materials and its ability to detect early micro-defects that traditional linear ultrasound cannot identify. The basic principle of this technology is that when ultrasonic waves propagate through materials containing micro-damage, they interact with the damage to generate higher harmonics (such as second and third harmonics). By measuring the amplitude of these harmonic signals and calculating the nonlinear coefficients, the damage state of the material can be quantitatively assessed.

[0020] However, the nonlinear harmonic signals generated by microscopic defects are extremely weak, typically 40 to 60 dB lower than the fundamental signal, and are easily drowned out by the system's background noise. To enhance this nonlinear effect, existing nonlinear ultrasonic testing systems generally employ short-pulse, high-amplitude excitation schemes. The RAM5000Snap system, mass-produced by RETIC in the United States, is a prime example. This system uses a Class A power amplifier to generate high-voltage pulses of hundreds or even thousands of volts, which are applied to the transmitting transducer to increase the harmonic amplitude at the receiving end. The receiving section uses a superheterodyne phase-sensitive receiver structure to extract the weak signal. Besides commercial systems, researchers also commonly use a scheme combining pre-amplification, analog-to-digital conversion, and digital lock-in amplification to detect nonlinear ultrasonic signals.

[0021] However, the aforementioned existing technologies have two inherent drawbacks. First, while high-amplitude excitation enhances the nonlinear response of the defect, it also significantly increases the nonlinear distortion of the power amplifier and transducer themselves. Taking a Class A power amplifier as an example, its distortion increases sharply with the increase of input voltage, and the harmonics it generates at the output of several hundred volts can reach several percentage points of the fundamental frequency. At the same time, the wafer size of the high-frequency transducer is small and cannot withstand high-voltage excitation; otherwise, not only will the distortion be severe, but the transducer may also be directly damaged. These system nonlinearities superimposed on the real nonlinear signals generated by the defect seriously interfere with the accuracy of the detection results. Second, the inherent power conversion efficiency of Class A power amplifiers is extremely low (usually less than 50%), and the device heats up severely, making it unable to stably transmit long pulse signals. Taking the RAM5000Snap system as an example, its maximum transmit pulse width is only 200 microseconds. Under this short pulse constraint, the receiver filter must have sufficient bandwidth to meet the settling time requirements. The minimum bandwidth of the actually configured filter is 50 kHz, corresponding to a high thermal noise level (measured noise floor of about -124 dBm), which further limits the ability to extract weak harmonic signals. Furthermore, although researchers have tried improved schemes such as the mixing method and the zero-frequency method, as well as digital signal processing techniques to suppress noise and system interference, the mixing method is complex and requires high phase accuracy, the zero-frequency method has lower signal amplitude and is sensitive to environmental vibration, and digital signal processing techniques have limited adaptability to noise types. None of these methods can fundamentally solve the problem of accurately extracting weak nonlinear defect signals.

[0022] Therefore, the core technical challenge of existing nonlinear ultrasonic testing technologies is how to accurately extract weak nonlinear harmonic signals from a high-noise environment while avoiding nonlinear interference from the system itself, thereby achieving highly sensitive detection of early-stage micro-damage in materials. To address this, this application provides a nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method, apparatus, computer equipment, and storage medium. By transmitting long-pulse signals and employing ultra-narrowband receiving filtering, the signal-to-noise ratio at the receiving end is improved while reducing the excitation amplitude.

[0023] The nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, the ultrasonic testing device 102 communicates with the data processing server 104 via a network. A data storage system can store the data that the data processing server 104 needs to process. The data storage system can be integrated into the data processing server 104 or placed in the cloud or on another network server. The ultrasonic testing device 102 emits millisecond-level long-pulse ultrasonic waves, receives the echo signals, and sequentially performs pre-selection filtering, low-noise amplification, up-conversion, bandpass filtering, down-conversion, ultra-narrowband crystal filter filtering, and quadrature phase-locked loop detection to extract the amplitudes of the fundamental wave signal and the target harmonic signal. The data processing server 104 calculates the relative nonlinear coefficient based on the amplitudes of the fundamental wave signal and the target harmonic signal to quantitatively assess the material damage state of the tested object. The ultrasonic testing device and the data processing server transmit the extracted amplitude data via a wired or wireless network. The ultrasonic testing device 102 can be, but is not limited to, a dedicated nonlinear ultrasonic testing instrument, a multi-channel ultrasonic testing system, an embedded ultrasonic flaw detector, a portable ultrasonic testing terminal, or an online testing module integrated into an industrial robot. The data processing server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers, and is used to store historical detection data, train damage assessment models, or perform remote diagnostic analysis.

[0024] Firstly, existing nonlinear ultrasonic testing techniques, due to their use of short-pulse, high-amplitude excitation, suffer from significant inherent nonlinear interference and high thermal noise levels, making it difficult to accurately extract weak harmonic signals. Therefore, in one embodiment, such as... Figure 2 As shown, a nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method is provided, which is applied to... Figure 1 The ultrasonic testing equipment 102 in the example is used to illustrate the process, including the emission step, the fundamental wave detection step, the harmonic detection step, and the calculation and evaluation step. Details are as follows.

[0025] Transmission steps: Transmit millisecond-level long-pulse ultrasonic waves to excite fundamental ultrasonic waves in the object under test.

[0026] Specifically, in the transmission step, an excitation source generates a long-pulse ultrasonic signal with a duration in the millisecond range (e.g., 1 to 100 milliseconds). This long-pulse signal is converted into sound waves by a transmitting transducer and propagates inside the object under test (such as a metallic material component), exciting ultrasonic waves at the fundamental frequency. The fundamental frequency can be set according to the detection requirements, such as 1 MHz, 5 MHz, or 10 MHz. The transmission step is performed before each time an ultrasonic echo signal needs to be acquired to ensure that there is always sufficient sound field energy in the object under test.

[0027] Fundamental wave detection steps: When performing fundamental wave detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by the fundamental wave bandpass filter and the low-pass filter to obtain the fundamental wave pre-filtered signal.

[0028] Specifically, in the fundamental wave detection step, the aforementioned transmission step is first performed. After fundamental wave ultrasound is excited in the object under test, the receiving transducer acquires the ultrasonic echo signal that returns after propagating through the object under test. This echo signal contains the fundamental wave frequency component and any harmonic components that may be generated. To extract the fundamental wave component separately, the acquired ultrasonic echo signal is sequentially sent to a fundamental wave bandpass filter and a low-pass filter for pre-selection filtering. The passband center frequency of the fundamental wave bandpass filter is set to the fundamental wave frequency (e.g., the fundamental wave frequency). The bandwidth is wide enough to pass the fundamental signal but suppress higher harmonics; a low-pass filter further filters out high-frequency noise. After processing by these two filters, the output signal mainly retains the fundamental component, which is the fundamental pre-selected filter signal.

[0029] Harmonic detection steps: When performing harmonic detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a harmonic bandpass filter and a fundamental notch filter to obtain the harmonic pre-filtered signal.

[0030] Specifically, the harmonic detection step is similar to the fundamental frequency detection step, but the pre-selection filter configuration is different. The transmission step is performed first to acquire the ultrasonic echo signal. To extract the target harmonic (e.g., second or third harmonic), the echo signal is fed into a harmonic bandpass filter and a fundamental frequency notch filter. The passband center frequency of the harmonic bandpass filter is set to the target harmonic frequency (e.g., two or three times the fundamental frequency) to allow this harmonic component to pass through; the fundamental frequency notch filter is used to strongly attenuate the fundamental frequency component, preventing strong fundamental signals from entering subsequent circuits and causing saturation. After processing by these two filters, the output signal mainly retains the target harmonic component, thus obtaining the harmonic pre-selection filtered signal.

[0031] The calculation and evaluation steps are as follows: The fundamental pre-selected filtered signal is sequentially amplified with low noise, up-converted, band-pass filtered, down-converted, band-pass filtered by an ultra-narrowband crystal filter, and quadrature phase-locked loop (QLL) detection to extract the amplitude of the fundamental signal; the harmonic pre-selected filtered signal is sequentially amplified with low noise, up-converted, band-pass filtered, down-converted, band-pass filtered by an ultra-narrowband crystal filter, and QLL detection to extract the amplitude of the target harmonic signal; the relative nonlinear coefficient is calculated based on the amplitude of the fundamental signal and the amplitude of the target harmonic signal to quantitatively assess the material damage state of the tested object.

[0032] like Figure 3As shown in the figure, an overall structural block diagram of a detection system is presented. The transmitting channel includes: a gallium nitride-based high-efficiency power amplifier, a frequency-selective filter network, and a transmitting transducer; the receiving channel includes: a pre-selection filter (fundamental bandpass filter + low-pass filter, or harmonic bandpass filter + fundamental notch filter), a low-noise amplifier, a first mixer, a first intermediate frequency crystal filter, a second mixer, a second intermediate frequency ultra-narrowband crystal filter, a programmable amplifier, and a quadrature phase-locked detector. The figure also shows a digital acquisition and processing unit (FPGA), which controls the transmit pulse width, duty cycle, frequency self-test, and local oscillator frequency adjustment. Figure 3 The “receiving transducer” and “transmitting transducer” can be two separate probes, or they can share the same probe (switched through a transmit / receive isolation circuit). Figure 3 The signal flow not marked in the text is as follows: the echo signal passes through pre-selection filtering, low noise amplification, first mixing (up-conversion), first intermediate frequency filtering, second mixing (down-conversion), second intermediate frequency ultra-narrowband filtering, and quadrature phase-locked detection in sequence, and finally outputs the fundamental amplitude and harmonic amplitude to the calculation unit.

[0033] Specifically, in the calculation and evaluation steps, the same signal processing chain is performed on the fundamental pre-selected filtered signal and the harmonic pre-selected filtered signal obtained above. Taking the fundamental pre-selected filtered signal as an example: First, low-noise amplification is performed to amplify the weak signal to a level that can be processed by subsequent circuits. The noise figure of the low-noise amplifier should be as small as possible (e.g., less than 2). Then, up-conversion processing is performed, mixing the fundamental frequency signal with the local oscillator signal and shifting it to a higher first intermediate frequency (e.g., 45MHz). After up-conversion, a bandpass filter (e.g., a crystal filter with a bandwidth of 30kHz) is used to filter out the image frequency and out-of-band noise generated by the mixing. Next, down-conversion processing is performed, mixing the first intermediate frequency signal with the second local oscillator signal and shifting it to a lower second intermediate frequency (e.g., 455kHz). This second intermediate frequency signal is then bandpass filtered by an ultra-narrowband crystal filter with a bandwidth of 100Hz to 1000Hz (e.g., 300Hz), which can suppress thermal noise. The filtered signal is then sent to a quadrature phase-locked detector to extract the signal amplitude. Perform the same processing on the harmonic pre-selected filter signal to obtain the amplitude of the target harmonic signal.

[0034] In quadrature phase-locked loop (PLL) detection, a digital PLL amplifier can be used. Its working principle is as follows: the input signal is multiplied by a quadrature reference signal of the same frequency, and the in-phase component is obtained after low-pass filtering. and orthogonal components The amplitude of the signal is .

[0035] Finally, based on the amplitude of the fundamental wave signal Amplitude of the target harmonic signal (For second harmonics) or (For the third harmonic), calculate the relative nonlinearity coefficient. For example, for the second harmonic, the relative nonlinearity coefficient... For the third harmonic, By comparing the difference between this coefficient and the undamaged baseline value, the material damage state of the tested object can be quantitatively assessed (the more severe the damage, the larger the nonlinear coefficient).

[0036] Furthermore, in this embodiment, the bandwidth of the ultra-narrowband crystal filter is 100Hz~1000Hz, and satisfies... ,in The pulse width of the millisecond-level long pulse excitation signal. This represents the bandwidth of the ultra-narrowband crystal filter. This relationship ensures that the pulse width is long enough to meet the settling time requirements of the ultra-narrowband filter (typically, the settling time is approximately...). (multiples of), thus ensuring that the filter can be stably established and achieve the expected noise suppression effect.

[0037] This embodiment integrates the transmission step into the fundamental wave detection step and the harmonic detection step, ensuring that a long pulse transmission is performed independently before each detection, thus avoiding the logical contradiction of multiple receptions from a single transmission in traditional schemes. The long pulse (millisecond level) is combined with an ultra-narrowband crystal filter (100Hz~1000Hz) to utilize... The quantitative relationship allows the filter sufficient time to establish a stable output, thereby reducing the thermal noise floor of the receiving channel by more than 20 dB and improving the signal-to-noise ratio. Simultaneously, fundamental wave detection and harmonic detection employ different pre-selected filter configurations, effectively suppressing strong fundamental or harmonic interference and preventing saturation of subsequent circuits. The superheterodyne structure, consisting of up-conversion, bandpass filtering, and down-conversion, shifts the target signal to a fixed intermediate frequency suitable for extremely narrow-band filtering, enabling accurate extraction of weak harmonic amplitudes at the nanovolt level even with low excitation amplitudes. Calculating the relative nonlinear coefficient enables quantitative assessment of early-stage micro-damage in materials. This method improves detection sensitivity and repeatability while reducing its own nonlinearity.

[0038] The specific implementation of transmitting millisecond-level long-pulse ultrasonic waves in the transmission step directly affects the quality of the excitation signal and the power consumption of the system. Existing technologies using Class A power amplifiers are inefficient, generate significant heat, and struggle to stably generate long pulses. Therefore, in one embodiment, the transmission step includes: A high-efficiency power amplifier based on gallium nitride devices is used to drive the high-efficiency power amplifier with a rectangular wave of set duty cycle to generate millisecond-level long pulse ultrasonic waves, which are then applied to the transmitting transducer after frequency selection filtering.

[0039] like Figure 4 As shown in the figure, a high-efficiency, high-linearity pulse generation circuit based on gallium nitride is illustrated. Figure 4 The system includes: a DDS, a high-speed comparator, an analog switch, a gate drive circuit, a GaN MOSFET, a low-pass filter and a frequency-selective filter network, a transmitter transducer, a control unit, a pulse count control circuit, and a programmable voltage power supply. The control unit includes an FPGA and a duty cycle measurement feedback branch for implementing closed-loop duty cycle control; the pulse count control circuit is used to set the transmit pulse width. Figure 4 The pulse width control signal is calculated by the FPGA main control unit based on the ultra-narrow band filter bandwidth BWBW. The length of the pulse train is controlled by an analog switch enable terminal. Figure 4 Connections not explicitly marked: DC voltage output from the low-speed digital-to-analog converter The sine wave output from LC filter 1 is simultaneously fed into a high-speed comparator to generate a rectangular wave with a set duty cycle. This rectangular wave is then intercepted by an analog switch and drives the gate driver. The GaN half-bridge output is attenuated and shaped before being fed back to the FPGA duty cycle measurement module to form a closed loop.

[0040] Specifically, in the launch step, a high-efficiency power amplifier based on gallium nitride (GaN) devices is first constructed or employed. Gallium nitride devices possess characteristics such as high breakdown voltage, high electron mobility, and low on-resistance, making them suitable for fabricating Class D or Class E switching power amplifiers, with theoretical efficiencies exceeding 90%. In this embodiment, the input to the power amplifier is a rectangular wave signal with a set duty cycle (e.g., a duty cycle of 50% or...). The square wave (or rectangular wave) has a frequency equal to the required fundamental ultrasonic frequency (e.g., 5MHz). Driven by the rectangular wave, the power amplifier converts DC power into a high-voltage pulse output through a switching action, with the output signal having the same frequency as the rectangular wave. Due to its switching operation, the power amplifier has low power consumption and low heat generation, and can stably generate long pulse signals with millisecond-level durations.

[0041] The specific process for generating a high-voltage pulse wave is as follows: First, a sine wave data of a set frequency (e.g., 5MHz) is generated by the Direct Digital Synthesizer (DDS) core in the FPGA, and then fed into a high-speed digital-to-analog converter to generate a sine wave. High-order harmonics are then filtered out by an LC filter. This sine wave is then combined with a DC voltage generated by a low-speed digital-to-analog converter. (Unit: Volts) A rectangular wave with a set duty cycle is generated through comparison using a high-speed comparator. The pulse count control word of the FPGA main control unit generates a control signal with a time width of T to control the enable terminal of the analog switch, thereby setting the number of output pulses and obtaining a low-voltage pulse wave. This low-voltage pulse wave serves as the input to the gate driver, driving a gallium nitride-based half-bridge circuit to generate a high-voltage pulse wave. This high-voltage pulse wave is attenuated and shaped before being sent to the FPGA's duty cycle measurement module for monitoring, and then fed back to the duty cycle control unit to achieve closed-loop adjustment.

[0042] The high-voltage pulse signal contains a fundamental frequency component and multiple harmonic components. To obtain pure fundamental ultrasonic waves, the high-voltage pulse signal output from the power amplifier is fed into a frequency-selective filter network. This frequency-selective filter network can consist of a fundamental frequency series resonant filter (e.g., composed of an inductor L and a capacitor C connected in series, resonating at the fundamental frequency). The system consists of a fundamental frequency series resonant filter that allows the fundamental component to pass through while filtering out most of the higher harmonics; the target harmonic frequency notch further suppresses harmonic components in the excitation signal that have the same frequency as the target being detected (e.g., when measuring the second harmonic). The notch filter removes the second harmonic in the excitation. The filtered signal becomes a high-voltage excitation signal that is approximately sinusoidal, which is applied to the electrodes of the transmitting transducer (e.g., a piezoelectric ceramic transducer). The transducer converts it into ultrasonic waves and transmits them.

[0043] In practical applications, high-efficiency power amplifiers based on gallium nitride can employ various circuit topologies, such as half-bridge Class D amplifiers or Class E resonant amplifiers. Frequency-selective filtering networks can also take other forms, as long as they can effectively extract the fundamental component and suppress the target harmonics. This embodiment does not impose any limitations on this.

[0044] This embodiment utilizes a high-efficiency power amplifier based on gallium nitride devices to generate millisecond-level long-pulse high-voltage excitation signals under low power consumption and low temperature rise conditions, overcoming the limitation of Class A power amplifiers in being unable to transmit long pulses. The rectangular wave driving method with a set duty cycle simplifies the excitation generation circuit, while frequency-selective filtering (fundamental series resonance plus target harmonic notch filtering) ensures high fundamental purity of the signal applied to the transducer and effectively eliminates target harmonic interference from the excitation itself. These technical features collectively make the long-pulse narrowband detection method feasible in engineering, creating the preconditions for subsequent ultra-narrowband filtering.

[0045] Because the rectangular wave output by the power amplifier contains abundant harmonic components, some of which (such as the second or third harmonic) happen to coincide with the frequency of the target harmonic, this can severely interfere with the measurement of the defect harmonic. Therefore, in one embodiment, the duty cycle is set, including: When the target harmonic is the second harmonic, the duty cycle of the rectangular wave is set to 50%. When the target harmonic is the third harmonic, the duty cycle of the rectangular wave is set to... .

[0046] Specifically, according to Fourier series analysis, one period is A rectangular wave with amplitude A and duty cycle D has the following Fourier series expansion: in When the duty cycle D = 50%, When n is even Therefore, the coefficients of all even harmonics (n=2,4,6,…) are zero, meaning that the ideal 50% duty cycle rectangular wave does not contain second, fourth, or other even harmonics. When this method requires measuring the second harmonic as a detection signal, setting the duty cycle of the rectangular wave of the driving power amplifier to 50% results in almost no second harmonic being generated in the excitation signal. This avoids the second harmonic of the excitation itself being superimposed on the second harmonic generated by the defect, ensuring the authenticity of the detection results.

[0047] Similarly, when the duty cycle hour, It is zero when n is a multiple of 3 (because) Therefore, the coefficients of all integer multiples of 3 harmonics (n=3,6,9,…) are zero, i.e., ideal. The duty cycle rectangular wave does not contain third or sixth harmonics. When this method requires measuring the third harmonic as a detection signal, the duty cycle of the rectangular wave is set to... If this is achieved, almost no third harmonic is generated in the excitation signal, thus eliminating the interference of the excitation's own third harmonic.

[0048] In practical circuits, due to device non-ideals (such as dead time, switching edge asymmetry, gate drive mismatch, etc.), the actual duty cycle may have slight deviations. For example, when a 50% duty cycle is set but the actual value is 49%, the second harmonic amplitude may rise to 3% of the fundamental frequency. Therefore, closed-loop control is needed to accurately maintain the set value. This embodiment provides 50% and... The duty cycle is an optimal value derived from theoretical derivation.

[0049] This embodiment suppresses the target harmonic components generated by the power amplifier itself at the source by selecting specific duty cycles corresponding to different target harmonics, thus avoiding interference from nonlinearity in defect detection. This technique requires no additional filtering circuitry; it can be achieved simply by adjusting the driving waveform parameters, making it simple and effective. Furthermore, because the target harmonic components in the excitation signal are extremely low, the receiver can use a higher gain without worrying about its own harmonic saturation, thereby further improving the detection capability of weak defect signals. This technique requires no additional filtering circuitry; it can be achieved simply by adjusting the driving waveform parameters, making it simple and effective.

[0050] Although a theoretical duty cycle is set, the actual duty cycle of the high-voltage pulse wave output may deviate from the set value due to factors such as dead time, switching edge asymmetry, and gate drive voltage fluctuations in the actual circuit, leading to a decrease in the target harmonic suppression effect. For example, when a 50% duty cycle is set but the actual value is 49%, the second harmonic amplitude may rise to 3% of the fundamental frequency, severely affecting detection accuracy. Therefore, in one embodiment, the transmission step further includes: The actual duty cycle of the high-voltage pulse wave output by the high-efficiency power amplifier is compared with the set duty cycle. Based on the comparison result, the duty cycle of the rectangular wave is adjusted to lock the actual duty cycle at the set duty cycle.

[0051] Specifically, a closed-loop duty cycle control loop is added to the transmission process. First, a signal is separated from the high-voltage pulse wave output from the high-efficiency power amplifier (e.g., attenuated to a safe level via resistor division) and sent to the duty cycle measurement module. This module can use a high-speed comparator to shape the pulse and then use an FPGA to count and measure the high-level duration. (Unit: seconds) and period The ratio (in seconds) yields the actual duty cycle value. Then, the actual duty cycle is... Compared with the target duty cycle set by the user (e.g., 50% or) Compare and calculate the error value. Based on the error value, the feedback control unit (e.g., a proportional-integral-derivative controller in an FPGA) adjusts the duty cycle of the rectangular wave drive signal input to the power amplifier: if the actual duty cycle is too small ( If the duty cycle of the driving rectangular wave is too large, then increase the duty cycle; if it is too large ( If the actual duty cycle is less than the set duty cycle, the duty cycle of the driving rectangular wave will be reduced. This feedback adjustment continues until the difference between the actual duty cycle and the set duty cycle is less than a preset threshold (e.g., ±0.1%). At this point, the actual duty cycle is locked at the set value.

[0052] In practical engineering, the duty cycle measurement module can employ a high-speed analog-to-digital converter (ADC) for sampling and digital processing, or a dedicated time-to-digital converter (TD-SCDMA). Feedback adjustment can be achieved by changing the initial value of the phase accumulator or the comparison threshold of the digital waveform generator. Specifically, in the aforementioned generation process, the FPGA can precisely adjust the duty cycle of the rectangular wave by changing the DC voltage output of the low-speed ADC to alter the toggling point of the high-speed comparator. This embodiment does not impose specific limitations on this, as long as closed-loop locking can be achieved.

[0053] This embodiment uses closed-loop feedback control to compensate for duty cycle deviations caused by circuit non-ideals, ensuring that the actual high-voltage pulse wave's duty cycle is precisely equal to the theoretical optimal value. This mechanism maintains the target harmonic suppression effect in the actual circuit, avoids detection errors caused by duty cycle drift, and improves robustness.

[0054] The specific filters used for bandpass and low-pass filtering in the receiving link affect signal quality and the complexity of the detection system. Therefore, in one embodiment, a crystal filter is used for bandpass filtering after up-conversion; and an LC low-pass filter is used for filtering after low-noise amplification and before up-conversion.

[0055] Specifically, after up-conversion, a first intermediate frequency (IF) signal (e.g., 45MHz) is obtained. To filter out image frequencies, local oscillator leakage, and other out-of-band noise generated by mixing, a crystal filter is used for bandpass filtering. Crystal filters have very high quality factors, reaching tens of thousands or even higher, thus their passband width can be made very narrow (e.g., 30kHz), with good rectangularity and strong out-of-band rejection. Choosing a crystal filter ensures that only narrow-band components near the target frequency pass through the up-converted signal, suppressing broadband noise.

[0056] In addition, an LC low-pass filter is added after the low-noise amplifier and before the up-conversion. The signal amplified by the low-noise amplifier may contain high-frequency noise or harmonic components. If these components directly enter the mixer, they will cause intermodulation distortion and degrade the quality of the received signal. The LC low-pass filter consists of an inductor L and a capacitor C, and its cutoff frequency is... Setting the cutoff frequency to approximately 1.5 times the fundamental or harmonic frequency (for example, for a 5MHz fundamental frequency, the cutoff frequency is set to 30MHz) can effectively filter out high-frequency noise, while also offering advantages such as low insertion loss and low cost.

[0057] In practical applications, crystal filters can use devices with standard intermediate frequency (IF) values ​​such as 455kHz or 10.7MHz, or they can be customized according to specific frequencies. The component parameters of an LC low-pass filter need to be determined based on signal frequency and impedance matching calculations. This embodiment does not impose such limitations.

[0058] This embodiment achieves narrowband filtering by using a crystal filter with a quality factor after up-conversion, laying the foundation for further bandwidth compression by subsequent ultra-narrowband crystal filters; by adding an LC low-pass filter after low-noise amplification, out-of-band noise entering the mixer is reduced, intermodulation distortion is lowered, and the linearity of the receiving link is improved.

[0059] Ultra-narrowband crystal filters (bandwidth 100Hz~1000Hz) have extremely high quality factors, but their center frequency is easily affected by factors such as temperature changes and device aging, causing it to drift. If the local oscillator frequency remains constant, the drifted filter center frequency will mismatch with the signal frequency, leading to attenuation or even blockage of the useful signal. Therefore, in one embodiment, this method further includes: Before performing fundamental frequency detection or harmonic detection, a frequency sweep test is performed on the ultra-narrowband crystal filter to determine the actual passband center frequency of the ultra-narrowband crystal filter; the local oscillator frequency of the down-conversion is adjusted so that the signal center frequency before bandpass filtering of the ultra-narrowband crystal filter is equal to the actual passband center frequency.

[0060] like Figure 5 As shown in the figure, a detailed circuit implementation block diagram of the embodiment provided in this application is illustrated, including FPGA internal modules (clock control, duty cycle measurement, duty cycle control, first direct digital frequency synthesizer (DDS) core, second DDS core, third DDS core, frequency compensation control unit, digital lock-in amplifier (LIA), analog-to-digital converter (ADC), multiple digital-to-analog converters (including 16-bit and 24-bit DACs), LC filters, analog switches, gate drivers, GaN half-bridge, transmitter transducer, receiver pre-selection filter, low-noise amplifier, up-conversion mixer, first intermediate frequency selective amplifier circuit, down-conversion mixer, second intermediate frequency programmable amplifier, ultra-narrowband crystal filter, channel selection relay, etc. The figure fully illustrates the circuit connections for transmitter pulse generation, closed-loop duty cycle control, receiver superheterodyne processing, ultra-narrowband filtering, and frequency self-testing and locking. The frequency self-test path is as follows: the frequency compensation control unit controls the third DDS core to generate a frequency sweep signal, which is sent to the ultra-narrow band crystal filter through the channel selection relay. The output is sampled by the ADC and the amplitude is detected by the digital lock-in amplifier LIA. The FPGA records the amplitude at different frequencies and finds the peak frequency as the actual passband center frequency. Then, the frequency control word of the down-converted local oscillator is adjusted so that the center frequency of the down-converted signal is equal to the actual center frequency.

[0061] Specifically, a self-test procedure is executed before each formal fundamental or harmonic detection. In this procedure, the frequency compensation control unit in the FPGA controls the second intermediate frequency channel selection signal relay to connect a sweep signal source (e.g., a sine wave generated by a DDS) to the input of the ultra-narrowband crystal filter. Then, the frequency control word of the DDS core is changed to achieve continuous frequency variation (sweep), covering the expected passband of the ultra-narrowband crystal filter (e.g., 455kHz ± 1kHz). Simultaneously, a digital lock-in amplifier detects the output amplitude-frequency characteristics of the filter, recording the output amplitude at different frequencies. The frequency corresponding to the point of maximum amplitude is the actual passband center frequency. .

[0062] Obtain the actual passband center frequency Then, adjust the local oscillator frequency (i.e., the second local oscillator frequency) of the down-conversion stage. (unit: Hertz), making the center frequency of the signal input to the ultra-narrowband crystal filter equal to Specifically, assume that the original received signal is up-converted to a fixed first intermediate frequency. (e.g., 45MHz), the originally designed down-conversion local oscillator frequency was... ,in This is the nominal value of the second intermediate frequency (e.g., 455kHz). The actual center frequency is now measured to be... Then adjust This ensures that the frequency of the down-converted signal is exactly equal to... This ensures that the signal falls precisely in the center of the passband of the ultra-narrowband crystal filter. This adjustment can be achieved by changing the division ratio of the DDS core or the phase-locked loop via an FPGA.

[0063] In a practical detection system, frequency sweep detection can be implemented using an FPGA-controlled direct digital frequency synthesizer (DDS), while amplitude detection can be performed by a digital lock-in amplifier. The local oscillator frequency can be adjusted by changing the frequency control word of the DDS or the division ratio of the phase-locked loop. This embodiment does not limit this approach.

[0064] This embodiment determines the real-time center frequency of the ultra-narrowband crystal filter through self-test scanning and dynamically adjusts the down-conversion local oscillator frequency to ensure the signal is always aligned with the filter's passband center, effectively overcoming temperature drift and aging issues. This mechanism ensures the actual bandwidth utilization of the ultra-narrowband filter and avoids attenuation of the useful signal due to frequency mismatch.

[0065] The specific type of high-efficiency power amplifier and the noise figure of the low-noise amplifier have a significant impact on performance. Therefore, in one embodiment, the high-efficiency power amplifier is a Class D or Class E power amplifier, and the noise figure of the amplifier used for low-noise amplification is less than 2.

[0066] Specifically, the high-efficiency power amplifier can employ either a Class D or Class E power amplifier topology. Class D power amplifiers use a half-bridge or full-bridge switching structure, converting DC power to AC output through pulse width modulation or direct rectangular wave drive. Their theoretical efficiency can reach 100%, and their actual efficiency is typically above 85%. Class E power amplifiers are single-ended switching amplifiers that utilize a resonant network to achieve zero-voltage switching, achieving efficiencies of over 90%. Both types are based on gallium nitride (GaN) devices, enabling millisecond-level long pulse transmission with low heat generation, making them suitable for the application scenario described in this embodiment.

[0067] Amplifiers used for low-noise amplification (i.e., low-noise amplifiers) have a noise figure NF (dimensionless, usually expressed in decibels) of less than 2. The noise figure is a measure of the additional noise generated by the amplifier itself, defined as the amplifier's input signal-to-noise ratio. With output signal-to-noise ratio The ratio: A smaller NF value indicates lower noise introduced by the amplifier. In this embodiment, (i.e., less than 2 dB) means that the amplifier's own noise contribution is very small, which is beneficial for maintaining the original signal-to-noise ratio of weak signals. In practical applications, commercially available low-noise amplifier chips (such as chips with a noise figure of 0.8 dB) can be used, or low-noise circuits built with discrete transistors can be employed. As long as the following conditions are met... That's all.

[0068] In practical applications, Class D or Class E power amplifiers can be selected based on the appropriate topology according to output power, operating frequency, and cost. Specific models of low-noise amplifiers can include Analog Devices' ADL5523, Mini-Circuits' PSA series, etc., but this application does not limit these, as long as the noise figure specification is met.

[0069] This embodiment ensures high-efficiency, low-heat-loss generation of the growth pulse excitation signal by limiting the power amplifier to Class D or Class E; and ensures a sufficiently low noise floor for the receiving link by limiting the noise figure of the low-noise amplifier to less than 2. The combined effect of these two factors enables high-sensitivity detection even with small to medium excitation amplitudes.

[0070] The following section uses a specific testing example to illustrate the complete workflow of the method described in this application for detecting 10MHz second harmonics using 5MHz ultrasonic excitation.

[0071] In this example, the transmitting transducer uses a single-crystal longitudinal wave probe with a center frequency of 5MHz (e.g., the 5P10 customized by Shantou Ultrasonic), and the receiving transducer uses a single-crystal longitudinal wave probe with a center frequency of 10MHz (e.g., the 10P6 customized by Shantou Ultrasonic). An ultra-narrowband crystal filter with a bandwidth of 300Hz and a center frequency of 455kHz is selected (the system also offers six selectable frequencies: 100Hz, 1kHz, 3kHz, 7.5kHz, and 15kHz). The pulse width is set to... seconds (approximately 33.3 milliseconds), satisfying Second.

[0072] Step a: Self-test of the center frequency of the second intermediate frequency ultra-narrow band crystal filter.

[0073] The FPGA frequency compensation control unit controls the second intermediate frequency channel selection signal relay to switch to the output of the LC filter to receive a swept sine wave signal. Frequency sweeping is achieved by changing the frequency control word of the DDS core 3. The amplitude-frequency characteristic of the second intermediate frequency filter output is detected by a digital lock-in amplifier, and the actual passband center frequency of the crystal filter is measured. (Unit: Hertz)

[0074] Step b: Generate the fundamental signal.

[0075] The DDS core 1 in the FPGA (model XC7K410T) generates 5MHz sine wave data, writes it to the 16-bit high-speed digital-to-analog converter (model DAC5681) to generate 5MHz sine wave, and uses LC filter 1 to filter out high-order harmonics.

[0076] Step c: Set the duty cycle to generate a rectangular wave.

[0077] The duty cycle control unit writes a DC voltage to a 24-bit low-speed digital-to-analog converter (model DAC1282). (Unit: Volt), the sine wave filtered by LC filter 1 is compared with the sine wave by a high-speed comparator (model LTC6752) to generate a rectangular wave 1 with a duty cycle of 50% (the duty cycle is set to 1 when measuring the third harmonic). ).

[0078] Step d: Pulse width control.

[0079] The pulse count control word of the FPGA main control unit is set to generate a control signal with a time width of TT to control the enable terminal of the analog switch (model TMUX1072) to set the number of output pulses and obtain low-voltage pulse wave 1.

[0080] Step e: High-voltage pulse generation.

[0081] The low-voltage pulse wave 1 is used as the input to the gate driver (model LMG1025) to drive a gallium nitride-based half-bridge circuit to generate a high-voltage pulse wave 1.

[0082] Step f: Closed-loop duty cycle adjustment.

[0083] The high-voltage pulse wave 1 is attenuated and shaped before being sent to the FPGA duty cycle measurement module to monitor the actual duty cycle. The 24-bit low-speed digital-to-analog converter, which feeds back to the duty cycle control unit, adjusts the actual duty cycle of the high-voltage pulse wave 1 to the set value in a closed loop. (50% or ( ), to suppress even harmonics or third harmonics.

[0084] Step g: Frequency selective filtering.

[0085] The high-voltage pulse wave 1 is passed through a high-power LC filter (fundamental series resonant filter) with a cutoff frequency of 7.5MHz and a notch filter (target harmonic frequency notch filter) with a center frequency of 10MHz to further suppress the second harmonic component in the excitation signal, thus obtaining a high-voltage sinusoidal excitation signal.

[0086] Step h: Excite the emission.

[0087] A 5MHz ultrasonic transducer (5P10) is driven by a high-voltage sinusoidal excitation signal to generate 5MHz ultrasonic waves that are applied to the object under test.

[0088] Step i: Harmonic reception.

[0089] Second harmonics were detected using a 10MHz ultrasonic transducer (10P6).

[0090] Step j: Harmonic pre-selection filtering.

[0091] The received signal is passed through a second harmonic bandpass filter (center frequency 10MHz) and a fundamental notch filter (center frequency 5MHz) to suppress the strong fundamental wave and pass through the second harmonic.

[0092] Step k: Low-noise amplification and pre-filtering.

[0093] The signal is amplified using a low-noise amplifier (model PSA39+) and then filtered by an LC low-pass filter with a cutoff frequency of 30MHz.

[0094] Step 1: First frequency converter upconversion and first intermediate frequency filter.

[0095] The signal was up-converted to 45MHz using a mixer (model AD835), with the local oscillator frequency being [missing information]. MHz. Then it is filtered by a crystal filter with a center frequency of 45MHz and a bandwidth of 30kHz, and then amplified 10 times by a high-frequency low-noise amplifier (model OPA847).

[0096] Step m: Second mixing downconversion and ultra-narrowband filtering.

[0097] Use a mixer (model AD835) to downconvert the signal to the actual center frequency of the 455kHz crystal filter measured in step a. The signal is then amplified to 40% to 90% of the full-scale voltage range of the analog-to-digital converter by a programmable amplifier (composed of an ADA4898-1YRDZ and an ADG658 analog switch, with eight gain levels of 1, 2, 5, 10, 20, 50, 100, and 200 times). Finally, bandpass filtering is performed using an ultra-narrow band crystal filter (300Hz bandwidth).

[0098] Step n: Quadrature phase-locked loop (QLL) detection.

[0099] The amplitude of the target frequency (10MHz) was measured using a quadrature digital lock-in amplifier. .

[0100] Step o: Fundamental wave reception (another mode).

[0101] The fundamental frequency can be detected using a 5MHz ultrasonic transducer (5P6), or the amplitude of the 5MHz fundamental frequency can be measured directly using a 10MHz ultrasonic transducer without changing the probe.

[0102] Step p: Fundamental wave pre-selection filtering.

[0103] When detecting the fundamental frequency, the pre-selected filter is configured as a fundamental bandpass filter (center frequency 5MHz) and a low-pass filter to allow the fundamental frequency to pass while suppressing higher harmonics.

[0104] Step q: Fundamental signal processing.

[0105] Repeat steps k to n, changing the target frequency to 5MHz, and obtain the amplitude of the fundamental signal. .

[0106] Step r: Calculate the nonlinear coefficients.

[0107] Based on the measured fundamental amplitude and second harmonic amplitude Calculate the relative nonlinear coefficients The test object is compared with the non-destructive baseline value to determine whether there are defects.

[0108] In this example, the FPGA's highest operating frequency should be greater than 10 times the highest excitation frequency (e.g., greater than 50MHz), the high-speed digital-to-analog converter (DAC) used in the DDS should have a conversion rate greater than 5 times the highest excitation frequency (e.g., greater than 25MHz), and the high-speed comparator's propagation delay should be less than 1% of the ultrasonic excitation period (e.g., for 5MHz, with a period of 200 nanoseconds, the delay should be less than 2 nanoseconds). The noise figure (NF) of the low-noise amplifier should be less than 2. The analog device bandwidth should be greater than 10 times the circuit's operating frequency.

[0109] This example demonstrates that under small to medium amplitude excitation (tens of volts or even a few volts), with a 300Hz ultra-narrow band filter, the equivalent input noise of the receiving circuit is below -144dBm, improving the detection sensitivity compared to existing systems and enabling effective detection of micron-sized internal cracks in cemented carbide.

[0110] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps.

[0111] Secondly, based on the same inventive concept, this application also provides an apparatus for implementing the aforementioned nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method. The solution provided by this apparatus is similar to the implementation described in the above method; therefore, the specific limitations in one or more apparatus embodiments provided below can be found in the above-described limitations of the nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method, and will not be repeated here.

[0112] In one embodiment, such as Figure 6 As shown, the device includes: a transmission module, a fundamental wave detection module, a harmonic detection module, and a calculation and evaluation module. Among them: Transmitting module: Transmits millisecond-level long-pulse ultrasonic waves to excite fundamental ultrasonic waves in the object under test; Fundamental wave detection module: When performing fundamental wave detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by the fundamental wave bandpass filter and the low-pass filter to obtain the fundamental wave pre-filtered signal; Harmonic detection module: When performing harmonic detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a harmonic bandpass filter and a fundamental notch filter to obtain the harmonic pre-filtered signal; The computational evaluation module performs the following steps sequentially on the fundamental pre-selected filtered signal: low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and quadrature phase-locked loop (QLL) detection to extract the amplitude of the fundamental signal. Similarly, it performs the same process on the harmonic pre-selected filtered signal, extracting the amplitude of the target harmonic signal. Based on the amplitudes of the fundamental and target harmonic signals, it calculates the relative nonlinear coefficients to quantitatively assess the material damage state of the tested object. Among them, the bandwidth of the ultra-narrowband crystal filter is 100Hz~1000Hz, and satisfies ,in The pulse width of the millisecond-level long pulse excitation signal. This represents the bandwidth of the ultra-narrowband crystal filter.

[0113] In one embodiment, the transmitting module includes a transmitting unit.

[0114] The transmitting unit uses a high-efficiency power amplifier based on gallium nitride devices to drive the high-efficiency power amplifier with a rectangular wave of set duty cycle to generate millisecond-level long pulse ultrasonic waves, which are then applied to the transmitting transducer after frequency selection filtering.

[0115] In one embodiment, the transmitting unit includes a duty cycle setting subunit.

[0116] The duty cycle setting subunit is used to set the duty cycle of the rectangular wave to 50% when the target harmonic is the second harmonic. The duty cycle setting subunit is used to set the duty cycle of the rectangular wave to 50% when the target harmonic is the third harmonic. .

[0117] In one embodiment, the transmitter module includes a duty cycle locking unit.

[0118] The duty cycle locking unit is used to compare the actual duty cycle of the high-voltage pulse wave output by the high-efficiency power amplifier with the set duty cycle, and adjust the duty cycle of the rectangular wave according to the comparison result, so that the actual duty cycle is locked at the set duty cycle.

[0119] In one embodiment, the bandpass filtering after upconversion is implemented using a crystal filter, and the calculation and evaluation module is also used to perform filtering using an LC low-pass filter after low-noise amplification and before upconversion.

[0120] In one embodiment, the device further includes a frequency sweep detection module.

[0121] The frequency sweep detection module is used to perform frequency sweep detection on the ultra-narrowband crystal filter before fundamental wave detection or harmonic detection to determine the actual passband center frequency of the ultra-narrowband crystal filter; and to adjust the local oscillator frequency of the down-conversion so that the signal center frequency before bandpass filtering of the ultra-narrowband crystal filter is equal to the actual passband center frequency.

[0122] In one embodiment, the transmitting unit further includes an amplifier configuration subunit.

[0123] The amplifier configuration subunit is used to employ a Class D power amplifier or a Class E power amplifier as a high-efficiency power amplifier, and to configure the noise figure of the amplifier used for low-noise amplification to be less than 2.

[0124] Each module in the aforementioned nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0125] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 7 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores all data required for the device's operation. The network interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a nonlinear ultrasonic long-pulse ultra-narrowband harmonic detection method.

[0126] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0127] Thirdly, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described in the above method embodiments.

[0128] Fourthly, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of any of the methods described in the above method embodiments.

[0129] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0130] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0131] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method, characterized in that, The method includes: Transmission steps: Transmit millisecond-level long-pulse ultrasonic waves to excite fundamental ultrasonic waves in the object under test; Fundamental wave detection step: When performing fundamental wave detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a fundamental wave bandpass filter and a low-pass filter to obtain the fundamental wave pre-filtered signal; Harmonic detection steps: When performing harmonic detection, the transmission step is executed to obtain the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a harmonic bandpass filter and a fundamental notch filter to obtain the harmonic pre-filtered signal. The calculation and evaluation steps are as follows: The fundamental pre-selected filtered signal is sequentially subjected to low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and quadrature phase-locked loop (QLL) detection to extract the amplitude of the fundamental signal; the harmonic pre-selected filtered signal is sequentially subjected to low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and QLL detection to extract the amplitude of the target harmonic signal; the relative nonlinear coefficient is calculated based on the amplitudes of the fundamental signal and the target harmonic signal to quantitatively assess the material damage state of the tested object; The ultra-narrowband crystal filter has a bandwidth of 100Hz~1000Hz and satisfies... ,in The pulse width of the millisecond-level long pulse excitation signal is given. The bandwidth of the ultra-narrowband crystal filter is given.

2. The method according to claim 1, characterized in that, The launch step includes: A high-efficiency power amplifier based on gallium nitride devices is used to drive the high-efficiency power amplifier with a rectangular wave of set duty cycle to generate the millisecond-level long pulse ultrasonic wave, which is then applied to the transmitting transducer after frequency selection filtering.

3. The method according to claim 2, characterized in that, The set duty cycle includes: When the target harmonic is the second harmonic, the duty cycle of the rectangular wave is set to 50%. When the target harmonic is the third harmonic, the duty cycle of the rectangular wave is set to... .

4. The method according to claim 3, characterized in that, The launch step also includes: The actual duty cycle of the high-voltage pulse wave output by the high-efficiency power amplifier is compared with the set duty cycle. Based on the comparison result, the duty cycle of the rectangular wave is adjusted to lock the actual duty cycle at the set duty cycle.

5. The method according to claim 1, characterized in that, The bandpass filtering after the upconversion is implemented using a crystal filter; after the low-noise amplification and before the upconversion, the filtering is further performed using an LC low-pass filter.

6. The method according to any one of claims 1 to 5, characterized in that, The method further includes: Before performing the fundamental frequency detection or the harmonic detection, the ultra-narrowband crystal filter is subjected to frequency sweep detection to determine the actual passband center frequency of the ultra-narrowband crystal filter; the local oscillator frequency of the down-conversion is adjusted so that the signal center frequency before bandpass filtering of the ultra-narrowband crystal filter is equal to the actual passband center frequency.

7. The method according to claim 2, characterized in that, The high-efficiency power amplifier is a Class D power amplifier or a Class E power amplifier, and the noise figure of the amplifier used for the low-noise amplification is less than 2.

8. A nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection device, characterized in that, The device includes: Transmitting module: Transmits millisecond-level long-pulse ultrasonic waves to excite fundamental ultrasonic waves in the object under test; Fundamental wave detection module: When performing fundamental wave detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a fundamental wave bandpass filter and a low-pass filter to obtain a fundamental wave pre-filtered signal; Harmonic detection module: When performing harmonic detection, the transmission step is executed to acquire the ultrasonic echo signal, and the ultrasonic echo signal is pre-filtered by a harmonic bandpass filter and a fundamental notch filter to obtain a harmonic pre-filtered signal; The calculation and evaluation module performs the following steps sequentially on the fundamental pre-selected filtered signal: low-noise amplification, up-conversion, bandpass filtering, down-conversion, bandpass filtering via an ultra-narrowband crystal filter, and quadrature phase-locked loop detection, to extract the amplitude of the fundamental signal. Similarly, it performs the same steps on the harmonic pre-selected filtered signal, extracting the amplitude of the target harmonic signal. Based on the amplitudes of the fundamental signal and the target harmonic signal, it calculates a relative nonlinear coefficient to quantitatively assess the material damage state of the tested object. The ultra-narrowband crystal filter has a bandwidth of 100Hz~1000Hz and satisfies... ,in The pulse width of the millisecond-level long pulse excitation signal is given. The bandwidth of the ultra-narrowband crystal filter is given.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the nonlinear ultrasonic long-pulse ultra-narrow band harmonic detection method according to any one of claims 1 to 7.

Images