TDLAS second harmonic software demodulation method and device

Abstract

The invention provides a TDLAS (tunable diode laser absorption spectroscopy) second harmonic software demodulation method, which comprises the steps of modulation signal acquisition, signal conversion, equivalent sampling and signal reconstruction, mobile fitting harmonic demodulation, signal smoothing, parabola fitting peak extraction and the like. The peak value is extracted by combining moving average smoothing and parabola fitting, high-precision and low-cost real-time measurement of the gas concentration of the TDLAS system is realized by a full-software scheme, meanwhile, the complexity of the system is reduced, and the method is suitable for dynamic measurement and analysis of the gas concentration.

Description

Technical Field

[0001] The present invention relates to the technical field of spectral analysis, in particular to a software demodulation method for the second harmonic of TDLAS. Background Art

[0002] Tunable Diode Laser Absorption Spectroscopy (TDLAS) technology uses current and temperature to precisely tune and control a laser, scans one or several absorption lines of a gas to be measured, and inversely calculates the concentration of the gas to be measured based on the change in laser intensity before and after the characteristic absorption of the incident light by the gas to be measured. It is a conventional spectral analysis technology.

[0003] In a traditional TDLAS (Tunable Diode Laser Absorption Spectroscopy) system, there are mainly two demodulation methods: hardware demodulation and FPGA demodulation. However, both methods have their respective limitations: The hardware demodulation method has a complex circuit, difficult parameter adjustment, high cost, poor adaptability, and is affected by the fundamental frequency harmonic. The signal can only be amplified after envelope demodulation and is greatly affected by low-frequency power supply noise. Although the FPGA demodulation method can improve the processing speed, it increases the additional hardware cost, requires a high-speed ADC acquisition card and an FPGA hardware algorithm chip, and the calculation result also needs to be re-simulated and filtered, which reduces the envelope peak to a certain extent. Similarly, the signal can only be amplified after envelope demodulation, affecting the measurement accuracy.

[0004] Chinese Patent Application No. CN202410313945.2 discloses a multi-gas concentration synchronous measurement device based on digital modulation and demodulation of TDLAS, including: a laser driver, a DFB laser, a beam combiner, an optical fiber collimator, a gas absorption cell, a gas distribution system, a photodetector, an analog-to-digital converter, an FPGA, and a digital-to-analog converter; the digital-to-analog converter is connected to the laser driver and the FPGA, with its input end connected to the FPGA for receiving the low-frequency scanning signal and the high-frequency modulation signal generated by the modulation signal generator in the FPGA, and its output end connected to the laser driver for outputting the converted analog signal to the laser driver. The present invention uses the method of multiple light sources and a single detector. Multiple signals only need to adjust a set of optical paths, improving the system integration. At the same time, starting from the software aspect, it avoids the noise interference of electronic components in the traditional digital lock-in amplifier circuit and solves the problem that most current TDLAS gas detection systems measure only a single gas and the integration of multi-gas detection technology is relatively low. However, it still has problems such as insufficient cross-interference suppression and poor real-time performance.

[0005] In view of this, the applicant has proposed this software demodulation method for the second harmonic of TDLAS, which replaces the hardware demodulation module with a software algorithm, reduces the dependence on high-speed ADC and FPGA chips, and significantly reduces the system cost. It solves the deficiencies existing in the prior art. Summary of the Invention

[0006] The present invention provides a TDLAS second-harmonic software demodulation method and device, which solves the problems of complex hardware demodulation and high cost in the prior art. At the same time, the envelope solution of the software algorithm greatly improves the flexibility of data processing and the peak point has no attenuation, greatly reducing the hardware cost and improving the measurement accuracy. The technical solution of the present invention is realized as follows:

[0007] A TDLAS second-harmonic software demodulation method includes the following steps:

[0008] Drive a tunable diode laser to emit laser with a specific wavelength and obtain an equivalent sampling signal. Perform sinusoidal moving fitting on the equivalent sampling signal through a moving fitting harmonic demodulation algorithm to obtain the envelope curve of the second harmonic, and use a sliding average filtering algorithm to smooth the second harmonic signal; on the smoothed envelope curve, select the data segment near the peak point, calculate the vertex position through parabola fitting to eliminate local extreme interference; according to Lambert-Beer's law, establish a linear relationship between the second harmonic amplitude and the gas concentration, and invert the target gas concentration through the calibration coefficient.

[0009] As a preferred technical solution, by strictly setting the proportional relationship between the sampling frequency and the signal frequency, the actually undersampled data has the same function as the oversampled data. Assume the signal frequency is f x , and the sampling frequency is f s , in m cycles of the signal, n points can be sampled, where m and n are relatively prime, ensuring that m·f s = n·f x , which is equivalent to increasing the sampling frequency by m times, ensuring that the low-frequency sampling system can obtain high-frequency sampling results, where m, n, f x , f s need to be reasonably set according to the system clock. By reconstructing n points into 1 cycle, equivalent sampling can be achieved. The reconstruction method is as follows:

[0010] r′(mod(i*m,n)+1) = r(i + 1) (1)

[0011] In the formula: r is the original signal, r′ is the reconstructed signal, m is the number of signal cycles, n is the number of sampling points, i = 0~n - 1, and mod represents taking the modulus.

[0012] As a preferred technical solution, the moving fitting harmonic demodulation algorithm specifically includes the following steps:

[0013] Step 1) After sampling the signal, obtain a corresponding sequence of sampling values Y = {y1,..., y i ,..., y N′} and represent it as:

[0014]

[0015] Where: Y i is discrete sampling data; i is the sampling sequence, i = 0 to N - 1, N is the sampling length; A is the amplitude; Δ i = 2πfi / f s , f is the frequency of the input signal, f s is the sampling frequency; is the initial phase; ε i is the sampling error; Let Equation (2) can be further expressed as

[0016] y i = a cos Δ i + b sin Δ i + c + ε i (3)

[0017] Step 2) For a certain segment of data {y k ,..., y k+m-1} starting from a certain point in the sampling sequence, a total of m data points are used for the least squares fitting of trigonometric functions. m is usually set to the number of cycle points of equivalent sampling. From the least squares linear fitting, the fitting parameters ak, bk, and ck of this segment can be expressed as:

[0018]

[0019] where, is:

[0020]

[0021] Thus, the fitting amplitude and fitting phase of this segment are obtained as

[0022]

[0023] Step 3) Starting from the starting data, the method of moving sine fitting segment by segment can be used to demodulate the continuous amplitude and phase information of the sequence, reducing the computational load; in the TDLAS gas concentration detection, the sequence signal sampled by AD contains a fundamental wave with a continuously changing amplitude and high-order harmonic signals, expressed as:

[0024]

[0025] Considering the sampling phase 2δ i of the second harmonic, that is, Δ i = 2δ i , and then through the sine fitting method and the idea of moving segment by segment, the envelope curve of the second harmonic can be obtained:

[0026]

[0027] Step 4) Change Δ i value, and the amplitude envelope curves of the fundamental harmonic and other higher harmonics can also be obtained.

[0028] As a preferred technical solution, by using a moving average filtering algorithm, the second harmonic signal is smoothed, that is, the input array is operated on through a sliding window of a fixed size, the average value of the elements in the window is calculated, and by utilizing the characteristic of the window sliding, the algorithm only needs to update the sum of the window in each iteration, thereby efficiently realizing signal smoothing, where the smoothing is performed for multiple iterations.

[0029] As a preferred technical solution, in order to further eliminate the influence of local extrema on the measurement accuracy, for the smoothed envelope curve, parabola fitting is implemented using N points before and after the envelope maximum point, N≥300, and the vertex of the fitted parabola is calculated. Among them, let the fitted quadratic polynomial function be

[0030] y = ax 2 + bx + c (9)

[0031] where a, b, and c are fitting parameters,

[0032] and its vertex is:

[0033]

[0034] A spectral analysis device for performing a TDLAS second harmonic software demodulation method, comprising:

[0035] A signal sampling device for collecting the original signal, and by strictly setting the proportional relationship between the sampling frequency and the signal frequency, making the actually under-sampled data have the same function as the over-sampled data;

[0036] A signal processing device that intelligently analyzes the equivalent sampling signal collected by the signal sampling device through moving fitting and peak optimization techniques;

[0037] An interaction device: providing a friendly interaction interface and interacting with the user on the results of the second harmonic software demodulation.

[0038] Compared with the prior art, the present solution has the following beneficial effects:

[0039] (1) Cost and complexity reduction: By replacing the hardware demodulation module with a software algorithm, the dependence on high-speed ADC and FPGA chips is reduced, and the system cost is significantly reduced.

[0040] (2) Improvement in measurement accuracy and stability: The equivalent sampling and moving fitting algorithms avoid signal attenuation, suppress noise and zero drift, and extract the second harmonic peak more accurately; the smoothing process and parabolic fitting further optimize the signal-to-noise ratio and enhance the measurement repeatability.

[0041] (3) Enhancement of real-time performance and flexibility: The recursive calculation reduces data storage and computational volume, meeting the requirements of high-speed real-time processing; the algorithm parameters are adjustable to adapt to the concentration detection of different gases (such as oxygen and methane), with strong versatility.

[0042] (4) Optimization of anti-interference ability: The multi-component gas second harmonic separation algorithm effectively suppresses cross-interference and is applicable to complex gas environments. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0044] Figure 1(a) is the first equivalent sampling diagram;

[0045] Figure 1(b) is the equivalent sampling Figure 2 ;

[0046] Figure 2 is the system diagram of the TDLAS gas measurement method based on moving fitting harmonic demodulation of the present invention;

[0047] Figure 3 is the signal sampling diagram;

[0048] Figure 4 is the flowchart of a TDLAS second harmonic software demodulation method of the present invention;

[0049] Figure 5 is the comparison diagram between the actual undersampled signal and the equivalent sampling;

[0050] Figure 6 is the comparison diagram of the second harmonic before and after smoothing;

[0051] Figure 7 is the parabola fitting second harmonic peak point diagram;

[0052] Figure 8 is the second harmonic diagram at different concentrations;

[0053] Figure 9 is the multiple measurement result diagram at different concentrations;

[0054] Figure 10Schematic diagram of the linear relationship between the second harmonic amplitude and the oxygen concentration;

[0055] Figure 11 Schematic diagram of the measurement signal-to-noise ratio. Specific implementation mode

[0056] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work shall fall within the protection scope of the present invention.

[0057] Refer to Figure 4 , the present invention provides a TDLAS second harmonic software demodulation method, including the following steps:

[0058] (1) System initialization, configure the hardware parameters of the STM32 single-chip microcomputer (such as timers, ADCs, DACs), generate a DDS look-up table (LUT), and prepare the modulation signal waveform.

[0059] (2) Modulation signal generation and laser driving, generate a 40 Hz triangular wave (controlled by TIM5) and a 78.9 kHz sine wave (controlled by TIM7) through DDS technology, convert them into analog signals through DAC1, and drive the DFB laser to emit modulated laser.

[0060] (3) Optical signal absorption and electrical signal conversion, after the laser passes through the gas absorption cell, the remaining unabsorbed light is converted into an electrical signal by a photodetector, and after high-pass filtering and amplification, it is input to the ADC for sampling (triggered by TIM3, frequency about 545 kHz).

[0061] (4) Equivalent sampling and signal reconstruction, perform periodic reconstruction on the ADC sampling data according to the co-prime ratio relationship (m·f s =n·f x ) to generate an equivalent oversampled discrete signal sequence.

[0062] (5) Moving fitting harmonic demodulation, select data segments one by one, apply the least squares sine fitting algorithm to demodulate the amplitude and phase of the second harmonic, reduce the computational complexity through recursive calculation, and extract the envelope curve.

[0063] (6) Moving average smoothing process, perform moving average filtering (window size adjustable) on the second harmonic envelope curve, suppress noise interference, and retain the main features of the signal.

[0064] (7) Parabolic fitting peak extraction, on the smoothed envelope curve, select the data segment near the peak point, calculate the vertex position through quadratic polynomial fitting, and eliminate the local extreme value error.

[0065] (8) Gas concentration inversion: Based on the Lambert-Beer law, using the linear relationship between the amplitude of the second harmonic and the concentration, and combining with the calibration coefficient to invert the target gas concentration.

[0066] (9) Result output and cyclic measurement: Output the current concentration value, and the system returns to the initial state to prepare for the next measurement cycle.

[0067] Specifically, it includes the following processes:

[0068] By strictly setting the proportional relationship between the sampling frequency and the signal frequency, the actually undersampled data has the same function as the oversampled data. Assuming the signal frequency is f x , the sampling frequency is f s , in m cycles of the signal, n points can be sampled, where m and n are relatively prime. Ensure that m·f s = n·f x , which is equivalent to increasing the sampling frequency by m times, ensuring that the low-frequency sampling system can obtain high-frequency sampling results. Among them, m, n, f x , f s need to be reasonably set according to the system clock. By reconstructing n points into 1 cycle, equivalent sampling can be achieved. The reconstruction method is as follows:

[0069] r′(mod(1i*m,n)+1) = r(i+1) (1)

[0070] In the formula: r is the original signal, r′ is the reconstructed signal, m is the number of signal cycles, n is the number of sampling points, i = 0~n-1, and mod represents taking the modulus.

[0071] As a preferred technical solution, the moving fitting harmonic demodulation algorithm specifically includes the following steps:

[0072] Step 1) After sampling the signal, obtain a corresponding sequence of sampling values Y = {y1,..., y i ,..., y N′} is expressed as:

[0073]

[0074] In the formula: Y i is the discrete sampling data; i is the sampling sequence, i = 0~N-1, N is the sampling length; A is the amplitude; Δ i = 2πfi / f s , f is the frequency of the input signal, f s is the sampling frequency; is the initial phase; ε i is the sampling error; Let a =, b =, Equation (2) can be further shown as

[0075] y i =acosΔ i +bsinΔ i +c+ε i (3)

[0076] Step 2) For a certain segment of data {y k ,...,y k+m-1} starting from a certain point in the sampling sequence, a total of m data points are used for the least squares fitting of trigonometric functions. m is usually set to the number of period points of equivalent sampling. From the least squares linear fitting, the fitting parameters ak, bk, and ck of this segment can be expressed as:

[0077]

[0078] where, is:

[0079]

[0080] Thus, the fitting amplitude and fitting phase of this segment are obtained as

[0081]

[0082] Step 3) Starting from the starting data, a method of moving sine fitting segment by segment can be used to demodulate the continuous amplitude and phase information of the sequence, so as to reduce the calculation amount; in the TDLAS gas concentration detection, the sequence signal sampled by AD contains a fundamental wave with continuously changing amplitude and high-order harmonic signals, which is expressed as:

[0083]

[0084] Considering the sampling phase 2δ i of the second harmonic, that is, Δ i =2δ i , and then through the sine fitting method and the idea of moving segment by segment, the envelope curve of the second harmonic can be obtained:

[0085]

[0086] Step 4) By changing the value of Δ i , the amplitude envelope curves of the fundamental harmonic and other high-order harmonics can also be obtained.

[0087] As a preferred technical solution, by using a moving average filtering algorithm, the second harmonic signal is smoothed, that is, the input array is operated on through a sliding window of a fixed size, and the average value of the elements in the window is calculated. Utilizing the characteristic of window sliding, the algorithm only needs to update the sum of the window in each iteration, thereby efficiently realizing signal smoothing, where the smoothing times are carried out through multiple iterations.

[0088] As a preferred technical solution, in order to further eliminate the influence of local extrema on the measurement accuracy, for the smoothed envelope curve, parabola fitting is implemented using N points before and after the envelope maximum point, where N≥300, and the vertex of the fitted parabola is calculated. Among them, the quadratic polynomial function obtained by fitting is set as

[0089] y = ax 2 + bx + c (9)

[0090] where a, b, and c are fitting parameters,

[0091] and its vertex is:

[0092]

[0093] Among them, the system hardware design diagram of the TDLAS gas measurement method based on moving fitting harmonic demodulation is as Figure 2 shown. The system chip selects STM32F407. In the initial stage, the direct digital synthesis (DDS) technology is used to generate signals through the DDS look-up table (LUT). The signals are transmitted to the digital-to-analog converter (DAC1) through the DMA method. These converters convert the signals in digital form into analog signals. Among them, DAC1 is used as the reference voltage and input to the reference voltage of the digital-to-analog conversion chip TLV5614. Sine waves, triangular waves, and bias signals required for generating control lasers are generated through SPI communication to drive the lasers and control them to emit light of a specific wavelength. After the laser interacts with the target gas molecules in the optical cell, the remaining optical signals are captured by the photoelectric sensor and converted into electrical signals. These electrical signals are first processed by two high-pass filters and amplified on the main board to enhance the signal intensity. Subsequently, the signals are sampled by the ADC

[0094] In software design, the TIM5 of the single-chip microcomputer triggers an interrupt to perform SPI communication with the TLV5614, generating a triangular wave with a frequency of 40 Hz. The pre-allocation value of TIM5 is set to 1, and the auto-reload value is set to 8400. The TIM7 of the single-chip microcomputer controls the DAC to generate a sine wave. The pre-allocation value of TIM7 is set to 1, and the auto-reload value is set to 14. The number of DAC points in one cycle is set to 76. The DAC inputs a sine signal with a frequency of 78.9 KHz to the reference signal of the CD port of the TLV5614 through the DMA method. In this way, only a fixed value needs to be written to it during initialization, and the output signal of port C can change with the reference signal, that is, output a sine signal with a frequency of 78.9 KHz. In addition, the TIM3 of the single-chip microcomputer triggers the ADC for sampling. The pre-allocation value of TIM3 is set to 1, and the auto-reload value is set to 154. The sampling frequency of the ADC is about 545 KHz. A total of 1657 cycles are sampled in 21 ms; 11449 points are sampled. Ensuring that the scanning interval can be scanned and having a certain redundancy can provide enough data for the differential processing of the half cycle.

[0095] Figure 3 The signal sampled by the ADC. Figure 4 For the system flowchart, Figure 5 For the actual undersampling signal and the equivalent sampling diagram, Figure 6 For the comparison diagram of the second harmonic before and after smoothing, Figure 7 For the parabola fitting diagram of the peak points of the second harmonic, Figure 8 For the second harmonic diagrams at different concentrations, Figure 9 For the multiple measurement diagrams at different concentrations, Figure 10 For the linear relationship between the second harmonic amplitude and the oxygen concentration, Figure 11 For the measurement signal-to-noise ratio diagram.

[0096] Summarizing the above results, the technical problems solved and advantages of this solution are as follows:

[0097] (1) Solved the problems of complex hardware demodulation and high cost. At the same time, the envelope solution of the software algorithm greatly improves the flexibility of data processing and the peak points have no attenuation, greatly reducing the hardware cost and improving the measurement accuracy;

[0098] (2) Solved the problems of large computational amount and long solution time. The amplitude-phase algorithm of sine fitting has high precision, but the piecewise fitting of envelope calculation is time-consuming and requires a large system storage space, which is not suitable for high-speed real-time processing. The sine moving fitting demodulation method adopts the forward recurrence idea. Under the condition of not requiring re-fitting, it uses the intermediate matrix parameters of the previous fitting to calculate the fitting results of the data segment starting from the next point, and obtains new intermediate matrix parameters until the fitting results of each segment during the movement are obtained. This method greatly reduces the computational amount and the solution time of the envelope curve, and improves the real-time performance of signal processing.

[0099] (3) Improved the measurement accuracy and stability. Moving fitting can maximize the suppression of the influence of zero drift and improve the measurement stability.

[0100] (4) Strong adaptability. Multiple information of a signal can be synchronously extracted. By using the fundamental harmonic light intensity to replace the background light intensity to calculate the gas concentration, the accuracy is improved.

[0101] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A TDLAS second harmonic software demodulation method, characterized in that, It includes the following steps: Drive a tunable diode laser to emit laser with a specific wavelength and obtain an equivalent sampling signal. Use the moving fitting harmonic demodulation algorithm to perform sinusoidal moving fitting on the equivalent sampling signal to obtain the envelope curve of the second harmonic. Use the sliding average filtering algorithm to smooth the second harmonic signal; on the smoothed envelope curve, select the data segment near the peak point, calculate the vertex position through parabolic fitting to eliminate the interference of local extrema; according to Lambert-Beer's law, establish a linear relationship between the second harmonic amplitude and the gas concentration, and invert the target gas concentration through the calibration coefficient.

2. The TDLAS second harmonic software demodulation method according to claim 1, characterized in that By strictly setting the proportional relationship between the sampling frequency and the signal frequency, the actually under-sampled data has the same function as the over-sampled data. Assume the signal frequency is f x , and the sampling frequency is f s . In m cycles of the signal, n points can be sampled, where m and n are relatively prime, ensuring that m·f s = n·f x , that is, the sampling frequency is increased by m times, and the low-frequency sampling system can obtain the high-frequency sampling result. Among them, m, n, f x , f s need to be reasonably set according to the system clock. By reconstructing n points into 1 cycle, equivalent sampling can be achieved. The reconstruction method is as follows: r′(mod(i*m,n)+1)=r(i+1) (1) In the formula: r is the original signal, r′ is the reconstructed signal, m is the number of signal periods, n is the number of sampling points, i = 0 to n - 1, and mod represents taking the modulus.

3. The TDLAS second harmonic software demodulation method according to claim 1, wherein The moving fitting harmonic demodulation algorithm specifically includes the following steps: Step 1) After sampling the signal, a corresponding sequence of sampling values Y = {y1,..., y i ,..., y N′} is expressed as: Where: Y i is discrete sampling data; i is the sampling sequence, i = 0 to N - 1, N is the sampling length; A is the amplitude; Δ i = 2πfi / f s , f is the frequency of the input signal, f s is the sampling frequency; is the initial phase; ε i is the sampling error; Let Equation (2) can be further expressed as y i =acosΔ i +bsinΔ i +c+ε i (3) Step 2) For a certain segment of data {y k ,..., y k+m-1} starting from a certain point in the sampling sequence, a total of m data points are subjected to least-squares fitting of trigonometric functions. m is set to the number of cycle points of equivalent sampling. From the least-squares linear fitting, the fitting parameters ak, bk, and ck of this segment can be expressed as: Among them, is:[[]]END]] Furthermore, the fitting amplitude and fitting phase of this segment are obtained as follows: Step 3) Starting from the starting data, use the method of segment-by-segment moving sine fitting to demodulate the continuous amplitude and phase information of the sequence to reduce the calculation amount; in the TDLAS gas concentration detection, the sequence signal sampled by AD contains a fundamental wave with a continuously changing amplitude and high-order harmonic signals, and its expression is: Considering the sampling phase 2δ of the second harmonic i , that is, Δ i = 2δ i , and then by using the sine fitting method and the idea of moving segment by segment, the envelope curve of the second harmonic can be obtained: Step 4) Changing Δ i can also obtain the amplitude envelope curves of the fundamental harmonic and other higher harmonics.

4. The TDLAS second harmonic software demodulation method according to claim 1, wherein By using the sliding average filtering algorithm, the second harmonic signal is smoothed, that is, the input array is operated through a sliding window of a fixed size, the average value of the elements in the window is calculated, and by using the characteristic of the window sliding, the algorithm only needs to update the sum of the window in each iteration, so as to efficiently realize signal smoothing, where the smoothing is performed through multiple iterations.

5. The TDLAS second harmonic software demodulation method according to claim 1, characterized in that In order to further eliminate the influence of local extrema on the measurement accuracy, for the smoothed envelope curve, apply parabola fitting to the N points before and after the envelope maximum point, N≥300, and calculate the vertex of the fitted parabola. Among them, assume that the fitted quadratic polynomial function is y = ax 2 + bx + c (9) where a, b, and c are fitting parameters, Its vertex is:

6. A spectral analysis device for performing a TDLAS second harmonic software demodulation method as described in any one of claims 1 to 5 above, characterized in that, It includes: A signal sampling device, which is used to collect the original signal, and by strictly setting the ratio relationship between the sampling frequency and the signal frequency, make the actually undersampled data have the same function as the oversampled data; A signal processing device, which intelligently analyzes the equivalent sampling signal collected by the signal sampling device through moving fitting and peak optimization techniques; An interaction device: provides a friendly interaction interface and interacts with the user on the result of the second harmonic software demodulation.

Images