A method and system for laser carbon monoxide monitoring

Abstract

The application relates to the technical field of industrial monitoring, and discloses a laser carbon monoxide monitoring method and a monitoring system, which comprises the following steps: scanning a carbon monoxide absorption spectrum line by alternately outputting low-depth and high-depth modulation signals through a laser driver; extracting photoelectric signal harmonics of different channels and calculating a ratio; inversely calculating low-depth and high-depth initial concentrations in combination with environmental temperature and pressure parameters; calculating a deviation parameter according to the difference between the two and converting the deviation parameter into an equivalent absorption line width; then generating an adaptive modulation signal and feeding back the adaptive modulation signal to a driving module to perform re-scanning; finally, using a newly obtained harmonic ratio and the equivalent absorption line width to substitute into an inversion model to obtain a corrected concentration, and outputting a final result after applying a maximum change rate limit. The application realizes dynamic compensation of measurement errors caused by self-exhibition, and significantly improves the accuracy of single-optical-path gas monitoring under harsh working conditions.

Description

Technical Field

[0001] This invention relates to the field of industrial monitoring technology, and more specifically, to a laser carbon monoxide monitoring method and system. Background Technology

[0002] Currently, laser absorption spectroscopy is widely used for online monitoring of carbon monoxide (CO) in industrial combustion processes and emissions. However, in combustion chambers near the outlet or industrial furnace combustion channels, where the combustion is influenced by wall cooling films, local quenching, and high-temperature exhaust gas mixing, the limitations of the operating conditions make single-path laser monitoring more suitable than multi-projection tomography. Under these complex conditions, the gas distribution within the channel is not completely uniform, and CO accumulates locally near the wall or in the quenching zone. Furthermore, influenced by factors such as wall heat transfer and freezing oxidation, the temperature distribution and CO concentration distribution within the channel are not synchronized; lower temperature regions may correspond to higher CO concentrations. Existing conventional methods, when processing single-path monitoring signals, often treat the medium throughout the entire optical path as a completely uniform state for equivalent linewidth calculations. However, the collisional broadening of the gas is actually a superposition of external gas broadening and self-broadening. When a local high concentration of CO co-occurs with a lower temperature region, the temperature amplification will occur simultaneously at the CO enrichment site. This will inevitably lead to a significant deviation between the actual equivalent linewidth along the optical path and the calculated linewidth under the condition of uniformity.

[0003] Furthermore, traditional harmonic online monitoring schemes typically use only a single, fixed modulation depth to drive the laser for scanning. This conventional approach results in a lack of differentiation in the system's response to the same hidden linewidth shift, making it impossible for the instrument to effectively identify the hidden bias caused by self-broadening. Therefore, when faced with complex, non-uniform combustion channels, existing single-path monitoring schemes incorrectly map this parameter drift caused by spatial distribution covariance into concentration calculation errors, ultimately leading to severely distorted concentration inversion results due to treating the medium as completely homogeneous. How to transform the hidden self-broadening bias into observable control parameters and compensate for it within the single-path monitoring framework under such harsh conditions, thereby improving the accuracy of the final output concentration, is a core technical challenge that urgently needs to be solved in the industry. Summary of the Invention

[0004] This invention provides a laser carbon monoxide monitoring method and system, which solves the technical problems mentioned in the background art.

[0005] In a first aspect, a laser-based carbon monoxide monitoring method includes:

[0006] The laser driver alternately outputs low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles to drive the laser to scan the carbon monoxide absorption spectrum.

[0007] The laser is received and converted into a photoelectric signal. The photoelectric signal is amplified by phase lock, and the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal are extracted respectively. The ratio of low-depth harmonics to high-depth harmonics is calculated.

[0008] By combining the environmental temperature and pressure parameters, the low-depth harmonic ratio and the high-depth harmonic ratio are substituted into the concentration inversion model to obtain the low-depth initial concentration and the high-depth initial concentration.

[0009] The deviation parameter is calculated based on the difference between the initial concentration at low depth and the initial concentration at high depth, and the deviation parameter is converted into an equivalent absorption linewidth;

[0010] An adaptive modulation signal is generated based on the equivalent absorption linewidth and fed back to the laser driver. The adaptive modulation signal is then used to rescan and obtain the adaptive harmonic ratio.

[0011] Substituting the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model, the corrected concentration is obtained;

[0012] Apply a maximum rate of change limit to the corrected concentration and output the final carbon monoxide concentration.

[0013] Secondly, a laser carbon monoxide monitoring system, comprising the steps of a laser carbon monoxide monitoring method as described in any one of the claims, including:

[0014] The alternating scanning module is used to drive the laser to scan the carbon monoxide absorption spectrum by alternately outputting low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles through the laser driver.

[0015] The ratio calculation module is used to receive laser light and convert it into photoelectric signal, perform phase-locked amplification on the photoelectric signal, extract the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal respectively, and calculate the low-depth harmonic ratio and the high-depth harmonic ratio.

[0016] The initial inversion module is used to combine the environmental temperature and pressure parameters, substitute the low-depth harmonic ratio value and the high-depth harmonic ratio value into the concentration inversion model respectively, and obtain the low-depth initial concentration and the high-depth initial concentration accordingly.

[0017] An equivalent conversion module is used to calculate the deviation parameter based on the difference between the low-depth initial concentration and the high-depth initial concentration, and convert the deviation parameter into an equivalent absorption linewidth;

[0018] An adaptive feedback module is used to generate an adaptive modulation signal based on the equivalent absorption linewidth and feed it back to the laser driver, and use the adaptive modulation signal to rescan and obtain the adaptive harmonic ratio value.

[0019] The correction inversion module is used to substitute the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model to obtain the corrected concentration.

[0020] The result output module is used to apply a maximum rate of change limit to the corrected concentration and output the final carbon monoxide concentration.

[0021] The beneficial effects of this invention are as follows: Through a dual-modulation depth alternating scanning mechanism, this invention transforms the self-expanding covariance distribution state hidden under the non-uniform co-location conditions of the temperature field and carbon monoxide concentration field into an observable deviation parameter, overcoming the inherent theoretical defects caused by directly equating the beam penetration medium to a uniform field in traditional single-path monitoring; furthermore, by using this deviation parameter to dynamically calculate the equivalent absorption linewidth and constructing an adaptive modulation depth closed-loop feedback control mechanism, the calibration mapping model is directly corrected from the bottom layer, achieving error compensation; with a minimal hardware configuration of single-path, tomographic imaging, and extreme value change rate limiting to filter abnormal jumps, this invention fundamentally solves the measurement distortion problem caused by local gas enrichment and low-temperature quenching under harsh working conditions such as industrial combustion and high-temperature exhaust gas emissions, significantly improving the robustness of online monitoring and the absolute accuracy of the final output results under complex interference environments. Attached Figure Description

[0022] Figure 1 This is a flowchart of a laser carbon monoxide monitoring method according to the present invention;

[0023] Figure 2 This is a comparison chart of the dual-depth and adaptive correction concentration of the present invention. Detailed Implementation

[0024] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0025] Example 1: As Figure 1 As shown, a laser carbon monoxide monitoring method includes:

[0026] The laser driver alternately outputs low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles to drive the laser to scan the carbon monoxide absorption spectrum.

[0027] The laser is received and converted into a photoelectric signal. The photoelectric signal is amplified by phase lock, and the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal are extracted respectively. The ratio of low-depth harmonics to high-depth harmonics is calculated.

[0028] By combining the environmental temperature and pressure parameters, the low-depth harmonic ratio and the high-depth harmonic ratio are substituted into the concentration inversion model to obtain the low-depth initial concentration and the high-depth initial concentration.

[0029] The deviation parameter is calculated based on the difference between the initial concentration at low depth and the initial concentration at high depth, and the deviation parameter is converted into an equivalent absorption linewidth;

[0030] An adaptive modulation signal is generated based on the equivalent absorption linewidth and fed back to the laser driver. The adaptive modulation signal is then used to rescan and obtain the adaptive harmonic ratio.

[0031] Substituting the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model, the corrected concentration is obtained;

[0032] Apply a maximum rate of change limit to the corrected concentration and output the final carbon monoxide concentration.

[0033] Preferably, the laser is driven to scan the carbon monoxide absorption spectrum by alternately outputting low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles via a laser driver, including:

[0034] Obtain the center wavenumber of the absorption line of the carbon monoxide absorption spectrum, as well as the slow scan rate, scan time, and modulation angular frequency;

[0035] In the first and second scan cycles belonging to the adjacent scan cycles, the low modulation depth and the high modulation depth are respectively substituted into the instantaneous output wavenumber formula to obtain the first instantaneous output wavenumber as the low-depth modulation signal and the second instantaneous output wavenumber as the high-depth modulation signal.

[0036] The formula for the instantaneous output wavenumber is:

[0037]

[0038] In the formula, The status identifier is Instantaneous output wavenumber at time; This is a status identifier, representing the first scan cycle or the second scan cycle; Indicates the center wavenumber of the absorption line; This indicates the slow scan rate; Indicates the scan time; The status identifier is modulation depth at time, when Representing the first scan cycle For the low modulation depth, when Representing the second scan cycle The high modulation depth; Indicates the modulation angular frequency; This represents the cosine function.

[0039] The center wavenumber of the absorption line is the center wavenumber corresponding to the carbon monoxide absorption spectrum, and it is the core reference wavenumber for laser scanning. It can be acquired by querying a spectral library combined with laser frequency sweep calibration, or it can be determined through standard gas absorption experiments.

[0040] The slow scan rate is the rate at which the laser output wavenumber changes slowly over time; it is a parameter controlling the rate at which the laser sweeps across the carbon monoxide absorption spectrum. A preferred rate is 0.5 to 2 cm⁻¹ per second. This rate range ensures that the laser fully sweeps across the absorption spectrum while avoiding insufficient signal acquisition due to excessively fast scanning, thus meeting the real-time monitoring needs of industrial environments.

[0041] Scan time is the time it takes for the laser to output a modulated signal for a single scan; it is a time parameter characterizing a single frame scan. It can be acquired through the timing module of the laser driver or determined by the sampling timestamps of the data acquisition card.

[0042] The modulation angular frequency is the angular frequency at which the output wavenumber of the laser is modulated at high frequency. It is preferably 6.28 to 31.4 megaradians per second. This range corresponds to a modulation frequency of 1 to 5 megahertz, which not only meets the high-frequency modulation requirements of laser absorption spectroscopy technology, but also avoids hardware drive losses caused by excessively high frequencies, and is compatible with the response characteristics of conventional photodetectors.

[0043] The low modulation depth is the modulation depth parameter used to generate the low-depth modulation component in the first scan cycle. It is preferably 1.2 to 1.8 times the reference linewidth. This range ensures basic matching between the low-depth modulation signal and the absorption linewidth, giving the harmonic signal a good signal-to-noise ratio, while reserving differential response space for the high modulation depth signal.

[0044] High modulation depth is the modulation depth parameter used to generate high-depth modulation components in the second scan cycle. It is preferably 2.4 to 3.2 times the reference linewidth. This range allows the high-depth modulation signal to form a differential match with the absorption linewidth. When combined with a low modulation depth signal, it can effectively reveal hidden linewidth structural features without causing harmonic signal distortion due to excessive modulation depth.

[0045] The first instantaneous output wavenumber is the wavenumber output by the laser in real time during the first scanning cycle. It is a low-depth modulation signal formed by the superposition of the absorption line center wavenumber, the slow scanning component, and the low-depth modulation component.

[0046] The second instantaneous output wavenumber is the wavenumber output by the laser in real time during the second scanning cycle. It is a high-depth modulation signal formed by the superposition of the absorption line center wavenumber, the slow scanning component, and the high-depth modulation component.

[0047] In detail, for the same carbon monoxide absorption spectral line, alternating low- and high-depth modulation signals are output in adjacent scanning cycles. Specifically, two modulation depths are used to create differentiated responses to the same hidden linewidth shift. In practice, the laser output wavenumber is decomposed into the absorption line center wavenumber, slow scan component, and modulation component. The low- and high-depth modulation signals are constructed based on the same center wavenumber, slow scan rate, and modulation angular frequency, differing only in modulation depth. For example, the low-depth modulation signal is output in odd-numbered scanning cycles, and the high-depth modulation signal is output in even-numbered scanning cycles. Two adjacent frames form a set of matching frames to ensure the consistency of the scene between the two sets of scanning results. This design is not simply noise reduction, but rather transforms the hidden self-expanding covariance structure into an observable signal difference, laying the foundation for subsequent feature extraction. Furthermore, the same scanning benchmark can eliminate systematic errors caused by slow scanning and high-frequency modulation, ensuring that the signal difference is only caused by the difference in modulation depth.

[0048] In detail, the center wavenumber of the absorption line needs to be determined by calculating the optimal spectral index within a low-interference narrow spectral window in the 4.85-micron neighborhood of the carbon monoxide fundamental frequency. Specifically, the line selected should have the highest ratio of the carbon monoxide peak absorption cross-section to the sum of the equivalent absorption cross-sections of water, carbon dioxide, and instrument background, and possess parameters for self-broadening, external broadening, and usable temperature index, such as the P20 line. The slow scan rate and modulation angular frequency can be adjusted according to the laser model and photodetector response characteristics. In conventional industrial monitoring, the slow scan rate is set to 1 cm⁻¹ / s, and the modulation angular frequency to 12.56 MHz / s. The reference linewidth for low and high modulation depths is determined by experiments with zero gas or standard gas of known concentration. It was found that when the reference linewidth is 0.05 cm to the power of -1, the low modulation depth is 0.06 to 0.09 cm to the power of -1, and the high modulation depth is 0.12 to 0.16 cm to the power of -1. The duration of adjacent scanning cycles is set to the time it takes for the laser to scan a single absorption spectral line, which is generally 0.01 to 0.1 seconds. During execution, the odd-even cycle alternation rule is strictly followed, and the scanning frames are not matched across groups. The laser is a tunable semiconductor laser, and the output power is adapted to the optical path loss in the industrial environment, which is generally 5 to 20 milliwatts. The slow scanning component and the modulation component are directly linearly superimposed without weighting distinction. After superposition, they are directly used as the wavenumber control signal of the laser.

[0049] Preferably, the laser light is received and converted into a photoelectric signal, the photoelectric signal is amplified by lock-in, and the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal are extracted respectively. The ratio of the low-depth harmonic to the ratio of the high-depth harmonic are calculated, including:

[0050] Obtain the phase-locked integration window length and integration start time;

[0051] In the first and second scanning cycles that belong to the adjacent scanning cycles, the received photoelectric signals are substituted into the digital phase-locked loop formula according to the phase-locked integration window length, the integration start time, the modulation angular frequency and the scanning time, respectively, to obtain the first harmonic in-phase component, the first harmonic quadrature component, the second harmonic in-phase component and the second harmonic quadrature component corresponding to the first scanning cycle and the second scanning cycle.

[0052] Substitute the corresponding first-order harmonic in-phase component, first-order harmonic quadrature component, second-order harmonic in-phase component, and second-order harmonic quadrature component into the amplitude calculation formula to obtain the first-order harmonic amplitude and second-order harmonic amplitude corresponding to the first scanning period and the second scanning period, respectively.

[0053] Substitute the corresponding first-order harmonic amplitude and second-order harmonic amplitude into the ratio calculation formula to obtain the low-depth harmonic ratio corresponding to the first scanning period, and obtain the high-depth harmonic ratio corresponding to the second scanning period.

[0054] The digital phase-locked loop formula, the amplitude calculation formula, and the ratio calculation formula are respectively:

[0055]

[0056]

[0057]

[0058]

[0059] In the formula, The status identifier is And the order is The harmonic in-phase components, according to and The value of corresponds to the first harmonic in-phase component or the second harmonic in-phase component corresponding to the first scanning period or the second scanning period. The status identifier is And the order is The harmonic quadrature components, according to and The value of corresponds to the first or second harmonic quadrature component or the second harmonic quadrature component corresponding to the first or second scanning period. This indicates the length of the phase-locked integration window; Indicates the start time of the integration; The status identifier is The photoelectric signal at that time; Indicates the scan time; Indicates the harmonic order; Indicates the modulation angular frequency; Represents the cosine function; Represents the sine function; The status identifier is And the order is The harmonic amplitude, according to and The value of corresponds to the first harmonic amplitude or the second harmonic amplitude corresponding to the first scanning period or the second scanning period. The status identifier is The harmonic ratio at time, when Representing the first scan cycle For the low-depth harmonic ratio value, when Representing the second scan cycle The high-depth harmonic ratio value; This is a status identifier, representing either the first scan cycle or the second scan cycle.

[0060] The photoelectric signal is a voltage signal converted by a photodetector after receiving laser light passing through the monitoring area. It corresponds to scanning laser light with low-depth and high-depth modulation signals, respectively. It can be acquired using an infrared photodiode combined with a signal amplification circuit, and then converted into a digital signal by a high-speed data acquisition card.

[0061] The phase-locked integration window length is the length of the time window for digital phase-locked integration of photoelectric signals, and it is a core parameter for controlling the integration noise reduction effect and signal real-time performance. It is preferably 8 to 16 times the modulation period. This range ensures that the integration window covers an integer number of modulation periods, effectively reducing spectral leakage, while preventing a decrease in signal real-time performance due to an excessively long integration window, thus meeting the real-time requirements of industrial online monitoring.

[0062] The integration start time is the initial moment when phase-locked integration of the photoelectric signal begins. It is a parameter that ensures synchronization between the phase-locked integration and the high-frequency modulation signal. Preferably, it is the start time of the high-frequency modulation signal's period. This value ensures strict synchronization between the phase-locked integration process and the modulation signal, avoiding harmonic component extraction errors caused by phase deviation and guaranteeing the accuracy of harmonic characteristics.

[0063] The first harmonic in-phase component is the voltage component that is in phase with the first-order reference signal, obtained by phase-locked integration of the photoelectric signal.

[0064] The first-order harmonic quadrature component is a voltage component orthogonal to the first-order reference signal obtained after phase-locked integration of the photoelectric signal.

[0065] The second-order harmonic in-phase component is the voltage component that is in phase with the second-order reference signal, obtained by phase-locked integration of the photoelectric signal.

[0066] The second-order harmonic quadrature component is a voltage component orthogonal to the second-order reference signal obtained after phase-locked integration of the photoelectric signal.

[0067] The amplitude of a first harmonic is the intensity value of a first harmonic signal calculated from the in-phase component and the quadrature component of the first harmonic, and is used to reflect the magnitude of the first harmonic signal.

[0068] The second harmonic amplitude is the intensity value of the second harmonic signal calculated from the in-phase component and the quadrature component of the second harmonic, and is used to reflect the magnitude of the second harmonic signal.

[0069] The low-depth harmonic ratio is the ratio of the second-order harmonic amplitude to the first-order harmonic amplitude corresponding to a low-depth modulated signal.

[0070] The high-depth harmonic ratio is the ratio of the second-order harmonic amplitude to the first-order harmonic amplitude corresponding to a high-depth modulated signal.

[0071] In detail, the photoelectric signals corresponding to the low-depth and high-depth modulation signals are subjected to independent digital phase-locked amplification. To ensure the comparability of the processing results, the two sets of signals use the same phase-locked integration window length, integration start time, and phase-locked parameters. During processing, the amplitude of each harmonic is calculated by squaring and taking the square root to avoid the loss of information from a single in-phase or quadrature component. The harmonic ratio is then obtained by dividing the second harmonic amplitude by the first harmonic amplitude. This calculation method can effectively eliminate common-mode interference caused by laser intensity fluctuations. Furthermore, the obtained low-depth and high-depth harmonic ratios are not averaged or fused, but the numerical difference between the two is deliberately preserved. For example, the low-depth harmonic ratio is 0.22, and the high-depth harmonic ratio is 0.31. This numerical difference is the external signal manifestation of the self-expanding covariance structure in the combustion scenario, rather than being regarded as a measurement error to be eliminated.

[0072] In detail, the digital phase-locked amplification uses a dedicated phase-locked module built with a digital signal processing chip. During phase calibration, zero carbon monoxide standard gas is first introduced for laser scanning to adjust the reference phase of the phase-locked module, minimizing the value of the first-order harmonic quadrature component and maximizing the value of the first-order harmonic amplitude. Then, low-concentration carbon monoxide standard gas is introduced to confirm that the second-order harmonic signal phase is not reversed. The sampling rate of the photoelectric signal must be no less than 40 times the modulation frequency. For example, if the modulation frequency is 2 MHz, the sampling rate must be set to 80 MHz or higher to ensure the stability of digital phase-locking and envelope reconstruction. Three harmonic in-phase and quadrature components are extracted. The moving average method of sampling points is used for noise filtering to reduce the interference of random noise on harmonic components; the photoelectric conversion device is an infrared photodiode adapted to the 4.85-micron band, and the amplification factor of the signal amplification circuit is set to 100 to 1000 times to adapt to the detection of weak light signals after the laser passes through the monitoring area in the industrial site; if the harmonic amplitude calculation results in a negative value, the amplitude is directly set to zero and it is determined that the signal acquisition of that frame is abnormal and is not included in subsequent calculations; the integration start time is triggered by the modulation synchronization signal output by the laser driver, so that the integration process is strictly synchronized with the high-frequency modulation signal, eliminating the processing error caused by phase deviation.

[0073] Preferably, by combining environmental temperature and pressure parameters, the low-depth harmonic ratio and the high-depth harmonic ratio are substituted into the concentration inversion model to obtain the corresponding low-depth initial concentration and high-depth initial concentration, including:

[0074] Obtain the ambient temperature and ambient pressure from the ambient temperature and pressure parameters;

[0075] Obtain the zero-concentration baseline ratio, primary concentration calibration coefficient, and secondary concentration calibration coefficient corresponding to the first and second scan cycles belonging to the adjacent scan cycles under the ambient temperature and ambient pressure, respectively.

[0076] Substitute the corresponding zero-concentration baseline ratio, the first concentration calibration coefficient, the second concentration calibration coefficient, and the corresponding harmonic ratio into the concentration inversion model to obtain the low-depth initial concentration corresponding to the first scanning cycle and the high-depth initial concentration corresponding to the second scanning cycle, respectively.

[0077] The concentration inversion model is as follows:

[0078]

[0079] In the formula, The status identifier is The initial concentration at that time, when Representing the first scan cycle For the initial concentration at the low depth, when Representing the second scan cycle The initial concentration at the high depth; The status identifier is The initial concentration calibration coefficient at that time; The status identifier is The secondary concentration calibration coefficient at that time; The status identifier is The harmonic ratio at time, when Representing the first scan cycle For the low-depth harmonic ratio value, when Representing the second scan cycle The high-depth harmonic ratio value; The status identifier is The zero-concentration baseline ratio at that time; This is a status identifier, representing either the first scan cycle or the second scan cycle.

[0080] Ambient temperature refers to the overall gas temperature within the monitoring area. It can be acquired through contact temperature sensors located at the same position within the monitoring area, or through laser-assisted absorption line inversion.

[0081] Ambient pressure refers to the overall gas pressure within the monitoring area. It can be collected by pressure sensors located at the same location within the monitoring area, or obtained through pressure transmitter modules in industrial settings.

[0082] The zero-concentration baseline ratio is the ratio of the second-order to the first-order harmonic amplitudes under the corresponding temperature and pressure conditions in the absence of carbon monoxide. It serves as the baseline reference parameter for concentration inversion. It can be determined by repeatedly scanning and collecting data under the corresponding temperature and pressure conditions with a zero-carbon-monoxide standard gas, and then calculating the average value.

[0083] The primary concentration calibration coefficient is a calibration parameter characterizing the linear correlation between the harmonic ratio and carbon monoxide concentration. It can be obtained by scanning with standard gases of different concentrations at corresponding temperature and pressure, followed by data fitting and calculation.

[0084] The secondary concentration calibration coefficient is a calibration parameter characterizing the nonlinear relationship between the harmonic ratio and carbon monoxide concentration. It can be obtained by scanning with standard gases of different concentrations at corresponding temperature and pressure, followed by data fitting and calculation.

[0085] The initial concentration at low depth is the uncorrected carbon monoxide mole fraction obtained by substituting the low-depth harmonic ratio into the concentration inversion model.

[0086] The initial concentration at high depth is the uncorrected carbon monoxide mole fraction obtained by substituting the high-depth harmonic ratio into the concentration inversion model.

[0087] In detail, ambient temperature and pressure are used as pre-compensation conditions for concentration inversion. For different temperature and pressure combinations, pre-calibrated corresponding calibration coefficient sets are called. The low-depth and high-depth harmonic ratios use their own independent zero-concentration baseline ratios, first-order concentration calibration coefficients, and second-order concentration calibration coefficients, instead of sharing a single set of calibration coefficients. During inversion, a quadratic concentration inversion model containing first-order and second-order terms is used. The initial concentration is accurately solved by the root-finding formula, instead of the traditional single linear model. For example, at a temperature of 800 Kelvin and a pressure of 1 standard atmosphere, the low-depth harmonic ratio is substituted into the corresponding temperature and pressure calibration coefficients to obtain a low-depth initial concentration of 0.002, and the high-depth harmonic ratio is substituted into the independent calibration coefficients at the same temperature and pressure to obtain a high-depth initial concentration of 0.0025. The numerical difference between the two sets of initial concentrations is deliberately retained without averaging. This difference is the original data for subsequent identification of the self-expanding covariance structure, rather than directly fusing the harmonic ratios to output a single concentration value.

[0088] In detail, the establishment of calibration coefficients requires experiments to be conducted at different temperature and pressure nodes. Temperature nodes are set at 50 Kelvin intervals, and pressure nodes at 0.2 standard atmospheres intervals. At least six different concentrations of carbon monoxide standard gas are introduced at each temperature and pressure node. After scanning to obtain the harmonic ratios, a quadratic model is fitted to obtain the corresponding calibration coefficients, and a two-dimensional temperature and pressure calibration table is constructed. During actual operation, the calibration coefficients at the current temperature and pressure are obtained through bilinear interpolation. When the absolute value of the quadratic concentration calibration coefficient is less than 1 multiplied by 10 to the power of -6, it is determined that the quadratic term can be ignored, and the linear inversion model is switched to directly use harmonics. The initial concentration is obtained by dividing the difference between the ratio and the zero-concentration baseline by the first concentration calibration coefficient. Environmental temperature and pressure parameters are preferably collected by temperature and pressure sensors at the same location in the monitoring area to reduce system coupling complexity. After the calibration coefficient is established, intermediate concentration standard gas is introduced for verification. The relative error must be less than 3%, and the calibration coefficient is recalibrated every 3 months. When solving for the initial concentration using the root-finding formula, negative solutions are discarded, and only non-negative solutions are retained as valid initial concentrations. During the standard gas experiment, the gas flow rate is controlled at 500 ml / min. After the standard gas is introduced, it is stabilized for 5 minutes before laser scanning to ensure the stability and accuracy of the experimental data.

[0089] Preferably, the deviation parameter is calculated based on the difference between the initial concentration at low depth and the initial concentration at high depth, and the deviation parameter is converted into an equivalent absorption linewidth, including:

[0090] Substitute the high-depth initial concentration and the low-depth initial concentration into the parameter calculation formula to obtain the deviation parameter and the average initial concentration;

[0091] Obtain reference pressure, reference temperature, external broadening factor, self-broadening factor, temperature index, primary correction factor, and secondary correction factor;

[0092] Substituting the environmental pressure, the environmental temperature, the reference pressure, the reference temperature, the external broadening factor, the self-broadening factor, the temperature index, the average initial concentration, the deviation parameter, the first correction factor, and the second correction factor into the linewidth calculation formula, we obtain the uniform assumed linewidth and the equivalent absorption linewidth.

[0093] The parameter calculation formula and the line width calculation formula are respectively:

[0094]

[0095]

[0096]

[0097]

[0098] In the formula, Indicates the deviation parameter; This indicates the initial concentration at high depth; This indicates the initial concentration at the low depth; This represents the average initial concentration; This represents the uniform assumed line width; This indicates the environmental pressure; This indicates the reference pressure; This represents the external broadening factor; This represents the self-expansion coefficient; This indicates the reference temperature; This indicates the ambient temperature; This indicates the temperature index; This represents the equivalent absorption linewidth; This represents the first-order correction factor; This represents the second-order correction coefficient.

[0099] The deviation parameter is a scalar that characterizes the numerical difference between the initial concentration at low depth and the initial concentration at high depth.

[0100] The average initial concentration is the arithmetic mean of the initial concentrations at low depth and the initial concentrations at high depth.

[0101] The reference pressure is the base pressure selected in the calculation of spectral linewidth, and it is a parameter that eliminates the influence of pressure dimensions. The preferred value is 1 standard atmosphere, which is the universal pressure reference for spectroscopic research, unifies the pressure reference standard for linewidth calculation, and is compatible with the parameter system of existing spectral libraries.

[0102] The reference temperature is the standard temperature selected in the calculation of spectral linewidth, and it is a parameter to eliminate the influence of temperature on spectral line broadening. A preferred value is 296 Kelvin, which is the standard reference temperature for ambient conditions, a commonly used value in the field of spectroscopy, and can match the existing calibration standards for broadening coefficients.

[0103] The exogenous broadening factor is the spectral line broadening factor caused by collisions between carbon monoxide molecules and other gas molecules. It characterizes the baseline width of exogenous collisions under reference temperature and pressure conditions. It can be obtained by searching the latest spectral libraries or through actual measurements in gas collision experiments under high temperature and pressure.

[0104] The self-broadening coefficient is the spectral line broadening factor caused by collisions between carbon monoxide molecules. It characterizes the baseline width of self-collisions under reference temperature and pressure conditions. It can be obtained by searching the latest spectral libraries or through actual measurements in collision experiments with pure carbon monoxide gas.

[0105] The temperature index is an exponential parameter characterizing how spectral line broadening changes with temperature. It can be obtained by interpolating parameters of adjacent quantum number lines using a spectral library, or by fitting experimental data on spectral line broadening at different temperatures.

[0106] The primary correction factor is a coefficient used to linearly correct the uniformity assumption linewidth. It is a parameter that maps deviation parameters to a linear correction amount for linewidth. Preferably, it is between 0.5 and 2.0. This range can adapt to the self-expanding covariance characteristics of most industrial combustion scenarios, achieving accurate correction of the linear portion of the linewidth and avoiding under- or over-correction.

[0107] The secondary correction factor is a coefficient used to nonlinearly correct the uniform assumed linewidth. It maps the nonlinear characteristics of the deviation from the parameter to the linewidth correction amount. Preferably, it is between 0.1 and 0.8. This range can compensate for the nonlinear characteristics of the self-expanding covariance in complex combustion scenarios. Combined with the primary correction factor, it can achieve accurate calculation of the equivalent absorption linewidth.

[0108] The uniform assumption linewidth is the Lorentz linewidth of the carbon monoxide absorption spectrum calculated when the monitoring area is considered as a homogeneous medium. It is the baseline value for linewidth correction.

[0109] The equivalent absorption linewidth is the Lorentz linewidth of the carbon monoxide absorption spectrum, obtained after correction by deviation parameters and first and second correction coefficients, reflecting the actual non-uniform scenario.

[0110] In detail, the deviation parameter is obtained by dividing the difference between the initial concentration at low depth and the initial concentration at high depth by the sum of the two. This transforms the spatial self-broadening covariance structure, which cannot be directly measured under a single optical path, into a calculable scalar. First, the linewidth under the assumption of a homogeneous medium is calculated as a correction benchmark by combining the ambient temperature and pressure reference broadening coefficient and the average initial concentration. Then, the benchmark linewidth is nonlinearly corrected using the first and second correction coefficients to obtain the equivalent absorption linewidth reflecting the actual scenario. This process comprehensively considers the structural characteristics of carbon monoxide enrichment and low-temperature co-location in the combustion scenario, and does not rely solely on temperature and pressure to correct the linewidth. For example, if the initial concentration at low depth is 0.002 and the initial concentration at high depth is 0.0025, the deviation parameter is calculated to be 0.11. If the linewidth under the homogeneous assumption is 0.04 cm to the power of negative 1, and combining the first correction coefficient of 1.0 and the second correction coefficient of 0.3, the equivalent absorption linewidth can be calculated to the power of negative 1 of 0.045 cm, thus realizing the transformation of the hidden structure into a calculable linewidth parameter.

[0111] In detail, the external broadening factor and the self-broadening factor are preferably obtained using the latest measured parameters of the selected carbon monoxide absorption line. If no completely corresponding parameters are available, the parameters of adjacent quantum number lines are obtained through linear interpolation. The unit of the broadening factor (characterizing the reference equivalent linewidth under reference conditions) must be uniformly set to the negative first power of centimeters, temperature is in Kelvin, and pressure is in standard atmospheric pressure. The primary and secondary correction factors are calibrated using two tandem tubular heating furnaces of equal length as calibration chambers. In practice, carbon monoxide standard gas at room temperature and low concentration is introduced into the first calibration chamber, while carbon monoxide standard gas at high temperature and high concentration is introduced into the second calibration chamber to simulate the local enrichment and non-uniform working conditions of actual industrial sites. Then, the theoretical equivalent absorption linewidth under the current setting of the two chambers is calculated using known spectral parameter database data and used as the true reference value. At the same time, the monitoring system outputs low-depth and high-depth modulation signals to scan the tandem calibration chamber and record the system data. The initial concentrations at low and high depths were measured, and the deviation parameter was calculated accordingly. The temperature and concentration settings of the two calibration chambers were continuously varied to obtain at least twenty pairs of corresponding theoretical equivalent absorption linewidths and measured deviation parameter data. Finally, the theoretical equivalent absorption linewidth was divided by the uniform assumed linewidth to obtain the linewidth correction ratio. Using the deviation parameter from the aforementioned twenty data pairs as the independent variable and the difference between the corresponding linewidth correction ratio and one as the dependent variable, a polynomial curve was fitted using the least squares method. The coefficient of the first term obtained from the fitting was used as the first-order correction coefficient of the system, and the coefficient of the second-order term obtained from the fitting was used as the second-order correction coefficient of the system. When calculating the deviation parameter, if the sum of the initial concentrations at low and high depths was less than 1% of the full scale, the deviation parameter was directly set to zero to avoid numerical explosion due to an excessively small denominator. In the linewidth calculation, the ambient pressure needed to be converted to a value in standard atmospheres, and the ambient temperature was directly expressed in Kelvin to ensure dimensional consistency in the calculation process. The accuracy of the equivalent absorption linewidth is verified by passing a standard gas with known temperature and concentration distribution. The relative error between the calculated and measured values ​​must be less than 5%. The effective range of the temperature index is 300 to 2000 Kelvin. When the temperature exceeds this range, the temperature index of the adjacent temperature range is used after linear interpolation correction.

[0112] Preferably, generating an adaptive modulation signal based on the equivalent absorption linewidth and feeding it back to the laser driver, then rescanning using the adaptive modulation signal to obtain the adaptive harmonic ratio, includes:

[0113] Obtain the dimensionless matching coefficients;

[0114] Substitute the equivalent absorption linewidth and the dimensionless matching coefficient into the modulation depth update formula to obtain the adaptive modulation depth.

[0115] Substituting the adaptive modulation depth, the absorption line center wavenumber, the slow scan rate, the scan time, and the modulation angular frequency into the modified output wavenumber formula, the third instantaneous output wavenumber is obtained as the adaptive modulation signal;

[0116] The laser is driven to rescan using the adaptive modulation signal to obtain the first harmonic amplitude and second harmonic amplitude corresponding to the rescan, and the second harmonic amplitude corresponding to the rescan is divided by the first harmonic amplitude to obtain the adaptive harmonic ratio.

[0117] The modulation depth update formula and the corrected output wavenumber formula are respectively:

[0118]

[0119]

[0120] In the formula, This indicates the adaptive modulation depth; This represents the dimensionless matching coefficient; This represents the equivalent absorption linewidth; This indicates the third instantaneous output wavenumber; Indicates the center wavenumber of the absorption line; This indicates the slow scan rate; Indicates the scan time; Indicates the modulation angular frequency; This represents the cosine function.

[0121] The dimensionless matching coefficient is a dimensionless coefficient that establishes a linear relationship between the equivalent absorption linewidth and the adaptive modulation depth. It is the core parameter for achieving precise matching between modulation depth and linewidth. The preferred value is 1.0 to 2.0. This range ensures the optimal matching relationship between modulation depth and equivalent absorption linewidth, maximizing the concentration sensitivity of harmonic signals and ensuring that the harmonic ratio remains monotonic across the entire carbon monoxide monitoring range, thus adapting to the dynamic changes in linewidth in industrial combustion scenarios.

[0122] The adaptive modulation depth is a modulation depth parameter obtained by multiplying the dimensionless matching coefficient by the equivalent absorption linewidth.

[0123] The third instantaneous output wavenumber is the real-time wavenumber formed by the superposition of the absorption line center wavenumber, the slow scan component, and the adaptive modulation component. It is the adaptive modulation signal output by the laser driver.

[0124] The adaptive harmonic ratio is the ratio of the second-order harmonic amplitude to the first-order harmonic amplitude calculated after the laser is re-scanned by the adaptive modulation signal.

[0125] In detail, the equivalent absorption linewidth is used as the basis for calculating the adaptive modulation depth. The modulation depth is transformed from a traditional fixed instrument parameter into an adaptive parameter driven by the structural characteristics of the combustion scene. A linear relationship between the two is established through a dimensionless matching coefficient, which simplifies the calculation while ensuring the rationality of the matching. The adaptive modulation signal is constructed entirely based on the same absorption linecenter wavenumber, slow scan rate, and modulation angular frequency as the low- and high-depth modulation signals. Only the modulation depth parameter is updated to ensure the continuity of the scan and the comparability of the results. This parameter is fed back to the laser driver to drive the laser to rescan. The adaptive harmonic ratio is then obtained using the same calculation method as the low- and high-depth harmonic ratio, forming a closed-loop feedback control mechanism for the modulation depth. For example, if the equivalent absorption linewidth is 0.045 cm to the power of -1 and the dimensionless matching coefficient is 1.5, the calculated adaptive modulation depth is 0.0675 cm to the power of -1. The generated adaptive modulation signal can accurately match the actual absorption linewidth, avoiding harmonic signal distortion and concentration inversion errors caused by the drift between the fixed modulation depth and the actual linewidth.

[0126] In detail, the determination of the dimensionless matching coefficient needs to be carried out independently at five to eight representative equivalent absorption linewidths. At any representative equivalent absorption linewidth, the dimensionless matching coefficient is sequentially increased from 0.5 to 3.0 in fixed steps of 0.1. For each set dimensionless matching coefficient value, two different concentrations of carbon monoxide standard gas at 20% and 80% of full scale are respectively introduced into the calibration chamber, and the two adaptive harmonic ratio values ​​output by the monitoring system under adaptive signal scanning are recorded respectively. Then, the absolute value of the difference between these two adaptive harmonic ratio values ​​is divided by... The quotient obtained by the concentration difference between the two standard gases is used as the concentration sensitivity corresponding to the dimensionless matching coefficient. Simultaneously, ten concentration test points are uniformly selected within the full scale range, and standard gases are sequentially introduced for testing. If the adaptive harmonic ratio measured by the system maintains a strictly increasing value with increasing gas concentration, the dimensionless matching coefficient is deemed to meet the monotonicity requirement. Finally, from all dimensionless matching coefficients that have passed monotonicity verification, the one with the largest corresponding concentration sensitivity value is selected as the dimensionless matching coefficient ultimately used by the system under the equivalent absorption linewidth. The timing sequence for switching from dual-modulation depth scanning to adaptive modulation depth scanning is as follows: low-depth frames and high-depth frames form a group of scan frames. After completing one group, the deviation parameter and equivalent absorption linewidth are immediately calculated, and the next frame directly switches to the adaptive modulation frame. This process is repeated cyclically in the order of low-depth frames, high-depth frames, and adaptive modulation frames. Rescanning the adaptive modulation signal is a single-frame scan. The valid frame determination criterion is that the signal-to-noise ratio (SNR) of the harmonic signal is greater than 10. If the SNR is lower than this value, the scan is deemed invalid, and the adaptive scan is repeated. If the adaptive harmonic ratio deviates from the average of the previous set of low and high depth harmonic ratios by more than 50%, it is considered an anomaly, and the adaptive harmonic ratio of the previous frame is directly used in subsequent calculations. The laser driver's response speed to the adaptive modulation signal is no less than 1 microsecond, and the output accuracy of the modulation depth is no less than 0.001 cm to the power of -1, adapting to fine changes in linewidth. The update frequency of the equivalent absorption linewidth is synchronized with the scanning frame cycle, updating once after each set of low and high depth frames is scanned, ensuring that the adaptive modulation depth can match linewidth changes in real time.

[0127] Preferably, the adaptive harmonic ratio and the equivalent absorption linewidth are substituted into the concentration inversion model to obtain the corrected concentration, including:

[0128] Based on the ambient temperature, the ambient pressure, and the equivalent absorption linewidth, the corrected zero-concentration baseline ratio, the corrected primary concentration calibration coefficient, and the corrected secondary concentration calibration coefficient are obtained.

[0129] Substitute the adaptive harmonic ratio, the corrected zero-concentration baseline ratio, the corrected first-order concentration calibration coefficient, and the corrected second-order concentration calibration coefficient into the concentration inversion model to obtain the corrected concentration;

[0130] The concentration inversion model is as follows:

[0131]

[0132] In the formula, This indicates the corrected concentration; This indicates the correction coefficient for the first concentration calibration. This represents the corrected secondary concentration calibration coefficient; This represents the adaptive harmonic ratio value; This represents the corrected zero-concentration baseline ratio.

[0133] The corrected zero-concentration baseline ratio, after considering the influence of equivalent absorption linewidth, is the ratio of the second-order harmonic amplitude to the first-order harmonic amplitude under ambient temperature and pressure conditions in the absence of carbon monoxide. It serves as the baseline reference parameter for corrected concentration inversion. Data can be collected by laser scanning with zero-carbon monoxide standard gas at different temperature and pressure linewidth nodes, fitted to construct a three-dimensional calibration table, and then obtained through interpolation.

[0134] The primary concentration calibration coefficient, after considering the influence of the equivalent absorption linewidth, is a calibration parameter characterizing the linear correlation between the harmonic ratio and carbon monoxide concentration. It can be obtained by scanning and collecting data with standard carbon monoxide gas of different concentrations at different temperature, pressure, and linewidth nodes, performing data fitting calculations, and then constructing a three-dimensional calibration table for interpolation.

[0135] The corrected secondary concentration calibration coefficient is a calibration parameter that characterizes the nonlinear relationship between the harmonic ratio and carbon monoxide concentration, taking into account the influence of the equivalent absorption linewidth. It can be obtained by scanning and collecting data with standard carbon monoxide gas of different concentrations at different temperature, pressure, and linewidth nodes, performing data fitting calculations, and then constructing a three-dimensional calibration table for interpolation.

[0136] The corrected concentration is the carbon monoxide mole fraction obtained by substituting the adaptive harmonic ratio and equivalent absorption linewidth into the corrected concentration inversion model. It is the core concentration parameter after correction by the self-expanding covariance structure characteristics.

[0137] In detail, the equivalent absorption linewidth is incorporated into the calibration dimension of concentration inversion, extending the traditional two-dimensional calibration model of ambient temperature and ambient pressure into a three-dimensional calibration model based on the equivalent absorption linewidth of ambient temperature and ambient pressure. The primary and secondary concentration calibration coefficients are then corrected based on the corrected zero-concentration baseline ratio obtained from this model. The influence of the self-expanding covariance structure is directly embedded into the concentration inversion model itself, rather than performing a simple posterior numerical correction on the initial concentration. Furthermore, when inverting the adaptive harmonic ratio, the same secondary inversion model framework as the initial concentration is used. By simply replacing the correction coefficients obtained from the three-dimensional calibration, the consistency of the inversion method is ensured and systematic errors are reduced. For example, under the operating conditions of an ambient temperature of 800 Kelvin and an ambient pressure of 1 standard atmosphere with an equivalent absorption linewidth of 0.045 cm to the power of -1, the corresponding correction calibration coefficients are obtained from the three-dimensional calibration table through trilinear interpolation. After substituting the adaptive harmonic ratio of 0.28, the corrected concentration is 0.0023. This process realizes the full-link correlation of the inversion of the concentration of the structural features of the scanned signal, making the inversion results more consistent with the real concentration of the actual non-uniform combustion scenario.

[0138] In detail, when establishing the three-dimensional calibration model for the corrected calibration coefficients, temperature nodes were set at 50 Kelvin intervals, pressure nodes at 0.2 standard atmospheres intervals, and equivalent absorption linewidth nodes at negative first-order intervals of 0.01 cm. At least six different concentrations of carbon monoxide standard gas were introduced under each three-dimensional node. After scanning to obtain the harmonic ratios, a quadratic model was fitted to obtain the corresponding corrected calibration coefficients. All node data were integrated to construct a three-dimensional calibration table. During actual operation, trilinear interpolation was used to obtain the corrected calibration coefficients for the equivalent absorption linewidth of the current ambient temperature, ambient pressure, and ambient pressure. Eight three-dimensional grid points around the current point were selected, and the interpolation at each point was calculated. The weights are summed to obtain the final correction coefficient value; after the three-dimensional calibration table is established, a standard gas of intermediate concentration is introduced for verification, and the relative error must be less than 3%, and the three-dimensional calibration table is recalibrated every 6 months; if the equivalent absorption linewidth exceeds the linewidth range of the calibration table, the correction coefficients of the calibration table boundary nodes are used for linear extrapolation to obtain the required coefficients; when the correction concentration is inverted by solving the root formula, negative solutions are directly discarded, and only non-negative solutions are retained as the effective correction concentration; during the three-dimensional calibration experiment, the standard gas flow rate is controlled at 500 ml / min, and after the standard gas is introduced, it is stabilized for 5 minutes before laser scanning is performed to ensure the stability and accuracy of the collected data.

[0139] Preferably, applying a maximum rate of change limit to the corrected concentration and outputting the final carbon monoxide concentration includes:

[0140] Obtain the output mole fraction, maximum rate of change, and sampling interval of the previous time step;

[0141] Substitute the corrected concentration, the output mole fraction at the previous moment, the maximum rate of change, and the sampling interval into the physical constraint calculation formula to obtain the current error term and the output mole fraction at this moment;

[0142] Substitute the output mole fraction at this moment into the unit conversion formula to obtain the final carbon monoxide concentration;

[0143] The physical constraint calculation formula and the unit conversion formula are as follows:

[0144]

[0145]

[0146]

[0147] In the formula, Indicates the discrete sampling sequence number; This indicates the current error term; This indicates the corrected concentration; This indicates the output mole fraction at the previous time step; This indicates the output mole fraction at this moment; Represents a symbolic function; Describes the minimum value function; This represents the absolute value of the current error term; This represents the maximum rate of change; Indicates the sampling interval; This indicates the final carbon monoxide concentration.

[0148] The output mole fraction at the previous moment is the carbon monoxide mole fraction output after the maximum rate of change limit in the previous sampling period, and it serves as the benchmark reference parameter for this concentration constraint. Historical data acquisition of the final output mole fraction from the previous sampling period can be retrieved through the data storage module.

[0149] The maximum rate of change is the maximum permissible change in the mole fraction of carbon monoxide per unit time in an industrial combustion scenario. It is preferably between 0.0001 and 0.001 per second. This range adapts to the actual physical rate of change of carbon monoxide in the combustion chamber and industrial furnace combustion channel. A value that is too large cannot suppress non-physical jumps caused by noise or abnormal frames, while a value that is too small will excessively restrict the actual concentration change, thus adapting to the actual needs of industrial online monitoring.

[0150] The sampling interval is the time interval between two consecutive carbon monoxide concentration outputs, and it is a parameter that correlates the maximum rate of change with the actual change. It is preferably 0.01 to 0.1 seconds, a range that matches the laser scanning cycle, ensuring real-time industrial monitoring while avoiding frequent concentration fluctuations caused by excessively short sampling intervals, and maintaining consistency with the scan frame rhythm.

[0151] The current error term is the difference between the current corrected concentration and the output mole fraction at the previous moment, and it is a calculated parameter characterizing the magnitude of concentration change.

[0152] The output mole fraction at this moment is the carbon monoxide mole fraction obtained after limiting the maximum rate of change, and it is the basic parameter for the final concentration conversion.

[0153] The final carbon monoxide concentration is a value obtained by converting the output mole fraction at this moment into parts per million (ppm) concentration units, and is the final monitoring result for industrial applications.

[0154] In detail, instead of using traditional signal smoothing filtering, a maximum rate of change limit is imposed on the corrected concentration based on the physical laws of industrial combustion scenarios. Specifically, the output mole fraction of the previous moment is used as the constraint benchmark. The difference between the corrected concentration and this benchmark is calculated to obtain the current error term. Then, the maximum rate of change is multiplied by the sampling interval to obtain the maximum allowable change. The sign function is used to determine the direction of concentration change, and the smaller value between the absolute value of the error and the maximum allowable change is selected as the actual change amplitude by combining the minimum value function. Finally, the output mole fraction of the current moment is obtained. The physical constraint of the mole fraction is completed first before converting it to a concentration in parts per million (ppm), avoiding data errors after unit conversion. For example, if the output mole fraction of the previous moment is 0.0023, the corrected concentration is 0.0026, the sampling interval is 0.05 seconds, the maximum rate of change is 0.002 per second, and the calculated maximum allowable change is 0.0001. At this time, the actual change amplitude is taken as 0.0001, and the output mole fraction of the current moment is 0.0024. This effectively avoids non-physical jumps in the corrected concentration caused by optical anomalies, making the output result more consistent with the actual process of combustion scenarios.

[0155] In detail, the maximum rate of change is primarily determined by calculating the ratio of the maximum injected molar flow rate of carbon monoxide within the monitored volume to the total molar amount of gas. If this data is unavailable on-site, historical monitoring data from the past 30 days is used, extracting the rate of change of the fastest rising edge of concentration as an estimate. The sampling interval is kept completely consistent with the laser scanning cycle. For example, if the laser single scan cycle is 0.05 seconds, the sampling interval is directly set to 0.05 seconds to ensure that the concentration output is synchronized with the scanning cycle. The initial output molar fraction is set according to the following rules: the initial low-depth concentration of the first frame is used during the first monitoring, and the molar fraction value of the most recent zero-point calibration is used after the device restarts or zero-point calibration. The rule for determining outlier frames is that if the adaptive harmonic ratio deviates from the median of the harmonic ratios of the most recent 10 frames by more than 3 times the standard deviation, the frame is determined to be an outlier frame, and the current corrected concentration is directly set to the output molar fraction of the previous moment. The final carbon monoxide concentration value retains 3 significant digits, and the data refresh frequency is synchronized with the sampling interval without additional delay. When the slope of concentration change exceeds 1.5 times the maximum rate of change for 5 consecutive frames, it is determined that a real abrupt change has occurred in the combustion scene. The maximum rate of change is temporarily increased to twice the original value. After the slope returns to the maximum rate of change range for 3 consecutive frames, the original value is adjusted back. When converting mole fraction to parts per million concentration, it is directly multiplied by 1,000,000. Rounding is only performed according to the significant figures requirement in the final output. The original values ​​are retained in the intermediate calculation process to avoid truncation errors.

[0156] like Figure 2 As shown, Figure 2 The results show that the reference concentration and the adaptively corrected concentration curves almost completely overlap, with the overall concentration steadily increasing from about 15 ppm to about 53 ppm, reflecting the true concentration trend. The initial concentration at low depths is consistently significantly lower than the reference concentration, slowly increasing from about 6 ppm to about 37 ppm, indicating a systematic underestimation bias. The initial concentration at high depths is consistently significantly higher than the reference concentration, continuously increasing from about 23 ppm to about 73 ppm, indicating a systematic overestimation bias. In summary, the adaptive correction method can effectively eliminate the systematic error of the initial concentration at dual-modulation depths, significantly improving the accuracy of carbon monoxide concentration monitoring.

[0157] Example 2: A laser carbon monoxide monitoring system, comprising the steps of any one of the laser carbon monoxide monitoring methods described herein, including:

[0158] The alternating scanning module is used to drive the laser to scan the carbon monoxide absorption spectrum by alternately outputting low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles through the laser driver.

[0159] The ratio calculation module is used to receive laser light and convert it into photoelectric signal, perform phase-locked amplification on the photoelectric signal, extract the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal respectively, and calculate the low-depth harmonic ratio and the high-depth harmonic ratio.

[0160] The initial inversion module is used to combine the environmental temperature and pressure parameters, substitute the low-depth harmonic ratio value and the high-depth harmonic ratio value into the concentration inversion model respectively, and obtain the low-depth initial concentration and the high-depth initial concentration accordingly.

[0161] An equivalent conversion module is used to calculate the deviation parameter based on the difference between the low-depth initial concentration and the high-depth initial concentration, and convert the deviation parameter into an equivalent absorption linewidth;

[0162] An adaptive feedback module is used to generate an adaptive modulation signal based on the equivalent absorption linewidth and feed it back to the laser driver, and use the adaptive modulation signal to rescan and obtain the adaptive harmonic ratio value.

[0163] The correction inversion module is used to substitute the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model to obtain the corrected concentration.

[0164] The result output module is used to apply a maximum rate of change limit to the corrected concentration and output the final carbon monoxide concentration.

[0165] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0166] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A laser carbon monoxide monitoring method, characterized in that, include: The laser driver alternately outputs low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles to drive the laser to scan the carbon monoxide absorption spectrum. The laser is received and converted into a photoelectric signal. The photoelectric signal is amplified by phase lock, and the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal are extracted respectively. The ratio of low-depth harmonics to high-depth harmonics is calculated. By combining the environmental temperature and pressure parameters, the low-depth harmonic ratio and the high-depth harmonic ratio are substituted into the concentration inversion model to obtain the low-depth initial concentration and the high-depth initial concentration. The deviation parameter is calculated based on the difference between the initial concentration at low depth and the initial concentration at high depth, and the deviation parameter is converted into an equivalent absorption linewidth; An adaptive modulation signal is generated based on the equivalent absorption linewidth and fed back to the laser driver. The adaptive modulation signal is then used to rescan and obtain the adaptive harmonic ratio. Substituting the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model, the corrected concentration is obtained; Apply a maximum rate of change limit to the corrected concentration to output the final carbon monoxide concentration.

2. The laser carbon monoxide monitoring method according to claim 1, characterized in that, The laser is driven to scan the carbon monoxide absorption spectrum by alternately outputting low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles via a laser driver, including: Obtain the center wavenumber of the absorption line of the carbon monoxide absorption spectrum, as well as the slow scan rate, scan time, and modulation angular frequency; The product of the slow scan rate and the scan time is taken as the slow scan component; The cosine function corresponding to the product of the modulation angular frequency and the scanning time is used as the cosine modulation basis. In the first scan cycle belonging to the adjacent scan cycle, the low modulation depth is multiplied by the cosine modulation base to obtain the low-depth modulation component, and the absorption line center wavenumber, the slow scan component and the low-depth modulation component are added to obtain the first instantaneous output wavenumber, which is used as the low-depth modulation signal. In the second scan cycle belonging to the adjacent scan cycle, the high modulation depth is multiplied by the cosine modulation base to obtain the high-depth modulation component, and the absorption line center wavenumber, the slow scan component and the high-depth modulation component are added to obtain the second instantaneous output wavenumber, which is used as the high-depth modulation signal.

3. The laser carbon monoxide monitoring method according to claim 2, characterized in that, The laser beam is received and converted into a photoelectric signal. The photoelectric signal is then amplified using a lock-in method. The first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal are extracted respectively. The ratio of the low-depth harmonic to the ratio of the high-depth harmonic are calculated, including: Obtain the first harmonic in-phase component, the first harmonic quadrature component, the second harmonic in-phase component, and the second harmonic quadrature component corresponding to the first scanning cycle. The square root of the sum of the square of the in-phase component of the first harmonic corresponding to the first scanning period and the square of the quadrature component of the first harmonic is taken to obtain the amplitude of the first harmonic corresponding to the first scanning period. The square root of the sum of the square of the in-phase component of the second harmonic corresponding to the first scanning period and the square of the quadrature component of the second harmonic is taken to obtain the amplitude of the second harmonic corresponding to the first scanning period. Divide the second harmonic amplitude corresponding to the first scan cycle by the first harmonic amplitude corresponding to the first scan cycle to obtain the low-depth harmonic ratio value. Using the same steps, the first harmonic amplitude and the second harmonic amplitude corresponding to the second scanning period are obtained, and the second harmonic amplitude corresponding to the second scanning period is divided by the first harmonic amplitude corresponding to the second scanning period to obtain the high-depth harmonic ratio.

4. The laser carbon monoxide monitoring method according to claim 3, characterized in that, Combining environmental temperature and pressure parameters, the low-depth harmonic ratio and the high-depth harmonic ratio are substituted into the concentration inversion model to obtain the corresponding low-depth initial concentration and high-depth initial concentration, including: Obtain the ambient temperature and ambient pressure from the ambient temperature and pressure parameters; Obtain the zero-concentration baseline ratio, primary concentration calibration coefficient, and secondary concentration calibration coefficient of the first scan cycle under the ambient temperature and ambient pressure. Subtract the zero-concentration baseline ratio from the low-depth harmonic ratio value and the ratio difference value of the first scan cycle to obtain the ratio difference value; Multiply constant four by the secondary concentration calibration coefficient of the first scan cycle and the ratio difference to obtain the compensation term; The square root of the sum of the square of the concentration calibration coefficient of the first scan cycle and the compensation term is obtained to obtain the root value term. Subtract the concentration calibration coefficient of the first scan cycle from the root value term to obtain the numerator term; Divide the molecule by twice the secondary concentration calibration coefficient of the first scan cycle to obtain the low-depth initial concentration; Using the same steps, the high-depth initial concentration is obtained by combining the zero-concentration baseline ratio under the ambient temperature and ambient pressure, the primary concentration calibration coefficient and the secondary concentration calibration coefficient, and the high-depth harmonic ratio value during the second scan cycle.

5. The laser carbon monoxide monitoring method according to claim 4, characterized in that, The deviation parameter is calculated based on the difference between the initial concentration at low depth and the initial concentration at high depth, and the deviation parameter is converted into an equivalent absorption linewidth, including: Obtain the initial difference and initial sum of the high-depth initial concentration and the low-depth initial concentration; Divide the initial difference by the initial sum to obtain the deviation parameter; divide the initial sum by two to obtain the average initial concentration; Obtain reference pressure, reference temperature, external broadening factor, self-broadening factor, temperature index, primary correction factor, and secondary correction factor; The uniform assumed linewidth is calculated based on the environmental pressure, the reference pressure, the external broadening factor, the self-broadening factor, the average initial concentration, the reference temperature, the environmental temperature, and the temperature index. The equivalent absorption linewidth is obtained by correcting the uniform assumption linewidth using the deviation parameter, the first correction coefficient, and the second correction coefficient.

6. The laser carbon monoxide monitoring method according to claim 5, characterized in that, An adaptive modulation signal is generated based on the equivalent absorption linewidth and fed back to the laser driver. The adaptive modulation signal is used to rescan and obtain the adaptive harmonic ratio, including: Obtain the dimensionless matching coefficients; Multiply the dimensionless matching coefficient by the equivalent absorption linewidth to obtain the adaptive modulation depth; The adaptive modulation component is obtained by multiplying the adaptive modulation depth by the cosine modulation base. The third instantaneous output wavenumber is obtained by adding the absorption line center wavenumber, the slow scan component, and the adaptive modulation component, and is used as the adaptive modulation signal. The laser is driven to rescan using the adaptive modulation signal, and the first-order harmonic amplitude and second-order harmonic amplitude corresponding to the rescan are obtained. The adaptive harmonic ratio is obtained by dividing the second harmonic amplitude corresponding to the rescan by the first harmonic amplitude corresponding to the rescan.

7. The laser carbon monoxide monitoring method according to claim 6, characterized in that, Substituting the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model yields the corrected concentration, including: Based on the ambient temperature, the ambient pressure, and the equivalent absorption linewidth, the corrected zero-concentration baseline ratio, the corrected primary concentration calibration coefficient, and the corrected secondary concentration calibration coefficient are obtained. Subtract the corrected zero-concentration baseline ratio from the adaptive harmonic ratio to obtain the adaptive ratio difference; Multiply the constant four by the corrected secondary concentration calibration coefficient and the difference in the adaptive ratio to obtain the correction compensation term; The square root of the square of the correction first concentration calibration coefficient and the correction compensation term is obtained to obtain the correction root value term; Subtract the first-order concentration calibration coefficient from the corrected root value term to obtain the corrected numerator term; The corrected concentration is obtained by dividing the corrected numerator by twice the corrected secondary concentration calibration coefficient.

8. The laser carbon monoxide monitoring method according to claim 7, characterized in that, Applying a maximum rate of change limit to the corrected concentration, the final carbon monoxide concentration is output, including: Obtain the output mole fraction, maximum rate of change, and sampling interval of the previous time step; Subtract the previous output mole fraction from the corrected concentration to obtain the current error term; Multiply the maximum rate of change by the sampling interval to obtain the maximum allowable change. Obtain the absolute value and sign value of the current error term; The actual change range is obtained by selecting the smaller value between the absolute value of the error and the maximum allowable change using a minimum value function. Multiply the error sign value by the actual change magnitude to obtain the signed change amount; Add the mole fraction output at the previous moment to the signed change to obtain the mole fraction output at the current moment; Multiply the current moment output mole fraction by one million to obtain the final carbon monoxide concentration.

9. A laser carbon monoxide monitoring system, implementing the steps of a laser carbon monoxide monitoring method as described in any one of claims 1-8, characterized in that, include: The alternating scanning module is used to drive the laser to scan the carbon monoxide absorption spectrum by alternately outputting low-depth modulation signals and high-depth modulation signals in adjacent scanning cycles through the laser driver. The ratio calculation module is used to receive laser light and convert it into photoelectric signal, perform phase-locked amplification on the photoelectric signal, extract the first and second harmonics corresponding to the low-depth modulation signal and the high-depth modulation signal respectively, and calculate the low-depth harmonic ratio and the high-depth harmonic ratio. The initial inversion module is used to combine the environmental temperature and pressure parameters, substitute the low-depth harmonic ratio value and the high-depth harmonic ratio value into the concentration inversion model respectively, and obtain the low-depth initial concentration and the high-depth initial concentration accordingly. An equivalent conversion module is used to calculate the deviation parameter based on the difference between the low-depth initial concentration and the high-depth initial concentration, and convert the deviation parameter into an equivalent absorption linewidth; An adaptive feedback module is used to generate an adaptive modulation signal based on the equivalent absorption linewidth and feed it back to the laser driver, and use the adaptive modulation signal to rescan and obtain the adaptive harmonic ratio value. The correction inversion module is used to substitute the adaptive harmonic ratio and the equivalent absorption linewidth into the concentration inversion model to obtain the corrected concentration. The result output module is used to apply a maximum rate of change limit to the corrected concentration and output the final carbon monoxide concentration.

Images