A liquid chromatograph signal acquisition noise suppression method and system

Abstract

The present application relates to the technical field of liquid chromatography signal processing, and discloses a liquid chromatograph acquisition signal noise suppression method and system, which comprises the following steps: acquiring an ultraviolet acquisition signal, a gradient organic phase ratio, an organic phase pump operation state parameter and a mixer volume; screening candidate component disturbance events and dividing gradient component intervals, determining event response delays and response windows; screening reference response windows, determining trifluoroacetic acid pair disturbance templates and template matching thresholds; and obtaining trifluoroacetic acid ripple components according to the amplitude ratio of effective response segments and corresponding side templates and deducting the trifluoroacetic acid ripple components from the ultraviolet acquisition signal. The present application can reduce the interference of trifluoroacetic acid related baseline fluctuation on the ultraviolet acquisition signal and reduce the false deduction of sample response.

Description

Technical Field

[0001] This invention relates to the field of liquid chromatography signal processing technology, specifically a method and system for suppressing noise in liquid chromatography signal acquisition. Background Technology

[0002] During gradient elution in liquid chromatography, changes in mobile phase composition, pumping fluctuations, and detector baseline drift can all contribute to noise in the UV acquisition signal. Existing techniques typically employ frequency domain filtering, baseline correction, blank spectrum subtraction, or improved mobile phase mixing to reduce noise.

[0003] However, in reverse-phase binary high-pressure gradient elution of trifluoroacetic acid in an aqueous phase and a trifluoroacetic acid-acetonitrile phase, minute compositional fluctuations are accompanied by transient redistribution of trifluoroacetic acid between the reverse-phase stationary and mobile phases, resulting in UV baseline fluctuations that vary with the gradient state. When these baseline fluctuations locally overlap with the sample response, existing methods struggle to distinguish between the two types of signals, easily leading to noise residue or synchronous attenuation of the sample response. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for suppressing noise in liquid chromatography signals, in order to solve the problem that it is difficult to accurately identify and suppress the baseline fluctuations of trifluoroacetic acid and the sample response when they locally overlap during reversed-phase binary high-pressure gradient elution of a mobile phase containing trifluoroacetic acid.

[0005] To achieve the above objectives, on the one hand, the present invention provides a method for suppressing noise in liquid chromatography acquisition signals, applicable to liquid chromatography analysis scenarios using a reversed-phase binary high-pressure gradient elution with an aqueous phase containing trifluoroacetic acid and an acetonitrile phase containing trifluoroacetic acid. The method includes the following steps: S1. Acquire the UV acquisition signal, gradient organic phase ratio, organic phase pump operating status parameters, and mixer volume during elution.

[0006] S2. Determine the pumping time and pumping volume based on the operating status parameters of the organic phase pump; determine the organic phase volume period based on the gradient organic phase ratio of the pumping volume and the corresponding pumping time; and identify pumping events with an organic phase volume period greater than the mixer volume as candidate composition disturbance events.

[0007] S3. Divide the gradient composition interval according to the gradient organic phase ratio, and perform event triggering segmentation of the ultraviolet acquisition signal according to the pumping time of the candidate composition disturbance event in the interval. Determine the event response delay and response window based on the repeated adjacent reverse responses in multiple signal segments.

[0008] S4. Perform local baseline correction on the response window, calculate the integral area difference and response interval of the positive and negative responses, and select multiple response windows that satisfy the area difference threshold and the interval consistency condition as reference response windows; determine the trifluoroacetic acid paired perturbation template and template matching threshold based on the multiple reference response windows.

[0009] S5. Match the positive and negative response segments of each response window with the corresponding side of the trifluoroacetic acid paired perturbation template, and determine the response segments that meet the template matching threshold as valid response segments; scale the trifluoroacetic acid paired perturbation template according to the amplitude ratio of the valid response segment and the corresponding side template to obtain the trifluoroacetic acid ripple component and subtract it from the ultraviolet acquisition signal; if no valid response segment is obtained, retain the ultraviolet acquisition signal of the corresponding response window.

[0010] Furthermore, the method for dividing the gradient composition interval according to the gradient organic phase ratio includes: Candidate component perturbation events are arranged in ascending order of gradient organic phase proportion, and subsequent candidate component perturbation events are added sequentially starting from the uncollected candidate component perturbation event with the smallest gradient organic phase proportion.

[0011] For each candidate component perturbation event added, the ratio of the difference between the maximum and minimum values ​​of the organic phase volume period corresponding to the added candidate component perturbation event to the median value of the organic phase volume period is calculated to obtain the period change degree. When the period change degree is greater than the period change threshold when subsequent candidate component perturbation events are added, the gradient organic phase proportion range corresponding to the previous candidate component perturbation event is determined as the initial gradient composition interval, and the interval is further divided from the next uncollected candidate component perturbation event to obtain multiple initial gradient composition intervals.

[0012] The number of candidate component perturbation events in each initial gradient component interval is counted. The initial gradient component intervals with the number of candidate component perturbation events less than the event number threshold are determined as the gradient component intervals to be merged. The periodicity of the gradient component interval to be merged and each adjacent gradient component interval after merging is calculated. The gradient component interval to be merged is merged with the adjacent gradient component interval with the smallest periodicity after merging.

[0013] The number of candidate component perturbation events within the merged gradient component interval is recounted. When the number of candidate component perturbation events is less than the event number threshold and there are adjacent gradient component intervals, adjacent gradient component intervals are merged to obtain multiple gradient component intervals.

[0014] Furthermore, the method of segmenting the ultraviolet acquisition signal into event-triggered segments based on the pumping time of candidate component disturbance events within the interval, and determining the event response delay and response window based on the repeated adjacent reverse responses in multiple signal segments, includes: Using the pumping time of each candidate component perturbation event within the gradient component interval as a time reference, the ultraviolet acquisition signal is segmented by event triggering and local trend correction to obtain multiple signal segments; the first response extreme value time of adjacent inverse responses within each signal segment is extracted, and the time difference between the first response extreme value time and the corresponding pumping time is calculated to obtain the candidate response time shift.

[0015] Candidate response time shifts are grouped according to the sampling interval of the ultraviolet acquisition signals. The proportion of the number of signal segments corresponding to each time shift group to the number of candidate component disturbance events in the gradient component interval is calculated to obtain the synchronization repetition rate. The median value of the candidate response time shift in the time shift group with the largest synchronization repetition rate and not less than the synchronization repetition rate threshold is determined as the event response delay.

[0016] The pumping time is added to the event response delay to obtain the response location time. The ultraviolet acquisition signal is divided according to the midpoint of adjacent response location times to obtain the response window corresponding to each candidate component disturbance event.

[0017] Furthermore, the method for calculating the integral area difference and response interval of the positive and negative responses, and selecting multiple response windows that satisfy the area difference threshold and the interval consistency condition as reference response windows includes: Integrate the positive and negative responses in the response window after local baseline correction to obtain the absolute values ​​of the integral areas of the positive and negative responses. Calculate the ratio of the absolute value of the difference between the integral areas of the positive and negative responses to the sum of the absolute values ​​of the integral areas of the positive and negative responses to obtain the relative difference of the integral areas.

[0018] The time difference between the extreme moments of the positive and negative responses is calculated to obtain the response interval; the median value of multiple response intervals within the same gradient component interval is calculated to obtain the baseline response interval; the ratio of the absolute value of the difference between the response interval and the baseline response interval to the baseline response interval is calculated to obtain the interval relative deviation; the response window whose relative difference of the integral area is not greater than the area difference threshold and whose interval relative deviation is not greater than the interval consistency threshold is determined as the reference response window.

[0019] Furthermore, the method for determining the trifluoroacetic acid paired perturbation template and template matching threshold based on multiple reference response windows includes: The reference response windows within the same gradient composition interval are grouped according to the order of occurrence of positive and negative responses; time alignment is performed based on the first extreme moment of the first response of each group of reference response windows, and amplitude normalization is performed based on the sum of the absolute values ​​of the integral areas of the positive and negative responses to obtain the normalized reference response windows.

[0020] Calculate the median value of the response values ​​of the normalized reference response window in the same group at each sampling time to obtain the trifluoroacetic acid paired perturbation templates in the corresponding response direction order; calculate the morphological deviation between the positive response segment of the normalized reference response window and the positive response side of the trifluoroacetic acid paired perturbation template, and the morphological deviation between the negative response segment of the normalized reference response window and the negative response side of the trifluoroacetic acid paired perturbation template, to obtain the positive reference morphological deviation set and the negative reference morphological deviation set.

[0021] The positive template matching threshold is determined based on the preset quantile value of the positive reference morphological deviation set, and the negative template matching threshold is determined based on the preset quantile value of the negative reference morphological deviation set. The positive template matching threshold and the negative template matching threshold constitute the template matching threshold.

[0022] Furthermore, the method of matching the positive and negative response segments of each response window with the corresponding sides of the trifluoroacetic acid paired perturbation template, and determining the response segments that meet the template matching threshold as valid response segments includes: Extract the positive and negative response segments from the response window respectively; calculate the amplitude ratio between the positive response segment and the positive response side of the trifluoroacetic acid paired perturbation template to obtain the positive candidate amplitude ratio; scale the positive response side of the trifluoroacetic acid paired perturbation template according to the positive candidate amplitude ratio to obtain the positive scaled template; calculate the morphological deviation between the positive response segment and the positive scaled template to obtain the positive morphological deviation.

[0023] Calculate the amplitude ratio between the negative response segment and the negative response side of the trifluoroacetic acid paired perturbation template to obtain the negative candidate amplitude ratio; scale the negative response side of the trifluoroacetic acid paired perturbation template according to the negative candidate amplitude ratio to obtain the negative scaled template; calculate the morphological deviation between the negative response segment and the negative scaled template to obtain the negative morphological deviation.

[0024] A positive response segment with a positive morphological deviation not greater than the positive template matching threshold is defined as a valid positive response segment; a negative response segment with a negative morphological deviation not greater than the negative template matching threshold is defined as a valid negative response segment.

[0025] Furthermore, when the response window contains only one of a valid positive response segment and a valid negative response segment, the method of scaling the trifluoroacetic acid paired perturbation template according to the amplitude ratio of the valid response segment and the corresponding side template includes: When the response window contains only valid positive response segments, the positive candidate amplitude ratio is determined as the perturbation amplitude ratio; when the response window contains only valid negative response segments, the negative candidate amplitude ratio is determined as the perturbation amplitude ratio; the trifluoroacetic acid paired perturbation templates that are consistent with the response direction order of the response window are scaled according to the perturbation amplitude ratio to obtain the trifluoroacetic acid ripple components corresponding to the candidate component perturbation events.

[0026] Furthermore, when the response window simultaneously contains both valid positive response segments and valid negative response segments, the method for scaling the trifluoroacetic acid paired perturbation template according to the amplitude ratio of the valid response segments and the corresponding side templates includes: Calculate the ratio of the absolute value of the difference between the positive candidate amplitude ratio and the negative candidate amplitude ratio to the average value of the positive candidate amplitude ratio and the negative candidate amplitude ratio to obtain the relative difference of amplitude ratio; determine the response window whose relative difference of amplitude ratio is not greater than the proportional consistency threshold as the bilaterally consistent response window; calculate the average value of the positive candidate amplitude ratio and the negative candidate amplitude ratio corresponding to the bilaterally consistent response window to obtain the perturbation amplitude ratio; scale the trifluoroacetic acid paired perturbation template that is consistent with the response direction order of the bilaterally consistent response window according to the perturbation amplitude ratio to obtain the trifluoroacetic acid ripple component corresponding to the candidate component perturbation event.

[0027] Based on the same inventive concept, in another aspect, the present invention also provides a noise suppression system for liquid chromatograph acquisition signals, the system comprising: The data acquisition module is used to acquire ultraviolet acquisition signals, gradient organic phase ratio, organic phase pump operating status parameters, and mixer volume during elution.

[0028] The disturbance screening module is used to determine the pumping time and pumping volume based on the operating status parameters of the organic phase pump; determine the organic phase volume period based on the gradient organic phase ratio of the pumping volume and the corresponding pumping time; and identify pumping events with an organic phase volume period greater than the mixer volume as candidate composition disturbance events.

[0029] The response localization module is used to divide the gradient composition interval according to the gradient organic phase ratio, perform event triggering segmentation of the ultraviolet acquisition signal according to the pumping time of the candidate component disturbance event in the interval, and determine the event response delay and response window based on the repeated adjacent reverse responses in multiple signal segments.

[0030] The template construction module is used to perform local baseline correction on the response window, calculate the integral area difference and response interval of the positive and negative responses, and select multiple response windows that meet the area difference threshold and the interval consistency condition as reference response windows; and determine the trifluoroacetic acid paired perturbation template and template matching threshold based on the multiple reference response windows.

[0031] The ripple suppression module is used to match the positive and negative response segments of each response window with the corresponding side of the trifluoroacetic acid paired perturbation template, and determine the response segments that meet the template matching threshold as valid response segments; the trifluoroacetic acid paired perturbation template is scaled according to the amplitude ratio of the valid response segment and the corresponding side template to obtain the trifluoroacetic acid ripple component and subtract it from the ultraviolet acquisition signal; when no valid response segment is obtained, the ultraviolet acquisition signal of the corresponding response window is retained.

[0032] Compared with the prior art, the beneficial effects of the present invention are: 1. By combining the operating parameters of the organic phase pump, the gradient organic phase ratio, and the mixer volume, candidate component perturbation events are screened. The trifluoroacetic acid ripple component is determined by using the event response timing, reference response window, and trifluoroacetic acid paired perturbation template in the same elution. This establishes a correspondence between the noise suppression process and the current gradient state and the actual pumping process, reducing the ripple residue generated by fixed frequency filtering, blank subtraction, and other methods when facing gradient changes. Furthermore, the corresponding UV acquisition signal is retained when no effective response segment is obtained, reducing the false attenuation of the sample response.

[0033] 2. After establishing the paired perturbation templates for trifluoroacetic acid, the matching status of the positive and negative response segments with the corresponding side templates is determined. When one side is effective, the amplitude ratio corresponding to the effective response segment is used to reconstruct the complete trifluoroacetic acid ripple component. When both sides are effective, the perturbation amplitude ratio is determined by the consistency between the positive candidate amplitude ratio and the negative candidate amplitude ratio. This ensures that even if the sample response and the trifluoroacetic acid ripple partially overlap, causing distortion of one side's response, the ripple reconstruction can still be completed using the effective response on the other side. Furthermore, the mutual verification of the two-sided responses reduces the probability of morphologically similar interference being mistakenly identified as trifluoroacetic acid ripple and subtracted. Attached Figure Description

[0034] Figure 1 This is a flowchart of a method for suppressing noise in a liquid chromatograph signal acquisition according to the present invention; Figure 2 This is a block diagram of a signal noise suppression system for liquid chromatographs according to the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Example 1: As Figure 1 As shown, this embodiment provides a method for suppressing noise in liquid chromatography (LC) acquisition signals, applied to LC analysis scenarios using a reversed-phase binary high-pressure gradient elution with an aqueous phase containing trifluoroacetic acid and an acetonitrile phase containing trifluoroacetic acid. The method includes: A binary high-pressure gradient liquid chromatograph was used, with a C18 packed column for the reversed-phase chromatography. The aqueous phase containing trifluoroacetic acid was a 0.10% (v / v) aqueous solution of trifluoroacetic acid, and the acetonitrile phase containing trifluoroacetic acid was a 0.10% (v / v) trifluoroacetic acid acetonitrile solution. The proportion of the gradient organic phase increased linearly from 10.0% to 22.0% within 0–720 s, the elution flow rate was maintained at 0.30 mL / min, the UV detection wavelength was 214 nm, the UV acquisition signal sampling frequency was 20 Hz, and the mixer volume was 75.0 μL. The effective stroke volume of the organic phase pump during this elution was 20.0 μL.

[0037] Acquire UV acquisition signals, gradient organic phase ratio, organic phase pump operating parameters, and mixer volume during elution.

[0038] The UV acquisition signal was recorded at 0.05s sampling intervals, with the sampling time and absorbance response value recorded. The gradient organic phase ratio was matched with the sampling time according to the gradient program saved in the HPLC controller. The organic phase pump operating status parameters were recorded with the effective stroke end time and effective stroke volume, with the effective stroke end time used as the pumping time and the effective stroke volume used as the pumping volume. The UV acquisition signal, gradient organic phase ratio, and organic phase pump operating status parameters were recorded according to a unified time base, with each sampling data corresponding to a timestamp. The organic phase pump operating status parameters were derived from the drive control records of the organic phase pump in the HPLC, using operating parameters that could determine the pumping time and pumping volume corresponding to each pumping event.

[0039] The pumping time and volume are determined based on the operating parameters of the organic phase pump; the organic phase volume period is determined based on the gradient organic phase ratio of the pumping volume and the corresponding pumping time; and pumping events with an organic phase volume period greater than the mixer volume are identified as candidate composition disturbance events.

[0040] No. The pumping volume corresponding to each pumping event is The gradient organic phase ratio conversion value corresponding to the pumping time is Organic phase volume period .

[0041] For example, the pumping time of candidate component perturbation event 14 is 415.874 s, corresponding to a gradient organic phase proportion of 16.9312%. μL, 118.125 > 75.0, meeting the screening criteria. The organic phase volume period characterizes the total mobile phase volume scale corresponding to a single organic phase pumping volume at the corresponding gradient organic phase ratio. When the organic phase volume period is greater than the mixer volume, the local compositional changes caused by the organic phase pumping state may not be sufficiently smoothed within the mixer. The corresponding pumping event therefore enters the candidate compositional perturbation event set. Whether an adjacent reverse response corresponding to the pumping event is formed is further determined by the subsequent event response delay determination process. A total of 28 candidate compositional perturbation events were screened in this 720s elution, numbered 1 to 28 in ascending order of gradient organic phase ratio.

[0042] Candidate component perturbation events are arranged in ascending order of gradient organic phase proportion, and subsequent candidate component perturbation events are added sequentially starting from the uncollected candidate component perturbation event with the smallest gradient organic phase proportion.

[0043] Candidate component perturbation events are sorted by gradient organic phase ratio as the sorting key, and if the gradient organic phase ratios are the same, the pumping time is used as the second sorting key. At the beginning of the partitioning, all candidate component perturbation events are set to an unsettled state. Each time an initial gradient component interval is formed, the candidate component perturbation events contained in the initial gradient component interval are removed from the unsettled set.

[0044] For each candidate component perturbation event added, the ratio of the difference between the maximum and minimum values ​​of the organic phase volume period corresponding to the added candidate component perturbation event to the median value of the organic phase volume period is calculated to obtain the period change degree. When the period change degree is greater than the period change threshold when subsequent candidate component perturbation events are added, the gradient organic phase proportion range corresponding to the previous candidate component perturbation event is determined as the initial gradient composition interval, and the interval is further divided from the next uncollected candidate component perturbation event to obtain multiple initial gradient composition intervals.

[0045] The maximum periodic value of the organic phase volume corresponding to the current set is The minimum periodic value of the organic phase volume is The median value of the organic phase volume period is Periodicity In this embodiment, the periodicity of candidate component perturbation events 1-3 is 0.1060; after adding candidate component perturbation event 4, the periodicity increases to 0.1520, which is greater than 0.15. Therefore, candidate component perturbation events 1-3 form the first initial gradient component interval. Following the same calculation relationship, five initial gradient component intervals are obtained corresponding to event number ranges 1-3, 4-7, 8-13, 14-21, and 22-28.

[0046] The number of candidate component perturbation events in each initial gradient component interval is counted. The initial gradient component intervals with the number of candidate component perturbation events less than the event number threshold are determined as the gradient component intervals to be merged. The periodicity of the gradient component interval to be merged and each adjacent gradient component interval after merging is calculated. The gradient component interval to be merged is merged with the adjacent gradient component interval with the smallest periodicity after merging.

[0047] Based on the event response delay statistics and the minimum number of candidate component perturbation events required for constructing the trifluoroacetic acid paired perturbation template, the event number threshold was determined to be 5. The initial gradient component intervals corresponding to event number ranges 1–3 and 4–7 contain 3 and 4 candidate component perturbation events respectively, both less than the event number threshold. Event number ranges 1–3 are located on the gradient initiation side, and only the adjacent gradient component intervals corresponding to event number ranges 4–7 exist; therefore, the two initial gradient component intervals are merged. After merging, there are 7 candidate component perturbation events, with a periodicity of 0.2726. When the gradient component interval to be merged is located between two gradient component intervals, the periodicity of the merged interval with the preceding and following gradient component intervals is calculated separately, and the merged result with the smaller periodicity is selected. When the gradient component interval to be merged is located at the end of the gradient range, it is merged with the only adjacent gradient component interval.

[0048] The number of candidate component perturbation events within the merged gradient component interval is recounted. When the number of candidate component perturbation events is less than the event number threshold and there are adjacent gradient component intervals, adjacent gradient component intervals are merged to obtain multiple gradient component intervals.

[0049] After merging, the event number ranges 1–7, 8–13, 14–21, and 22–28 constitute gradient composition intervals G1, G2, G3, and G4, respectively, with 7, 6, 8, and 7 candidate composition perturbation events, respectively. The periodicity change threshold is used to determine the boundaries of the initial gradient composition intervals; when merging gradient composition intervals, the degree of periodicity change is used as a comparison of the merging directions of adjacent gradient composition intervals, but not as a merging termination condition. In this embodiment, based on the organic phase volume periodicity fluctuation range corresponding to the organic phase pump operating status parameters under the same liquid chromatography analysis conditions, the upper limit of the relative periodicity change of the organic phase volume allowed to be classified into the same initial gradient composition interval is set to 15%, and the periodicity change threshold is determined to be 0.15.

[0050] If the number of candidate component perturbation events is less than the event number threshold and there are adjacent gradient component intervals, merging continues. If there are no adjacent gradient component intervals that can be merged and the number of candidate component perturbation events is still less than the event number threshold, a pairwise perturbation template for trifluoroacetic acid is not established using the corresponding gradient component interval. If the number of reference response windows in the same response direction order within the same gradient component interval is less than the reference response window number threshold, a pairwise perturbation template for trifluoroacetic acid in the corresponding response direction order is not established.

[0051] Using the pumping time of each candidate component perturbation event within the gradient component interval as a time reference, the ultraviolet acquisition signal is segmented by event triggering and local trend correction to obtain multiple signal segments; the first response extreme value time of adjacent inverse responses within each signal segment is extracted, and the time difference between the first response extreme value time and the corresponding pumping time is calculated to obtain the candidate response time shift.

[0052] The gradient composition interval G3 includes candidate composition perturbation events 14–21. The left boundary of the event trigger segment is taken as 1.00 s before the corresponding pumping time, and the right boundary is set according to the flow path transport time from the organic phase pumping position to the UV detection position in the liquid chromatograph. The flow path transport time is determined based on the effective flow path volume and elution flow rate between the organic phase pumping position and the UV detection position, so that the signal segment covers the possible time range of the response corresponding to the candidate composition perturbation event. The local trend is represented by a straight line, and the response values ​​at both ends of the line are taken as the median values ​​of the continuous sampling values ​​at the beginning and end of the signal segment, respectively. The two median values ​​correspond to the center time of the two continuous sampling intervals, forming a local trend line. The signal segment sampling value minus the local trend value at the same time participates in the search for adjacent reverse responses.

[0053] Local extrema are extracted from the signal segments after local trend correction according to the sampling time sequence. When the response signs of two adjacent local extrema are opposite, they form a set of adjacent reverse responses. When there are multiple sets of adjacent reverse responses within the same signal segment, the sum of the absolute values ​​of the two local extrema in each set of adjacent reverse responses is calculated. The adjacent reverse response with the largest sum of the absolute values ​​of the two local extrema is taken as the candidate adjacent reverse response for the corresponding signal segment. The first response extremum time is taken as the earlier local extremum time among the candidate adjacent reverse responses. A total of 8 candidate response time shifts are obtained for the signal segments corresponding to disturbance events 14-21. Among them, 5 candidate response time shifts fall into the same time shift group, and the remaining 3 candidate response time shifts fall into other time shift groups respectively.

[0054] Candidate response time shifts are grouped according to the sampling interval of the ultraviolet acquisition signals. The proportion of the number of signal segments corresponding to each time shift group to the number of candidate component disturbance events in the gradient component interval is calculated to obtain the synchronization repetition rate. The median value of the candidate response time shift in the time shift group with the largest synchronization repetition rate and not less than the synchronization repetition rate threshold is determined as the event response delay.

[0055] Using 0.05s as the grouping interval for candidate response time shifts, each time shift group is centered at an integer multiple of 0.05s, with the midpoint between two adjacent group centers serving as the group boundary. The left boundary is included in the corresponding time shift group, while the right boundary is excluded. Candidate response time shifts located at a common boundary are assigned to the time shift group with the larger time value. Based on the proportion of candidate component disturbance events that need to be covered by the synchronous response within the same gradient component interval, the synchronization repetition rate threshold is determined to be 0.50. Among the eight signal segments corresponding to candidate component disturbance events 14–21, the time shift group corresponding to the maximum synchronization repetition rate contains five signal segments, with a synchronization repetition rate of 5 / 8 = 0.625, which is greater than the synchronization repetition rate threshold of 0.50. Therefore, the median value of the candidate response time shift within the time shift group corresponding to the maximum synchronization repetition rate is determined as the event response delay. The candidate response time shift corresponding to candidate component disturbance event 17 does not belong to the time shift group corresponding to the maximum synchronization repetition rate. When the maximum synchronization repetition rate is less than the synchronization repetition rate threshold, the event response delay of the corresponding gradient component interval is uncertain, and no response window for the corresponding gradient component interval is generated.

[0056] The pumping time is added to the event response delay to obtain the response location time. The ultraviolet acquisition signal is divided according to the midpoint of adjacent response location times to obtain the response window corresponding to each candidate component disturbance event.

[0057] Event response delay is denoted as The pumping time of candidate component disturbance event 14 is 415.874s, and the response positioning time of candidate component disturbance event 14 is... The response location times of adjacent candidate disturbance events are obtained by adding the event response delay to the corresponding pumping time. The midpoint of two adjacent response location times is used as the common boundary of adjacent response windows. The midpoint of adjacent response location times is used as the left and right boundaries of intermediate response windows; the left boundary of the first response window and the right boundary of the last response window extend outward by half the interval between two adjacent response location times. When the boundary time is between two sampling times, the sampling time closer to the boundary time is taken.

[0058] Local baseline correction is performed on the response window. The positive and negative responses in the locally baseline corrected response window are integrated to obtain the absolute values ​​of the integrated areas of the positive and negative responses. The ratio of the absolute value of the difference between the absolute values ​​of the integrated areas of the positive and negative responses to the sum of the absolute values ​​of the integrated areas of the positive and negative responses is calculated to obtain the relative difference of the integrated areas. The positive and negative responses in the response window are the positive and negative responses in the candidate adjacent reverse responses within the corresponding signal segments, respectively.

[0059] In this embodiment, the left reference segment of the local baseline is taken 0.50–0.30 s before the response positioning time, and the right reference segment is taken 0.95–1.15 s after the response positioning time. The median values ​​of the sampled values ​​of the left and right reference segments are assigned to the center times of the two reference segments, respectively. Linear interpolation is used between the two center times, and the baseline slope corresponding to the adjacent center times is extended to the integration boundary outside the center times. The response value after local baseline correction is equal to the sampled value of the ultraviolet acquisition signal minus the local baseline value at the same sampling time. The integration boundary of adjacent inverse responses is taken as the zero-crossing time of the first response before, the positive-negative response transition position, and the second response after; when the zero-crossing is located between two sampling points, it is determined by linear interpolation of the two sampled values. When the first response before, the positive-negative response transition position, and the second response after form a complete zero-crossing time, the integration area of ​​the positive response and the integration area of ​​the negative response are both integrated using trapezoidal integration, and the absolute value of the integration area of ​​the negative response is taken after integration; when a complete zero-crossing time is not formed, the corresponding response window does not enter the reference response window set.

[0060] relative difference of integral area , This represents the area of ​​the positive response integral. This represents the absolute value of the integral area of ​​the negative response. and The integral area is determined by the sign of the response and is independent of the order in which the positive and negative responses appear. The unit of integral area is mA·s. When the local organic phase ratio changes for a short time, the distribution state of trifluoroacetic acid between the reversed stationary phase and the mobile phase shifts accordingly; after the local organic phase ratio recovers, the distribution state of trifluoroacetic acid returns to its original equilibrium direction, causing the UV response corresponding to the local composition change and recovery process to change in opposite directions. Within the same gradient composition interval, adjacent reverse responses that have undergone local baseline correction and are not significantly disturbed by the sample response usually show that the absolute values ​​of the integral areas of the positive and negative responses are close. The relative difference in integral areas is used to characterize the degree of closeness between the areas of the positive and negative responses.

[0061] The time difference between the extreme moments of the positive and negative responses is calculated to obtain the response interval; the median value of multiple response intervals within the same gradient component interval is calculated to obtain the baseline response interval; the ratio of the absolute value of the difference between the response interval and the baseline response interval to the baseline response interval is calculated to obtain the interval relative deviation; the response window whose relative difference of the integral area is not greater than the area difference threshold and whose interval relative deviation is not greater than the interval consistency threshold is determined as the reference response window.

[0062] Based on the discrete range of the relative difference of the integrated area and the discrete range of the response interval for paired responses of trifluoroacetic acid under the same liquid chromatography analysis conditions, the area difference threshold was determined to be 0.12, and the interval consistency threshold was determined to be 0.10. The baseline response interval for gradient component interval G3 is 0.65 s. The relative difference of the integrated area for candidate component perturbation event 14 is 0.0184, the response interval is 0.60 s, and the relative deviation of the interval is |0.60-0.65| / 0.65=0.0769. Since the relative difference of the integrated area is not greater than the area difference threshold and the relative deviation of the interval is not greater than the interval consistency threshold, the response window corresponding to candidate component perturbation event 14 is included in the reference response window set. The relative difference of the integrated area for candidate component perturbation event 17 is 0.1935, the response interval is 0.80 s, and the relative deviation of the interval is |0.80-0.65| / 0.65=0.2308. The response window corresponding to candidate component perturbation event 17 is not included in the reference response window set. The response interval for candidate component perturbation event 19 is 0.75 s, and the relative deviation of the interval is |0.75-0.65| / 0.65=0.1538, which is greater than the interval consistency threshold of 0.10. Therefore, the response window corresponding to candidate component perturbation event 19 will not be included in the reference response window set. The gradient component interval G3 ultimately retains the five reference response windows corresponding to candidate component perturbation events 14, 15, 18, 20, and 21.

[0063] Reference response windows within the same gradient component interval are grouped according to the order of occurrence of positive and negative responses. Time alignment is performed based on the first extreme value of each group's reference response window, and amplitude normalization is performed based on the sum of the absolute values ​​of the integral areas of the positive and negative responses to obtain normalized reference response windows. The proportions of organic phases within the same gradient component interval are similar. The elution flow rate, reversed-phase column, and mobile phase system remain unchanged in the same elution. The transfer conditions corresponding to the transient redistribution of trifluoroacetic acid between the reversed-phase stationary and mobile phases are similar. Adjacent reverse responses with the same response direction sequence exhibit repeatability in response intervals and local waveform morphology. Differences in the occurrence time and intensity of responses corresponding to perturbation events of different candidate components are eliminated by time alignment and amplitude normalization.

[0064] The five reference response windows of the gradient composition interval G3 all follow a positive-to-negative response direction order, forming a grouping based on the same response direction order. The median of the sum of the absolute values ​​of the positive and negative integral areas of the five reference response windows is 0.1198 mAU·s, with amplitude normalized... Each sampled value in the normalized reference response window is multiplied by .

[0065] Time alignment is performed with the extreme moment of the first response as the relative time 0s. Each reference response window is shifted by an integer number of sampling points at a sampling interval of 0.05s. The missing sampling positions at both ends caused by the shift are not included in the median value calculation of the corresponding sampling time.

[0066] Calculate the median value of the response values ​​of the normalized reference response window in the same group at each sampling time to obtain the trifluoroacetic acid paired perturbation templates in the corresponding response direction order; calculate the morphological deviation between the positive response segment of the normalized reference response window and the positive response side of the trifluoroacetic acid paired perturbation template, and the morphological deviation between the negative response segment of the normalized reference response window and the negative response side of the trifluoroacetic acid paired perturbation template, to obtain the positive reference morphological deviation set and the negative reference morphological deviation set.

[0067] The trifluoroacetic acid paired perturbation template uses a 0.05s time grid consistent with the UV acquisition signal. The gradient composition interval G3 corresponds to the trifluoroacetic acid paired perturbation template, which is positive first and then negative. It obtains a positive extreme value of 0.180mAU at a relative time of 0s and a negative extreme value of -0.174mAU at a relative time of 0.65s.

[0068] The morphological bias is calculated using the normalized root mean square bias. The sampled values ​​of a response segment of length m are denoted as... The corresponding template-side sampled value is denoted as Amplitude fitting coefficient morphological deviation When calculating the morphological deviation using the reference response window, the sampled values ​​of the normalized reference response window are used, and the positive and negative response sides are calculated separately. The positive reference morphological deviation of the gradient component interval G3 is distributed between 0.027 and 0.033, and the negative reference morphological deviation is distributed between 0.022 and 0.027.

[0069] The positive template matching threshold is determined based on the preset quantile value of the positive reference morphological deviation set, and the negative template matching threshold is determined based on the preset quantile value of the negative reference morphological deviation set. The positive template matching threshold and the negative template matching threshold constitute the template matching threshold.

[0070] The preset quantile value is set to 95%, and the nearest rank rule is used for the quantile value. The 95th quantile value corresponding to the five positive reference morphological deviations is 0.033, and the positive template matching threshold is set to 0.033; the 95th quantile value corresponding to the five negative reference morphological deviations is 0.027, and the negative template matching threshold is set to 0.027. The minimum number of reference response windows required to construct the trifluoroacetic acid paired perturbation template using the median value of the response value at each sampling time using multiple reference response windows is set to 3. When the number of reference response windows corresponding to the same response direction sequence within the same gradient composition interval is less than 3, trifluoroacetic acid paired perturbation templates for the corresponding response direction sequence are not generated, and no corresponding positive and negative template matching thresholds are set.

[0071] Extract the positive and negative response segments from the response window respectively; calculate the amplitude ratio between the positive response segment and the positive response side of the trifluoroacetic acid paired perturbation template to obtain the positive candidate amplitude ratio; scale the positive response side of the trifluoroacetic acid paired perturbation template according to the positive candidate amplitude ratio to obtain the positive scaled template; calculate the morphological deviation between the positive response segment and the positive scaled template to obtain the positive morphological deviation.

[0072] The positive response segment uses 13 sampling points from 0.30s before to 0.30s after the extreme value of the positive response, and the proportion of positive candidate amplitudes... The first step of forward scaling template Each sample value equals Positive morphological deviation .

[0073] Calculate the amplitude ratio between the negative response segment and the negative response side of the trifluoroacetic acid paired perturbation template to obtain the negative candidate amplitude ratio; scale the negative response side of the trifluoroacetic acid paired perturbation template according to the negative candidate amplitude ratio to obtain the negative scaled template; calculate the morphological deviation between the negative response segment and the negative scaled template to obtain the negative morphological deviation.

[0074] The negative response segment uses 13 sampling points from 0.30s before to 0.30s after the extreme moment of the negative response, and the proportion of negative candidate amplitudes... The negative scaling template's first... Each sample value equals Negative morphological deviation When calculating the proportion of negative candidate amplitudes, signed response values ​​are used for both the negative response segment and the negative response side of the trifluoroacetic acid paired perturbation template.

[0075] A positive response segment with a positive morphological deviation not greater than the positive template matching threshold is defined as a valid positive response segment; a negative response segment with a negative morphological deviation not greater than the negative template matching threshold is defined as a valid negative response segment.

[0076] The positive candidate amplitude ratio of candidate component perturbation event 16 is 1.38, and the positive shape deviation is 0.118; the negative candidate amplitude ratio is 1.04, and the negative shape deviation is 0.025. Candidate component perturbation event 16 only yields a valid negative response segment. The positive shape deviation of candidate component perturbation event 17 is 0.091, and the negative shape deviation is 0.079. Candidate component perturbation event 17 does not yield a valid response segment. The positive candidate amplitude ratio of candidate component perturbation event 19 is 1.20, and the positive shape deviation is 0.032; the negative candidate amplitude ratio is 1.16, and the negative shape deviation is 0.026. Candidate component perturbation event 19 yields both valid positive and valid negative response segments.

[0077] When the response window contains only valid positive response segments, the positive candidate amplitude ratio is determined as the perturbation amplitude ratio; when the response window contains only valid negative response segments, the negative candidate amplitude ratio is determined as the perturbation amplitude ratio; the trifluoroacetic acid paired perturbation templates are scaled according to the perturbation amplitude ratio and are consistent with the response direction order of the response window to obtain the trifluoroacetic acid ripple components corresponding to the candidate component perturbation events and subtract them from the ultraviolet acquisition signal.

[0078] The trifluoroacetic acid paired perturbation template with the same response direction order within the same gradient composition interval is formed by complete paired responses from multiple reference response windows. The trifluoroacetic acid paired perturbation template retains the relative amplitude relationship between the positive and negative response sides. When the candidate amplitude ratio obtained by matching the effective positive or negative response segment with the corresponding side of the trifluoroacetic acid paired perturbation template is used to scale the complete trifluoroacetic acid paired perturbation template, the other response side is scaled synchronously according to the relative amplitude relationship in the trifluoroacetic acid paired perturbation template.

[0079] Candidate component perturbation event 16 only has a valid negative response segment, and the perturbation amplitude ratio is taken as the negative candidate amplitude ratio of 1.04. After multiplying the trifluoroacetic acid paired perturbation template of gradient component interval G3 by 1.04, the positive extremum of the trifluoroacetic acid ripple component is 0.187mAU, and the negative extremum is -0.181mAU. The original positive local extremum of candidate component perturbation event 16 is 0.251mAU, and after subtracting the positive extremum of 0.187mAU, 0.064mAU is retained; the positive candidate amplitude ratio of 1.38 is not involved in the determination of the perturbation amplitude ratio of candidate component perturbation event 16.

[0080] When the response window simultaneously contains both valid positive and valid negative response segments, the absolute value of the difference between the positive and negative candidate amplitude ratios is calculated as a percentage of the average of the positive and negative candidate amplitude ratios, yielding a relative difference in amplitude ratios. Response windows with a relative difference in amplitude ratios not exceeding the proportional consistency threshold are defined as bilaterally consistent response windows. The average of the positive and negative candidate amplitude ratios corresponding to the bilaterally consistent response window is calculated to obtain the perturbation amplitude ratio. Trifluoroacetic acid paired perturbation templates, which follow the same response direction order as the bilaterally consistent response window, are scaled according to the perturbation amplitude ratio to obtain the trifluoroacetic acid ripple component corresponding to the candidate component perturbation event and subtracted from the ultraviolet acquisition signal.

[0081] The proportional consistency threshold is determined based on the preset quantile value of the set of relative differences in amplitude proportions corresponding to the reference response window. In this embodiment, the 95th quantile value is used and set to 0.08. The relative difference in amplitude proportion of candidate component perturbation event 19 is |1.20-1.16| / [(1.20+1.16) / 2]=0.0339, which is less than 0.08. Therefore, the corresponding response window is determined to be a two-sided consistent response window, and the perturbation amplitude proportion is (1.20+1.16) / 2=1.18. After multiplying the entire trifluoroacetic acid paired perturbation template by 1.18, the positive extreme value of the trifluoroacetic acid ripple component is 0.212mAU, and the negative extreme value is -0.205mAU.

[0082] When both valid positive and negative response segments exist simultaneously, and the relative difference in amplitude ratios exceeds the proportional consistency threshold, the proportions of positive morphological deviation to the positive template matching threshold and negative morphological deviation to the negative template matching threshold are calculated separately to obtain the positive and negative normalized morphological deviations. The smaller of the two, the candidate amplitude ratio, is determined as the perturbation amplitude ratio. Trifluoroacetic acid paired perturbation templates, consistent with the response direction order of the response window, are scaled according to the perturbation amplitude ratio. When the positive and negative normalized morphological deviations are equal, the average of the positive and negative candidate amplitude ratios is taken as the perturbation amplitude ratio.

[0083] If no valid response segment is obtained, the UV acquisition signal of the corresponding response window is retained.

[0084] Since no valid response segment was obtained for candidate component perturbation event 17, the trifluoroacetic acid ripple component within the corresponding response window was set to a zero-value sequence, while the UV acquisition signal sampling value remained unchanged. Each response window adopted a non-overlapping midpoint boundary; the reconstructed trifluoroacetic acid ripple component was only written to the corresponding response window, while the trifluoroacetic acid ripple component outside the response window was set to zero. The original UV acquisition signal sampling value was... The trifluoroacetic acid ripple component corresponding to the response window is Corrected sample value When the maximum synchronization repetition rate is less than the synchronization repetition rate threshold, the response delay for uncertain events is not specified, and no response window is generated for the corresponding gradient component interval; the corresponding gradient component interval is not processed. If a response window has been generated but a trifluoroacetic acid paired perturbation template is missing, or if a response window has been generated but no effective response segment is obtained, ... Take the sequence of zero values.

[0085] Example 2: Based on the same inventive concept, such as Figure 2 As shown, this embodiment also provides a noise suppression system for liquid chromatograph acquisition signals, the system comprising: The data acquisition module is used to acquire ultraviolet acquisition signals, gradient organic phase ratio, organic phase pump operating status parameters, and mixer volume during elution.

[0086] The disturbance screening module is used to determine the pumping time and pumping volume based on the operating status parameters of the organic phase pump; determine the organic phase volume period based on the gradient organic phase ratio of the pumping volume and the corresponding pumping time; and identify pumping events with an organic phase volume period greater than the mixer volume as candidate composition disturbance events.

[0087] The response localization module is used to divide the gradient composition interval according to the gradient organic phase ratio, perform event triggering segmentation of the ultraviolet acquisition signal according to the pumping time of the candidate component disturbance event in the interval, and determine the event response delay and response window based on the repeated adjacent reverse responses in multiple signal segments.

[0088] The template construction module is used to perform local baseline correction on the response window, calculate the integral area difference and response interval of the positive and negative responses, and select multiple response windows that meet the area difference threshold and the interval consistency condition as reference response windows; and determine the trifluoroacetic acid paired perturbation template and template matching threshold based on the multiple reference response windows.

[0089] The ripple suppression module is used to match the positive and negative response segments of each response window with the corresponding side of the trifluoroacetic acid paired perturbation template, and determine the response segments that meet the template matching threshold as valid response segments; the trifluoroacetic acid paired perturbation template is scaled according to the amplitude ratio of the valid response segment and the corresponding side template to obtain the trifluoroacetic acid ripple component and subtract it from the ultraviolet acquisition signal; when no valid response segment is obtained, the ultraviolet acquisition signal of the corresponding response window is retained.

[0090] It should be noted that the specific methods by which each module performs operations in the system described in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0091] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for suppressing signal noise in liquid chromatography, applied to a liquid chromatography analysis scenario using a reversed-phase binary high-pressure gradient elution with an aqueous phase containing trifluoroacetic acid and an acetonitrile phase containing trifluoroacetic acid, characterized in that, The method includes: Acquire UV acquisition signals, gradient organic phase ratio, organic phase pump operating status parameters, and mixer volume during elution; The pumping time and pumping volume are determined based on the operating status parameters of the organic phase pump; the organic phase volume period is determined based on the pumping volume and the gradient organic phase ratio corresponding to the pumping time; and pumping events with an organic phase volume period greater than the mixer volume are identified as candidate composition disturbance events. The gradient composition intervals are divided according to the ratio of the gradient organic phases. The ultraviolet acquisition signals are segmented by event triggering based on the pumping time of the candidate composition perturbation events within the intervals. The event response delay and response window are determined based on the repeated adjacent reverse responses in multiple signal segments. Local baseline correction is performed on the response window, the integral area difference and response interval of the positive and negative responses are calculated, and multiple response windows that satisfy the area difference threshold and the interval consistency condition are selected as reference response windows; the trifluoroacetic acid paired perturbation template and template matching threshold are determined based on the multiple reference response windows. The positive and negative response segments of each response window are matched with the corresponding side of the trifluoroacetic acid paired perturbation template. The response segments that meet the template matching threshold are determined as valid response segments. The trifluoroacetic acid paired perturbation template is scaled according to the amplitude ratio of the valid response segment and the corresponding side template to obtain the trifluoroacetic acid ripple component, which is then subtracted from the ultraviolet acquisition signal. If no valid response segment is obtained, the ultraviolet acquisition signal of the corresponding response window is retained.

2. The method for suppressing noise in a liquid chromatograph acquisition signal according to claim 1, characterized in that, The method for dividing gradient composition intervals according to the proportion of gradient organic phases includes: Candidate component perturbation events are arranged in ascending order of gradient organic phase proportion, and subsequent candidate component perturbation events are added sequentially starting from the uncollected candidate component perturbation event with the smallest gradient organic phase proportion. For each candidate component perturbation event added, the ratio of the difference between the maximum and minimum values ​​of the organic phase volume period corresponding to the added candidate component perturbation event to the median value of the organic phase volume period is calculated to obtain the period change degree. When the period change degree is greater than the period change threshold when subsequent candidate component perturbation events are added, the gradient organic phase ratio range corresponding to the previous candidate component perturbation event is determined as the initial gradient composition interval, and the interval is further divided from the next uncollected candidate component perturbation event to obtain multiple initial gradient composition intervals. The number of candidate component perturbation events in each initial gradient component interval is counted. The initial gradient component intervals with the number of candidate component perturbation events less than the event number threshold are determined as the gradient component intervals to be merged. The periodicity of the gradient component interval to be merged and each adjacent gradient component interval after merging is calculated. The gradient component interval to be merged is merged with the adjacent gradient component interval with the smallest periodicity after merging. The number of candidate component perturbation events within the merged gradient component interval is recounted. When the number of candidate component perturbation events is less than the event number threshold and there are adjacent gradient component intervals, adjacent gradient component intervals are merged to obtain multiple gradient component intervals.

3. The method for suppressing noise in a liquid chromatograph acquisition signal according to claim 2, characterized in that, The method of segmenting the ultraviolet acquisition signal into event-triggered segments based on the pumping time of candidate component disturbance events within the interval, and determining the event response delay and response window based on the repeated adjacent reverse responses in multiple signal segments, includes: Using the pumping time of each candidate component perturbation event within the gradient component interval as the time reference, the ultraviolet acquisition signal is segmented by event triggering and local trend correction to obtain multiple signal segments; the first response extreme value time of adjacent inverse responses within each signal segment is extracted, and the time difference between the first response extreme value time and the corresponding pumping time is calculated to obtain the candidate response time shift. Candidate response time shifts are grouped according to the sampling interval of the ultraviolet acquisition signals. The proportion of the number of signal segments corresponding to each time shift group to the number of candidate component disturbance events in the gradient component interval is calculated to obtain the synchronization repetition rate. The median value of the candidate response time shift in the time shift group with the largest synchronization repetition rate and not less than the synchronization repetition rate threshold is determined as the event response delay. The pumping time is added to the event response delay to obtain the response location time. The ultraviolet acquisition signal is divided according to the midpoint of adjacent response location times to obtain the response window corresponding to each candidate component disturbance event.

4. The method for suppressing noise in a liquid chromatograph acquisition signal according to claim 3, characterized in that, The method for calculating the integral area difference and response interval of positive and negative responses, and selecting multiple response windows as reference response windows where the integral area difference meets the area difference threshold and the response interval meets the interval consistency condition, includes: Integrate the positive and negative responses in the response window after local baseline correction to obtain the absolute values ​​of the integral areas of the positive and negative responses; calculate the ratio of the absolute value of the difference between the integral areas of the positive and negative responses to the sum of the absolute values ​​of the integral areas of the positive and negative responses to obtain the relative difference of the integral areas. The time difference between the extreme moments of the positive and negative responses is calculated to obtain the response interval; the median value of multiple response intervals within the same gradient component interval is calculated to obtain the baseline response interval; the ratio of the absolute value of the difference between the response interval and the baseline response interval to the baseline response interval is calculated to obtain the interval relative deviation; the response window whose relative difference of the integral area is not greater than the area difference threshold and whose interval relative deviation is not greater than the interval consistency threshold is determined as the reference response window.

5. A method for suppressing noise in a liquid chromatograph acquisition signal according to claim 4, characterized in that, The method for determining the trifluoroacetic acid paired perturbation template and template matching threshold based on multiple reference response windows includes: The reference response windows within the same gradient composition interval are grouped according to the order of appearance of positive and negative responses; time alignment is performed according to the first response extreme moment of each group of reference response windows, and amplitude normalization is performed according to the sum of the absolute values ​​of the integral areas of the positive and negative responses to obtain the normalized reference response windows. Calculate the median value of the response values ​​of the normalized reference response window in the same group at each sampling time to obtain the trifluoroacetic acid paired perturbation templates in the corresponding response direction order; calculate the morphological deviation between the positive response segment of the normalized reference response window and the positive response side of the trifluoroacetic acid paired perturbation template, and the morphological deviation between the negative response segment of the normalized reference response window and the negative response side of the trifluoroacetic acid paired perturbation template, to obtain the positive reference morphological deviation set and the negative reference morphological deviation set. The positive template matching threshold is determined based on the preset quantile value of the positive reference morphological deviation set, and the negative template matching threshold is determined based on the preset quantile value of the negative reference morphological deviation set. The positive template matching threshold and the negative template matching threshold constitute the template matching threshold.

6. The method for suppressing noise in a liquid chromatograph acquisition signal according to claim 5, characterized in that, The method of matching the positive and negative response segments of each response window with the corresponding sides of the trifluoroacetic acid paired perturbation template, and determining the response segments that meet the template matching threshold as valid response segments includes: Extract the positive and negative response segments from the response window respectively; calculate the amplitude ratio between the positive response segment and the positive response side of the trifluoroacetic acid paired perturbation template to obtain the positive candidate amplitude ratio; scale the positive response side of the trifluoroacetic acid paired perturbation template according to the positive candidate amplitude ratio to obtain the positive scaled template; calculate the morphological deviation between the positive response segment and the positive scaled template to obtain the positive morphological deviation; Calculate the amplitude ratio between the negative response segment and the negative response side of the trifluoroacetic acid paired perturbation template to obtain the negative candidate amplitude ratio; scale the negative response side of the trifluoroacetic acid paired perturbation template according to the negative candidate amplitude ratio to obtain the negative scaled template; calculate the shape deviation between the negative response segment and the negative scaled template to obtain the negative shape deviation. A positive response segment with a positive morphological deviation not greater than the positive template matching threshold is defined as a valid positive response segment; a negative response segment with a negative morphological deviation not greater than the negative template matching threshold is defined as a valid negative response segment.

7. A method for suppressing noise in a liquid chromatograph acquisition signal according to claim 6, characterized in that, When the response window contains only one of a valid positive response segment and a valid negative response segment, the method of scaling the trifluoroacetic acid paired perturbation template according to the amplitude ratio of the valid response segment and the corresponding side template includes: When the response window contains only valid positive response segments, the positive candidate amplitude ratio is determined as the perturbation amplitude ratio; when the response window contains only valid negative response segments, the negative candidate amplitude ratio is determined as the perturbation amplitude ratio; the trifluoroacetic acid paired perturbation templates that are consistent with the response direction order of the response window are scaled according to the perturbation amplitude ratio to obtain the trifluoroacetic acid ripple components corresponding to the candidate component perturbation events.

8. A method for suppressing noise in a liquid chromatograph acquisition signal according to claim 6, characterized in that, When the response window simultaneously contains both valid positive and valid negative response segments, the method for scaling the trifluoroacetic acid paired perturbation template according to the amplitude ratio of the valid response segment and the corresponding side template includes: Calculate the ratio of the absolute value of the difference between the positive candidate amplitude ratio and the negative candidate amplitude ratio to the average value of the positive candidate amplitude ratio and the negative candidate amplitude ratio to obtain the relative difference of amplitude ratio; determine the response window whose relative difference of amplitude ratio is not greater than the proportional consistency threshold as the bilaterally consistent response window; calculate the average value of the positive candidate amplitude ratio and the negative candidate amplitude ratio corresponding to the bilaterally consistent response window to obtain the perturbation amplitude ratio; scale the trifluoroacetic acid paired perturbation template that is consistent with the response direction order of the bilaterally consistent response window according to the perturbation amplitude ratio to obtain the trifluoroacetic acid ripple component corresponding to the candidate component perturbation event.

9. A noise suppression system for acquiring signals in a liquid chromatograph, used to perform the method according to any one of claims 1 to 8, characterized in that, The system includes: The data acquisition module is used to acquire ultraviolet acquisition signals, gradient organic phase ratio, organic phase pump operating status parameters, and mixer volume during elution. The disturbance screening module is used to determine the pumping time and pumping volume based on the operating status parameters of the organic phase pump; determine the organic phase volume period based on the gradient organic phase ratio of the pumping volume and the corresponding pumping time; and identify pumping events with an organic phase volume period greater than the mixer volume as candidate composition disturbance events. The response localization module is used to divide the gradient composition interval according to the gradient organic phase ratio, perform event triggering segmentation of the ultraviolet acquisition signal according to the pumping time of the candidate component perturbation event in the interval, and determine the event response delay and response window based on the repeated adjacent reverse responses in multiple signal segments. The template construction module is used to perform local baseline correction on the response window, calculate the integral area difference and response interval of the positive and negative responses, and select multiple response windows whose integral area difference meets the area difference threshold and whose response interval meets the interval consistency condition as reference response windows; and determine the trifluoroacetic acid paired perturbation template and template matching threshold based on the multiple reference response windows. The ripple suppression module is used to match the positive and negative response segments of each response window with the corresponding side of the trifluoroacetic acid paired perturbation template, and determine the response segments that meet the template matching threshold as valid response segments; the trifluoroacetic acid paired perturbation template is scaled according to the amplitude ratio of the valid response segment and the corresponding side template to obtain the trifluoroacetic acid ripple component and subtract it from the ultraviolet acquisition signal; when no valid response segment is obtained, the ultraviolet acquisition signal of the corresponding response window is retained.

Images