CN103439440B - A kind of prediction method for retention time of high performance liquid chromatographic peak - Google Patents

Abstract

The invention relates to a method for predicting peak retention time of high performance liquid chromatography. The method includes: measuring the standard retention time of various components of various samples, selecting two components in the target components of each sample as the double-labeled reference components of the sample, and obtaining the test results of the double-labeled reference components in the sample to be tested. The measured retention time in the product solution is obtained, and the measured retention time of other target components is obtained, and two-point verification and multi-point verification are performed. The high performance liquid chromatography peak retention time prediction method provided by the present invention can accurately predict the retention time of the chromatographic peaks of various components of the sample to be tested, and then perform qualitative analysis on the chromatographic peaks of the sample to be tested and identify the sample to be tested. The method provided by the invention has high predictive accuracy and a large number of applicable chromatographic columns, which is obviously superior to the existing relative retention time method.

Description

A high performance liquid chromatography peak retention time prediction method

technical field

The invention relates to a high performance liquid chromatography peak retention time prediction method, which belongs to the technical field of high performance liquid chromatography detection.

Background technique

One of the core tasks in the content determination of multi-index components of traditional Chinese medicine, fingerprint analysis, and related substance inspection of chemical drugs (or antibiotics) is to characterize a large number of target compounds. If pure chemical reference substances are used, the standard The preparation and supply of substances bring great difficulties and will also significantly increase the cost of testing. Therefore, it is of great significance and application value to use a small amount of chemical reference substances to characterize other chromatographic peaks by a certain method.

At present, the relative retention time method using a chemical reference substance is the most commonly used chromatographic peak qualitative method, which is widely used in the inspection of related substances in European, American and Chinese pharmacopoeias, and is also often used in the multi-index content determination research of the alternative mode. However, due to the difference between the chromatographic instrument and the chromatographic column, the deviation between the predicted value of the retention time and the measured value is often large. In practical applications, it is necessary to limit the type of chromatographic column, or choose the same type of chromatographic column with similar separation performance. inconvenient.

Contents of the invention

In order to solve the above-mentioned technical problems, the object of the present invention is to provide a method for predicting the peak retention time of high performance liquid chromatography, and carry out chromatographic peak qualitative by two chemical reference substances, compared with relative retention time method, this method can significantly reduce qualitative error, increase the number of columns used.

In order to achieve the above object, the present invention provides a high performance liquid chromatography (HPLC) peak retention time prediction method, which can be called a double-standard linear correction method, which includes the following steps:

(1) Use high-performance liquid chromatography to measure the actual retention time of various components of various samples on multiple different chromatographic columns. Perform linear fitting in two groups, draw the standard curve and calculate the correlation coefficient r, and eliminate the measured retention time (measured by the chromatographic column) r<0.995, the measured retention time (measured by the chromatographic column) with a large deviation and the measured retention time (measured by the chromatographic column) Measured) the actual measured retention time that leads to a change in the order of the chromatographic peaks, and the average value of the measured retention times measured by the remaining chromatographic columns is used as the standard retention time of various components;

Alternatively, the average of the measured retention times of the same component of the same sample measured on all chromatographic columns is used as the preliminary standard retention time, and the value of the preliminary standard retention time of all components is the same as that of all components of the same sample measured on all chromatographic columns. The measured retention times were linearly fitted, the standard curves were drawn and the correlation coefficient r was calculated, and the measured retention times (measured by the chromatographic column) r<0.995, the measured retention times (measured by the chromatographic column) with large deviations and ( chromatographic column) lead to the measured retention time of chromatographic peak sequence change, the average value of the measured retention time measured by the rest of the chromatographic column is used as the standard retention time of various components;

(2) Select two components among the target components of each sample as the double standard control components of the sample;

(3) Run the test solution of the sample to be tested and the reference solution of the double-standard control component on a chromatographic column and use high-performance liquid chromatography to obtain the chromatogram of the test solution. The reference substance of the component is qualitative to the chromatographic peak in the obtained chromatogram, obtains the measured retention time of the double standard control component in need testing solution;

(4) Establish a linear curve with the standard retention time of the double-labeled control component as the ordinate and the measured retention time as the abscissa, and obtain a two-point linear equation;

(5) Bring the standard retention time of the target component of the test sample except the double-labeled control component into the two-point linear equation to obtain the predicted retention time of the corresponding target component. Other chromatographic peaks in the chromatogram were qualitatively obtained to obtain the measured retention time of the corresponding target component;

If the absolute value of the deviation between the standard retention time and the measured retention time of all the target components of the test sample except the double standard control component is >t ₂ , then the prediction fails, replace the chromatographic column in step (3), and Repeat steps (3)-(5);

If the absolute value of the deviation between the standard retention time and the measured retention time of at least one target component among the target components of the test sample except the double-standard control component is ≤t ₂ , proceed to step (6);

(6) With the standard retention time of all target components whose deviations between the measured retention time and the predicted retention time are ≤ t ₂ and the standard retention time of the double-labeled control component as the ordinate, and the measured retention time as the abscissa, establish a multi-point relationship curve and obtain multiple point linear equation;

(7) Bring the standard retention time of the target component of the sample to be tested except the double-labeled control component into the multi-point linear equation, and calculate the new predicted retention time of the corresponding target component;

(8) Find the chromatographic peak closest to the new predicted retention time in the chromatogram of the test product, obtain the new measured retention time of the corresponding target component, and calculate the new measured retention time and new predicted retention time of each target component The absolute value of the deviation of the retention time to judge whether the multi-point linear equation includes all the target components: if the absolute value of all deviations ≤ t ₂ , go to the next step, if the absolute value of deviation > t ₂ , return In step (6), the standard retention time of the target components whose absolute value of deviation ≤ _t2 calculated in step (8) and the double-standard control component is used as the ordinate, and the measured retention time is used as the abscissa to re-establish a multi-point relationship curve and To get a new multi-point linear equation, repeat steps (6)-(8);

(9) Check whether the absolute value of the deviation between the standard retention time of all target components and the new predicted retention time is ≤t _m , where t _m <t ₂ : if yes, the prediction is successful; if not, the prediction fails; prediction fails , replace the chromatographic column in step (3), and repeat steps (3)-(9).

The purpose of the above-mentioned method provided by the present invention is to obtain the retention time of the chromatographic peaks of the various components of the sample to be tested, and to qualitatively carry out the chromatographic peaks, and utilize two chemical reference substances to achieve the qualitative determination of multiple target components in the sample to be tested. , the applicable samples can be traditional Chinese medicine, natural plant medicine, chemical medicine, food (such as health food) or antibiotics, etc., but not limited to this, any field that requires multi-component qualitative analysis by HPLC can be applied.

In the above method, preferably, when utilizing a high performance liquid chromatograph to measure the actual retention time of various components of various samples on multiple chromatographic columns, each chromatographic column is injected at least 3 times, and the obtained The RSD of the obtained measured retention time should be ≤2%.

In the above method, preferably, the standard retention time of the selected components of the double standard reference substance is close to both ends of the retention time period of the target component of the corresponding sample. The retention time period here refers to the retention time range of the chromatographic peaks of various components.

In the above method, preferably, the retention time period of the target component of the sample is equally divided into four segments, and the standard retention time of the selected double standard reference substance component is located near the first cut-off point and the third cut-off point.

In the above method, preferably, when selecting the standard retention time of the components of the double-standard reference substance, components that deviate greatly during the fitting process of the chromatographic column are not selected, and components that are cheap and easy to obtain are preferred.

In the above method, preferably, t ₂ <1 min, t _m <1 min; more preferably, t ₂ is 0.7 min, and t _m is 0.5 min.

In the above method, the more chromatographic columns are used for determination, the more data can be obtained, and the more data added, the smaller the error, but there is also an effect of diminishing returns. Preferably, in step (1), The standard retention time of various components is calculated from the measured retention time measured by 10-20 chromatographic columns. The standard retention time thus obtained is representative enough. When selecting the chromatographic column for the measurement of the measured retention time at the beginning, it should be Choose more chromatographic columns to meet this requirement of the data measured by using 10-20 chromatographic columns in the calculation here.

According to a specific embodiment of the present invention, the above-mentioned high performance liquid chromatography peak retention time prediction method can be carried out in the following manner, and its specific process can be as shown in Figure 1:

1. Column fitting

(1) Use high-performance liquid chromatography to measure the measured retention time of various components of various samples on multiple different chromatographic columns (reserve at least 3 decimal places), and the detection can use optimized chromatographic conditions (including column temperature, The type and ratio of flow rate and mobile phase can be selected according to the records of the Pharmacopoeia), and the sample should be injected at least three times during the detection process, and the RSD of the retention time should be satisfied≤2%, so as to control the error of the collected data;

Carry out linear fitting for the measured retention time of various components of the same sample measured on different chromatographic columns in two groups, draw the standard curve and calculate the correlation coefficient r, and eliminate the chromatographic column with large deviation and even cause the chromatographic peak sequence Changed columns, in general, these columns do not fit well with other columns, as shown by r < 0.995 or a certain (or some) points deviate significantly from the fitted line; removing these data can make The obtained standard retention time and other data are more accurate; when eliminating, you can classify the chromatographic columns, put together the ones that fit better with each other, and then select, try to choose a group with a large number;

After the elimination, the average retention time measured by the un-eliminated chromatographic column is used as the standard retention time (SRT for short) of each component, and the standard retention time is used as the predicted reference value, which should be calculated using The data measured by 10-20 chromatographic columns are carried out.

(2) Select two components among the target components of each sample as the double-standard control components of the sample; theoretically, any two components in the target should have the same effect as double standards, but in fact due to the chromatographic Differences in columns and chromatographs, structural differences between components, complexity of elution conditions, and occasional errors in chromatographic analysis, the choice of double standards will still cause differences. In addition, the difficulty and price of the reference substance must be taken into account. factor. The following are the principles that should be followed and taken into account in the selection of double standards: The double standards should be distributed at both ends of the retention time period of the target compound as much as possible. Nearby; try not to choose components that deviate greatly during the fitting process of the chromatographic column for double-standard components; try to choose cheap and easy-to-obtain components for double-standard components.

After obtaining the SRT of various components of various samples and the double-labeled reference components of various samples, they can be collected together as a standard set or pharmacopoeia. When it is necessary to predict the retention time of a chromatographic peak of a certain sample, you can Query the corresponding SRT and double-standard reference components from the standard set or pharmacopoeia, and then follow the steps below to predict:

2. Two predictions:

(3) Run the test solution of the sample to be tested and the reference solution of the double-standard reference component on a chromatographic column (it can be any chromatographic column, not necessarily the chromatographic column used when establishing the above-mentioned standard set or pharmacopoeia) And utilize the high-performance liquid chromatograph to detect, obtain the chromatogram of need testing solution, carry out qualitative to the chromatographic peak in the acquired chromatogram by the reference substance of double standard contrast composition, obtain double mark contrast composition in need testing solution The measured retention time in

(4) Establish a linear curve with the standard retention time of the double-labeled control component as the ordinate and the measured retention time as the abscissa, and obtain a two-point linear equation;

(5) Bring the standard retention time of the target component of the test sample except the double-labeled control component into the two-point linear equation to obtain the predicted retention time of the corresponding target component. Other chromatographic peaks in the chromatogram were qualitatively obtained to obtain the measured retention time of the corresponding target component;

If the absolute value of the deviation between the standard retention time and the measured retention time of all the target components in the test sample is >t ₂ , except for the double standard control component, the prediction fails, replace the chromatographic column in step (3), and repeat Steps (3)-(5);

If the absolute value of the deviation between the standard retention time and the measured retention time of at least one target component in the sample to be tested is ≤t ₂ , proceed to step (6) except for the double-standard control component.

3. Multi-point prediction:

(6) With the standard retention time of all target components whose deviations between the measured retention time and the predicted retention time are ≤ t ₂ and the standard retention time of the double-labeled control component as the ordinate, and the measured retention time as the abscissa, establish a multi-point relationship curve and obtain multiple point linear equation;

(7) Bring the standard retention time of the target component of the sample to be tested except the double-labeled control component into the multi-point linear equation, and calculate the new predicted retention time of the corresponding target component;

(8) Find the chromatographic peak closest to the new predicted retention time in the chromatogram of the test product, obtain the new measured retention time of the corresponding target component, and calculate the new measured retention time and new predicted retention time of each target component The absolute value of the deviation of the retention time to judge whether the multi-point linear equation includes all the target components: if the absolute value of all deviations ≤ t ₂ , enter the next step, if the absolute value of deviation > t ₂ , return Step (6), even if there is only one absolute value > t ₂ , return to step (6), and the standard retention time between the target component and the double-standard control component whose absolute value of all deviations ≤ t ₂ calculated in step (8) is The ordinate and the measured retention time are the abscissa to re-establish a multi-point relationship curve and obtain a new multi-point linear equation, repeat steps (6)-(8) until the established multi-point linear equation can include all points;

When the multi-point linear equation is established at the beginning of step (6), there may be part of the data of the target component that is not used because the absolute value of the deviation>t ₂ , when it is brought into the multi-point linear equation in step (7) When using the equation, the absolute value of the difference between the calculated new predicted retention time and the standard retention time may become ≤ t ₂ , therefore, when the absolute value of the deviation > t ₂ and return to step (6), The data used when establishing a new multi-point linear equation will be different from the previous ones, and the subsequent process will also be different. When all the data can satisfy the absolute value of the deviation ≤ t ₂ , you can enter the subsequent steps (9 ).

(9) Check whether the absolute value of the deviation between the standard retention time of all target components and the new predicted retention time is ≤t _m , and t _m <t ₂ : if yes, the prediction is successful; if not, the prediction fails; prediction fails , replace the chromatographic column in step (3), and repeat steps (3)-(9).

The accuracy of prediction can be improved through subsequent multi-point prediction. In addition, t ₂ >t _m can achieve the effect of wide entry and strict exit. On the one hand, it increases the number of applicable columns, and on the other hand, it also ensures the accuracy of prediction. As for the specific values of t ₂ and t _m above, it can be adjusted according to the situation. Adjustment, generally not more than 1min.

The high performance liquid chromatography peak retention time prediction method provided by the present invention can accurately predict the retention time of the chromatographic peaks of various components of the sample to be tested, and then perform qualitative analysis on the chromatographic peaks of the sample to be tested and identify the sample to be tested. The method provided by the invention has high predictive accuracy and a large number of applicable chromatographic columns, which is obviously superior to the existing relative retention time method.

Description of drawings

Fig. 1 is a flow chart of the method for predicting peak retention time of high performance liquid chromatography provided by the present invention.

Figure 2 is the fitted line graph of the two chromatographic columns.

Fig. 3a-Fig. 3c are the fitting results of 5 kinds of anthraquinones in rhubarb on 10 chromatographic columns.

Figures 4a-4c are linear relationship diagrams of the retention times of the five components of rhubarb measured on different chromatographic columns.

Fig. 5 is the chromatogram of need testing solution.

Detailed ways

In order to have a clearer understanding of the technical features, purposes and beneficial effects of the present invention, the technical solution of the present invention is described in detail below, but it should not be construed as limiting the scope of implementation of the present invention.

Example

This embodiment provides a high performance liquid chromatography peak retention time prediction method, which performs prediction of the chromatographic peak retention time of rhubarb, including the following steps:

1. Column fitting:

According to the optimized chromatographic conditions (according to the "Chinese Pharmacopoeia" 2010 edition a rhubarb content determination items (page 22), the specific parameters are methanol-0.1% phosphoric acid aqueous solution (85:15), detection wavelength 254nm, column temperature 30°C, flow rate 1.0mL·min ^-1 ), inject mixed reference substance or test solution on 11 C18 columns of different brands and models, and obtain five components (aloe-emodin, rhein, emodin , chrysophanol, emodin methyl ether) measured retention time on different chromatographic columns, inject samples at least three times and keep three decimal places, and remove chromatographic columns with RSD greater than 2%; there are 12 chromatographic columns used, respectively : Kromasil C18*, Sunfire C18**, AgilentTC-C18*, Inertsil C18**, Hyperisil C18*, Vensil MP C18**, Shimadzu VP C18**, Phenomen luna C18**, Agilent HC-C18*, Agilent Zorbax SB-C18*, Shiseido AQ C18*, Shiseido MG C18*, the above C18 columns are referred to as C1 to C12 (*.4.6mm×250mm×5μm, **.4.6mm×150mm×5μm), of which C1 to C11 are used for Column fitting, C12 is used as an unknown column for method validation (not involved in the calculation of SRT);

Perform linear fitting (two-two combination) of the measured retention times of the above five components measured by the C1 to C11 chromatographic columns, draw a standard curve, calculate the correlation coefficient r, and then eliminate the chromatographic columns that deviate greatly or even cause the order of the chromatographic peaks to change ( For example, the chromatographic column with r<0.995, and the chromatographic column whose individual points obviously deviate from the fitting line); for example, as shown in Figure 2, it is the fitting line of the two chromatographic columns C2 and C11. The two points (rhein) deviate greatly from the C2 chromatographic column. The two chromatographic columns cannot be put together, and one should be eliminated; through fitting, C1 to C10 are retained, and C11 is eliminated.

2. Determination of SRT:

On the two chromatographic systems, the content determination method under the rhubarb item of the "Chinese Pharmacopoeia" 2010 edition was used to simultaneously determine the five free anthraquinone components of aloe-emodin, rhein, emodin, chrysophanol and emodin methyl ether (above The 5 components are sorted by retention time from small to large), system 1: Shimadzu 2010A (corresponding to columns 1 and 2); system 2: waterse2695 (corresponding to columns 3 and 4) and 4 different brands and models of chromatographic columns, column 1 : kromasil C18 (250mm×4.6mm×5μm); Column 2: waters sunfire C18 (150mm×4.6mm×5μm); Column 3: agilentTC-C18 (250mm×4.6mm×5μm); Column 4: inertial C18 (150mm× 4.6mm×5μm), to verify the accuracy and applicability of the conclusion, the results are shown in Figure 4a-Figure 4c, where the horizontal and vertical coordinates are the retention times of the five components on any two chromatographic columns.

Fig. 3a-Fig. 3c are the fitting results of 5 kinds of anthraquinones of rhubarb on 10 chromatographic columns, the abscissa is the SRT of 5 kinds of anthraquinones, the ordinate is the measured retention time on each chromatographic column, and the linear equation of each chromatographic column and correlation coefficients are shown in Table 1.

Table 1 Linear equation and correlation coefficient of retention time on different chromatographic columns

Chromatographic column linear equation correlation coefficient C1 Y=1.7985x-2.5911 0.9987 C2 Y=0.7319x+0.1108 0.9998 C3 Y=0.6900x+2.4103 0.9962 C4 Y=1.0901x-0.9971 0.9985 C5 Y=1.0790x-0.0185 0.9987 C6 Y=1.0438x-0.0284 0.9963 C7 Y=0.9682x-0.3483 0.9998 C8 Y=0.5612x+0.7563 0.9982 C9 Y=1.1173x+0.4010 0.9999 C10 Y=0.8448x+0.6422 0.9978

Comparing Figure 3a-Figure 3c and Figure 4a-Figure 4c, it can be seen that the r of C2 increased from 0.9976 (r ² =0.9951) to 0.9998, the r of C4 increased from 0.9974 (r ² =0.9947) to 0.9985, and the r of C3 increased from 0.9977 (r ² =0.9954) decreased to 0.9962, which is a small probability event. From the overall data point of view, the advantage of using SRT is obvious.

The average value of the measured retention time measured by the above-mentioned 10 chromatographic columns that has not been excluded is used as the SRT of various components, and the SRT is used as the predicted reference value. The SRTs of the five components are respectively: aloe-emodin 4.081min, rhein 4.979min, emodin 7.380min, chrysophanol 9.385min and emodin methyl ether 12.715min, double standard selection aloe-emodin with the shortest retention time and emodin methyl ether with the longest retention time.

3. Two-point prediction

First, run the double standard reference solution (i.e. aloe-emodin reference solution and emodin reference substance) and the test solution (i.e. the solution of the sample to be tested) on a C18 column (C12, Shiseido MG C18*) , the chromatographic peaks in the chromatogram of the test product (as shown in Figure 5) were qualified by the reference substance, so as to obtain the measured retention time of the double-labeled components in the test product: aloe-emodin 4.970min, emodin methyl ether 16.000 min;

Then, with SRT as the vertical axis and the measured retention time as the horizontal axis, two points are obtained: aloe-emodin (4.970, 4.081) and emodin methyl ether (16.000, 12.715), and then a two-point linear equation is obtained: Y=0.7828X+ 0.1903;

Substitute the SRT values (corresponding Y values) of the other three anthraquinones into the above equation to obtain the predicted retention time: 6.117 min for rhein, 9.185 min for emodin, and 11.746 min for chrysophanol;

Finally, look for the chromatographic peak closest to the predicted value in the chromatogram of the test product (a is the chromatogram of the reference substance solution, b is the chromatogram of the rhubarb test solution, 1 and 5 are double standards, 2, 3, and 4 are target peak), and obtained the measured retention time, which are: rhein 5.978min, emodin 9.218min and chrysophanol 11.843min, the absolute value of the deviation between the measured retention time and predicted retention time is 0.139min, 0.033min and 0.097min in turn .

4. Multi-point prediction:

(1) Deviation judgment: The absolute values of the deviations of rhein, emodin, and chrysophanol obtained above are all less than 0.7min, so the prediction effect is good, and these three target substances are all involved in fitting multi-point linear equations;

(2) Fitting a multi-point linear equation: take SRT as the ordinate, and the measured retention time as the abscissa, and get 5 points: aloe-emodin (4.970, 4.081), rhein (5.978, 4.979), emodin (9.218, 7.380), chrysophanol (11.843, 9.385) and emodin methyl ether (16.000, 12.715), and then get a 5-point linear equation: Y=0.7759X+0.2580;

(3) Substitute the SRT values of the remaining three anthraquinones (rhein, emodin, and chrysophanol) into the equation again to obtain new predicted retention times: 6.085 min for rhein, 9.179 min for emodin, and 11.763 min for chrysophanol;

(4) Re-search for the chromatographic peaks closest to the new predicted retention times of the three anthraquinones (rhein, emodin, and chrysophanol) in the chromatogram of the test product, which are still peaks 2, 3, and 4. And get the new measured retention time of three kinds of anthraquinones (rhein, emodin, chrysophanol): rhein 5.978min, emodin 9.218min and chrysophanol 11.843min;

(5) Check whether the fitting of the multi-point linear equation includes all targets, that is, check whether the absolute values of the prediction deviations of all targets are ≤0.5min, and the three kinds of anthraquinones (rhein, emodin, chrysophanol) The absolute values of prediction deviations are 0.107min, 0.033min and 0.080min in turn, all less than 0.5min, therefore, the prediction is successful.

Comparison of prediction effect between double standard linear correction method and relative retention time method

Taking the analysis of five anthraquinones in rhubarb as an example, the advantages and disadvantages of the two prediction methods were compared. Tables 2 and 3 are the comparison results of the two methods on 10 chromatographic columns (the prediction of the relative retention time method takes emodin as a reference, and the average values of the relative retention times are 0.556, 0.678, 1.266 and 1.709), among which, in Table 3 The second column is the maximum absolute deviation between the predicted retention time and the measured retention time; the third column is the corresponding maximum relative deviation; the fourth column is the percentage of the total number of absolute deviations less than or equal to 0.5min; the fifth column is the applicable column Total number (when the absolute deviation of the predicted values of each component on a chromatographic column is less than or equal to 0.5min, the column is considered applicable); the sixth and seventh columns are the corresponding relative deviation percentage and the number of applicable columns.

Table 2 Absolute deviation/min of predicted value of retention time

Table 3 Comparison of the prediction results of the two methods

From the results shown in Table 2 and Table 3, it can be seen that the prediction accuracy of the double-standard linear calibration method is higher, and the number of applicable chromatographic columns is larger, which is obviously better than the relative retention time method.

Claims (7)

1. A high performance liquid chromatography peak retention time prediction method, comprising the following steps: (1) Utilize high-performance liquid chromatography to measure the actual retention time of various components of various samples on multiple different chromatographic columns, two groups of the actual measurement retention time of various components of the same sample measured on different chromatographic columns Carry out linear fitting in two groups, draw the standard curve and calculate the correlation coefficient r, and eliminate the measured retention time measured by the chromatographic column with r<0.995, and the measured retention time measured by the chromatographic column whose one or some points obviously deviate from the fitting line during the fitting. The measured retention time and the measured retention time of the chromatographic column that cause the order of the chromatographic peak to change, the average value of the measured retention time measured by the remaining chromatographic columns is used as the standard retention time of various components; Alternatively, the average of the measured retention times of the same component of the same sample measured on all chromatographic columns is used as the preliminary standard retention time, and the value of the preliminary standard retention time of all components is the same as that of all components of the same sample measured on all chromatographic columns. The measured retention time was linearly fitted, respectively, the standard curve was drawn and the correlation coefficient r was calculated, and the measured retention time measured by the chromatographic column with r<0.995, and the chromatographic column whose one or some points were obviously deviated from the fitting line during the fitting were eliminated. The actual measured retention time of the determination and the measured retention time of the chromatographic column that cause the order of the chromatographic peak to change, the average value of the measured retention time of the remaining chromatographic columns is used as the standard retention time of various components; (2) Select two components in the target components of each sample as the double standard reference components of the sample, and the standard retention time of the selected double standard reference substance components is close to the two ends of the retention time period of the target component of the sample. Select components that deviate significantly from the fitting line during the column fitting process; (3) run the need testing solution of the sample to be tested and the reference substance solution of the double-standard reference component on a chromatographic column and utilize high-performance liquid chromatography to detect, obtain the chromatogram of the need testing solution, and pass the double-standard control The reference substance of the component is qualitative to the chromatographic peak in the obtained chromatogram, obtains the measured retention time of the double standard control component in need testing solution; (4) set up a linear curve with the standard retention time of the double-standard reference component as the ordinate, and the measured retention time as the abscissa, and obtain a two-point linear equation; (5) Bring the standard retention time of the target component of the test sample except the double standard reference component into the two-point linear equation to obtain the predicted retention time of the corresponding target component, and according to the predicted retention time to the test product Other chromatographic peaks in the chromatogram were qualitatively obtained to obtain the measured retention time of the corresponding target component; If the absolute value of the deviation between the predicted retention time and the measured retention time of all the target components of the sample to be tested except the double standard control component is >t ₂ , then the prediction fails, and the chromatographic column in step (3) is replaced, and Repeat steps (3)-(5); If the absolute value of the deviation between the predicted retention time and the measured retention time of at least one target component among the target components of the test sample except the double-labeled control component is ≤ t ₂ , proceed to step (6); (6) The standard retention time of all the target components whose deviations between the actual measured retention time and the predicted retention time are ≤ t ₂ and the standard retention time of the double-standard control component are used as the ordinate, and the measured retention time is used as the abscissa to establish a multi-point relationship curve and obtain a multi-point relationship curve. point linear equation; (7) Bring the standard retention time of the target component of the sample to be tested except the double standard control component into the multi-point linear equation, and calculate the new predicted retention time of the corresponding target component; (8) Find the chromatographic peak closest to the new predicted retention time in the chromatogram of the test product, obtain the new measured retention time of the corresponding target component, and calculate the new measured retention time and new predicted retention time of each target component The absolute value of the deviation of the retention time to judge whether the multi-point linear equation includes all the target components: if the absolute value of all deviations ≤ t ₂ , enter the next step, if the absolute value of deviation > t ₂ , return Step ( ₆ ), the target component with the absolute value of all deviations≤t2 calculated in step (8) and the standard retention time of the double-standard control component are the ordinate, and the measured retention time is the abscissa to re-establish a multi-point relationship curve and To obtain a new multipoint linear equation, repeat steps (6)-(8); (9) Detect whether the absolute value of the deviation between the measured retention time of all target components and the new predicted retention time is ≤t _m , and t _m <t ₂ : if yes, the prediction is successful; if not, the prediction fails; the prediction fails , replace the chromatographic column in step (3), and repeat steps (3)-(9). 2. The HPLC peak retention time prediction method according to claim 1, wherein the sample is Chinese medicine, natural herbal medicine, chemical medicine, food or antibiotic. 3. high performance liquid chromatography peak retention time prediction method according to claim 1, wherein, when utilizing high performance liquid chromatography to measure the various components of various samples on the measured retention time of many different chromatographic columns, Each chromatographic column was injected at least 3 times, and the RSD of the obtained measured retention time was ≤2%. 4. The high performance liquid chromatography peak retention time prediction method according to claim 1, wherein, when selecting the standard retention time of the double-standard reference substance components, select cheap and easy-to-obtain components. 5. The method for predicting peak retention time of high performance liquid chromatography according to claim 1, wherein said t ₂ <1 min, t _m <1 min. 6. The high performance liquid chromatography peak retention time prediction method according to claim 5, wherein said t is _0.7min , and said _t is 0.5min. 7. the high performance liquid chromatography peak retention time prediction method according to claim 1, wherein, in step (1), the standard retention time of various compositions is that the measured retention time of the 10-20 chromatographic columns that adopts measures carries out calculate.