Cloud computing analysis system and method based on high performance liquid chromatograph

Abstract

The invention discloses a cloud computing analysis system and method based on a high performance liquid chromatograph, and belongs to the technical field of analytical chemistry and information. The system comprises a data acquisition module, a data transmission module, a cloud storage module, a data analysis module, a user interface module and a real-time analysis and feedback module. The data acquisition module is used for acquiring data detected by the high performance liquid chromatograph; the data transmission module is used for determining a protocol adopted by data transmission; the cloud storage module is used for storing the data transmitted to the cloud; the data analysis module is used for training an intelligent clustering analysis model and inputting data for automatic data analysis; the user interface module is used for providing an interface for a user to check an analysis result; and the real-time analysis and feedback module is used for carrying out real-time analysis and judgment on the data. Automatic data analysis is realized by using a machine learning training model, and manual intervention is reduced.

Description

Technical Field

[0001] The present invention relates to the fields of analytical chemistry and information technology, and in particular to a cloud computing analysis system and method based on a high performance liquid chromatograph. Background Art

[0002] High Performance Liquid Chromatography (HPLC) is an advanced analytical instrument used to separate, identify and quantitatively analyze components in a mixture.

[0003] The current mainstream methods of chemical gas analysis include chemiluminescence detection and spectral analysis. However, chemiluminescence detection relies on specific reagents and optical equipment, requires real-time monitoring by detection personnel, and requires a lot of manual intervention, resulting in low detection efficiency. Spectral analysis is often interfered by gas concentration and background noise during the detection process, resulting in insufficient sensitivity and affecting the accuracy of the results. These two methods are highly dependent on manual operation in the detection of gas classification, are difficult to automate, and lack the support of big data, which further limits the efficiency and accuracy of detection. Summary of the invention

[0004] The object of the present invention is to provide a cloud computing analysis system and method based on high performance liquid chromatography to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: A cloud computing analysis system based on high performance liquid chromatography, characterized in that: the system includes a data acquisition module, a data transmission module, a cloud storage module, a data analysis module, a user interface module and a real-time analysis and feedback module; The data acquisition module is used to collect data detected by the high-performance liquid chromatograph and make baseline adjustments to the data; the data transmission module is used to determine the protocol used for data transmission, and at the same time compress and encrypt the data after the baseline adjustment; the cloud storage module is used to store the data transmitted to the cloud and provide data backup and recovery options; the data analysis module is used to decrypt and decompress the data, input it into the trained intelligent clustering analysis model for data analysis, and use artificial intelligence algorithms to optimize the model to improve analysis efficiency and accuracy; the user interface module is used to provide users with an interface for viewing data analysis results and record logs; the real-time analysis and feedback module is used to analyze the data in real time, and make judgments when the analysis returns the results. If the error between the analysis result and the predicted result exceeds the sum of square errors, the analysis plan and model parameters are immediately adjusted for secondary analysis; The baseline adjustment is to remove noise and make the signal show a peak value; The output end of the data acquisition module is connected to the input end of the data transmission module; the output end of the data transmission module is connected to the input end of the cloud storage module; the output end of the cloud storage module is connected to the input end of the data analysis module; the output end of the data analysis module is connected to the input end of the user interface module; the output end of the user interface module is connected to the input end of the real-time analysis and feedback module.

[0006] The data acquisition module includes an HPLC instrument interface unit and a data processing unit; The HPLC instrument interface unit is used to provide an interface between the data acquisition module and the HPLC instrument, and transmit the data detected by the HPLC instrument to the analysis system; the data processing unit is used to perform baseline adjustment on the data; The output end of the HPLC instrument interface unit is connected to the input end of the data processing unit; the output end of the data processing unit is connected to the input end of the data transmission module.

[0007] The data transmission module includes a transmission protocol unit, a data compression unit and a data encryption unit; The transmission protocol unit adopts a secure transmission protocol HTTPS to ensure the security and integrity of data upload; the data compression unit is used to compress the collected data; the data encryption unit is used to encrypt and transmit the compressed data; The output end of the transmission protocol unit is connected to the input end of the data compression unit; the output end of the data compression unit is connected to the input end of the data encryption unit; the output end of the data encryption unit is connected to the input end of the cloud storage module.

[0008] The cloud storage module includes a data storage unit and a backup and recovery unit; The data storage unit is used to store the received data in the cloud; the backup and recovery unit is used to add a data backup and recovery mechanism, so that the data can be reviewed and downloaded for use when the data is abnormal or lost; The output end of the data storage unit is connected to the input end of the backup and recovery unit.

[0009] The data analysis module includes a data decryption unit, a data decompression unit, a data preprocessing unit, an intelligent clustering analysis model unit, a multimodal data fusion unit, a model optimization unit and an analysis result visualization unit; The data decryption unit is used to decrypt the encrypted data; the data decompression unit is used to decompress the decrypted data to obtain the transmitted data; the data preprocessing unit is used to perform baseline correction, peak detection and peak integration on the data; the intelligent clustering analysis model unit is used to transmit the data to the trained model for automatic analysis; the multimodal data fusion unit is used to fuse the high performance liquid chromatography data with the mass spectrometry data, the ultraviolet-visible spectrum data and the infrared spectrum data, and improve the accuracy and reliability of the analysis results through multimodal data analysis; the model optimization unit is used to adjust the parameters of the model according to the gas analysis data in the database; the result visualization unit is used to display the results of the data analysis in a visual way; The baseline correction is to remove the baseline drift from the chromatographic data and quantify and qualify the peaks in the chromatogram; The peak detection is to identify the peak from the corrected signal; The peak integration is to calculate the area of ​​each detected peak; The output end of the data decompression unit is connected to the input end of the data decompression unit; the output end of the data decompression unit is connected to the input end of the data preprocessing unit; the output end of the data preprocessing unit is connected to the input end of the intelligent clustering analysis model unit; the output end of the intelligent clustering analysis model unit is connected to the input end of the multimodal data fusion unit; the output end of the multimodal data fusion unit is connected to the input end of the model optimization unit; the output end of the model optimization unit is respectively connected to the input ends of the analysis result visualization unit and the intelligent clustering analysis model unit; the output end of the analysis result visualization unit is connected to the input end of the user interface module.

[0010] The user interface module includes an authorization management unit, a user interface unit and a monitoring and logging unit; The authorization management unit is used to authenticate the identity of the user who views the analysis results; the user interface unit is an interface presented to the user; the monitoring and log unit is used to record the operation, user and time of viewing the analysis results; The output end of the authorization management unit is connected to the input end of the user interface unit; the output end of the user interface unit is connected to the input end of the monitoring and logging unit; the output end of the monitoring and logging unit is connected to the input end of the real-time analysis and feedback module.

[0011] A cloud computing analysis method based on high performance liquid chromatography, the method comprising the following steps: S1. Analyze the sample using HPLC instrument, obtain chromatographic data, eliminate noise from the data, and convert the chromatographic data after noise elimination into digital signals by the detector, and record them in the local data system; S2, compress and encrypt the locally stored chromatographic data and upload it to the cloud storage platform, using the secure transmission protocol HTTPS to ensure the security and integrity of the data upload; S3, after decrypting and decompressing the received data, use the cloud computing platform to perform automated data preprocessing and data standardization; S4. Use big data analysis tools on the cloud computing platform to perform data modeling and analysis; apply the machine learning algorithm K-means clustering to perform gas prediction and classification on the pre-processed chromatographic data, and train the intelligent clustering analysis model; S5. Combine multimodal data fusion technology to fuse high performance liquid chromatography data, mass spectrometry data, UV-visible spectroscopy data and infrared spectroscopy data, and improve the accuracy and reliability of analysis results through multimodal data analysis; S6. Add a model optimization mechanism to the intelligent clustering analysis model, adjust the model parameters in combination with the database, and optimize the model; S7. Provide an interface for users to view analysis results, add a permission management mechanism, verify the permissions of users who view analysis results, and record logs; S8. Real-time analysis and feedback. After obtaining access rights, a judgment is made at the end of the analysis. If the error between the analysis result and the prediction exceeds the sum of square errors, the scheme and model parameters are immediately adjusted for secondary data analysis.

[0012] In step S3, The preprocessing includes baseline correction, peak detection and peak integration; the data standardization is performed on the same scale for each feature through a standard normal distribution with a mean of 0 and a standard deviation of 1; The baseline correction is to remove the baseline drift from the chromatographic data and quantify and qualify the peaks in the chromatogram; The peak detection is to identify the peak from the corrected signal; The peak integration is to calculate the area of ​​each detected peak.

[0013] In step S4, the intelligent clustering analysis model training specifically includes the following steps: S9-1. Data points , , , Divided into Group , , , , so that the data points within the same group are similar and the data points between different groups are different; Centroid calculation: in, For each group The center of mass, ; Yes Group The number of data points in ; Yes Group Data points in ; S9-2, Initialization: Select Group initial centroid , , , ; S9-3. Group assignment: For each data point , calculate its distance to each centroid and assign it to the group to which the nearest centroid belongs ; S9-4. Centroid update: Calculate the new centroid of each group and update: in, For each group The center of mass, ; Yes Group The number of data points in ; Yes Group Data points in ; S9-5, iteration: repeat step S9-3 and step S9-4 until the centroid no longer changes; The silhouette coefficient method is used to determine the number of groups in K-means clustering; specifically: When the number of data points When the initial number of groups Take 2, 3 and 4 respectively, when the number of data points When the initial number of groups Take 5, 6 and 7 respectively; for each value, execute the K-means algorithm, and calculate each The silhouette coefficient corresponding to the value : in It is a data point The average distance to other data points in the same group, It is a data point The average distance to the nearest data point in the other group; in, It is a data point The group you belong to; Yes Group The number of data points in ; It is a data point and data points The Euclidean distance between in, It is a data point The group you belong to; Yes Group The number of data points in ; It is a data point and data points The Euclidean distance between calculate The corresponding silhouette coefficient , the largest The corresponding This is the number of groups in the K-means clustering.

[0014] In steps S5-S8, The multimodal data fusion technology uses a high performance liquid chromatograph to obtain high performance liquid chromatography data, a mass spectrometer to obtain mass spectrum data, a UV-visible spectrometer to obtain UV-visible spectrum data, and an infrared spectrometer to obtain infrared spectrum data, fuses the four types of data, performs data analysis multiple times, and improves the accuracy and reliability of the analysis results through multimodal data analysis; The real-time analysis and feedback: obtain access rights, make judgments after the analysis is completed, and if the error between the analysis result and the prediction result exceeds the sum of square errors, immediately adjust the scheme and model parameters to perform secondary data analysis; if the error between the secondary analysis result and the prediction result is less than the sum of square errors, select the second analysis result and record it in the database; if the error between the secondary analysis result and the prediction result is greater than or equal to the sum of square errors, and the error between the secondary analysis result and the first analysis result is less than the sum of square errors, adjust the prediction result part of the intelligent clustering analysis model and record the log; The squared error and : in, represents the sum of squared errors; It is a data point The centroid of the group to which it belongs The Euclidean distance between ; Yes Group Data points in ; It is a data point The number of

[0015] Compared with the prior art, the beneficial effects achieved by the present invention are: 1. The invention uses cloud computing to automatically analyze large amounts of complex data, greatly improving the speed of data analysis; 2. The invention adds a real-time data analysis and feedback mechanism, which can analyze the data in the database and then adjust the analysis parameters to obtain more accurate analysis results; 3. This invention uses multimodal data fusion technology. Through multimodal data fusion technology, mass spectrometry data, ultraviolet-visible spectrum data and infrared spectrum data are compared, which can further improve the performance of the analysis system and the reliability of the analysis results, and provide users with a more comprehensive analysis solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 It is a flow chart of a cloud computing analysis system and method based on high performance liquid chromatography of the present invention; Figure 2 It is a schematic diagram of the steps of a cloud computing analysis method based on high performance liquid chromatography of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] See also Figure 1-Figure 2 , the present invention provides a technical solution: a cloud computing analysis system based on high performance liquid chromatography, characterized in that: the system includes a data acquisition module, a data transmission module, a cloud storage module, a data analysis module, a user interface module and a real-time analysis and feedback module; The data acquisition module is used to collect data detected by the high-performance liquid chromatograph and make baseline adjustments to the data; the data transmission module is used to determine the protocol used for data transmission, and at the same time compress and encrypt the data after the baseline adjustment; the cloud storage module is used to store the data transmitted to the cloud and provide data backup and recovery options; the data analysis module is used to decrypt and decompress the data, input it into the trained intelligent clustering analysis model for data analysis, and use artificial intelligence algorithms to optimize the model to improve analysis efficiency and accuracy; the user interface module is used to provide users with an interface for viewing data analysis results and record logs; the real-time analysis and feedback module is used to analyze the data in real time, and make judgments when the analysis returns the results. If the error between the analysis result and the predicted result exceeds the sum of square errors, the analysis plan and model parameters are immediately adjusted for secondary analysis; The baseline adjustment is to remove noise and make the signal show a peak value; The output end of the data acquisition module is connected to the input end of the data transmission module; the output end of the data transmission module is connected to the input end of the cloud storage module; the output end of the cloud storage module is connected to the input end of the data analysis module; the output end of the data analysis module is connected to the input end of the user interface module; the output end of the user interface module is connected to the input end of the real-time analysis and feedback module.

[0019] The data acquisition module includes an HPLC instrument interface unit and a data processing unit; The HPLC instrument interface unit is used to provide an interface between the data acquisition module and the HPLC instrument, and transmit the data detected by the HPLC instrument to the analysis system; the data processing unit is used to perform baseline adjustment on the data; The output end of the HPLC instrument interface unit is connected to the input end of the data processing unit; the output end of the data processing unit is connected to the input end of the data transmission module.

[0020] The data transmission module includes a transmission protocol unit, a data compression unit and a data encryption unit; The transmission protocol unit adopts a secure transmission protocol HTTPS to ensure the security and integrity of data upload; the data compression unit is used to compress the collected data; the data encryption unit is used to encrypt and transmit the compressed data; The output end of the transmission protocol unit is connected to the input end of the data compression unit; the output end of the data compression unit is connected to the input end of the data encryption unit; the output end of the data encryption unit is connected to the input end of the cloud storage module.

[0021] The cloud storage module includes a data storage unit and a backup and recovery unit; The data storage unit is used to store the received data in the cloud; the backup and recovery unit is used to add a data backup and recovery mechanism, so that the data can be reviewed and downloaded for use when the data is abnormal or lost; The output end of the data storage unit is connected to the input end of the backup and recovery unit.

[0022] The data analysis module includes a data decryption unit, a data decompression unit, a data preprocessing unit, an intelligent clustering analysis model unit, a multimodal data fusion unit, a model optimization unit and an analysis result visualization unit; The data decryption unit is used to decrypt the encrypted data; the data decompression unit is used to decompress the decrypted data to obtain the transmitted data; the data preprocessing unit is used to perform baseline correction, peak detection and peak integration on the data; the intelligent clustering analysis model unit is used to transmit the data to the trained model for automatic analysis; the multimodal data fusion unit is used to fuse the high performance liquid chromatography data with the mass spectrometry data, the ultraviolet-visible spectrum data and the infrared spectrum data, and improve the accuracy and reliability of the analysis results through multimodal data analysis; the model optimization unit is used to adjust the parameters of the model according to the gas analysis data in the database; the result visualization unit is used to display the results of the data analysis in a visual way; The baseline correction is to remove the baseline drift from the chromatographic data and quantify and qualify the peaks in the chromatogram; The peak detection is to identify the peak from the corrected signal; The peak integration is to calculate the area of ​​each detected peak; The output end of the data decompression unit is connected to the input end of the data decompression unit; the output end of the data decompression unit is connected to the input end of the data preprocessing unit; the output end of the data preprocessing unit is connected to the input end of the intelligent clustering analysis model unit; the output end of the intelligent clustering analysis model unit is connected to the input end of the multimodal data fusion unit; the output end of the multimodal data fusion unit is connected to the input end of the model optimization unit; the output end of the model optimization unit is respectively connected to the input ends of the analysis result visualization unit and the intelligent clustering analysis model unit; the output end of the analysis result visualization unit is connected to the input end of the user interface module.

[0023] The user interface module includes an authorization management unit, a user interface unit and a monitoring and logging unit; The authorization management unit is used to authenticate the identity of the user who views the analysis results; the user interface unit is an interface presented to the user; the monitoring and log unit is used to record the operation, user and time of viewing the analysis results; The output end of the authorization management unit is connected to the input end of the user interface unit; the output end of the user interface unit is connected to the input end of the monitoring and logging unit; the output end of the monitoring and logging unit is connected to the input end of the real-time analysis and feedback module.

[0024] A cloud computing analysis method based on high performance liquid chromatography, the method comprising the following steps: S1. Analyze the sample using HPLC instrument, obtain chromatographic data, eliminate noise from the data, and convert the chromatographic data after noise elimination into digital signals by the detector, and record them in the local data system; S2, compress and encrypt the locally stored chromatographic data and upload it to the cloud storage platform, using the secure transmission protocol HTTPS to ensure the security and integrity of the data upload; S3, after decrypting and decompressing the received data, use the cloud computing platform to perform automated data preprocessing and data standardization; S4. Use big data analysis tools on the cloud computing platform to perform data modeling and analysis; apply the machine learning algorithm K-means clustering to perform gas prediction and classification on the pre-processed chromatographic data, and train the intelligent clustering analysis model; S5. Combine multimodal data fusion technology to fuse high performance liquid chromatography data, mass spectrometry data, UV-visible spectroscopy data and infrared spectroscopy data, and improve the accuracy and reliability of analysis results through multimodal data analysis; S6. Add a model optimization mechanism to the intelligent clustering analysis model, adjust the model parameters in combination with the database, and optimize the model; S7. Provide an interface for users to view analysis results, add a permission management mechanism, verify the permissions of users who view analysis results, and record logs; S8. Real-time analysis and feedback. After obtaining access rights, a judgment is made at the end of the analysis. If the error between the analysis result and the prediction exceeds the sum of square errors, the scheme and model parameters are immediately adjusted for secondary data analysis.

[0025] In step S3, The preprocessing includes baseline correction, peak detection and peak integration; the data standardization is performed on the same scale for each feature through a standard normal distribution with a mean of 0 and a standard deviation of 1; The baseline correction is to remove the baseline drift from the chromatographic data and quantify and qualify the peaks in the chromatogram; The peak detection is to identify the peak from the corrected signal; The peak integration is to calculate the area of ​​each detected peak.

[0026] In step S4, the intelligent clustering analysis model training specifically includes the following steps: S9-1. Data points , , , Divided into Group , , , , so that the data points within the same group are similar and the data points between different groups are different; Centroid calculation: in, For each group The center of mass, ; Yes Group The number of data points in ; Yes Group Data points in ; S9-2, Initialization: Select Group initial centroid , , , ; S9-3. Group assignment: For each data point , calculate its distance to each centroid and assign it to the group to which the nearest centroid belongs ; S9-4. Centroid update: Calculate the new centroid of each group and update: in, For each group The center of mass, ; Yes Group The number of data points in ; Yes Group Data points in ; S9-5, iteration: repeat step S9-3 and step S9-4 until the centroid no longer changes; The silhouette coefficient method is used to determine the number of groups in K-means clustering; specifically: When the number of data points When the initial number of groups Take 2, 3 and 4 respectively, when the number of data points When the initial number of groups Take 5, 6 and 7 respectively; for each value, execute the K-means algorithm, and calculate each The silhouette coefficient corresponding to the value : in It is a data point The average distance to other data points in the same group, It is a data point The average distance to the nearest data point in the other group; in, It is a data point The group you belong to; Yes Group The number of data points in ; It is a data point and data points The Euclidean distance between in, It is a data point The group you belong to; Yes Group The number of data points in ; It is a data point and data points The Euclidean distance between calculate The corresponding silhouette coefficient , the largest The corresponding This is the number of groups in the K-means clustering.

[0027] In steps S5-S8, The multimodal data fusion technology uses a high performance liquid chromatograph to obtain high performance liquid chromatography data, a mass spectrometer to obtain mass spectrum data, a UV-visible spectrometer to obtain UV-visible spectrum data, and an infrared spectrometer to obtain infrared spectrum data, fuses the four types of data, performs data analysis multiple times, and improves the accuracy and reliability of the analysis results through multimodal data analysis; The real-time analysis and feedback: obtain access rights, make judgments after the analysis is completed, and if the error between the analysis result and the prediction result exceeds the sum of square errors, immediately adjust the scheme and model parameters to perform secondary data analysis; if the error between the secondary analysis result and the prediction result is less than the sum of square errors, select the second analysis result and record it in the database; if the error between the secondary analysis result and the prediction result is greater than or equal to the sum of square errors, and the error between the secondary analysis result and the first analysis result is less than the sum of square errors, adjust the prediction result part of the intelligent clustering analysis model and record the log; The squared error and : in, represents the sum of squared errors; It is a data point The centroid of the group to which it belongs The Euclidean distance between ; Yes Group Data points in ; It is a data point The number of

[0028] In the embodiments of the present invention, the hardware facilities used are: high performance liquid chromatograph (HPLC), computer and data acquisition system, network connection equipment (router, modem), cloud server and cloud storage platform and data processing and analysis software; Materials used: samples to be analyzed, mobile phase solvents (water, methanol), and standard samples for calibration; Test results: test result label 1, test result label 2; Sample analysis: The sample to be analyzed is injected into the HPLC instrument and the components are separated by the chromatographic column; Data recording: The chromatogram data, including retention time, peak area and peak height parameters, are recorded through the data acquisition system; the data recording format is {sample number, retention time (min), peak area, peak height}; {"1","5.2","15000","300"} {"2","7.8","12000","250"} {"3","5.3","15500","310"} {"4","9.1","13000","270"} {"5","5.1","14900","295"} {"6","7.7","12300","255"} {"7","9.0","12800","265"} According to the silhouette coefficient method , Take 2, 3 and 4; Calculate the silhouette coefficient , and ; , , ; The number of K-means clustering groups is 2; The two groups of cluster centers are: {“ ','5.2','15000','300'} {“ ','7.8','12000','250'} , sample "1" is divided into Group; Calculate the results in sequence: Sample "1" to The distance to The distance is 3000.03, assigned to Group; Sample "2" to The distance to The distance is 0, assigned to Group; Sample "3" to The distance is:

[0029] arrive The distance is:

[0030] Assign to Group; Sample "4" to The distance is:

[0031] arrive The distance is:

[0032] Assign to Group; Sample "5" to The distance is:

[0033] arrive The distance is:

[0034] Assign to Group; Sample "6" to The distance is:

[0035] arrive The distance is:

[0036] Assign to Group; Sample "7" to The distance is:

[0037] arrive The distance is:

[0038] Assign to Group; The mean of samples "1", "3" and "5": Retention time (min):

[0039] Peak area:

[0040] Peak height:

[0041] Samples "2" "4" "6" "7": Retention time (min):

[0042] Peak area:

[0043] Peak height:

[0044] The updated two groups of cluster centers are: {“ ','5.2','15133.33','301.67'} {“ ','8.4','12525','260'} Recalculate the Euclidean distance and group: Sample "1" to The distance is:

[0045] arrive The distance is: Assign to Group; Sample "2" to The distance is:

[0046] arrive The distance is:

[0047] Assign to Group; Sample "3" to The distance is:

[0048] arrive The distance is:

[0049] Assign to Group; Sample "4" to The distance is:

[0050] arrive The distance is:

[0051] Assign to Group; Sample "5" to The distance is:

[0052] arrive The distance is:

[0053] Assign to Group; Sample "6" to The distance is:

[0054] arrive The distance is:

[0055] Assign to Group; Sample "7" to The distance is:

[0056] arrive The distance is: Assign to Group; Update and calculate each cluster center: {“ ','5.2','15133.33','301.67'} {“ ','8.4','12525','260'} No change occurs, and the iteration ends. The result is: the data record format is {sample number, retention time (min), peak area, peak height, test result label}; {"1","5.2","15000","300","Tag1"} {"2","7.8","12000","250","Label 2"} {"3","5.3","15500","310","Tag 1"} {"4","9.1","13000","270","Tag 2"} {"5","5.1","14900","295","Tag 1"} {"6","7.7","12300","255","Tag 2"} {"7","9.0","12800","265","Tag 2"} Prediction results; data record format is {sample number, retention time (min), peak area, peak height, test result label}; {"1","5.3","15000","290","Tag 1"} {"2","7.2","13000","250","Tag 2"} {"3","5.3","15500","300","Tag 1"} {"4","8.6","13500","270","Tag 2"} {"5","5.6","14700","285","Tag 1"} {"6","7.1","13300","285","Tag 2"} {"7","8.5","12100","225","Tag 2"} Calculate the sum of squared errors: in, represents the sum of squared errors; It is a data point The centroid of the group to which it belongs The Euclidean distance between ; Yes Group Data points in ; It is a data point The number of

[0057] The error between the analysis result and the prediction result is , which is less than the sum of square errors, so there is no need for secondary analysis; Analysis results: {"1","5.2","15000","300","Tag1"} {"2","7.8","12000","250","Label 2"} {"3","5.3","15500","310","Tag 1"} {"4","9.1","13000","270","Tag 2"} {"5","5.1","14900","295","Tag 1"} {"6","7.7","12300","255","Tag 2"} {"7","9.0","12800","265","Tag 2"} Gas analysis results are obtained through multimodal data fusion and data analysis feedback.

[0058] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0059] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A cloud computing analysis system based on high performance liquid chromatography, characterized in that: The system includes a data acquisition module, a data transmission module, a cloud storage module, a data analysis module, a user interface module and a real-time analysis and feedback module; The data acquisition module is used to collect data detected by the high-performance liquid chromatograph and make baseline adjustments to the data; the data transmission module is used to determine the protocol used for data transmission, and at the same time compress and encrypt the data after the baseline adjustment; the cloud storage module is used to store the data transmitted to the cloud and provide data backup and recovery functions; the data analysis module is used to train the intelligent clustering analysis model, decrypt and decompress the data transmitted to the cloud, and input it into the trained intelligent clustering analysis model for data analysis; the user interface module is used to provide an interface for users to view the analysis results and record logs; the real-time analysis and feedback module is used to analyze and judge the data in real time. If the error between the analysis result and the prediction result exceeds the sum of square errors, the analysis plan and model parameters are adjusted for secondary analysis; The baseline adjustment is to remove noise and display the signal peak; The output end of the data acquisition module is connected to the input end of the data transmission module; the output end of the data transmission module is connected to the input end of the cloud storage module; the output end of the cloud storage module is connected to the input end of the data analysis module; the output end of the data analysis module is connected to the input end of the user interface module; the output end of the user interface module is connected to the input end of the real-time analysis and feedback module.

2. The cloud computing analysis system based on high performance liquid chromatography according to claim 1, characterized in that: The data acquisition module includes an HPLC instrument interface unit and a data processing unit; The HPLC instrument interface unit is used to provide an interface between the data acquisition module and the HPLC instrument, and transmit the data detected by the HPLC instrument to the system; the data processing unit is used to perform baseline adjustment on the data; The output end of the HPLC instrument interface unit is connected to the input end of the data processing unit; the output end of the data processing unit is connected to the input end of the data transmission module.

3. The cloud computing analysis system based on high performance liquid chromatography according to claim 1, characterized in that: The data transmission module includes a transmission protocol unit, a data compression unit and a data encryption unit; The transmission protocol unit adopts a secure transmission protocol HTTPS to ensure the security and integrity of data upload; The data compression unit is used to compress the collected data; the data encryption unit is used to encrypt and transmit the compressed data; The output end of the transmission protocol unit is connected to the input end of the data compression unit; the output end of the data compression unit is connected to the input end of the data encryption unit; the output end of the data encryption unit is connected to the input end of the cloud storage module.

4. The cloud computing analysis system based on high performance liquid chromatography according to claim 1, characterized in that: The cloud storage module includes a data storage unit and a backup and recovery unit; The data storage unit is used to store the received data in the cloud; The backup and recovery unit is used to add a data backup and recovery mechanism, so that the data can be reviewed and downloaded for use when data is abnormal or lost; The output end of the data storage unit is connected to the input end of the backup and recovery unit.

5. The cloud computing analysis system based on high performance liquid chromatography according to claim 1, characterized in that: The data analysis module includes a data decryption unit, a data decompression unit, a data preprocessing unit, an intelligent clustering analysis model unit, a multimodal data fusion unit, a model optimization unit and an analysis result visualization unit; The data decryption unit is used to decrypt the encrypted data; the data decompression unit is used to decompress the decrypted data; the data preprocessing unit is used to perform baseline correction, peak detection and peak integration on the data; the intelligent clustering analysis model unit is used to transfer the preprocessed data to the trained model for automatic analysis; the multimodal data fusion unit is used to fuse the high performance liquid chromatography data, mass spectrometry data, ultraviolet-visible spectrum data and infrared spectrum data, and improve the accuracy and reliability of the analysis results through multimodal data analysis; the model optimization unit is used to adjust the parameters of the model according to the gas analysis data in the database; The result visualization unit is used to display the result of data analysis in a visual way; The baseline correction is to remove the baseline drift from the chromatographic data and to quantify and characterize the peaks in the chromatogram; the peak detection is to identify the peaks from the corrected signal; the peak integration is to calculate the area of ​​each detected peak; The output end of the data decryption unit is connected to the input end of the data decompression unit; the output end of the data decompression unit is connected to the input end of the data preprocessing unit; the output end of the data preprocessing unit is connected to the input end of the intelligent clustering analysis model unit; the output end of the intelligent clustering analysis model unit is connected to the input end of the multimodal data fusion unit; The output end of the multimodal data fusion unit is connected to the input end of the model optimization unit; the output end of the model optimization unit is respectively connected to the input ends of the analysis result visualization unit and the intelligent clustering analysis model unit; the output end of the analysis result visualization unit is connected to the input end of the user interface module.

6. The cloud computing analysis system based on high performance liquid chromatography according to claim 1, characterized in that: The user interface module includes an authorization management unit, a user interface unit and a monitoring and logging unit; The authorization management unit is used to authenticate the identity of the user who views the analysis results; the user interface unit is an interface presented to the user; the monitoring and log unit is used to record the user and time who views the analysis results; The output end of the authorization management unit is connected to the input end of the user interface unit; the output end of the user interface unit is connected to the input end of the monitoring and logging unit; the output end of the monitoring and logging unit is connected to the input end of the real-time analysis and feedback module.

7. A cloud computing analysis method based on high performance liquid chromatography, characterized in that: The method comprises the following steps: S1. Analyze the sample using HPLC instrument, obtain chromatographic data, eliminate noise from the data, and convert the chromatographic data after noise elimination into digital signals by the detector, and record them in the local data system; S2, compress and encrypt the locally stored chromatographic data and upload it to the cloud storage platform, using the secure transmission protocol HTTPS to ensure the security and integrity of the data upload; S3, after decrypting and decompressing the received data, use the cloud computing platform to perform automated data preprocessing and data standardization; S4. Use big data analysis tools on the cloud computing platform to perform data modeling and analysis, apply the machine learning algorithm K-means clustering to perform gas prediction and classification on the pre-processed chromatographic data, and train the intelligent clustering analysis model; S5. Combine multimodal data fusion technology to fuse high performance liquid chromatography data, mass spectrometry data, UV-visible spectroscopy data and infrared spectroscopy data, and improve the accuracy and reliability of analysis results through multimodal data analysis; S6. Add a model optimization mechanism to the intelligent clustering analysis model, adjust the model parameters in combination with the database, and optimize the model; S7. Provide an interface for users to view analysis results, add a permission management mechanism, verify the permissions of users who view analysis results, and record logs; S8. Real-time analysis and feedback. After obtaining access rights, a judgment is made at the end of the analysis. If the error between the analysis result and the prediction exceeds the sum of square errors, the scheme and model parameters are immediately adjusted for secondary data analysis.

8. The cloud computing analysis method based on high performance liquid chromatography according to claim 7, characterized in that: In step S3, The preprocessing includes baseline correction, peak detection and peak integration; the data standardization is performed on the same scale for each feature through a standard normal distribution with a mean of 0 and a standard deviation of 1; The baseline correction is to remove the baseline drift from the chromatographic data and quantify and qualify the peaks in the chromatogram; The peak detection is to identify the peak from the corrected signal; The peak integration is to calculate the area of ​​each detected peak.

9. The cloud computing analysis method based on high performance liquid chromatography according to claim 7, characterized in that: In step S4, the intelligent clustering analysis model training specifically includes the following steps: S9-1. Data points , , , Divided into Group , , , , The values ​​were calculated using the silhouette coefficient method; Centroid calculation: in, For each group The center of mass, ; Yes Group The number of data points in ; Yes Group Data points in ; S9-2, Initialization: Select Group initial centroid , , , ; S9-3. Group assignment: For each data point , calculate its distance to each centroid and assign it to the group to which the nearest centroid belongs ; S9-4. Centroid update: Calculate the new centroid of each group and update: in, For each group The center of mass, ; Yes Group The number of data points in ; Yes Group Data points in ; S9-5, iteration: repeat step S9-3 and step S9-4 until the centroid no longer changes; The silhouette coefficient method is used to determine the number of groups in K-means clustering; specifically: When the number of data points When the initial number of groups Take 2, 3 and 4 respectively, when the number of data points When the initial number of groups Take 5, 6 and 7 respectively; for each value, execute the K-means algorithm, and calculate each The silhouette coefficient corresponding to the value : in It is a data point The average distance to other data points in the same group, It is a data point The average distance to the nearest data point in the other group; in, It is a data point The group you belong to; Yes Group The number of data points in ; It is a data point and data points The Euclidean distance between in, It is a data point The group you belong to; Yes Group The number of data points in ; It is a data point and data points The Euclidean distance between calculate The corresponding silhouette coefficient , the largest The corresponding This is the number of groups in the K-means clustering.

10. The cloud computing analysis method based on high performance liquid chromatography according to claim 7, characterized in that: In steps S5-S8, The multimodal data fusion technology is to obtain high-performance liquid chromatography data using a high-performance liquid chromatograph, obtain mass spectrum data using a mass spectrometer, obtain ultraviolet-visible spectrum data using an ultraviolet-visible spectrometer, and obtain infrared spectrum data using an infrared spectrometer, fuse the four types of data, perform data analysis multiple times, and improve the accuracy and reliability of the analysis results through multimodal data analysis; The real-time analysis and feedback: obtain access rights, make judgments after the analysis is completed, and if the error between the analysis result and the prediction result exceeds the sum of square errors, immediately adjust the scheme and model parameters to conduct secondary data analysis; If the error between the secondary analysis result and the predicted result is less than the sum of square errors, the second analysis result is selected and recorded in the database. If the error between the secondary analysis result and the predicted result is greater than or equal to the sum of square errors, and the error between the secondary analysis result and the first analysis result is less than the sum of square errors, the predicted result part of the intelligent clustering analysis model is adjusted and the log is recorded. The squared error and : in, represents the sum of squared errors; It is a data point The centroid of the group to which it belongs The Euclidean distance between ; Yes Group Data points in ; It is a data point The number of

Images