Laboratory instrument data automatic acquisition and analysis method and system

Abstract

The invention discloses a laboratory instrument data automatic acquisition and analysis method and system, and relates to the technical field of laboratory instrument acquisition and analysis, and the method comprises the steps: obtaining sample identification information, automatically sorting and distributing the sample identification information to different analyzers, uploading detection data to a central database in real time, and associating a sample number, a detection timestamp and an instrument number; the method comprises the following steps: constructing a multi-dimensional feature space based on historical detection data, performing sub-cluster division by adopting a density clustering algorithm, establishing a cross-index logic constraint relationship in combination with an expert rule base, mapping real-time detection data to the multi-dimensional feature space, judging a data exception type according to sub-cluster affiliation and the logic constraint relationship, and when judging that the data is exceptional, judging that the data is abnormal. And triggering a secondary detection process, automatically dispatching a mechanical arm to send the sample back to a specified instrument to execute differentiated reinspection operation, and generating a comprehensive report containing detection process tracing. According to the invention, the automation level of detection is improved, the labor cost is reduced, the detection quality and efficiency are improved, and the laboratory detection work efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of experimental instrument acquisition and analysis, and particularly to a method and system for automatic acquisition and analysis of laboratory instrument data. Background Art

[0002] With the continuous growth of laboratory testing requirements and the rapid development of automation technology, more and more laboratories have begun to introduce automated equipment to complete the transfer and testing of samples. For example, oil chromatography analysis, moisture analysis, dielectric loss analysis, breakdown voltage analysis, etc. are usually applied to routine tests in industrial oil product analysis, power system fault diagnosis, and other energy fields. These instruments are responsible for different testing items in the automated production line and can realize the transfer and docking of samples between various instruments through robotic arms or automatic sorting systems.

[0003] However, in actual operation, the existing technology often only collects and processes data from a single testing instrument, lacking means for comprehensive correlation analysis of the testing results of multiple instruments. The data of each instrument is usually stored separately in different management software or databases, and the testing results are judged and compared manually. Once there are conflicts in the testing results (for example, the values of the same batch of samples in different testing items have logical contradictions) or abnormal data caused by instrument failures, laboratory personnel need to manually check and re-inspect multiple times, which not only increases the workload but also easily leads to missed inspections and misjudgments. In addition, the lack of cross-instrument data correlation and automatic error correction functions makes it difficult to fully utilize the existing historical testing data to continuously monitor and improve the testing quality, and it is impossible to capture and correct potential errors in real time. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are as follows: Through cross-instrument correlation analysis of various testing data such as oil chromatography, moisture, dielectric loss, and breakdown voltage, and using intelligent algorithms to realize real-time anomaly identification and dynamic error correction of data, the accuracy and credibility of the testing results are significantly improved; it can be deeply coupled with automatic transfer and sorting equipment to build a closed-loop testing process from sample reception to abnormal re-inspection; when the testing data deviates or conflicts, secondary testing can be automatically performed or the expert knowledge base can be triggered to assist in judgment without manual intervention, greatly reducing the labor input and reducing the risk of misjudgment.

[0006] To solve the above technical problems, a method for automatic acquisition and analysis of laboratory instrument data is proposed, including,

[0007] Obtain sample identification information, automatically sort and allocate it to different analyzers, upload the testing data to the central database in real time, and associate the sample number, testing timestamp, and instrument number;

[0008] Construct a multi-dimensional feature space based on historical detection data. Use the density clustering algorithm to divide sub-clusters for the detection values of oil chromatography, moisture, dielectric loss, and breakdown voltage, and establish cross-index logical constraint relationships in combination with an expert rule library. Map the real-time detection data to the multi-dimensional feature space, determine the type of data anomaly according to the sub-cluster attribution and logical constraint relationships. When an anomaly is determined, trigger a secondary detection process, automatically dispatch the robotic arm to send the sample back to the specified instrument for differential re-inspection operations. Integrate the multi-dimensional detection data and anomaly marking information through a visualization interface, and generate a comprehensive report including the traceability of the detection process.

[0009] As a preferred solution of the automatic data acquisition and analysis method for laboratory instruments described in the present invention, among them: the different analyzers include an oil chromatography analyzer, a moisture analyzer, a dielectric loss analyzer, and a breakdown voltage analyzer;

[0010] The real-time uploaded detection data includes that the data interface adopts a dynamic protocol adaptation mechanism, uniformly parses the heterogeneous data frames of the protocol through a protocol conversion gateway, and generates a unique check code based on the hash algorithm. During the sorting process, through the secondary matching check of the mapping relationship with the task queue, when the check fails, a redundant communication link is triggered for retransmission.

[0011] As a preferred solution of the automatic data acquisition and analysis method for laboratory instruments described in the present invention, among them: the density clustering algorithm includes dynamically calculating the neighborhood radius according to the mean and standard deviation of the historical data sample spacing, combined with a scaling factor, and determining the minimum sample number based on the total amount of historical data and the preset number of sub-clusters, and initializing the sub-cluster center using an improved K-Means++ algorithm;

[0012] The formula for calculating the neighborhood radius is:

[0013]

[0014] Among them, is the neighborhood radius, is the mean of the historical data sample spacing, is the standard deviation of the historical data sample spacing, is the scaling factor;

[0015] The formula for determining the minimum sample number is:

[0016]

[0017] Among them, is the total amount of historical data, k is the preset number of sub-clusters, and m is the minimum sample number;

[0018] The formula for initializing the sub-cluster center is expressed as:

[0019]

[0020] Among them, is the centroid of the sub-cluster, is the sub-cluster, x is the sample point, j is the variable index, and k is the preset number of sub-clusters.

[0021] As a preferred solution of the automatic data acquisition and analysis method for laboratory instruments according to the present invention, wherein: the sub-cluster division includes determining a reasonable fluctuation range of the detection value based on the sub-cluster standard deviation and the confidence coefficient, storing the multi-variable logical constraint relationship. When the dielectric loss value exceeds the first threshold, the breakdown voltage must be lower than the second threshold, and the abnormal scenario is dynamically matched through the fuzzy inference engine to trigger the error correction process;

[0022] The establishment of the cross-index logical constraint relationship includes training a negative correlation model between the dielectric loss value and the breakdown voltage based on historical data to obtain the regression coefficient and the prediction interval. If the real-time detection value exceeds the prediction interval and exceeds the tolerance threshold, it is marked as a logical conflict and the cross-instrument calibration is triggered;

[0023] The multi-variable logical constraint relationship is expressed as:

[0024]

[0025] Among them, and are the linkage thresholds, D is the dielectric loss value, and B is the breakdown voltage;

[0026] The negative correlation model is expressed as:

[0027]

[0028] Among them, D is the dielectric loss value, B is the breakdown voltage, and b are the regression coefficients, and g is the residual term.

[0029] As a preferred solution of the automatic data acquisition and analysis method for laboratory instruments according to the present invention, wherein: the execution of the differential re-inspection operation includes dynamically adjusting the instrument cleaning time according to the ratio of the abnormal score to the preset maximum abnormal value, and selecting the cleaning, calibration, and environmental control parameter adjustment schemes from the preset strategy library for different abnormal types;

[0030] The formula for dynamically adjusting the instrument cleaning time is expressed as:

[0031]

[0032] Among them, is the reference cleaning time, is the maximum adjustment amount, E is the current abnormal score, is the system preset maximum abnormal value, is the adjusted cleaning time.

[0033] As a preferred solution of a method for automatically collecting and analyzing laboratory instrument data according to the present invention, wherein: the preset policy library includes calculating the posterior probability based on the historical abnormal scenario type and the success rate of the re-inspection action, selecting the optimal re-inspection policy, and when the new abnormal scenario does not match the existing policy, calling the expert knowledge base to supplement the rules and updating the policy library;

[0034] The preset policy library uses a Bayesian network to generate re-inspection instructions, and the conditional probability is expressed as:

[0035]

[0036] where A is the re-inspection action and S is the abnormal scenario type, is the probability of the re-inspection action A occurring given the abnormal scenario type S, is the probability of observing the abnormal scenario type S given the re-inspection action A, is the prior probability of the re-inspection action A occurring and is the probability of the abnormal scenario type S occurring.

[0037] As a preferred solution of a method for automatically collecting and analyzing laboratory instrument data according to the present invention, wherein: the generation of a comprehensive report including the traceability of the detection process includes synchronously displaying the oil chromatogram, moisture curve and breakdown voltage trend through a parallel coordinate system, supporting multi-dimensional data comparison, and marking the abnormal risk level with color grading according to the real-time abnormal density distribution, and the threshold is dynamically adjusted based on the upper quartile and interquartile range of the real-time data;

[0038] The generation of a comprehensive report including the traceability of the detection process further includes pushing structured alarm information to the laboratory terminal when the abnormal score exceeds the preset threshold after the secondary detection, and the alarm priority is dynamically adjusted according to the logarithmic relationship between the abnormal score and the number of re-inspections;

[0039] The abnormal score formula is expressed as:

[0040]

[0041] where, is the weight coefficient of the detection index, is the abnormal membership value of the i-th detection index, i is the variable index, and F is the abnormal score; when F is greater than or equal to 0.8, it is determined that an alarm needs to be pushed; when F is less than 0.8, it is determined to be normal and no alarm needs to be pushed.

[0042] As a preferred solution of a laboratory instrument data automatic acquisition and analysis system according to the present invention, it is characterized in that it includes a data acquisition and sorting module, a data processing and analysis module, a differential re-inspection operation module, and a report generation and visualization module.

[0043] The data acquisition and sorting module is used to automatically obtain sample identification information, sort and allocate samples to different analyzers, upload detection data to the central database in real time, and generate a unique verification code through a dynamic protocol adaptation mechanism and a hash algorithm.

[0044] The data processing and analysis module includes a density clustering algorithm unit, a logical constraint relationship establishment unit, and a secondary detection process trigger unit. It is used to construct a multi-dimensional feature space based on historical detection data, use the density clustering algorithm to divide the detection values into sub-clusters, establish cross-index logical constraint relationships in combination with an expert rule library, map real-time detection data to the multi-dimensional feature space, determine the type of data anomaly according to the sub-cluster attribution and logical constraint relationships, and trigger a secondary detection process when it is determined as an anomaly.

[0045] The differential re-inspection operation module includes a differential re-inspection operation execution unit, an instrument cleaning time dynamic adjustment unit, and a parameter adjustment scheme selection unit. It is used to automatically dispatch the robotic arm to send the sample back to the specified instrument to perform differential re-inspection operations after determining data anomalies, dynamically adjust the instrument cleaning time according to the ratio of the anomaly score to the preset maximum anomaly value, and select a suitable cleaning, calibration, and environmental control parameter adjustment scheme from the preset policy library.

[0046] The report generation and visualization module is used to integrate multi-dimensional detection data and anomaly marking information through a visualization interface, generate a comprehensive report including detection process traceability, and provide users with intuitive and comprehensive detection results and analysis.

[0047] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of a method for automatic acquisition and analysis of laboratory instrument data.

[0048] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps of a method for automatic acquisition and analysis of laboratory instrument data.

[0049] Advantages of the present invention: By automatically obtaining sample identification information and sorting it into different analyzers, and uploading and associating the detection data in real time, the present invention realizes the efficient collection and management of data, ensuring the integrity and traceability of the data; constructs a multi-dimensional feature space based on historical data and uses a density clustering algorithm for sub-cluster division, and establishes cross-index logical constraint relationships in combination with an expert rule base, effectively improving the determination accuracy of data anomaly types and reducing false positives and missed detections; after mapping the real-time detection data to the multi-dimensional feature space, triggers a secondary detection process according to the sub-cluster attribution and logical constraint relationships, and executes differential re-inspection operations by automatically scheduling the robotic arm, significantly improving the pertinence and efficiency of re-inspection; integrates multi-dimensional detection data and anomaly marking information through a visualization interface to generate a comprehensive report including the traceability of the detection process, providing users with intuitive and comprehensive analysis results and enhancing the convenience of data interpretation; in addition, the application of technologies such as dynamic protocol adaptation mechanism, hash algorithm verification, and redundant communication link retransmission greatly enhances the stability and security of data transmission; dynamically adjusts the instrument cleaning time and selects parameter adjustment schemes based on anomaly scores, optimizing the instrument maintenance process and extending the service life of the instrument; the preset policy library combines with the Bayesian network to generate re-inspection instructions, ensuring the optimization of re-inspection strategies, and at the same time, the supplementary rule update of the expert knowledge base enhances the adaptability and flexibility of the system; the present invention improves the automation level of detection, reduces labor costs, improves the detection quality and efficiency, and brings significant benefits to laboratory detection work. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0051] Figure 1 It is the overall flowchart of a method for automatically collecting and analyzing laboratory instrument data provided by an embodiment of the present invention.

[0052] Figure 2 It is the system scheme flowchart of a system for automatically collecting and analyzing laboratory instrument data provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present invention.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0055] Example 1, referring to Figure 1 , which is the first embodiment of the present invention. This embodiment provides a method for automatically collecting and analyzing laboratory instrument data, including:

[0056] S1: Obtain sample identification information, automatically sort and allocate it to different analyzers, upload the detection data to the central database in real time, and associate the sample number, detection timestamp, and instrument number.

[0057] The different analyzers include an oil chromatograph analyzer, a moisture analyzer, a dielectric loss analyzer, and a breakdown voltage analyzer;

[0058] The real-time upload of detection data includes that the data interface adopts a dynamic protocol adaptation mechanism, uniformly parses the heterogeneous data frames of the protocol through a protocol conversion gateway, and generates a unique checksum based on the hash algorithm. During the sorting process, the mapping relationship with the task queue is verified through secondary matching. When the verification fails, a redundant communication link is triggered for retransmission.

[0059] S2: Construct a multi-dimensional feature space based on historical detection data, use the density clustering algorithm to divide sub-clusters for the detection values of oil chromatography, moisture, dielectric loss, and breakdown voltage, and establish cross-index logical constraint relationships in combination with an expert rule base.

[0060] Further, the density clustering algorithm includes dynamically calculating the neighborhood radius according to the mean and standard deviation of the historical data sample spacing, combining a scaling factor, and determining the minimum number of samples based on the total amount of historical data and the preset number of sub-clusters. The sub-cluster centers are initialized using an improved K-Means++ algorithm;

[0061] The formula for calculating the neighborhood radius is:

[0062]

[0063] Where, is the neighborhood radius, is the mean of the historical data sample spacing, is the standard deviation of the historical data sample spacing, is the scaling factor;

[0064] The formula for determining the minimum number of samples is:

[0065]

[0066] where, is the total amount of historical data, k is the preset number of sub-clusters, and m is the minimum number of samples;

[0067] The formula for initializing the sub-cluster center is expressed as:

[0068]

[0069] where, is the centroid of the sub-cluster, is the sub-cluster, x is the sample point, j is the variable index, and k is the preset number of sub-clusters.

[0070] It should be noted that the dynamic parameters reduce the interference of outliers and avoid overfitting caused by fixed thresholds. K-Means++ optimizes the centroid initialization, accelerates the clustering convergence, and improves the accuracy.

[0071] Furthermore, the sub-cluster division includes determining a reasonable fluctuation range of the detection value based on the sub-cluster standard deviation and the confidence coefficient, storing the multivariate logical constraint relationship. When the dielectric loss value exceeds the first threshold, the breakdown voltage must be lower than the second threshold, and the abnormal scenario is dynamically matched through the fuzzy inference engine to trigger the error correction process;

[0072] The establishment of the cross-index logical constraint relationship includes training a negative correlation model between the dielectric loss value and the breakdown voltage based on historical data to obtain the regression coefficient and the prediction interval. If the real-time detection value exceeds the prediction interval and the tolerance threshold, it is marked as a logical conflict and the cross-instrument calibration is triggered;

[0073] The multivariate logical constraint relationship is expressed as:

[0074]

[0075] where, and are the linkage thresholds, D is the dielectric loss value, and B is the breakdown voltage;

[0076] The negative correlation model is expressed as:

[0077]

[0078] where D is the dielectric loss value, B is the breakdown voltage, and b are the regression coefficients, and g is the residual term.

[0079] It should be noted that through the dual-path complementarity of data-driven (clustering) and experience-driven (rules), the misjudgment of a single method is reduced, and the sub-cluster division is updated with historical data to adapt to the detection characteristics of different batches of samples.

[0080] S3: Map the real-time detection data to a multi-dimensional feature space, determine the type of data anomaly according to the sub-cluster attribution and logical constraint relationship. When it is determined as an anomaly, trigger the secondary detection process, and automatically dispatch the robotic arm to send the sample back to the specified instrument to perform a differential re-inspection operation.

[0081] Furthermore, the execution of the differential re-inspection operation includes dynamically adjusting the instrument cleaning time according to the ratio of the anomaly score to the preset maximum anomaly value, and selecting cleaning, calibration, and environmental control parameter adjustment schemes from the preset strategy library for different types of anomalies;

[0082] The formula for dynamically adjusting the instrument cleaning time is expressed as:

[0083]

[0084] Among them, is the reference cleaning time, is the maximum adjustment amount, E is the current anomaly score, is the system preset maximum anomaly value, is the adjusted cleaning time.

[0085] It should be noted that the differential re-inspection strategy is optimized for different types of anomalies to avoid the waste of resources in non-discriminatory re-inspection; the secondary detection eliminates accidental errors and improves the credibility of the results.

[0086] Even further, the preset strategy library includes calculating the posterior probability based on the success rate of historical anomaly scenario types and re-inspection actions, selecting the optimal re-inspection strategy. When a new anomaly scenario does not match the existing strategy, call the expert knowledge base to supplement the rules and update the strategy library;

[0087] The preset strategy library uses a Bayesian network to generate re-inspection instructions, and the conditional probability is expressed as:

[0088]

[0089] Among them, A is the re-inspection action, S is the anomaly scenario type, is the probability of the re-inspection action A occurring given the anomaly scenario type S, is the probability of observing the anomaly scenario type S given the re-inspection action A, is the prior probability of the re-inspection action A occurring, and is the probability of the anomaly scenario type S occurring.

[0090] S4: Integrate multi-dimensional detection data and anomaly marking information through a visualization interface, and generate a comprehensive report including traceability of the detection process.

[0091] The generation of the comprehensive report including traceability of the detection process includes synchronously displaying the oil chromatogram, moisture curve, and breakdown voltage trend through a parallel coordinate system, supporting multi-dimensional data comparison, and marking the anomaly risk level with color grading according to the real-time anomaly density distribution. The threshold is dynamically adjusted based on the upper quartile and interquartile range of the real-time data;

[0092] The generation of the comprehensive report including traceability of the detection process also includes pushing structured alarm information to the laboratory terminal when the anomaly score exceeds the preset threshold after secondary detection. The alarm priority is dynamically adjusted according to the logarithmic relationship between the anomaly score and the number of retest times;

[0093] The anomaly score formula is expressed as:

[0094]

[0095] Wherein, is the weight coefficient of the detection index, is the anomaly membership value of the i-th detection index, i is the variable index, and F is the anomaly score. When F is greater than or equal to 0.8, it is determined that an alarm needs to be pushed; when F is less than 0.8, it is determined to be normal and no alarm needs to be pushed.

[0096] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

[0097] Example 2, referring to Figure 2 , which is the second embodiment of the present invention. This embodiment provides an automatic data acquisition and analysis system for laboratory instruments, including a data acquisition and sorting module, a data processing and analysis module, a differential retest operation module, and a report generation and visualization module.

[0098] The data acquisition and sorting module is used to automatically obtain sample identification information, sort and allocate samples to different analyzers, upload detection data to the central database in real time, and generate a unique verification code through a dynamic protocol adaptation mechanism and a hash algorithm.

[0099] The data processing and analysis module includes a density clustering algorithm unit, a logical constraint relationship establishment unit, and a secondary detection process trigger unit, which are used to construct a multi-dimensional feature space based on historical detection data, divide the detection values into sub-clusters using the density clustering algorithm, establish cross-index logical constraint relationships in combination with an expert rule base, map real-time detection data to the multi-dimensional feature space, determine the type of data anomaly according to the sub-cluster attribution and logical constraint relationships, and trigger the secondary detection process when an anomaly is determined.

[0100] The differential re-inspection operation module includes a differential re-inspection operation execution unit, an instrument cleaning time dynamic adjustment unit, and a parameter adjustment scheme selection unit, which are used to automatically dispatch the robotic arm to send the sample back to the specified instrument to perform the differential re-inspection operation after determining data anomalies, dynamically adjust the instrument cleaning time according to the ratio of the anomaly score to the preset maximum anomaly value, and select a suitable cleaning, calibration, and environmental control parameter adjustment scheme from the preset strategy library.

[0101] The report generation and visualization module is used to integrate multi-dimensional detection data and anomaly marking information through a visualization interface, generate a comprehensive report including the traceability of the detection process, and provide users with intuitive and comprehensive detection results and analysis.

[0102] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

[0103] Embodiment 3, the third embodiment of the present invention, which is different from the previous two embodiments in that:

[0104] If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the essence of the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0105] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered as a definitional sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in conjunction with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0106] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0107] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

Claims

1. A method for automatic collection and analysis of laboratory instrument data, characterized in that: include, Obtain sample identification information, automatically sort and distribute to different analyzers, upload test data to the central database in real time, and associate sample number, test timestamp and instrument number; A multi-dimensional feature space is constructed based on historical detection data, and a density clustering algorithm is used to divide the oil chromatography, moisture, dielectric loss and breakdown voltage detection values ​​into sub-clusters, and a cross-indicator logical constraint relationship is established in combination with an expert rule base; Map the real-time detection data to the multi-dimensional feature space, determine the data anomaly type according to the sub-cluster affiliation and logical constraint relationship, and trigger the secondary detection process when it is determined to be abnormal. Automatically dispatch the robotic arm to send the sample back to the designated instrument for differentiated re-inspection. Integrate multi-dimensional detection data and abnormal marking information through a visual interface, and generate a comprehensive report including detection process traceability.

2. The method for automatic data collection and analysis of laboratory instruments according to claim 1, characterized in that: The different analyzers include an oil chromatograph, a moisture analyzer, a dielectric loss analyzer, and a breakdown voltage analyzer; The real-time uploading of detection data includes a data interface that adopts a dynamic protocol adaptation mechanism, uniformly parsing heterogeneous data frames of the protocol through a protocol conversion gateway, and generating a unique check code based on a hash algorithm. During the sorting process, a mapping relationship between the check and the task queue is checked through a secondary matching process. When the check fails, a redundant communication link retransmission is triggered.

3. A method for automatic collection and analysis of laboratory instrument data as claimed in claim 2, characterized in that: The density clustering algorithm includes dynamically calculating the neighborhood radius based on the mean and standard deviation of the historical data sample spacing combined with the scaling factor, determining the minimum number of samples based on the total amount of historical data and the preset number of sub-clusters, and initializing the sub-cluster center using the improved K-Means++ algorithm; The formula for calculating the neighborhood radius is: ,in, is the neighborhood radius, is the mean value of historical data sample intervals, is the standard deviation of historical data sample intervals, is the scaling factor; The formula for determining the minimum number of samples is: ,in, is the total amount of historical data, k is the number of preset subclusters, and m is the minimum number of samples; The formula for initializing the subcluster center is expressed as: ,in, is the centroid of the subcluster, is a sub-cluster, x is a sample point, j is a variable index, and k is the preset number of sub-clusters.

4. A method for automatic collection and analysis of laboratory instrument data as claimed in claim 3, characterized in that: The sub-cluster division includes determining a reasonable fluctuation range of the detection value based on the sub-cluster standard deviation and the confidence coefficient, storing a multi-variable logical constraint relationship, when the dielectric loss value exceeds a first threshold, the breakdown voltage must be lower than a second threshold, and dynamically matching abnormal scenarios through a fuzzy reasoning engine to trigger an error correction process; The establishment of the cross-indicator logical constraint relationship includes training a negative correlation model between dielectric loss value and breakdown voltage based on historical data to obtain a regression coefficient and a prediction interval. If the real-time detection value exceeds the prediction interval and exceeds the tolerance threshold, it is marked as a logical conflict, triggering a cross-instrument calibration; The multivariable logical constraint relationship is expressed as: ,in, and is the linkage threshold, D is the dielectric loss value, and B is the breakdown voltage; The negative correlation model is expressed as: , where D is the dielectric loss value, B is the breakdown voltage, and b are regression coefficients, and g is the residual term.

5. A method for automatic collection and analysis of laboratory instrument data as claimed in claim 4, characterized in that: The performing of the differentiated re-inspection operation includes dynamically adjusting the instrument cleaning time according to the ratio of the abnormality score to the preset maximum abnormality value, and selecting cleaning, calibration and environmental control parameter adjustment schemes from a preset strategy library for different abnormality types; The formula for dynamically adjusting the instrument cleaning time is expressed as: ,in, is the baseline cleaning time, is the maximum adjustment amount, E is the current anomaly score, Preset the maximum abnormal value for the system, is the adjusted cleaning time.

6. A method for automatic collection and analysis of laboratory instrument data as claimed in claim 5, characterized in that: The preset strategy library includes calculating the posterior probability and selecting the optimal re-inspection strategy based on the historical abnormal scenario type and the success rate of the re-inspection action. When the new abnormal scenario does not match the existing strategy, the expert knowledge base is called to supplement the rules and the strategy library is updated; The preset strategy library uses Bayesian network to generate re-inspection instructions, and the conditional probability is expressed as: , where A is the recheck action, S is the abnormal scenario type, is the probability of recheck action A occurring under a given abnormal scenario type S, is the probability of observing abnormal scene type S under a given recheck action A, is the prior probability of the recheck action A occurring, is the probability of abnormal scenario type S occurring.

7. A method for automatic collection and analysis of laboratory instrument data as claimed in claim 6, characterized in that: The generation of a comprehensive report including traceability of the detection process includes synchronously displaying the oil chromatogram, moisture curve and breakdown voltage trend through a parallel coordinate system, supporting multi-dimensional data comparison, marking the abnormal risk level with color classification according to the real-time abnormal density distribution, and dynamically adjusting the threshold based on the upper quartile and interquartile range of the real-time data; The generating of the comprehensive report including the traceability of the testing process also includes pushing structured alarm information to the laboratory terminal when the abnormality score after the second test exceeds the preset threshold, and the alarm priority is dynamically adjusted according to the logarithmic relationship between the abnormality score and the number of retests; The anomaly score formula is expressed as: ,in, is the weight coefficient of the detection index, is the abnormal membership value of the i-th detection indicator, i is the variable index, and F is the abnormal score; when F is greater than or equal to 0.8, it is determined that an alarm needs to be pushed; when F is less than 0.8, it is determined to be normal and no alarm needs to be pushed.

8. A system using the method for automatic data collection and analysis of laboratory instruments as claimed in any one of claims 1 to 7, characterized in that: It includes data collection and sorting module, data processing and analysis module, differentiated re-inspection operation module and report generation and visualization module; The data acquisition and sorting module is used to automatically obtain sample identification information, sort and distribute samples to different analyzers, upload test data to the central database in real time, and generate a unique verification code through a dynamic protocol adaptation mechanism and a hash algorithm; The data processing and analysis module includes a density clustering algorithm unit, a logic constraint relationship establishment unit and a secondary detection process triggering unit, which are used to construct a multi-dimensional feature space based on historical detection data. The density clustering algorithm is used to divide the detection values ​​into sub-clusters, and the cross-indicator logical constraint relationship is established in combination with the expert rule base. The real-time detection data is mapped to the multi-dimensional feature space, and the data anomaly type is determined according to the sub-cluster affiliation and logical constraint relationship. When it is determined to be abnormal, the secondary detection process is triggered; The differential re-inspection operation module includes a differential re-inspection operation execution unit, an instrument cleaning time dynamic adjustment unit and a parameter adjustment scheme selection unit, which is used to automatically dispatch a robotic arm to send the sample back to a designated instrument to perform a differential re-inspection operation after determining that the data is abnormal, and dynamically adjust the instrument cleaning time according to the ratio of the abnormality score to the preset maximum abnormality value, and select a suitable cleaning, calibration and environmental control parameter adjustment scheme from a preset strategy library; The report generation and visualization module is used to integrate multi-dimensional detection data and abnormal marking information through a visual interface, generate a comprehensive report including detection process traceability, and provide users with intuitive and comprehensive detection results and analysis.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for automatic collection and analysis of laboratory instrument data described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for automatic collection and analysis of laboratory instrument data described in any one of claims 1 to 7 are implemented.

Images