Chemistry laboratory information management system

Abstract

The application relates to a chemical detection laboratory information management system, which comprises a standard management module, a process engine module, a data acquisition module and a quality control module; the standard management module analyzes standard texts through a structured deconstruction unit to construct a detection knowledge graph; the process engine module is coupled with the standard management module, and according to the detection standard identification of a commissioned task, the knowledge graph data are called to generate an instantiated process template; the data acquisition module comprises an instrument interface adaptation unit and an edge preprocessing unit arranged in a data link between detection instruments and the quality control module, and the edge preprocessing unit respectively acquires original data streams and performs real-time compliance verification; the quality control module is signal-connected with the edge preprocessing unit and the process engine module, generates control instructions according to verification results, suspends the process and activates abnormal processing when unqualified, and allows the process to proceed after being qualified. The application solves the structural contradiction between unstructured standard texts and process automation, and the timing dislocation problem between quality control and data acquisition.

Description

Technical Field

[0001] This invention relates to the field of laboratory information management technology, and in particular to a chemical testing laboratory information management system. Background Technology

[0002] With the deepening of digital transformation in the testing and inspection industry, Laboratory Information Management Systems (LIMS) have become the core infrastructure for chemical testing laboratories to achieve sample management, data acquisition, and report generation. Existing LIMS platforms, by solidifying business process nodes, realize the basic flow of testing tasks and the storage and management of electronic original records (ELNs), thereby improving the laboratory's document management level to a certain extent.

[0003] However, the field of chemical testing is characterized by significant method dependence and standard complexity. Currently, chemical testing laboratories face a structural bottleneck in the digital transformation of standards and specifications during the process of information management. Specifically, national and industry-issued testing standards are typically in the form of natural language text, with key metadata such as testing method parameters, quality control indicators, and judgment thresholds scattered across multiple chapters, lacking a unified structured description standard. Existing LIMS systems, when processing such heterogeneous standard texts, mainly rely on manual parsing and manual input to build testing method libraries. This process is not only inefficient but also prone to errors in parameter extraction or version confusion due to operators' misunderstandings of standard clauses, directly leading to inaccurate configuration of subsequent testing processes. Simultaneously, the high complexity of chemical testing business processes poses a severe challenge to the workflow orchestration capabilities of existing systems. Chemical testing involves multiple interconnected stages, including sample pretreatment, instrumental analysis, and data processing, with the technical parameters and quality control requirements of each stage tightly coupled to specific clauses in the testing standards. Existing workflow engine-based LIMS platforms typically use pre-set rigid workflow templates, making it difficult to dynamically reconstruct workflows based on sample matrix type, target analyte characteristics, and the specific constraints of the corresponding testing standards. When faced with complex matrix samples or new methods that have been modified according to standards, the system often fails to automatically adapt the logical dependencies between detection steps and the rules for inserting quality control nodes, resulting in rigid process configurations that make it difficult to meet the stringent requirements of the chemical testing field for compliance of detection methods and traceability of data quality.

[0004] Furthermore, there is a significant gap in the existing systems regarding the reliability assurance mechanisms for chemical detection data. The raw data output by detection instruments typically includes multi-dimensional unstructured information such as chromatograms and spectral signals. However, existing LIMS systems primarily focus on the physical storage and simple display of data, lacking the ability to verify data quality in real time for chemical detection characteristics. During data transfer, the compliance determination of key quality control parameters (such as parallel sample deviation, spike recovery rate, and detection limit applicability) relies on manual offline calculation and verification. This approach not only lags behind the detection process but also makes it difficult to implement effective quality control at the source of data generation, increasing the risk to the reliability of detection results. Summary of the Invention

[0005] Based on this, it is necessary to provide a chemical testing laboratory information management system to address the structural contradiction between the unstructured nature of the aforementioned standard texts and the need for automated testing processes, as well as the technical problem of timing misalignment between quality control nodes and data acquisition nodes.

[0006] The present invention provides a chemical testing laboratory information management system, including a standard management module, a process engine module, a data acquisition module, and a quality control module; The standard management module is configured with a standard text structure deconstruction unit, which is used to perform clause-level semantic parsing on the detection standard text to extract detection method metadata, and to construct a detection knowledge graph based on the detection method metadata. The node attributes of the detection knowledge graph include detection condition parameters, quality control rule thresholds, and judgment logic relationships. The process engine module is data coupled with the standard management module. The process engine module includes a process template dynamic generation unit. The process template dynamic generation unit is configured to call the corresponding detection method metadata from the detection knowledge graph according to the target detection standard identifier associated with the received commissioned detection task, and automatically reconstruct the detection process node sequence and the quality control node embedding order according to the quality control rule threshold and judgment logic relationship, so as to generate an instantiated process template that matches the target detection standard. The data acquisition module includes an instrument interface adaptation unit and an edge preprocessing unit. The instrument interface adaptation unit is used to acquire the raw data stream output by the testing instrument through multi-protocol conversion. The edge preprocessing unit is deployed in the data link between the testing instrument and the quality control module. The edge preprocessing unit is configured to parse the raw data stream in real time to extract detection feature values, and to perform compliance verification on the detection feature values ​​according to the quality control rule threshold corresponding to the current process node in the instantiated process template. The quality control module is signal-connected to the edge preprocessing unit and the process engine module. The quality control module is configured to generate process control instructions based on the compliance verification results. When the detection feature value does not meet the quality control rule threshold, the process control instruction is used to trigger the process engine module to suspend the current flow of the instantiated process template and activate the quality control exception handling node. Only after the detection feature value meets the quality control rule threshold will the process engine module allow the instantiated process template to flow to the next detection process node.

[0007] In one embodiment, the standard text structure deconstruction unit is equipped with a natural language processing engine. This engine is configured to perform clause-level semantic parsing on the detection standard text to extract detection method metadata. The detection method metadata includes detection condition parameters, quality control rule thresholds, and judgment logic relationships. The engine is then used to construct the detection knowledge graph in the form of a resource description framework. The nodes of the detection knowledge graph include detection method nodes and quality control rule nodes. The attributes of the quality control rule nodes include a parallel sample relative deviation threshold, a qualified recovery rate interval, and a method detection limit. The edges of the detection knowledge graph include judgment logic edges, used to characterize the conditional triggering relationship between the detection method nodes and the quality control rule nodes. The process template dynamic generation unit further includes a matrix type identification subunit. This subunit is configured to parse the sample matrix description text in the commissioned testing task to identify the matrix complexity level, and, based on the matrix complexity level, call the corresponding preprocessing step node sequence from the detection knowledge graph to adapt to the digestion, extraction, or purification process nodes required for complex matrix samples.

[0008] In one embodiment, the process engine module further includes a task load balancing and dispatching unit and a task auto-claiming unit; the task load balancing and dispatching unit is configured to obtain the real-time task queue length of each detection device. Matching coefficient with the qualification of the testing method And calculate the task assignment priority probability. The entrusted testing task is assigned to the experimental task node associated with the testing device that has the highest assignment priority probability, wherein, , The total number of testing devices. and The first The matching coefficient of the testing method qualification of each testing device and the length of the real-time task queue. To sum and traverse the index of the detection devices, =1,2,…,n; The task auto-assignment unit is configured to obtain the detection capability qualification vector of the experimenters. With current task queue saturation And calculate the matching degree function. , in, This is the qualification requirement vector for the commissioned testing task. To maximize task capacity, when the matching degree When the preset threshold is exceeded, the commissioned testing task will be automatically assigned to the corresponding experimental personnel.

[0009] In one embodiment, the edge preprocessing unit is configured with an outlier detection subunit, a chromatographic data processing subunit, and a curve fitting evaluation subunit; the outlier detection subunit is configured to use the isolated forest algorithm to calculate the outlier of the detected feature values, and when the detected feature value is a parallel sample measurement value, to perform a Grubbs test to calculate the statistic. , in, This is the Grubbs test statistic. For the first One measured value, The average of the measured values. The sample standard deviation of the measured values; When the statistic Greater than the Grubbs threshold At that time, the corresponding abnormal measurement values ​​are removed and a retest process is triggered; the chromatographic data processing subunit is configured to perform baseline correction and peak identification on the chromatographic data in the original data stream, and extract the chromatographic peak area. With retention time The retention time Standard retention time of the corresponding standard in the detection knowledge graph Compare the results, and when the relative deviation... When the threshold is exceeded, the chromatographic peak is identified as abnormal and a retest process is triggered; when the comparison is successful, the sample concentration is calculated based on the standard curve parameters in the detection knowledge graph. , in, The slope The intercept; The curve fitting evaluation subunit is configured to perform curve fitting using a Logistic four-parameter equation when the standard curve parameters correspond to a nonlinear response interval. The Logistic four-parameter equation is: , in, This is the instrument response value. This represents the sample concentration value. The upper limit asymptotic value, The lower bound asymptotic value, The half-maximum response concentration, Slope factor; The parameters were iteratively optimized using the Levenberg-Marquardt algorithm, and the Akaike information criterion was adopted. Evaluate the goodness of fit of the model to determine the optimal fitting curve, where For the number of parameters, This represents the maximum likelihood function value.

[0010] In one embodiment, the quality control module further includes an uncertainty assessment unit configured to identify uncertainty source components of the detection feature value, including Type A assessment components. Compared with Category B assessment scores Based on the uncertainty propagation law, when the detected characteristic value is the measurement result of a linear model, its combined standard uncertainty is calculated using the following formula. , , in, For measurement function, For input quantity, For standard uncertainty, For covariance, Total number of inputs.

[0011] In one embodiment, a blockchain evidence storage module is further included. This module is data-coupled with the data acquisition module and the quality control module. The blockchain evidence storage module is configured to perform hash operations on the original data stream and the detection feature values ​​to generate data fingerprints, and to package the data fingerprints and the timing information of the process control instructions to generate blocks and write them into the consortium blockchain. The blockchain evidence storage module is also configured with a smart contract unit. This smart contract unit is configured to automatically trigger a contract event when the conformity verification result is unqualified, and to write the processing record of the quality control anomaly handling node and the digital signature of the reviewer into the next block of the consortium blockchain.

[0012] In one embodiment, the data acquisition module further includes a sample tracking unit, which is equipped with an RFID reader and a temperature and humidity sensor. The RFID reader is used to read electronic tags on the sample container to obtain sample identification, and the temperature and humidity sensor is used to collect temperature and humidity data of the sample storage environment. The sample tracking unit also includes a stability prediction subunit, which is configured to calculate the degradation rate constant of the target analyte in the sample based on the temperature and humidity data and the Arrhenius equation, and predict the sample's shelf-life decay curve based on the degradation rate constant. When the predicted concentration deviation exceeds the allowable range, a retest reminder instruction is generated. The Arrhenius equation is... , in, This represents the degradation rate constant of the target analyte in the sample. Pre-exponential factor, For activation energy, The gas constant is... This refers to absolute temperature.

[0013] In one embodiment, the system further includes a commission management module, which is coupled to an external LIMS platform via a standardized interface. The commission management module is configured to parse the structured data of the commission order pushed by the external LIMS platform, extract sample information, testing items, and testing standard identifiers, and automatically trigger the standard management module to call the testing method metadata based on the testing standard identifiers. The quality control module also includes a data rounding unit, which is configured to identify the required number of significant digits based on the testing method metadata in the testing knowledge graph, and execute rounding rules (rounding to the nearest even number). When the leftmost digit of the number to be discarded is 5 and there are no digits after it or all digits are zero, if the last digit to be retained is even, it is discarded; if it is odd, it is rounded up.

[0014] In one embodiment, the process engine module employs a domain-specific language interpreter, which is configured to parse the primitive identifiers of the detection process nodes in the visual process orchestration diagram. The primitive identifiers include preprocessing primitives, instrument analysis primitives, and quality control insertion primitives. The interpreter converts the primitive identifiers into executable code and verifies the logical dependencies of the detection process nodes according to chemical logic constraint rules to generate the instantiated process template.

[0015] In one embodiment, the system further includes a statistical analysis module and a system management module. The statistical analysis module is data-coupled with the quality control module and is configured to perform multi-dimensional statistical queries and data analysis based on the timing information of the detection feature values ​​and the process control instructions. It supports the generation of custom report templates and the automatic generation of reports periodically, and realizes the export and printing functions of statistical reports. The system management module is configured to perform permission settings and access control management for departments, positions and personnel, provide system log recording and audit trail functions, and adopt high availability and disaster recovery technologies to deploy the system architecture to ensure stable system operation and data security.

[0016] The aforementioned chemical testing laboratory information management system performs clause-level semantic parsing of testing standard texts through standard text structure deconstruction units to extract testing method metadata and construct a testing knowledge graph. This achieves automated, structured extraction and associated storage of testing standard metadata, overcoming the inefficiencies and susceptibility to parameter extraction errors or version confusion caused by manual standard text parsing. The workflow engine module automatically reconstructs the sequence of testing workflow nodes and the embedding order of quality control nodes based on the quality control rule thresholds and judgment logic relationships in the testing knowledge graph. This generates an instantiated workflow template matching the target testing standard, enabling the testing workflow to dynamically adapt to the logical dependencies between testing steps based on the testing standard identifier of the commissioned sample. This solves the technical problem of existing LIMS platforms using preset rigid workflow templates, which struggle to dynamically reconstruct workflows according to the specific constraints of testing standards. The edge preprocessing unit performs real-time parsing and compliance verification of the raw data stream in the data link between the testing instrument and the quality control module. Combined with the quality control module, it generates process control instructions based on the verification results to control the flow status of the instantiated process template. This implements quality interception and process suspension at the source of data generation, eliminating the timing misalignment between the quality control node and the data acquisition node in the existing system, and ensuring the reliability of the test data and the compliance requirements of the test process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a chemical testing laboratory information management system according to one embodiment; Figure 2 This is a schematic diagram of a process engine module for one embodiment; Figure 3This is a schematic diagram of an edge processing unit according to one embodiment.

[0019] Figure label: 110. Standard Management Module; 112. Standard Text Structure Deconstruction Unit; 120. Process Engine Module; 122. Process Template Dynamic Generation Unit; 124. Task Load Balancing Allocation Unit; 126. Task Automatic Assignment Unit; 130. Data Acquisition Module; 132. Instrument Interface Adaptation Unit; 134. Edge Preprocessing Unit; 140. Quality Control Module. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] It should be noted that when a component is referred to as being "fixed to" or "set on" another component, it can be directly on the other component or there may be an intermediate component. When a component is considered to be "connected to" another component, it can be directly connected to the other component or there may be an intermediate component present. The terms "vertical," "horizontal," "upper," "lower," "left," "right," and similar expressions used in this specification are for illustrative purposes only and do not represent the only possible implementation.

[0022] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0023] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature and the second feature are in indirect contact through an intermediate medium. Furthermore, "above," "over," and "on top" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0024] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0025] The following is combined with Figures 1-3 The present invention describes a chemical testing laboratory information management system.

[0026] like Figure 1 As shown, in one embodiment, a chemical testing laboratory information management system includes a standard management module 110, a process engine module 120, a data acquisition module 130, and a quality control module 140.

[0027] The standard management module 110 is equipped with a standard text structure deconstruction unit 112. The standard text structure deconstruction unit 112 is used to perform clause-level semantic parsing on the detection standard text to extract detection method metadata, and to construct a detection knowledge graph based on the detection method metadata. The node attributes of the detection knowledge graph include detection condition parameters, quality control rule thresholds, and judgment logic relationships. The structure deconstruction unit uses natural language processing technology to perform in-depth parsing on national and industry detection standards that exist in the form of natural language text. It transforms the detection method parameters, quality control indicators, and judgment thresholds scattered in multiple chapters of the standard text into structured data, and constructs a knowledge graph containing detection method nodes, quality control rule nodes, and judgment logic edges. This realizes the automated extraction and associated storage of detection standard metadata, overcoming the inefficiency and parameter extraction errors caused by manual parsing.

[0028] Specifically, the implementation of the natural language processing engine includes the following steps: First, a pre-trained chemical domain named entity recognition model is used to identify key entities in standard text, such as "sample amount," "extraction solvent," "column temperature," "parallel sample deviation limit," and "recovery range." This model can be fine-tuned on a labeled standard text corpus (entity types include: detection object, instrument parameters, reagents, operation steps, calculation formulas, and thresholds). Second, dependency parsing and rule templates are used to extract relationships between entities. For example, "determination: weigh 1.0g sample (accurate to 0.1mg)" is parsed into structured triples (operation: weighing, object: sample, parameters: mass = 1.0g, accuracy = 0.1mg). Finally, all extracted triples are stored in the form of a resource description framework to construct a knowledge graph. The attribute set of the "detection method node" includes: pretreatment method, instrument conditions, and calculation formula references; the attribute set of the "quality control rule node" includes: rule type (e.g., parallel sample, spiked recovery), threshold, and applicable steps. The decision logic edges are associated through predefined attributes. For example, the quality control rule node is linked to a specific detection method node through the applyToStep attribute.

[0029] The standard text structuring and deconstruction unit 112 is also equipped with a version control subunit to handle updates to the detection standard. When a new version of the standard text is input, the version control subunit automatically identifies the changed clauses and generates a difference report by calculating the edit distance between the old and new versions of the knowledge graph (based on the number of additions, deletions, and modifications to nodes and edges). For nodes that only involve changes in numerical thresholds (such as updates to method detection limit values), the system manages the coexistence of multiple versions through the version attribute in the graph and calls the latest version by default when the process template is dynamically generated, while retaining the historical version traceability link. For standard text in scanned PDF format, the standard text structuring and deconstruction unit 112 first performs text recognition through an OCR engine (such as Tesseract or a deep learning-based text detection model), then extracts the table structure through a layout analysis algorithm (such as converting "Table 1 Precision Data" into a structured data table), and finally inputs it into an NLP engine for semantic parsing.

[0030] The process engine module 120 is data-coupled with the standard management module 110. The process engine module 120 includes a process template dynamic generation unit 122. The process template dynamic generation unit 122 is configured to call the corresponding detection method metadata from the detection knowledge graph according to the target detection standard identifier associated with the received commissioned detection task, and automatically reconstruct the detection process node sequence and the embedding order of quality control nodes according to the quality control rule threshold and judgment logic relationship, and generate an instantiated process template that matches the target detection standard. This process engine abandons the preset rigid process template used by the existing LIMS platform, dynamically adapts the logical dependency relationship between detection steps according to the detection standard identifier clearly stated in the commission, and automatically inserts quality control nodes at key quality control points.

[0031] The workflow of the process template dynamic generation unit 122 is as follows: The system predefines a set of basic process node types (such as "sample pretreatment", "instrument setup", "data calculation", "report generation") and their configurable attributes. Upon receiving the target detection standard identifier, the process template dynamic generation unit 122 queries the knowledge graph to obtain all "detection method nodes" corresponding to the standard and their sequential relationships (via the hasNextStep attribute), mapping them to basic node instances. Simultaneously, it queries all "quality control rule nodes" associated with these method nodes via the applyToStep attribute. Based on the "applicable steps" attribute in the rule node, the quality control node (such as "perform parallel sample determination" or "calculate spiked recovery rate") is embedded after the corresponding detection method node. For example, if a rule in the knowledge graph specifies that "parallel sample deviation calculation" is required after the "instrument setup" step, then a "quality control calculation - parallel sample" node will be automatically inserted after the "instrument setup" node in the generated instantiated process template. The matrix type identification subunit is implemented through a text classification model. This model is trained based on sample descriptions (such as "soil", "wastewater", "edible oil") in historical order forms and manually labeled complexity levels. It outputs complexity levels (such as "high", "medium", "low") and dynamically adjusts the node sequence of the preprocessing part in the process template according to a preset rule mapping table (such as mapping "high complexity" to the digestion, purification and other preprocessing node sequences marked "suitable for complex matrices" in the knowledge graph).

[0032] The process engine module 120 maintains a process state machine, including states: initialization, running, suspended, completed, and exception. When the quality control module 140 issues a "suspend" command, the state machine sets the current node state to suspended and persists the current context (including collected data, verification results, and timestamps) to the database. When a "continue" command is received, the context is restored from the database, and execution continues from the suspended node. In the task load balancing and dispatching unit 124, the qualification matching coefficient of the detection method is... The calculation method is as follows =(Number of equipment calibration items covered / Total number of items required for the task) × Equipment calibration status coefficient (1 for not expired, 0.5 for about to expire, and 0 for expired); Task queue length This represents the number of tasks currently pending on the device; the calculated result. The largest device is the target-assigned device. In the task automatic assignment unit 126, the experimental personnel's detection capability qualification vector C is constructed through the types of detection projects completed by the personnel in the past and their proficiency scores. The proficiency scores are calculated based on the historical task completion quality (inverse of the anomaly rate) and speed. The qualification requirement vector R is represented by the knowledge graph node vector of the detection method required by the task. The matching degree threshold is usually set to 0.75.

[0033] The data acquisition module 130 includes an instrument interface adapter unit 132 and an edge preprocessing unit 134. The instrument interface adapter unit 132 is used to acquire the raw data stream output by the testing instrument through multi-protocol conversion. The edge preprocessing unit 134 is deployed in the data link between the testing instrument and the quality control module 140. The edge preprocessing unit 134 is configured to perform real-time parsing of the raw data stream to extract detection feature values, and to perform compliance verification of the detection feature values ​​according to the quality control rule threshold corresponding to the current process node in the instantiated process template. The edge preprocessing unit performs quality verification at the data source and moves the quality control node to the front end of the data acquisition.

[0034] Specifically, the instrument interface adaptation unit 132 maintains a protocol driver library that supports common communication protocols such as serial communication, TCP / IP, OPC UA, and proprietary protocols from various vendors (through reverse engineering or vendor-provided SDK encapsulation). Based on a preset device configuration table, this unit assigns appropriate drivers to each testing instrument, uniformly converting the raw byte streams or messages output by the instruments into internally defined standard data objects (containing fields such as timestamp, device ID, raw values, and units). The edge preprocessing unit 134, as an independent service or containerized application, is deployed on the industrial control computer or edge server where the instrument resides. It subscribes to the standard data stream from the instrument interface adaptation unit 132. The preprocessing unit 134 maintains the context of the currently executing "instantiated process template," which is pushed by the process engine module 120 at the start of the task. When the data stream arrives, the preprocessing unit 134 determines which type of verification needs to be performed based on the current node information in the context. For example, if the current node is "parallel sample determination," the Grubbs test is triggered; if it is "chromatographic analysis," the chromatographic peak processing flow is triggered. The threshold used for verification is obtained directly from the "quality control rule node" attribute associated with the current node.

[0035] The edge preprocessing unit 134 communicates with the instrument interface adapter unit 132 via MQTT protocol or HTTP long connection to subscribe to the raw data stream in real time. The standard data object adopts JSON format and contains the following fields: {"instrumentId":"GC-001","timestamp":"2024-01-15T09:30:00Z", "rawData":{"chromatogram":[...],"peakList":[...]},"sampleId":"S20240115001"}. When extracting detection feature values, if it is chromatographic data, baseline correction is performed by derivative method or moving average method, and peak identification is performed by second derivative method or waveform recognition algorithm to extract peak area and retention time; if it is spectral data, absorbance value at a specific wavelength is extracted. In the Grubbs test, the critical value... Based on the number of measurements and significance level (Usually taken as 0.05) is determined by querying a pre-stored Grubbs critical value table, or according to the formula. Calculation, where for The critical value of the distribution.

[0036] The quality control module 140 is signal-connected to the edge preprocessing unit 134 and the process engine module 120. The quality control module 140 is configured to generate process control instructions based on the compliance verification results. When the detected feature value does not meet the quality control rule threshold, the process control instruction is used to trigger the process engine module 120 to suspend the current flow of the instantiated process template and activate the quality control exception handling node. Only after the detected feature value meets the quality control rule threshold can the process engine module 120 allow the instantiated process template to flow to the next detection process node. This effectively intercepts quality at the source of data generation, eliminates the timing misalignment between the quality control node and the data acquisition node, and ensures the compliance of the detection process and the traceability of data quality.

[0037] A process control instruction is a structured message that includes at least: instruction type (e.g., "suspend" or "continue"), target process instance ID, current node ID, verification result (including specific values ​​and thresholds), and timestamp. The quality control module 140 publishes this instruction to a designated topic via a message middleware (e.g., RabbitMQ, Kafka). The process engine module 120 subscribes to this topic. Upon receiving a "suspend" instruction, it sets the status of the corresponding process instance to "suspend" and inserts an "exception handling" task into its task queue. This task typically requires human intervention for review or retesting. Simultaneously, the system interface provides a prominent prompt. After completing the exception handling (e.g., confirming data or re-injecting samples), the human can trigger a "retry verification" or "force pass" operation through the interface. This operation generates a new process control instruction ("continue"), which the engine uses to resume the process. The uncertainty assessment unit reads the predefined template of "uncertainty sources" associated with the detection method node in the knowledge graph (the template lists typical sources, such as sample weighing, standard solution preparation, instrument repeatability, etc.), and combines it with the historical measurement data sequence required for Type A assessment (provided by the data acquisition module) and the instrument calibration certificate information required for Type B assessment (stored in the system equipment management database) to automatically fill in the values ​​of each component in the calculation formula.

[0038] When calculating the combined standard uncertainty, the uncertainty assessment unit uses partial derivatives... The calculation, if the measurement function For simple algebraic expressions (such as If the function is nonlinear, the partial derivatives are automatically calculated using symbolic differentiation; if it is a complex nonlinear function, numerical differentiation methods (such as the central difference method) are used for calculation, with a step size of [missing information]. Take as input quantity Standard uncertainty 0.5 times, that is Covariance The determination is based on historical data statistics; if there is no correlation between input quantities, it is set to 0. The quality control anomaly handling node includes the following processing options: re-inspection, data correction (authorization required), deviation acceptance (electronic signature of the technical person in charge and recording of reasons), and equipment calibration inspection; the processing record includes anomaly description, processing measures, handler, and processing timestamp, and is written as a mandatory field to the blockchain evidence storage module.

[0039] The standard text structuring and deconstruction unit 112 is equipped with a natural language processing engine. This engine is configured to perform clause-level semantic parsing on the standard text to extract metadata about the detection method. This metadata includes detection condition parameters, quality control rule thresholds, and judgment logic relationships. A detection knowledge graph is constructed using a resource description framework. The nodes in the detection knowledge graph include detection method nodes and quality control rule nodes. The attributes of the quality control rule nodes include the relative deviation threshold for parallel samples, the acceptable range for spiked recovery rate, and the method detection limit. The edges of the detection knowledge graph include judgment logic edges, used to represent the conditional triggering relationship between the detection method nodes and the quality control rule nodes. The process template dynamic generation unit 122 also includes a matrix type identification subunit. This subunit is configured to parse the sample matrix description text in the commissioned detection task to identify the matrix complexity level. Based on the matrix complexity level, it calls the corresponding preprocessing step node sequence from the detection knowledge graph to adapt to the digestion, extraction, or purification process nodes required for complex matrix samples. This allows for customized detection processes based on sample type, achieving refined control of detection processes for different matrix samples.

[0040] like Figure 2 As shown, in one embodiment, the process engine module 120 further includes a task load balancing and dispatching unit 124 and a task auto-acceptance unit 126; the task load balancing and dispatching unit 124 is configured to obtain the real-time task queue length of each detection device. Matching coefficient with the qualification of the testing method And calculate the task assignment priority probability. The commissioned testing tasks are assigned to the experimental task nodes associated with the testing equipment that has the highest assignment priority probability. , To detect the total number of devices, the system automatically assigns and batches experimental tasks, balancing the workload of each device. and The first The matching coefficient of the testing method qualification of each testing device and the length of the real-time task queue. For summation and traversal of the detection device index =1,2,…,n; The task auto-assignment unit 126 is configured to acquire the detection capability qualification vector of the experimenters. With current task queue saturation And calculate the matching degree function. ,in, This is a vector of qualification requirements for commissioned testing tasks. To maximize task capacity, when the matching degree When the preset threshold is exceeded, the commissioned testing task will be automatically assigned to the corresponding experimental personnel, realizing flexible configuration and efficient collaboration of task flow.

[0041] like Figure 3 As shown, in one embodiment, the edge preprocessing unit 134 is configured with an outlier detection subunit, a chromatographic data processing subunit, and a curve fitting evaluation subunit; the outlier detection subunit is configured to use the isolated forest algorithm to calculate the outlier of the detected feature values, and when the detected feature value is a parallel sample measurement value, to perform a Grubbs test to calculate the statistic. ,in, This is the Grubbs test statistic. For the first One measured value, The average of the measured values. This represents the sample standard deviation of the measured values.

[0042] When the statistic Greater than the Grubbs threshold At that time, the corresponding abnormal measured values ​​are removed and the retest process is triggered, automatically identifying and processing abnormal values ​​in the raw data, reducing the workload of manual verification; the chromatographic data processing subunit is configured to perform baseline correction and peak identification on the chromatographic data in the raw data stream, and extract the chromatographic peak area. With retention time Retention time Standard retention time of corresponding standards in the detection knowledge graph Compare the results, and when the relative deviation... When the threshold is exceeded, the chromatographic peak is identified as abnormal and a retest process is triggered; when the comparison is successful, the sample concentration is calculated based on the standard curve parameters in the detection knowledge graph. ,in, The slope To achieve the intercept, some device data can be automatically collected and imported, reducing manual input errors.

[0043] The implementation of the Isolation Forest algorithm includes: constructing multiple random binary trees, each recursively partitioning the data by randomly selecting features and random split values ​​until samples are isolated or the tree height limit is reached; calculating the average path length of each sample point, with shorter path lengths considered outliers; algorithm parameters include: the number of trees (usually 100) and the subsampling size (usually 256). In the chromatographic data processing subunit, the preset threshold for retention time comparison is typically set to... 0.05 min or 2 (Choose the more lenient one); if multiple chromatographic peaks exist simultaneously, their identities are confirmed by matching the peak area ratio pattern with that of the standard chromatogram.

[0044] The curve fitting evaluation subunit is configured to perform curve fitting using the Logistic four-parameter equation when the standard curve parameters correspond to a nonlinear response interval. The Logistic four-parameter equation is: ,in, This is the instrument response value. This represents the sample concentration value. The upper limit asymptotic value, The lower bound asymptotic value, The half-maximum response concentration, The slope factor is used; the parameters are iteratively optimized using the Levenberg-Marquardt algorithm, and the Akaike information criterion is adopted. Evaluate the goodness of fit of the model to determine the optimal fitting curve, where For the number of parameters, The maximum likelihood function value is used to automatically select the optimal fitting model and calculate the sample concentration, ensuring the accuracy of the detection data.

[0045] The initial parameters of the Levenberg-Marquardt algorithm are determined using a genetic algorithm or grid search to avoid local optima; the iteration termination condition is: the rate of change of the parameters is less than... Or the change in the sum of squared residuals is less than 0.01. Maximum likelihood function value Calculated by assuming the residuals follow a normal distribution, i.e. ,in These are measured values. These are the fitted values. This is an estimate of the standard deviation.

[0046] In one embodiment, the quality control module 140 further includes an uncertainty assessment unit configured to identify uncertainty source components of the detection feature value, including Type A assessment components. Compared with Category B rating scores Based on the law of uncertainty propagation, when the detection characteristic value is the measurement result of a linear model, the combined standard uncertainty is calculated using the following formula. , ,in, For measurement function, For input quantity, For standard uncertainty, For covariance, The total number of input quantities enables automated assessment of detection uncertainty, improving data reliability.

[0047] In one embodiment, the chemical testing laboratory information management system further includes a blockchain evidence storage module. This module is data-coupled with the data acquisition module and the quality control module. The blockchain evidence storage module is configured to perform hash operations on the raw data stream and testing feature values ​​to generate data fingerprints, and then package the data fingerprints with the timing information of process control instructions to generate blocks that are written to the consortium blockchain. This blockchain evidence storage module is also equipped with a smart contract unit. This smart contract unit is configured to automatically trigger a contract event when the conformity verification result is unqualified, writing the processing record of the quality control anomaly handling node and the digital signature of the reviewer into the next block of the consortium blockchain. This establishes an effective data verification and traceability mechanism, ensuring the immutability of the testing data and the auditability of the entire process.

[0048] The blockchain evidence storage module runs as a microservice within the system. Internally, it maintains a lightweight blockchain client connected to a pre-built consortium blockchain network (e.g., based on Hyperledger Fabric). Data fingerprints are generated using the SHA-256 algorithm. When packaging blocks, in addition to data fingerprints and time-series information, the hash value of the previous block, process instance ID, and current node ID are included, forming a chain structure. The smart contract unit is deployed on the consortium blockchain, and the contract code predefines event listening conditions (e.g., the `checkResult` field in the received message is "unqualified"). When the edge preprocessing unit 134 or the quality control module 140 submits a message containing an unqualified result through the service interface of the blockchain evidence storage module, the contract is automatically triggered, submitting a new transaction containing the message's key fields, a timestamp, and a subsequent review record submitted by the experimenter (signed with their private key) to the blockchain network. After consensus, this transaction is written into a new block.

[0049] The pseudocode logic of the smart contract includes: the function `recordQualityEvent(string memorysampleId, string memory checkResult, bytes memory signature)`, which first verifies the signer's identity (whether they are an authorized quality inspector). If `checkResult` is "unqualified", it creates an event `QualityException` and stores `sampleId`, the current block timestamp, and the transaction hash in the mapping table `qualityLogs`; the function `getQualityHistory(string memory sampleId)` allows external queries to retrieve the complete quality control history of the sample. The consortium blockchain uses the PBFT (Practical Byzantine Fault Tolerance) consensus mechanism, with a block time set to 1-2 seconds. When generating data fingerprints, the original data stream is first serialized (e.g., in Protobuf format) and then hashed to ensure that identical data content generates unique fingerprints.

[0050] In one embodiment, the data acquisition module 130 further includes a sample tracking unit, which is equipped with an RFID reader and a temperature and humidity sensor. The RFID reader is used to read the electronic tag on the sample container to obtain the sample identification, and the temperature and humidity sensor is used to collect temperature and humidity data of the sample storage environment. The sample tracking unit also includes a stability prediction subunit, which is configured to calculate the degradation rate constant of the target analyte in the sample based on the temperature and humidity data and the Arrhenius equation, and predict the sample's shelf-life decay curve based on the degradation rate constant. When the predicted concentration deviation exceeds the allowable range, a retest reminder instruction is generated. The Arrhenius equation is... ,in, This represents the degradation rate constant of the target analyte in the sample. Pre-exponential factor, For activation energy, The gas constant is Using absolute temperature, this sample tracking unit tracks the entire lifecycle of the retained sample, including retention time, location, and conditions, enabling full-cycle digital management of the sample from receipt, processing, testing to disposal.

[0051] The sample tracking unit is linked to a sample information database, recording each sample's unique ID, target analyte, and initial concentration. Storage requirements. The stability prediction subunit embeds a physicochemical parameter database that stores the pre-exponential factors of common analytes in different matrices. and activation energy Typical values ​​(derived from publicly available literature or internal accumulation). This subunit periodically reads data from the temperature and humidity sensor to calculate the absolute temperature. The physicochemical parameters of the analytes in the sample information are then retrieved and substituted into the Arrhenius equation. Calculate the degradation rate constant under the current conditions. ,in This is the gas constant. Combined with the initial concentration... and stored time According to the first-order dynamic equation Predict the current concentration. When A re-inspection reminder will be generated at that time. This refers to the "stability tolerance threshold" obtained from the node of this detection method in the knowledge graph.

[0052] The RFID reader uses UHF RFID (860-960MHz), and the electronic tag adopts the EPCGen2 standard. The stored sample identification is a unique 128-bit code, mapped one-to-one with the system sample ID. The temperature and humidity sensor uses a digital sensor (such as SHT30), communicating with the edge gateway via I2C or Modbus RTU protocol. The data acquisition frequency is set to once every 15 minutes and stored in a time-series database (such as InfluxDB). If the temperature exceeds the preset range during storage (e.g., ...), the system will detect the temperature and humidity. In addition to recording, the system also triggers immediate alarms. For multi-temperature segment storage (such as...), back The stability prediction subunit uses a segmented calculation method to calculate the degradation amount in each temperature range and then sum them up.

[0053] In one embodiment, the chemical testing laboratory information management system further includes a commission management module. This module is coupled to an external LIMS platform via a standardized interface. It is configured to parse the structured data of commission orders pushed by the external LIMS platform, extract sample information, testing items, and testing standard identifiers. Based on the testing standard identifiers, it automatically triggers the standard management module to call the testing method metadata, achieving automatic capture of commission information and seamless integration of business commission management with sample management and testing standard management. The quality control module 140 also includes a data rounding unit. This unit is configured to identify the required number of significant digits based on the testing method metadata in the testing knowledge graph and execute rounding rules. When the leftmost digit of the number to be discarded is 5 and there are no subsequent digits or all are zero, if the last digit to be retained is even, it is discarded; if it is odd, it is rounded up. This satisfies the statistical unbiasedness requirement for chemical testing data rounding, ensuring the standardization of the data output format and the accurate reproduction of the testing results. The process engine module 120 employs a domain-specific language interpreter. This interpreter is configured to parse the primitive identifiers of the detection process nodes in the visualized process orchestration diagram. These primitive identifiers include preprocessing primitives, instrument analysis primitives, and quality control insertion primitives. The interpreter converts these primitive identifiers into executable code and verifies the logical dependencies of the detection process nodes according to chemical logic constraint rules, generating instantiated process templates. This domain-specific language interpreter, based on a low-code platform architecture, enables agile development and rapid feature iteration of the detection process, reducing user learning costs. The system also includes a statistical analysis module and a system management module. The statistical analysis module is data-coupled with the quality control module and is configured to perform multi-dimensional statistical queries and data analysis based on the timing information of detection feature values ​​and process control instructions. It supports the generation of custom report templates and periodic automatic report generation, and provides export and printing functions for statistical reports. The system management module is configured to implement permission settings and access control management for departments, positions, and personnel, provides system log recording and audit trail functions, and employs high availability and disaster recovery technologies to deploy the system architecture, ensuring stable system operation and data security.

[0054] The standardized interface provided by the entrustment management module is a RESTful API based on the HTTP protocol, receiving entrustment data pushed by external systems in JSON format. Before performing rounding, the data rounding unit first queries the "significant digits" or "decimal places" attribute under the current detection method node in the knowledge graph as the basis for rounding. The domain-specific language interpreter's core is a set of predefined syntax rules and converters. Each element in the visualization orchestration diagram corresponds to a JSON structure in the background, describing its type (Type), configuration parameters (Config), and connection relationships (Connections). The interpreter traverses this JSON structure, calls the corresponding code generation template based on the Type field (e.g., the "pretreatment-extraction" template generates a series of parameter settings and instrument control instructions), and converts Connections into flow control logic (e.g., sequence, branching). Chemical logic constraint rules are implemented in the form of a rule engine; for example, a rule could be "quality control inserted elements must be located after the instrument analysis elements they monitor," and the interpreter will check and report errors during the conversion process. The system management module adopts a role-based access control model, with permission granularity configurable down to specific function menus and data types. The high availability architecture is achieved by deploying a load balancer and at least two application server instances, with the database using master-slave replication.

[0055] The specific syntax definition of a domain-specific language (DSL) includes: the node definition statement NODE.<node_id> TYPE <type>CONFIG { ...}, the connection definition statement CONNECT<from_node> TO<to_node> CONDITION <expr>];in <type>This includes enumerated values ​​such as "Digestion", "Extraction", "GC_Analysis", and "QC_Parallel". Chemical logic constraints also include: prohibiting circular dependencies (detecting loops by traversing the graph), checking necessary quality control nodes (ensuring quality control follows critical steps), and checking for resource conflicts (such as preventing the same equipment from being used by two processes simultaneously). The RESTful API's JSON message format has a predefined schema, including fields: order number (string), sample list (array), detection item code (string, associated knowledge graph standard identifier), and required completion time (ISO 8601 format). Database master-slave replication uses an asynchronous replication mode, with the master handling write operations and the slave handling read operations, achieving automatic failover through Keepalived or a similar mechanism. System logs use a structured log format (such as JSON), recording user operations, data changes before and after, timestamps, and IP addresses, stored in Elasticsearch for auditing and retrieval.

[0056] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0057] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.< / type> < / expr> < / type>

Claims

1. A chemical testing laboratory information management system, characterized in that, It includes a standard management module, a process engine module, a data acquisition module, and a quality control module; The standard management module is configured with a standard text structure deconstruction unit, which is used to perform clause-level semantic parsing on the detection standard text to extract detection method metadata, and to construct a detection knowledge graph based on the detection method metadata. The node attributes of the detection knowledge graph include detection condition parameters, quality control rule thresholds, and judgment logic relationships. The process engine module is data coupled with the standard management module. The process engine module includes a process template dynamic generation unit. The process template dynamic generation unit is configured to call the corresponding detection method metadata from the detection knowledge graph according to the target detection standard identifier associated with the received commissioned detection task, and automatically reconstruct the detection process node sequence and the quality control node embedding order according to the quality control rule threshold and judgment logic relationship, so as to generate an instantiated process template that matches the target detection standard. The data acquisition module includes an instrument interface adaptation unit and an edge preprocessing unit. The instrument interface adaptation unit is used to acquire the raw data stream output by the testing instrument through multi-protocol conversion. The edge preprocessing unit is deployed in the data link between the testing instrument and the quality control module. The edge preprocessing unit is configured to parse the raw data stream in real time to extract detection feature values, and to perform compliance verification on the detection feature values ​​according to the quality control rule threshold corresponding to the current process node in the instantiated process template. The edge preprocessing unit is configured with an outlier detection subunit, a chromatographic data processing subunit, and a curve fitting evaluation subunit. The outlier detection subunit is configured to use the isolated forest algorithm to calculate the outlier of the detected feature values, and when the detected feature value is a parallel sample measurement value, to perform a Grubbs test to calculate the statistic. ， in, This is the Grubbs test statistic. For the first One measured value, The average of the measured values. The sample standard deviation of the measured values; When the statistic Greater than the Grubbs threshold At that time, the corresponding abnormal measurement values ​​are removed and a retest process is triggered; the chromatographic data processing subunit is configured to perform baseline correction and peak identification on the chromatographic data in the original data stream, and extract the chromatographic peak area. With retention time The retention time Standard retention time of the corresponding standard in the detection knowledge graph Compare the results, and when the relative deviation... When the threshold is exceeded, the chromatographic peak is identified as abnormal and a retest process is triggered; when the comparison is successful, the sample concentration is calculated based on the standard curve parameters in the detection knowledge graph. ， in, The slope The intercept; The curve fitting evaluation subunit is configured to perform curve fitting using a Logistic four-parameter equation when the standard curve parameters correspond to a nonlinear response interval. The Logistic four-parameter equation is: ， in, This is the instrument response value. This represents the sample concentration value. The upper limit asymptotic value, The lower bound asymptotic value, The half-maximum response concentration, Slope factor; The parameters were iteratively optimized using the Levenberg-Marquardt algorithm, and the Akaike information criterion was adopted. Evaluate the goodness of fit of the model to determine the optimal fitting curve, where For the number of parameters, The maximum likelihood function value, The quality control module is signal-connected to the edge preprocessing unit and the process engine module. The quality control module is configured to generate process control instructions based on the compliance verification results. When the detection feature value does not meet the quality control rule threshold, the process control instruction is used to trigger the process engine module to suspend the current flow of the instantiated process template and activate the quality control exception handling node. Only after the detection feature value meets the quality control rule threshold will the process engine module allow the instantiated process template to flow to the next detection process node.

2. The chemical testing laboratory information management system according to claim 1, characterized in that, The standard text structure deconstruction unit is equipped with a natural language processing engine. The natural language processing engine is configured to perform clause-level semantic parsing on the detection standard text to extract detection method metadata. The detection method metadata includes detection condition parameters, quality control rule thresholds, and judgment logic relationships, and constructs the detection knowledge graph in the form of a resource description framework. The nodes of the detection knowledge graph include detection method nodes and quality control rule nodes. The attributes of the quality control rule nodes include the relative deviation threshold of parallel samples, the qualified range of spiked recovery rate, and the method detection limit value. The edges of the detection knowledge graph include judgment logic edges, which are used to characterize the conditional triggering relationship between the detection method nodes and the quality control rule nodes. The process template dynamic generation unit also includes a matrix type identification subunit. The matrix type identification subunit is configured to parse the sample matrix description text in the commissioned testing task to identify the matrix complexity level, and call the corresponding preprocessing step node sequence from the testing knowledge graph according to the matrix complexity level to adapt to the digestion, extraction or purification process nodes required for complex matrix samples.

3. The chemical testing laboratory information management system according to claim 1, characterized in that, The process engine module also includes a task load balancing and dispatching unit and a task auto-claiming unit; the task load balancing and dispatching unit is configured to obtain the real-time task queue length of each detection device. Matching coefficient with the qualification of the testing method And calculate the task assignment priority probability. The entrusted testing task is assigned to the experimental task node associated with the testing device that has the highest assignment priority probability, wherein, ， The total number of testing devices. and The first The matching coefficient of the testing method qualification of each testing device and the length of the real-time task queue. To sum and traverse the index of the detection devices, =1,2,…,n; The task auto-assignment unit is configured to obtain the detection capability qualification vector of the experimenters. With current task queue saturation And calculate the matching degree function. ， in, This is the qualification requirement vector for the commissioned testing task. To maximize task capacity, when the matching degree When the preset threshold is exceeded, the commissioned testing task will be automatically assigned to the corresponding experimental personnel.

4. The chemical testing laboratory information management system according to claim 1, characterized in that, The quality control module further includes an uncertainty assessment unit, which is configured to identify the uncertainty source components of the detection feature value, including Type A assessment components. Compared with Category B assessment scores Based on the uncertainty propagation law, when the detected characteristic value is the measurement result of a linear model, its combined standard uncertainty is calculated using the following formula. , ， in, For measurement function, For input quantity, For standard uncertainty, For covariance, This represents the total number of input values.

5. The chemical testing laboratory information management system according to claim 1, characterized in that, It also includes a blockchain evidence storage module, which is data-coupled with the data acquisition module and the quality control module. The blockchain evidence storage module is configured to perform hash operations on the raw data stream and the detection feature value to generate a data fingerprint, and package the data fingerprint and the timing information of the process control instructions to generate a block and write it into the consortium blockchain. The blockchain evidence storage module is also configured with a smart contract unit, which is configured to automatically trigger a contract event when the result of the conformity verification is unqualified, and write the processing record of the quality control anomaly handling node and the digital signature of the reviewer into the next block of the consortium blockchain.

6. The chemical testing laboratory information management system according to claim 1, characterized in that, The data acquisition module further includes a sample tracking unit, which is equipped with an RFID reader and a temperature and humidity sensor. The RFID reader is used to read electronic tags on the sample container to obtain sample identification, and the temperature and humidity sensor is used to collect temperature and humidity data of the sample storage environment. The sample tracking unit also includes a stability prediction subunit, which is configured to calculate the degradation rate constant of the target analyte in the sample based on the temperature and humidity data and the Arrhenius equation, and predict the sample's shelf-life decay curve based on the degradation rate constant. When the predicted concentration deviation exceeds the allowable range, a retest reminder instruction is generated. The Arrhenius equation... in, This represents the degradation rate constant of the target analyte in the sample. Pre-exponential factor, For activation energy, The gas constant is This refers to absolute temperature.

7. The chemical testing laboratory information management system according to claim 1, characterized in that, The system also includes a commission management module, which is coupled with an external LIMS platform through a standardized interface. This module is configured to parse the structured data of the commission order pushed by the external LIMS platform, extract sample information, testing items, and testing standard identifiers, and automatically trigger the standard management module to call the testing method metadata based on the testing standard identifiers. The quality control module also includes a data rounding unit, which is configured to identify the required number of significant digits based on the testing method metadata in the testing knowledge graph, and execute rounding rules (rounding to the nearest even number). When the leftmost digit of the number to be discarded is 5 and there are no subsequent digits or all are zero, if the last digit to be retained is even, it is discarded; if it is odd, it is rounded up.

8. The chemical testing laboratory information management system according to claim 1, characterized in that, The process engine module employs a domain-specific language interpreter, which is configured to parse the primitive identifiers of the detection process nodes in the visual process orchestration diagram. The primitive identifiers include preprocessing primitives, instrument analysis primitives, and quality control insertion primitives. The interpreter converts the primitive identifiers into executable code and verifies the logical dependencies of the detection process nodes according to chemical logic constraint rules, thereby generating the instantiated process template.

9. The chemical testing laboratory information management system according to claim 1, characterized in that, The system also includes a statistical analysis module and a system management module. The statistical analysis module is data-coupled with the quality control module and is configured to perform multi-dimensional statistical queries and data analysis based on the timing information of the detection feature values ​​and the process control instructions. It supports the generation of custom report templates and the automatic generation of reports periodically, and realizes the export and printing functions of statistical reports. The system management module is configured to manage the permissions and access control of departments, positions and personnel, provide system log recording and audit trail functions, and adopts high availability and disaster recovery technologies to deploy the system architecture to ensure stable system operation and data security.

Images