Drug temperature and humidity digital monitoring and quality tracing system based on full life cycle

Abstract

The invention provides a drug temperature and humidity digital monitoring and quality tracing system based on a full life cycle, and relates to the technical field of digital monitoring, and the system comprises the steps: firstly collecting the temperature and humidity data of the whole link of drug production, storage, transportation to terminal sales, and obtaining a standardized data set through the verification and packaging of an industrial standard library; performing secondary screening through primary screening in combination with a dynamic temperature and humidity screening model to obtain a precise abnormal data set; associating drug full-cycle multi-class information, and constructing a temperature and humidity-drug state mapping relation based on a time sequence association algorithm to form a full-dimension association data set; based on the data set, combining with a quality risk assessment model for early warning, realizing full-period traceability through a reverse traceability algorithm, and generating an assessment result and a traceability link; and finally, relevant data are converted into a dynamic temperature and humidity monitoring curve, a quality risk early warning report and a full-life-cycle tracing atlas, and an interactive report is assembled, so that the temperature and humidity state of the medicine can be checked conveniently.

Description

Technical Field

[0001] This invention relates to the field of digital monitoring technology, specifically a digital monitoring and quality traceability system for drug temperature and humidity based on the entire life cycle. Background Technology

[0002] As special commodities, pharmaceuticals are highly susceptible to abnormal fluctuations in temperature and humidity throughout their entire lifecycle, from production, storage, and transportation to final sales. These fluctuations can easily lead to changes in drug composition, reduced efficacy, and even the generation of harmful substances. Existing traceability systems primarily focus on recording drug distribution information, failing to effectively link temperature and humidity data with drug quality status. This hinders accurate risk warnings based on temperature and humidity anomalies and makes it difficult to quickly and accurately trace back from the end consumer to key temperature and humidity nodes throughout the entire lifecycle. This poses a significant challenge to drug quality control and fails to meet the pharmaceutical industry's demand for end-to-end temperature and humidity monitoring and precise traceability for high-quality development.

[0003] Therefore, a digital monitoring and quality traceability system for drug temperature and humidity based on the entire life cycle is provided. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide a digital monitoring and quality traceability system for drug temperature and humidity based on the entire life cycle.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a digital monitoring and quality traceability system for drug temperature and humidity based on the entire life cycle, the system comprising: a full-cycle data acquisition module, a temperature and humidity screening module, a data fusion processing module, a quality traceability analysis module, and a visualization display module; The full-cycle data acquisition module is used to acquire temperature and humidity data of pharmaceuticals at each stage of their entire lifecycle, from production, storage, transportation to terminal sales. It combines the collected temperature and humidity data with the pharmaceutical industry temperature and humidity standard library to perform format verification and standardized packaging, resulting in a full-cycle standardized temperature and humidity dataset. The temperature and humidity filtering module is used to perform a first-stage filtering on the full-cycle standardized temperature and humidity dataset, and then perform a second-stage filtering on the data selected in the first-stage filtering, combined with the dynamic temperature and humidity filtering model, to obtain an accurate temperature and humidity anomaly dataset. The data fusion processing module is used to associate batch information, production information, warehousing information, and transportation information of the entire life cycle of the drug with the precise temperature and humidity anomaly dataset. Based on the time-series association algorithm, it constructs a temperature and humidity-drug state mapping relationship to obtain a full-dimensional associated dataset. The quality traceability analysis module, based on the aforementioned full-dimensional associated dataset, combines a drug quality risk assessment model to provide early warning of temperature and humidity exceeding standards, and uses a reverse traceability algorithm to achieve traceability tracking from the end drug to the temperature and humidity data throughout its entire life cycle, thereby obtaining quality risk assessment results and traceability link information; The visualization module generates dynamic temperature and humidity monitoring curves, quality risk early warning reports, and full life cycle traceability maps based on the accurate temperature and humidity anomaly dataset, quality risk assessment results, and traceability link information. These are then assembled into an interactive traceability monitoring report, facilitating access to the temperature and humidity status of the medicine.

[0006] Furthermore, the process of filtering the full-cycle standardized temperature and humidity dataset includes: The system iterates through the temperature and humidity data in the full-cycle standardized temperature and humidity dataset and compares it with the temperature and humidity threshold ranges of corresponding drugs under different scenarios in the pharmaceutical industry temperature and humidity standard library. It obtains the temperature and humidity data of the corresponding drugs that exceed or fall below the reasonable temperature and humidity threshold range throughout the entire life cycle, and marks the corresponding collection time of the comparison results to obtain the first screening and positioning time.

[0007] Furthermore, the process of performing a secondary screening of the data selected in the first screening, combined with a dynamic temperature and humidity screening model, includes: A dynamic temperature and humidity screening model is constructed to obtain the humidity jump threshold for different drug scenarios; based on the humidity jump threshold and the time of a screening location, the corresponding compliance anomaly data is obtained; based on the temperature and humidity data and compliance anomaly data corresponding to the time of a screening location, a precise temperature and humidity anomaly dataset is obtained.

[0008] Furthermore, the process of constructing a dynamic temperature and humidity screening model includes: Acquire temperature and humidity data from several historical collection periods, repeat the filtering process once, acquire the first filtering positioning time corresponding to several historical collection periods, and then acquire temperature and humidity data corresponding to several adjacent collection times adjacent to the corresponding first filtering positioning time. Based on the temperature and humidity data of several adjacent collection times corresponding to a historical collection period, the continuous jump rate of the temperature and humidity data at the corresponding collection time is calculated. Grouping and labeling several sets of corresponding continuous jump rates, denoted as ; For natural numbers; The continuous jump rate of the group is used as sample data, and Less than The natural numbers are used to obtain the mean of the sample data, which is denoted as the sample set; the continuous jump rate of the remaining groups is used as the test set; and a training sample set is formed based on the sample set and the test set. A standard screening model is constructed based on a long short-term memory network. The training sample set is then input into the standard screening model to train it. The resulting standard screening model is denoted as the dynamic temperature and humidity screening model.

[0009] Furthermore, the process of constructing a temporal correlation matching algorithm includes: Using the batch number of each data point in the accurate temperature and humidity anomaly dataset as the primary matching condition, the full lifecycle association information corresponding to the same batch of drugs is filtered out. Using the process code of the corresponding data as the secondary matching condition, the temperature and humidity anomaly data is initially associated with the association information of the corresponding process. Using the collection timestamp of the corresponding data as the tertiary matching condition, the time overlap between the temperature and humidity anomaly collection time and the association information of the corresponding process is calculated, and a time matching threshold is set. When the temperature and humidity anomaly collection time is within the time interval of the association information, or the deviation from the key time node of the association information is within the threshold range, it is determined that the match is successful.

[0010] Furthermore, the process of conducting quality traceability analysis on the full-dimensional associated dataset includes: Construct a drug quality risk assessment model; based on the quality risk assessment results of the drug quality risk assessment model, classify the risk warning of exceeding temperature and humidity standards; construct a reverse traceability algorithm to generate traceability link information.

[0011] Furthermore, the process of constructing a reverse tracing algorithm includes: Using the batch number or traceability code of the terminal drug as the traceability entry point, the batch number of the drug is obtained by scanning the traceability code on the terminal drug packaging; using the obtained batch number as the search key, information on all life cycle stages and temperature and humidity data corresponding to the batch of drugs is retrieved in the full-dimensional associated dataset; following the reverse order of terminal sales-transportation-warehousing-production, the circulation nodes of the batch of drugs are sorted out and the information corresponding to each node is extracted.

[0012] Furthermore, the process of generating dynamic temperature and humidity monitoring curves, quality risk early warning reports, and full lifecycle traceability maps based on the aforementioned precise temperature and humidity anomaly dataset, quality risk assessment results, and traceability link information includes: The system extracts full lifecycle temperature and humidity data for the same batch of drugs from a precise temperature and humidity anomaly dataset, sorts them chronologically, and renders a dynamic temperature and humidity monitoring curve. It also extracts core information for each batch of drugs from the quality risk assessment results and renders a risk warning table. The system statistically analyzes global risk level distribution data and renders a risk distribution dashboard. When a new risk warning is generated, the rendering engine automatically triggers report updates to achieve real-time display of warning information. Finally, it extracts three-layer structure data of nodes, events, and data from traceability information and renders a directed traceability graph according to the flow sequence of production-warehousing-transportation-sales.

[0013] Compared with existing technologies, the beneficial effects of this invention are as follows: This system effectively solves the problems of incomplete temperature and humidity monitoring, inaccurate data processing, and weak traceability in existing technologies by constructing a digital monitoring and quality traceability system for drug temperature and humidity throughout the entire life cycle. The full-cycle data acquisition module realizes the comprehensive acquisition and standardized processing of temperature and humidity data at all stages of drug production, laying a unified and reliable data foundation for subsequent analysis; the temperature and humidity screening module adopts a two-level screening mechanism, combined with a dynamic screening model, to improve the accuracy of abnormal data identification and avoid the problems of false positives and false negatives caused by fixed threshold screening; the data fusion processing module constructs a mapping relationship between temperature and humidity and drug status through a time-series correlation algorithm, realizing the deep fusion of multi-dimensional data; the quality traceability analysis module realizes accurate early warning of temperature and humidity exceeding the standard risk and full-chain reverse traceability, providing precise targeting for drug quality control; the visualization display module improves the system's usability and information transmission efficiency through intuitive dynamic curves, early warning reports, and traceability maps. Overall, the system achieves full-cycle coverage of drug temperature and humidity monitoring, accurate identification of abnormal data, early warning of quality risks, and clear presentation of traceability links. It strengthens the quality control capabilities throughout the entire drug lifecycle, ensures drug quality and safety, and improves the level of intelligent quality supervision in the pharmaceutical industry. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0015] Figure 1 This is a schematic diagram of a module for a digital monitoring and quality traceability system for drug temperature and humidity based on the entire life cycle. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0017] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] like Figure 1 As shown, a digital monitoring and quality traceability system for drug temperature and humidity based on the entire life cycle is provided. The system includes, but is not limited to, a full-cycle data acquisition module, a temperature and humidity screening module, a data fusion and processing module, a quality traceability analysis module, and a visualization display module. The full-cycle data acquisition module is used to acquire temperature and humidity data of pharmaceuticals at each stage of their entire lifecycle, from production, storage, transportation to terminal sales. It combines the collected temperature and humidity data with the pharmaceutical industry temperature and humidity standard library to perform format verification and standardized packaging, resulting in a full-cycle standardized temperature and humidity dataset. In practical applications, temperature and humidity data can be obtained at each stage based on the specific scenario of the entire drug lifecycle. Different scenarios correspond to different acquisition targets, acquisition devices, and acquisition requirements, as detailed below: For example, in the pharmaceutical manufacturing scenario, the ambient temperature and humidity of each key workstation in the entire pharmaceutical manufacturing process, as well as the process-related temperature and humidity, can be used as the temperature and humidity data to be collected; in the pharmaceutical warehousing scenario, the ambient temperature and humidity of each area of ​​the pharmaceutical storage warehouse can be used as the temperature and humidity data to be collected; in the pharmaceutical transportation scenario, the real-time temperature and humidity inside the cold chain carrier (such as refrigerated trucks, refrigerated boxes, etc.) during the pharmaceutical transportation process, as well as the ambient temperature and humidity along the transportation route, can be used as the temperature and humidity data to be collected. The process of acquiring temperature and humidity data in different scenarios includes, but is not limited to, scenario positioning and data range determination, deployment of acquisition equipment, configuration of acquisition parameters, and preliminary data acquisition and labeling. In an exemplary embodiment, the process of acquiring temperature and humidity data in a pharmaceutical manufacturing scenario includes: identifying the specific process steps corresponding to the pharmaceutical manufacturing process, determining the key workstations where temperature and humidity data need to be collected, covering the entire production process including raw material synthesis, formulation production, and packaging, with core data including environmental temperature and humidity at each workstation, and process-related temperature and humidity (such as material-related temperature and humidity in the drying process and cavity temperature and humidity in the sterilization process). Based on the spatial layout of the production workshop, process characteristics, and temperature and humidity monitoring accuracy requirements, corresponding acquisition equipment is deployed: industrial-grade high-precision temperature and humidity sensors are installed at key workstations such as the mixing area, reaction area, drying area, filling area, and packaging area, with sensor accuracy meeting production process requirements (e.g., temperature accuracy ±0.1℃, humidity accuracy ±1%RH); embedded temperature and humidity acquisition modules are deployed inside process equipment such as sterilizers and drying ovens to achieve direct acquisition of temperature and humidity within the equipment; all acquisition equipment is connected to the industrial control network of the production workshop, supporting real-time data transmission. Based on the requirements of the production process, the data acquisition frequency is configured as follows: For dynamic processes (such as reaction and filling), the acquisition frequency is set to once per minute to ensure the capture of temperature and humidity fluctuations during the process; for static processes (such as static drying), the acquisition frequency can be adjusted to once every 5-10 minutes to reduce data redundancy while ensuring data integrity. Data acquisition trigger conditions are configured synchronously; when temperature and humidity values ​​approach the upper / lower limit of the process threshold, the acquisition frequency is automatically increased to once every 30 seconds to improve monitoring accuracy. The acquisition equipment collects temperature and humidity data in real time according to the configured parameters, synchronously associating the acquisition timestamp, equipment number, workstation code, and corresponding drug production batch number, etc., and initially identifies each piece of acquired data to ensure that the data is traceable to a specific production stage and drug batch. The acquired data is transmitted to the local data acquisition gateway in the production workshop through the industrial control network to complete the initial acquisition of temperature and humidity data for the production scenario.

[0019] The process of acquiring temperature and humidity data in other scenarios is set according to actual needs, and the general process is similar to that described above, so we will not go into too much detail.

[0020] In practical applications, based on a pre-defined standard association method, the existing standard text data of the pharmaceutical industry is structured and parsed to extract core fields such as temperature and humidity threshold ranges, storage requirements, impact levels of exceeding limits, and data recording specifications corresponding to different drug dosage forms. A field mapping table is then established to construct a pharmaceutical industry temperature and humidity standard library for the corresponding drugs. Based on the constructed pharmaceutical industry temperature and humidity standard library, the raw temperature and humidity data collected throughout the entire lifecycle of the corresponding drugs undergoes multi-dimensional format verification: First, field integrity verification: by comparing with the standard library field mapping table, each piece of collected data is checked to see if it contains required fields such as temperature and humidity values, collection timestamps, equipment numbers, and associated drug batch numbers. Data lacking required fields is marked as incomplete data. Second, data type verification: the data type of each field is checked to see if it conforms to the standard library specifications. For example, temperature and humidity values ​​must be numeric, collection timestamps must be time-based, and equipment numbers must be character-based. Data with mismatched types is marked as abnormal data. Third, compliance verification: temperature and humidity data exceeding or falling below a certain proportional relationship (such as one or two times) of the temperature and humidity thresholds for the corresponding drugs in the standard library are removed.

[0021] For valid data that passes multi-dimensional format verification, standardization and encapsulation are performed according to the specifications of the pharmaceutical industry temperature and humidity standard library: First, unit standardization: all temperature and humidity values ​​are uniformly converted to the "℃ (temperature) / %RH (humidity)" unit format specified by the standard library, such as converting Fahrenheit temperature data to Celsius and converting percentage decimal humidity data to percentage integers; Second, time format standardization: timestamps generated by different acquisition devices are uniformly converted to the UTC+8 standard time format (yyyy-MM-ddHH:mm:ss) specified by the standard library to ensure consistency in the time dimension; Third, location information standardization: referring to the spatial coding specifications in the standard library, location information such as warehouse partition names, transportation route nodes, and terminal sales location names are included; Fourth, association and encapsulation: according to the unified dimension of drug batch - life cycle stage - acquisition time - standard threshold - actual temperature and humidity value - acquisition device information, the standardized field data are associated and integrated, and each data is assigned a unique full-cycle data acquisition identifier (ID), and a verification pass mark and corresponding standard library association number are attached.

[0022] The standardized and encapsulated data from various scenarios are globally integrated, and incomplete and invalid data that fails verification are removed to form a full-lifecycle standardized temperature and humidity dataset that is traceable to all stages of the entire lifecycle. This provides a high-quality data foundation for the subsequent abnormal data screening and deviation correction of the temperature and humidity screening module.

[0023] The temperature and humidity filtering module is used to perform a first-stage filtering on the full-cycle standardized temperature and humidity dataset, and then perform a second-stage filtering on the data selected in the first-stage filtering, combined with the dynamic temperature and humidity filtering model, to obtain an accurate temperature and humidity anomaly dataset. In practical applications, the screening process includes the following steps: A1: Traverse through the temperature and humidity data in the full-cycle standardized temperature and humidity dataset; A2: Compare the temperature and humidity threshold ranges of the corresponding drugs under different scenarios in the pharmaceutical industry temperature and humidity standard library to obtain the temperature and humidity data of the corresponding drugs throughout their entire life cycle that exceed or fall below the reasonable temperature and humidity threshold range, and mark the corresponding collection time of the comparison results to obtain the first screening and positioning time.

[0024] In steps A1 to A2, abnormal temperature and humidity data are obtained from the full-cycle standardized temperature and humidity dataset, and the corresponding collection time is located. In order to improve the quality of digital monitoring of drug temperature and humidity throughout the entire life cycle, it is necessary to filter the temperature and humidity data of several adjacent collection times, which will result in secondary filtering. In order to improve the filtering efficiency, intelligent algorithms are further adopted, which can improve both efficiency and accuracy.

[0025] It should be noted that in step A2, the reasonable temperature and humidity threshold range refers to the data collected by the corresponding sensor under normal working conditions; In practical applications, the process of secondary screening includes the following steps: B1: Construct a dynamic temperature and humidity screening model to obtain the humidity jump threshold for different drugs in different scenarios; B2: Obtain the corresponding compliance anomaly data based on the humidity jump threshold and the time of a single screening and location; B3: Based on the temperature and humidity data and compliance anomaly data corresponding to the time of a single screening and positioning, obtain an accurate temperature and humidity anomaly dataset.

[0026] In steps B1 to B3, the humidity jump thresholds of multiple corresponding drug types in corresponding scenarios are obtained based on the constructed dynamic temperature and humidity screening model. By screening and locating the time, the corresponding compliant abnormal data are further screened out. Through data fusion, a precise temperature and humidity abnormal dataset for the corresponding drug in different scenarios is formed.

[0027] In this embodiment, there are issues with data accuracy and processing efficiency in the digital monitoring of drug temperature and humidity throughout its entire life cycle. These issues are addressed through a two-layer screening mechanism and intelligent algorithms.

[0028] In practical applications, the process of constructing a dynamic temperature and humidity screening model includes the following steps: C1: Obtain temperature and humidity data from several historical collection periods, repeat the filtering process once, obtain the first filtering positioning time corresponding to several historical collection periods, and then obtain the temperature and humidity data corresponding to several adjacent collection times adjacent to the corresponding first filtering positioning time. C2: Calculate the continuous jump rate of temperature and humidity data at the corresponding acquisition time based on several adjacent acquisition times corresponding to a historical acquisition cycle. C3: Group and label several sets of corresponding continuous jump rates, denoted as... ; For natural numbers; The continuous jump rate of the group is used as sample data, and Less than The natural numbers are used to obtain the mean of the sample data, which is denoted as the sample set; the continuous jump rate of the remaining groups is used as the test set; and a training sample set is formed based on the sample set and the test set. C4: Based on the Long Short-Term Memory network, a standard screening model is constructed; and the training sample set is input into the standard screening model to train the standard screening model. The standard screening model after training is denoted as the dynamic temperature and humidity screening model.

[0029] It should be noted that in step C2, the continuous jump rate refers to the ratio of the absolute difference between the temperature and humidity data differences between several adjacent collection times before a screening and positioning time and the temperature and humidity data differences between several adjacent collection times after a screening and positioning time to the median value of the standard temperature and humidity range at the corresponding collection time.

[0030] Steps C1 to C4 describe the model building process for the Long Short-Term Memory (LSTM) network. The parameter settings for the model are consistent with those for other models based on LTM networks, except for the differences in data processing and evaluation criteria. All other settings are adaptive and will not be elaborated further.

[0031] For example, the standard temperature range is 3~5℃. Temperature data from the three time points preceding a single screening and positioning time are taken, such as 4℃, 3℃, and 5℃; temperature data from the three time points following a single screening and positioning time are also taken, such as 3℃, 3℃, and 5℃. The temperature differences between the first three time points are 1℃ and -2℃; the temperature differences between the last three time points are 0℃ and -2℃. The median value of the standard temperature and humidity range is 4℃, therefore the corresponding continuous temperature jump rate is: .

[0032] The data fusion processing module is used to associate batch information, production information, warehousing information, and transportation information of the entire life cycle of the drug with the precise temperature and humidity anomaly dataset. Based on the time-series association algorithm, it constructs a temperature and humidity-drug state mapping relationship to obtain a full-dimensional associated dataset. In practical applications, the core dimensions of drug lifecycle-related information are clearly defined. Data integration with heterogeneous systems such as Manufacturing Execution System (MES), Warehouse Management System (WMS), and Transportation Management System (TMS) is achieved through API interfaces and direct database connections, enabling the synchronous access and caching of drug lifecycle-related information and precise temperature and humidity anomaly datasets. This drug lifecycle-related information includes: batch basic information, including but not limited to batch number, dosage form, specifications, production execution standards, and expiration date; production information, including but not limited to production process nodes, start and end times of each process, process parameters, and operator numbers; warehousing information, including but not limited to warehouse zoning codes, inbound time, outbound time, storage location, and warehouse management personnel; and transportation information, including but not limited to cold chain carrier number, transportation route nodes, departure time, arrival time, transportation personnel information, and cold chain equipment operating status.

[0033] In practical applications, the steps for constructing a time-series correlation matching algorithm based on the timeline features of each stage of the drug's entire lifecycle include: D1: Using the batch number of the drug in each data point in the precise temperature and humidity anomaly dataset as the primary matching condition, filter out the full life cycle association information of the same batch of drugs; D2: Using the corresponding data's process code as the secondary matching condition, preliminarily associate the abnormal temperature and humidity data with the associated information of the corresponding process (such as production drying process, cold storage area, and transport refrigerated truck); D3: Using the data collection timestamp as the third-level matching condition, calculate the time overlap between the abnormal temperature and humidity collection time and the corresponding link-related information, set a time matching threshold, and determine that the match is successful when the abnormal temperature and humidity collection time is within the time interval of the related information, or the deviation from the key time node of the related information is within the threshold range.

[0034] It should be noted that, for abnormal situations in the time-series correlation process (such as multiple correlation information corresponding to the same timestamp, or no matching correlation information for temperature and humidity data), the following abnormal handling rules are set: for data with many-to-one matching, mark "correlation information redundancy" and retain all correlation records; for data without matching, mark "correlation information missing" and add a note field indicating "information of the corresponding stage not obtained".

[0035] In practical applications, based on the time-series correlation results and combined with the pharmaceutical industry's temperature and humidity standard library, a mapping rule library for the mapping relationship between temperature and humidity and drug status is constructed. In this embodiment, the core dimensions of the mapping rules include: duration of anomaly, magnitude of anomaly deviation (in percentage terms), and drug sensitivity. It should be noted that the mapping rule library for generating the temperature and humidity-drug status mapping relationship contains core fields such as drug batch number, process code, temperature and humidity anomaly collection time, anomaly type, deviation range, duration, and drug quality status label, so as to achieve accurate association between temperature and humidity anomaly data and drug quality status.

[0036] In practical applications, a full-dimensional association dataset is constructed based on the mapping rule base, the corresponding accurate temperature and humidity anomaly dataset, and the drug's full life cycle association information. The dataset follows a unified dimension of batch-process-time-temperature and humidity anomaly-drug status-association information.

[0037] It should be noted that the integrated full-dimensional related dataset undergoes global deduplication and consistency verification: First, deduplication is performed to remove duplicate data caused by system integration or matching rules; second, consistency verification is performed to check whether there are logical conflicts in the related information of the same batch of drugs at different stages (such as the warehousing entry time being later than the transportation arrival time). Conflicting data is marked as "information logic abnormal" and pushed to the manual review node.

[0038] The quality traceability analysis module, based on the aforementioned full-dimensional associated dataset, combines a drug quality risk assessment model to provide early warning of temperature and humidity exceeding standards, and uses a reverse traceability algorithm to achieve traceability tracking from the end drug to the temperature and humidity data throughout its entire life cycle, thereby obtaining quality risk assessment results and traceability link information; In practical applications, the steps for quality traceability analysis of a full-dimensional correlated dataset are as follows: G1: Constructing a drug quality risk assessment model; G2: Based on the quality risk assessment results of the drug quality risk assessment model, classify the risk warning level for exceeding temperature and humidity limits; G3: Construct a reverse tracing algorithm to generate tracing link information.

[0039] In step G1, a drug quality risk assessment model is constructed based on historical data from the full-dimensional correlation dataset. The model input consists of four dimensions: anomaly type, anomaly duration, anomaly deviation magnitude, and drug dosage form sensitivity coefficient. The output is the quality risk level (Level I / II / III, corresponding to low / medium / high risk). The model construction and training process is as follows: The weights of each input feature were determined using the Analytic Hierarchy Process (AHP), including the drug dosage form sensitivity coefficient, the weight of the magnitude of the abnormal deviation, the weight of the duration of the abnormality, and the weight of the type of temperature and humidity abnormality. Anomalies in temperature and humidity from historical collection periods are extracted from the full-dimensional correlated dataset. Combined with drug quality inspection results (qualified / unqualified / pending re-inspection), a training sample set containing input features and risk level labels is constructed. This training sample set is input into the model, and the Gradient Boosting Decision Tree (GBDT) algorithm is used for model training. Model parameters are adjusted through cross-validation to optimize the risk level prediction accuracy. A test sample set is selected to validate the trained model. When the model's prediction accuracy reaches a preset threshold, the model is deployed to the quality traceability analysis module to achieve real-time risk assessment of the full-dimensional correlated dataset. The full-dimensional correlated dataset at the current collection time is input into the drug quality risk assessment model to obtain the quality risk assessment result at the current collection time. The specific implementation process is not detailed here.

[0040] In step G2, based on the output of the drug quality risk assessment model, a three-level risk warning mechanism is constructed, including: Level I warning (low risk): The model outputs a low risk level, corresponding to scenarios where the degree of temperature and humidity abnormality is slight and has no significant impact on drug quality. The warning method is system log recording and synchronous push to the terminals of the corresponding management personnel; Level II warning (medium risk): The model outputs a medium risk level, corresponding to scenarios where temperature and humidity abnormalities may affect the stability of drug quality. The warning method is system pop-up reminder + SMS notification, and at the same time, a quality re-inspection suggestion is generated, specifying the re-inspection items, re-inspection standards, and re-inspection responsible persons; Level III warning (high risk): The model outputs a high risk level, corresponding to scenarios where temperature and humidity abnormalities may lead to drug quality non-compliance. The warning method is system emergency pop-up + multi-level personnel SMS notification + email notification, and at the same time, the drug control process is triggered, the sales and delivery of this batch of drugs are suspended, and the isolation and sealing procedure is initiated.

[0041] For example, if a batch of injectable drugs experiences a temperature exceeding the standard by 0.5°C for 10 minutes during transportation, the label will be classified as low-risk; if the same batch of injectable drugs experiences a temperature exceeding the standard by 3°C for 2 hours during transportation, the label will be classified as high-risk.

[0042] In step G3, to address the traceability needs of end-product drugs, a reverse traceability algorithm based on batch numbers is designed. The core logic is a reverse derivation mechanism of end-product drug identification - batch number - full life cycle stages - temperature and humidity data. The algorithm implementation process includes: G31: Using the batch number or traceability code of the terminal drug as the traceability entry point, basic information such as the drug batch number, manufacturer, and expiration date can be obtained by scanning the traceability code on the terminal drug packaging; G32: Using the batch number obtained from parsing as the search key, retrieve all life cycle information and temperature and humidity data corresponding to this batch of medicines in the full-dimensional associated dataset; G33: Following the reverse order of terminal sales-transportation-warehousing-production, sort out the flow nodes of this batch of drugs and extract the time, location, responsible person and abnormal temperature and humidity information of each node; G34: During the reverse tracing process, it automatically locates the link, time and specific abnormality of temperature and humidity anomalies, and associates them with the corresponding quality risk level to achieve accurate location of abnormal nodes.

[0043] In step G3, based on the execution result of the reverse traceability algorithm, full lifecycle traceability link information is generated. The link information is organized in a three-layer structure of node-event-data, including a node layer, an event layer, and a data layer. The node layer includes terminal sales nodes, transportation nodes, warehousing nodes, and production nodes. Each node is labeled with a unique node code, location information, and responsible person. The event layer records key events for each node, such as production process start / completion events, warehousing inbound / outbound events, transportation start / arrival events, and terminal sales outbound events. Each event is associated with a corresponding timestamp and operation record. The data layer associates the temperature and humidity monitoring data, equipment operating status data, and quality risk assessment results of each node, achieving precise binding of events and data.

[0044] It should be noted that the traceability information is standardized and packaged to generate a traceability report that complies with the drug traceability management standards. The report supports export in multiple formats such as PDF and Excel, meeting the needs of regulatory authorities for verification, internal enterprise control, and consumer inquiries.

[0045] The visualization module generates dynamic temperature and humidity monitoring curves, quality risk early warning reports, and full life cycle traceability maps based on the accurate temperature and humidity anomaly dataset, quality risk assessment results, and traceability link information. These are then assembled into an interactive traceability monitoring report, facilitating access to the temperature and humidity status of the medicine.

[0046] In practical applications, the entire lifecycle temperature and humidity data of the same batch of drugs is extracted from a precise temperature and humidity anomaly dataset, sorted chronologically, and rendered into a dynamic temperature and humidity monitoring curve. The time interval and risk level of the abnormal data are also extracted to provide a clear visual representation of the abnormal process. Core information for each batch of drugs is extracted from the quality risk assessment results and rendered into a risk warning table. Global risk level distribution data is statistically analyzed and rendered into a risk distribution dashboard. When a new risk warning is generated, the rendering engine automatically triggers report updates, enabling real-time display of warning information. A three-tiered structure of nodes, events, and data is extracted from the traceability chain information and rendered into a directed traceability graph according to the production-warehousing-transportation-sales flow sequence. Nodes are assigned corresponding color labels based on their risk level, enabling rapid identification of risk nodes.

[0047] It should be noted that, for the visualization rendering needs of large amounts of data, data sharding loading technology is adopted to shard the data according to time intervals or batches, so as to achieve fast loading and smooth interaction of visualization components.

[0048] Based on the input drug batch number or traceability code, the system automatically configures the report title and basic drug batch information; it embeds dynamic temperature and humidity monitoring curves, quality risk early warning reports, and full life cycle traceability maps into the visualization component display area of ​​the report template according to a preset layout, supporting custom adjustments to component position and size; it embeds the core information of the full-dimensional related dataset in the data details area, allowing users to quickly jump to the corresponding data details by clicking on anomalies or nodes in the visualization components; and it automatically generates disposal suggestions based on the quality risk assessment results, including risk control measures, re-inspection procedures, and responsibility traceability suggestions.

[0049] The interactive traceability monitoring report supports online viewing, export in multiple formats, and sharing. By scanning the drug traceability code on the terminal, users can quickly access the corresponding traceability monitoring report, enabling one-stop query of drug temperature and humidity status and quality risks.

[0050] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0051] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0052] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more electronic devices to execute all or part of the steps of the methods described in the various embodiments of this application.

[0053] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0054] In the several embodiments provided in this application, it should be understood that the disclosed application can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0055] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0056] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0057] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A full-life-cycle-based drug temperature and humidity digital monitoring and quality tracing system, characterized in that, The system comprises a full-cycle data acquisition module, a temperature and humidity screening module, a data fusion processing module, a quality traceability analysis module, and a visual display module. The full-cycle data acquisition module is configured to acquire temperature and humidity data of each link in the full life cycle of a drug from production, storage, transportation to terminal sales, and perform format verification and standardization packaging on the acquired temperature and humidity data in combination with a temperature and humidity standard library of the drug industry to obtain a full-cycle standardized temperature and humidity data set. The temperature and humidity screening module is configured to perform a first screening on the full-cycle standardized temperature and humidity data set, perform a second screening on the data screened in the first screening in combination with a dynamic temperature and humidity screening model, and obtain a precise temperature and humidity abnormal data set. The data fusion processing module is configured to associate batch information, production information, storage information, transportation information of the drug in the full life cycle with the precise temperature and humidity abnormal data set, construct a temperature and humidity-drug state mapping relationship based on a time sequence association algorithm, and obtain a full-dimension associated data set. The quality traceability analysis module is configured to perform temperature and humidity over-standard risk early warning based on the full-dimension associated data set in combination with a drug quality risk assessment model, and realize traceability tracking from terminal drugs to full-life-cycle temperature and humidity data based on a reverse traceability algorithm to obtain quality risk assessment results and traceability link information. The visual display module is configured to generate a dynamic temperature and humidity monitoring curve, a quality risk early warning report, and a full-life-cycle traceability atlas according to the precise temperature and humidity abnormal data set, the quality risk assessment results, and the traceability link information, and assemble them into an interactive traceability monitoring report to facilitate access to the temperature and humidity state of the drug. 2.The whole life cycle based medicine temperature and humidity digital monitoring and quality tracing system according to claim 1, characterized in that, The process of performing a first screening on the full-cycle standardized temperature and humidity data set comprises: traversing the temperature and humidity data in the full-cycle standardized temperature and humidity data set, comparing the temperature and humidity data with the temperature and humidity threshold range of the corresponding drug in different scenarios in the temperature and humidity standard library of the drug industry, obtaining the temperature and humidity data of the corresponding drug in the full life cycle that exceeds or is lower than the reasonable temperature and humidity threshold range, marking the corresponding collection time of the comparison result, and obtaining the first screening positioning time. 3.The whole life cycle based medicine temperature and humidity digital monitoring and quality tracing system according to claim 2, characterized in that, The process of performing a second screening on the data screened in the first screening in combination with the dynamic temperature and humidity screening model comprises: constructing a dynamic temperature and humidity screening model to obtain the humidity jump threshold in different scenarios of different drugs, obtaining the corresponding compliance abnormal data according to the humidity jump threshold and the first screening positioning time, and obtaining the precise temperature and humidity abnormal data set according to the temperature and humidity data corresponding to the first screening positioning time and the compliance abnormal data.

4. The system according to claim 3, wherein, The process of constructing a dynamic temperature and humidity screening model comprises: acquiring temperature and humidity data of several historical collection cycles, repeating the process of the first screening to obtain the first screening positioning time corresponding to the several historical collection cycles, and then acquiring the temperature and humidity data corresponding to the several adjacent collection times adjacent to the corresponding first screening positioning time; calculating the continuous jump rate of the temperature and humidity data corresponding to the collection time according to the temperature and humidity data of the several adjacent collection times corresponding to one historical collection cycle; The corresponding continuous jump rates of several groups are grouped and labeled as ; is a natural number; the continuous jump rates of the group are taken as sample data, and is a natural number less than , and the sample data mean is obtained using the sample data, which is denoted as a sample set; the continuous jump rates of the remaining groups are taken as a test set; and the training sample set is composed according to the sample set and the test set; The standard screening model is constructed based on a long short-term memory network, and the training sample set is input into the standard screening model to train the standard screening model, and the standard screening model after training is recorded as a dynamic temperature and humidity screening model. 5.The whole life cycle based medicine temperature and humidity digital monitoring and quality tracing system according to claim 4, characterized in that, The process of constructing the time sequence correlation matching algorithm includes: Taking the drug batch number of each data in the accurate temperature and humidity abnormal data set as a primary matching condition, the full life cycle correlation information corresponding to the same batch of drugs is screened out; taking the link code of the corresponding data as a secondary matching condition, the temperature and humidity abnormal data and the correlation information of the corresponding link are preliminarily associated; taking the collection time stamp of the corresponding data as a tertiary matching condition, the time overlap degree of the temperature and humidity abnormal collection time and the correlation information of the corresponding link is calculated, and a time matching threshold is set, when the temperature and humidity abnormal collection time is within the time interval of the correlation information, or the deviation from the key time node of the correlation information is within the threshold range, it is determined that the matching is successful. 6.The whole life cycle based medicine temperature and humidity digital monitoring and quality tracing system according to claim 5, characterized in that, The process of quality traceability analysis on the full-dimensional correlation data set includes: A drug quality risk assessment model is constructed; the temperature and humidity over-standard risk early warning is graded according to the quality risk assessment result of the drug quality risk assessment model; a reverse traceability algorithm is constructed, and then traceability link information is generated. 7.The whole life cycle based medicine temperature and humidity digital monitoring and quality tracing system according to claim 6, characterized in that, The process of constructing the reverse traceability algorithm includes: Taking the batch number or traceability code of the terminal drug as the traceability entrance, the batch number of the drug is obtained by scanning the traceability code on the packaging of the terminal drug; taking the batch number obtained by analysis as a search key, searching all life cycle link information and temperature and humidity data corresponding to the batch of drugs in the full-dimensional correlation data set; according to the reverse order of terminal sales-transportation-warehousing-production, the flow nodes of the batch of drugs are combed, and the information corresponding to each node is extracted. 8.The system according to claim 7, wherein, The process of generating a dynamic temperature and humidity monitoring curve, a quality risk early warning report, and a full life cycle traceability map according to the accurate temperature and humidity abnormal data set, the quality risk assessment result, and the traceability link information includes: The full life cycle temperature and humidity data of the same batch of drugs is extracted from the accurate temperature and humidity abnormal data set, sorted in chronological order, and rendered to generate a dynamic temperature and humidity monitoring curve; the core information of each batch of drugs is extracted from the quality risk assessment result, and a risk early warning table is rendered; the global risk level distribution data is counted, and a risk distribution dashboard is rendered; when a new risk early warning occurs, the report is automatically updated by the rendering engine, and the real-time display of the early warning information is realized; the node, event, and data three-layer structure data are extracted from the traceability link information, and a directed traceability map is rendered according to the flow order of production-warehousing-transportation-sales.

Images