A system and method for checking leakage of products and quantities in pharmaceutical channels based on statistical analysis

Abstract

The present invention relates to the field of pharmaceutical supply chain management and intelligent data analysis technology, and in particular to a system and method for checking leakage of pharmaceutical products and quantities in pharmaceutical channels based on statistical analysis. A complete pharmaceutical supply chain management system is constructed through a multimodal sensor module, a data lake module, an intelligent analysis and verification module, a prediction and optimization module, a feedback and decision support module, and a visualization and monitoring module. The system collects and processes pharmaceutical circulation data in real time and in a standardized manner, identifies anomalies using intelligent analysis and causal reasoning, and formulates optimal management strategies through multi-objective optimization. The application of feedback mechanism and machine learning enables the system to dynamically adjust management strategies and improve the responsiveness and transparency of the supply chain. Finally, the system realizes efficient and accurate management of the entire process of pharmaceuticals from production to terminal stores, providing reliable guarantee for the safety and efficiency of pharmaceutical channels.

Description

Technical Field

[0001] The present invention relates to the technical field of pharmaceutical supply chain management and intelligent data analysis, and in particular to a system and method for checking leakage of pharmaceutical products and quantities within pharmaceutical channels based on statistical analysis. Background Art

[0002] In the pharmaceutical industry, drug leakage and quantity discrepancies not only lead to economic losses, but also threaten drug supply security and patient medication safety. In order to meet these challenges, it is urgent to develop a drug leakage and quantity verification system within the pharmaceutical channel to ensure the accuracy and safety of drugs from production to the terminal store. Methods based on statistical analysis can provide scientific data support and effective strategy adjustments for this verification.

[0003] The existing technology (Chinese invention patent, publication number: CN116308080A, name: A method, system, terminal device and storage medium for drug purchase and sales supervision) mainly relies on the analysis of sales inventory record data, and identifies abnormal sales behavior by judging the compliance of inventory data with preset standards. Although this method can detect anomalies to a certain extent, it relies on fixed preset standards and cannot dynamically adapt to changing market environments and complex supply chain conditions. In addition, the existing technology mainly focuses on data recording and simple logical judgment, lacks in-depth analysis and intelligent prediction of anomalies, resulting in deficiencies in real-time and flexibility. Summary of the invention

[0004] In view of the many problems existing in the above-mentioned prior art, the present invention provides a system and method for checking the leakage amount of leaked products in pharmaceutical channels based on statistical analysis. The present invention collects the environment, location and transportation data in the pharmaceutical supply chain in real time through a multimodal sensor module, and transmits the data to a data lake module for standardization and semantic integration to form integrated semantic data that can be used for in-depth analysis; then, the intelligent analysis and verification module uses association rule mining and adaptive statistical reasoning models to analyze the data to identify anomalies and evaluate the trust of supply chain nodes; then, the prediction and optimization module uses a genetic algorithm to achieve multi-objective optimization based on the generated causal data, so as to formulate the best supply chain management strategy; finally, the feedback and decision support module adjusts the optimization strategy through machine learning, and uses the visualization and monitoring module to achieve real-time monitoring and adjustment of the supply chain; through this full-chain management and analysis, the present invention not only improves the efficiency and reliability of the pharmaceutical supply chain, but also enhances the response capability to abnormal events and the scientific nature of decision support.

[0005] A system for checking leakage of products and quantities in pharmaceutical channels based on statistical analysis, comprising:

[0006] Multimodal sensor module, which is used to collect multi-dimensional data of drugs at various supply chain links, including environmental status, location data and transportation conditions, and generate multimodal sensor data;

[0007] A data lake module, communicating with the multimodal sensor module, receiving multimodal sensor data, performing data standardization and semantic level integration, and generating integrated semantic data;

[0008] Intelligent analysis and verification module, including:

[0009] Heterogeneous data interaction unit, used to perform data interaction analysis on integrated semantic data and generate interaction analysis result data;

[0010] An adaptive statistical reasoning unit, used to analyze the interactive analysis result data, identify anomalies and generate statistical reasoning data and node trust data;

[0011] Prediction and optimization modules, including:

[0012] A causal inference unit, used for performing nonlinear causal analysis based on statistical inference data to generate causal relationship data;

[0013] An optimization analysis unit, used to perform multi-objective optimization based on causal relationship data and generate optimization strategy data;

[0014] Feedback and decision support module, which is used to receive optimization strategy data, implement feedback iteration mechanism to generate decision support data, and support real-time supply chain management adjustments;

[0015] The visualization and monitoring module is used to display decision support data and abnormal alarm data to achieve real-time monitoring and transparent management of the supply chain.

[0016] Preferably, the multimodal sensor module comprises:

[0017] Environmental sensors are used to collect environmental status data of drugs, including temperature and humidity;

[0018] Location sensor, used to obtain geographical location information of drugs;

[0019] Transport status sensors are used to monitor dynamic conditions of drugs during transportation, including vibration and acceleration.

[0020] Preferably, the data lake module further includes a data cleaning unit for performing noise filtering and format conversion on the multimodal sensor data to generate standardized data input into the semantic level integration process.

[0021] Preferably, the heterogeneous data interaction unit of the intelligent analysis and verification module analyzes the integrated semantic data through an association rule mining algorithm, identifies potential correlations between data, and generates interactive analysis result data.

[0022] Preferably, the adaptive statistical reasoning unit uses a Bayesian reasoning model to process the interactive analysis result data to generate statistical reasoning data, wherein the node trust data is calculated by the following formula:

[0023]

[0024] Among them, Trust i is the trust value of node i; P(E i |H) is the event E observed when hypothesis H holds i The probability of H; P(H) is the prior probability of hypothesis H; P(E i ) is event E i The marginal probability of .

[0025] Preferably, the causal reasoning unit of the prediction and optimization module analyzes statistical reasoning data based on a graphical model to generate causal relationship data for identifying the causal relationship between various factors in the supply chain.

[0026] Preferably, the optimization analysis unit implements multi-objective optimization through a genetic algorithm to generate optimization strategy data to maximize supply chain efficiency while minimizing costs and risks. The optimization objective function can be expressed as:

[0027]

[0028] Among them, Optimize is the optimization target; w i is the weight; f i (x) is the objective function value.

[0029] Preferably, the feedback and decision support module adjusts the optimization strategy data through a machine learning algorithm to generate updated decision support data to support real-time adjustment of supply chain management.

[0030] Preferably, the visualization and monitoring module uses a Web-based interactive interface to provide users with visualization tools for viewing decision support data, abnormal alarm data and supply chain dynamics, so as to achieve real-time response and processing of abnormal situations.

[0031] A method for checking the amount of leaked products in a pharmaceutical channel based on statistical analysis, using the system, comprises the following steps:

[0032] At each supply chain link, multi-dimensional data of drugs, including environmental status, location data and transportation conditions, is collected through multi-modal sensor modules to generate multi-modal sensor data;

[0033] Transmit multimodal sensor data to the data lake module, perform data standardization and semantic level integration, and generate integrated semantic data;

[0034] In the heterogeneous data interaction unit, the integrated semantic data is analyzed for data interaction to generate interaction analysis result data; in the adaptive statistical reasoning unit, the interaction analysis result data is analyzed to identify abnormal situations and generate statistical reasoning data and node trust data;

[0035] In the causal reasoning unit, nonlinear causal analysis is performed based on statistical reasoning data to generate causal relationship data; in the optimization analysis unit, multi-objective optimization analysis is performed based on the causal relationship data to generate optimization strategy data;

[0036] Receive optimization strategy data, implement feedback iteration mechanism, generate decision support data, and support real-time supply chain management adjustments;

[0037] In the visualization and monitoring module, decision support data and abnormal alarm data are displayed to achieve real-time monitoring and transparent management of the supply chain.

[0038] Compared with the prior art, the advantages and beneficial effects of the present invention are:

[0039] The present invention realizes the real-time collection and monitoring of multi-dimensional data in the pharmaceutical supply chain through multi-modal sensor technology, thereby improving the accuracy of anomaly detection;

[0040] The present invention uses data lake technology to achieve standardized data processing and semantic level integration, thereby improving the depth and consistency of data analysis;

[0041] The present invention realizes abnormal analysis and node trust evaluation based on association rules through intelligent analysis and verification technology, and enhances the ability to identify supply chain risks;

[0042] The present invention realizes multi-objective optimization based on causality through prediction and optimization technology, and optimizes the supply chain management strategy;

[0043] In the feedback and decision support module, the present invention realizes dynamic adjustment of strategies through machine learning algorithms, thereby improving the flexibility and response speed of supply chain management;

[0044] The present invention realizes real-time monitoring and transparent management of the supply chain through visualization and monitoring technical means, and enhances the timeliness and effectiveness of abnormal response. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a structural block diagram of the system of the present invention;

[0046] Figure 2 Schematic diagram of a multimodal sensor module in the present invention;

[0047] Figure 3 It is a schematic diagram of data flow in the present invention;

[0048] Figure 4 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0049] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0050] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0051] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0052] like Figure 1 As shown, a system for checking leakage of products and quantities in pharmaceutical channels based on statistical analysis includes:

[0053] like Figure 2-Figure 3 As shown, the multimodal sensor module is used to collect multi-dimensional data of drugs at various supply chain links, including environmental status, location data and transportation conditions, and generate multimodal sensor data;

[0054] Preferably, the multimodal sensor module comprises:

[0055] Environmental sensors are used to collect environmental status data of drugs, including temperature and humidity;

[0056] Location sensor, used to obtain geographical location information of drugs;

[0057] Transport status sensors are used to monitor dynamic conditions of drugs during transportation, including vibration and acceleration.

[0058] In the present invention, the multimodal sensor module is a key component, which is responsible for collecting multi-dimensional data of drugs in various supply chain links, including environmental status, location data and transportation conditions, and generating multimodal sensor data. The module is composed of a variety of sensors, including environmental sensors, location sensors and transportation status sensors, which work together to provide comprehensive status information during the circulation of drugs. Environmental sensors are responsible for measuring the temperature and humidity of the environment where the drugs are located, ensuring that the drugs are stored and transported within the specified temperature and humidity range to maintain their safety and effectiveness. Position sensors usually use GPS technology to obtain real-time geographic location information of drugs. By tracking the circulation path of drugs in real time, managers can ensure that the drugs are transported according to the planned route and can quickly locate the location of drugs when abnormal situations occur. The transportation status sensor is composed of an accelerometer and a gyroscope, which is responsible for monitoring the movement state of drugs during transportation, including the acceleration, tilt and rotation of the vehicle. These data are used to evaluate the impact on drugs during transportation and prevent damage to drugs caused by severe vibration.

[0059] The multimodal sensor module provides real-time environmental and dynamic information to achieve comprehensive monitoring and analysis of the status of drugs in the supply chain, and can detect abnormal conditions early, such as excessive vibration or abnormal temperature during transportation, which are important factors that lead to the decline of drug quality or even failure. Through real-time monitoring of the environment and location during the circulation of drugs, the module ensures that drugs are stored and transported under appropriate conditions, maintains the safety and effectiveness of drugs, and improves the management efficiency and transparency of the entire supply chain. For example, during the transportation of vaccines, since vaccines are extremely sensitive to temperature changes, temperature sensors are used to monitor the temperature of the environment in which vaccines are located during cold chain transportation in real time to ensure that vaccines are always within a suitable temperature range and avoid vaccine failure caused by temperature out of control. In addition, the position sensor tracks the transportation route of vaccines in real time to ensure the safety and transparency of the transportation process, while the transportation status sensor helps identify potential transportation risks and prevent vaccine damage caused by bumps by monitoring vibration during transportation. The comprehensive application of this multimodal sensor module can not only ensure the quality and safety of vaccines, but also improve the management efficiency of vaccines during circulation. Therefore, the multimodal sensor module significantly improves the overall efficiency of the leakage verification system within the pharmaceutical channel by real-time collection, analysis and feedback of multi-dimensional data of drugs, ensuring the effective management and safe use of drugs.

[0060] A data lake module, communicating with the multimodal sensor module, receiving multimodal sensor data, performing data standardization and semantic level integration, and generating integrated semantic data;

[0061] Preferably, the data lake module further includes a data cleaning unit for performing noise filtering and format conversion on the multimodal sensor data to generate standardized data input into the semantic level integration process.

[0062] In the present invention, the data lake module is a core component, which is responsible for receiving multi-dimensional sensor data from the multimodal sensor module, performing data standardization and semantic level integration, and generating integrated semantic data. This module communicates with the multimodal sensor module to ensure the real-time and integrity of the data. The primary task of the data lake module is to handle data cleaning. The data cleaning unit performs noise filtering and format conversion on the received multimodal sensor data, thereby eliminating errors and inconsistencies in the data. Through this processing, standardized data is generated so that it can be further analyzed and used. The core of this step is to ensure that all data is input into the next processing stage in a consistent format and high quality.

[0063] After data cleaning, the data lake module performs semantic level integration. This process uses semantic analysis technology to semantically match and integrate data from different sources to form a higher-level data view, called integrated semantic data. This integration is not just a simple data summary, but through the association of semantic levels, it identifies and connects the correlation between data, providing a global perspective to analyze the circulation status of drugs in the supply chain. The operation of this module greatly improves the accuracy and depth of data analysis, allowing subsequent intelligent analysis and verification steps to be processed and decided based on more comprehensive and accurate data.

[0064] For example, in actual applications, after the data lake module receives data from temperature sensors, GPS location sensors, and transportation status sensors, it first performs data cleaning to filter out noise data caused by sensor failure or signal interference, and unifies the output formats of different sensors into a standardized format. Then, through semantic hierarchical integration, the data lake module can associate the temperature, location, and transportation status data of a batch of drugs at different time points to form a complete circulation path view. Through this view, managers can easily identify whether the drugs have experienced non-compliant temperature changes or path deviations during transportation. In this case, the drugs may be affected, and the system can automatically generate alarms to notify relevant personnel for processing. Through standardization and semantic integration, the data lake module significantly improves the system's data processing capabilities and analysis depth, providing a solid data foundation for the entire leaked product and leaked quantity verification system.

[0065] Intelligent analysis and verification module, including:

[0066] Heterogeneous data interaction unit, used to perform data interaction analysis on integrated semantic data and generate interaction analysis result data;

[0067] Preferably, the heterogeneous data interaction unit of the intelligent analysis and verification module analyzes the integrated semantic data through an association rule mining algorithm, identifies potential correlations between data, and generates interactive analysis result data.

[0068] In the present invention, the intelligent analysis and verification module plays a key role in data analysis and anomaly detection, of which the heterogeneous data interaction unit is its core component. This unit is responsible for conducting in-depth data interaction analysis on the integrated semantic data from the data lake module and generating interaction analysis result data. Through this analysis, the system can identify the potential correlation between data, thereby helping to discover anomalies in the drug circulation process. To achieve this goal, the heterogeneous data interaction unit adopts an association rule mining algorithm, which is a technology commonly used in data mining and can mine meaningful association rules from large data sets. The algorithm can identify potential correlations and patterns between different data sets, providing a basis for subsequent anomaly detection and decision-making.

[0069] In the present invention, the heterogeneous data interaction unit can identify factors that may lead to missing products or inconsistent quantities during the circulation of drugs by analyzing and integrating semantic data. For example, during the transportation of drugs, the heterogeneous data interaction unit can identify the correlation between certain specific transportation paths or states and the high incidence of drug damage by analyzing location data and transportation status data. This mining of association rules enables the system to identify high-risk transportation links in advance, so that additional preventive measures can be taken in these links to reduce the risk of drug damage.

[0070] In the embodiment, suppose the system finds that a certain type of medicine is often damaged or the quantity is inconsistent when it is transported to a specific area. Through the analysis of the heterogeneous data interaction unit, the system recognizes that these events are significantly associated with specific areas of the transportation route and high temperature weather conditions. Therefore, the system can recommend adjusting the transportation time or taking additional refrigeration measures to reduce the risk of damage. This data-based intelligent analysis significantly improves the efficiency and accuracy of supply chain management, helping managers to proactively respond to potential problems and optimize logistics processes.

[0071] In general, the intelligent analysis and verification module uses advanced data mining technology to conduct in-depth analysis of integrated semantic data, identify patterns and associations hidden behind complex data, and provide strong technical support for the effective verification of leaked products and quantities in the drug channel. This data-driven approach not only improves the accuracy of anomaly detection, but also provides valuable insights for supply chain management decisions.

[0072] An adaptive statistical reasoning unit, used to analyze the interactive analysis result data, identify anomalies and generate statistical reasoning data and node trust data;

[0073] Preferably, the adaptive statistical reasoning unit uses a Bayesian reasoning model to process the interactive analysis result data to generate statistical reasoning data, wherein the node trust data is calculated by the following formula:

[0074]

[0075] Among them, Trust i is the trust value of node i; P(E i |H) is the event E observed when hypothesis H holds i The probability of H; P(H) is the prior probability of hypothesis H; P(E i ) is event E i The marginal probability of .

[0076] In the present invention, the adaptive statistical reasoning unit is a key component in the intelligent analysis and verification module, and its main function is to conduct in-depth analysis of the interactive analysis result data to identify anomalies in the supply chain and generate statistical reasoning data and node trust data. This unit uses the Bayesian reasoning model to process data, providing a flexible and accurate method for anomaly detection and trust assessment. The Bayesian reasoning model is a statistical method that dynamically updates the assessment of the probability of an event by combining prior knowledge and current data observations. This method is particularly suitable for dynamic and complex environments such as supply chains because it can adjust the trust in nodes and the accuracy of anomaly identification in real time based on new data.

[0077] In the application, the adaptive statistical reasoning unit uses the interactive analysis result data to evaluate whether the behavior of each supply chain node conforms to the normal mode through the Bayesian reasoning model. Specifically, the unit evaluates the reliability of the node by calculating the trust value of each node. The above trust value formula reflects the process of Bayesian reasoning to evaluate the trust of node behavior after combining prior knowledge with actual observation data.

[0078] The adaptive statistical reasoning unit can dynamically adjust the trust value of each node, allowing the system to quickly identify potential high-risk nodes and abnormal events, improving the sensitivity and accuracy of abnormality detection. For example, during the drug distribution process, if the drug damage rate of a distribution center suddenly increases, through the analysis of the adaptive statistical reasoning unit, the system can quickly identify the center as a high-risk node and reduce its trust value to the warning level. This warning not only helps managers take timely measures to solve the problem, but also optimizes the overall efficiency and security of the supply chain.

[0079] For example, in a pharmaceutical supply chain, the adaptive statistical reasoning unit identifies abnormal vibration patterns in certain specific vehicles during transportation by analyzing the interactive analysis result data of the transportation status sensor. Through the Bayesian reasoning model, the system dynamically adjusts the trust value of these vehicles, lowers the trust level of these vehicles, and prompts managers to inspect and maintain these vehicles. In this way, by updating the trust of the nodes in real time, the adaptive statistical reasoning unit not only improves the system's anomaly detection capabilities, but also effectively reduces the potential risks of leakage and damage, and improves the safety and reliability of the drug circulation process. Through the application of the adaptive statistical reasoning unit, the system can accurately identify anomalies and optimize management decisions in a complex and changing supply chain environment, providing a solid guarantee for the safety of drug channels.

[0080] Prediction and optimization modules, including:

[0081] A causal inference unit, used for performing nonlinear causal analysis based on statistical inference data to generate causal relationship data;

[0082] Preferably, the causal reasoning unit of the prediction and optimization module analyzes statistical reasoning data based on a graphical model to generate causal relationship data for identifying the causal relationship between various factors in the supply chain.

[0083] In the present invention, the prediction and optimization module is responsible for identifying and analyzing the causal relationship between various factors in the supply chain in order to optimize management strategies and decisions. The causal reasoning unit in this module uses nonlinear causal analysis technology to process statistical reasoning data and generate causal relationship data from it. This process is based on the analysis method of the graphical model, which can intuitively show the mutual influence relationship between various links in the supply chain, helping managers to better understand and predict potential risks and problems in the drug circulation process.

[0084] The basic principle of causal reasoning is to construct a causal graph model, taking each factor in the supply chain as a node and connecting them through edges to represent the causal relationship between them. This graph model can not only reveal the direct causal relationship between factors, but also help identify indirect effects and potential chain reactions. During the analysis process, the causal reasoning unit uses an algorithm to reason about the graph model to determine which factors are the root cause of missing products or inconsistent quantities. In this way, the system can provide more targeted suggestions based on statistical reasoning data to support optimization decisions in supply chain management.

[0085] In practical applications, for example, when analyzing the damage of a drug during transportation, the causal reasoning unit can identify the causal relationship between transportation routes, environmental conditions, and transportation methods by analyzing various types of sensor data. If a high damage rate is found on a specific transportation route, the causal reasoning unit may find that this is due to the combined effect of high temperature and frequent bumps. By generating causal relationship data, the system can recommend changing the transportation route or adding refrigeration measures to effectively reduce the risk of damage. This causal analysis-based recommendation is not only based on data analysis results, but also combined with a deep understanding of the complex dynamics of the supply chain, thereby improving the scientific nature and effectiveness of decision-making.

[0086] A significant effect of the causal reasoning unit is that it can simplify complex supply chain information into easy-to-understand causal relationship maps, providing managers with clear decision-making basis. In addition, by identifying the key factors that affect the safety and efficiency of drug circulation, the system can take measures in advance to prevent potential problems and optimize the operational efficiency of the entire supply chain. Ultimately, the causal reasoning unit not only improves the system's predictive ability by deeply exploring the causal relationship between data, but also provides innovative solutions for supply chain optimization. In drug channel management, this unit can significantly reduce the occurrence of missed products and damage incidents, ensuring the quality of drugs and the reliability of supply.

[0087] An optimization analysis unit, used to perform multi-objective optimization based on causal relationship data and generate optimization strategy data;

[0088] Preferably, the optimization analysis unit implements multi-objective optimization through a genetic algorithm to generate optimization strategy data to maximize supply chain efficiency while minimizing costs and risks. The optimization objective function can be expressed as:

[0089]

[0090] Among them, Optimize is the optimization target; w i is the weight; f i (x) is the objective function value.

[0091] In the present invention, the role of the optimization analysis unit is to use the causal relationship data to perform multi-objective optimization to generate optimization strategy data. Its main goal is to develop a set of strategies that can maximize supply chain efficiency while minimizing costs and risks by analyzing the causal relationship of various factors in the supply chain. The optimization analysis unit achieves multi-objective optimization through genetic algorithms, which is a flexible and powerful optimization tool that can effectively solve multi-objective optimization problems in complex systems.

[0092] Genetic algorithm is a heuristic search algorithm based on natural selection and genetic mechanism, which is suitable for solving complex optimization problems. In this system, the genetic algorithm continuously evolves the population of candidate solutions to find the strategy that can best meet multiple optimization objectives. Specifically, based on the above optimization objective function, the system can find the optimal balance between multiple objectives (such as cost, risk and efficiency).

[0093] In an embodiment, after the optimization analysis unit receives the causal relationship data generated by the causal reasoning unit, it first determines the key indicators that need to be optimized in supply chain management, such as transportation costs, delivery time, and quality risks. Then, these indicators are optimized by multi-objective optimization through genetic algorithms to explore the best strategy combination under different weight configurations. For example, if the system identifies that certain transportation routes have lower costs but higher risks, through multi-objective optimization, the system may recommend adjusting the route to reduce the risk, even if it means a slight increase in cost. Such an optimization strategy not only takes into account the current data and causal relationships, but also combines the pursuit of global efficiency of the supply chain.

[0094] Through this optimization analysis, the system can develop a drug circulation strategy that is both economical and safe and reliable in a complex supply chain environment. This process not only helps companies strike a balance between operating costs and risk control, but also improves the response speed and service level of the overall supply chain. Ultimately, the optimization analysis unit provides scientific decision-making support for managers of pharmaceutical channels through a data-driven multi-objective optimization method, ensuring the timely supply and quality assurance of drugs, and gaining a significant advantage in a highly competitive market environment.

[0095] Feedback and decision support module, which is used to receive optimization strategy data, implement feedback iteration mechanism to generate decision support data, and support real-time supply chain management adjustments;

[0096] Preferably, the feedback and decision support module adjusts the optimization strategy data through a machine learning algorithm to generate updated decision support data to support real-time adjustment of supply chain management.

[0097] In the present invention, the feedback and decision support module is an important part of the system to achieve dynamic adjustment and optimization. The main function of this module is to receive the optimization strategy data from the optimization analysis unit and generate new decision support data by implementing the feedback iteration mechanism to support real-time supply chain management adjustment. Its core lies in using feedback loops and machine learning algorithms to dynamically adjust the optimization strategy data to generate updated decision support data.

[0098] The Feedback and Decision Support Module dynamically adjusts the optimization strategy through machine learning algorithms, which can effectively handle the ever-changing environment and conditions in supply chain management. Machine learning algorithms allow the system to continuously update and improve the optimization strategy based on historical data and real-time feedback, enabling the system to quickly adapt to new situations and challenges. This approach not only improves the accuracy and efficiency of decision-making, but also enhances the system's adaptive capabilities, making supply chain management more flexible and efficient.

[0099] In practice, the feedback and decision support module is able to monitor various activities in the supply chain and adjust strategies based on real-time feedback. For example, in a pharmaceutical supply chain, if the demand for a certain drug suddenly increases, the system can identify this trend through machine learning algorithms and adjust the supply chain strategy to ensure timely replenishment and avoid shortages. This capability is particularly important because the demand for drugs may be affected by a variety of factors, including seasonal changes, epidemic outbreaks, and market competition. By adjusting the supply chain strategy in a timely manner, the system can effectively avoid market losses caused by inaccurate forecasts or untimely responses.

[0100] In addition, the module also supports global optimization of the supply chain. By integrating various types of feedback data, the system can identify weak links in the supply chain and make suggestions for improvement. For example, the system may suggest optimizing transportation routes, adjusting inventory management strategies, or improving supplier selection to improve overall efficiency and service levels. Through this feedback-driven optimization method, the system can not only improve the operating efficiency of the supply chain, but also maintain a competitive advantage in a complex market environment.

[0101] In general, the feedback and decision support module provides intelligent decision support for managers of pharmaceutical channels through real-time monitoring and dynamic adjustment, ensuring efficient operation and flexible response of the supply chain. By combining machine learning and feedback iteration mechanisms, the module can continuously optimize supply chain management strategies and provide reliable guarantees for the safe, effective and timely supply of drugs.

[0102] The visualization and monitoring module is used to display decision support data and abnormal alarm data to achieve real-time monitoring and transparent management of the supply chain.

[0103] Preferably, the visualization and monitoring module uses a Web-based interactive interface to provide users with visualization tools for viewing decision support data, abnormal alarm data and supply chain dynamics, so as to achieve real-time response and processing of abnormal situations.

[0104] In the present invention, the visualization and monitoring module is an important component for achieving transparent and efficient monitoring of supply chain management. This module is responsible for displaying decision support data and abnormal alarm data, and providing real-time monitoring and management capabilities for the supply chain. Through the Web-based interactive interface, users can easily view the dynamic operation status of the supply chain, identify abnormal situations, and respond and handle them quickly.

[0105] The principle of the visualization and monitoring module is based on information visualization technology, which converts complex data into easy-to-understand graphical representations. In this way, users can intuitively observe the operation of the supply chain, including key indicators such as inventory levels, transportation status, and abnormal alerts. The interactive interface allows users to browse data in different ways, such as filtering, sorting, and zooming to view specific details. This interface design not only improves the accessibility of information, but also enables users to customize views according to actual needs, so as to more effectively monitor and manage the supply chain.

[0106] In practice, this module can help managers quickly identify and handle abnormal situations. For example, if a batch of medicines has an abnormal temperature during transportation, the system will immediately issue an alarm through a visual interface and provide detailed location information and temperature change charts to help managers take quick action. Users can view the specific transportation route, current environmental conditions, and possible solutions through the interface to ensure that the problem can be solved in a timely manner. Through this real-time monitoring and response mechanism, the system can effectively reduce the risks and losses caused by abnormal situations.

[0107] In addition, the visualization and monitoring module also supports supply chain performance analysis and optimization. By analyzing historical data, users can identify bottlenecks and improvement opportunities in the supply chain. For example, the system may show that the delivery time of certain transportation routes is longer than other routes, prompting managers to re-evaluate and optimize logistics strategies. This data-driven optimization capability can not only improve the efficiency of the supply chain, but also provide a solid foundation for corporate decision-making.

[0108] In short, the visualization and monitoring module significantly enhances the management capabilities of the leakage verification system in the pharmaceutical channel by providing comprehensive real-time monitoring and information visualization tools. Through an intuitive interface and powerful data analysis functions, the module provides supply chain managers with instant decision support to ensure the safe, effective and timely delivery of drugs. This transparent management not only improves the efficiency and reliability of the supply chain, but also enhances the competitiveness of enterprises in the market.

[0109] like Figure 4 As shown, a method for checking the amount of leakage of pharmaceutical products in a pharmaceutical channel based on statistical analysis, using the system, includes the following steps:

[0110] In each supply chain link, the multi-dimensional data of drugs, including environmental status, location data and transportation conditions, are collected through multi-modal sensor modules to generate multi-modal sensor data; In each supply chain link, the system collects multi-dimensional data of drugs through multi-modal sensor modules, including environmental status, location data and transportation conditions, to generate multi-modal sensor data. The multi-modal sensor module uses multiple sensors to work together to monitor the storage and transportation conditions of drugs in real time to ensure the accuracy and comprehensiveness of the data. Through these sensors, the system can obtain detailed status information of drugs during the circulation process, providing basic data for subsequent analysis and decision-making.

[0111] The multimodal sensor data is transferred to the data lake module, and the data is standardized and integrated at the semantic level to generate integrated semantic data; once the data collection is completed, the multimodal sensor data will be transferred to the data lake module. In the data lake module, the data is standardized and integrated at the semantic level to generate integrated semantic data. The main function of the data lake module is to clean, format and integrate data to eliminate noise and inconsistencies in the data, thereby forming a unified data format. Through semantic level integration, the system can identify and associate the relationships between different data sources, providing rich semantic information for intelligent analysis. This process makes subsequent data analysis more efficient and accurate.

[0112] In the heterogeneous data interaction unit, the integrated semantic data is interactively analyzed to generate interactive analysis result data; in the adaptive statistical reasoning unit, the interactive analysis result data is analyzed to identify abnormal situations and generate statistical reasoning data and node trust data; the integrated semantic data enters the intelligent analysis and verification module, in which the heterogeneous data interaction unit interactively analyzes the integrated semantic data to generate interactive analysis result data. Through association rule mining and data fusion, the system can identify potential correlations between data and mine hidden patterns and trends. Next, the adaptive statistical reasoning unit further analyzes the interactive analysis result data, identifies abnormal situations and generates statistical reasoning data and node trust data. This module uses methods such as the Bayesian reasoning model to perform probability analysis and trust evaluation on data, which can effectively identify abnormal nodes and potential risks, and provide reliable risk assessment and trust management for the system.

[0113] In the causal reasoning unit, nonlinear causal analysis is performed based on statistical reasoning data to generate causal relationship data; in the optimization analysis unit, multi-objective optimization analysis is performed based on causal relationship data to generate optimization strategy data; in the causal reasoning unit, the system performs nonlinear causal analysis based on statistical reasoning data to generate causal relationship data. The causal reasoning unit reveals the causal relationship between various factors in the supply chain by constructing a causal graph model to help managers understand the complex causal network. On this basis, the optimization analysis unit implements multi-objective optimization analysis through genetic algorithms to generate optimization strategy data. The goal of multi-objective optimization is to balance cost, risk and efficiency and formulate the best supply chain management strategy. This process seeks the optimal solution for supply chain management by adjusting the weights of different factors.

[0114] Receive optimization strategy data, implement feedback iteration mechanism, generate decision support data, and support real-time supply chain management adjustments; the generated optimization strategy data will be passed to the feedback and decision support module, which implements feedback iteration mechanism, generates decision support data, and supports real-time supply chain management adjustments. Through machine learning and feedback loops, the system can dynamically adjust strategies based on real-time data to enhance the flexibility and responsiveness of the supply chain.

[0115] In the visualization and monitoring module, decision support data and abnormal alarm data are displayed to achieve real-time monitoring and transparent management of the supply chain. In the visualization and monitoring module, the system displays decision support data and abnormal alarm data to achieve real-time monitoring and transparent management of the supply chain. The Web-based interactive interface allows users to intuitively view and analyze supply chain dynamics and respond quickly to abnormal situations. Through this full-process management and monitoring, the present invention achieves efficient and accurate management of the entire process of drugs from production to terminal stores, providing reliable protection for the safety and efficiency of pharmaceutical channels.

[0116] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware.

[0117] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A system for checking the leakage of products and quantities in pharmaceutical channels based on statistical analysis, characterized in that: include: Multimodal sensor module, which is used to collect multi-dimensional data of drugs at various supply chain links, including environmental status, location data and transportation conditions, and generate multimodal sensor data; A data lake module, communicating with the multimodal sensor module, receiving multimodal sensor data, performing data standardization and semantic level integration, and generating integrated semantic data; Intelligent analysis and verification module, including: Heterogeneous data interaction unit, used to perform data interaction analysis on integrated semantic data and generate interaction analysis result data; An adaptive statistical reasoning unit, used to analyze the interactive analysis result data, identify anomalies and generate statistical reasoning data and node trust data; Prediction and optimization modules, including: A causal inference unit, used for performing nonlinear causal analysis based on statistical inference data to generate causal relationship data; An optimization analysis unit, used to perform multi-objective optimization based on causal relationship data and generate optimization strategy data; Feedback and decision support module, which is used to receive optimization strategy data, implement feedback iteration mechanism to generate decision support data, and support real-time supply chain management adjustments; The visualization and monitoring module is used to display decision support data and abnormal alarm data to achieve real-time monitoring and transparent management of the supply chain.

2. The system for checking leakage of pharmaceutical products and quantities based on statistical analysis according to claim 1 is characterized in that: The multimodal sensor module comprises: Environmental sensors are used to collect environmental status data of drugs, including temperature and humidity; Location sensor, used to obtain geographical location information of drugs; Transport status sensors are used to monitor dynamic conditions of drugs during transportation, including vibration and acceleration.

3. The system according to claim 1, characterized in that The data lake module further includes a data cleaning unit for performing noise filtering and format conversion on the multimodal sensor data to generate standardized data input into the semantic level integration process.

4. The system according to claim 1, characterized in that The heterogeneous data interaction unit of the intelligent analysis and verification module analyzes the integrated semantic data through an association rule mining algorithm, identifies potential correlations between data, and generates interactive analysis result data.

5. The system according to claim 1, characterized in that The adaptive statistical reasoning unit uses the Bayesian reasoning model to process the interactive analysis result data to generate statistical reasoning data, wherein the node trust data is calculated by the following formula: Among them, Trust i is the trust value of node i; P(E i |H) is the event E observed when hypothesis H holds i The probability of H; P(H) is the prior probability of hypothesis H; P(E i ) is event E i The marginal probability of .

6. The system according to claim 1, characterized in that The causal reasoning unit of the prediction and optimization module analyzes statistical reasoning data based on a graphical model to generate causal relationship data for identifying the causal relationship between various factors in the supply chain.

7. The system according to claim 1, characterized in that The optimization analysis unit implements multi-objective optimization through genetic algorithms and generates optimization strategy data to maximize supply chain efficiency while minimizing costs and risks. The optimization objective function can be expressed as: Among them, Optimize is the optimization target; w i is the weight; f i (x) is the objective function value.

8. The system according to claim 1, characterized in that The feedback and decision support module adjusts the optimization strategy data through a machine learning algorithm to generate updated decision support data to support real-time adjustment of supply chain management.

9. The system according to claim 1, characterized in that The visualization and monitoring module uses a web-based interactive interface to provide users with visualization tools for viewing decision support data, abnormal alarm data and supply chain dynamics, so as to achieve real-time response and processing of abnormal situations.

10. A method for checking the amount of leakage of pharmaceutical products in a pharmaceutical channel based on statistical analysis, characterized in that: The system according to any one of claims 1 to 9 comprises the following steps: At each supply chain link, multi-dimensional data of drugs, including environmental status, location data and transportation conditions, is collected through multi-modal sensor modules to generate multi-modal sensor data; Transmit multimodal sensor data to the data lake module, perform data standardization and semantic level integration, and generate integrated semantic data; In the heterogeneous data interaction unit, the integrated semantic data is analyzed for data interaction to generate interaction analysis result data; in the adaptive statistical reasoning unit, the interaction analysis result data is analyzed to identify abnormal situations and generate statistical reasoning data and node trust data; In the causal reasoning unit, nonlinear causal analysis is performed based on statistical reasoning data to generate causal relationship data; in the optimization analysis unit, multi-objective optimization analysis is performed based on the causal relationship data to generate optimization strategy data; Receive optimization strategy data, implement feedback iteration mechanism, generate decision support data, and support real-time supply chain management adjustments; In the visualization and monitoring module, decision support data and abnormal alarm data are displayed to achieve real-time monitoring and transparent management of the supply chain.

Images