Customs laboratory full-process safety management intelligent decision-making system and method

Abstract

The invention discloses a customs laboratory full-process safety management intelligent decision system and method, and belongs to the technical field of intelligent decision, and the system comprises a data collection module which is used for collecting multi-dimensional data of a customs laboratory in real time; the digital twinning construction module is used for constructing a virtual model based on the acquired data and generating multi-dimensional prediction data; the security risk analysis module is used for analyzing the prediction data to identify security risk factors and association relationships thereof; the intelligent decision generation module is used for predicting the risk probability and generating an intelligent decision scheme by utilizing a machine learning model in combination with the security risk factors and the incidence relation; and the decision scheme pushing module is used for accurately pushing the decision scheme to a manager according to a preset rule. According to the invention, the security risk early warning and management optimization of the whole process of the customs laboratory are realized through an intelligent means, the management efficiency is improved, the security risk is reduced, and the stable operation of the laboratory is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent decision-making systems, and more specifically, to an intelligent decision-making system and method for the whole-process safety management of a customs laboratory. Background Art

[0002] At present, with the increasingly prosperous global trade, customs laboratories shoulder the important mission of ensuring the quality and safety of imported and exported goods, and safeguarding national economic interests and public safety. The work of customs laboratories involves a large number of complex detection processes, covering multiple key links such as personnel operation, equipment operation, sample processing, reagent and consumable use, and environmental control.

[0003] Currently, the safety management of most customs laboratories still relies on traditional manual inspections and simple recording methods, with low efficiency and obvious limitations. For example, for the supervision of personnel operation behaviors, it mainly relies on regular training and on-site spot checks, making it difficult to achieve real-time and comprehensive monitoring, and it is difficult to effectively prevent the risk of personnel operation errors. In terms of equipment management, only regular maintenance and repair after a failure can be used to ensure the operation of the equipment, and it is impossible to predict equipment failures in advance. Once the equipment suddenly fails, it may lead to delays in detection work and affect the customs clearance efficiency.

[0004] In sample management, the tracking of sample source information, storage conditions, and detection progress mainly relies on manual records, which are prone to information errors or omissions, increasing the risk of sample contamination. In terms of reagent and consumable management, the inventory is mainly counted manually, and it is difficult to grasp the inventory quantity and expiration date in real time, and it is easy to have a shortage of reagents and consumables, affecting the experimental progress. For laboratory environmental parameters, such as temperature, humidity, air pressure, and light, although some laboratories have installed simple monitoring equipment, they cannot conduct in-depth analysis and early warning of environmental anomalies.

[0005] With the rapid development of emerging technologies such as big data, the Internet of Things, and artificial intelligence, all industries are actively exploring digital transformation and intelligent upgrading. The safety management of customs laboratories also urgently needs to introduce advanced technologies to realize the transformation from traditional management models to intelligent and refined management models to meet the growing trade demands and continuously improving safety management standards.

[0006] Therefore, how to provide an intelligent decision-making system and method that can achieve whole-process safety management is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0007] In view of this, the present invention provides an intelligent decision-making system and method for the whole-process safety management of customs laboratories. By combining technical means such as data collection, digital twin modeling, intelligent risk analysis, intelligent decision-making generation, real-time push, and dynamic optimization, a comprehensive, efficient, and intelligent intelligent decision-making system for the whole-process safety management of customs laboratories is constructed to improve the safety management level of customs laboratories, enhance management efficiency, optimize resource allocation, and achieve the continuous optimization and adaptive ability of the system through a dynamic feedback mechanism to ensure the safe and efficient operation of the laboratories.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] On the one hand, the present invention provides an intelligent decision-making system for the whole-process safety management of customs laboratories, including:

[0010] A data collection module for real-time collecting multi-dimensional data of customs laboratories, including personnel data, equipment operation status data, sample data, reagent and consumable data, and environmental data;

[0011] A digital twin construction module for constructing a virtual model of the customs laboratory based on the collected multi-dimensional data, and predicting the multi-dimensional data based on the virtual model to obtain multi-dimensional prediction data;

[0012] A safety risk analysis module for analyzing the multi-dimensional prediction data to determine safety risk factors and their correlation relationships;

[0013] An intelligent decision-making generation module for generating corresponding intelligent decision-making schemes according to the safety risk factors and the correlation relationships, in combination with the multi-dimensional prediction data, by using a machine learning model;

[0014] A decision-making scheme push module for real-time pushing the intelligent decision-making scheme to the management personnel of the customs laboratory;

[0015] A system update and optimization module for establishing a feedback mechanism to dynamically update and optimize the virtual model and the machine learning model according to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making scheme.

[0016] Preferably, the safety risk analysis module includes:

[0017] A classification unit for identifying the key constituent units and business scenarios of the customs laboratory according to the multi-dimensional prediction data;

[0018] A feature extraction unit, configured to extract basic security risk factors of the customs laboratory according to the business scenario, and extract security risk characterization factors, overall security risk factors of the customs laboratory, and unit security risk factors of the key constituent units according to the basic security risk factors;

[0019] A relationship construction unit, configured to construct a first association relationship, a second association relationship, and a third association relationship between the basic security risk factors and the security risk characterization factors, the overall security risk factors, and the unit security risk factors respectively.

[0020] Preferably, the intelligent decision-making generation module includes:

[0021] A model construction unit, configured to construct a machine learning model based on the basic security risk factors and perform training to predict the risk probability;

[0022] A risk prediction unit, configured to input the multi-dimensional prediction data into the machine learning model for risk prediction;

[0023] An optimization unit, configured to introduce the overall security risk factors, the unit security risk factors, and the first association relationship, the second association relationship, and the third association relationship to optimize the risk probability;

[0024] A decision-making unit, configured to generate an intelligent decision-making scheme according to the risk probability.

[0025] Preferably, the decision-making scheme push module includes:

[0026] A push condition determination unit, configured to read a preset push rule and determine a first condition that the recipients of each decision-making data in the intelligent decision-making scheme need to meet according to the push rule;

[0027] A first screening unit, configured to select recipients who meet the first condition from each recipient;

[0028] A sending unit, configured to send each decision-making data to the corresponding recipient.

[0029] Preferably, the system update and optimization module includes:

[0030] A feedback data collection unit, configured to obtain the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making scheme;

[0031] A second screening unit, configured to perform a primary screening process on the operation feedback information and the execution feedback information to obtain target feedback information;

[0032] A risk ranking unit, configured to construct a feedback risk ranking model according to the operation feedback information and the execution feedback information, and train the feedback risk ranking model;

[0033] A third screening unit, configured to perform secondary screening on the target feedback information based on the pre-trained feedback risk ranking model to obtain valid feedback information;

[0034] A scoring unit, configured to perform scoring on the valid feedback information based on the pre-trained feedback risk ranking model to obtain a feedback score corresponding to the valid feedback information;

[0035] An updating unit, configured to determine final feedback information according to the feedback score, and dynamically update and optimize the virtual model and the majority of machine learning models based on the final feedback information.

[0036] On the other hand, the present invention provides an intelligent decision-making method for the full-process safety management of a customs laboratory, including the following steps:

[0037] Real-time collect multi-dimensional data of the customs laboratory, including personnel data, equipment operation status data, sample data, reagent and consumable data, and environmental data;

[0038] Based on the collected multi-dimensional data, use digital twin technology to construct a virtual model of the customs laboratory, and predict the multi-dimensional data based on the virtual model to obtain multi-dimensional prediction data;

[0039] Analyze the multi-dimensional prediction data in the virtual model to determine safety risk factors and their correlation relationships;

[0040] According to the safety risk factors and the correlation relationships, combined with the multi-dimensional prediction data, use a machine learning model to generate corresponding intelligent decision-making solutions;

[0041] Push the intelligent decision-making solution to the management personnel of the customs laboratory in real time;

[0042] Dynamically update and optimize the virtual model and the machine learning model according to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making solution.

[0043] Preferably, analyzing the multi-dimensional prediction data in the virtual model to determine safety risk factors and their correlation relationships includes:

[0044] Identify the key constituent units and business scenarios of the customs laboratory according to the multi-dimensional prediction data;

[0045] Extract the basic security risk factors of the customs laboratory according to the business scenario, and extract the security risk characterization factors, overall security risk factors of the customs laboratory, and unit security risk factors of the key constituent units according to the basic security risk factors;

[0046] Construct the first correlation relationship, the second correlation relationship, and the third correlation relationship between the basic security risk factors and the security risk characterization factors, the overall security risk factors, and the unit security risk factors respectively.

[0047] Preferably, according to the security risk factors and the correlation relationships, combined with the multi-dimensional prediction data, use a machine learning model to generate corresponding intelligent decision-making solutions, including:

[0048] Construct a machine learning model based on the basic security risk factors and train it to predict the risk probability;

[0049] Input the multi-dimensional prediction data into the machine learning model for risk prediction;

[0050] Introduce the overall security risk factors, the unit security risk factors, and the first correlation relationship, the second correlation relationship, and the third correlation relationship to optimize the risk probability;

[0051] Generate an intelligent decision-making solution according to the risk probability.

[0052] Preferably, push the intelligent decision-making solution to the management personnel of the customs laboratory in real time, including:

[0053] Read the preset push rules, and determine the first conditions that the recipients of each decision-making data in the intelligent decision-making solution need to meet according to the push rules;

[0054] Select the recipients who meet the first conditions from each recipient;

[0055] Send each decision-making data to the corresponding recipient.

[0056] Preferably, dynamically update and optimize the virtual model and the machine learning model according to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making solution, including:

[0057] Obtain the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making solution;

[0058] Perform a primary screening process on the operation feedback information and the execution feedback information to obtain the target feedback information;

[0059] Construct a feedback risk ranking model based on the operation feedback information and the execution feedback information, and train the feedback risk ranking model;

[0060] Perform secondary screening on the target feedback information based on the pre-trained feedback risk ranking model to obtain effective feedback information;

[0061] Perform a scoring process on the effective feedback information based on the pre-trained feedback risk ranking model to obtain a feedback score corresponding to the effective feedback information;

[0062] Determine the final feedback information according to the feedback score, and dynamically update and optimize the virtual model and the majority machine learning model based on the final feedback information.

[0063] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a full-process safety management intelligent decision-making system and method for a customs laboratory. The multi-dimensional data such as personnel, equipment, samples, reagent consumables, and environment are obtained in real time through a data acquisition module, and a virtual model is constructed using digital twin technology to monitor and predict the operation status of the laboratory in real time. The safety risk analysis module analyzes the data in the virtual model, identifies safety risk factors and their correlation relationships, and discovers potential hidden dangers in advance. The system adopts a machine learning model, combines risk factors and correlation relationships, generates an intelligent decision-making plan, and provides scientific and accurate decision-making support for management personnel. The push module sends the decision-making plan to relevant personnel in real time according to preset rules to ensure timely information transmission and quick response. At the same time, the system collects operation and execution feedback information through a feedback mechanism, dynamically updates and optimizes the virtual model and the machine learning model to adapt to changes in the laboratory operation environment, and continuously improves the performance and adaptability of the system. The present invention realizes the full-process safety management of the customs laboratory through intelligent means, improves management efficiency, optimizes resource allocation, and ensures the long-term effectiveness and flexibility of the system through a dynamic feedback and optimization mechanism, providing a strong guarantee for the safe operation of the customs laboratory. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings according to the provided drawings without creative efforts.

[0065] Figure 1 It is a schematic structural diagram provided by the present invention;

[0066] Figure 2 It is a schematic flow diagram provided by the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0067] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0068] The embodiment of the present invention discloses an intelligent decision-making system for the whole process safety management of customs laboratories, such as Figure 1 As shown, including:

[0069] The data acquisition module is used to collect multi-dimensional data of the customs laboratory in real time, including personnel data, equipment data, sample data, reagent and consumables data, and environmental data; among them, personnel data includes the entry and exit records, operation behavior data, and qualification information of personnel in the customs laboratory; equipment data includes the equipment's operating status data, usage frequency data, and maintenance record data; sample data includes sample source information, storage condition data, and detection progress data; reagent and consumables data includes reagent and consumables inventory quantity data, expiration date data, and usage record data; environmental data includes environmental parameter data such as temperature, humidity, air pressure, and light.

[0070] The digital twin construction module is used to build a virtual model of the customs laboratory based on the collected multi-dimensional data using digital twin technology, and to predict the multi-dimensional data based on the virtual model to obtain multi-dimensional prediction data; the model has a real-time data update function, which can synchronously map various data in the physical laboratory to the virtual model, realize the synchronous interaction between the physical laboratory and the virtual model, and comprehensively and accurately reflect the operating status and safety status of the customs laboratory. At the same time, the virtual model also has functions such as virtual simulation, fault prediction, remote monitoring and collaborative interaction, which provide strong support for subsequent security risk analysis and intelligent decision-making.

[0071] The safety risk analysis module analyzes the multi-dimensional prediction data in the virtual model to determine the safety risk factors and their correlations, including pre-processing operations such as data cleaning and feature extraction to remove noise and invalid data and extract key features related to safety risks. Then, it mines the safety risk laws and patterns hidden in the data to identify potential safety risk factors, including the risk of human operation errors, equipment failures, sample contamination, reagent and consumable shortages, and environmental abnormalities.

[0072] The intelligent decision-making generation module generates corresponding intelligent decision-making solutions based on security risk factors and correlations, combined with multi-dimensional prediction data, and using machine learning models;

[0073] The decision-making solution push module real-time pushes the intelligent decision-making solutions to the management personnel of the customs laboratory; and visually real-time pushes the generated intelligent decision-making solutions to the management personnel and relevant operators of the customs laboratory through mobile terminals, computers or other display devices. The visual display forms include various forms such as text descriptions, chart displays, warning signals, etc., to ensure that the management personnel and operators can quickly and accurately understand and execute the decision-making solutions and take corresponding safety management measures in a timely manner.

[0074] The system update and optimization module is used to establish a feedback mechanism and dynamically update and optimize the virtual model and machine learning model according to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making solutions. The update and optimization contents include adjusting the data collection strategy, correcting the virtual model, optimizing the algorithm parameters, and improving the decision-making rules, etc., to adapt to the changes in the business of the customs laboratory and the improvement of safety management requirements, and continuously improve the accuracy and adaptability of safety management decisions.

[0075] Specifically, the security risk analysis module includes:

[0076] The classification unit is used to identify the key constituent units and business scenarios of the customs laboratory according to the multi-dimensional prediction data;

[0077] The feature extraction unit is used to extract the basic security risk factors of the customs laboratory according to the business scenarios, and extract the security risk characterization factors, overall security risk factors and unit security risk factors of the key constituent units of the customs laboratory according to the basic security risk factors; on the basis of classification, further deeply analyze each business scenario and extract the basic features related to security. These basic features are called basic security risk factors, which are the direct factors that may cause safety problems in the laboratory operation, such as the failure rate of equipment, the probability of human operation errors, the danger of reagents, etc. Based on these basic factors, further extract the indicators that can characterize the overall security status of the laboratory, that is, the security risk characterization factors. At the same time, overall security risk factors and unit security risk factors are respectively extracted for the overall operation situation and key constituent units of the laboratory. These factors reflect the security risk status of the laboratory from different levels and provide comprehensive data support for subsequent risk assessment.

[0078] The relationship construction unit is used to construct the first association relationship, the second association relationship and the third association relationship between the basic security risk factors and the security risk characterization factors, overall security risk factors and unit security risk factors respectively. By constructing these relationships, it can clearly show how the basic risk factors affect the characterization factors, overall factors and unit factors, thus forming a complete risk conduction network. This networked risk relationship model helps to deeply understand the propagation path and influence range of risks and provides a scientific basis for subsequent risk prediction and decision-making.

[0079] Furthermore, the intelligent decision-making generation module includes:

[0080] A model construction unit, which is used to construct and train a machine learning model based on basic security risk factors to predict the risk probability; the basic security risk factors are the key factors that may cause security problems during the operation of the laboratory, such as equipment failure rate, probability of personnel operation errors, danger of reagents, etc. These factors are used as input features of the model to train the machine learning model so that it can learn the internal relationship between the risk factors and security risks. Through training with a large amount of historical data and real-time data, the model can accurately predict the probability of risk occurrence, providing data support for subsequent risk assessment and decision-making.

[0081] A risk prediction unit, which is used to input multi-dimensional prediction data into the machine learning model for risk prediction; the risk prediction unit inputs multi-dimensional prediction data into the trained machine learning model for real-time risk prediction. The multi-dimensional prediction data includes personnel data, equipment operation status data, sample data, reagent and consumable data, environmental data, etc. These data are generated through digital twin technology and can reflect the operation status of the laboratory in real time. The machine learning model calculates the risk probability at each time point based on the input multi-dimensional prediction data, thus providing real-time risk warning information for the management personnel. This process can help the management personnel discover potential security hazards in advance and take measures to intervene in a timely manner.

[0082] An optimization unit, which is used to introduce the overall security risk factor, unit security risk factor, as well as the first correlation relationship, the second correlation relationship, and the third correlation relationship to optimize the risk probability; introduce the overall security risk factor, unit security risk factor, as well as the first correlation relationship, the second correlation relationship, and the third correlation relationship to optimize the preliminarily predicted risk probability. The overall security risk factor reflects the overall security status of the laboratory operation, the unit security risk factor evaluates the key constituent units of the laboratory, and the correlation relationship shows the mutual influence between different risk factors. By comprehensively considering these factors, the optimization unit can adjust and correct the risk probability to make it more in line with the actual operation situation of the laboratory. For example, if the failure rate of a certain device is relatively high and the device is associated with multiple key business processes, the optimization unit will adjust the risk probability according to these correlation relationships to more accurately reflect the impact of the device failure on the overall security.

[0083] Decision-making unit, which is used to generate intelligent decision-making solutions based on risk probabilities. The generated decision-making solutions include, but are not limited to, equipment maintenance plans, personnel training suggestions, optimization of sample processing procedures, reagent and consumable management strategies, etc. The decision-making solutions are designed to help managers optimize resource allocation and improve operational efficiency while ensuring laboratory safety. For example, if the risk probability of a certain piece of equipment is high, the decision-making solution may recommend early equipment maintenance or replacement; if there are high safety risks in a certain operation process, the decision-making solution may recommend additional training for relevant personnel. Through these intelligent decision-making solutions, managers can scientifically address various safety risks in the laboratory and ensure the smooth operation of the laboratory.

[0084] Furthermore, the decision-making solution push module includes:

[0085] Push condition determination unit, which is used to read the pre-set push rules and determine the first conditions that the recipients of each decision-making data in the intelligent decision-making solution need to meet according to the push rules; this unit is responsible for reading the pre-set push rules. These rules define the conditions that the recipients of each decision-making data in the intelligent decision-making solution need to meet, such as the recipient's scope of responsibilities, permission level, department where they are located, etc. By clarifying the push conditions, it ensures that the decision-making solution can be accurately pushed to the managers who most need it, avoiding information redundancy and mis-pushing, and improving the efficiency and pertinence of information transmission. The push rules can be customized based on the organizational structure and management process of the laboratory. For example, decision-making solutions related to equipment maintenance are pushed to the person in charge of the equipment management department, while personnel training suggestions are pushed to the human resources department.

[0086] First screening unit, which is used to select the recipients who meet the first conditions from each recipient; according to the conditions set by the push condition determination unit, it screens out the recipients who meet the conditions from all possible recipients. Through screening, it ensures that only managers who meet the push conditions can receive the relevant decision-making solutions, avoiding unnecessary information being received by irrelevant personnel. Specifically, the detailed information of the recipients can be obtained through the laboratory's personnel management system and matched with the push conditions to screen out a list of recipients who meet the conditions.

[0087] Sending unit, which is used to send each decision-making data to the corresponding recipient. The sending unit can push the decision-making solution to the recipient through various communication channels (such as email, instant messaging tools, text messages, etc.) to ensure the timely transmission and receipt of information.

[0088] In another embodiment, the system update and optimization module includes:

[0089] Feedback data collection unit, which is used to obtain the operation feedback information of the customs laboratory and the implementation feedback information of managers on the intelligent decision-making solution;

[0090] By collecting feedback information, the system can understand the actual effects of decision-making solutions and new problems in the laboratory operation, providing data support for subsequent optimization. The feedback information can be collected through various channels such as the laboratory management system, the feedback forms of management personnel, and the equipment operation logs.

[0091] The second screening unit is used to perform a primary screening process on the operation feedback information and the execution feedback information to obtain the target feedback information; through screening, it ensures that the feedback information for subsequent processing is valuable and relevant, improving the processing efficiency of the system. Data cleaning and preprocessing techniques can be used to classify, denoise, and screen the feedback information to extract valuable information.

[0092] The risk ranking unit is used to construct a feedback risk ranking model based on the operation feedback information and the execution feedback information, and train the feedback risk ranking model; through the feedback risk ranking model, the risks in the feedback information are quantified and ranked to help the system identify the most important risk points. Machine learning algorithms (such as decision trees, random forests, etc.) can be used to analyze the feedback information, construct the feedback risk ranking model, and train and optimize it through historical data.

[0093] The third screening unit is used to perform a secondary screening process on the target feedback information based on the pre-trained feedback risk ranking model to obtain the effective feedback information;

[0094] Through the feedback risk ranking model, the risks in the feedback information are quantified and ranked to help the system identify the most important risk points. This unit can use machine learning algorithms (such as decision trees, random forests, etc.) to analyze the feedback information, construct the feedback risk ranking model, and train and optimize it through historical data.

[0095] The scoring unit is used to perform a scoring process on the effective feedback information based on the pre-trained feedback risk ranking model to obtain the feedback score corresponding to the effective feedback information; through the feedback risk ranking model, the risks in the feedback information are quantified and ranked to help the system identify the most important risk points. Machine learning algorithms (such as decision trees, random forests, etc.) can be used to analyze the feedback information, construct the feedback risk ranking model, and train and optimize it through historical data.

[0096] The update unit is used to determine the final feedback information according to the feedback score, and dynamically update and optimize the virtual model and most machine learning models based on the final feedback information. This unit can select the most important feedback information as the update basis according to the feedback score to adjust and optimize the virtual model and machine learning models. For example, if a piece of feedback information indicates that the failure rate of a certain device is relatively high, the system can adjust the parameters of the device in the virtual model and retrain the machine learning model to improve the accuracy of risk prediction.

[0097] On the other hand, the present invention provides an intelligent decision-making method for the whole-process safety management of a customs laboratory, as Figure 2 shown, including the following steps:

[0098] Collect multi-dimensional data of the customs laboratory in real time, including personnel data, equipment operation status data, sample data, reagent and consumable data, and environmental data;

[0099] Based on the collected multi-dimensional data, use digital twin technology to construct a virtual model of the customs laboratory, and predict the multi-dimensional data based on the virtual model to obtain multi-dimensional prediction data;

[0100] Analyze the multi-dimensional prediction data in the virtual model to determine safety risk factors and their correlation relationships;

[0101] According to the safety risk factors and their correlation relationships, combined with the multi-dimensional prediction data, use a machine learning model to generate corresponding intelligent decision-making solutions;

[0102] Push the intelligent decision-making solutions to the management personnel of the customs laboratory in real time;

[0103] According to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making solutions, dynamically update and optimize the virtual model and the machine learning model.

[0104] Furthermore, analyzing the multi-dimensional prediction data in the virtual model to determine safety risk factors and their correlation relationships includes:

[0105] Identify the key constituent units and business scenarios of the customs laboratory according to the multi-dimensional prediction data;

[0106] Extract the basic safety risk factors of the customs laboratory according to the business scenarios, and extract the safety risk characterization factors, overall safety risk factors and unit safety risk factors of the key constituent units of the customs laboratory according to the basic safety risk factors;

[0107] Construct the first correlation relationship, the second correlation relationship and the third correlation relationship between the basic safety risk factors and the safety risk characterization factors, the overall safety risk factors and the unit safety risk factors respectively.

[0108] Even further, according to the safety risk factors and their correlation relationships, combined with the multi-dimensional prediction data, using a machine learning model to generate corresponding intelligent decision-making solutions includes:

[0109] Construct a machine learning model according to the basic safety risk factors and train it to predict the risk probability;

[0110] Input the multi-dimensional prediction data into the machine learning model for risk prediction;

[0111] Introduce the overall security risk factor, the unit security risk factor, as well as the first correlation relationship, the second correlation relationship, and the third correlation relationship to optimize the risk probability;

[0112] Generate an intelligent decision-making plan based on the risk probability.

[0113] Specifically, push the intelligent decision-making plan to the management personnel of the customs laboratory in real time, including:

[0114] Read the pre-set push rules, and determine the first conditions that the recipients of each decision-making data in the intelligent decision-making plan need to meet according to the push rules;

[0115] Select the recipients who meet the first conditions from each recipient;

[0116] Send each decision-making data to the corresponding recipient.

[0117] Preferably, dynamically update and optimize the virtual model and the machine learning model according to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making plan, including:

[0118] Obtain the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making plan;

[0119] Perform a primary screening process on the operation feedback information and the execution feedback information to obtain the target feedback information;

[0120] Construct a feedback risk ranking model based on the operation feedback information and the execution feedback information, and train the feedback risk ranking model;

[0121] Perform a secondary screening process on the target feedback information based on the pre-trained feedback risk ranking model to obtain the effective feedback information;

[0122] Perform a scoring process on the effective feedback information based on the pre-trained feedback risk ranking model to obtain the feedback score corresponding to the effective feedback information;

[0123] Determine the final feedback information according to the feedback score, and dynamically update and optimize the virtual model and most machine learning models based on the final feedback information.

[0124] Each embodiment in this specification is described in a progressive manner. The key points of each embodiment are the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method part.

[0125] The foregoing description of the disclosed embodiments enables those skilled in the art to practice or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent decision-making system for the whole process safety management of customs laboratories, characterized by: include: Data collection module, used to collect multi-dimensional data of customs laboratories in real time, including personnel data, equipment operation status data, sample data, reagent and consumables data, and environmental data; A digital twin construction module, used to construct a virtual model of the customs laboratory based on the collected multi-dimensional data using the digital twin technology, and to predict the multi-dimensional data based on the virtual model to obtain multi-dimensional prediction data; A safety risk analysis module analyzes the multi-dimensional prediction data to determine safety risk factors and correlation relationships; An intelligent decision-making generation module generates a corresponding intelligent decision-making plan based on the security risk factors and the association relationships in combination with the multi-dimensional prediction data and using a machine learning model; A decision-making solution push module pushes the intelligent decision-making solution to the managers of the customs laboratory in real time; The system update optimization module is used to establish a feedback mechanism to dynamically update and optimize the virtual model and the machine learning model based on the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making plan.

2. According to claim 1, a customs laboratory full-process safety management intelligent decision-making system is characterized by: The security risk analysis module includes: A classification unit, used for identifying key components and business scenarios of the customs laboratory according to the multi-dimensional prediction data; A feature extraction unit, configured to extract the basic security risk factor of the customs laboratory according to the business scenario, and extract the security risk characterization factor, the overall security risk factor and the unit security risk factor of the key component unit of the customs laboratory according to the basic security risk factor; The relationship construction unit is used to construct a first association relationship, a second association relationship and a third association relationship between the basic safety risk factor and the safety risk characterization factor, the overall safety risk factor and the unit safety risk factor respectively.

3. According to claim 2, a customs laboratory full-process safety management intelligent decision-making system is characterized by: The intelligent decision-making module includes: A model building unit, used to build a machine learning model based on the basic security risk factors and perform training to predict risk probability; A risk prediction unit, used for inputting the multi-dimensional prediction data into a machine learning model to perform risk prediction; An optimization unit, used for introducing the overall safety risk factor and the unit safety risk factor as well as the first association relationship, the second association relationship and the third association relationship to optimize the risk probability; A decision-making unit is used to generate an intelligent decision-making plan according to the risk probability.

4. According to claim 1, a customs laboratory full-process safety management intelligent decision-making system is characterized by: The decision-making solution push module includes: A push condition determination unit, used to read a preset push rule and determine a first condition that needs to be satisfied by a recipient of each decision data in the intelligent decision-making scheme according to the push rule; A first screening unit, used to select a recipient that meets a first condition from among the recipients; The sending unit is used to send each decision data to the corresponding receiving party.

5. According to claim 1, a customs laboratory full-process safety management intelligent decision-making system is characterized by: The system update optimization module includes: A feedback data collection unit, used to obtain operational feedback information of the customs laboratory and executive feedback information of the management personnel on the intelligent decision-making scheme; A second screening unit, configured to perform a screening process on the operation feedback information and the execution feedback information to obtain target feedback information; A risk ranking unit, configured to construct a feedback risk ranking model according to the operation feedback information and the execution feedback information, and train the feedback risk ranking model; A second screening unit, configured to perform secondary screening processing on the target feedback information based on the pre-trained feedback risk ranking model to obtain valid feedback information; A scoring unit, configured to score the valid feedback information based on the pre-trained feedback risk ranking model to obtain a feedback score corresponding to the valid feedback information; An updating unit is used to determine final feedback information according to the feedback score, and dynamically update and optimize the virtual model and the machine learning model based on the final feedback information.

6. An intelligent decision-making method for the whole process safety management of customs laboratories, characterized in that: The following steps are involved: Real-time collection of multi-dimensional data of customs laboratories, including personnel data, equipment operation status data, sample data, reagent and consumables data, and environmental data; Based on the collected multi-dimensional data, a virtual model of the customs laboratory is constructed using the digital twin technology, and the multi-dimensional data is predicted based on the virtual model to obtain multi-dimensional prediction data; Analyze the multi-dimensional prediction data in the virtual model to determine the safety risk factors and their correlations; According to the security risk factors and the association relationships, combined with the multi-dimensional prediction data, a machine learning model is used to generate a corresponding intelligent decision-making solution; Push the intelligent decision-making solution to the managers of the customs laboratory in real time; The virtual model and the machine learning model are dynamically updated and optimized based on the operation feedback information of the customs laboratory and the execution feedback information of the intelligent decision-making plan by the management personnel.

7. According to claim 6, a customs laboratory full-process safety management intelligent decision-making method is characterized by: Analyze the multi-dimensional prediction data in the virtual model to determine the safety risk factors and their correlations, including: Identify key components and business scenarios of the customs laboratory based on the multi-dimensional prediction data; According to the business scenario, extract the basic security risk factor of the customs laboratory, and according to the basic security risk factor, extract the security risk characterization factor, the overall security risk factor and the unit security risk factor of the key component unit of the customs laboratory; A first association relationship, a second association relationship and a third association relationship are constructed between the basic safety risk factor and the safety risk characterization factor, the overall safety risk factor and the unit safety risk factor respectively.

8. According to claim 7, a customs laboratory full-process safety management intelligent decision-making method is characterized by: According to the security risk factors and the association relationships, combined with the multi-dimensional prediction data, a machine learning model is used to generate a corresponding intelligent decision-making solution, including: Constructing and training a machine learning model based on the basic safety risk factors to predict risk probability; Inputting the multidimensional prediction data into a machine learning model to perform risk prediction; Introducing the overall safety risk factor and the unit safety risk factor as well as the first association relationship, the second association relationship and the third association relationship to optimize the risk probability; Generate an intelligent decision-making plan based on the risk probability.

9. According to claim 6, a customs laboratory full-process safety management intelligent decision-making method is characterized by: Push the intelligent decision-making solution to the managers of the customs laboratory in real time, including: Reading a preset push rule, and determining a first condition that a recipient of each decision data in the intelligent decision-making scheme needs to meet according to the push rule; Selecting a receiver that satisfies the first condition from among the receivers; Send each decision data to the corresponding recipient.

10. According to claim 6, a customs laboratory full-process safety management intelligent decision-making method is characterized by: According to the operation feedback information of the customs laboratory and the execution feedback information of the management personnel on the intelligent decision-making plan, the virtual model and the machine learning model are dynamically updated and optimized, including: Obtain operational feedback from customs laboratories and feedback from managers on the implementation of the intelligent decision-making solution; Performing a screening process on the operation feedback information and the execution feedback information to obtain target feedback information; Constructing a feedback risk ranking model according to the operation feedback information and the execution feedback information, and training the feedback risk ranking model; Performing secondary screening on the target feedback information based on the pre-trained feedback risk ranking model to obtain effective feedback information; Scoring the effective feedback information based on the pre-trained feedback risk ranking model to obtain a feedback score corresponding to the effective feedback information; Final feedback information is determined according to the feedback score, and the virtual model and the majority of machine learning models are dynamically updated and optimized based on the final feedback information.

Images