Key management system and method based on medical block chain

Abstract

The invention relates to the technical field of information security, in particular to a key management system and method based on a medical block chain, and the system comprises a state evaluation module, a reverse tracking module, a foresight analysis module, an intelligent alarm threshold adjustment module, a resource priority scheduling module, a reliability analysis and fault tolerance module, and a process optimization and decision support module. According to the method, risk assessment and classification are effectively carried out and a key access control strategy is adjusted by adopting a support vector machine, a decision tree algorithm and a deep learning algorithm, a key use trend is predicted by utilizing an autoregression integral moving average model and a long-short-term memory network, and the key access control strategy is optimized through an isolated forest and a deep learning network. An alarm threshold value is dynamically adjusted, the early warning capability of the system for potential security threats is improved, a key management process is optimized through a genetic algorithm and a particle swarm optimization technology, a security decision and a response strategy are executed in combination with a rule engine and a machine learning model, and the response efficiency and the decision accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the field of information security technology, and in particular to a key management system based on medical blockchain and a method thereof. Background Art

[0002] The field of information security technology is dedicated to protecting information from unauthorized access, disclosure, modification, inspection or destruction. The core goals of information security include ensuring the confidentiality, integrity and availability of information. In the medical field, information security technology involves the protection of patient privacy data, the secure storage and transmission of medical records, and secure communications between medical service providers.

[0003] Among them, the key management system based on medical blockchain is a system that uses blockchain technology to manage encryption keys, aiming to improve the security of medical data processing and storage. The purpose of the system is to protect the confidentiality and integrity of medical data by effectively managing patient keys to prevent data leakage or unauthorized access. The system is designed to achieve a highly secure data encryption and decryption process, ensuring that only authorized users can access sensitive medical information, thereby protecting patient privacy and enhancing data security.

[0004] Traditional blockchain key management systems have obvious deficiencies in real-time security status assessment, risk prediction and management, abnormal behavior detection, and resource scheduling. Traditional systems rely on static security policies and manually set thresholds, lack in-depth analysis of user behavior and the ability to predict key usage trends, resulting in the system's response to security threats not being timely or accurate enough, and resource allocation and process management being relatively rigid and lacking in flexibility. They cannot be dynamically adjusted according to real-time system status and security requirements, limiting the system's ability to respond to complex security threats, resulting in higher risks for medical data and affecting the quality of medical services and the protection of patient privacy. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and propose a key management system and method based on medical blockchain.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions: a key management system based on medical blockchain includes a status assessment module, a reverse tracking module, a forward-looking analysis module, an intelligent alarm threshold adjustment module, a resource priority scheduling module, a reliability analysis and fault tolerance module, and a process optimization and decision support module;

[0007] The status assessment module uses support vector machines and decision tree algorithms to conduct real-time security status assessment of the medical system based on security performance indicators and user blockchain key usage information. It uses deep learning algorithms to analyze user behavior, conduct risk assessment and classification, and adjust key access control policies to generate security status assessment results.

[0008] The reverse tracking module uses decision trees and cluster analysis algorithms based on the security status assessment results to analyze the user's blockchain key usage and access path, and combines support vector machines to analyze user behavior patterns, conduct reverse tracking, optimize security policies, and generate tracking and analysis results;

[0009] The forward-looking analysis module uses the autoregressive integral moving average model and long short-term memory network to perform time series analysis based on the security status assessment results and the tracking and analysis results, predict the key usage trend, formulate security enhancement measures, optimize resource allocation, and generate trend predictions and security enhancement measures;

[0010] The intelligent alarm threshold adjustment module uses isolation forest and deep learning networks to detect abnormal system behavior based on trend prediction and security enhancement measures, dynamically adjusts the alarm threshold according to the detection results, and generates alarm threshold adjustment information;

[0011] The resource priority scheduling module uses the long short-term memory network and random forest algorithm to predict resource demand and access pressure based on the alarm threshold adjustment information, adjusts resource allocation and processing priority according to the prediction results, optimizes system response, and generates resource scheduling strategies;

[0012] The reliability analysis and fault tolerance module evaluates system reliability based on resource scheduling strategies, formulates fault tolerance mechanisms including redundant storage and automatic fault switching, and generates adaptive fault tolerance strategies through fault tree analysis and Monte Carlo simulation;

[0013] The process optimization and decision support module is based on an adaptive fault-tolerant strategy, applies genetic algorithms and particle swarm optimization to optimize the blockchain key management process, combines rule engines and machine learning models, executes security decisions and response strategies through decision trees and support vector machines, and generates optimized processes and decision information.

[0014] As a further solution of the present invention, the security status assessment results include security levels, potential risk points, and user behavior risk scores; the tracking and analysis results are specifically blockchain key illegal use paths, potential security vulnerability lists, and abnormal user behavior patterns; the trend prediction and security enhancement measures include warning time periods, security monitoring enhancement areas, and resource reallocation plans; the alarm threshold adjustment information includes abnormal behavior detection sensitivity, alarm level standards, and alarm response strategies; the resource scheduling strategies are specifically resource allocation priorities, resource security measures, and key operation resource retention strategies; the adaptive fault-tolerant strategies include redundant key storage solutions, automatic fault switching mechanisms, and key component monitoring points; the optimization processes and decision-making information are specifically process execution efficiency improvement points, automated security decision rules, and exception handling automatic response processes.

[0015] As a further solution of the present invention, the state assessment module includes a safety index analysis submodule, a user behavior assessment submodule, and a risk level determination submodule;

[0016] The security index analysis submodule uses a support vector machine algorithm based on security performance indicators and user blockchain key usage information to monitor and analyze the information security performance indicators of the medical system in real time, evaluate the current security status of the medical system, and generate security performance indicator analysis results;

[0017] The user behavior assessment submodule uses a decision tree algorithm based on the security performance index analysis results to analyze the user's behavior patterns, including key usage frequency and access patterns, identify potential risk behaviors and abnormal operations, and generate user behavior analysis results;

[0018] The risk level determination submodule performs risk assessment and classification of user behavior based on the user behavior analysis results through a deep learning algorithm, adjusts the key access control strategy according to the assessment results, optimizes the security of the system, and generates a security status assessment result.

[0019] As a further solution of the present invention, the reverse tracing module includes a key usage analysis submodule, an access path tracing submodule, and a security vulnerability identification submodule;

[0020] The key usage analysis submodule analyzes the user's blockchain key usage based on the security status assessment result and uses a decision tree algorithm to identify the compliance and security of blockchain key usage and generate a key usage analysis result;

[0021] The access path tracking submodule uses a cluster analysis algorithm based on the key usage analysis results to track the blockchain key access path, identify the risk nodes and sensitive areas passed through during the use of the blockchain key, and generate access path tracking results;

[0022] The security vulnerability identification submodule uses a support vector machine based on the access path tracking results to analyze user behavior patterns, perform reverse tracking, identify potential security vulnerabilities and improper operations, optimize security policies, and generate tracking and analysis results.

[0023] As a further solution of the present invention, the forward-looking analysis module includes a trend analysis submodule, a risk prediction submodule, and a strategy formulation submodule;

[0024] The trend analysis submodule applies the autoregressive integral moving average model and the long short-term memory network based on the security status assessment results and the tracking and analysis results to perform time series analysis, analyze historical key usage data, identify usage trends and patterns, and perform trend prediction to generate trend analysis results;

[0025] The risk prediction submodule uses deep learning models, including convolutional neural networks and recurrent neural networks, based on trend analysis results to analyze key usage trends, identify abnormal patterns and risk factors in trends, predict future security risks, and generate risk prediction results;

[0026] Based on the risk prediction results, the strategy formulation submodule adopts multi-attribute decision-making method and fuzzy set theory to optimize security policies and resource allocation, formulate countermeasures to potential security threats, adjust resource allocation, optimize the security performance and efficiency of the system, and generate trend predictions and security enhancement measures.

[0027] As a further solution of the present invention, the intelligent alarm threshold adjustment module includes an abnormal behavior detection submodule, a threshold adjustment strategy submodule, and an alarm generation submodule;

[0028] The abnormal behavior detection submodule uses the isolation forest algorithm to detect abnormal behavior based on trend prediction and security enhancement measures, constructs a forest by randomly selecting features and randomly selecting cut values, identifies rare events and abnormal behaviors, and generates abnormal behavior detection results;

[0029] The threshold adjustment strategy submodule adopts an adaptive learning algorithm based on the abnormal behavior detection results, dynamically updates the model according to the real-time data stream through online learning and incremental learning technology, matches the changes in behavior patterns, optimizes the alarm threshold, dynamically adjusts the threshold of the alarm system, and generates a threshold adjustment strategy;

[0030] The alarm generation submodule is based on the threshold adjustment strategy. When abnormal behavior that meets the adjusted threshold is detected, the event-driven programming model is used to automatically generate alarm information, notify the system administrator to take corresponding safety measures, and generate alarm threshold adjustment information.

[0031] As a further solution of the present invention, the resource priority scheduling module includes a resource demand prediction submodule, a priority adjustment submodule, and a resource allocation submodule;

[0032] The resource demand prediction submodule uses a long short-term memory network to analyze historical resource usage data and access patterns based on alarm threshold adjustment information, predicts resource demand and access pressure, applies a random forest algorithm, analyzes the impact of multiple factors on resource demand, predicts future trends in resource demand, and generates resource demand prediction results;

[0033] The priority adjustment submodule adopts a priority queue algorithm based on the resource demand prediction result, analyzes the urgency and importance of differentiated resource requests, adjusts the priority of resource allocation and processing, and generates a priority adjustment strategy;

[0034] The resource allocation submodule uses a dynamic resource allocation algorithm based on a priority adjustment strategy to dynamically adjust the allocation of system resources through a minimum connection number algorithm. It dynamically adjusts the allocation of computing resources, storage resources, and network bandwidth according to the current system status and priority strategy, optimizes system response and processing capabilities, and generates a resource scheduling strategy.

[0035] As a further solution of the present invention, the reliability analysis and fault tolerance module includes a reliability assessment submodule, a fault simulation submodule, and a fault tolerance strategy design submodule;

[0036] The reliability assessment submodule uses the fault tree analysis method based on the resource scheduling strategy to assess the reliability of system components, identifies key weak links by analyzing the factors and paths leading to system failures, and generates reliability assessment results;

[0037] The fault simulation submodule uses the Monte Carlo simulation method based on the reliability evaluation results to simulate the performance of the system under differentiated fault conditions, evaluate the effectiveness of the preset fault tolerance mechanism, and use random sampling technology to evaluate the behavior and performance of the system in the face of various random events, and generate fault simulation results;

[0038] The fault-tolerant strategy design submodule uses RAID technology to disperse and store data on multiple disks based on fault simulation results to achieve redundant storage, and uses a heartbeat detection mechanism to achieve automatic fault switching and generate an adaptive fault-tolerant strategy.

[0039] As a further solution of the present invention, the process optimization and decision support module includes a process analysis submodule, a decision logic design submodule, and an automated execution submodule.

[0040] The process analysis submodule uses a genetic algorithm based on an adaptive fault-tolerant strategy to evaluate and analyze the blockchain key management process, identify bottlenecks and inefficient links in the process by simulating the natural selection process, and generate process analysis results;

[0041] The decision logic design submodule uses a particle swarm optimization method based on the process analysis results to optimize the decision process by simulating the social behavior of bird flocks, and uses a rule engine to automatically perform process optimization and safety decisions according to preset rules to generate optimized decision logic;

[0042] The automated execution submodule is based on optimized decision logic, selects decision paths according to differentiated scenarios through decision trees, uses support vector machines to classify and identify security threat patterns, automatically adjusts system configurations and response measures, and generates optimized processes and decision information.

[0043] A key management method based on medical blockchain, the key management method based on medical blockchain is executed based on the above-mentioned key management system based on medical blockchain, and includes the following steps:

[0044] S1: Based on security performance indicators and user blockchain key usage information, support vector machine algorithm and decision tree algorithm are used to analyze the current security status of the medical system and generate security status analysis results;

[0045] S2: Based on the security status analysis results, apply deep learning algorithms to analyze user behaviors, identify potential risks and abnormal operations, and generate risk identification and user behavior analysis results;

[0046] S3: Based on the risk identification and user behavior analysis results, combined with decision tree and cluster analysis algorithms, reverse tracing of key usage is performed to identify illegal usage paths and potential security vulnerabilities, and generate reverse tracing analysis results;

[0047] S4: Based on the reverse tracing analysis results, the autoregressive integrated moving average model and the long short-term memory network are used to predict the key usage trend, formulate security enhancement measures, and generate trend prediction results;

[0048] S5: Based on the trend prediction results, an isolation forest algorithm and a deep learning network are used to detect abnormal behavior, and the alarm threshold is dynamically adjusted according to the detection results to generate an alarm threshold adjustment strategy;

[0049] S6: Based on the alarm threshold adjustment strategy, use the long short-term memory network and random forest algorithm to predict resource demand and access pressure, adjust resource allocation and processing priority, and generate resource scheduling optimization solutions;

[0050] S7: Based on the resource scheduling optimization scheme, the system reliability is evaluated through fault tree analysis and Monte Carlo simulation, and redundant storage and automatic fault switching mechanisms are formulated to generate adaptive fault tolerance strategies;

[0051] S8: Based on the adaptive fault-tolerant strategy, genetic algorithm and particle swarm optimization are applied to optimize the key management process. The rule engine, decision tree and support vector machine are combined to automatically execute security decisions and response strategies and generate optimized processes and decision information.

[0052] Compared with the prior art, the advantages and positive effects of the present invention are:

[0053] In the present invention, by adopting support vector machines, decision tree algorithms and deep learning algorithms, accurate assessment of the real-time security status of the medical system and in-depth analysis of user behavior are achieved, risk assessment and classification are effectively performed, and the key access control strategy is adjusted. The autoregressive integral sliding average model and long short-term memory network are used to predict the key usage trend, and anomaly detection is performed through isolation forest and deep learning networks, and the alarm threshold is dynamically adjusted, which significantly improves the system's early warning ability for potential security threats. The key management process is optimized through genetic algorithms and particle swarm optimization technology, and security decisions and response strategies are executed in combination with rule engines and machine learning models, which greatly improves the system's response efficiency and decision accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a system flow chart of the present invention;

[0055] Figure 2 It is a schematic diagram of the system framework of the present invention;

[0056] Figure 3 is a flow chart of the state assessment module of the present invention;

[0057] Figure 4 It is a flow chart of the reverse tracking module of the present invention;

[0058] Figure 5 is a flow chart of the forward-looking analysis module of the present invention;

[0059] Figure 6 It is a flow chart of the intelligent alarm threshold adjustment module of the present invention;

[0060] Figure 7 It is a flow chart of the resource priority scheduling module of the present invention;

[0061] Figure 8 It is a flow chart of the reliability analysis and fault tolerance module of the present invention;

[0062] Fig. 9 It is a flow chart of the process optimization and decision support module of the present invention;

[0063] Fig.10 It is a schematic diagram of the method steps of the present invention. DETAILED DESCRIPTION

[0064] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0065] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0066] Embodiment 1

[0067] See also Figure 1 to Figure 2 ,The present invention provides a technical solution: a key management system based on medical blockchain includes a status assessment module, a reverse tracking module, a forward-looking analysis module, an intelligent alarm threshold adjustment module, a resource priority scheduling module, a reliability analysis and fault tolerance module, and a process optimization and decision support module;

[0068] The status assessment module uses support vector machines and decision tree algorithms to conduct real-time security status assessment of the medical system based on security performance indicators and user blockchain key usage information. It uses deep learning algorithms to analyze user behavior, conduct risk assessment and classification, and adjust key access control policies to generate security status assessment results.

[0069] The reverse tracking module uses decision trees and cluster analysis algorithms to analyze the user's blockchain key usage and access paths based on the security status assessment results. It also uses support vector machines to analyze user behavior patterns, conduct reverse tracking, optimize security policies, and generate tracking and analysis results.

[0070] Based on the security status assessment results and the tracking and analysis results, the forward-looking analysis module applies the autoregressive integral moving average model and the long short-term memory network to perform time series analysis, predict key usage trends, formulate security enhancement measures, optimize resource allocation, and generate trend predictions and security enhancement measures;

[0071] The intelligent alarm threshold adjustment module uses isolation forest and deep learning networks to detect abnormal system behaviors based on trend prediction and security enhancement measures, dynamically adjusts the alarm threshold according to the detection results, and generates alarm threshold adjustment information;

[0072] The resource priority scheduling module uses the long short-term memory network and random forest algorithm to predict resource demand and access pressure based on the alarm threshold adjustment information. According to the prediction results, it adjusts resource allocation and processing priority, optimizes system response, and generates resource scheduling strategies;

[0073] The reliability analysis and fault tolerance module evaluates system reliability based on resource scheduling strategies through fault tree analysis and Monte Carlo simulation, formulates fault tolerance mechanisms, including redundant storage and automatic fault switching, and generates adaptive fault tolerance strategies;

[0074] The process optimization and decision support module optimizes the blockchain key management process based on an adaptive fault-tolerant strategy, applies genetic algorithms and particle swarm optimization, combines rule engines and machine learning models, executes security decisions and response strategies through decision trees and support vector machines, and generates optimized processes and decision information.

[0075] The security status assessment results include security level, potential risk points, and user behavior risk score. The tracking and analysis results include illegal use paths of blockchain keys, lists of potential security vulnerabilities, and abnormal user behavior patterns. Trend prediction and security enhancement measures include warning time periods, security monitoring enhancement areas, and resource reallocation plans. Alarm threshold adjustment information includes abnormal behavior detection sensitivity, alarm level standards, and alarm response strategies. Resource scheduling strategies include resource allocation priorities, resource protection measures, and key operation resource retention strategies. Adaptive fault-tolerant strategies include redundant key storage solutions, automatic fault switching mechanisms, and key component monitoring points. Optimization process and decision-making information include process execution efficiency improvement points, automated security decision-making rules, and automatic response processes for exception handling.

[0076] In the status assessment module, support vector machine and decision tree algorithms are used for refinement operations based on security performance indicators and user blockchain key usage information. The support vector machine algorithm classifies data by constructing one or more hyperplanes to evaluate the real-time security status of the medical system, while the decision tree algorithm identifies risk patterns in user behavior by learning decision rules from data. The deep learning algorithm analyzes user behavior, automatically learns and extracts data features through the network hierarchy, and performs risk assessment and classification. Not only does it accurately assess the current security status, but it also dynamically adjusts the key access control policy through intelligent methods, generates security status assessment results that include security levels, potential risk points, and user behavior risk scores, and provides comprehensive and dynamic security protection for the system.

[0077] In the reverse tracking module, the security status assessment results are used as input data, and the decision tree and cluster analysis algorithms are combined to conduct an in-depth analysis of the user's blockchain key usage. The decision tree algorithm is used to identify the regularities and abnormal behaviors in the use of keys, while the cluster analysis analyzes the key access path to find risk nodes and sensitive areas in the use process. The support vector machine algorithm analyzes user behavior patterns and accurately identifies abnormal usage and potential security vulnerabilities. This enables the system to conduct effective reverse tracking, optimize security policies, and generate tracking and analysis results including illegal use paths of blockchain keys, lists of potential security vulnerabilities, and abnormal user behavior patterns, significantly improving the security protection capabilities of the medical blockchain system.

[0078] The forward-looking analysis module uses the autoregressive integrated moving average model and long short-term memory network to perform time series analysis on the key usage trend through the security status assessment results and tracking and analysis results. The autoregressive integrated moving average model is used to analyze and predict trends and seasonal changes in time series, while the long short-term memory network can process and predict time dependencies in long series data. These two algorithms jointly predict the trend of future key usage and provide data support for the formulation of security enhancement measures and the optimization of resource allocation. Trend predictions and security enhancement measures including warning time periods, security monitoring enhancement areas, and resource reallocation plans are generated, which enhances the medical blockchain system's ability to predict and respond to future security threats.

[0079] The intelligent alarm threshold adjustment module uses trend prediction and security enhancement measures as the basis to detect anomalies in system behavior through isolation forest and deep learning network. The isolation forest algorithm identifies anomalies in the data set by randomly selecting features and cut points, which is suitable for identifying rare events and abnormal behaviors; the deep learning network improves the accuracy of anomaly detection by learning the deep features of the data. It enables the system to dynamically adjust the alarm threshold according to the abnormal behavior monitored in real time, and generates alarm threshold adjustment information including abnormal behavior detection sensitivity, alarm level standards and alarm response strategies, which greatly improves the security monitoring capability and response speed of the medical blockchain system.

[0080] The resource priority scheduling module uses the long short-term memory network and random forest algorithm to predict resource demand and access pressure based on the alarm threshold adjustment information. The long short-term memory network predicts future resource demand trends by learning historical data on resource usage, while the random forest algorithm analyzes the impact of different factors on resource demand to improve the accuracy of the prediction. Based on the prediction results, the system can adjust resource allocation and processing priorities, and generate a resource scheduling strategy that includes resource allocation priorities, resource safeguards, and key operation resource reservation strategies, effectively optimizing system response and resource utilization efficiency.

[0081] The reliability analysis and fault tolerance module uses resource scheduling strategies, fault tree analysis and Monte Carlo simulation to evaluate system reliability. Fault tree analysis identifies and evaluates the various causes of system failures, while Monte Carlo simulation simulates different failure situations through random sampling techniques to evaluate the behavior and performance of the system in the face of various random events. A fault tolerance mechanism including redundant storage and automatic fault switching was developed, and an adaptive fault tolerance strategy was generated, which greatly improved the reliability and stability of the medical blockchain system.

[0082] The process optimization and decision support module optimizes the blockchain key management process based on an adaptive fault-tolerant strategy, combined with genetic algorithms and particle swarm optimization. Genetic algorithms simulate natural selection and genetic mechanisms to optimize problem solutions, while particle swarm optimization simulates the social behavior of bird flocks to find the optimal solution. The optimization method combines rule engines, decision trees, and support vector machine algorithms to not only improve the efficiency of process execution, but also ensure the efficient operation and security of the system through automated security decision-making and response strategies, and generates optimized processes and decision information for process execution efficiency improvement points, automated security decision-making rules, and automatic response processes for exception handling.

[0083] See also Figure 2 and Figure 3 ,The state assessment module includes a security indicator analysis submodule, a user behavior assessment submodule, and a risk level determination submodule;

[0084] The security indicator analysis submodule uses the support vector machine algorithm based on security performance indicators and user blockchain key usage information to monitor and analyze the information security performance indicators of the medical system in real time, evaluate the current security status of the medical system, and generate security performance indicator analysis results;

[0085] The user behavior assessment submodule uses a decision tree algorithm based on the security performance index analysis results to analyze the user's behavior patterns, including key usage frequency and access patterns, identify potential risk behaviors and abnormal operations, and generate user behavior analysis results;

[0086] The risk level determination submodule conducts risk assessment and classification of user behaviors based on the user behavior analysis results through deep learning algorithms, adjusts the key access control strategy according to the assessment results, optimizes the security of the system, and generates security status assessment results.

[0087] In the security index analysis submodule, by collecting the security performance indicators of the medical system and the user blockchain key usage information, the support vector machine algorithm is used for real-time monitoring and analysis. The support vector machine algorithm is mainly used to solve classification and regression problems. It constructs one or more hyperplanes in high-dimensional space to achieve data classification or regression analysis. The support vector machine algorithm first extracts features from the security performance indicators and key usage information, and then uses the features as input to build a classification model to determine the security status of the system. It can effectively identify the security risks and potential risks in the system. The generated security performance indicator analysis results describe the current security status of the system in detail, including information such as security level, key risk indicators and their thresholds. For system administrators, this result is an important basis for judging the overall security status of the system and taking corresponding security measures.

[0088] In the user behavior assessment submodule, by analyzing the results of the security performance index analysis, the decision tree algorithm is used to conduct an in-depth analysis of the user's behavior pattern. The decision tree is a commonly used data mining algorithm. By creating a model, the instance is divided from the root to the leaf node, and each leaf node corresponds to a decision result. The algorithm will generate a series of rules to distinguish normal behavior from abnormal behavior based on the data features in the security performance index analysis results, such as key usage frequency and access mode. By evaluating user behavior, potential risk behaviors and abnormal operations can be identified. The generated user behavior analysis results list potential risk users, abnormal behavior patterns and related risk scores in detail. It is crucial for risk level determination and security policy adjustment, and improves the accuracy and response speed of the system to user behavior management.

[0089] In the risk level determination submodule, based on the user behavior analysis results, the user behavior is assessed and classified through the deep learning algorithm. By building a multi-layer neural network model, the deep learning algorithm can automatically learn the high-level features of the data and effectively identify and classify complex user behavior patterns. In this module, the deep learning algorithm first learns the characteristics of various behavior patterns in the user behavior analysis results, and then evaluates and classifies the risk level based on the learned characteristics. It not only improves the accuracy of risk identification, but also proposes specific response strategies for risk behaviors of different levels. The final security status assessment results provide a comprehensive risk management solution for the medical system, including risk levels, potential risk points, and targeted key access control policy adjustment suggestions. It significantly improves the ability of the medical blockchain system in risk prevention and control and emergency response, ensuring the security of medical data and the protection of user privacy.

[0090] See also Figure 2 and Figure 4 ,The reverse tracing module includes a key usage analysis submodule, an access path tracing submodule, and a security vulnerability identification submodule;

[0091] The key usage analysis submodule uses a decision tree algorithm based on the security status assessment results to analyze the user's blockchain key usage, identify the compliance and security of blockchain key usage, and generate key usage analysis results;

[0092] The access path tracking submodule uses a cluster analysis algorithm based on the key usage analysis results to track the blockchain key access path, identify the risk nodes and sensitive areas passed during the use of the blockchain key, and generate access path tracking results;

[0093] The security vulnerability identification submodule uses support vector machines based on the access path tracking results to analyze user behavior patterns, conduct reverse tracking, identify potential security vulnerabilities and improper operations, optimize security policies, and generate tracking and analysis results.

[0094] In the key usage analysis submodule, the decision tree algorithm is used to conduct a detailed analysis of the user's blockchain key usage through the security status assessment results. The decision tree algorithm can efficiently handle classification problems with its clear logical structure and easy-to-understand decision-making process. The algorithm takes the data in the security status assessment results as input, and the data includes information such as the user's key usage frequency, usage time, and usage scenarios. The decision tree constructs a tree model with this information. Each node represents a decision rule, and each branch represents a result of the decision, ultimately achieving the purpose of classification. It can identify non-compliant or high-security-risk key usage behaviors, and the generated key usage analysis results list compliance and security assessments in detail, providing a basis for risk management and policy adjustments.

[0095] In the access path tracking submodule, based on the key usage analysis results, a cluster analysis algorithm is used to track the blockchain key access path. The cluster analysis algorithm can group the samples in the data set into several categories, and the samples in each category have high similarity in certain attributes, thereby achieving natural classification of data. The algorithm takes the access path data in the key usage analysis results as input, and clusters them into different risk levels by calculating the similarity between different access paths. This analysis helps identify risk nodes and sensitive areas that may be passed through during the use of keys, and the generated access path tracking results provide important information for system security monitoring and risk prevention.

[0096] In the security vulnerability identification submodule, based on the access path tracing results, the support vector machine algorithm is used to conduct in-depth analysis of user behavior patterns to achieve reverse tracing and identification of security vulnerabilities. Support vector machine is a supervised learning method that is particularly suitable for high-dimensional data classification problems. It divides data into different categories by finding the optimal hyperplane. This module uses the support vector machine algorithm to process the access path tracing results, identify abnormal behaviors that are significantly different from normal behavior patterns, and then reveal potential security vulnerabilities and improper operations. Not only does it optimize security policies, but it also generates detailed tracking and analysis results, including a list of identified security vulnerabilities and recommended repair measures, which significantly improves the security protection capabilities of the medical blockchain system.

[0097] See also Figure 2 and Figure 5 ,The forward-looking analysis module includes a trend analysis submodule, a risk prediction submodule, and a strategy formulation submodule;

[0098] The trend analysis submodule uses the autoregressive integral moving average model and long short-term memory network to perform time series analysis based on the security status assessment results and the tracking and analysis results, analyze historical key usage data, identify usage trends and patterns, and perform trend forecasts to generate trend analysis results;

[0099] The risk prediction submodule uses deep learning models, including convolutional neural networks and recurrent neural networks, based on trend analysis results to analyze key usage trends, identify abnormal patterns and risk factors in trends, predict future security risks, and generate risk prediction results;

[0100] Based on the risk prediction results, the strategy formulation submodule uses multi-attribute decision-making methods and fuzzy set theory to optimize security policies and resource allocation, formulate responses to potential security threats, adjust resource allocation, optimize the security performance and efficiency of the system, and generate trend forecasts and security enhancement measures.

[0101] In the trend analysis submodule, the autoregressive integrated moving average model and long short-term memory network are used to perform time series analysis based on the security status assessment results and tracking and analysis results. The autoregressive integrated moving average model is specifically used to analyze and predict linear time series data. It identifies long-term trends and seasonal changes in data by modeling the autoregression, difference, and moving average of historical data. The long short-term memory network is a recurrent neural network that is suitable for processing and predicting nonlinear long-term dependencies in time series. In this module, the autoregressive integrated moving average model is used to analyze the trends and patterns of historical key usage data, and then the long short-term memory network is used to deeply learn the complex patterns and hidden features in the data for trend prediction. The generated trend analysis results can accurately reveal the dynamic changes in key usage and provide a scientific basis for subsequent risk prediction and security strategy formulation.

[0102] In the risk prediction submodule, deep learning models are used for analysis based on trend analysis results. The deep learning models here include convolutional neural networks and recurrent neural networks. Convolutional neural networks are good at processing data with strong spatial correlation and extracting local features in time series data through convolutional layers; recurrent neural networks can process time dependencies in sequence data and learn the dynamic changes of time series. The combination of these two models can identify abnormal patterns and risk factors in key usage trends from trend analysis results and predict future security risks. The generated risk prediction results not only describe potential security threats in detail, but also provide an important basis for formulating corresponding preventive measures.

[0103] In the strategy formulation submodule, based on the risk prediction results, the multi-attribute decision-making method and fuzzy set theory are used to optimize the security strategy and resource allocation. The multi-attribute decision-making method is a method of making decisions by comprehensively considering multiple decision attributes, which is suitable for solving complex decision-making problems with multiple goals and constraints. Fuzzy set theory introduces fuzzy concepts to deal with uncertainty and ambiguity in the decision-making process, making the decision results more in line with the actual situation. Through these two methods, the module can comprehensively consider various risk factors in the risk prediction results, scientifically formulate countermeasures for potential security threats, and reasonably adjust resource allocation to optimize the overall security performance and operation efficiency of the system. The generated trend prediction and security enhancement measures not only clarify the future security monitoring focus and resource allocation plan, but also provide strong support for the continuous security operation of the medical blockchain system.

[0104] See also Figure 2 and Figure 6 ,The intelligent alarm threshold adjustment module includes an abnormal behavior detection submodule, a threshold adjustment strategy submodule, and an alarm generation submodule;

[0105] The abnormal behavior detection submodule uses the isolation forest algorithm to detect abnormal behaviors based on trend prediction and security enhancement measures. It constructs a forest by randomly selecting features and randomly selecting cut values, identifies rare events and abnormal behaviors, and generates abnormal behavior detection results.

[0106] The threshold adjustment strategy submodule is based on the abnormal behavior detection results and adopts an adaptive learning algorithm. Through online learning and incremental learning technology, it dynamically updates the model according to the real-time data stream, matches the changes in behavior patterns, optimizes the alarm threshold, dynamically adjusts the threshold of the alarm system, and generates a threshold adjustment strategy;

[0107] The alarm generation submodule is based on the threshold adjustment strategy. When abnormal behavior that meets the adjusted threshold is detected, it uses the event-driven programming model to automatically generate alarm information, notify the system administrator to take corresponding security measures, and generate alarm threshold adjustment information.

[0108] In the abnormal behavior detection submodule, the behavior data in the system is analyzed by the isolation forest algorithm, aiming to identify and mark abnormal events in the behavior pattern. The isolation forest algorithm is particularly suitable for dealing with the anomaly detection problem of high-dimensional data. Its core idea is to randomly select features and segmentation values ​​to construct a decision tree, and build a "forest" through the decision tree. During the construction process, data points of normal behavior usually require more segmentation to be "isolated", while abnormal data is often isolated faster due to its rare nature. In this way, the algorithm can effectively distinguish between normal and abnormal behaviors. In this submodule, the input data format includes the user's key usage frequency, access time, access path, etc. After the data is processed by the isolation forest algorithm, a detailed report containing the identified abnormal behavior can be generated. It provides an accurate reference basis for subsequent threshold adjustment and security response measures, and significantly improves the system's response ability to potential security threats.

[0109] In the threshold adjustment strategy submodule, an adaptive learning algorithm is used to dynamically adjust the threshold based on the abnormal behavior detection results. The adaptive learning algorithm can update the model in real time according to the changes in the data stream to match the changes in the behavior pattern, thereby optimizing the threshold of the alarm system. The algorithm continuously adjusts the decision threshold through online learning and incremental learning techniques to ensure that the system maintains the best detection performance in a changing environment. The sensitivity and response level of the alarm can be dynamically adjusted based on the real-time monitored behavior data and the identified abnormal patterns, and a policy document containing the optimized alarm threshold can be generated. This document describes in detail the alarm threshold corresponding to each behavior pattern, providing system administrators with a flexible and efficient security management tool to ensure that the system can respond quickly and accurately to various potential risks.

[0110] In the alarm generation submodule, based on the threshold adjustment strategy, the event-driven programming model is used to automatically generate alarm information. The event-driven programming model is a programming paradigm that triggers corresponding operations when the system detects the occurrence of a predefined event. The "event" in the module refers to abnormal behavior that meets the adjusted threshold. When the system monitors behaviors that meet the alarm conditions in real time, it can automatically generate alarm information and promptly notify the system administrator of the information to prompt him to take corresponding security measures. It not only greatly reduces the need for manual intervention, but also ensures that security incidents can be responded to in the shortest time. The generated alarm threshold adjustment information includes detailed alarm levels, behavioral patterns that trigger alarms, recommended response measures, etc., providing system administrators with real-time and accurate security management guidance, significantly improving the security protection capabilities of the medical blockchain system.

[0111] See also Figure 2 and Figure 7 ,The resource priority scheduling module includes a resource demand prediction submodule, a priority adjustment submodule, and a resource allocation submodule;

[0112] The resource demand prediction submodule uses the long short-term memory network to analyze historical resource usage data and access patterns based on the alarm threshold adjustment information, predicts resource demand and access pressure, applies the random forest algorithm, analyzes the impact of multiple factors on resource demand, predicts the trend of future resource demand changes, and generates resource demand prediction results;

[0113] The priority adjustment submodule uses the priority queue algorithm based on the resource demand prediction results to analyze the urgency and importance of differentiated resource requests, adjust the priority of resource allocation and processing, and generate a priority adjustment strategy;

[0114] The resource allocation submodule uses a dynamic resource allocation algorithm based on the priority adjustment strategy and the minimum connection number algorithm to dynamically adjust the allocation of system resources. According to the current system status and priority strategy, it dynamically adjusts the allocation of computing resources, storage resources and network bandwidth, optimizes system response and processing capabilities, and generates resource scheduling strategies.

[0115] In the resource demand prediction submodule, the historical resource usage data and access patterns are comprehensively analyzed through the long short-term memory network and the random forest algorithm. The long short-term memory network, a special recurrent neural network, is suitable for processing and predicting long-term dependency problems in time series data. The long short-term memory network captures the dynamic trend of resource demand by learning the time series characteristics of historical resource usage data. The random forest algorithm is applied to analyze and identify the impact of different factors, such as user access frequency, system operation type, etc., on resource demand. The random forest constructs multiple decision trees and summarizes the prediction results to improve the accuracy and stability of the prediction. The resource demand prediction results are generated, which describe in detail the changing trend of resource demand in the future period, helping system administrators to plan resources reasonably and ensure efficient and stable operation of the system.

[0116] The priority adjustment submodule uses the priority queue algorithm to adjust the priority of resource allocation and processing based on the resource demand prediction results. The priority queue algorithm is an algorithm that effectively manages and schedules multiple tasks or requests, and assigns a priority to each task based on the urgency and importance of the task. The algorithm evaluates the urgency and importance of each resource request, and then dynamically adjusts the priority of resource allocation based on the evaluation results. It can ensure that critical tasks and high-priority tasks can obtain the required resources first, thereby optimizing the response speed and processing power of the entire system. The generated priority adjustment strategy provides the system with a flexible and efficient resource management solution, which significantly improves the system's ability to respond to resource demands in complex scenarios.

[0117] The resource allocation submodule uses the dynamic resource allocation algorithm and the minimum connection number algorithm to dynamically adjust system resources based on the priority adjustment strategy. The dynamic resource allocation algorithm adjusts the resource allocation strategy in real time according to the current state of the system and resource requirements to adapt to the actual needs of system operation. The minimum connection number algorithm focuses on optimizing the allocation of network resources by allocating new requests to the server with the least current connections to balance the load of each server. The algorithm comprehensively considers multiple resource types such as computing resources, storage resources, and network bandwidth, and dynamically optimizes resource allocation based on the priority adjustment strategy and the current system status. The generated resource scheduling strategy specifies the allocation plan for each type of resource in detail, effectively improving the processing efficiency and stability of the system, and ensuring that the system can meet the service needs of users when resources are limited.

[0118] See also Figure 2 and Figure 8 ,Reliability analysis and fault tolerance module includes reliability evaluation submodule, fault simulation submodule, and fault tolerance strategy design submodule;

[0119] The reliability assessment submodule uses the fault tree analysis method based on the resource scheduling strategy to assess the reliability of system components. It identifies key weak links and generates reliability assessment results by analyzing the factors and paths that lead to system failures.

[0120] The fault simulation submodule uses the Monte Carlo simulation method based on the reliability evaluation results to simulate the system performance under differentiated fault conditions, evaluate the effectiveness of the preset fault tolerance mechanism, and use random sampling technology to evaluate the system's behavior and performance when facing various random events, and generate fault simulation results;

[0121] The fault-tolerant strategy design submodule is based on the fault simulation results and uses RAID technology to store data in multiple disks to achieve redundant storage. It uses the heartbeat detection mechanism to achieve automatic fault switching and generate an adaptive fault-tolerant strategy.

[0122] In the reliability assessment submodule, the reliability of system components is assessed based on the resource scheduling strategy through the fault tree analysis method. Fault tree analysis is a top-down analysis method that displays various ways in which events lead to failure in the form of logic diagrams, thereby determining the failure modes and causes in the system. By collecting the operating data and resource scheduling strategies of system components, and then building a fault tree based on the information, all factors and paths that lead to system failures are identified. Logical operators such as "and", "or", and "not" are used to analyze the logical relationship between failure factors and identify the key weak links of the system. Not only does it reveal the potential causes of system failures, but it also evaluates the reliability of each component. The generated reliability assessment results list the weak links and failure probabilities of the system in detail, providing an important basis for subsequent fault prevention and system optimization.

[0123] In the fault simulation submodule, the Monte Carlo simulation method is used to simulate the system performance under different fault conditions based on the reliability assessment results. Monte Carlo simulation is a numerical calculation method based on random sampling, which estimates the system behavior by building a probability model and performing a large number of random sampling. In this submodule, according to the reliability assessment results, a variety of fault scenarios are set, and the system performance under various fault conditions is simulated by randomly generating system parameters under the scenarios. It can evaluate the system's response capability to different types of faults and the effectiveness of the fault-tolerant mechanism. The generated fault simulation results provide a comprehensive evaluation of the behavior and performance of various random events encountered by the system in actual operation, providing an empirical basis for the design and optimization of the system's fault-tolerant strategy.

[0124] In the fault-tolerant strategy design submodule, RAID technology and heartbeat detection mechanism are used to design an adaptive fault-tolerant strategy based on the fault simulation results. RAID technology achieves the purpose of redundant storage by distributing data on multiple disks, which can ensure the integrity and availability of data even if some disks fail. The heartbeat detection mechanism is used to monitor the status of system components. Once a component loses its response, the system automatically switches to the backup component to ensure the continuous operation of the system. According to the fault simulation results, the appropriate RAID level is selected and the heartbeat detection parameters are set to design a fault-tolerant strategy that can effectively utilize system resources and quickly respond to faults. The generated adaptive fault-tolerant strategy details the data storage and fault switching schemes, significantly improving the reliability and robustness of the system.

[0125] See also Figure 2 and Fig. 9 ,The process optimization and decision support module includes the process ,analysis submodule, the decision logic design submodule, and the automated ,execution submodule.

[0126] The process analysis submodule uses genetic algorithms based on adaptive fault-tolerant strategies to evaluate and analyze the blockchain key management process. By simulating the natural selection process, it identifies bottlenecks and inefficient links in the process and generates process analysis results.

[0127] The decision logic design submodule uses particle swarm optimization method based on the process analysis results to optimize the decision process by simulating the social behavior of bird flocks. It uses the rule engine to automatically perform process optimization and safety decisions according to preset rules to generate optimized decision logic.

[0128] The automated execution submodule is based on optimized decision logic. It selects decision paths according to differentiated situations through decision trees, uses support vector machines to classify and identify security threat patterns, automatically adjusts system configurations and response measures, and generates optimized processes and decision information.

[0129] In the process analysis submodule, the blockchain key management process is deeply evaluated and optimized through genetic algorithms. Genetic algorithm is a search algorithm that simulates the biological evolution process. It searches for the optimal solution in the solution space through operations such as selection, crossover, and mutation. By defining the process efficiency and effect as the fitness function, each link of the blockchain key management process is regarded as an individual, and each operation step and decision point in the process is regarded as the gene of the individual. By simulating the process of natural selection, the algorithm iteratively selects the process configuration with high fitness, and the crossover and mutation operations introduce new process variants to explore a better process configuration space. Until the optimal process configuration is found, the bottlenecks and inefficient links in the process are identified and improved. The generated process analysis results list the optimization points and recommended improvement measures in the process in detail, providing a scientific basis and direct guidance for the efficient operation of the system.

[0130] Based on the process analysis results, the decision logic design submodule uses the particle swarm optimization method to optimize the decision logic by simulating the social behavior of bird flocks. Particle swarm optimization is an optimization tool based on group collaboration. It regards each solution as a particle in the search space and guides the group to move to the optimal solution area through information sharing between particles. In this submodule, the logic configuration of each decision point is regarded as a particle, the initial decision configuration is set according to the process analysis results, and the decision logic is continuously adjusted using the information exchange mechanism between particles to find the best decision path. The rule engine is used to automatically perform process optimization and security decisions. The optimized decision logic finally generated not only takes into account process efficiency, but also ensures the security of the system, providing strong logical support for automated process execution.

[0131] The automated execution submodule automatically adjusts the system configuration and response measures based on the optimized decision logic through decision trees and support vector machines. As a simple and intuitive classification algorithm, the decision tree selects the decision path according to the rule tree defined in the optimized decision logic, and is suitable for handling decision problems with clear rules. The support vector machine is used to classify and identify security threat patterns. It realizes data classification in high-dimensional space by constructing one or more hyperplanes, and has high accuracy for complex security threat identification scenarios. Combining the advantages of decision trees and support vector machines, different security threats are automatically identified and system configurations and security response measures are quickly adjusted. The generated optimization process and decision information not only improve the system's automated management capabilities, but also enhance the system's security protection level, ensuring the stable and reliable operation of the medical blockchain system.

[0132] See also Fig.10 , a key management method based on medical blockchain, the key management method based on medical blockchain is executed based on the above-mentioned key management system based on medical blockchain, and includes the following steps:

[0133] S1: Based on security performance indicators and user blockchain key usage information, support vector machine algorithm and decision tree algorithm are used to analyze the current security status of the medical system and generate security status analysis results;

[0134] S2: Based on the security status analysis results, deep learning algorithms are applied to analyze user behaviors, identify potential risks and abnormal operations, and generate risk identification and user behavior analysis results;

[0135] S3: Based on the results of risk identification and user behavior analysis, combined with decision tree and cluster analysis algorithms, reverse tracing of key usage is performed to identify illegal usage paths and potential security vulnerabilities, and generate reverse tracing analysis results;

[0136] S4: Based on the results of reverse tracing analysis, the autoregressive integrated moving average model and long short-term memory network are used to predict the key usage trend, formulate security enhancement measures, and generate trend prediction results;

[0137] S5: Based on the trend prediction results, the isolation forest algorithm and deep learning network are used to detect abnormal behavior, and the alarm threshold is dynamically adjusted according to the detection results to generate an alarm threshold adjustment strategy;

[0138] S6: Based on the alarm threshold adjustment strategy, use the long short-term memory network and random forest algorithm to predict resource demand and access pressure, adjust resource allocation and processing priority, and generate resource scheduling optimization solutions;

[0139] S7: Based on the resource scheduling optimization scheme, the system reliability is evaluated through fault tree analysis and Monte Carlo simulation, and redundant storage and automatic fault switching mechanisms are formulated to generate adaptive fault tolerance strategies;

[0140] S8: Based on the adaptive fault-tolerant strategy, genetic algorithm and particle swarm optimization are applied to optimize the key management process. The rule engine, decision tree and support vector machine are combined to automatically execute security decisions and response strategies and generate optimized processes and decision information.

[0141] By combining the support vector machine algorithm and the decision tree algorithm, the current security status of the medical system can be accurately analyzed, providing a higher level of protection for medical data. This refined security status assessment can timely discover potential security threats to the system, thereby effectively preventing data leakage or illegal access and ensuring the confidentiality and integrity of patient information. The application of deep learning algorithms makes user behavior analysis more in-depth, not only identifying obvious risk behaviors, but also digging out hidden abnormal patterns, improving the system's ability to identify complex attack behaviors. Such a behavior analysis mechanism enables the medical blockchain system to identify and respond to potential security threats at an early stage, strengthening the security of medical data. The implementation of reverse tracing analysis, based on decision trees and clustering analysis algorithms, can not only track the illegal use path of keys, but also discover potential security vulnerabilities in the system. This reverse tracing mechanism provides the possibility of timely patching security vulnerabilities and preventing future security threats, significantly improving the security and stability of the medical blockchain system. The execution of trend prediction, through the autoregressive integral sliding average model and long short-term memory network, can not only predict the trend of key usage, but also formulate targeted security enhancement measures based on the trend. It can actively respond to future changes and potential risks, thereby maintaining the long-term security and reliability of the medical blockchain system. The abnormal behavior detection and dynamic adjustment of alarm thresholds through isolation forest algorithm and deep learning network enhance the system's ability to detect abnormal behavior, and can flexibly adjust the alarm level according to actual conditions, improving the system's security response efficiency and accuracy. The implementation of the resource scheduling optimization plan ensures the efficient utilization and deployment of system resources, ensures that the system can still operate stably in the face of high concurrent requests, and ensures the continuity and reliability of medical services. The formulation and implementation of adaptive fault-tolerant strategies enhance the system's tolerance for faults, ensure that the medical blockchain system can still maintain service continuity in the face of hardware failures or network problems, and reduce the potential risk of service interruption. Through genetic algorithms and particle swarm optimization, the optimization of key management processes, and the use of rule engines, decision trees and support vector machines to automatically execute security decisions and response strategies, significantly improve the system's automated management capabilities and operational efficiency, reduce the burden on human resources, and also ensure the security and stability of the medical blockchain system.

[0142] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A key management system based on medical blockchain, characterized by: The system includes a status assessment module, a reverse tracking module, a forward-looking analysis module, an intelligent alarm threshold adjustment module, a resource priority scheduling module, a reliability analysis and fault tolerance module, and a process optimization and decision support module; The status assessment module uses support vector machines and decision tree algorithms to conduct real-time security status assessment of the medical system based on security performance indicators and user blockchain key usage information. It uses deep learning algorithms to analyze user behavior, conduct risk assessment and classification, and adjust key access control policies to generate security status assessment results. The reverse tracking module uses decision trees and cluster analysis algorithms based on the security status assessment results to analyze the user's blockchain key usage and access path, and combines support vector machines to analyze user behavior patterns, conduct reverse tracking, optimize security policies, and generate tracking and analysis results; The forward-looking analysis module uses the autoregressive integral moving average model and long short-term memory network to perform time series analysis based on the security status assessment results and the tracking and analysis results, predict the key usage trend, formulate security enhancement measures, optimize resource allocation, and generate trend predictions and security enhancement measures; The intelligent alarm threshold adjustment module uses isolation forest and deep learning networks to detect abnormal system behavior based on trend prediction and security enhancement measures, dynamically adjusts the alarm threshold according to the detection results, and generates alarm threshold adjustment information; The resource priority scheduling module uses the long short-term memory network and random forest algorithm to predict resource demand and access pressure based on the alarm threshold adjustment information, adjusts resource allocation and processing priority according to the prediction results, optimizes system response, and generates resource scheduling strategies; The reliability analysis and fault tolerance module evaluates system reliability based on resource scheduling strategies, formulates fault tolerance mechanisms including redundant storage and automatic fault switching, and generates adaptive fault tolerance strategies through fault tree analysis and Monte Carlo simulation; The process optimization and decision support module is based on an adaptive fault-tolerant strategy, applies genetic algorithms and particle swarm optimization to optimize the blockchain key management process, combines rule engines and machine learning models, executes security decisions and response strategies through decision trees and support vector machines, and generates optimized processes and decision information.

2. The key management system based on medical blockchain according to claim 1 is characterized in that: The security status assessment results include security level, potential risk points, and user behavior risk scores. The tracking and analysis results specifically include illegal use paths of blockchain keys, lists of potential security vulnerabilities, and abnormal user behavior patterns. The trend prediction and security enhancement measures include warning time periods, security monitoring enhancement areas, and resource reallocation plans. The alarm threshold adjustment information includes abnormal behavior detection sensitivity, alarm level standards, and alarm response strategies. The resource scheduling strategy specifically includes resource allocation priorities, resource protection measures, and key operation resource retention strategies. The adaptive fault-tolerant strategy includes redundant key storage solutions, automatic fault switching mechanisms, and key component monitoring points. The optimization process and decision-making information specifically include process execution efficiency improvement points, automated security decision-making rules, and exception handling automatic response processes.

3. The key management system based on medical blockchain according to claim 1 is characterized in that: The status assessment module includes a security index analysis submodule, a user behavior assessment submodule, and a risk level determination submodule; The security index analysis submodule uses a support vector machine algorithm based on security performance indicators and user blockchain key usage information to monitor and analyze the information security performance indicators of the medical system in real time, evaluate the current security status of the medical system, and generate security performance indicator analysis results; The user behavior assessment submodule uses a decision tree algorithm based on the security performance index analysis results to analyze the user's behavior patterns, including key usage frequency and access patterns, identify potential risk behaviors and abnormal operations, and generate user behavior analysis results; The risk level determination submodule performs risk assessment and classification of user behavior based on the user behavior analysis results through a deep learning algorithm, adjusts the key access control strategy according to the assessment results, optimizes the security of the system, and generates a security status assessment result.

4. The key management system based on medical blockchain according to claim 1 is characterized in that: The reverse tracking module includes a key usage analysis submodule, an access path tracking submodule, and a security vulnerability identification submodule; The key usage analysis submodule analyzes the user's blockchain key usage based on the security status assessment result and uses a decision tree algorithm to identify the compliance and security of blockchain key usage and generate a key usage analysis result; The access path tracking submodule uses a cluster analysis algorithm based on the key usage analysis results to track the blockchain key access path, identify the risk nodes and sensitive areas passed through during the use of the blockchain key, and generate access path tracking results; The security vulnerability identification submodule uses a support vector machine based on the access path tracking results to analyze user behavior patterns, perform reverse tracking, identify potential security vulnerabilities and improper operations, optimize security policies, and generate tracking and analysis results.

5. The key management system based on medical blockchain according to claim 1 is characterized in that: The forward-looking analysis module includes a trend analysis submodule, a risk prediction submodule, and a strategy formulation submodule; The trend analysis submodule applies the autoregressive integral moving average model and the long short-term memory network based on the security status assessment results and the tracking and analysis results to perform time series analysis, analyze historical key usage data, identify usage trends and patterns, and perform trend prediction to generate trend analysis results; The risk prediction submodule uses deep learning models, including convolutional neural networks and recurrent neural networks, based on trend analysis results to analyze key usage trends, identify abnormal patterns and risk factors in trends, predict future security risks, and generate risk prediction results; Based on the risk prediction results, the strategy formulation submodule adopts multi-attribute decision-making method and fuzzy set theory to optimize security policies and resource allocation, formulate countermeasures to potential security threats, adjust resource allocation, optimize the security performance and efficiency of the system, and generate trend predictions and security enhancement measures.

6. The key management system based on medical blockchain according to claim 1 is characterized in that: The intelligent alarm threshold adjustment module includes an abnormal behavior detection submodule, a threshold adjustment strategy submodule, and an alarm generation submodule; The abnormal behavior detection submodule uses the isolation forest algorithm to detect abnormal behavior based on trend prediction and security enhancement measures, constructs a forest by randomly selecting features and randomly selecting cut values, identifies rare events and abnormal behaviors, and generates abnormal behavior detection results; The threshold adjustment strategy submodule adopts an adaptive learning algorithm based on the abnormal behavior detection results, dynamically updates the model according to the real-time data stream through online learning and incremental learning technology, matches the changes in behavior patterns, optimizes the alarm threshold, dynamically adjusts the threshold of the alarm system, and generates a threshold adjustment strategy; The alarm generation submodule is based on the threshold adjustment strategy. When abnormal behavior that meets the adjusted threshold is detected, the event-driven programming model is used to automatically generate alarm information, notify the system administrator to take corresponding safety measures, and generate alarm threshold adjustment information.

7. The key management system based on medical blockchain according to claim 1 is characterized in that: The resource priority scheduling module includes a resource demand prediction submodule, a priority adjustment submodule, and a resource allocation submodule; The resource demand prediction submodule uses a long short-term memory network to analyze historical resource usage data and access patterns based on alarm threshold adjustment information, predicts resource demand and access pressure, applies a random forest algorithm, analyzes the impact of multiple factors on resource demand, predicts future trends in resource demand, and generates resource demand prediction results; The priority adjustment submodule adopts a priority queue algorithm based on the resource demand prediction result, analyzes the urgency and importance of differentiated resource requests, adjusts the priority of resource allocation and processing, and generates a priority adjustment strategy; The resource allocation submodule uses a dynamic resource allocation algorithm based on a priority adjustment strategy to dynamically adjust the allocation of system resources through a minimum connection number algorithm. It dynamically adjusts the allocation of computing resources, storage resources, and network bandwidth according to the current system status and priority strategy, optimizes system response and processing capabilities, and generates a resource scheduling strategy.

8. The key management system based on medical blockchain according to claim 1 is characterized in that: The reliability analysis and fault tolerance module includes a reliability assessment submodule, a fault simulation submodule, and a fault tolerance strategy design submodule; The reliability assessment submodule uses the fault tree analysis method based on the resource scheduling strategy to assess the reliability of system components, identifies key weak links by analyzing the factors and paths leading to system failures, and generates reliability assessment results; The fault simulation submodule uses the Monte Carlo simulation method based on the reliability evaluation results to simulate the performance of the system under differentiated fault conditions, evaluate the effectiveness of the preset fault tolerance mechanism, and use random sampling technology to evaluate the behavior and performance of the system in the face of various random events, and generate fault simulation results; The fault-tolerant strategy design submodule uses RAID technology to disperse and store data on multiple disks based on fault simulation results to achieve redundant storage, and uses a heartbeat detection mechanism to achieve automatic fault switching and generate an adaptive fault-tolerant strategy.

9. The key management system based on medical blockchain according to claim 1 is characterized in that: The process optimization and decision support module includes a process analysis submodule, a decision logic design submodule, and an automated execution submodule. The process analysis submodule uses a genetic algorithm based on an adaptive fault-tolerant strategy to evaluate and analyze the blockchain key management process, identify bottlenecks and inefficient links in the process by simulating the natural selection process, and generate process analysis results; The decision logic design submodule uses a particle swarm optimization method based on the process analysis results to optimize the decision process by simulating the social behavior of bird flocks, and uses a rule engine to automatically perform process optimization and safety decisions according to preset rules to generate optimized decision logic; The automated execution submodule is based on optimized decision logic, selects decision paths according to differentiated scenarios through decision trees, uses support vector machines to classify and identify security threat patterns, automatically adjusts system configurations and response measures, and generates optimized processes and decision information.

10. A key management method based on medical blockchain, characterized in that: The key management system based on medical blockchain according to any one of claims 1 to 9 is implemented, comprising the following steps: Based on security performance indicators and user blockchain key usage information, support vector machine algorithm and decision tree algorithm are used to analyze the current security status of the medical system and generate security status analysis results; Based on the security status analysis results, deep learning algorithms are applied to analyze user behaviors, identify potential risks and abnormal operations, and generate risk identification and user behavior analysis results; Based on the risk identification and user behavior analysis results, combined with decision tree and cluster analysis algorithms, reverse tracing of key usage is performed to identify illegal usage paths and potential security vulnerabilities, and generate reverse tracing analysis results; Based on the reverse tracing analysis results, the autoregressive integrated moving average model and the long short-term memory network are used to predict the key usage trend, formulate security enhancement measures, and generate trend prediction results; Based on the trend prediction results, an isolation forest algorithm and a deep learning network are used to detect abnormal behavior, and the alarm threshold is dynamically adjusted according to the detection results to generate an alarm threshold adjustment strategy; Based on the alarm threshold adjustment strategy, long short-term memory network and random forest algorithm are used to predict resource demand and access pressure, adjust resource allocation and processing priority, and generate resource scheduling optimization solutions; Based on the resource scheduling optimization scheme, the system reliability is evaluated through fault tree analysis and Monte Carlo simulation, and redundant storage and automatic fault switching mechanisms are formulated to generate adaptive fault-tolerant strategies; Based on the adaptive fault-tolerant strategy, genetic algorithm and particle swarm optimization are applied to optimize the key management process. The rule engine, decision tree and support vector machine are combined to automatically execute security decisions and response strategies and generate optimized processes and decision information.

Images