Lightweight process management method

Abstract

The invention relates to the technical field of process management, and discloses a lightweight process management method, which comprises the following steps of: defining a business process and analyzing the business process through a process engine; a graphical process design interface is provided, and a user can dynamically configure business process nodes through dragging and clicking operations; the business process is disassembled into a plurality of task nodes, and each task node is packaged into an independent function module through modularization design; the task nodes are combined according to needs and executed in parallel on different processors, and efficient execution of the task nodes is ensured; the execution state of the business process is monitored in real time, and potential problems are found and processed in time through log recording and an exception processing mechanism. According to the technical scheme of supporting dynamic adjustment of the task nodes and optimizing parameters based on the historical abnormal mode, execution parameters of the task nodes are dynamically corrected, association logic between the nodes is adjusted in real time, and the effects that process management is flexible and adapts to complex service scenes, and operation stability is improved are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of process management, and in particular to a lightweight process management method. Background Art

[0002] With the rapid development of Internet technology and the expansion of enterprise scale, the demand for business process automation is growing. Traditional workflow management systems are often complex in function and huge in architecture, which not only increases the difficulty of development and maintenance, but also makes it difficult to quickly adapt to changing business needs. Lightweight process management design is to use simple, flexible and efficient process design and execution methods when conducting process management, avoiding unnecessary links and cumbersome approvals, so as to better adapt to the rapidly changing environment, and improve the organization's response speed and operational efficiency by optimizing and streamlining business processes. When implementing lightweight process management design, it is necessary to follow the principles of user-centricity, efficiency and simplicity, flexibility and scalability, automation and intelligence, and design and implement according to the steps of demand analysis, process decomposition, process optimization, automation design, testing and feedback.

[0003] Current process management systems usually rely on preset rules or fixed parameters to identify and handle anomalies in the process, which is effective in simple business scenarios. However, in complex or dynamically changing task environments, the rules may not cover all abnormal situations, resulting in anomalies that are difficult to detect in a timely manner or misclassified. In addition, existing systems usually rely on manual intervention to handle identified anomalies, which not only delays the timeliness of problem solving, but also easily leads to process interruptions or inconsistent processing results due to human factors.

[0004] In traditional process management systems, the configuration of process nodes is usually static, and the execution parameters of the nodes are determined during process design. During operation, they cannot be adjusted according to actual business needs or abnormal situations. This fixed node configuration method is extremely passive in scenarios where business needs change rapidly or abnormalities occur frequently, resulting in the system being unable to flexibly respond to the dynamic adjustment needs of the business. At the same time, the linkage relationship and resource allocation between nodes are usually preset and cannot be optimized in real time according to actual operating conditions.

[0005] In view of the deficiencies of the prior art, the present invention provides a lightweight process management method. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a lightweight process management method, which solves the problem that in traditional process management systems, the configuration of process nodes is usually static, the execution parameters of the nodes are determined during process design, and cannot be adjusted according to actual business needs or abnormal situations during operation.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A lightweight process management method comprises the following steps: Define business processes and parse the business processes through a process engine; Provides a graphical process design interface, where users can dynamically configure business process nodes by dragging and clicking; The business process is decomposed into multiple task nodes, and each task node is encapsulated as an independent functional module through component design; the task nodes are combined as needed and executed in parallel on different processors to ensure efficient execution of the task nodes; Monitor the execution status of business processes in real time, and promptly discover and handle potential problems through logging and exception handling mechanisms; conduct intelligent analysis based on historical execution data, automatically optimize the execution order, resource allocation and load balancing of task nodes, and improve system efficiency.

[0008] Furthermore, the business process is defined and parsed through the process engine. The process engine can accurately identify each node of the process and perform efficient parsing to ensure the accuracy and stability of process execution. A graphical process design interface is provided, and users can dynamically configure business process nodes by dragging and clicking. The graphical interface has an intuitive visual design, and users can flexibly adjust the node order, properties and connection relationships in the interface to achieve rapid construction and optimization of business processes. The business process is decomposed into multiple task nodes, and each task node is encapsulated as an independent functional module through componentized design. The modular design of task nodes supports independent development, debugging and reuse, ensuring high cohesion and low coupling of nodes, facilitating the expansion and maintenance of subsequent processes, and combining the task nodes as needed and running on different processors Parallel execution improves task execution efficiency through multi-threading or distributed computing technology, ensures that task nodes can complete resource-intensive and computing-intensive operations in a short time, monitors the execution status of business processes in real time, and promptly discovers and handles potential problems through logging and exception handling mechanisms. The system can record each execution link of the process and automatically alert, record and repair abnormal situations. It also supports the generation of detailed log reports for subsequent analysis and tracing, and performs intelligent analysis based on historical execution data to automatically optimize the execution order, resource allocation and load balancing of task nodes to improve system efficiency. The system analyzes historical process data through machine learning and data mining technology, dynamically predicts execution bottlenecks, and reasonably allocates system resources to ensure efficient and stable process execution and meet the continuous changes in business needs.

[0009] Preferably, role-based dynamic permission configuration is provided, and users can flexibly set the approver type, approval method and access rights according to the requirements of the node, ensuring that the permission control of different users meets the process requirements. Role-based dynamic permission configuration is provided, and users can flexibly set the approver type, approval method and access rights according to the requirements of the node, ensuring that the permission control of different users meets the process requirements; The deployment, initiation and approval of the process are implemented through program code and can support rapid integration into existing business systems.

[0010] Furthermore, role-based dynamic permission configuration is provided. Users can flexibly set the approver type, approval method and access rights according to the requirements of the node. The system supports dynamic allocation of permissions according to the needs of different nodes in the business process, including specifying approvers as specific roles, user groups or individuals, ensuring that each node can accurately control the scope of operator permissions during execution to avoid unauthorized operations or permission vacancies. The approval method can be flexibly set to single-person approval, multi-person approval, sequential approval or parallel approval, and supports dynamic adjustment of approval rules according to the complexity of the business, so as to achieve refined management and efficient execution of the approval process. Access permission configuration supports role-based control and real-time permission adjustment. The system can update access permissions in real time according to user roles, task assignments and node configurations to ensure that users have the necessary operating permissions in each node, while strictly restricting the access and operating behaviors of irrelevant personnel to improve system security and compliance. The deployment, initiation and approval of the process are implemented through program code. The system provides standardized interfaces and process engines, supports business codes to quickly call deployment process templates, and completes the initialization and initiation of processes through automated programs. Users can customize deployment rules according to business needs to achieve efficient execution and automated operation of the process. The system can be quickly integrated with existing business systems, and realize data interoperability and process collaboration with external systems through standard API interfaces, ensuring seamless connection between the process management system and the company's existing information platform, further improving the overall operating efficiency and data consistency of business processes.

[0011] Preferably, the process engine supports multi-version management, and users can switch between different versions of process templates according to business needs to ensure the flexibility and scalability of the process.

[0012] Furthermore, the process engine supports multi-version management. Users can switch between different versions of process templates according to business needs. The system provides version control functions, allowing users to manage and track different versions during process design, testing and release stages, ensuring that historical versions coexist and run independently with the current version, avoiding abnormal process execution or data loss due to version changes. Users can view detailed information of each version of the process through the system interface, including version number, release time, modified content and associated business nodes, to facilitate version comparison and rollback operations. When business needs change, users can quickly switch to the adapted new version of the process template to ensure the continuous operation and stability of the business process. The process engine provides automated tools to support process migration and data synchronization during version updates, ensuring the integrity and consistency of business data during process upgrades. Users can achieve a smooth transition without interrupting the currently running process. The system also supports multi-version parallel operation based on business scenarios. Administrators can specify different departments or tasks to use different versions of process templates according to specific business needs, to achieve personalized configuration and refined control of process management, and to ensure that the flexibility and scalability of the process can adapt to the ever-changing development needs of the business.

[0013] Preferably, multiple approval methods are supported, including single-person approval, multi-person approval, sequential approval, and parallel approval, and the approval method can be dynamically adjusted during the process execution to meet the needs of different business scenarios.

[0014] Furthermore, there are multiple approval methods, including single-person approval, multi-person approval, sequential approval, and parallel approval. The system allows users to flexibly set approval methods during the process configuration stage to meet the differentiated needs of different business processes for approval links. At the same time, it supports dynamic adjustment of approval methods during process execution to ensure that the process can flexibly respond to actual business changes. Single-person approval is suitable for simple business scenarios, and the designated approver can quickly complete the approval operation to improve process execution efficiency; multi-person approval can be set to multi-person joint approval or voting approval mode to ensure the accuracy of approval and democracy of decision-making in complex business scenarios. Sequential approval supports approval one by one in the preset node order. After each node completes approval, it automatically flows to the next node. It is suitable for business scenarios with clear levels and processes. Business needs with fixed steps; parallel approval allows multiple approval nodes to perform approval operations at the same time, reducing approval waiting time and improving the overall efficiency of process execution. It is suitable for business scenarios that require fast processing or parallel evaluation. The system provides dynamic adjustment of approval rules. The approval method can be flexibly changed according to actual business needs during the process execution. The administrator or the system can trigger automatic switching of the approval method according to preset conditions to ensure the continuity and efficiency of the process. The system also supports complete tracking of approval history records. Users can view the execution status, approvers and decision results of each approval node in real time, provide data support for process monitoring and optimization, ensure the transparency and controllability of process execution, and fully meet the diverse needs of approval processes in different business scenarios.

[0015] Preferably, task nodes can be added or deleted dynamically, and support seamless insertion of new task nodes during process execution, ensuring continuous execution and flexible adjustment of the process.

[0016] Furthermore, task nodes can be added or deleted dynamically. The system supports users to flexibly adjust task nodes during process design and execution according to business needs, ensuring that the real-time variability of the process structure is highly matched with the business scenario. Through the visual interface or interface operation, users can quickly insert new task nodes at any position in the process. The insertion of nodes will not affect the executed part of the process, ensuring the continuous and stable operation of the process. The dynamic deletion function allows users to remove task nodes that are no longer needed. The system will automatically adjust the node connection and execution logic of the process to ensure the integrity and continuity of the process and avoid process interruption or execution errors due to node changes. The addition and deletion of task nodes supports seamless operation. The system can be hot-swapped during process execution, ensuring that the business process can achieve flexible expansion and adjustment of functions without interrupting execution, while ensuring that the task after the node is inserted can take effect immediately and participate in the execution of the current process. The task node configuration supports a variety of functional parameters, including node name, task type, execution priority, resource allocation and trigger conditions. Users can make personalized adjustments based on actual business needs to meet the diverse requirements of complex business processes for node functions. The system also provides version recording and backtracking functions for node changes. Users can track the dynamic adjustment history of task nodes and roll back to the previous process configuration when necessary to ensure the controllability and security of process design and execution.

[0017] Preferably, the exception handling mechanism includes automatically repairing errors in task execution and notifying relevant users, analyzing abnormal patterns through machine learning algorithms and adjusting the execution parameters of task nodes to avoid similar errors in the future.

[0018] Furthermore, the exception handling mechanism includes automatically repairing errors in task execution and notifying relevant users. The system can monitor the task status in real time during the execution of the task node. When an error or abnormal situation is detected, the exception handling mechanism is automatically triggered to quickly locate and repair the error through preset rules or custom scripts to ensure that the process can continue to execute without being affected by the abnormality. The system also has an intelligent notification function. When the automatic repair cannot solve the problem, the abnormal information will be pushed to the relevant users or managers in time, and the user will be reminded to intervene manually through various methods such as email, SMS, system notification, etc. to ensure that the problem is solved in time and avoid affecting the overall process progress. The abnormal pattern is analyzed by machine learning algorithm, and the system The system can conduct in-depth learning and analysis of historical error data, identify the patterns and root causes of anomalies, generate an exception handling model and optimize the processing logic, further improving the system's automated error repair capabilities. The system will dynamically adjust the execution parameters of task nodes based on the analysis results, including task execution order, resource allocation, trigger conditions and timeouts, optimize the operating logic of task nodes, reduce the probability of anomalies and improve the stability of task execution. The exception handling mechanism also provides detailed logging functions to record the time, node location, error type and processing results of the exception, and generate visual reports for users to analyze and trace back, helping users to understand the weak links in system operation and continuously optimize business processes.

[0019] The formula of Gaussian mixture model for anomaly detection is: Probability density function of Gaussian mixture model: Where: p(x) represents the probability density of data point x, which measures the possibility that x belongs to the training data distribution and is used to determine whether it is abnormal. When p(x) is less than a certain threshold (such as ∈), x is considered to be abnormal; Represents the execution parameters or feature vector of the current task node, such as task execution time, resource consumption, error rate, etc., with a dimension of d.

[0020] K is the number of mixture components of Gaussian distributions, that is, the data is considered to come from K different Gaussian distributions.

[0021] π k The mixing coefficient of the k-th Gaussian distribution represents the weight of each distribution and satisfies It reflects the proportion of the kth Gaussian distribution in the overall data distribution.

[0022] The mean vector of the kth Gaussian distribution represents the center of the distribution and reflects the characteristics of normal operating parameters.

[0023] Preferably, business processes are defined through a no-code / low-code platform, and users can quickly build and adjust processes through a graphical interface. The platform provides intelligent prompts to avoid configuration errors and improve process building efficiency.

[0024] Furthermore, by defining business processes through a codeless / low-code platform, users can quickly build and adjust processes through a graphical interface. The system provides intuitive operation methods such as dragging, clicking, and connecting, so that users can easily design business processes without writing complex codes, and quickly build and adjust configurations such as process nodes, approval rules, and trigger conditions, greatly reducing the technical threshold for process design. The platform supports a variety of visualization tools, including flowcharts, state diagrams, node trees, etc. Users can preview the overall structure and execution logic of the process design in real time to ensure that the process is clear, intuitive, and easy to manage. The platform has a built-in intelligent prompt function. When users configure nodes or rules, it can automatically detect potential common problems such as logical errors, duplicate nodes, and unconnected nodes, and provide optimization suggestions to help users correct problems in time during the construction process to avoid affecting the normal operation of the process due to configuration errors. The system supports the process template library function and provides industry-wide standard process templates. Users can quickly start process design based on templates. At the same time, it supports custom saving and sharing of templates to facilitate the reuse and standardized management of processes. The platform has a process version management function. Users can save multiple historical versions when adjusting process configurations. It supports one-click rollback and comparison operations to ensure the security and controllability of process adjustments. The no-code / low-code platform also provides real-time testing and simulation execution functions. Users can perform simulation runs before releasing the process to verify the logical integrity and execution effect of the process, and ensure that the process runs efficiently in actual deployment. The platform greatly improves the efficiency of process construction and shortens the business process launch cycle through drag-and-drop interaction and intelligent auxiliary tools. At the same time, it reduces technical dependence, ensures that both business personnel and developers can efficiently participate in process design and optimization, and comprehensively improve the flexibility and operability of process management.

[0025] Preferably, the process management system supports seamless integration with external systems, realizes data interaction through standardized interfaces, and ensures collaborative operation and data consistency between business systems.

[0026] Furthermore, the process management system supports seamless integration with external systems and realizes data interaction through standardized interfaces. The system provides a variety of data interaction methods such as rich API interfaces, Web services, message queues, etc., to ensure rapid docking with third-party business systems such as ERP, CRM, OA, and financial management systems, and realize cross-system data synchronization, process linkage, and business collaborative operation. The process management system can support a variety of protocol standards, including HTTP, HTTPS, SOAP, RESTful, etc., to ensure secure and reliable data transmission between systems, and support two-way data communication to ensure that data from external systems can be imported into the process management platform in real time, and the process execution results can also be synchronously transmitted back to the external system to achieve closed-loop management of data. The system integration process supports dynamic mapping and data conversion functions, which can automatically adapt and map according to the business data structure of the external system, solve the problem of inconsistent data formats between different systems, and ensure the accuracy and consistency of data exchange. The system provides a security authentication mechanism, supports multiple authentication methods based on Token, OAuth, digital signatures, etc., to ensure access security and operational compliance during data interaction, and supports encrypted transmission to avoid data leakage or tampering risks. The process management system also supports event linkage and triggering mechanisms with external systems. It automatically starts process tasks or nodes by monitoring the operations or status changes of external systems, further improving the intelligence and automation of process execution. During the integration process, the system provides visual integrated monitoring functions, allowing users to view the success rate, error information and log records of data interaction in real time, facilitating rapid location and repair of problems, and ensuring the stability and continuity of system operation. Through seamless integration with external systems, the process management system can break through the data barriers of various business platforms of the enterprise, realize cross-system data sharing and business collaboration, improve the overall efficiency and information level of enterprise process management, and ensure collaborative operation and data consistency between business systems.

[0027] Preferably, process monitoring and logging functions are provided to track process execution status in real time, record task execution information and exception handling in detail, and support real-time synchronization of log data to the cloud for multi-device access and analysis.

[0028] Furthermore, it provides process monitoring and log recording functions. The system can track the process execution status in real time, accurately display the current status of each task node, execution time, approver information and task completion status, ensure that users can fully grasp the execution dynamics of the process and quickly discover potential problems, and record task execution information in detail, including task initiation time, node running time, resource consumption, execution results and related data flow, to provide a detailed basis for process operation optimization and problem troubleshooting. The log records of exception handling cover exception types, trigger conditions, automatic repair conditions, manual processing records and final solutions. The system can generate visual reports to display exception frequency, location distribution and trend changes, to help users analyze exception patterns and formulate improvement strategies. Log data supports real-time synchronization to the cloud. The system can ensure data security through encrypted transmission and support multiple devices. Access, users can view process logs and monitoring data anytime and anywhere, which is convenient for mobile office and remote management. The log function provides search and filtering tools. Users can quickly retrieve required information according to time range, node name, exception type and other conditions to improve the efficiency of log analysis. The system also supports automatic archiving and classified storage of log data, which is convenient for long-term preservation and historical backtracking. At the same time, it provides log export functions in multiple formats to meet the needs of external auditing and data sharing. Through intelligent log analysis, the system can use machine learning algorithms to mine bottleneck nodes and high-frequency anomalies in process execution, provide accurate data support for process optimization, and further improve the execution efficiency and stability of the process. The process monitoring and logging functions provided not only realize the full visualization and efficient management of process operation, but also provide strong guarantees for the company's compliance, problem-solving and process optimization capabilities.

[0029] Preferably, access control based on security authentication and authorization mechanisms is provided, multi-factor authentication is supported, and access rights are dynamically adjusted according to user behavior to ensure the security and compliance of the process execution.

[0030] Furthermore, advanced authentication and permission management strategies are adopted to ensure that only authorized users can access specific resources and perform corresponding operations, preventing unauthorized access and potential security threats. Multi-factor authentication (MFA) is supported, which requires users to provide multiple independent verification information (such as passwords, SMS verification codes, biometrics, etc.) to confirm their identity, thereby significantly improving the security of the system and reducing the risk caused by the leakage of a single credential. The access rights are dynamically adjusted according to the user's behavior. The system monitors the user's operation behavior and access mode in real time, and uses behavioral analysis algorithms to detect abnormal activities, such as abnormal login locations, abnormal times or abnormal operation frequencies. Once suspicious behavior is detected, the system will automatically adjust or limit the user's access rights to prevent potential security threats and ensure the security and compliance of the process execution process. The system complies with industry security standards and regulatory requirements, implements strict security policies and access control measures, ensures the integrity, confidentiality and availability of data during the execution of business processes, meets compliance requirements, and provides an audit log function to record the user's access and operation history for security audits and problem tracing. By comprehensively using security mechanisms such as multi-factor authentication, dynamic permission adjustment and behavior monitoring, the system can effectively prevent unauthorized access and potential security threats, ensuring the secure execution of business processes and the reliability of data.

[0031] The present invention provides a lightweight process management method, which has the following beneficial effects: 1. The present invention adopts a technical solution that supports dynamic adjustment of task nodes and optimizes parameters based on historical abnormal patterns. By dynamically correcting the execution parameters of task nodes and adjusting the association logic between nodes in real time, the technical effect of flexible adaptation of process management to complex business scenarios and improved operational stability is achieved. Compared with the technical solution in the prior art in which process node parameters are fixed and the ability to dynamically adjust the nodes in execution is lacking, the problem that task nodes cannot be flexibly configured with business changes and abnormal states may cause system interruptions is solved, ensuring the continuous and stable operation of the process management system in changing business scenarios.

[0032] 2. The present invention adopts an anomaly detection and repair technical solution based on Gaussian mixture model and machine learning algorithm, and achieves the technical effect of accurate identification and automatic repair of task anomalies by modeling and dynamically adjusting the execution parameters of task nodes. Compared with the technical solutions in the prior art that rely on manual investigation of anomalies or handle anomalies based on fixed rules, it solves the shortcomings of traditional anomaly handling methods that rely on human experience, have poor real-time performance, and cannot adapt to complex business scenarios, and significantly improves the system's automated processing capabilities and intelligence level.

[0033] 3. The present invention uses a machine learning algorithm to analyze historical abnormal patterns and optimize the resource allocation of task nodes, achieving the technical effect of intelligent optimization of process node execution efficiency and resource load balancing. Compared with the technical solution in the prior art where resource allocation is static and difficult to handle high-concurrency scenarios, it solves the problems of low task execution efficiency and resource waste in traditional systems, further improves the overall operating efficiency of the system in highly complex business scenarios, and ensures the efficiency and reliability of process operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of the present invention; Figure 2 Editing schematic diagrams for the process of the present invention; Figure 3 The schematic diagram of the process approval of the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the specification 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.

[0036] Please see attached Figure 1 - Attachment Figure 3 , an embodiment of the present invention provides a lightweight process management method, comprising the following steps: Define business processes and parse the business processes through a process engine; Provides a graphical process design interface, where users can dynamically configure business process nodes by dragging and clicking; The business process is decomposed into multiple task nodes, and each task node is encapsulated as an independent functional module through component design; the task nodes are combined as needed and executed in parallel on different processors to ensure efficient execution of the task nodes; Monitor the execution status of business processes in real time, and promptly discover and handle potential problems through logging and exception handling mechanisms; conduct intelligent analysis based on historical execution data, automatically optimize the execution order, resource allocation and load balancing of task nodes, and improve system efficiency.

[0037] In the specific implementation process, the process management method of the present invention includes the following core technical links: anomaly detection of task nodes, dynamic parameter adjustment and overall process optimization. The system detects anomalies in real time, performs automatic repair, and dynamically adjusts the node logic and resource allocation in the process through comprehensive analysis of the task node operating parameters and historical data, thereby realizing efficient operation and intelligent management of the process.

[0038] Firstly, the present invention uses anomaly detection technology based on Gaussian mixture model to monitor and evaluate the running status of process task nodes in real time. The running parameters of each task node (such as execution time, resource occupancy rate, error rate, etc.) are regarded as multidimensional feature vectors. By constructing a Gaussian mixture model (GMM), historical running data are clustered into multiple normal mode distributions, and the running area of ​​normal task nodes is defined. When the probability density of the real-time monitoring data point is lower than the set threshold, the task node is judged to be in an abnormal state. The Gaussian mixture model can dynamically adapt to a variety of running modes in a complex task environment by optimizing its parameters (mean, covariance matrix and mixing coefficient), thereby improving the accuracy of anomaly detection.

[0039] After detecting an anomaly, the system of the present invention automatically performs parameter adjustment of the task node. Specifically, the anomaly detection module provides the gradient information of the deviation of the anomaly point from the normal distribution, and the system uses an optimization algorithm (such as the gradient descent method) to dynamically adjust the parameters of the task node and gradually correct it to the normal operating area. The adjustment formula is as follows: Among them, x new is the adjusted task parameter, -logp(x) represents the abnormal loss function of the current parameter, is the gradient of the parameter adjustment direction, and η is the learning rate, which is used to control the adjustment step size. Through this method, the system can quickly repair abnormal task nodes without affecting the continuity of the process, thereby ensuring the normal execution of the process.

[0040] In addition, the present invention continuously optimizes the configuration of task nodes and process logic through cluster analysis and pattern mining of historical abnormal data. The abnormality detection module will generate a large amount of abnormal pattern data during operation, including information such as the frequency of abnormal occurrence, distribution characteristics of task nodes, and repair effects. Through clustering algorithms (such as K-Means or DBSCAN), the abnormal patterns are classified into different types. The system further optimizes the task node configuration of the process according to these patterns. For example, for frequently occurring resource bottlenecks, the system can dynamically adjust the resource allocation parameters or execution priority of the node to ensure efficient execution of tasks in high concurrency scenarios.

[0041] In terms of process optimization, the system of the present invention adopts a dynamic resource scheduling algorithm to adjust the resource allocation strategy in real time according to the operating data and historical performance of the task nodes. The system introduces a multi-objective optimization method, combines the completion time, resource utilization and abnormality rate of the task nodes, and constructs a comprehensive optimization objective function. The optimization process is iteratively solved through a genetic algorithm or a particle swarm algorithm to obtain the best resource allocation plan. In the process of multiple task nodes, the system dynamically adjusts the task execution order and parallelism according to task dependencies and resource requirements, thereby achieving efficient scheduling and execution of the overall process.

[0042] At the same time, the system of the present invention provides detailed process monitoring and log recording functions during the execution process. The system records each execution of the task node in real time, including the task start time, execution status, exception information and repair results, etc. The log recording data supports cloud synchronization. Users can intuitively view the execution status of the process and the exception handling results through the system interface. The system also provides a visual report function for exception analysis. Users can understand the exception distribution, repair efficiency and overall process optimization effect in the form of charts.

[0043] The present invention realizes the full process intelligence and efficiency of process management through the above technical means. Specifically, the system can not only automatically detect and repair anomalies in task nodes, but also continuously optimize process configuration and resource scheduling through data analysis to ensure the efficient operation of processes in complex business environments. The technical content disclosed in the present invention is sufficient and detailed, which enables technical personnel in this field to reproduce the present invention without creative work and apply it to actual process management scenarios.

[0044] First of all, the present invention relates to a process management method based on a machine learning algorithm, and in particular, through the comprehensive application of Gaussian mixture models, anomaly detection and dynamic adjustment mechanisms, real-time identification and repair of anomalies, optimization and adjustment of task nodes, and improvement of system stability are achieved in process management. In the overall technical implementation process, step S1 plays a key role in anomaly detection and real-time monitoring of task status, and provides basic data support for subsequent task repair and resource optimization. Specifically, the core of step S1 is based on feature extraction and analysis of task execution data, constructing an anomaly detection model through machine learning methods, dynamically judging the task status, and triggering subsequent process optimization operations in combination with the actual operating environment.

[0045] In this embodiment, step S1 models the execution parameters of the process task nodes and performs real-time anomaly detection through a Gaussian mixture model (GMM). The execution parameters of the task nodes include but are not limited to execution time, resource occupancy rate, error rate and other features. These parameters constitute a multidimensional feature vector of the task nodes. As an option, the system uses historical task execution data to train a Gaussian mixture model to fit the normal distribution of the feature vector x as a mixture of K Gaussian distributions, each with mean μ k , covariance matrix Σ k and the mixing coefficient π k Characterize, for example, the probability density function of the Gaussian mixture model is expressed as: In one possible implementation, the running status of the task node is determined by calculating the probability density p(x) of x in the Gaussian mixture model. When p(x) is less than the set threshold ∈, the task node is considered to be abnormal. It should be noted that the selection of the threshold ∈ can be dynamically adjusted according to the business scenario and abnormality tolerance. For example, a lower threshold can be selected in a critical task node to detect anomalies with higher sensitivity, while the threshold can be appropriately relaxed in a low-priority task.

[0046] Specifically, the anomaly detection process of task nodes is divided into a training phase and a detection phase. In the training phase, the system clusters historical task data through unsupervised learning and uses the maximum likelihood estimation method to optimize the parameters of the Gaussian mixture model, including the mean μ k , covariance matrix Σ k and the mixing coefficient π k ,In the detection phase, the real-time task data is input into the ,Gaussian mixture model, and its probability density p(x) is calculated and ,compared with the threshold ∈, to determine whether an anomaly occurs.

[0047] In some embodiments, when an abnormal state is detected, the system generates an abnormal feature vector and detection results, including the number of the abnormal node, the timestamp, and the degree of deviation from the normal distribution, etc. These data will be recorded in the system's log file and used for subsequent task repair and resource scheduling analysis. For example, if the execution time of a task node deviates from the normal range, the system will record the time deviation and historical mean of the node to support subsequent abnormal pattern analysis.

[0048] As an option, the Gaussian mixture model can be combined with other machine learning algorithms to further improve the accuracy of detection. For example, the output of the Gaussian mixture model can be input into a support vector machine (SVM) or a neural network to identify fine-grained patterns of task node anomalies based on more complex decision boundaries. This multi-level detection approach can significantly improve the accuracy of anomaly detection in complex scenarios, especially when the distribution of multi-dimensional features is asymmetric.

[0049] It can be understood that the anomaly detection process in step S1 is not limited to the analysis of static data, but can also be combined with time series models for dynamic trend prediction. For example, by using the feature change trend in historical data, an ARIMA model or LSTM model is constructed to predict the future state of the task node. When the predicted value deviates significantly from the normal range, an early warning of the anomaly is issued and a repair mechanism is triggered.

[0050] It should be noted that the anomaly detection process in step S1 is closely related to the subsequent task parameter adjustment and process optimization. Through the detected abnormal features, the system can directly generate the initial input for parameter optimization, such as the deviation direction and amplitude of the abnormal point, which is used to guide the optimization direction in the subsequent gradient descent method. In addition, the anomaly detection results can also provide data support from a global perspective for the overall optimization of the process, such as readjusting the resource scheduling strategy or optimizing the priority configuration of the task node based on the statistical results of the abnormal distribution.

[0051] Through the above description, step S1 realizes real-time anomaly detection of process task nodes, provides basic technical support for the efficient and stable operation of the process management system, and lays a solid foundation for anomaly repair and task optimization in subsequent steps. It should be noted that those skilled in the art can make necessary adjustments to the specific implementation based on the above disclosed content to ensure that it adapts to the needs of different business scenarios without departing from the core technical ideas of the present invention.

[0052] Step S2 in the present invention assumes the important responsibility of automatically repairing the detected anomalies. Its goal is to ensure that the abnormal task can quickly return to normal operation by dynamically adjusting the parameters of the task node, and at the same time provide an effective data basis for subsequent process optimization. Step S2 takes the abnormal features detected in step S1 as input, and uses the optimization algorithm of machine learning combined with the execution logic of the task node to complete the repair operation of the anomaly through iterative calculation and dynamic adjustment. In this process, the continuity of the task and the stability of the system are fully guaranteed.

[0053] In this embodiment, step S2 mainly dynamically adjusts the parameters of abnormal task nodes through optimization algorithms. Specifically, when step S1 detects that the task node is in an abnormal state, it will generate abnormal feature data, including the feature vector x of the abnormal point, the degree of deviation from the normal distribution, and related gradient information. Using these data, the system adopts an optimization strategy based on the gradient descent method to gradually adjust the execution parameters of the abnormal point to the normal distribution area. The specific adjustment formula is as follows: Among them, x new represents the adjusted task parameters, logp(x) is the loss function of the outlier point, and represents the degree of abnormality of the current parameters; is the gradient of the loss function with respect to the parameters, reflecting the direction in which the anomaly deviates from the normal distribution; η is the learning rate, which controls the step size of each adjustment. In a possible implementation, the system will constrain the adjustment process in combination with the business characteristics of the task node. For example, for some key task nodes, a smaller learning rate η may be set to ensure the stability of the parameter adjustment process; and for nodes with higher real-time requirements, the learning rate can be appropriately increased to speed up the anomaly repair process. It should be noted that the adjustment process is iterative, and the system will recalculate the probability density p(x) of the task parameters after each adjustment, and determine whether it reaches the normal distribution area based on the adjusted results, thereby terminating the adjustment.

[0054] In some embodiments, in order to further improve the accuracy of anomaly repair, the system will combine the pattern analysis results of historical anomaly data during the repair process. Specifically, the system will use a clustering algorithm (such as K-Means) to classify historical anomaly points, thereby identifying typical patterns of different types of anomalies. When repairing the current anomaly, the system can directly reference the corresponding repair strategy based on the pattern classification results of the anomaly point, such as adjusting the priority or fixed offset value of certain specific parameters to accelerate the repair efficiency and ensure the effectiveness of the repair.

[0055] As an option, the repair process of the task node can also be combined with reinforcement learning technology to achieve dynamic optimization. In this embodiment, the system models the repair process as a decision-making process, with the state of the task node as the environmental input and the repair action as the policy output. The optimal parameter adjustment strategy is gradually learned through a reinforcement learning algorithm (such as Q learning or deep reinforcement learning). For example, the system can design a reward function with the normalization degree and resource consumption of the task node after repair as the main indicators, and complete the dynamic optimization of the repair process by maximizing the reward function.

[0056] It should be noted that the parameter adjustment process of step S2 is not limited to the current abnormal task node, but can also be extended to other task nodes associated with the abnormal node. For example, when a task node is abnormal due to insufficient resources, the system can indirectly alleviate the resource pressure of the abnormal node by adjusting the resource occupancy of other nodes, thereby completing the repair. This global repair strategy can effectively avoid the problem of task chain imbalance caused by local repair.

[0057] After the task parameters are repaired, the system will also record the results of the repair process, including the parameter status before and after the repair, the specific strategies used for the repair, and the time taken for the repair. These data will be stored in the system log and used as input data for subsequent abnormal pattern analysis and process optimization.

[0058] It can be understood that the dynamic parameter adjustment mechanism of step S2 is an important basis for realizing intelligent process management. By combining real-time exception data with historical pattern analysis, the system can not only quickly and effectively repair exceptions, but also provide more accurate data support for subsequent resource optimization and process scheduling, thereby further improving the overall operation efficiency and stability of the system. It should be noted that technical personnel in this field can appropriately expand or modify the parameter adjustment strategy based on the above-mentioned disclosure to adapt to different business scenarios without departing from the core idea of ​​the present invention.

[0059] In the present invention, step S3 is responsible for further using the system's exception handling and parameter adjustment results for process optimization, ensuring that the resource allocation, task node logic and execution order of the entire process achieve global optimization. Its main goal is to improve the overall execution efficiency and resource utilization of the process through comprehensive evaluation of the process operation status and iterative adjustment of the scheduling strategy. In this context, step S3 is closely related to the aforementioned steps S1 and S2. By integrating exception detection and repair data, based on optimization algorithms and historical data analysis results, the process logic and resource configuration are dynamically adjusted, thereby achieving global process optimization.

[0060] In this embodiment, the core of step S3 is to construct an optimization objective function of the process and obtain the best process execution plan through algorithm iteration. Specifically, the optimization objective function comprehensively considers multiple key indicators, including the execution time of the task node, resource usage, abnormality rate, and overall process throughput. Exemplarily, the optimization objective function can be expressed as: F=α·T total +β·R usage +γ·A rate -δ·Q throughput Among them, T total Represents the total execution time of the process; R usage Indicates resource occupancy rate; A rate Indicates the abnormality rate; Q throughput It represents the throughput of the process. The coefficients α, β, γ, and δ are weight parameters used to adjust the emphasis of the optimization goal according to business needs.

[0061] In one possible implementation, in order to solve the above optimization objective function, the system uses a genetic algorithm or a particle swarm algorithm for iterative solution. Specifically, the system first generates an initial solution set (i.e., the initial execution plan of the process), and then selects a better solution by evaluating the fitness value of each solution (i.e., the objective function value). Then, a new generation of solution sets is generated through operations such as crossover and mutation. The above process is repeated until the objective function value converges. It should be noted that the dependency relationship between task nodes can be introduced as a constraint condition in the optimization process. For example, some task nodes must be executed in sequence, or some nodes need to share resources, so as to ensure the feasibility of the optimization results.

[0062] As an option, the process optimization in step S3 can be combined with the results of the abnormal pattern analysis to further refine the optimization plan. Specifically, the system will count the distribution of different types of abnormalities in the historical abnormal data. For example, if some task nodes have a high abnormality rate under specific resource allocation, the optimization process can give priority to adjusting the resource allocation strategy or task priority of the node. At the same time, for other task nodes associated with the abnormal node, the system will dynamically adjust its execution time window or resource configuration according to the association relationship to ensure that the impact of the abnormality on the overall process is minimized.

[0063] In some embodiments, step S3 can also dynamically adjust the optimization plan in combination with the real-time data of the task nodes. For example, when the resource occupancy rate of a task node during the process is close to the system threshold, the system can reallocate resources or adjust the task execution order in advance according to the optimization results to avoid execution blockage of subsequent task chains. For example, for high-priority tasks, the system will give priority to ensuring their resource allocation, and for low-priority tasks, resources can be freed up by delaying execution or reducing resource occupancy rate.

[0064] It should be noted that the optimization result of step S3 not only plays a role in the current process operation, but will also be stored in the system's historical database for subsequent optimization model training and improvement. For example, the system can compare and analyze the effects of different optimization schemes and iteratively update the weight parameters of the optimization objective function, thereby further improving the accuracy and robustness of future process optimization.

[0065] It can be understood that the optimization process in step S3 is global and dynamic. By seamlessly integrating anomaly detection, task repair and process optimization, the system not only achieves rapid response to anomalies, but also improves the overall efficiency and resource utilization of process operation through global optimization strategies. At the same time, step S3 provides an efficient and reliable solution for process management in multi-task and high-concurrency scenarios. It should be particularly emphasized that the optimization algorithm and implementation method disclosed in the present invention are sufficient to enable technical personnel in this field to implement the technical solution without doing creative work, and can be reasonably expanded according to specific business scenarios.

[0066] Step S4 plays an important role in result monitoring and feedback in the overall process of the present invention. Through real-time recording and evaluation of process execution status, exception handling effects, and process optimization results, step S4 provides comprehensive visualization support and data feedback mechanism for the operation of the system. This step not only ensures the transparency of system operation, but also provides key data basis for subsequent process adjustment and optimization strategies. The design of step S4 complements the aforementioned steps S1 to S3. Through in-depth recording and analysis of the execution process, the advantages of the present invention in process management are further consolidated.

[0067] In this embodiment, step S4 realizes detailed tracking of the entire process of process execution by building a global log recording and monitoring system. Specifically, the system collects the execution data of each task node in real time, including but not limited to key information such as the task start time, completion time, execution duration, resource usage, and abnormal status. These data are recorded in the log in a structured form to facilitate subsequent backtracking and analysis. For example, the system generates a record for each task node in the following format: Log(i) = {T start ,T end ,R usage ,E status ,A fix ,O opt} Where: T start Indicates the start time of the task, T end represents the completion time of the task, R usage Indicates the resource occupancy rate of the task; E status Indicates the abnormal state of the task, A fix Indicates the result of abnormal repair, O opt Indicates the resource configuration after task optimization.

[0068] In one possible implementation, the system synchronizes the above log data to cloud storage in real time, supporting users to access and manage through multiple devices. For example, system administrators can view the execution status of task nodes and the operation of the overall process through a visual interface. The interface includes a variety of chart forms such as line charts, bar charts, and pie charts to intuitively display information such as execution time distribution, resource utilization, and exception handling results.

[0069] As an option, the system's monitoring function also includes a detailed evaluation of the results of exception handling. For example, for each exception repair, the system will record the start and end time of the repair and calculate the repair time. At the same time, the system will evaluate whether the execution parameters of the task node have been restored to the normal range after the repair and generate a repair effectiveness report. If a repair result does not meet expectations, the system will automatically mark the exception handling process and store its log data as a special case for subsequent improvement analysis of abnormal patterns.

[0070] In some embodiments, step S4 also integrates a real-time alarm mechanism. When the status or execution data of a task node exceeds the set safety threshold during the process operation, the system will send an alarm to the relevant users through email, SMS or system notification. For example, when the resource utilization rate R usage When the value continues to be higher than the set upper limit, the system will generate alarm information in real time and suggest reallocating resources or adjusting task priorities. This alarm mechanism effectively improves the system's response speed and the initiative of exception handling.

[0071] It should be noted that the log records in step S4 are not only used for real-time monitoring, but also provide key data support for process optimization. For example, the system can discover potential problems in resource allocation or execution order of certain task nodes by comparing and analyzing multiple process operation logs, thereby providing more accurate input data for the optimization algorithm. In addition, historical log data can also be used to build task prediction models, such as using time series analysis technology to predict the execution time of future tasks or the probability of abnormal occurrence, thereby providing forward-looking decision support for process scheduling.

[0072] It can be understood that the monitoring and feedback mechanism of step S4 not only ensures the transparency and controllability of the system operation, but also improves the intelligence level of process management in a data-driven manner. By comprehensively recording and analyzing the execution data of task nodes, abnormal repair results, and process optimization effects, the system can continuously improve the efficiency and stability of process operation, while providing reliable support for process management in complex business scenarios. It should be particularly noted that the technical solution disclosed in the present invention is fully applicable to a variety of business scenarios, and relevant technical personnel can make appropriate adjustments to the specific implementation based on the above content without departing from the core idea of ​​the present invention.

[0073] Example 1: Multi-node anomaly detection and repair in intelligent manufacturing In the intelligent manufacturing system, multiple task nodes (such as production equipment scheduling, material transportation and assembly process) work together to achieve efficient production. Due to the influence of equipment load changes and sudden tasks, some task nodes are prone to execution anomalies, which affects the overall production efficiency. The present invention realizes intelligent management and efficiency improvement of the production system through anomaly detection, dynamic repair and process optimization.

[0074] Implementation steps: First, the system monitors the operating status of each task node (such as task execution time, resource utilization, and success rate, etc.) in real time, and uses these parameters as the multi-dimensional operating data of the node. The system builds an anomaly detection model based on historical normal operation data. This model can dynamically identify abnormal conditions that may exist during task execution. For example, when a device times out due to insufficient resource allocation and the system detects that the execution time of the task exceeds the normal range, it considers that the node is abnormal.

[0075] Subsequently, for the detected abnormal task nodes, the system automatically repairs them by dynamically adjusting relevant parameters. Specific repair measures include adjusting the priority of the task, increasing the allocated resources or re-planning the task path. For example, for task nodes with insufficient resources, the system will dynamically increase the CPU or memory resource allocation according to the current device load, or transfer some tasks to devices with lower load to alleviate the abnormal state.

[0076] After the repair is completed, the system comprehensively optimizes the task node and its associated nodes. The system re-plans the order of task execution based on the task dependencies to ensure the smooth execution of high-priority tasks while balancing the resource usage of other nodes. After the optimization is completed, the system records the parameter status and process operation efficiency before and after the repair to provide data support for subsequent task planning.

[0077] Example 2: Resource Scheduling and Dynamic Optimization of Cloud Computing Platform In a cloud computing platform, multiple tenants run a large number of tasks at the same time. The dynamic changes in task loads may lead to uneven resource allocation, thereby causing system performance problems. The present invention ensures the stable operation of tasks in a high-concurrency environment through an anomaly detection and repair mechanism combined with a resource scheduling optimization strategy.

[0078] Implementation steps: First, the system monitors the task execution status of each tenant and obtains the resource usage data of the task (such as CPU usage, memory occupancy and network bandwidth). The system uses the anomaly detection module to determine whether there is a resource overload problem. For example, when the CPU occupancy of a tenant's task continues to exceed the set threshold and the network bandwidth usage reaches a bottleneck, the system detects that the tenant's task node is abnormal and marks it as a resource overload.

[0079] For detected resource anomalies, the system automatically triggers a repair mechanism and adjusts the resource allocation strategy of the abnormal node. Specifically, the system can free up more resources for high-priority tasks by reducing the resource allocation of low-priority tasks, or balance resource usage by reallocating tasks. For example, the system transfers tasks from servers with higher resource utilization to servers with lower resource utilization while maintaining the timeliness of task execution.

[0080] After the repair is completed, the system further optimizes the overall resource allocation strategy. The goal of the optimization is to ensure the balance of resource use while maximizing the throughput of tasks. The system adjusts the resource allocation rules based on the current task load distribution and historical abnormal patterns, such as increasing the resource upper limit of high-priority tasks and limiting the maximum resource usage of low-priority tasks, thereby improving the overall performance of the system.

[0081] Example 3: Process optimization and feedback mechanism in logistics system In an automated logistics system, multiple distribution nodes need to collaborate to complete the sorting, loading and transportation of goods. Due to the imbalance of distribution tasks or the impact of emergencies, some nodes may experience task accumulation or delays, affecting the overall logistics efficiency. The present invention significantly improves the operating efficiency and reliability of the logistics system through anomaly detection, task repair and full process optimization.

[0082] Implementation steps: First, the system records the task execution data of each distribution node in real time, including key parameters such as task start time, completion time, and task delay duration. The system analyzes these data through the anomaly detection module and finds that the task completion rate of some nodes is lower than the threshold set by the system, or there are continuous task delays. The system determines that the node is abnormal. For example, in sorting center A, due to insufficient human resources, cargo sorting tasks are piled up. The system recognizes that the task delay of this node exceeds the normal range and generates an abnormality report.

[0083] For abnormal nodes, the system repairs them by dynamically adjusting task allocation and scheduling paths. For example, the system dispatches additional resources from the adjacent sorting center B to support sorting center A in completing the cargo sorting task; or re-plans the distribution path and diverts part of the cargo directly to other nodes for processing to alleviate the problem of task backlog.

[0084] After the repair is completed, the system further optimizes the overall logistics process. By analyzing historical task execution data and exception handling records, the system readjusts the task allocation strategy of each node. For example, it allocates more resources to key nodes or adds backup lines during peak hours, and sets task caps for nodes with low task volumes to avoid wasting resources. The system monitors and records the optimized operation process to ensure the execution efficiency and distribution balance of tasks.

[0085] Summarize The above three embodiments fully demonstrate the wide application capabilities of the present invention in complex process management. In the intelligent manufacturing system, the present invention realizes the anomaly detection and dynamic repair of multiple nodes, ensuring the efficient operation of production; in the cloud computing platform, through dynamic resource scheduling optimization, it effectively solves the resource allocation problem in a high-concurrency environment; in the logistics system, through task repair and process optimization, the task execution efficiency and the flexibility of logistics scheduling are greatly improved. These embodiments not only demonstrate the core technical advantages of the present invention, but also show its applicability and scalability in multiple industry scenarios, providing strong technical support and solutions for complex process management systems.

[0086] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A lightweight process management method, characterized in that: The following steps are involved: Define business processes and parse the business processes through a process engine; Provides a graphical process design interface, where users can dynamically configure business process nodes by dragging and clicking; The business process is broken down into multiple task nodes, and each task node is encapsulated as an independent functional module through component design; The task nodes are combined as needed and executed in parallel on different processors to ensure efficient execution of the task nodes; Monitor the execution status of business processes in real time, and promptly discover and handle potential problems through logging and exception handling mechanisms; Intelligent analysis is performed based on historical execution data to automatically optimize the execution order, resource allocation, and load balancing of task nodes to improve system efficiency.

2. The lightweight process management method according to claim 1, characterized in that: Provides dynamic permission configuration based on roles. Users can flexibly set the approver type, approval method and access rights according to the requirements of the node to ensure that the permission control of different users meets the process requirements; The deployment, initiation and approval of the process are implemented through program code and can support rapid integration into existing business systems.

3. The lightweight process management method according to claim 1, characterized in that: The process engine supports multi-version management. Users can switch between different versions of process templates according to business needs to ensure the flexibility and scalability of the process.

4. The lightweight process management method according to claim 1, characterized in that: It supports multiple approval methods, including single-person approval, multi-person approval, sequential approval, and parallel approval. The approval method can be dynamically adjusted during the process execution to meet the needs of different business scenarios.

5. The lightweight process management method according to claim 1, characterized in that: Task nodes can be added or deleted dynamically, and support seamless insertion of new task nodes during process execution, ensuring continuous execution and flexible adjustment of the process.

6. The lightweight process management method according to claim 1, characterized in that: The exception handling mechanism includes automatically repairing errors in task execution and notifying relevant users, analyzing abnormal patterns through machine learning algorithms and adjusting the execution parameters of task nodes to avoid similar errors in the future.

7. The lightweight process management method according to claim 1, characterized in that: By defining business processes through a no-code / low-code platform, users can quickly build and adjust processes through a graphical interface. The platform provides intelligent prompts to avoid configuration errors and improve process building efficiency.

8. The lightweight process management method according to claim 1, characterized in that: The process management system supports seamless integration with external systems, realizes data interaction through standardized interfaces, and ensures collaborative operation and data consistency between business systems.

9. The lightweight process management method according to claim 1, characterized in that: It provides process monitoring and logging functions, real-time tracking of process execution status, detailed recording of task execution information and exception handling, and supports real-time synchronization of log data to the cloud for multi-device access and analysis.

10. The lightweight process management method according to claim 1, characterized in that: Provides access control based on security authentication and authorization mechanisms, supports multi-factor authentication, and dynamically adjusts access rights based on user behavior to ensure the security and compliance of process execution.

Images