Work plan and task management system

Abstract

The invention provides a work plan and task management system, which belongs to the technical field of work plan tasks and comprises a portal layer, an application program interface gateway layer, a business service layer, a public service layer and a data layer. And the application program interface gateway layer realizes intelligent management of the interface through a self-adaptive division module, wherein the intelligent management comprises dependency analysis, granularity calculation, aggregation optimization and recombination execution. The business service layer focuses on work plans and task management, and realizes collaboration through flexible interface interaction. The public service layer provides basic support for user authentication, log recording, message queue, service registration, configuration management and the like. And the data layer adopts a diversified storage strategy, integrates a relational database, a cache database, a time sequence database and distributed file storage, and meets data processing requirements of different scenes. The problem that in the prior art, interface definition and calling relations between micro-services are difficult to manage in a self-adaptive mode is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of work plan and task, and in particular, relates to a work plan and task management system. Background Art

[0002] In the current work management field, work plan and task management have become key components of efficient enterprise operations. Traditional work plan management usually relies on manual preparation and tracking, which has problems such as low efficiency, information islands and lack of flexibility. With the continuous expansion of enterprise scale and the increasing complexity of business, single manual management has been unable to meet the needs of enterprises for work plan and task management. This requires enterprises to establish a more intelligent and adaptive work plan and task management system. At present, most work plan and task management systems adopt a Web-based centralized management mode, using a single database and application to implement work plan formulation, task decomposition, progress tracking and other functions. This centralized architecture has many problems: first, when the scale of the enterprise and the complexity of the business increase, a single database and application service is likely to become a performance bottleneck and it is difficult to support large-scale concurrent access; second, various work plan and task information is stored in isolated systems, and it is difficult to achieve information sharing and collaboration between departments; third, the system has poor scalability and flexibility, and it is difficult to quickly adapt to changing business needs.

[0003] In order to solve the above problems, the industry has begun to try to adopt a work plan and task management system based on a microservice architecture. The microservice architecture divides the entire system into multiple independent service units, each of which is responsible for a specific business function and interacts in a loosely coupled manner. This architecture has good scalability, flexibility and maintainability. However, there are still some problems with the existing microservice-based work plan and task management system: first, the interface definition and call relationship between each microservice is difficult to manage automatically, requiring a lot of manual configuration and maintenance; second, when business needs change, the system cannot adaptively adjust the interface definition and service deployment, requiring manual large-scale reconstruction; third, the system lacks a global perspective, making it difficult to monitor and uniformly manage the overall work plan execution in real time.

[0004] That is to say, the existing technology has the problem of difficulty in adaptively managing the interface definition and calling relationship between microservices. Summary of the invention

[0005] In view of this, the present invention provides a work plan and task management system, which can solve the problem in the prior art that it is difficult to adaptively manage interface definitions and calling relationships between microservices.

[0006] The present invention is achieved in that: The present invention provides a work plan and task management system, which is implemented based on a multi-layer microservice architecture system. The multi-layer microservice architecture system includes a portal layer, an application program interface gateway layer, a business service layer, a public service layer and a data layer: the application program interface gateway layer includes an application program interface adaptive partitioning module, an application program interface monitoring module, an application program interface registration module and an application program interface routing module, wherein the application program interface adaptive partitioning module includes an interface dependency analysis unit, an interface granularity calculation unit, an interface aggregation optimization unit and an interface reorganization execution unit; the business service layer includes a work plan module and a task management module, and the work plan module and the task management module perform interface interaction through the application program interface adaptive partitioning module; the public service layer includes a user authentication service module, a log service module, a message queue service module, a service registration center module and a configuration center module; the data layer includes a relational database module, a cache database module, a time series database module and a distributed file storage module.

[0007] The work plan module is used to perform the following steps: S11, receiving a work plan request submitted by a user, wherein the work plan request includes a plan title, a plan description, a plan type, a plan start time, a plan end time, and a plan priority; S12, generating a work plan unique identifier based on the work plan request, calling the interface dependency analysis unit of the application program interface adaptive partitioning module, and analyzing the dependency relationship between the work plan module and the task management module; S13, according to the dependency analysis result, the interface granularity calculation unit calculates the optimal granularity value of the interface related to the work plan module, and writes the data of the work plan request into the relational database module; S14, calling the interface aggregation optimization unit to perform an aggregation operation on the interfaces in the work plan module whose similarity reaches a preset threshold, and generating an optimized work plan interface set; S15, based on the optimized work plan interface set, executing work plan interface reorganization through the interface reorganization execution unit, and updating the work plan interface information in the service registration center module; S16, calling the message queue service module to send work plan notification information to the work plan participating users; S17, monitoring the execution status information of the work plan in real time, and writing the execution status information into the time series database module; S18, generating a work plan progress report based on the execution status information, and writing the work plan progress report into the distributed file storage module; S19. When the execution status of the work plan changes, the application program interface adaptive partitioning module is triggered to re-execute the interface partitioning evaluation.

[0008] Furthermore, the task management module is used to perform the following steps: S21, receiving the unique identifier of the work plan, calling the application program interface adaptive partitioning module to obtain the latest interface partitioning scheme; S22, creating a task decomposition interface based on the interface division scheme, and decomposing the work plan into multiple subtasks; S23, calling the interface dependency analysis unit to analyze the dependency relationship between the multiple subtasks; S24, determining the optimal granularity values ​​of the interfaces of the multiple subtasks through the interface granularity calculation unit according to the dependency relationship between the multiple subtasks; S25, calling the interface aggregation optimization unit to perform optimization operations on the multiple subtask interfaces to generate an optimized subtask interface set; S26, performing a reorganization operation on the optimized subtask interface set through the interface reorganization execution unit, and assigning task execution personnel to the multiple subtasks; S27, calling the user authentication service module to verify the operation authority of the task executor, and writing the plurality of subtask information into the relational database module; S28, monitoring the execution status of the plurality of subtasks in real time, and when detecting that the task execution status is abnormal, triggering the application program interface adaptive partitioning module to perform an interface optimization operation; S29, generating a task execution report, and updating the overall completion progress of the work plan through the reorganized subtask interface set.

[0009] Furthermore, the interface dependency analysis unit is used to collect interface call link data, analyze interface call frequency, and calculate the coupling degree between interfaces; The interface granularity calculation unit is used to calculate the optimal interface granularity, evaluate the interface response time, and analyze the consistency of interface data based on the business domain model; The interface aggregation optimization unit is used to identify high-similarity interfaces, merge redundant interfaces, and optimize interface parameter structures; The interface reorganization execution unit is used to generate new interface definitions, update service registration information, and execute interface switching strategies.

[0010] Furthermore, the interface dependency analysis unit is used to perform the following steps: S31, collecting interface call data between the work plan module and the task management module, including call timestamp, call direction, call parameters and response results; S32, constructing an interface call link diagram, marking the timing relationship and data flow of the interface call; S33, statistically analyzing the interface call frequency, and calculating the number of times each interface is called per unit time; S34, using a weighted directed graph algorithm to construct an interface dependency matrix, calculating the importance and dependency strength of the interface through a PageRank algorithm, and assigning a weight coefficient to each interface; S35, analyzing the interface data exchange mode and identifying synchronous call and asynchronous call scenarios; S36, using the cosine similarity algorithm to calculate the overlap of data structures between interfaces, and combining Shannon entropy to evaluate data transmission efficiency; S37. Use Dijkstra's shortest path algorithm to identify the key call path, and use Floyd-Warshall algorithm to analyze the optimal call link between all interfaces; S38. Use a topological sorting algorithm to detect circular dependencies and use a strongly connected component algorithm to identify potential circular dependency groups.

[0011] Furthermore, the interface granularity calculation unit is used to perform the following steps: S41, extracting the business domain models of the work plan module and the task management module; S42. Use a hierarchical clustering algorithm to group entity relationships and construct an entity dependency tree; S43, calculating the interface data transmission volume and evaluating the interface response time; S44. Divide the interface responsibilities based on the business domain boundaries; S45, use the multi-objective optimization algorithm NSGA-II to build an evaluation index system to balance the three goals of response time, data consistency and maintenance cost; S46, using simulated annealing algorithm to find the optimal granularity threshold and avoiding local optimum by dynamically adjusting temperature parameters; S47. Genetic algorithm is used to simulate the impact of different granularities on system performance, and the optimal solution is evaluated through the fitness function.

[0012] Furthermore, the interface aggregation optimization unit is used to perform the following steps: S51, parsing the interface definition document of the work plan module and the task management module; S52, extracting interface signature features, including input parameters, output parameters, and business operation types; S53, use K-means clustering algorithm to preliminarily group interfaces, and combine with DBSCAN algorithm to identify abnormal interfaces; S54, use the Word2Vec algorithm to analyze the semantic similarity of the interface and use the LDA topic model to classify business operations; S55, using decision tree algorithm C4.5 to identify dependencies between parameters and optimize parameter structure; S56. Use spectral clustering algorithm to merge interfaces to ensure the cohesion of the aggregated interfaces; S57, verifying the service integrity of the aggregation interface; S58, output interface aggregation optimization solution.

[0013] Furthermore, the interface reorganization execution unit is used to perform the following steps: S61. Use the critical path method to develop an interface reorganization plan and assess the time and resource requirements for reorganization; S62, generating a new interface definition document, including an interface description, parameter description, and a call example; S63, registering new interface information in the service registration center module; S64. Establish a mapping relationship between the old and new interfaces to ensure backward compatibility; S65. Use Monte Carlo simulation algorithm to evaluate the risks of different grayscale release strategies and formulate the optimal release plan; S66. Use anomaly detection algorithms to monitor the operation status of new interfaces and detect anomalies in a timely manner; S67. Perform sample analysis on performance indicators to evaluate the statistical significance of the recombination effect.

[0014] Furthermore, in the step of using an anomaly detection algorithm to monitor the running status of the new interface, the anomaly detection algorithm used is an isolation forest algorithm.

[0015] Furthermore, in the step of sampling and analyzing the performance indicators, the Bootstrap algorithm is used.

[0016] The following are the descriptions of other modules or layers: 1. Portal layer: The portal layer is used to provide a unified user access entrance and is responsible for user interface display, user interaction processing, front-end data verification, responsive layout adaptation, single sign-on authentication, session state management, front-end resource loading optimization, user operation log recording, and interface theme customization to ensure that multi-terminal users obtain a consistent operating experience.

[0017] 2. Application Programming Interface Gateway Layer: 2.1 Application Programming Interface Monitoring Module: The application programming interface monitoring module is used to collect interface call data in real time, analyze interface performance indicators, monitor interface availability status, count interface call trends, identify abnormal call behaviors, generate interface health reports, send alarm notification information, record interface change history, and track interface call links.

[0018] 2.2 Application Programming Interface Registration Module: The application programming interface registration module is used to manage interface metadata information, maintain interface version records, process interface registration requests, verify interface parameter specifications, assign interface access rights, update interface status information, synchronize interface configuration data, manage interface life cycle, and provide interface query services.

[0019] 2.3 Application Programming Interface Routing Module: The application programming interface routing module is used to process interface call requests, perform request load balancing, implement dynamic service discovery, manage routing forwarding rules, process circuit breaker and degradation strategies, perform request current limiting control, record routing log information, maintain routing table information, and optimize routing forwarding performance.

[0020] 3. Public service layer: 3.1 User authentication service module: The user authentication service module is used to verify user identity information, manage user authority data, process identity token issuance, maintain authentication rule configuration, execute access control policy, record authentication log information, manage user session status, synchronize user authentication data, and process single sign-on requests.

[0021] 3.2 Log service module: The log service module is used to collect system operation logs, record user operation logs, store interface call logs, manage log storage strategies, perform log analysis tasks, generate log statistical reports, provide log query services, maintain log backup data, and monitor log exception information.

[0022] 3.3 Message Queue Service Module: The message queue service module is used to process asynchronous message communication, manage message topic subscriptions, execute message distribution tasks, maintain message queue status, ensure reliable message delivery, process message retry strategies, monitor queue performance indicators, record message processing logs, and optimize message processing efficiency.

[0023] 3.4 Service Registry Center Module: The service registry center module is used to manage service registration information, maintain service instance status, process service discovery requests, perform service health checks, update service routing tables, synchronize service configuration data, record service change history, provide service query interfaces, and optimize service discovery performance.

[0024] 3.5 Configuration Center Module: The configuration center module is used to manage system configuration information, maintain configuration version records, process configuration update requests, verify configuration data formats, synchronize configuration change notifications, record configuration modification logs, perform configuration backup tasks, provide configuration query services, and ensure configuration consistency.

[0025] 4. Data layer: 4.1 Relational database module: The relational database module is used to store core business data, manage data table structure, maintain data index information, perform data query optimization, ensure data consistency, handle data backup and recovery, record data change logs, manage data access permissions, and optimize data storage performance.

[0026] 4.2 Cache database module: The cache database module is used to store temporary data information, manage cache expiration policies, perform cache synchronization updates, maintain cache data consistency, process cache invalidation requests, optimize cache hit rates, monitor cache usage status, record cache access logs, and provide cache query services.

[0027] 4.3 Time series database module: The time series database module is used to store time series data, manage data retention policies, perform data downsampling, maintain data compression storage, process time series data queries, optimize query response time, monitor data writing performance, record data processing logs, and provide data analysis interfaces.

[0028] 4.4 Distributed file storage module: The distributed file storage module is used to manage file storage space, maintain file metadata, perform file backup tasks, process file read and write requests, optimize storage space utilization, monitor storage node status, record file operation logs, provide file retrieval services, and ensure high data availability.

[0029] Compared with the prior art, the work plan and task management system provided by the present invention has the following beneficial effects: 1. Adaptive interface management: The system has a built-in application interface adaptive partitioning module, which can automatically analyze the dependencies and call characteristics between microservices, dynamically adjust the granularity and aggregation strategy of the interface, and maximize the performance and maintainability of the system. When business needs change, the system can adaptively re-evaluate the interface definition and complete the interface reorganization without manual intervention, which improves the flexibility of the system.

[0030] 2. Real-time monitoring and unified management: The system integrates the work plan module and the task management module, which can monitor the execution status of the work plan in real time and generate progress reports. Through the unified management platform, managers can fully grasp the execution status of various work plans, find problems in time and make adjustments, thus improving work efficiency.

[0031] 3. High availability and easy scalability: Based on the microservice architecture, the functional modules of the system are loosely coupled and easy to expand and deploy independently. When the system load fluctuates, the corresponding service instances can be quickly added to improve the availability of the system. At the same time, new business needs can also be easily realized by adding new microservice modules, which improves the scalability of the system.

[0032] 4. Powerful data support: The system integrates multiple data storage modules such as relational database, cache database, time series database and distributed file storage, which can meet various data storage requirements in the process of work planning and task management. Rich data support ensures the reliability and scalability of the system.

[0033] In summary, the work plan and task management system proposed in the present invention solves many problems existing in the existing work plan management systems under the centralized and microservice architectures through adaptive interface management, real-time monitoring and unified management, as well as powerful data support capabilities, greatly improves the performance, scalability and flexibility of the system, and solves the problem in the prior art that it is difficult to adaptively manage interface definitions and call relationships between microservices. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 A flowchart of the steps to be performed for the work plan module; Figure 2 A flowchart of the steps executed by the task management module; Figure 3 a flowchart of steps performed for an interface dependency analysis unit; Figure 4 A flowchart of the steps performed by the interface granularity computing unit; Figure 5 A flow chart of steps performed by an interface aggregation optimization unit; Figure 6 Flowchart of the steps performed by the execution unit for the interface reorganization. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0036] The present invention provides a work plan and task management system, which is implemented based on a multi-layer microservice architecture system. The multi-layer microservice architecture system includes a portal layer, an application interface gateway layer, a business service layer, a public service layer and a data layer: The application program interface gateway layer includes an application program interface adaptive division module, an application program interface monitoring module, an application program interface registration module and an application program interface routing module, wherein the application program interface adaptive division module includes an interface dependency analysis unit, an interface granularity calculation unit, an interface aggregation optimization unit and an interface reorganization execution unit; The business service layer includes a work plan module and a task management module, and the work plan module and the task management module interact with each other through an application program interface adaptive partition module; The public service layer includes user authentication service module, log service module, message queue service module, service registration center module and configuration center module; The data layer includes relational database module, cache database module, time series database module and distributed file storage module.

[0037] like Figure 1-6 As shown, they are flowcharts of the execution steps of the work plan module, the task management module, the interface dependency analysis unit, the interface granularity calculation unit, the interface aggregation optimization unit, and the interface reorganization execution unit. The specific implementation methods of the key processes in these modules are described in detail below: The specific implementation of the work plan module is described as follows: The specific implementation of step S11 is as follows: first, the system receives a work plan request submitted by a user, which includes information such as plan title, plan description, plan type, plan start time, plan end time, and plan priority. Next, the system generates a unique work plan identifier based on this information for subsequent work plan management.

[0038] The specific implementation method of step S12 is: the system calls the interface dependency analysis unit in the application interface adaptive partitioning module to analyze the interface dependency relationship between the work plan module and the task management module. Specifically, the unit will collect interface call data between the two modules, including call timestamp, call direction, call parameters, response results and other information. Then, the system constructs an interface call link diagram to mark the timing relationship and data flow of the interface call. Next, the weighted directed graph algorithm is used to calculate the importance and dependency strength of each interface, and a weight coefficient is assigned to each interface. At the same time, the system will also analyze the interface data exchange mode to identify synchronous call and asynchronous call scenarios. Finally, the cosine similarity algorithm is used to calculate the overlap of the data structure between interfaces, combined with Shannon entropy to evaluate the data transmission efficiency, and the topological sorting algorithm is used to detect circular dependencies, and the strongly connected component algorithm is used to identify potential circular dependency groups.

[0039] The specific implementation method of step S13 is as follows: First, the system calculates the optimal granularity value of the interface related to the work plan module through the interface granularity calculation unit according to the interface dependency analysis results obtained in the previous step. Specifically, the unit will extract the business domain model of the work plan module and the task management module, group the entity relationship using the hierarchical clustering algorithm, and build an entity dependency tree. Then, the interface data transmission volume is calculated and the interface response time is evaluated. Next, the interface responsibility scope is divided based on the business domain boundary. After that, the multi-objective optimization algorithm NSGA-II is used to construct an evaluation index system to balance the three goals of response time, data consistency and maintenance cost. Finally, the simulated annealing algorithm is used to find the optimal granularity threshold, and the local optimum is avoided by dynamically adjusting the temperature parameters. The genetic algorithm is used to simulate the impact of different granularities on system performance, and the optimal solution is evaluated through the fitness function. After obtaining the optimal interface granularity, the system writes the data of the work plan request into the relational database module.

[0040] The specific implementation method of step S14 is: first, the system calls the interface aggregation optimization unit to perform aggregation operations on the interfaces in the work plan module whose similarity reaches a preset threshold. The unit will parse the interface definition documents of the work plan module and the task management module, and extract interface signature features, including input parameters, output parameters, business operation types, etc. Then, the K-means clustering algorithm is used to preliminarily group the interfaces, and the DBSCAN algorithm is combined to identify abnormal interfaces. Next, the Word2Vec algorithm is used to analyze the semantic similarity of the interfaces, and the LDA topic model is used to classify business operations. After that, the decision tree algorithm C4.5 is used to identify the dependency between parameters and optimize the parameter structure. Finally, the spectral clustering algorithm is used to merge interfaces to ensure that the aggregated interface has high cohesion and verify the business integrity of the aggregated interface. Through this series of operations, the system generates an optimized work plan interface set.

[0041] The specific implementation method of step S15 is: based on the optimized work plan interface set, the system executes the work plan interface reorganization through the interface reorganization execution unit. The unit first uses the critical path method (CPM) to formulate an interface reorganization plan and evaluate the reorganization time and resource requirements. Then, a new interface definition document is generated, including interface description, parameter description, call examples, etc. Next, the new interface information is registered in the service registration center module. After that, a mapping relationship between the new and old interfaces is established to ensure backward compatibility. Finally, the Monte Carlo simulation algorithm is used to evaluate the risks of different gray release strategies, formulate the optimal release plan, and use the anomaly detection algorithm Isolation Forest to monitor the running status of the new interface and detect anomalies in time. At the same time, the Bootstrap algorithm is used to sample and analyze the performance indicators to evaluate the statistical significance of the reorganization effect.

[0042] The specific implementation method of step S16 is: the system calls the message queue service module to send work plan notification information to the participating users of the work plan. Specifically, the system will push the relevant information of the work plan, such as the plan title, plan description, plan start time, plan end time, etc., to the participating users in an asynchronous manner through the message queue service module, so that they can understand the relevant information of the work plan in a timely manner.

[0043] The specific implementation method of step S17 is: the system monitors the execution status information of the work plan in real time, and writes this status information into the time series database module. Specifically, the system will collect various execution indicators of the work plan, such as completion progress, time spent, problems encountered, etc., and store them in the time series database in a time series manner for subsequent analysis and report generation.

[0044] The specific implementation method of step S18 is: the system generates a work plan progress report based on the execution status information of the work plan, and writes the report into the distributed file storage module. Specifically, the system will analyze the overall execution of the work plan, including the plan progress, resource consumption, risk events, etc., and present it in the form of a report, which is stored in the distributed file storage system for relevant personnel to consult at any time.

[0045] The specific implementation method of step S19 is: when the execution status of the work plan changes, the system will trigger the application interface adaptive partitioning module to re-execute the interface partitioning evaluation. This ensures that during the execution of the work plan, the interface definition can be dynamically adjusted as business needs change to maintain optimal system performance. Specifically, the system will recalculate the interface dependencies, granularity values, aggregation optimization schemes, etc., and perform interface reorganization operations to ensure that the interface design can adapt to changing business needs.

[0046] The specific implementation of the task management module is described as follows: The specific implementation of step S21 is: the task management module receives the unique identifier of the work plan, and calls the application program interface adaptive partitioning module to obtain the latest interface partitioning scheme, so as to ensure that the task management module can create task decomposition based on the latest interface definition.

[0047] The specific implementation method of step S22 is: based on the interface division scheme, the task management module creates a task decomposition interface to decompose the work plan into multiple subtasks. This requires the system to split the work plan into multiple independently executable subtasks according to the specific content of the work plan, and define a corresponding interface for each subtask.

[0048] The specific implementation of step S23 is: the task management module calls the interface dependency analysis unit to analyze the dependency relationship between multiple subtasks. Specifically, the system will collect the interface call data between subtasks, build a call link graph, and use a weighted directed graph algorithm to calculate the importance and dependency strength of each subtask interface, and assign a weight coefficient to each interface. In this way, the execution order and data flow between subtasks can be clarified.

[0049] The specific implementation method of step S24 is: according to the dependency relationship between subtasks, the task management module determines the optimal granularity value of multiple subtask interfaces through the interface granularity calculation unit. Specifically, the system extracts the business domain model of the subtask, uses the hierarchical clustering algorithm to group the entity relationship, and constructs the entity dependency tree. Then, the interface data transmission volume and response time are calculated, and the interface responsibility scope is divided based on the business domain boundary. Next, the multi-objective optimization algorithm NSGA-II is used to construct an evaluation index system to balance the three objectives of response time, data consistency and maintenance cost. Finally, the simulated annealing algorithm is used to find the optimal granularity threshold, and the genetic algorithm is used to simulate the impact of different granularities on system performance to obtain the optimal interface granularity.

[0050] The specific implementation method of step S25 is: the task management module calls the interface aggregation optimization unit, performs optimization operations on multiple subtask interfaces, and generates an optimized subtask interface set. Specifically, the unit will parse the definition document of the subtask interface, extract the interface signature features, and then use the K-means clustering algorithm to preliminarily group the interfaces, and combine the DBSCAN algorithm to identify abnormal interfaces. Then, the Word2Vec algorithm is used to analyze the semantic similarity of the interface, and the LDA topic model is used to classify business operations. After that, the decision tree algorithm C4.5 is used to identify the dependency relationship between parameters and optimize the parameter structure. Finally, the spectral clustering algorithm is used to merge interfaces to ensure that the aggregated interface has high cohesion and verify the business integrity of the aggregated interface.

[0051] The specific implementation method of step S26 is: the task management module performs a reorganization operation on the optimized subtask interface set through the interface reorganization execution unit, and assigns task execution personnel to multiple subtasks. Specifically, the unit first uses the critical path method (CPM) to formulate an interface reorganization plan and evaluate the reorganization time and resource requirements. Then, a new subtask interface definition document is generated, and the new interface information is registered in the service registration center module. Next, a mapping relationship between the new and old interfaces is established to ensure backward compatibility. Finally, the Monte Carlo simulation algorithm is used to evaluate the risks of different grayscale release strategies, formulate the optimal release plan, and use the anomaly detection algorithm Isolation Forest to monitor the running status of the new interface and detect anomalies in time. At the same time, the Bootstrap algorithm is used to sample and analyze the performance indicators to evaluate the statistical significance of the reorganization effect.

[0052] The specific implementation of step S27 is: the task management module calls the user authentication service module to verify the operation authority of the task executor, and then writes multiple subtask information into the relational database module. This ensures that only authorized users can execute the corresponding tasks, and also persists the task information in the relational database for subsequent query and management.

[0053] The specific implementation method of step S28 is: the task management module monitors the execution status of multiple subtasks in real time, and when an abnormal task execution status is detected, the application interface adaptive partitioning module is triggered to perform interface optimization operations. Specifically, the system will collect various execution indicators of subtasks in real time, such as completion progress, time spent, problems encountered, etc., and analyze abnormal situations. Once an abnormality is found, the system will recalculate the interface dependency, granularity value, aggregation optimization plan, etc., and perform interface reorganization operations to ensure that the interface design can cope with changing business needs.

[0054] The specific implementation method of step S29 is: the task management module generates a task execution report, and updates the overall completion progress of the work plan through the reorganized subtask interface set. Specifically, the system will analyze the overall execution of the subtasks, including the planned progress, resource consumption, risk events, etc., and present them in the form of a report. At the same time, the system will also update the overall completion progress of the work plan based on the execution of the subtasks, providing data support for the upper-level work plan management.

[0055] 1. In step S13, the optimal granularity value of the work plan interface is calculated: ; In the formula, is the interface granularity value; The number of business operations included in the interface; is the maximum number of business operations; is the interface cohesion; is the maximum cohesion; is the interface coupling degree; is the maximum coupling degree; is the weight coefficient and satisfies .

[0056] 2. In step S14, interface similarity calculation: ; In the formula, is the parameter similarity; is the functional similarity; is the performance characteristic similarity; is the weight coefficient and satisfies .

[0057] 3. In step S17, the execution status is monitored and scored in real time: ; In the formula, The progress has been completed; is the total progress; is the current exception number; is the maximum number of exceptions allowed; for the remaining time; is the maximum remaining time; is the weight coefficient and satisfies .

[0058] 4. In step S22, the task decomposition granularity is calculated: ; In the formula, For the The granularity value of each subtask; is the complexity of the subtask; is the maximum complexity; is the dependent quantity; is the maximum number of dependencies; To estimate working hours; is the maximum working hours; is the weight coefficient and satisfies .

[0059] 5. In step S23, subtask dependency strength is calculated: ; In the formula, is the data dependency; is the timing dependency; is the resource dependency; is the weight coefficient and satisfies .

[0060] 6. In step S24, the subtask interface granularity optimization objective function is: ; In the formula, It is a 0-1 variable, indicating whether to merge the subtasks; is the total number of subtasks; is the balance coefficient.

[0061] 7. In step S26, the task allocation matching degree is calculated: ; In the formula, For skill matching; is the current load rate; is the maximum load rate; is the relevant experience value; is the maximum experience value; is the weight coefficient and satisfies .

[0062] 8. In step S28, task execution status abnormality detection: ; In the formula, For actual progress; To plan progress; is the current exception number; is the abnormal threshold; is the delay time; To plan time; is the weight coefficient and satisfies .

[0063] 9. In step S29, the overall completion schedule of the work plan is calculated: ; In the formula, For the The degree of completion of each subtask; is the subtask weight, and the calculation formula is: ; In the formula, They are the complexity, working hours and number of dependencies of the subtasks respectively; The total complexity, total work time and total number of dependencies for all subtasks; is the weight coefficient and satisfies .

[0064] The construction principles and significance of these equations are as follows: the interface granularity value comprehensively considers the number of business operations, cohesion and coupling, and balances the functional integrity and independence of the interface; the interface similarity evaluates the similarity of the interface from three dimensions: parameters, functions and performance, and provides a basis for interface aggregation; the execution status monitoring score combines the progress completion, the number of exceptions and the remaining time to comprehensively reflect the execution status; the task decomposition granularity is based on complexity, dependency and man-hour estimation to ensure the rationality of subtask division; the dependency strength calculation considers the dependencies in the three dimensions of data flow, timing and resources to guide task scheduling; the task allocation matching evaluation considers personnel skills, current load and relevant experience to optimize resource allocation; the state anomaly detection identifies abnormal situations through progress deviation, number of exceptions and delay time, and supports timely intervention; the overall progress calculation adopts a weighted average method, and the weight setting considers the complexity, man-hour and dependency of the task.

[0065] The specific implementation of the interface dependency analysis unit is described as follows: The system first embeds a distributed tracing identifier in the application interface call link. The identifier adopts a globally unique identifier format and consists of four parts: timestamp, machine identifier, process identifier and serial number. The system implements full tracking of the call link by injecting the tracing identifier into the request header of the interface call request and response. The collected interface call data includes the call timestamp accurate to milliseconds, the service instance identifier of the caller and the callee, the interface path, the call method type, the complete structure of the request parameters, the response status code, the response data structure, and the call time. The system stores the collected data in a time series database according to the time series, and establishes a multi-level index to support efficient data query and analysis. During the data collection process, the system adopts an adaptive sampling strategy. Under normal load conditions, the sampling frequency is once per second. When the system load exceeds 80%, the sampling frequency is automatically reduced to once every 5 seconds to reduce the impact of monitoring on system performance. When constructing the interface call link graph, the system uses an adjacency matrix to represent the directed graph structure, and the element values ​​in the matrix represent the call weights between interfaces. When traversing the call link through the depth-first search algorithm, the system uses a recursive method and uses an access tag array to avoid the problem of dead loops caused by cyclic calls. During the access process of each node, the system records the access time, access depth and parent node information of the node for subsequent path analysis. When the system statistically analyzes the interface call frequency, it uses a sliding window method based on the time decay factor. The window size is set to 1 hour and the window sliding step is 5 minutes. For each time window, the system calculates the weighted call frequency. The number of calls in the most recent time period has a higher weight. The weight factor decays exponentially with time, and the decay coefficient is set to 0.1. When the system constructs the interface dependency matrix, it first initializes an N×N square matrix, where N is the total number of interfaces. The element values ​​in the matrix are calculated by weighting the three factors of call frequency, data transmission volume and business importance, with weight ratios of 0.4, 0.3 and 0.3 respectively. When the system uses the improved web page ranking algorithm to calculate the importance of the interface, a damping coefficient of 0.85 is introduced, and the calculation is performed through the power iteration method. When the Euclidean distance between the results of two adjacent iterations is less than 0.0001, it is considered to have reached convergence. When analyzing the interface data exchange mode, the system detects the response time distribution characteristics of the interface call and marks the interfaces with an average response time of less than 100 milliseconds and a variance of less than 10 milliseconds as synchronous calls, and marks other interfaces as asynchronous calls. For asynchronous calls, the system further analyzes its callback mechanism, including the timeout setting, retry strategy, and compensation mechanism of the callback interface. When the system uses the cosine similarity algorithm to calculate the degree of overlap of the data structure between interfaces, it first converts the input and output parameters of the interface into a feature vector. The dimension of the vector consists of all possible parameter names, and the parameter value is the frequency of the parameter appearing in the interface.The cosine similarity threshold is set to 0.8. When the similarity of two interfaces exceeds the threshold, the system will mark the two interfaces as candidate merge objects. When calculating the data transmission efficiency, the system uses the Shannon entropy formula to calculate the information entropy of the interface data. The higher the entropy value, the greater the uncertainty of the data and the more information needs to be transmitted. When the system uses the Dijkstra shortest path algorithm to identify the key call path, the interface response time is used as the weight of the edge, and it is implemented through the priority queue optimization algorithm with a time complexity of O(ElogV), where E is the number of call relationships and V is the number of interfaces. When using the Floyd Wascher algorithm to analyze the optimal call link between all interfaces, the system constructs a distance matrix and a predecessor matrix, and calculates the shortest path between any two interfaces through a dynamic programming method. When the system uses the topological sorting algorithm to detect circular dependencies, it uses a variant of the depth-first search algorithm to maintain a stack to record the access path. When it is found that the successor node of a node is already in the stack, a circular dependency is detected. When identifying circular dependency groups through the strongly connected component algorithm, the system uses the Kosaraju algorithm, which first performs a depth-first search on the graph, records the completion time of the nodes, and then performs a second depth-first search on the transpose of the graph, visiting the nodes in reverse order of completion time to obtain all strongly connected components.

[0066] The specific implementation method of the interface granularity calculation unit is described as follows: the system first extracts the core business entities and their relationships from the work plan module and the task management module by combining static code analysis and runtime monitoring. In the static analysis stage, the system parses the database table structure, domain object class definition and business interface definition, and identifies entity attributes, association relationships and business constraints. In the runtime monitoring stage, the system records the creation, modification and deletion operations of entity objects, and analyzes the access mode and data flow path between entities. When the system uses the hierarchical clustering algorithm to group entity relationships, it first calculates the similarity matrix between entities. The similarity calculation considers the three dimensions of the overlap of entity attributes, the correlation of business operations and the temporal nature of data access. The system uses the average link method as the clustering strategy, and constructs a hierarchical clustering tree through a bottom-up iterative merging process. During the clustering process, the system dynamically adjusts the clustering threshold, and the initial threshold is set to 0.6. When the silhouette coefficient of the clustering result is less than 0.5, the system automatically adjusts the threshold until a suitable clustering effect is obtained. When calculating the interface data transmission volume, the system implants a traffic statistics probe in the network layer to record the request and response data packet size within 1 hour. Statistical data include average transmission volume, peak transmission volume, packet size distribution, and transmission rate change trend. The system calculates statistical indicators every 5 minutes and stores the results in a time series database for subsequent trend analysis and capacity planning. When evaluating the interface response time, the system deploys distributed performance monitoring probes to collect the end-to-end response time of interface calls. Monitoring indicators include network transmission time, service processing time, and resource waiting time. The system calculates the average, standard deviation, 95th percentile, and 99th percentile values ​​of these indicators, and sets response time thresholds. When exceeding the threshold, an alarm is triggered. When dividing the scope of interface responsibilities, the system organizes business functions according to aggregate roots and bounded contexts based on the principles of domain-driven design. The system analyzes the correlation between business operations, uses association rule mining algorithms to identify high-frequency operation combinations, and encapsulates these operation combinations in the same interface. When the system uses the non-dominated sorting genetic algorithm 2 to build a multi-objective evaluation system, the three objectives of interface response time, data consistency, and maintenance cost are converted into fitness functions. The response time target is calculated based on performance monitoring data, the data consistency target is evaluated by analyzing data version conflicts and update delays, and the maintenance cost target considers code complexity and the scope of change impact. The system sets the population size to 100, the number of evolution generations to 500, and adopts a binary tournament selection strategy with a single-point crossover probability of 0.8 and a uniform mutation probability of 0.1. In each generation of evolution, the system uses the crowding comparison operator to maintain the diversity of solutions to ensure a uniformly distributed Pareto front. When the system uses the simulated annealing algorithm to find the optimal granularity threshold, it adopts an adaptive temperature adjustment strategy. The initial temperature is set to 100, and the cooling rate is automatically adjusted according to the current state of the system. The cooling rate ranges from 0.8 to 0.99.100 iterations are performed at each temperature, and the Metropolis acceptance criterion is used to decide whether to accept the new solution. The system defines a state transition function and randomly walks in the solution space. The state transition step size decreases as the temperature decreases, ensuring that there is a greater probability of jumping out of the local optimum when the temperature is high, and focusing on searching near the optimal solution when the temperature is low. Finally, the system uses a genetic algorithm to simulate the impact of different granularities on system performance, encodes the granularity parameters as real chromosomes, sets the population size to 200, and the evolutionary generation to 1000. The fitness function comprehensively considers the system's average response time, peak response time, resource utilization, throughput, and error rate, and calculates the fitness value by weighted summation. The system adopts an elite retention strategy to ensure that the best individuals in each generation will not be lost during the evolution process. By running the algorithm multiple times and performing statistical analysis on the results, the most stable granularity scheme is selected as the final optimization scheme.

[0067] The specific implementation method of the interface aggregation optimization unit is described as follows: The system first builds an interface definition parser to support the recognition and parsing of mainstream interface description specifications, including open application interface specifications, network service description languages, and interface description languages. The parser uses a recursive descent algorithm to parse the interface definition document into an abstract syntax tree, from which the complete metadata information of the interface is extracted. When extracting interface signature features, the system uses a feature engineering method to convert the various dimensional features of the interface into numerical vectors. For input parameters and output parameters, the system analyzes their data type hierarchy, the mandatory and optional attributes of the parameters, the value constraints of the parameters, and the dependencies between the parameters. The system uses a morphological restoration algorithm to process parameter names, and identifies the semantic features of the parameters through word segmentation and part-of-speech tagging. For the classification of business operation types, the system preliminarily classifies the interfaces into four basic types of addition, deletion, modification, and query based on a verb dictionary and rule templates. For compound operations, the system further decomposes them into a combination of basic operations. The system uses an improved K-means clustering algorithm to preliminarily group interfaces. First, the optimal number of cluster centers is determined by silhouette coefficient analysis, which is generally set between 3 and 7. The system uses the K-means addition algorithm to select the initial cluster center, and improves the stability of the clustering results by maximizing the distance between cluster centers. During the clustering process, the system uses the cosine distance to measure the similarity between interface feature vectors, and iterates until the moving distance of the cluster center is less than the preset threshold of 0.001 or the maximum number of iterations is 300. When the system combines the density clustering algorithm to identify abnormal interfaces, it first calculates the local density of each sample point and the distance to the high-density point. The scanning radius is set to 0.5 times the average distance between samples in the feature space, and the minimum number of samples is set to 2% of the total number of samples. The system marks interfaces with low local density and far distance from high-density points as abnormal points. These interfaces usually have special business semantics or do not meet the design specifications of the system. When the system uses the word vector algorithm to analyze the semantic similarity of the interface, it first builds a corpus of interface names and parameter names and uses the word vector model for training. The dimension of the word vector is set to 100, and considering the characteristics of interface naming, the context window size is set to 5 words. The system uses negative sampling to optimize the training process, with the negative sample ratio set to 5 and the number of training rounds to 100. At the same time, the system uses the implicit Dirichlet distribution topic model to cluster interfaces. The number of topics is determined by perplexity analysis and is generally set between 8 and 12. The system uses the Gibbs sampling algorithm for parameter estimation, and the number of iterations is set to 1000. For each interface, the probability distribution of its belonging to each topic is calculated. When identifying the dependency between parameters, the system uses the improved C4.5 decision tree algorithm, taking the occurrence of parameters as features and the dependency between parameters as the classification target. The system uses the information gain ratio as the metric for feature selection and constructs a decision tree recursively. To prevent overfitting, the system sets the minimum number of samples to 5, the maximum tree depth to 10, and uses the pessimistic pruning algorithm to optimize the decision tree.Based on the classification results of the decision tree, the system extracts explicit and implicit dependencies between parameters to optimize the parameter structure. When the system uses the spectral clustering algorithm to merge interfaces, it first constructs a similarity matrix between interfaces. The similarity calculation comprehensively considers the functional similarity, parameter structure similarity and call pattern similarity of the interfaces. The system selects the radial basis function as the kernel function, and the kernel parameter σ determines the optimal value through grid search. The system calculates the eigenvectors of the Laplace matrix through eigenvalue decomposition, selects the eigenvectors corresponding to the first k smallest eigenvalues ​​to construct a new feature space, where the k value is determined by eigenvalue difference analysis. In the new feature space, the system uses the K-means algorithm for final clustering to obtain the interface group to be merged. When verifying the business integrity of the aggregated interface, the system constructs a test case set that covers various call scenarios and boundary conditions of the interface. The test cases include three dimensions: functional verification, performance verification and exception handling verification. The system uses regression testing to ensure that the aggregated interface meets the original business requirements, including data consistency check, transaction integrity verification and business rule compliance test. Finally, the system generates a detailed interface aggregation optimization plan, including the specific implementation steps of interface merging, parameter mapping rules, calling method adjustment suggestions and related system configuration change instructions.

[0068] The specific implementation method of the interface reorganization execution unit is described as follows: When the system uses the critical path method to formulate an interface reorganization plan, it first decomposes the reorganization task into the smallest execution unit, including interface design, code refactoring, test verification, and release deployment. The system constructs a task dependency graph and uses an adjacency list to represent the front-end and back-end relationships between tasks. Through two traversals of forward recursion and backward recursion, the earliest start time, earliest completion time, latest start time, and latest completion time of each task are calculated to determine the time float of the task. The system identifies the task sequence with zero float time as the critical path, and these tasks directly affect the total duration of the project. When generating a new interface definition document, the system adopts a template-driven approach to automatically generate an interface description document based on a predefined document template. The document content includes basic information of the interface (such as interface name, version number, access address, etc.), detailed interface description (including function description, usage scenarios and precautions, etc.), request parameter description (including parameter name, type, whether it is required, default value and value range, etc.), response data structure (including the data format of successful response and failed response), and call sample code. When the system registers new interface information in the service registration center, it adopts a double-write mechanism to maintain the interface registration information of both the new and old versions. The system sets an expiration time for each interface version, which is 30 days by default. During the expiration time, the system supports the calling of both the new and old versions. Before the expiration time arrives, the system monitors the calling of the interface to ensure that all callers have completed the upgrade. When establishing the mapping relationship between the new and old interfaces, the system creates a detailed mapping configuration file, including mapping rules at the parameter level (such as parameter name mapping, type conversion rules, etc.), conversion logic of response data (including data structure conversion and field mapping, etc.), and mapping relationship of error codes. The system manages these mapping rules uniformly through the configuration center, supporting dynamic updates and real-time effectiveness. When the system uses the Monte Carlo simulation algorithm to evaluate the risks of different grayscale release strategies, it simulates the system behavior under different traffic distribution schemes by building a system performance model. During the simulation process, the system randomly generates 10,000 samples, each of which contains different traffic characteristics (such as request rate, data volume, etc.) and system status (such as server load, network delay, etc.). The system evaluates the risk level of each grayscale solution through the statistical distribution of performance indicators, and selects the solution with the lowest risk as the final release strategy. When the system uses the isolation forest algorithm for anomaly detection, it constructs an isolation tree by random hyperplane segmentation. The system sets the contamination rate of anomaly detection to 0.1, which indicates the expected proportion of abnormal samples. During the sampling process, the system randomly selects features and segmentation points to construct multiple isolation trees, and the sample size of each tree is set to 256. The system calculates the average path length of the sample to obtain the anomaly score. The higher the score, the more likely it is an anomaly. Finally, the system uses the self-help method to perform sampling analysis on the performance indicators, and calculates the confidence interval of the performance indicators by repeating random sampling 1,000 times.The system sets the significance level to 0.05. When the lower bound of the confidence interval is greater than the preset performance improvement target, it is considered that the interface reorganization has achieved the expected effect. In this way, the system ensures that the effect of interface reorganization is statistically significant.

[0069] 1. In step S34 of the interface dependency analysis unit, the PageRank algorithm is used to calculate the importance of the interface, which is specifically expressed as follows: ; In the formula, For interface PageRank value; To call the interface A collection of interfaces; For interface The number of interfaces called externally; is the damping coefficient, the value range is [0,1], and the default value is 0.85; The total number of interfaces.

[0070] 2. In step S36, the cosine similarity calculates the interface data structure overlap: ; In the formula, are the characteristic vectors of the two interfaces; Interface No. eigenvalues; is the feature dimension.

[0071] 3. Shannon entropy evaluates data transmission efficiency: ; In the formula, is information entropy; For the The probability of class data appearing; The number of data types.

[0072] 4. In the interface granularity calculation unit step S45, the multi-objective optimization evaluation index is calculated: ; In the formula, is the comprehensive assessment score; is the response time; is the maximum allowed response time; It is an indicator of data consistency; is the maximum consistency index; For maintenance costs; is the maximum allowable maintenance cost; is the weight coefficient and satisfies .

[0073] 5. In step S54 of the interface clustering optimization unit, Word2Vec semantic similarity calculation: ; In the formula, is the target word; is the context window size; is the input word vector; is the output word vector; is the vocabulary size.

[0074] 6. In step S57, business integrity verification score: ; In the formula, For service coverage; For call chain integrity; For parameter completeness; is the weight coefficient and satisfies .

[0075] 7. In step S65 of the interface reorganization execution unit, grayscale release risk assessment: ; In the formula, is the failure probability; For recovery time; For the scope of influence; is the weight coefficient and satisfies .

[0076] 8. In step S37 of the interface dependency analysis unit, the Dijkstra shortest path algorithm calculates the optimal call link: ; In the formula, From source node to node The shortest distance; is the intermediate node; For Node arrive The weight of a task can be expressed as response time or resource consumption.

[0077] 9.Floyd-Warshall algorithm calculates the global optimal path: ; In the formula, Indicates that before Nodes, slave nodes To Node The shortest path length.

[0078] 10. In step S46 of the interface granularity calculation unit, the simulated annealing algorithm searches for the optimal granularity threshold: ; In the formula, is the probability of accepting a new solution; is the change in the objective function value; is the Boltzmann constant; is the current temperature, and , is the initial temperature, is the cooling coefficient (0.8~0.99), is the number of iterations.

[0079] 11. In step S47, the fitness function of the genetic algorithm is: ; In the formula, is the average response time; It is an indicator of data consistency; is the system complexity; is the weight coefficient and satisfies .

[0080] 12. Roulette probability calculation for selection operator: ; In the formula, For individuals Probability of being selected; For individuals The fitness value of ; For population size.

[0081] 13. In step S53 of the interface aggregation optimization unit, the DBSCAN algorithm density calculation is: ; In the formula, For point of The set of points in the neighborhood; For the dataset; is the Euclidean distance; is the neighborhood radius.

[0082] 14. In step S54, the Dirichlet distribution of the LDA topic model: ; In the formula, For the subject distribution; is a hyperparameter; is the number of topics; is the gamma function.

[0083] 15. In step S55, the information gain rate of decision tree C4.5 is calculated as: ; Where information gain: ; Split information: ; In the formula, is the attribute to be split; is the sample set; The attribute value is A subset of samples.

[0084] 16. In step S66, the Isolation Forest anomaly score is calculated: ; In the formula, is the anomaly score; For sample The average path length; The sample size is The average path length of a binary tree; is the path length of the isolated tree.

[0085] 17. In step S67, the confidence interval estimate of Bootstrap resampling: ; In the formula, is the parameter estimate; is the critical value of the standard normal distribution; is the Bootstrap standard error, , is the number of resampling times, For the The parameter estimates for the resampled values.

[0086] The construction principles of these equations are as follows: The PageRank algorithm adopts an iterative calculation method to construct a directed graph through the calling relationship between interfaces, calculates the importance of each interface, and takes into account the frequency of interface calls and the importance of the caller; cosine similarity measures the similarity of interface features by calculating the vector angle, which is suitable for similarity calculation in high-dimensional feature space; Shannon entropy reflects the amount of information in data transmission. The larger the entropy value, the more dispersed the data distribution and the lower the transmission efficiency; the multi-objective optimization evaluation index adopts a weighted summation method, normalizes the indicators of different dimensions and conducts a comprehensive evaluation, which is convenient for scheme comparison and selection; Word2Vec uses a neural network model to learn word vector representation and captures the semantic relationship of words through context prediction; the business integrity score considers the three dimensions of service coverage, call chain and parameters, and adopts a weighted method to comprehensively evaluate the rationality of interface aggregation; the grayscale release risk assessment comprehensively considers the failure probability, recovery time and impact range, and is used to select the optimal release strategy. The Dijkstra algorithm and the Floyd-Warshall algorithm use the idea of ​​dynamic programming to gradually update the shortest path and ensure the global optimality of the call link; the simulated annealing algorithm introduces random disturbances to avoid falling into the local optimality, and the temperature parameter controls the convergence process of the algorithm; the fitness function of the genetic algorithm comprehensively considers the system performance, data consistency and complexity, and optimizes the interface granularity through evolutionary operations; the DBSCAN algorithm defines clustering based on density connectivity, which can discover clusters of any shape and is suitable for identifying abnormal interfaces; the LDA topic model uses the Dirichlet distribution to model the topic-word distribution and capture the semantic characteristics of the interface; the C4.5 decision tree algorithm uses the information gain rate as the splitting criterion to avoid bias towards attributes with more values; the Isolation Forest isolates abnormal points by randomly dividing the space and calculates the path length required for isolation to measure the degree of abnormality; the Bootstrap method estimates the statistical characteristics of the parameters through resampling to evaluate the reliability of the reorganization effect.

[0087] Specifically, the principle of the present invention is: the work plan and task management system of the present invention adopts a multi-layer microservice architecture, mainly including a portal layer, an application interface gateway layer, a business service layer, a public service layer and a data layer. Among them, the application interface gateway layer is the core of the entire system and is responsible for adaptive management of the interfaces between microservices.

[0088] The API gateway layer consists of four key modules: 1. Interface adaptive partitioning module: This module is responsible for dynamically evaluating and adjusting the interface definitions between microservices. It first collects the call data of the interface, including call time, direction, parameters, response results and other information, and constructs the interface call link graph. Then, the weighted directed graph algorithm is used to calculate the importance and dependency strength of each interface, and a weight coefficient is assigned to each interface. At the same time, the data exchange mode of the interface is analyzed to distinguish between synchronous and asynchronous call scenarios. Next, cosine similarity and Shannon entropy are used to evaluate the data transmission efficiency between interfaces, providing a basis for subsequent interface optimization.

[0089] 2. Interface granularity calculation unit: This unit is responsible for determining the optimal granularity of the interface. It first extracts the business domain model and uses a hierarchical clustering algorithm to build an entity dependency tree. Then, it calculates the data transmission volume and response time of the interface, and divides the interface responsibilities based on the business domain boundaries. Next, the multi-objective optimization algorithm NSGA-II is used to balance indicators such as response time, data consistency, and maintenance cost to obtain the optimal interface granularity threshold. Finally, simulated annealing and genetic algorithms are used to verify the impact of different granularities on system performance and determine the optimal interface granularity.

[0090] 3. Interface aggregation optimization unit: This unit is responsible for aggregating and optimizing interfaces. It first parses the interface definition document and extracts the signature features of the interface. Then, the K-means and DBSCAN algorithms are used to preliminarily group the interfaces and identify abnormal interfaces. Next, the Word2Vec and LDA models are used to analyze the semantic similarity and business operation types of the interfaces to provide a basis for interface aggregation. After that, the decision tree algorithm is used to optimize the interface parameter structure and reduce redundant parameters. Finally, the spectral clustering algorithm is used to merge highly similar interfaces to ensure the cohesion of the aggregated interfaces.

[0091] 4. Interface Reorganization Execution Unit: This unit is responsible for executing the interface reorganization plan. It first uses the critical path method to develop a reorganization plan and evaluate time and resource requirements. Then, it generates a new interface definition document, registers the new interface information in the service registration center, and establishes a mapping relationship between the new and old interfaces to ensure backward compatibility. Next, the Monte Carlo simulation algorithm is used to evaluate the risks of different grayscale release strategies and select the optimal solution. At the same time, an anomaly detection algorithm is used to monitor the operating status of the new interface and detect anomalies in a timely manner. Finally, the Bootstrap algorithm is used to perform statistical analysis on the reorganization effect to ensure significant improvement in performance indicators.

[0092] Through the collaboration of the above four key modules, the application interface gateway layer can adaptively manage the interface definition and call relationship between microservices within the system. When business needs change, this layer can dynamically evaluate and optimize the interface design to ensure that the system always maintains high performance and scalability.

[0093] In order to better understand and implement the present invention, an embodiment of a specific application scenario of the present invention is provided below: a construction installation company undertook a curtain wall installation project of a 300-meter-high super-high-rise building, which involved the cooperation of multiple construction units and suppliers and was required to be completed in 8 months. In order to improve the efficiency of project management, the construction company adopted the work plan and task management system of the present invention for project management.

[0094] First, the project manager creates a general project plan through the system's work plan module. The work plan request contains the following information: the plan title is "300-meter super-high-rise building curtain wall installation project", the plan description lists the project overview, technical requirements and quality standards in detail, the plan type is "general construction plan", the plan start time is March 1, 2024, the plan end time is October 31, 2024, and the plan priority is "highest". The system generates a unique identifier "PROJECT202403001".

[0095] Based on the above work plan request, the system analyzes the dependency relationship between the work plan and task management through the interface dependency analysis unit. The analysis results show that the project needs to decompose the construction plan into five stages: design confirmation, material procurement, construction preparation, main construction and completion acceptance. There is a strict time sequence dependency relationship between each stage. The specific dependency analysis data is shown in Table 1 below.

[0096] Table 1 Dependency analysis data table

[0097] Based on the dependency analysis results, the system calculated the optimal interface granularity value and determined that the work plan interface should be divided into four main interfaces: plan creation, plan update, progress query, and status monitoring. After interface aggregation optimization, a complete work plan interface set was formed and registered in the service registration center.

[0098] After receiving the work plan identifier, the task management module decomposes the overall plan into multiple specific construction tasks based on the latest interface division scheme. The system first decomposes the tasks in the design confirmation stage. The specific task list is shown in Table 2 below.

[0099] Table 2 Task list

[0100] The task decomposition results of the material procurement stage are shown in Table 3 below.

[0101] Table 3 Procurement Task List

[0102] The task breakdown of the construction preparation phase is shown in Table 4 below.

[0103] Table 4 Construction preparation task list

[0104] The main construction phase is divided into multiple construction areas according to the floors. The specific tasks of each area are shown in Table 5 below.

[0105] Table 5 Construction area task list

[0106] The system analyzes the dependencies between tasks through the interface dependency analysis unit and determines the optimal granularity value of the task interface using the interface granularity calculation unit. The interface aggregation optimization unit optimizes similar tasks and manages tasks with the same characteristics using a unified interface.

[0107] During the project execution, the system monitors the execution status of each task in real time. When an abnormal situation is detected at the construction site, the system automatically triggers the interface optimization operation. For example, during the installation of the curtain wall on the 15th floor, the construction was suspended due to external weather conditions. The system detected that the T014 task status was abnormal, and immediately sent an early warning message to the relevant personnel through the message queue service module, and automatically adjusted the schedule of subsequent tasks.

[0108] The system regularly generates work plan progress reports, which include plan completion status, quality control indicators, safety production data, etc. Table 6 shows the main data of a monthly progress report.

[0109] Table 6 Monthly Progress Report

[0110] The system dynamically adjusts the task allocation strategy based on the progress report data. For construction areas that are lagging behind, the system automatically proposes optimization suggestions, including adding construction personnel, adjusting working hours, optimizing construction procedures, etc. For example, in order to catch up with the construction schedule, the system recommends using parallel operations in the construction of the 16th to 20th floors, changing the original serial keel installation and glass installation tasks to cross-operation.

[0111] During the project implementation process, the system strictly controls the operating permissions of personnel at all levels through the user authentication service module. Construction personnel can only view and perform specific tasks assigned to them, and project managers can view and manage all tasks, but sensitive information involving costs and contracts requires a higher level of approval permissions.

[0112] All operation records are stored and managed through the log service module. The system uses a time series database to record the changing trends of various indicators, including construction progress, material consumption, man-hours, etc. These data provide an important basis for subsequent project optimization and experience summary.

[0113] By adopting the system of the present invention, the curtain wall installation project was successfully completed within a construction period of 8 months. The system's automated management and intelligent optimization functions significantly improved project management efficiency, reduced manual intervention, and ensured construction quality. The specific effects are shown in the following aspects: First, through the system's intelligent task decomposition and resource allocation, the work efficiency of project personnel has increased by about 25%, and labor costs have been reduced by about 15%. Secondly, the real-time monitoring and early warning mechanism enables problems in the construction process to be discovered and solved in a timely manner, and the quality defect rate has been reduced by about 30% compared with the traditional management model. Thirdly, the system's data analysis function provides strong support for management decisions, keeping the overall cost of the project within the budget and saving about 5% of material costs. Finally, the electronic management and automated generation of project documents have reduced clerical work time by about 60%.

[0114] This embodiment fully demonstrates the application value of the present invention in the management of large-scale construction projects. The various functions and modules of the system cooperate closely to form a complete project management closed loop, providing effective technical support for improving the management level of construction projects.

[0115] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A work plan and task management system, characterized in that: It is implemented based on a multi-layer microservice architecture system, which includes a portal layer, an application interface gateway layer, a business service layer, a public service layer, and a data layer: The application program interface gateway layer includes an application program interface adaptive division module, an application program interface monitoring module, an application program interface registration module and an application program interface routing module, wherein the application program interface adaptive division module includes an interface dependency analysis unit, an interface granularity calculation unit, an interface aggregation optimization unit and an interface reorganization execution unit; The business service layer includes a work plan module and a task management module, and the work plan module and the task management module interact with each other through the application program interface adaptive partitioning module; The public service layer includes a user authentication service module, a log service module, a message queue service module, a service registration center module and a configuration center module; The data layer includes a relational database module, a cache database module, a time series database module and a distributed file storage module.

2. A work plan and task management system according to claim 1, characterized in that: The work plan module is used to perform the following steps: S11, receiving a work plan request submitted by a user, wherein the work plan request includes a plan title, a plan description, a plan type, a plan start time, a plan end time, and a plan priority; S12, generating a work plan unique identifier based on the work plan request, calling the interface dependency analysis unit of the application program interface adaptive partitioning module, and analyzing the dependency relationship between the work plan module and the task management module; S13, according to the dependency analysis result, the interface granularity calculation unit calculates the optimal granularity value of the interface related to the work plan module, and writes the data of the work plan request into the relational database module; S14, calling the interface aggregation optimization unit to perform an aggregation operation on the interfaces in the work plan module whose similarity reaches a preset threshold, and generating an optimized work plan interface set; S15, based on the optimized work plan interface set, executing work plan interface reorganization through the interface reorganization execution unit, and updating the work plan interface information in the service registration center module; S16, calling the message queue service module to send work plan notification information to the work plan participating users; S17, monitoring the execution status information of the work plan in real time, and writing the execution status information into the time series database module; S18, generating a work plan progress report based on the execution status information, and writing the work plan progress report into the distributed file storage module; S19. When the execution status of the work plan changes, the application program interface adaptive partitioning module is triggered to re-execute the interface partitioning evaluation.

3. A work plan and task management system according to claim 2, characterized in that: The task management module is used to perform the following steps: S21, receiving the unique identifier of the work plan, calling the application program interface adaptive partitioning module to obtain the latest interface partitioning scheme; S22, creating a task decomposition interface based on the interface division scheme, and decomposing the work plan into multiple subtasks; S23, calling the interface dependency analysis unit to analyze the dependency relationship between the multiple subtasks; S24, determining the optimal granularity values ​​of the interfaces of the multiple subtasks through the interface granularity calculation unit according to the dependency relationship between the multiple subtasks; S25, calling the interface aggregation optimization unit to perform optimization operations on the multiple subtask interfaces to generate an optimized subtask interface set; S26, performing a reorganization operation on the optimized subtask interface set through the interface reorganization execution unit, and assigning task execution personnel to the multiple subtasks; S27, calling the user authentication service module to verify the operation authority of the task executor, and writing the plurality of subtask information into the relational database module; S28, monitoring the execution status of the plurality of subtasks in real time, and when detecting that the task execution status is abnormal, triggering the application program interface adaptive partitioning module to perform an interface optimization operation; S29, generating a task execution report, and updating the overall completion progress of the work plan through the reorganized subtask interface set.

4. A work plan and task management system according to claim 3, characterized in that: The interface dependency analysis unit is used to collect interface call link data, analyze interface call frequency, and calculate the coupling degree between interfaces; The interface granularity calculation unit is used to calculate the optimal interface granularity, evaluate the interface response time, and analyze the consistency of interface data based on the business domain model; The interface aggregation optimization unit is used to identify high-similarity interfaces, merge redundant interfaces, and optimize interface parameter structures; The interface reorganization execution unit is used to generate new interface definitions, update service registration information, and execute interface switching strategies.

5. A work plan and task management system according to claim 4, characterized in that: The interface dependency analysis unit is used to perform the following steps: S31, collecting interface call data between the work plan module and the task management module, including call timestamp, call direction, call parameters and response results; S32, constructing an interface call link diagram, marking the timing relationship and data flow of the interface call; S33, statistically analyzing the interface call frequency, and calculating the number of times each interface is called per unit time; S34, using a weighted directed graph algorithm to construct an interface dependency matrix, calculating the importance and dependency strength of the interface through a PageRank algorithm, and assigning a weight coefficient to each interface; S35, analyzing the interface data exchange mode and identifying synchronous call and asynchronous call scenarios; S36, using the cosine similarity algorithm to calculate the overlap of data structures between interfaces, and combining Shannon entropy to evaluate data transmission efficiency; S37. Use Dijkstra's shortest path algorithm to identify the key call path, and use Floyd-Warshall algorithm to analyze the optimal call link between all interfaces; S38. Use a topological sorting algorithm to detect circular dependencies and use a strongly connected component algorithm to identify potential circular dependency groups.

6. A work plan and task management system according to claim 5, characterized in that: The interface granularity calculation unit is used to perform the following steps: S41, extracting the business domain models of the work plan module and the task management module; S42. Use a hierarchical clustering algorithm to group entity relationships and construct an entity dependency tree; S43, calculating the interface data transmission volume and evaluating the interface response time; S44. Divide the interface responsibilities based on the business domain boundaries; S45, use the multi-objective optimization algorithm NSGA-II to build an evaluation index system to balance the three goals of response time, data consistency and maintenance cost; S46, using simulated annealing algorithm to find the optimal granularity threshold and avoiding local optimum by dynamically adjusting temperature parameters; S47. Use genetic algorithm to simulate the impact of different granularities on system performance, and evaluate the optimal solution through fitness function.

7. A work plan and task management system according to claim 6, characterized in that: The interface aggregation optimization unit is used to perform the following steps: S51, parsing the interface definition document of the work plan module and the task management module; S52, extracting interface signature features, including input parameters, output parameters, and business operation types; S53, use K-means clustering algorithm to preliminarily group interfaces, and combine with DBSCAN algorithm to identify abnormal interfaces; S54, use the Word2Vec algorithm to analyze the semantic similarity of the interface and use the LDA topic model to classify business operations; S55, using decision tree algorithm C4.5 to identify dependencies between parameters and optimize parameter structure; S56. Use spectral clustering algorithm to merge interfaces to ensure the cohesion of the aggregated interfaces; S57, verifying the service integrity of the aggregation interface; S58, output interface aggregation optimization solution.

8. A work plan and task management system according to claim 7, characterized in that: The interface reorganization execution unit is used to execute the following steps: S61. Use the critical path method to develop an interface reorganization plan and assess the time and resource requirements for reorganization; S62, generating a new interface definition document, including interface description, parameter description, and calling examples; S63, registering new interface information in the service registration center module; S64. Establish a mapping relationship between the old and new interfaces to ensure backward compatibility; S65. Use Monte Carlo simulation algorithm to evaluate the risks of different grayscale release strategies and formulate the optimal release plan; S66. Use anomaly detection algorithms to monitor the operation status of new interfaces and detect anomalies in a timely manner; S67. Perform sample analysis on performance indicators to evaluate the statistical significance of the recombination effect.

9. A work plan and task management system according to claim 8, characterized in that: In the step of using an anomaly detection algorithm to monitor the running status of the new interface, the anomaly detection algorithm used is an isolation forest algorithm.

10. A work plan and task management system according to claim 9, characterized in that: In the step of sampling and analyzing the performance indicators, the Bootstrap algorithm is used.

Images