CN-121998595-A - Multi-item parallel oriented information system supervision collaborative management method and system

CN121998595ACN 121998595 ACN121998595 ACN 121998595ACN-121998595-A

Abstract

The application discloses a multi-project parallel oriented information system supervision collaborative management method and a system, which relate to the technical field of intelligent collaborative scheduling and comprise the steps of obtaining supervision matters of parallel implementation projects, dividing the supervision matters in each project into supervision task units, and obtaining project state data corresponding to the supervision task units; the method comprises the steps of carrying out parallel scheduling on the supervision task units of each item based on item state data, determining the execution sequence of the supervision tasks, triggering multi-role cooperative execution of the supervision tasks according to the execution sequence, carrying out synchronous management on the execution states of the supervision tasks, collecting feedback information in the execution process of the supervision tasks, carrying out dynamic adjustment on the supervision cooperative flow, carrying out association analysis on the feedback information of each item, and triggering cross-item cooperative treatment according to an analysis result.

Inventors

LAI SHIWEI
MIAO QIAN
LI SIYI
LIU MAO

Assignees

中极华盛工程咨询有限公司

Dates

Publication Date: 20260508
Application Date: 20260129

Claims (10)

1. A multi-item parallel oriented information system supervision cooperative management method is characterized by comprising the following steps: Acquiring supervision matters of parallel implementation projects, dividing the supervision matters in each project into supervision task units, and acquiring project state data corresponding to the supervision task units; based on project state data, performing parallel scheduling on the supervision task units of each project, and determining the execution sequence of the supervision tasks; triggering the multi-role cooperative execution of the supervision tasks according to the execution sequence, and synchronously managing the execution states of the supervision tasks; Collecting feedback information in the execution process of the supervision task, and dynamically adjusting the supervision cooperative flow; and carrying out association analysis based on feedback information of each item, and triggering cross-item co-treatment according to analysis results.
2. The collaborative management method for supervision of a multi-project parallel information system according to claim 1, wherein the acquiring of supervision matters of parallel implementation projects divides the supervision matters in each project into supervision task units and acquires project state data corresponding to the supervision task units is as follows: Acquiring historical supervision execution data formed in the supervision execution process of the completed project; Processing the history supervision execution data, and extracting a supervision execution behavior sequence corresponding to the supervision matters; analyzing the supervision execution behavior sequence, identifying execution fragments repeatedly appearing in different projects, and taking the repeated execution fragments as candidate supervision task structures; Mapping the candidate proctoring task structure into proctoring task units, wherein each proctoring task unit corresponds to an independent executable proctoring operation set and associates corresponding proctoring item identifiers; in the current parallel implementation project, splitting the supervision matters based on the supervision task units to generate a supervision task unit set; Matching the supervision task unit set with the belonging supervision item identifier to obtain a project and a task range corresponding to the supervision task unit, and obtaining real-time state data of the current parallel implementation project; and associating the extracted real-time state data with each supervision task unit based on the matching result to form project state data of the supervision task units.
3. The method for collaborative management of supervision of a multi-project parallel information system according to claim 2, wherein the mapping of candidate supervision task structures to supervision task units is specifically as follows: Generating a corresponding structure identifier according to the supervision execution action identifiers and the arrangement sequence thereof contained in the candidate supervision task structure; Based on the historical supervision execution data, judging whether the candidate supervision task structure completes execution in a state of not depending on other supervision execution actions outside the structure in the execution process; when the candidate supervision task structure meets independent executable conditions, obtaining an execution starting action and an execution ending action corresponding to the structure, and forming an execution boundary; collecting the proctoring operation records corresponding to the proctoring execution action identifiers in the execution boundary to form a proctoring operation set; Registering the supervision operation set as a whole as a supervision task unit, and distributing task unit identifiers for the supervision task unit; And based on the source information of the candidate supervision task structure, the generated supervision task units and the corresponding supervision item identifications are stored in a correlated mode.
4. The method for collaborative management of supervision of a multi-project-oriented parallel information system according to claim 1, wherein the parallel scheduling is performed on the supervision task units of each project based on project status data, and the execution sequence of the supervision tasks is determined, specifically as follows: Analyzing the dependency relationship and conflict state among all the supervision task units according to the project state data; Dividing the supervision task units into task groups which can be executed in parallel according to the dependency relationship and the conflict analysis result and combining the priority of the supervision task units and project state data; outputting the execution sequence of the supervision task units in each task group according to the task priority, the project state and the resource availability to form an executable task sequence list; Integrating the sequences of the task groups into a supervision task execution sequence; based on the supervision task execution sequence, carrying out association analysis on different project task groups, identifying cross-project conflicts, and triggering cross-project optimization strategies.
5. The method for collaborative management of information system supervision oriented to multi-project parallelism according to claim 4, wherein the task units are divided into task groups capable of being executed in parallel by combining the priority of the task units and project status data according to the dependency relationship and the conflict analysis result, specifically comprising the following steps: constructing a directed acyclic graph according to the supervision task units and the dependency relationship thereof; performing topological sorting on the directed acyclic graph, and identifying the executable sequence of the supervision task units; Identifying a critical task unit and a parallelizable task unit by a critical path method according to the topological ordering result and the directed acyclic graph weight; integrating the state information of each parallelizable task unit into a multidimensional feature vector to form a feature space of a supervision task unit; and clustering the supervision task units in the multidimensional feature space to form a task group capable of being executed in parallel.
6. The method for collaborative management of information system supervision oriented to multi-project parallelism according to claim 1, wherein the method is characterized by triggering multi-role collaborative execution of supervision tasks according to execution sequence and synchronously managing execution states of the supervision tasks, and specifically comprises the following steps: Constructing a task dependent network; Analyzing the historical supervision execution behavior sequence to obtain the logic precedence relation of each supervision task unit, and writing the logic precedence relation into a task dependency network to be used as a pre-dependency; Judging whether the task units have execution conditions or not according to the self node attribute and the pre-dependency completion state in the task dependency network, broadcasting execution willingness in a local neighborhood, and simultaneously sensing the resource occupation and the execution state of other task units in the neighborhood to form a local information interaction network; Iterating the local information interaction network, and if the local information interaction network detects that resource conflict or dependence conflict occurs in a task unit, performing neighborhood collaboration according to the request; Continuously carrying out local iteration, gradually and globally developing data accumulation and self-organizing behaviors of a local information interaction network, and naturally gathering parallelizable task units into task groups to form a global executable task sequence; And sensing state change through a local information interaction network in the execution process according to a task group formed by global emergence and a global executable task sequence, and dynamically adjusting through a local self-organizing rule.
7. The method for collaborative management of information system supervision oriented to multi-project parallelism according to claim 6, wherein each task unit judges whether the task unit has execution conditions according to its own node attribute and pre-dependency completion status, and broadcasts execution will in a local neighborhood, and simultaneously perceives resource occupation and execution status of other task units in the neighborhood to form a local information interaction network, specifically as follows: Analyzing the sequence constraint relation among all the supervision task units according to the supervision operation set corresponding to the supervision matters and the execution sequence thereof, and establishing association between task unit identifiers with sequence constraint or finished dependency relation to form a task unit dependency relation record; generating a corresponding pre-dependency task unit identification set for each supervision task unit according to the task unit dependency record; In the scheduling stage, acquiring the node attribute of each task unit, and loading a front-end dependent task unit identifier set corresponding to the task unit as an execution judgment basis; detecting the execution state of each pre-dependent task unit in a pre-dependent task unit identification set, determining that the task unit has an execution condition when all pre-dependent task units in the set are detected to be in a completion state, otherwise, marking the self state as a waiting state and continuously monitoring the dependent state change; When the task unit is judged to have the execution condition, corresponding execution intention information is generated; Broadcasting the execution willingness information to task units in the local neighborhood of the task unit by the task unit; The task unit receives execution willingness information and real-time state information broadcast by other task units in the local neighborhood, and stores the received information as a neighborhood state record; And carrying out association analysis on other task units related to the current task unit based on the neighborhood state record, establishing a dependency relationship side between task units with a front dependency relationship, establishing a resource conflict relationship side between task units with a resource competition or sharing relationship, and generating a corresponding local connection structure by taking the task units as nodes to form a local information interaction network.
8. The method for collaborative management of supervision of a multi-project parallel information system according to claim 1, wherein feedback information is collected during execution of a supervision task, and a collaborative process of supervision is dynamically adjusted, specifically as follows: Receiving feedback information from each task unit based on a local information interaction network, and associating the feedback information with a corresponding task unit identifier, a task group to which the feedback information belongs and a global executable task sequence to form a real-time feedback data set; based on the real-time feedback data set, comparing the planned execution state and the actual execution state of the task unit, identifying an abnormal state, and determining an abnormal influence range; When an abnormal state is detected, triggering a dynamic adjustment mechanism of the supervision cooperative flow; in the process of re-judging the execution conditions, if the pre-dependent task units are not completed, marking the corresponding task units as delay states, and informing the related task units to synchronously adjust the execution sequence through a local information interaction network; if the resource conflict is detected, reallocating the execution resources in the local neighborhood based on the feedback information; the adjustment result is broadcast in real time through the local information interaction network and synchronously updated to the global executable task sequence and the task group state.
9. The collaborative management method for supervision of a multi-project parallel information system according to claim 1, wherein the feedback information based on each project is subjected to association analysis, and cross-project collaborative treatment is triggered according to the analysis result, specifically as follows: Based on a real-time feedback data set formed in the process of executing the supervision task, extracting the execution state of task units in each project, and performing standardized processing to form a feedback feature set; Constructing a cross-project association relation model according to the feedback feature set; based on a cross-project association model, carrying out association analysis on abnormal task units and abnormal influence ranges thereof of different projects, judging cross-project cooperative conflict, and determining project sets and task unit sets related to the cross-project cooperative conflict; the method comprises the steps of evaluating the overall influence intensity of cross-project conflict by combining the current execution stage, task priority distribution and residual critical path length of each related project to judge cross-project cooperative conflict, generating cross-project cooperative treatment triggering conditions, and triggering a cross-project cooperative treatment mechanism when an evaluation result exceeds a preset cooperative threshold; after triggering the cross-project collaborative handling mechanism, carrying out joint rearrangement on task groups and task units of related projects based on a cross-project association network; And synchronously feeding back the cross-project co-processing result to the real-time feedback data set of the affected project and the global executable task sequence through the local information interaction network.
10. A system using a multi-item parallel oriented information system supervision cooperative management method as claimed in any one of claims 1-9, characterized by comprising a partitioning module, a parallel scheduling module, a synchronization management module, an adjusting module and a cooperative disposal module, wherein the modules are connected; The dividing module is used for acquiring the supervision matters of the parallel implementation projects, dividing the supervision matters in each project into supervision task units and acquiring project state data corresponding to the supervision task units; The parallel scheduling module is used for scheduling the supervision task units of each item in parallel based on the item state data and determining the execution sequence of the supervision tasks; The synchronous management module is used for triggering the multi-role cooperative execution of the supervision tasks according to the execution sequence and synchronously managing the execution states of the supervision tasks; The adjusting module is used for collecting feedback information in the execution process of the supervision task and dynamically adjusting the supervision cooperative flow; And the collaborative treatment module is used for carrying out association analysis based on feedback information of each item and triggering collaborative treatment of the cross item according to an analysis result.

Description

Multi-item parallel oriented information system supervision collaborative management method and system Technical Field The invention relates to the technical field of intelligent collaborative scheduling, in particular to a multi-item parallel oriented information system supervision collaborative management method and system. Background With the continuous expansion of the construction scale of the information system, parallel implementation of a plurality of information system projects becomes normal, and higher requirements are put on the organization mode and the coordination capability of the supervision work. The existing information system supervision management mode takes a single item as a core, manual experience is relied on to split and schedule supervision matters, the supervision tasks are coarse in granularity and nonuniform in structure, and effective multiplexing and overall arrangement are difficult to achieve in a multi-item parallel scene. Meanwhile, the traditional supervision flow adopts a centralized and static plan management mode, so that project state change and execution feedback are difficult to timely sense, and problems of low scheduling efficiency, delayed cooperative response, difficulty in timely processing of project-crossing conflicts and the like are easy to occur in the face of complex task dependence and frequent resource competition. Therefore, a collaborative management method for managing information systems oriented to multi-project parallelism is needed to achieve fine modeling, parallel scheduling and dynamic collaboration of management tasks, and improve overall efficiency and controllability of management of information systems in a multi-project parallelism environment. The enterprise multi-project collaborative management method and system based on data security analysis, for example, disclosed by the invention patent with publication number of CN120670116A, comprises the following steps of identifying key task nodes based on data security requirements, evaluating resource constitution and execution deviation to form a task tension strength index, analyzing path propulsion fluctuation and interruption conditions according to the task tension strength index to obtain stability degree, identifying a call aggregation structure of the key resources, judging task time sequence, resource overlapping and dependency difference among paths to generate conflict characteristic quantity, constructing a sequencing rule by combining propulsion state and conflict relation, and determining task collaborative level. By incorporating data security requirements into scheduling decisions, a cross-project state recognition mechanism is established, task tension is evaluated by combining task output field integrity and scheduling stage offset, so that structure sensitive tasks are captured, propulsion fluctuation and interruption frequency are superimposed in a path, task chain stability is measured, and scheduling uneven areas are positioned. For example, the invention patent with the bulletin number of CN120723479B discloses a task scheduling method and a task scheduling system based on multi-agent cooperation, which are characterized in that firstly, a target task demand text input from outside is received, structural analysis processing is performed to obtain a task element set, then a task planning module of a main agent is called to plan and disassemble the task element set to generate a sub task sequence set, sub task units are in one-to-one correspondence with preset professional agent types, the sub task units are allocated to matched professional agent for execution according to professional attribute identifiers of the sub task units to generate sub task execution instructions, each professional agent executes sub task operation based on the instructions to generate a sub task execution result set and feeds back the sub task execution result set to the main agent, and finally the main agent executes multi-source data integration processing on the sub task execution result set to generate a final task result file meeting task output requirements, so that task scheduling efficiency, quality and flexibility can be improved, and complex task demands can be adapted. In the above disclosed technical solution, at least the following technical problems exist: In the traditional technical scheme, the supervision task scheduling is mainly based on manual planning or static rules under the condition of multiple project parallelism, the dynamic adaptation of project state change is difficult, feedback utilization in the supervision process is insufficient, abnormal processing is mostly corrected afterwards, and risk diffusion is difficult to be restrained timely. The present invention proposes a solution to the above-mentioned problems. Disclosure of Invention In order to overcome the defects in the prior art, the embodiment of the invention provides a multi-project parallel informati