Intelligent dynamic topology reconfiguration system and method based on multi-modal event driving

Abstract

This invention discloses an intelligent dynamic topology reconfiguration system and method based on multimodal event-driven approaches, overcoming the shortcomings of existing technologies such as static fixation, inefficient and error-prone manual operation, and lack of dynamic reconfiguration capabilities. The system includes: an event perception engine, enabling multi-source access and protocol adaptation, and performing unified access processing and protocol conversion for various events from multimodal event sources; an event classifier, classifying events into different levels and transmitting high-impact events to a dynamic decision engine; a dynamic decision engine, performing rule chain reasoning and impact range prediction, determining the rules and impact range for permission topology reconfiguration based on high-impact events; and a permission reconfiguration service module, executing permission topology reconfiguration operations for the administrative tree based on the output of the dynamic decision engine.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and dynamic topology, specifically to the technology of permission management and dynamic topology reconstruction in a street cloud platform, and particularly to the technology of intelligent dynamic topology reconstruction of a five-level administrative tree (street level, community level, grid level, building level, and unit level) based on multimodal event-driven architecture, in order to optimize the data access and operation permission management of multiple entities (supervisory end, property management community, community self-governing organization, etc.). Background Technology

[0002] In the process of digital management of streets, the street cloud platform, as a digital management carrier integrating multiple entities such as regulatory ends (e.g., street offices, functional departments), property management communities, and community self-governing organizations, needs to achieve orderly management of data access and operation permissions among these entities through a permission organization tree. Currently, the regulatory hierarchy permission management of existing street cloud platforms generally adopts a static tree structure to represent the permission relationships between entities. Once this static structure is built during the system initialization phase, its hierarchical relationships, permission attributes, and association rules remain fixed and cannot be flexibly adjusted according to actual business needs.

[0003] As the street-level cloud platform continues to expand its services, the number of regulatory levels is increasing, and the number of participating entities, such as property management communities and community self-governing organizations, is also growing, making permission relationships increasingly complex. Against this backdrop, static permission organization trees cannot reflect the dynamic changes in permission relationships in a timely and accurate manner, leading to a significant reduction in permission management efficiency. In practical applications, problems such as permission confusion and abnormal data access frequently occur, seriously affecting the management effectiveness of the street-level cloud platform.

[0004] While some permission management methods exist in the current technical field that can update permissions, most of these methods rely on manual operation to modify permission configurations. Staff must manually locate permission nodes, adjust permission attributes, and verify the configuration results. This process is not only cumbersome and inefficient but also prone to errors due to human negligence. More importantly, these methods can only achieve partial updates of single permission attributes and cannot dynamically reconstruct the overall topology of the permission organization tree, completely failing to meet the core requirements of the street-level cloud platform for real-time, flexible, and accurate permission management.

[0005] In summary, the existing street cloud platform permission management technology has three major shortcomings: First, the permission organization tree structure is static and fixed, which cannot adapt to the dynamic changes in business needs; second, permission updates rely on manual operation, which is inefficient and prone to errors; and third, it lacks dynamic topology reconstruction capabilities, making it difficult to solve problems such as permission chaos and abnormal data access. These problems all need to be addressed and solved through new technical solutions. Summary of the Invention

[0006] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form to prepare for the more detailed descriptions that follow.

[0007] The purpose of this invention is to solve the above-mentioned problems and provide an intelligent dynamic topology reconstruction system and method based on multimodal event-driven architecture. This system enables the administrative tree to dynamically adjust with multimodal events to adapt to business changes. It can automatically complete permission updates and topology reconstruction without manual operation, effectively solving the problems of permission confusion and abnormal data access. It also specifically improves the shortcomings of existing technologies, such as static solidification, inefficient and error-prone manual operation, and lack of dynamic reconstruction capabilities.

[0008] The technical solution of this invention is as follows: This invention discloses an intelligent dynamic topology reconstruction system based on multimodal event-driven architecture. The system includes: an event perception engine, an event grading system, a dynamic decision engine, and a permission reconstruction service module, wherein: The event-aware engine enables multi-source access and protocol adaptation, and performs unified access processing and protocol conversion for various events from multimodal event sources. The event classifier classifies events into different levels and transmits high-impact events to the dynamic decision engine. The dynamic decision engine performs rule chain reasoning and impact range prediction, and determines the rules and impact range of permission topology reconstruction based on high-impact events; The permission restructuring service module performs permission topology restructuring operations on the administrative tree based on the output of the dynamic decision engine.

[0009] According to an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven technology of the present invention, the unified access processing and protocol conversion of the event-aware engine further includes: For structured events: Database change capture technology is used to monitor and capture changes in the permission configuration table and supervision operation log table in the backend database in real time, encapsulate them into standardized event formats, and push them to the event classifier through a message queue; For unstructured events: text data in unstructured events is transmitted to an unstructured database; audio data in unstructured events is converted into text and then used to generate event data, which is then pushed to the event grader. For sensor events: Video data collected by sensors is analyzed to generate alarm events and transmitted to the cloud; noise data collected by sensors is actively reported by the sensors to the cloud as alarm events.

[0010] According to an embodiment of the intelligent dynamic topology reconstruction system based on multimodal event-driven according to the present invention, the event classifier's classification and differentiation processing includes: natural language processing emotion and keyword analysis, sensor threshold judgment, and cross-validation.

[0011] According to an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven according to the present invention, the system further includes a log library, and the event classifier transmits the medium / low impact events distinguished by the classification to the log library for storage.

[0012] According to an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven methods of the present invention, the dynamic decision engine further includes: The event association rule base is stored in a decision tree structure. Each rule contains triggering conditions, association level, and associated permission type. The rule matching engine uses a depth-first search algorithm to traverse the event-related rule base, matches rules that meet the conditions, determines the affected administrative tree level and specific nodes based on the matching rules, generates a list of affected nodes, and pushes it to the permission refactoring service module.

[0013] According to an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven according to the present invention, the permission reconfiguration service module is further configured to perform incremental topology adjustment, which is used to perform node addition / deletion and permission attribute update operations only on the affected nodes and their associated nodes.

[0014] According to an embodiment of the intelligent dynamic topology reconstruction system based on multimodal event-driven according to the present invention, the permission reconstruction service module is further configured to perform conflict resolution, using a periodic scanning and real-time triggering mechanism to detect conflicting permissions, automatically resolve conflicts according to priority rules by calling the conflict resolution rule base, and generate a manual arbitration work order to resolve conflicts manually when automatic resolution is not possible, so as to ensure that the permission configuration of the administrative tree is conflict-free after reconstruction.

[0015] According to an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven according to the present invention, the permission reconfiguration service module is further configured to perform local reconfiguration and global synchronization. The scope of local reconfiguration is determined according to the list of affected nodes as the subtree where the affected nodes are located. The local reconfiguration operation is performed using a transaction mechanism and the service is ensured to be uninterrupted through cached routing. After the reconfiguration is completed, a synchronization message is sent to all servers through a message queue.

[0016] This invention also discloses a multimodal event-driven intelligent dynamic topology reconfiguration method, which operates on the multimodal event-driven intelligent dynamic topology reconfiguration system described above, and the method includes: Step 1: Receive multimodal event sources; Step 2: Distinguish the types of events transmitted from the multimodal event sources, identifying them as structured events, unstructured events, and sensor events respectively; Step 3: Perform semantic classification processing on the various events after type judgment and preliminary processing; Step 4: Based on the semantic classification results of the event, calculate the overall impact level of the event and classify the levels accordingly; Step 5: Initiate the dynamic restructuring process. First, for high-impact events, locate the range of nodes that affect them in the administrative tree. Step 6: Based on the location results of the affected area, perform incremental topology adjustment operations; Step 7: After incremental topology adjustment, call the conflict resolution rule base to handle any possible permission conflicts; Step 8: If the conflict is resolved, perform a partial reconstruction operation and global synchronization to ensure data consistency across the entire system. If the conflict is not resolved, resolve the conflict manually through the manual arbitration interface. Step 9: Determine if the simulation test verification is effective. If the test verification is effective, update the data in the cache and notify the relevant parties. If the test verification is ineffective, roll back the reconstruction operation and issue an alarm.

[0017] Compared with the prior art, the present invention has the following advantages: 1. The present invention provides a five-level administrative tree intelligent dynamic topology reconstruction system and method for street cloud permissions based on multimodal event-driven technology. This system can automatically sense multi-source heterogeneous events (structured / unstructured / sensor data), dynamically adjust the node relationships and permission attributes of the five-level administrative tree through event semantic classification and impact range prediction, and achieve real-time and accurate updates of permission relationships.

[0018] 2. This invention improves the efficiency of permission management on the street cloud platform through an intelligent dynamic reconstruction algorithm (incremental update + conflict resolution), ensuring that the permissions of various regulatory terminals, property management communities and other entities are clear and accurate, and avoiding permission confusion and abnormal data access caused by delayed event response or incorrect permission configuration.

[0019] 3. The core innovation of this invention lies in: ① Semantic hierarchical triggering mechanism for multimodal events: Key technical points: It integrates structured / unstructured / sensor-based multi-source events, and calculates the comprehensive impact level (high / medium / low) of events through NLP text analysis, sensor threshold detection, and event association rule base. It only triggers permission tree reconstruction for high-impact events.

[0020] Beneficial effects: It solves the problems of traditional solutions being highly dependent on a single event type and having a high rate of false triggering, significantly reduces the number of invalid refactorings (expected to be reduced by more than 60%), and improves system resource utilization and response accuracy.

[0021] ② Dynamic topology reconstruction algorithm based on incremental updates (including conflict resolution rule base): Key technical points: The system adopts a local subtree incremental adjustment strategy (modifying only the affected nodes and their relationships), combined with a conflict resolution rule base (automatically handling overlapping permissions) and a parent-child node permission inheritance mechanism, to achieve dynamic reconstruction with O(k) complexity (where k is the number of affected nodes).

[0022] Beneficial effects: Compared with the traditional full reconstruction scheme (complexity O(n²)), the computational efficiency is improved by more than 70%, while avoiding the risk of service interruption caused by global reconstruction and ensuring the real-time and consistency of permission relationships.

[0023] ③ Adaptive elastic scaling strategy for a five-level administrative tree: Key technical points: The number of hierarchical nodes in the administrative tree (e.g., splitting / merging grids) and the scope of permissions (e.g., decentralizing or centralizing permissions) are dynamically adjusted based on multimodal events (e.g., population density changes, scope of impact of emergencies), and are automatically executed through a predefined flexible rule base (e.g., "population density > threshold → split grid").

[0024] Beneficial effects: It enables the permission structure to flexibly adapt to changes in business needs (such as community expansion or emergency situations), improves the adaptability of the street cloud platform to complex scenarios, and ensures management efficiency and user experience. Attached Figure Description

[0025] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0026] Figure 1 A structural diagram of an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven architecture of the present invention is shown.

[0027] Figure 2 A flowchart of an embodiment of the intelligent dynamic topology reconfiguration method based on multimodal event-driven methods of the present invention is shown. Detailed Implementation

[0028] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should be noted that the aspects described below with reference to the accompanying drawings and specific embodiments are merely exemplary and should not be construed as limiting the scope of protection of the present invention in any way.

[0029] Figure 1 The structure of an embodiment of the intelligent dynamic topology reconfiguration system based on multimodal event-driven methods of the present invention is shown.

[0030] See Figure 1 The system in this embodiment includes: an event awareness engine, an event classifier, a dynamic decision engine, and a permission refactoring service module.

[0031] This embodiment of the system connects to an external multimodal event source, providing three types of event inputs: structured events (including permission configuration logs and supervisory operation records), unstructured events (including APP text / voice complaints and public opinion crawler data), and sensor events (including camera pedestrian flow and noise / smoke alarms).

[0032] The event awareness engine of the multimodal event source access system. The event awareness engine realizes multi-source access and protocol adaptation, and performs unified access processing and protocol conversion for various events transmitted from multimodal event sources.

[0033] Specifically, the event-aware engine enables multi-source access and protocol adaptation: (1) For structured events: Database Change Capture (CDC) technology is used to monitor changes in the permission configuration table and regulatory operation log table in the backend database (such as MySQL or PostgreSQL) of the street cloud platform in real time. For example, when a community-level node adds a "property dispute handling" identifier to its permission field, a CDC tool (such as Debezium) can capture the change record and encapsulate it into a standardized JSON format event (including event ID, changed table name, new and old field values, and operation timestamp), which is then pushed to the event grader via a Kafka message queue.

[0034] For regulatory operation logs, operation log tracking code is embedded in the front end of the management system used by street offices and functional departments to record user operation behaviors (such as the operator, operation type, and node ID involved in adjusting the grid-level permission scope), and uploaded to the log server (such as ELK Stack) in batches every minute. Then, the log parsing service extracts key information to generate structured events.

[0035] (2) For unstructured events: Citizen complaint text collection: In the complaint module of the street APP, set up text input box format validation (such as restricting sensitive words and requiring question category labels). After the user submits a complaint, the text data is transmitted to the backend in real time and stored in an unstructured database (such as MongoDB), which triggers the NLP processing flow.

[0036] Voice message collection: Integrate third-party speech recognition (ASR) interfaces (such as iFlytek API). After users record voice messages through the APP, the audio files are first uploaded to cloud storage (such as Alibaba Cloud OSS). The ASR interface asynchronously converts the audio into text. After the conversion is completed, event data containing the original voice link, converted text, and recognition confidence is generated and pushed to the event grader.

[0037] (3) For sensor events: At the community pedestrian monitoring camera terminal, an edge computing gateway is deployed. The camera collects video streams in real time. The edge gateway uses AI video analysis algorithms (such as YOLO object detection + people counting model) to count the number of people in the monitoring area every 5 seconds. When the number of people exceeds the threshold of 50 people, the gateway generates an alarm event containing the camera ID, monitoring area coordinates, the time of exceeding the threshold, and the real-time number of people, and transmits it to the cloud event receiving service via the MQTT protocol.

[0038] The noise sensor uses LoRa wireless communication technology and collects noise values ​​every 10 seconds. When the noise values ​​exceed 70 decibels for three consecutive times, the sensor actively reports an alarm event (including sensor number, collection time, and noise value change curve) to the sensor data receiving module of the street cloud platform.

[0039] The event awareness engine connects to the event classifier and transmits processed event data to it. The event classifier uses NLP (Natural Language Processing) sentiment and keyword analysis, sensor threshold judgment, cross-validation, and other methods to classify events, distinguishing between high-impact events and medium / low-impact events. High-impact events are then transmitted to the dynamic decision engine. Preferably, the system in this embodiment also includes a log library, to which the event classifier transmits medium / low-impact events. The log library receives and stores the medium / low-impact event data from the event classifier for subsequent statistical analysis or traceability queries.

[0040] Specifically, for unstructured events, NLP technology is used to analyze the sentiment tendency (such as the degree of negative sentiment) in text or speech, and keywords strongly related to access control (such as "access error," "data access anomaly," "urgent complaint," etc.) are extracted. For sensor events, the collected values ​​(such as crowd density, noise decibels, and smoke concentration) are compared with preset sensor thresholds (such as crowd density exceeding the community carrying capacity threshold, noise exceeding the resident disturbance threshold, and smoke reaching the fire warning threshold). For structured events, information such as the type of access control configuration change and the urgency of the regulatory actions are analyzed. Then, through a cross-validation mechanism, the analysis results of the three types of events are combined with multi-dimensional impact factors (event urgency, frequency, number of data sources, and scope of impact) to calculate the comprehensive impact level of the event, classifying it into high (≥0.8), medium (0.4-0.7), and low (<0.4) impact levels. High-impact events are transmitted to the dynamic decision engine.

[0041] The following is an example of establishing a multi-dimensional impact factor evaluation system. The factors and their weights are shown in the table below:

[0042] The formula for calculating the overall impact level is: Overall Score = Urgency Score × 0.3 + Frequency Score × 0.2 + Number of Sources Score × 0.2 + Scope of Impact Score × 0.3. A score ≥ 0.8 is considered a high-impact event, 0.4-0.7 is a medium-impact event, and < 0.4 is a low-impact event.

[0043] Application of grading results: High-impact events: Immediately push to the dynamic reconstruction engine and trigger a level 1 system alarm (notify the street emergency management specialist via SMS and APP push). Medium-impact events: Stored in the event buffer queue, processed in batches every 30 minutes, only updating permission configurations, without triggering topology reconstruction; Low-impact events: These are only recorded in the event log library for subsequent access control and data analysis, and no real-time operations are performed.

[0044] The event grading unit connects to the dynamic decision engine. The dynamic decision engine performs rule chain reasoning and impact scope prediction, determines the rules and impact scope for permission topology reconstruction based on high-impact events, and transmits the decision-made data to the permission reconstruction service module.

[0045] Specifically, the dynamic decision engine has a built-in event association rule base, which is stored in a decision tree structure. Each rule includes triggering conditions (e.g., event type = noise exceeding standards, event area = XX community), association level (e.g., grid level, building level), and associated permission type (e.g., environmental supervision permission, emergency response permission). For example, the rule "noise exceeding standards and area is community A" has an association level of all grid-level nodes under community A and an associated permission of environmental supervision permission. The event association rule base supports manual maintenance and automatic updates. When the system detects a new type of event (e.g., complaints about charging pile malfunctions of new energy vehicles), it generates candidate rules by analyzing historical similar events, which are then added to the rule base after being reviewed by the administrator.

[0046] The rule matching engine and event classifier input the key attributes of high-impact events (such as event type, region, and time) into the rule matching engine. The rule matching engine uses a depth-first search algorithm to traverse the event-related rule base and match rules that meet the conditions. Based on the matching rules, the affected administrative tree levels and specific nodes are determined, a list of affected nodes (including node ID, level, and current permissions) is generated, and pushed to the permission refactoring service module.

[0047] The permission restructuring service module performs permission topology restructuring operations on a five-level administrative tree based on the output of the dynamic decision engine.

[0048] The specific processing method of the permission refactoring service module is as follows: I. Implementation of Incremental Topology Adjustment The permission restructuring service module first initiates an incremental topology adjustment process, which only adjusts the affected nodes and their associated nodes as determined by the dynamic decision engine, without affecting unaffected administrative tree nodes. This includes two types of operations: node addition / deletion and permission attribute updates. (a) Implementation of adding / deleting nodes Node addition: When an event requires the addition of a new node (such as adding a "Emergency Access Management" child node), the system first queries the attributes of the parent node (such as a building-level node) (such as permission scope and associated entities), and generates the basic information of the new node based on the parent node attributes. The node ID is generated according to the rule of "parent node ID + child node sequence number", and the level is set to the parent node level + 1. The basic permissions are inherited from the parent node's basic permissions. Then, specific permissions (such as emergency access approval rights) are added to the new node according to the event requirements. Finally, the new node information is completely inserted into the administrative tree database, and the child node list of the parent node is updated synchronously to ensure that the hierarchical association between the parent node and the new child node takes effect.

[0049] Node Deletion: When an event requires the deletion of a node (such as the original grid node becoming invalid after a grid is split), the system first checks whether the node has child nodes. If child nodes exist, the child nodes are migrated to the new parent node (such as the new grid node after the split) to ensure that the permission inheritance chain of the child nodes is not interrupted. After the migration is completed, the record of the original node in the administrative tree database is deleted, and all related relationships (such as the list of child nodes of the parent node, the records related to the original node in the permission allocation table) and permission inheritance chain are updated to avoid the permission chain being broken due to node deletion.

[0050] (II) Implementation of Permission Attribute Update

[0051] Permission Upgrade / Downgrade: For example, when the community-level "Property Complaint Handling" permission needs to be upgraded from "Normal Approval" to "Emergency Direct Access to Street Level", the system first queries the approval process configuration corresponding to this permission (e.g., the original process is community review → street filing, the upgraded process is community submission → street direct approval), modifies the approval node, approval time limit and other fields corresponding to this permission in the permission configuration table; after the fields are modified, the permission data in the Redis cache is updated synchronously to ensure that all subsequent access requests for this permission are executed according to the new approval process.

[0052] Adding / Removing Permissions: When adding a permission to a node (e.g., adding the "Public Facility Inspection" permission to a grid-level node), the system adds an association record between the node ID and the "Public Facility Inspection" permission in the permission allocation table, and updates the node's permission bitmap (using binary bits to indicate whether each permission is enabled, e.g., setting the binary bit corresponding to "Public Facility Inspection" from "0" to "1"); when removing a permission from a node, the system deletes the association record between the node and the target permission in the permission allocation table, and sets the corresponding permission's binary bit in the permission bitmap from "1" to "0", ensuring that the permission status is synchronized in real time.

[0053] II. Conflict Resolution Implementation

[0054] After the incremental topology adjustment is completed, the permission reconstruction service module initiates the conflict resolution process to ensure that there are no conflicts in the permission configuration of the reconstructed administrative tree through conflict detection and resolution: (a) Conflict Detection Methods The system employs a dual mechanism of "periodic scanning + real-time triggering" to detect conflicting permissions: On the one hand, the permission configuration table is scanned periodically every 10 seconds, and a hash table is used to store the mapping relationship of "node - permission - associated subject". When multiple subjects are detected to correspond to the same (node, permission) combination, it is determined to be a permission conflict (such as multiple monitoring terminals having "delete data" permission for the same grid node); on the other hand, when the topology adjustment module performs permission update operations (such as adding or upgrading permissions), it automatically verifies whether the updated data will generate new permission conflicts. If a conflict is detected, the conflict resolution process is immediately triggered to avoid the retention of conflicting permissions.

[0055] (ii) Conflict Resolution Implementation

[0056] Automatic resolution: The system calls the conflict resolution rule library and handles conflicts according to preset priority rules—the hierarchical priority follows "street level > community level > grid level", and the subject type priority follows "supervisory end > property management community > self-governing organization"; the system retains the permissions of the subject with the highest priority according to the priority rules, automatically cancels the conflict permissions of other subjects, and generates a conflict resolution log. The log records in detail the conflict node ID, conflict permission type, resolution result and operation time, which is convenient for subsequent traceability.

[0057] Manual arbitration: When a conflict cannot be resolved automatically (such as when two peer-level regulatory agencies have conflicting permissions for the same node), the system automatically generates a manual arbitration work order. The work order includes conflict details (such as conflicting nodes, conflicting permissions, and involved entities) and candidate solutions (such as permission sharing or time-segmented authorization), and pushes it to the work platform of the street-level permission management specialist. After the specialist reviews the work order and selects a solution, the system performs permission adjustments according to the selected solution. After the adjustment is completed, the conflict resolution log and permission configuration data are updated.

[0058] III. Local Reconstruction and Global Implementation Simultaneously

[0059] After conflict resolution, the permission refactoring service module enters the partial refactoring and global synchronization phase to ensure the efficiency of the refactoring operation and data consistency: (I) Determining the scope of local reconstruction Based on the list of affected nodes output by the dynamic decision engine, the system further determines the local reconstruction scope to be the subtree where the affected nodes are located (such as the third-level grid nodes to the fifth-level unit nodes of a community), and generates a reconstruction scope list. By limiting the reconstruction scope, the system avoids performing invalid operations on unaffected nodes (such as nodes in other communities), which greatly reduces the system's computational overhead and resource consumption.

[0060] (II) Refactoring Execution and Performance Optimization

[0061] The system employs a transaction mechanism to perform partial reconstruction operations, encapsulating a series of operations such as adding / deleting nodes and updating permissions into a single transaction unit to ensure data consistency during the reconstruction process. If any step within a transaction fails, all executed operations are rolled back to the pre-reconstruction state; if all operations succeed, they are committed uniformly. Simultaneously, during reconstruction, data access requests to affected nodes utilize a caching routing strategy, temporarily routing requests to the pre-reconstruction Redis cache to avoid business interruption. After reconstruction, the old cache data is immediately cleared, and the new administrative tree topology and permission configuration are written to the Redis cache to ensure subsequent requests retrieve the latest data.

[0062] (III) Global Synchronization Mechanism

[0063] After the reconstruction operation is completed, the system sends a synchronization message through the Kafka message queue. The message contains key information such as the reconstruction node ID and the new topology version number, and is synchronized to all servers (such as the management system on the monitoring end and the APP backend service). After receiving the synchronization message, each server immediately updates the administrative tree data in its local cache and returns a synchronization confirmation message to the permission reconstruction service module. The system counts the number of confirmation messages in real time. When it is confirmed that all servers have completed the synchronization, the permission topology reconstruction process is marked as officially completed, ensuring that the entire system uses a consistent administrative tree topology and permission configuration.

[0064] An example of adaptive elastic scaling of a five-level administrative tree is shown below.

[0065] When the system detects a 20% increase in population density within a community in a short period (based on dynamic updates of population registration data from the property management system), the following adjustments will be automatically triggered: Hierarchical expansion: Add sub-grid nodes under the original three-level grid nodes (such as splitting the original 1 grid into 2), and migrate some building-level nodes of the original grid (such as buildings 4-6) to the new sub-grids to distribute the management pressure.

[0066] Decentralization of authority: The authority for "public facility repair application" that originally required community-level approval has been decentralized to the new sub-grid level node (the approval level has been shortened from 3 levels to 2 levels), improving response efficiency.

[0067] In summary, the above system can achieve the following effects: Real-time performance: Through the event triggering mechanism, the system can detect permission change events in real time and immediately start the dynamic reconstruction process of the permission organization tree to ensure the real-time update of permission relationships.

[0068] Flexibility: The dynamic reconstruction algorithm can flexibly adjust the structure of the permission organization tree according to different permission change event types (such as regulatory authorization, community authorization, etc.) and specific content, so as to meet the complex business needs of the street cloud platform.

[0069] Accuracy: During the process of reconstructing the permission organization tree, the system will conduct strict permission checks and conflict handling to ensure that the permission scope and type of each node are accurate and to avoid permission confusion and abnormal data access issues.

[0070] High efficiency: Through database operations and cache synchronization mechanisms, the system can quickly update the permission organization tree, reducing the impact on system performance and improving the efficiency of permission management.

[0071] The system of this invention can be applied to a street-level cloud platform, involving entities including regulatory bodies (street offices, functional departments), property management communities, community self-governing organizations, and resident users. The permission relationships between these entities are represented and managed through a five-level administrative tree (street level → community level → grid level → building level → unit level), with permission types including data reading, operation approval, and emergency response.

[0072] Figure 2 It shows the basis Figure 1 The illustrated system is a flowchart of an embodiment of a multimodal event-driven intelligent dynamic topology reconfiguration method. Please refer to [link to relevant documentation]. Figure 2 The following is a detailed description of the real-time steps of the method in this embodiment.

[0073] Step 1: Multimodal event source: As the starting point of the process, three types of event inputs are provided: structured events, unstructured events, and sensor events.

[0074] Step 2: Event type determination: Distinguish the types of events transmitted from the multimodal event source and identify them as structured events, unstructured events, and sensor events.

[0075] Structured event processing (database logs / monitoring operations): For structured events, process event data related to database logs and monitoring operations.

[0076] Unstructured event processing (APP complaint text / voice ASR): For unstructured events, process APP complaint text and voice complaint data converted through voice ASR technology.

[0077] Sensor event processing (camera / noise sensor alarms): This function processes alarm event data generated by sensors such as cameras and noise sensors.

[0078] Step 3: Event Classifier: Perform semantic classification on various events after type judgment and preliminary processing to prepare for subsequent calculation of comprehensive impact level.

[0079] Step 4: Calculate the overall impact level: Based on the event semantic classification results, calculate the overall impact level of the event and distinguish between high-impact events and medium / low-impact events.

[0080] High-impact event handling (triggering dynamic refactoring): If the event is a high-impact event, the dynamic refactoring process in step 5 is triggered.

[0081] Medium / Low Impact Event Handling (Logging - No Refactoring): If the event is a medium / low impact event, simply log the event to the log database and do not perform dynamic refactoring.

[0082] Step 5: Initiate the dynamic reconstruction process and first locate the scope of impact: For high-impact events, locate the range of nodes that affect them in the five-level administrative tree.

[0083] Step 6: Execute the incremental topology adjustment sub-process: Based on the impact range location results, perform incremental topology adjustment operations, including adding / deleting nodes, updating permission attributes, etc.

[0084] Step 7: Conflict resolution rule base: After incremental topology adjustment, call the conflict resolution rule base to handle possible permission conflicts.

[0085] Step 8: Determine if the conflict is resolved. If the conflict is resolved, perform a partial refactoring operation and global synchronization to ensure data consistency across the entire system. If the conflict is not resolved, resolve it manually through the manual arbitration interface.

[0086] Step 9: Determine whether the simulation test is effective. The judgment criteria are, for example, whether the effectiveness is ≥ a set value, such as 99.9%.

[0087] Update Redis cache & notify relevant parties: If the validity is ≥99.9%, that is, the test verification is effective, then update the data in the Redis cache and notify relevant parties.

[0088] Rollback and Refactoring & Alerts: If the validity is less than 99.9%, meaning the test verification is invalid, the rollback and refactoring operation will be rolled back and an alert will be issued.

[0089] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.

[0090] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

[0091] The various illustrative logic blocks, modules, and circuits described in conjunction with the embodiments disclosed herein can be implemented or performed using a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general-purpose processor may be a microprocessor, but in alternatives, it may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0092] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor such that the processor can read and write information to / from the storage medium. In an alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and storage medium may reside as discrete components in the user terminal.

[0093] In one or more exemplary embodiments, the described functionality may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functionality may be stored or transmitted as one or more instructions or code on or through a computer-readable medium. A computer-readable medium includes both computer storage media and communication media, encompassing any medium that facilitates the transfer of a computer program from one location to another. A storage medium may be any available medium accessible to a computer. By way of example and not limitation, such a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and is accessible to a computer. Any connection is also legitimately referred to as a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of a medium. As used in this article, disk and disc include compact discs (CDs), laser discs, optical discs, digital multi-purpose discs (DVDs), floppy disks, and Blu-ray discs. Disks typically reproduce data magnetically, while discs reproduce data optically using lasers. Combinations of these should also be included within the scope of computer-readable media.

[0094] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multimodal event-driven intelligent dynamic topology reconfiguration system, characterized in that, The system includes: an event-aware engine, an event grader, a dynamic decision engine, and a permission refactoring service module, among which: The event-aware engine enables multi-source access and protocol adaptation, and performs unified access processing and protocol conversion for various events from multimodal event sources. The event classifier classifies events into different levels and transmits high-impact events to the dynamic decision engine. The dynamic decision engine performs rule chain reasoning and impact range prediction, and determines the rules and impact range of permission topology reconstruction based on high-impact events; The permission restructuring service module performs permission topology restructuring operations on the administrative tree based on the output of the dynamic decision engine. The dynamic decision engine further includes: The event association rule base is stored in a decision tree structure. Each rule contains triggering conditions, association level, and associated permission type. The rule matching engine uses a depth-first search algorithm to traverse the event association rule library, matches rules that meet the conditions, determines the affected administrative tree level and specific nodes based on the matching rules, generates a list of affected nodes, and pushes it to the permission reconstructing service module. The unified access processing and protocol conversion of the event-aware engine further includes: For structured events: Database change capture technology is used to monitor and capture changes in the permission configuration table and supervision operation log table in the backend database in real time, encapsulate them into standardized event formats, and push them to the event classifier through a message queue; For unstructured events: text data in unstructured events is transmitted to an unstructured database; audio data in unstructured events is converted into text and then used to generate event data, which is then pushed to the event grader. For sensor events: Video data collected by sensors is analyzed to generate alarm events and transmitted to the cloud; noise data collected by sensors is actively reported by the sensors to the cloud as alarm events.

2. The intelligent dynamic topology reconfiguration system based on multimodal event-driven operation according to claim 1, characterized in that, The event grading system's grading and differentiation processes include: using natural language processing for sentiment and keyword analysis, sensor threshold judgment, and cross-validation.

3. The intelligent dynamic topology reconfiguration system based on multimodal event-driven operation according to claim 1, characterized in that, The system also includes a log library, where the event grader transmits medium / low impact events, which are classified into different levels, to the log library for storage.

4. The intelligent dynamic topology reconfiguration system based on multimodal event-driven operation according to claim 1, characterized in that, The permission refactoring service module is further configured to perform incremental topology adjustments, which are used to add / delete nodes and update permission attributes only on the affected nodes and their associated nodes.

5. The intelligent dynamic topology reconfiguration system based on multimodal event-driven operation according to claim 1, characterized in that, The permission refactoring service module is further configured to resolve conflicts. It uses a periodic scanning and real-time triggering mechanism to detect conflicting permissions. It automatically resolves conflicts by calling the conflict resolution rule base according to priority rules. If the conflict cannot be resolved automatically, it generates a manual arbitration work order for manual intervention to resolve the conflict, ensuring that there are no conflicts in the permission configuration of the administrative tree after the refactoring.

6. The intelligent dynamic topology reconfiguration system based on multimodal event-driven operation according to claim 1, characterized in that, The permission reconstruction service module is further configured to perform local reconstruction and global synchronization. The scope of local reconstruction is determined by the list of affected nodes, which is the subtree where the affected nodes are located. The local reconstruction operation is performed using a transaction mechanism and the service is guaranteed to be uninterrupted through cached routing. After the reconstruction is completed, a synchronization message is sent to all servers through a message queue.

7. A method for intelligent dynamic topology reconstruction based on multimodal event-driven mechanisms, characterized in that, The method operates on the intelligent dynamic topology reconfiguration system based on multimodal event-driven mechanisms as described in any one of claims 1 to 6, and the method comprises: Step 1: Receive multimodal event sources; Step 2: Distinguish the types of events transmitted from the multimodal event sources, identifying them as structured events, unstructured events, and sensor events respectively; Step 3: Perform semantic classification processing on the various events after type judgment and preliminary processing; Step 4: Based on the semantic classification results of the event, calculate the overall impact level of the event and classify the levels accordingly; Step 5: Initiate the dynamic restructuring process. First, for high-impact events, locate the range of nodes that affect them in the administrative tree. Step 6: Based on the location results of the affected area, perform incremental topology adjustment operations; Step 7: After incremental topology adjustment, call the conflict resolution rule base to handle any possible permission conflicts; Step 8: If the conflict is resolved, perform a partial reconstruction operation and global synchronization to ensure data consistency across the entire system. If the conflict is not resolved, resolve the conflict manually through the manual arbitration interface. Step 9: Determine if the simulation test verification is effective. If the test verification is effective, update the data in the cache and notify the relevant parties. If the test verification is ineffective, roll back the reconstruction operation and issue an alarm.

Images