A student card cooperative perception and hierarchical response method for campus security events

Abstract

The application provides a student ID cooperative perception and hierarchical response method for campus security events, and relates to the field of data processing. In the method, by constructing a student ID association record set, an order mother flow graph, and a student ID attached flow segment set, the false shell is gradually stripped and an inverse order silk chain, an intention cavity area, and a latent event original nucleus are generated. In combination with the shell removal, the hierarchical response, and the result backfilling, the closed-loop identification and dynamic response of the hidden abnormal behavior in the campus security event are realized. By implementing the technical solution provided by the application, the behavior deviation can be extracted and finely measured from the micro level under the premise of maintaining the consistency of the macro behavior, so that the abnormal behavior camouflage problem can be effectively identified.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a student ID collaborative perception and hierarchical response method for campus security incidents. Background Technology

[0002] With the continuous improvement of campus informatization and intelligent security systems, multi-source perception and behavior analysis methods based on student ID cards have been widely applied in the identification and early warning of campus security incidents. Existing technologies typically collect multi-source data such as location observation records, access control trigger records, and passageway entry / exit records corresponding to student ID cards to construct student ID card behavior chains. These chains are then matched and analyzed based on preset normal behavior templates to identify abnormal behavior patterns. In this type of technical solution, the normal behavior template is usually constructed based on historical large-scale group behavior statistics, enabling effective identification of behaviors that significantly deviate from the norm at the levels of macro-path, passage time, and regional distribution, thereby meeting campus security management needs to a certain extent.

[0003] However, in real-world campus settings, a more subtle problem of camouflaging abnormal behavior exists. Individuals mimic normal behavioral paths or utilize group cover to make their student ID behavior chains appear consistent with normal behavioral templates at a macro level, but exhibit subtle deviations at a micro level, such as abnormal stopping positions, abnormal neighborhood relationships, abnormal walking rhythms, or abnormal boundary-attaching behaviors. Because current technologies primarily rely on macro-path matching and statistical threshold determination, they struggle to effectively characterize and quantify these micro-deviations, often leading to these weakly abnormal behaviors being misjudged as normal and thus undetected. Furthermore, in densely populated areas, there is a high degree of overlap between multiple student ID behavior chains, further masking individual abnormal behavior with group behavior. This makes it difficult for anomaly detection methods based on individual trajectories to distinguish between individual anomalies and normal group fluctuations, thereby reducing overall recognition accuracy.

[0004] Therefore, how to extract behavioral deviations from the micro level and perform precise measurement while maintaining consistency in macro-behavior, so as to effectively identify the problem of disguised abnormal behavior, has become an urgent technical problem to be solved. Summary of the Invention

[0005] This application provides a student ID collaborative perception and hierarchical response for campus security incidents, which facilitates the extraction and fine measurement of behavioral deviations at the micro level while maintaining macro-level behavioral consistency, so as to effectively identify the problem of abnormal behavior disguise.

[0006] The first aspect of this application provides a student ID collaborative perception and hierarchical response method for campus security incidents. The method includes: acquiring location observation records, access control trigger records, passage arrival / departure records, peer neighborhood records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to the student ID, to form a student ID-related record set for campus security incidents; constructing an order flow graph based on the student ID-related record set, and mapping the student ID-related record set to the order flow graph to generate a student ID-related flow fragment set; performing shell extraction processing on the student ID-related flow fragment set to obtain an extracted... The extracted results are processed by performing order anti-entropy weaving to generate reverse-order threads; the reverse-order threads are then processed by performing substitute projection to obtain projection results, and the projection results are then processed by performing intent cavity mining to generate intent cavity regions; circumstantial evidence recall is performed around the intent cavity regions to obtain recall results, and latent event incubation is performed based on the recall results to generate latent event kernels; based on the latent event kernels, deshelling and perturbation processing and hierarchical response determination processing are performed to output student ID card hierarchical response results, and the student ID card hierarchical response results are then fed back to form a closed-loop evolutionary student ID card collaborative perception and hierarchical response process.

[0007] A second aspect of this application provides a student ID card collaborative sensing and hierarchical response device for campus security incidents. The device includes an acquisition module and a processing module. The acquisition module is used to acquire location observation records, access control trigger records, passage arrival / departure records, peer neighborhood records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to the student ID card, to form a set of student ID card-related records for campus security incidents. The processing module is used to construct an order flow graph based on the student ID card-related record set and map the student ID card-related record set to the order flow graph to generate a set of student ID card-related flow segments. The processing module is also used to perform shell extraction processing on the set of student ID card-related flow segments. The processing module obtains an extraction result and performs an order anti-entropy weaving process on the extraction result to generate a reverse-order silk chain. The processing module also performs a substitute projection process on the reverse-order silk chain to obtain a projection result, and performs an intent cavity mining process on the projection result to generate an intent cavity region. The processing module further performs a corroborating evidence recall process around the intent cavity region to obtain a recall result, and performs a latent event incubation process based on the recall result to generate a latent event kernel. The processing module also performs a shell removal and wake-up process and a graded response determination process based on the latent event kernel to output a student ID card graded response result, and performs a backfeedback of the student ID card graded response result to form a closed-loop evolutionary student ID card collaborative perception and graded response process.

[0008] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, and both the user interface and the network interface are used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method described above.

[0009] A fourth aspect of this application provides a non-transitory computer-readable storage medium storing instructions that, when executed, perform the method described above.

[0010] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages: By constructing a layer-by-layer convergent processing chain from the student ID-related record set to the latent event kernel and then to the hierarchical response results, and through mechanisms such as shell extraction, order anti-entropy weaving, and substitute projection, macroscopically normal but microscopically abnormal behaviors are stripped from the order shell and made explicit into identifiable structures. Simultaneously, the intention cavity region and circumstantial evidence feedback clusters are used to elevate individual behaviors to event-level structures. Anomaly verification and risk classification are completed through the linkage of shell removal and hierarchical response judgment processing, and a closed-loop evolution is formed through result feedback. This significantly improves the accuracy of identifying disguised and weakly abnormal behaviors and enhances the system's adaptive capability. Therefore, it is convenient to extract behavioral deviations at the micro level and perform fine-grained measurement while maintaining macroscopic behavioral consistency, thereby achieving effective identification of the problem of disguised abnormal behavior. Attached Figure Description

[0011] Figure 1 A flowchart illustrating a student ID collaborative perception and hierarchical response method for campus security incidents provided in this application embodiment; Figure 2 A schematic diagram of a student ID card collaborative sensing and hierarchical response device for campus security incidents provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0012] Explanation of reference numerals in the attached figures: 21. Acquisition module; 22. Processing module; 31. Processor; 32. Communication bus; 33. User interface; 34. Network interface; 35. Memory. Detailed Implementation

[0013] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0014] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0015] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. In addition, the terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0016] To address the aforementioned technical problems, this application provides a student ID collaborative perception and hierarchical response method for campus security incidents, referring to... Figure 1 , Figure 1 This is a flowchart illustrating a student ID collaborative perception and hierarchical response method for campus security incidents, provided in an embodiment of this application. The method is applied to a server and includes steps S110 to S160, as follows:

[0017] S110. Obtain the location observation records, access control trigger records, passage arrival and departure records, peer neighborhood records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to the student ID card, so as to form a set of student ID card-related records for campus security incidents.

[0018] Specifically, a server refers to the core computing platform responsible for aggregating, processing, modeling, and outputting multi-source student ID data. Its essence is not limited to a single physical device, but can manifest as a combination of centralized servers, distributed computing clusters, or cloud computing platforms. The student ID master index is a unique identifier for the same student ID within the multi-source sensing system, used to uniformly aggregate location observation records, access control trigger records, channel arrival / departure records, peer neighborhood records, class time records, dormitory schedule records, campus activity records, and space control records onto the same identity platform. Location observation records are used to characterize the location observation results of student ID cards over continuous time. Access control trigger records are used to characterize the trigger events of card swiping, identification, or authorization contact between student ID cards and access control nodes. Channel arrival / departure records are used to characterize the moment state of student ID cards entering or leaving channel nodes. Colleague neighbor records are used to characterize the accompanying relationship formed between student ID cards and other student ID cards in local space. Class time records are used to characterize the time intervals and teaching venue constraints corresponding to the course schedule. Dormitory schedule records are used to characterize the entry and exit time patterns and dormitory return constraints related to dormitories. Campus activity records are used to characterize the spatiotemporal aggregation patterns corresponding to the start, duration, and end of activities. Space control records are used to characterize the open, restricted, and temporary adjustment states of specific spatial semantic locations at different time periods. The core of the attribution binding process is to eliminate the identity splitting problem caused by inconsistent identifier formats, different upload delays, and different record granularities between different record sources. Therefore, the server first extracts the original identifier fields from each record, then constructs the attribution similarity, and merges records with attribution similarity higher than a preset threshold under the same student ID card master index. Attribution similarity can be expressed as:

[0019] in Indicates the similarity of attribution, used to determine whether the current record should be merged into the target student ID primary index; , , and The weight coefficients represent different discrimination factors and are used to adjust the influence of original identifier consistency, temporal proximity, spatial proximity and sequential continuity in the attribution binding. Each weight coefficient can be set according to the campus equipment deployment situation, and usually the sum is 1. Indicates the consistency of the original identifier, reflecting the degree of correspondence between card numbers, device identification numbers, or authentication numbers in different records; Indicates temporal proximity, used to reflect the degree of closeness between the times when two records occurred; Indicates spatial proximity, used to reflect the degree of proximity between corresponding positions of two records; This indicates continuity and reflects whether two records have a natural succession relationship in behavioral evolution. After the server completes the unified merging of multi-source records based on this attribution similarity, it obtains the original attribution results organized around the same student ID master index, laying a unified identity foundation for the subsequent formation of the student ID time-series record chain.

[0020] After completing the attribution binding process, the server performs time alignment processing on various records merged under the same student ID master index to form a student ID time-series record chain under a unified time benchmark. The student ID time-series record chain refers to a continuous time behavior chain formed after reordering and continuously organizing various records according to a unified time benchmark. Since location observation records are typically high-frequency continuous sampling records, access control trigger records are typically instantaneous event records, and passage arrival / departure records typically include two boundary moments: entry and departure. However, classroom time records, dormitory schedule records, campus activity records, and space control records are represented as duration intervals. Therefore, the server needs to first uniformly transcribe all records into a time expression structure compatible with start time, end time, and duration intervals, and then perform clock offset correction and upload latency compensation. The time-corrected record time can be represented as:

[0021] in Indicates the first Each record is a correction time under a unified time base; Indicates the first The original record time of each record; This indicates the amount of clock offset correction for the device, used to eliminate time discrepancies caused by inconsistencies in the internal clocks of different devices; This represents the network transmission delay correction amount, used to correct for the time lag that occurs during the process of uploading records from the terminal to the server; This represents the system synchronization compensation amount, used to unify the time base differences between different business systems. After the server completes time correction, it sorts the records according to the corrected start and end times and identifies the time connection relationship between adjacent records. This allows the location observation records to form a sequential sequence with access control trigger records and passage arrival / departure records, and allows classroom period records, dormitory schedule records, campus activity records, and space control records to be attached to the corresponding behavior time periods. The resulting student ID time sequence record chain is no longer a collection of isolated time points, but rather represents the continuous behavioral progression of the student ID over time.

[0022] After forming the student ID time-series record chain, the server performs spatial positioning processing on the chain to map various records to preset spatial semantic locations and form a student ID spatial record chain. Spatial semantic location refers to a location unit in the campus space with clear behavioral and order-constraining meanings, such as the entrance to a teaching building, dormitory building, canteen, library, floor transition points, connecting corridors, access control areas, boundary zones, outer edges of sensitive areas, open convergence areas, and sparse transition areas. The student ID spatial record chain refers to a continuous spatial behavioral chain of various records within the student ID time-series record chain after their location has been assigned within a unified spatial semantic framework. During implementation, the server first constructs a spatial semantic location dictionary based on the campus electronic map and spatial control records. Then, it maps the coordinate observation results from the positioning observation records to the nearest spatial semantic location. Simultaneously, it combines access control trigger records and channel arrival / departure records to correct spatial boundary assignments, avoiding misjudging outside a door as inside entry. Spatial positioning can be achieved through a combination of distance and topological constraints. The location matching score can be expressed as:

[0023] in Indicates the first The record matched the first The spatial semantic location score; , and The weighting coefficients for different criteria are used to adjust the influence of proximity, topological reachability, and boundary consistency in spatial placement. Indicates the first The spatial coordinates corresponding to the record and the first The distance between the centers of spatial semantic locations; This represents the distance decay parameter, used to control the magnitude of the impact of distance changes on the matching score; Represents topological reachability, used to characterize whether the positions before and after the current record can be naturally reached through actual campus paths. A spatial semantic location; This indicates boundary consistency, used to characterize whether the current record conforms to the access control boundary, passage boundary, and spatial control boundary. The server selects the position with the highest placement score that meets the spatial control constraints as the final spatial semantic position, and overlays the neighboring records of the same row onto the corresponding spatial semantic position to characterize the current student ID's state at that position: solitary state, accompanying state, group edge state, or group occlusion state.

[0024] After forming the student ID spatial record chain, the server performs event semantic annotation processing on the student ID spatial record chain to generate a student ID behavior record chain with behavioral semantic states. Behavioral semantic states refer to state expressions with behavioral meaning extracted from multi-source records and spatial semantic locations, such as approach behavior, traversing behavior, stopping behavior, turning back behavior, edge-keeping behavior, backflow behavior, attempted entry behavior, completed entry behavior, failed entry behavior, passageway movement behavior, establishing accompaniment behavior, dissolving accompaniment behavior, short-term neighbor exchange behavior, and group occlusion behavior. The student ID behavior record chain is a continuous chain of behavioral semantics formed by the server after interpreting continuous spatial changes in the student ID spatial record chain together with background constraints. During implementation, the server comprehensively judges behavior based on the magnitude of positional changes at adjacent times, the duration of stopping, changes in access control trigger states, changes in peer neighborhoods, and records from class periods, dormitory schedules, campus activities, and spatial control. For example, if a student ID card maintains the same spatial semantic position for multiple consecutive moments and the duration exceeds a dwelling threshold, it is labeled as dwelling behavior; if a student ID card moves continuously from the entrance of the teaching building along the connecting corridor and conforms to the flow of people during breaks, it is labeled as passing through behavior; if a student ID card repeatedly approaches the access control node in the access control area but fails to pass through, it is labeled as attempting to enter behavior. The server can also construct behavior discrimination indicators for continuous position changes and dwelling states:

[0025] in The behavior discrimination index represents the current time period and is used to help distinguish the semantic states of behaviors such as passing through, stopping, turning back, and following the edge; , , and The weight coefficients representing different discrimination factors; It indicates the intensity of location change within a time period, used to characterize the magnitude and continuity of student ID card movement; This indicates a persistent feature, used to characterize the degree to which a student ID card remains continuously at the same semantic location in space; It represents the characteristics of neighborhood changes and is used to characterize the frequency of the establishment, dissolution, or switching of accompanying relationships; The background consistency feature is used to characterize the degree of consistency between the current behavior and records from class time periods, dormitory schedules, campus activities, and space management. After the server completes the semantic annotation of the event based on the combination of behavior discrimination indicators and rules, the student ID behavior record chain forms a unified expression from the original spatial trajectory to the clear behavior state, providing stable behavior granularities for subsequent fragment aggregation processing.

[0026] After obtaining the student ID behavior record chain, the server performs fragment aggregation processing on the chain to generate student ID associated record fragments. A student ID associated record fragment refers to a local behavioral unit formed by a student ID card within a continuous time interval, encompassing the same spatial semantic location combination, the same behavioral semantic state combination, and the same background constraint state combination. The purpose of fragment aggregation processing is to compress high-frequency, fine-grained behavioral records into continuous units with overall semantics, facilitating the subsequent construction of higher-level order relationships. During implementation, the server sequentially reads each behavioral record from the student ID behavior record chain. If adjacent behavioral records are temporally continuous, belong to the same functional area or naturally adjacent area in spatial semantic location, exhibit the same behavioral trend in behavioral semantic state, and maintain consistent background constraint states, they are aggregated into the same student ID associated record fragment. Conversely, if there is a significant change in behavioral trend, a cross-functional jump in spatial semantic location, or a switch in background constraint states, the current student ID associated record fragment is terminated, and a new fragment is generated. The aggregation boundary can be constrained by a fragment consistency score, which can be expressed as:

[0027] in A consistency score indicating whether two adjacent behavioral records should be classified into the same student ID-related record segment; , , and These represent the weighting coefficients corresponding to temporal continuity, spatial continuity, behavioral consistency, and contextual consistency. It indicates the degree of temporal continuity between two adjacent behavior records, and is used to reflect whether there is a significant time break; It indicates the degree of spatial continuity between two adjacent behavior records, and is used to reflect whether they are in the same or reasonably adjacent spatial semantic positions; This indicates the consistency of behavior between two adjacent behavior records, reflecting whether they belong to the same behavioral trend; This indicates the consistency of background between two adjacent behavioral records, reflecting whether classroom time records, dormitory schedule records, campus activity records, and space control records remain continuous. After the server completes aggregation based on the consistency score, each student ID-related record fragment simultaneously contains the student ID master index, time interval, spatial semantic location, behavioral semantic state, neighborhood state, and background constraint state, becoming a more suitable structured intermediate object for subsequent analysis.

[0028] After generating student ID card associated record fragments, the server performs continuous continuation processing on these fragments to form a set of student ID card associated records with temporal continuation, spatial transfer, behavioral continuity, neighborhood evolution, and background switching relationships. The set of student ID card associated records refers to a complete set of records for the same student ID card organized according to a clear continuation logic over a relatively long observation period. Temporal continuation is used to characterize whether the next student ID card associated record fragment is continuous in time after the previous one ends; spatial transfer is used to characterize whether the spatial semantic positional change between two adjacent student ID card associated record fragments conforms to the actual campus pathways and order topology; behavioral continuity is used to characterize whether the behavioral trend between two consecutive student ID card associated record fragments has a natural extension; neighborhood evolution is used to characterize whether there are interpretable changes in accompanying relationships, group edge relationships, and group occlusion relationships; and background switching is used to characterize whether classroom time records, dormitory schedule records, campus activity records, and spatial control records trigger a switch in environmental constraints. During implementation, the server first sorts all student ID-related record fragments by time, then identifies the five types of relationships mentioned above for each adjacent fragment, constructing a connection graph between the fragments. To measure the overall connection strength between two fragments, a connection score can be constructed:

[0029] in This indicates the overall continuity score between two adjacent student ID record segments; , , , and Indicates the importance weight of various types of succession relationships; It indicates the temporal sequence and is used to reflect the degree of temporal connection between two segments; It indicates spatial transfer relationships and is used to reflect whether the spatial transformation between segments is reasonable; It indicates the continuity of behavior and is used to reflect whether the trends of previous and subsequent behaviors are naturally connected; It represents the neighborhood evolution relationship, used to reflect whether the accompanying objects and neighborhood structures have interpretable evolution; This indicates the background switching relationship, reflecting whether changes in environmental constraints support the switching between segments. The server establishes connections for segments with acceptance scores above a threshold, and treats segments with acceptance scores below the threshold as new behavioral starting points or abnormal breakpoints. The resulting set of student ID card association records not only preserves the continuous behavioral trajectory of the student ID card in the multi-source perception system, but also explicitly preserves the connection logic between segments, thus directly supporting subsequent construction of the main order graph, generation of student ID card supplementary segment sets, and shell extraction processing.

[0030] S120. Construct an order mother stream graph based on the student ID card associated record set, and map the student ID card associated record set to the order mother stream graph to generate a student ID card attached stream fragment set.

[0031] Specifically, the process begins by extracting order primitives from the student ID card-related record set to generate a set of order mother flow primitives. Order primitive extraction refers to the process of extracting the smallest normal order units with recurring, stable, and consistent background characteristics from a large number of student ID card-related record fragments. Order mother flow primitives are basic behavioral structural units that can stably represent the normal flow of campus behavior within a specific time interval, spatial semantic location, and specific background constraints. Each student ID card-related record fragment in the student ID card-related record set contains a time interval, spatial semantic location, behavioral semantic state, neighborhood state, and background constraint state. Therefore, the server first classifies and statistically analyzes all student ID card-related record fragments within an observation window jointly defined by classroom records, dormitory schedule records, campus activity records, and spatial control records. Then, it identifies behavioral combinations with high recurrence frequency, stable fragment continuity, and continuous occurrence across multiple observation periods, abstracting these behavioral combinations as order mother flow primitives. Spatial semantic location refers to location units with clear order connotations, such as the entrances to teaching buildings, dormitories, canteens, libraries, floor transition points, connecting corridors, access control areas, boundary zones, outer edges of sensitive areas, open convergence areas, and sparse transition areas. Background constraint state refers to the environmental state formed under the combined influence of classroom time constraints, dormitory schedule constraints, activity organization constraints, and spatial control constraints. To measure whether a certain type of student ID card-related record fragment is suitable as an order-based primitive, an order-based primitive extraction scoring system can be constructed:

[0032] in The score for extracting order primitives is used to measure whether the current combination of behaviors meets the conditions to become an order mother stream primitive. , , and The weight coefficients of different evaluation factors are used to adjust the influence of repetition degree, continuity stability degree, background consistency degree and neighborhood consistency degree in the extraction of order primitives. The values ​​of each weight coefficient can be set according to the campus business scenario. This indicates a recurring feature, used to reflect the frequency of the current combination of behaviors across multiple observation periods. The more frequently it occurs, the more likely it is to represent normal order. This indicates the stability of the connection between the current action combination and the preceding and following actions. The more stable the connection, the more suitable it is as a basic unit of order. This indicates a consistent background feature, used to reflect the degree of consistency between the current combination of behaviors and records from classroom periods, dormitory schedules, campus activities, and space management. This represents the neighborhood consistency feature, used to reflect whether the current behavior combination exhibits a stable accompanying, spreading, or declining pattern in the neighborhood state. The server includes behavior combinations with an extraction score higher than the extraction threshold into the order source primitive set, thereby obtaining the basic objects required for subsequent order relationship weaving processing.

[0033] After generating the set of order-based primitives, the server performs order relation weaving on the set to generate an order-based relation network. Order relation weaving refers to the process of establishing sequential succession relationships, parallel distribution relationships, accompanying diffusion relationships, stationary succession relationships, boundary fallback relationships, and restricted transfer relationships among the order-based primitives. The order-based relation network is a network structure composed of multiple order-based primitives and their interconnections, used to describe the evolution of the normal order flow on campus. In implementation, the server first calculates the co-occurrence of any two order-based primitives in different time windows based on the continuous succession results in the student ID card association record set. Then, combining the actual connectivity between spatial semantic locations, the permissibility of background constraint states, and the evolutionary consistency of neighborhood states, it determines whether a legitimate order relationship exists between the two order-based primitives. The relationships are categorized into several types: Successive succession, which refers to the relationship where a preceding order-matrix element typically leads to a subsequent order-matrix element; Parallel distribution, which refers to the relationship where two order-matrix elements occupy different spatial semantic positions within the same time window but collectively constitute the overall order flow; Accompanying diffusion, which refers to the relationship where a group expands its range as it migrates from one order-matrix element to another; Stationary succession, which refers to the relationship where transit-type order-matrix elements naturally transition to stationary-type order-matrix elements; Boundary retreat, which refers to the relationship where some behaviors naturally retreat to the boundary zone or transition area after the end of high-density flow; and Restricted transfer, which refers to order migration relationships that are only allowed under specific spatial control conditions. To measure the strength of the relationship between two order-matrix elements, a relationship weaving score can be constructed.

[0034] in The relation weaving score is used to measure whether an order relation should be established between two order parent stream primitives. , , and Indicates the weight coefficients of different evaluation items; This indicates the co-occurrence characteristic, used to reflect the frequency with which two order parent stream primitives appear together in the historical student ID related record set in either sequential or parallel form. It represents spatial topological features, used to reflect whether there is a real reachable path or natural connection between the spatial semantic locations of two order mother stream primitives; This indicates background compatibility characteristics, used to reflect whether two order parent stream primitives have a common environmental basis for being established under classroom time records, dormitory schedule records, campus activity records, and space control records; This represents the consistent evolutionary characteristics, reflecting whether two order-based mother stream primitives have a natural transitional relationship in terms of neighborhood states, behavioral trends, and density changes. The server establishes connections between order-based mother stream primitives based on relationship weaving scores, thereby forming a directional and binding order-based mother stream relationship network.

[0035] After forming the order mother flow relationship network, the server performs graph shaping processing based on the network to generate an order mother flow graph. Graph shaping processing refers to the process of organizing the order mother flow primitives and order relationships in the order mother flow relationship network into a graph model that can represent the overall normal order flow structure of the campus. The order mother flow graph is a dynamic order expression structure composed of order mother flow primitives and their order relationships under different time intervals, different spatial semantic locations, and different background constraints. In implementation, the server does not simply stack all order mother flow primitives and order relationships, but rather layers and merges them according to different order themes such as teaching switching flow, dormitory entry and exit flow, canteen dining flow, library stay flow, activity exit flow, and nighttime return flow. At the same time, each order mother flow primitive is assigned a time interval identifier, spatial semantic location identifier, behavioral semantic state identifier, neighborhood state identifier, and background constraint identifier, and each order relationship is assigned a receiving direction identifier, transfer condition identifier, and boundary restriction identifier. The direction of inheritance refers to the direction of the order flow represented by the order relationship; the transfer condition indicator refers to the temporal, spatial, or background conditions that must be met for a certain order relationship to be established; the boundary restriction indicator refers to the restrictive information that may prohibit or weaken a certain order relationship under a specific spatial control state. To evaluate the structural integrity of the formed map, a map integrity index can be constructed:

[0036] in The map completeness index is used to measure whether the order flow map can fully express the normal order flow structure of the campus. , , and Indicates the weight coefficients of different evaluation factors; Indicates the degree of coverage, used to reflect the sufficiency of the order mother flow map in covering the main spatial semantic locations and main time intervals; It indicates the degree of connectivity and is used to reflect whether the connecting network between the order mother stream primitives is complete; Indicates the degree of consistency, used to reflect whether the spectrum maintains a stable order across different observation periods; This indicates the degree of control adaptation, reflecting whether the map is consistent with the current spatial control records. After the map completeness meets the requirements, the server determines the generated result as the order mother map and uses it as the order reference structure for subsequent mapping of student ID card associated record fragments.

[0037] After the order-based flow graph is generated, the server maps student ID card-related record fragments from the student ID card-related record set to the order-based flow graph to generate corresponding mapping results. Mapping refers to the process of establishing a correspondence between student ID card-related record fragments and a specific order-based flow primitive or order relationship in the order-based flow graph. The mapping result refers to the order attribution result determined by the server for each student ID card-related record fragment, indicating whether the fragment is stably attached to a specific order-based flow primitive, migrates along a certain order relationship, or only forms an edge attachment at the order boundary. In practice, the server first filters candidate order-based flow primitives and candidate order relationships within the same observation window in the order-based flow graph based on the time interval of the student ID card-related record fragments. Then, it performs multi-dimensional matching based on the spatial semantic location, behavioral semantic state, neighborhood state, and background constraint state to determine the degree of consistency between the current student ID card-related record fragment and the candidate order objects. Stable attachment refers to the student ID-related record fragment maintaining consistency with a certain order-mainstream primitive across five dimensions: time, space, behavior, neighborhood, and context. Marginal attachment refers to the student ID-related record fragment maintaining consistency with a certain order-mainstream primitive across most dimensions, but exhibiting slight deviations in local behavior or neighborhood. Order migration refers to the student ID-related record fragment not corresponding solely to a single order-mainstream primitive, but rather existing in a transitional relationship between two order-mainstream primitives. To measure the degree of mapping matching, a mapping score can be constructed.

[0038] in This represents a mapping score, used to measure the degree of matching between a student ID-related record fragment and a candidate order object; , , , and Represents the weight coefficients corresponding to time matching, spatial matching, behavior matching, neighborhood matching, and background matching; Indicates time matching degree, which reflects the degree of consistency between the time interval of the student ID-related record segment and the time interval of the candidate order object; Spatial matching degree is used to reflect the degree of proximity or consistency between the spatial semantic location of the student ID-related record fragment and the spatial semantic location of the candidate order object. The behavior matching degree reflects the consistency between the behavioral semantic state of the student ID-related record fragment and the behavioral pattern represented by the candidate order object. The neighborhood matching degree reflects the consistency between the solitary, accompanying, group edge, or group occlusion states of the student ID-related record fragments and the neighborhood patterns of the candidate order objects. The background matching score reflects the consistency between the background constraint state of the student ID-related record fragment and the environment in which the candidate order object is established. The server takes the candidate order object with the highest mapping score and that meets the mapping threshold as the final mapping result, thereby completing the assignment mapping of the student ID-related record fragment to the main order graph.

[0039] After obtaining the mapping results, the server performs attached flow segmentation processing on the student ID card associated record fragments based on the mapping results to generate a set of student ID card attached flow fragments. Attached flow segmentation processing refers to the process of combining multiple adjacent student ID card associated record fragments into higher-level attached behavior segments according to the continuity of their affiliation, migration coherence, and behavioral continuity in the order flow graph. A set of student ID card attached flow fragments refers to a collection of several student ID card attached flow fragments, where each fragment represents a local behavioral process of a student ID card attaching to the normal order flow within a certain continuous time period. A student ID card attached flow fragment refers to a continuous attached behavior unit formed by student ID card associated record fragments continuously attaching to the same order flow primitive or naturally migrating along a legitimate order relationship in the order flow graph. During implementation, the server first organizes the mapping results according to the student ID card master index, and then checks in chronological order whether the mapping affiliation of adjacent student ID card associated record fragments is continuous, whether the order relationship is coherent, and whether the behavioral trend remains in the same direction. If several adjacent student ID card associated record segments are continuously mapped to the same order stream primitive, or can naturally migrate along the legal order relationship in the order stream graph, and belong to the same behavioral direction in terms of behavioral semantic state, then they are merged into the same student ID card attached stream segment. If a sudden change occurs in spatial semantic location, behavioral semantic state, neighborhood state, or background constraint state during the merging process, then the current student ID card attached stream segment ends and a new student ID card attached stream segment begins. To measure whether adjacent mapping results should be merged into the same student ID card attached stream segment, an attached stream continuity score can be constructed:

[0040] in The attached stream continuity score is used to measure whether adjacent mapping results should be merged into the same attached stream segment of the student ID card; , , and These represent the weighting coefficients corresponding to the continuity of order primitives, the coherence of order relations, the continuity of behavior, and the stability of state. It represents the continuity of the order primitive, and is used to reflect whether adjacent student ID card related record fragments are continuously attached to the same order mother stream primitive; It indicates the continuity of order relationships and is used to reflect whether there are inheritance or migration relationships allowed by the order mother graph between adjacent student ID card related record segments; Indicates behavioral continuity, used to reflect whether adjacent student ID record segments maintain consistency in behavioral trends such as traversing, stopping, falling back, or spreading; The stability of the state is represented by the score indicating whether there have been no destructive changes in the neighboring state and background constraint state of adjacent student ID card associated record segments. After the server completes the attached flow segmentation process based on the attached flow continuity score, it obtains the set of student ID card attached flow segments. This set of student ID card attached flow segments directly inherits the previous mapping results, can clearly characterize the way student ID cards are attached to the normal order flow, and provides a unified, continuous, and ordered semantic input object for subsequent shell extraction processing.

[0041] S130. Perform shell extraction processing on the student ID card attached fragment set to obtain the extraction result, and perform order anti-entropy weaving processing on the extraction result to generate reverse order silk chain.

[0042] Specifically, the server first performs a shell-borrowing candidate screening process on the student ID card-related flow fragment set to generate a shell-borrowing candidate fragment set. The student ID card-related flow fragment set characterizes the continuous attachment behavior of student ID cards in the main flow graph of the order. The shell-borrowing candidate screening process identifies which student ID card-related flow fragments, while superficially still attached to the normal order flow, have significantly dependent on the support of an external order shell for their attachment stability. The order shell refers to the external normal order conditions that can provide a superficially normal explanation for the student ID card-related flow fragments, including group passage shells, access control window shells, activity dispersal shells, channel receiving shells, and boundary fallback shells. During implementation, for each student ID card-related flow fragment, the server jointly examines its group occlusion dependency, access control window dependency, main flow residual dependency, boundary fallback dependency, background constraint dependency, and the continuity and stability of the flow, and constructs a shell-borrowing candidate comprehensive discrimination function. This function is not a simple addition, but simultaneously considers the interactive coupling between dependency strengths, the time persistence effect, and the spatial adjacency amplification effect, enabling more sensitive identification of shell-borrowing behavior hidden behind the superficially normal flow. The comprehensive discriminant function for backdoor listing candidates can be expressed as:

[0043] in, Indicates the first The comprehensive score of the student ID card incidental fragment is used to measure the overall dependence of the student ID card incidental fragment on the outer shell of order. The higher the score, the more likely it is to belong to the incidental fragment. Indicates the first The weighting coefficients of class-dependent components are used to adjust the contribution intensity of different order shell dependency factors in the overall score; Indicates the first The student ID card fragment is in the... The dependency quantity on the class order shell, when When the corresponding group occlusion dependency, when The corresponding access control window dependency quantity, when When the corresponding mainstream residual dependency is, The corresponding boundary fall-back dependency, when The corresponding background constraint dependency; This represents the nonlinear amplification index of various dependencies, used to enhance the distinguishing ability under high dependency conditions, and its value is greater than or equal to 1. The duration enhancement term is used to characterize the duration of the student ID card's accompanying stream segment in the shell-borrowing state and the degree of abnormal deviation from the duration of the normal accompanying stream. The spatial adjacency enhancement term is used to characterize the degree of adhesion between the student ID card's attached segment and highly sensitive locations such as the access control front area, boundary zone, outer edge of sensitive area, and floor transition point; This represents a dependency coupling enhancement term, used to characterize the amplification effect when multiple shell dependency factors exist simultaneously. For example, when group occlusion dependency and access control window dependency occur simultaneously, the shell borrowing discrimination strength will be increased. , and These represent the adjustment coefficients for the temporal persistence enhancement term, spatial adjacency enhancement term, and dependency coupling enhancement term, respectively. Indicates the first The drift stability of each student ID card drift segment is used to characterize its surface attachment stability in the order mother current map. This represents the reference threshold for the stability of the drift, used to distinguish between a normally stable drift and a pseudo-stable drift that depends on the shell for maintenance. This represents the compression coefficient in the denominator, used to control the suppression strength of the attached stream stability on the overall score. The server includes the attached stream segments of student ID cards with overall scores higher than the screening threshold into the attached stream segment set for potential attached streams, thus completing the first round of screening for potential attached stream behaviors.

[0044] After generating a set of candidate shell-borrowing segments, the server performs shell dependency decomposition processing on the candidate segment set to generate a shell dependency identifier set. Shell dependency decomposition processing clarifies which order shells each candidate segment depends on, the degree of dependency, and the time interval and spatial semantic location where the dependency relationship mainly occurs. The shell dependency identifier set records the shell type, shell strength, and shell location of each candidate segment, serving as the direct basis for subsequent order shell stripping processing. In practice, the server constructs a complete order interpretation field for each candidate segment, then sequentially removes group occlusion conditions, access control window conditions, activity dissipation conditions, channel connection conditions, and boundary fallback conditions, comparing the consistency differences between the removed and order parent flow maps. Simultaneously, to avoid localized random fluctuations caused by single removals, the server also calculates the joint removal impact between various order shells, thereby identifying the dominant and secondary dependent shells. The shell dependency strength function can be expressed as:

[0045] in, Indicates the first The candidate fragment for backdoor listing is the first The dependency strength of the class-order shell is used to quantify the degree of dependency of the current shell-borrowing candidate fragment on a specific order shell; Indicates the first The overall consistency between a candidate fragment for shell borrowing and the mother flow map of order under the condition of a complete order shell. The higher the consistency, the closer it is to the normal order interpretation on the surface. Indicates the removal of the first After the class order shell, the first The consistency between candidate segments of shell acquisition and the mother flow map; This represents the path explanation loss, used to characterize the loss after removing the first path. After the class order shell, the degree of lack of explanation in the path continuation of the current shell candidate fragment; Represents the neighborhood explanation loss, used to characterize the removal of the th After the class order shell, the degree of explanation missing in the current shell candidate fragment regarding the accompaniment relationship or the occlusion relationship; This represents the access control interpretation loss, used to characterize the removal of the first... Behind the shell of the class order, the degree of lack of explanation for the current candidate fragment for shell acquisition in terms of entry timing, reasonableness of stay, or non-entry behavior; , and These represent the enhancement coefficients for path interpretation loss, neighborhood interpretation loss, and access control interpretation loss, respectively. Indicates the first Class order shell and the first The coupling weights between class-order shells are used to characterize the substitutability or synergy between different order shells; the overall term in the denominator is used to suppress spurious high dependency values ​​caused by synchronous fluctuations of other shells, making the dominant shell dependency easier to identify. The server generates a set of shell dependency identifiers based on the various dependency strengths, so that each shell-borrowing candidate fragment has a dominant shell dependency identifier, a secondary shell dependency identifier, and a corresponding dependency strength.

[0046] After forming the shell dependency identifier set, the server performs order shell stripping processing on the shell-borrowing candidate fragment set based on the shell dependency identifier set to generate a naked fragment set as the extraction result. Order shell stripping processing is used to separate the explanatory components provided by the normal order shell from the student ID card's own behavioral components in the shell-borrowing candidate fragments. The naked fragment set is used to retain the independent behavioral residuals that still exist after the student ID card loses the support of the order shell, and is the direct input for subsequent naked association reorganization processing. Naked fragments refer to the attachment, hysteresis, probing, reversal, cross-layer borrowing, and edge sliding behavioral fragments that remain after removing normal order explanations such as group occlusion, access control residuals, mainstream residuals, channel acceptance, and boundary fallback. In implementation, the server adopts a differentiated stripping strategy for different shell-borrowing candidate fragments based on the shell dependency identifier set, and decomposes the original attached behavior into shell explanation components and naked behavioral components. Considering the continuous propagation of the shell explanation components in time and space, the server uses a spatiotemporal coupling decomposition function to strip them. The naked residual function can be expressed as:

[0047] in, Indicates the first The intensity of the naked meaning residual of a candidate segment after stripping away the outer shell of order is used to quantify the degree to which the real behavioral tendencies are preserved in the candidate segment. and They represent the first The start and end times of each candidate segment for backdoor listing; Indicates the first A candidate for a backdoor listing at the moment The original behavior state vector can be composed of multi-dimensional behavior features such as position offset, dwell intensity, return degree, accompanying state and access control relationship. Indicates the first A candidate for a backdoor listing at the moment The explanatory state vector provided by the shell of order; The magnitude of the difference between the original behavior state vector and the shell interpretation state vector is used to characterize the behavior residual that cannot be explained by the order shell at that moment; This represents the difference amplification index, used to enhance the contribution of high deviation behavior to the residual amount; The group occlusion intensity function is used to characterize time. The degree to which the surrounding crowd suppresses the prominence of the student ID; This represents the main residual current intensity function, used to characterize time. The degree to which the surrounding normal order supports the explanation of student ID behavior; The access control window strength function is used to characterize time. The extent to which continued access control or residual authorization supports student ID behavior; , and These represent the suppression coefficients of group occlusion intensity, mainstream residual intensity, and access control residual window intensity, respectively. The server retains behavioral fragments with naked residual intensity higher than the extraction threshold and merges them into a naked fragment set, thereby completing the transformation from shell-borrowing candidate fragments to extraction results.

[0048] After generating a set of naked intention fragments, the server performs naked intention association recombination processing on the set to generate naked intention fragment chains. Naked intention association recombination processing reorganizes scattered naked intention fragments according to the continuity of latent intentions, transforming the originally intermittent naked intention manifestation process into a continuous chain with an evolutionary direction. A naked intention fragment chain represents a continuous sequence of naked intentions formed by the same student ID card around the same potential target location, the same probing direction, or the same latent intention. In practice, the server first categorizes and organizes naked intention fragments according to the student ID card master index, and then comprehensively considers the temporal proximity, spatial adjacency, target tendency consistency, probing pattern repetition, and behavioral energy transfer between naked intention fragments to determine which naked intention fragments should be recombined into the same naked intention fragment chain. The degree of behavioral energy transfer refers to whether the probing trend already revealed in the previous naked intention fragment continues or strengthens in the subsequent naked intention fragment. The naked intention recombination scoring function can be expressed as:

[0049] in, Indicates the first The first nude scene and the first The recombination score between naked fragments is used to measure whether they should be connected to the same naked fragment chain; Indicates temporal proximity, used to reflect the closeness between the times when two naked segments occur; It indicates the degree of spatial adjacency, used to reflect the proximity and accessibility between the spatial semantic locations of two naked semantic fragments; It indicates the degree of consistency of the target and is used to reflect whether two naked meaning segments point to the same access control node, the outer edge of a sensitive area, a floor transition point, or a passage corner area; Indicates the degree of repetition of trial patterns, used to reflect the similarity between two naked meaning segments in behaviors such as stopping, attaching, moving around, and turning back; It indicates the degree of behavioral energy transfer and is used to characterize whether the latent tendency in the previous naked intention segment continues and is enhanced in the subsequent naked intention segment; , , , and These represent the weighting coefficients for temporal proximity, spatial adjacency, target convergence, repetition of trial patterns, and transfer of behavioral energy, respectively. This indicates the magnitude of change in background constraints between two naked image segments, used to characterize the degree of change in classroom time records, dormitory schedule records, campus activity records, or space control records; This represents the magnitude of neighborhood state changes between two naked fragments, used to characterize the degree of changes in associated states, group edge states, and group occlusion states. The server sequentially concatenates naked fragments with recombination scores higher than the recombination threshold to generate a naked fragment chain.

[0050] After forming the naked meaning fragment chain, the server performs order consumption identification processing on the naked meaning fragment chain to generate an anti-entropy unit set. Order consumption identification processing is used to identify the specific ways in which the naked meaning fragment chain gradually weakens the normal order explanatory power in the order mother flow graph. The anti-entropy unit set is used to record the order reverse consumption actions implemented by the student ID during the incubation process, and is the basic particle for subsequent directional weaving processing. An anti-entropy unit refers to the smallest behavioral unit that can independently characterize a decrease in order explanatory power, such as early decoupling anti-entropy unit, delayed decoupling anti-entropy unit, secondary adjoining anti-entropy unit, boundary attachment anti-entropy unit, residual window retention anti-entropy unit, reverse return anti-entropy unit, cross-layer bypass anti-entropy unit, and crack traversal anti-entropy unit. During implementation, the server calculates the explanatory loss of each naked meaning fragment in the naked meaning fragment chain on the order mother flow graph in terms of path continuity, temporal rhythm, spatial occupancy, neighborhood cooperation, and background self-consistency, and maps the loss pattern to the corresponding anti-entropy unit category. To characterize this order consumption process, the server constructs an anti-entropy intensity function. The anti-entropy intensity function can be expressed as:

[0051] in, Indicates the first The anti-entropy strength of each candidate anti-entropy behavior is used to measure the degree to which the behavior weakens the explanatory power of order. Indicates the first The weighting coefficients of the loss term are explained in the class definition; Indicates the first The candidate anti-entropy behavior in the th... The loss in the class order explanation dimension, when The corresponding path carrying loss amount, when The corresponding time rhythm loss, when The corresponding space occupation loss when The corresponding neighborhood cooperative loss amount, when Corresponding background self-consistent loss amount; The nonlinear amplification index represents various loss quantities and is used to amplify the impact of high-loss behavior on the total anti-entropy intensity. This indicates a repeated trial enhancement term, used to characterize whether the candidate anti-entropy behavior occurs repeatedly in multiple time windows; This represents the target drift enhancement term, used to characterize whether the candidate anti-entropy behavior exhibits a continuous drift trend from the normal order region to the sensitive region, boundary zone, or access control front region; This indicates a latent enhancement term, used to characterize whether the candidate anti-entropy behavior remains in a low-significance but persistent latent state over a long period of time; , and These represent the adjustment coefficients for the repeated trial enhancement term, the target drift enhancement term, and the latent persistence enhancement term, respectively. The server determines the anti-entropy unit category based on the anti-entropy strength and loss pattern, and includes behaviors that meet the identification criteria into the anti-entropy unit set.

[0052] After obtaining the set of anti-entropy units, the server performs directional weaving processing based on the set to generate reverse-order chains. Directional weaving connects multiple anti-entropy units into a continuous reverse-evolutionary chain of order, following the direction of continuously decreasing explanatory power. The reverse-order chains characterize how a student ID card evolves from superficially conforming to normal order to revealing its true latent intentions, serving as direct input for subsequent proxy projection processing. In practice, the server first determines the original location of each anti-entropy unit in the order matrix, its dominant loss dimension, and its target direction. Then, it analyzes whether there is a directional connection between different anti-entropy units, progressing step-by-step from high-explanatory appendages to low-explanatory naked intentions. If a student ID card first shows an anti-entropy unit that prematurely detaches, then a boundary-attached anti-entropy unit, then a residual anti-entropy unit, and finally a reverse-returning anti-entropy unit, it indicates that it is continuously consuming normal order explanatory power, and the server weaves it into the same reverse-order chain. If multiple anti-entropy units, although distributed in different time windows, repeatedly appear around the outer edge of the same sensitive area, the same access control front area, or the same floor transition point, the server connects them across windows through target convergence consistency and anti-entropy progression. The directional weaving scoring function can be expressed as:

[0053] in, Indicates the first The anti-entropy unit and the first The directional weaving score between anti-entropy units is used to measure whether they should be connected to the same reverse-order filament chain; Indicates directional consistency, used to reflect whether two anti-entropy units are consistent in the direction of weakening order explanatory power; It represents temporal continuity and is used to reflect whether two anti-entropy units are continuous in time or have the conditions for cross-window connection; It represents spatial adjacency, used to reflect whether there is an adjacency or transferability relationship between the spatial semantic locations of two anti-entropy units; This indicates that the goals tend to be consistent, and is used to reflect whether two anti-entropy units are pointing to the same potential target location; It represents the progressiveness of anti-entropy, used to reflect whether two anti-entropy units form a progressive relationship from weak to strong in terms of the intensity of order consumption; , , , , , and These represent the weighting coefficients of each factor; This represents the overall amplification index, used to enhance the contribution of high-matched anti-entropy units to the weaving score; and They represent the first The anti-entropy unit and the first The parameters of the manifest stage of each anti-entropy unit are used to characterize the sequence of stages it occupies during the latent evolution process; and They represent the first The anti-entropy unit and the first The dominant shell dependency parameter corresponding to each anti-entropy unit is used to characterize which type of order shell stripping it mainly originates from. The server completes the sequential connection and cross-window connection of the anti-entropy units based on the directional weaving score, and finally generates the reverse-order silk chain.

[0054] S140. Perform substitute projection processing on the reversed silk chain to obtain the projection result, and perform intention cavity mining processing on the projection result to generate the intention cavity region.

[0055] Specifically, the server first performs an interpretation consistency decomposition process on the reverse-order silk chain to generate a set of interpretation-consistent segments and a set of interpretation-missing segments. The reverse-order silk chain refers to a continuous chain structure woven from multiple anti-entropy units in a direction that continuously weakens the explanatory power of order, used to characterize the evolution of a student ID card from superficially attaching to the normal order to gradually revealing its latent intentions. The interpretation consistency decomposition process involves re-mapping each segment of the reverse-order silk chain onto the main flow graph of order, determining whether it can still be continuously explained by the normal order structure, and thus decomposing the reverse-order silk chain into interpretable and uninterpretable parts. The set of interpretation-consistent segments refers to the set of segments in the reverse-order silk chain that, although anti-entropy changes have occurred, can still find a continuous path and reasonable background support in the main flow graph of order; the set of interpretation-missing segments refers to the set of segments in the reverse-order silk chain that can no longer be continuously explained by the main flow graph of order, and whose behavior no longer depends on the logic of normal order. During implementation, the server calculates the path continuity consistency, temporal rhythm consistency, spatial occupancy consistency, neighborhood coordination consistency, and background constraint consistency for each reverse-order silk chain segment. These consistency results are then combined to form an interpretation consistency score, which is compared with a preset interpretation threshold to complete the chain segment decomposition. The interpretation consistency score function can be expressed as:

[0056] in, Indicates the first The consistency of interpretation of each reversed filament chain segment is used to measure whether the segment can still be interpreted by the order mother flow map; Indicates the first The weighting coefficients of class consistency factors are used to adjust the contribution intensity of different order explanation dimensions to the degree of explanation consistency; Indicates the first The segment in the first... Consistency quantity in the class order interpretation dimension, when The corresponding path carries a consistent quantity, when The time corresponds to the consistency of the time rhythm, when The corresponding space occupancy consistency quantity, when The corresponding neighborhood consensus quantity, when The consistency quantity corresponding to the background constraints; The non-linear amplification index represents various consistency quantities and is used to enhance the ability to distinguish between high-consistency chain segments and low-consistency chain segments. This represents the coupling enhancement term between multidimensional consistency, used to reflect the overall amplification effect when multiple explanatory dimensions are kept consistent simultaneously; Indicates the first The path continuity of a chain segment in the order mother flow map is used to characterize whether the chain segment can find a continuous preceding and following order continuity path. This represents a reference threshold for path integrity, used to distinguish between a path that is interpretable and a path that is mismatched. Indicates the first The background matching completeness between each chain segment and classroom time records, dormitory schedule records, campus activity records, and space management records is used to characterize whether the chain segment is still within the range of normal background support. This represents the reference threshold for background matching completeness, used to distinguish between background interpretable states and background missing states; This represents the compression factor, used to control the adjustment strength of path continuity completeness and background matching completeness on overall interpretation consistency. The server includes chain segments with interpretation consistency higher than the decomposition threshold into the interpretation consistent segment set, and chain segments with interpretation consistency lower than the decomposition threshold into the interpretation missing segment set, thus completing the first step of structural decomposition of the reversed silk chain.

[0057] After generating the set of consistent interpretation segments, the server performs a substitute chain construction process based on the consistent interpretation segment set to generate a substitute chain. A substitute chain is a continuous chain formed by reorganizing the chain segments in the consistent interpretation segment set, used to represent the normal order appearance structure that the student ID card still maintains to the outside world. The substitute chain construction process refers to the process of reconnecting the chain segments in the consistent interpretation segment set according to the original order continuity direction, while maintaining the interpretability of the main order graph, and identifying which chain segments play a masking role and which play a continuity role. Masking refers to a consistent interpretation segment surrounding a missing interpretation segment in time or space, making the missing interpretation segment difficult to directly identify; continuity refers to a consistent interpretation segment maintaining legitimate transfer within the main order graph, thereby continuously disguising the student ID card behavior within the normal order flow. During implementation, the server first performs order mapping on each chain segment in the set of consistent interpretation segments, repositioning them onto the corresponding order source primitives or order relations in the order source graph. Then, it establishes substitute chain connections based on the temporal continuity, spatial continuity, background consistency, and order continuity direction between chain segments. To quantify the substitute maintenance capability of a particular consistent interpretation segment within the substitute chain, a substitute maintenance strength function can be constructed:

[0058] in, Indicates the first The Stand maintenance strength of a consistent segment is used to measure the ability of that consistent segment to maintain a normal, orderly appearance within the Stand chain. Indicates the first Weighting coefficients of maintenance factors; Indicates the first The consistent explanation in the first paragraph Performance on maintenance factors, when The corresponding path is continuously maintained when The corresponding behavior is consistent with the maintenance quantity, when The corresponding background self-consistency maintenance quantity, when The corresponding neighborhood stability maintenance quantity; The nonlinear amplification index represents various maintenance quantities and is used to enhance the identification effect of high maintenance capacity segments; This indicates the occlusion enhancement term, which characterizes the degree to which the consistent segment of the explanation surrounds the adjacent missing segments of the explanation in time and space; The term "order continuation enhancement" is used to characterize the ability of the explanatory consistency segment to provide continuous continuation between preceding and following order parent stream primitives. and These represent the adjustment coefficients for the occlusion enhancement term and the order continuation enhancement term, respectively. The server assigns roles to the interpretation consistency segments based on the substitute maintenance strength and connects interpretation consistency segments with continuous continuation relationships into substitute chains, making the substitute chains an important outer structure for subsequent substitute traction identification processing and interpretation gap identification processing.

[0059] After generating the set of missing explanatory segments, the server performs ontology chain construction processing based on the set to generate an ontology chain. An ontology chain is a continuous chain formed by reorganizing the segments in the set of missing explanatory segments according to the actual behavioral trends of the student ID card. It represents the latent behavioral mainline that persists even after the student ID card loses its normal order shell support. The ontology chain construction process refers to the process of connecting multiple missing explanatory segments into a latent behavioral main chain, no longer relying on the normal continuity logic of the order matrix, but solely based on the temporal proximity, spatial adjacency, target tendency, and explicit continuity relationships between the missing explanatory segments. Explicit continuity relationships refer to the continued existence or further enhancement of attachment, lag, retreat, or probing behaviors already exposed in the previous missing explanatory segment in the subsequent missing explanatory segment in the same manner. During implementation, the server calculates the intensity of its deviation from the order explanation, its approach intensity to the outer edge of sensitive areas or the access control front area, and its degree of repeated exposure across time windows for each missing explanatory segment, and combines these quantities for ontology chain connection determination. To quantify the dominance of a missing segment in the ontology chain, an ontology exposure strength function can be constructed:

[0060] in, Indicates the first The ontological exposure intensity of a missing explanatory segment is used to measure the core degree to which the missing explanatory segment represents the real latent behavior in the ontological chain. Indicates the first Weighting coefficients for class-specific factors; Indicates the first The missing explanatory segment is in the... The magnitude of the revealed factor, when When the corresponding boundary is attached, the exposure amount is as follows: When the corresponding access control test exposure amount, when The corresponding reverse foldback exposure amount, when The corresponding amount of lag and hysteresis is displayed at that time; Nonlinear amplification index representing various exposure quantities; This indicates a repetitive exposure enhancement term, used to characterize the degree to which the explained missing segment recurs across multiple time intervals; This indicates a convergence enhancement term, used to characterize the degree to which the missing segment of the explanation continuously approaches the outer edge of the sensitive area, the area before the access control, or the floor transition point; This indicates the suppression term for occlusion residue, used to characterize how much occlusion effect provided by the substitute chain still remains in the missing segment of the explanation. The stronger the occlusion, the more the exposure of the subject needs to be suppressed. , and These represent the adjustment coefficients for the repetitive exposure enhancement term, the proximity enhancement term, and the occlusion residual suppression term, respectively. The server constructs an ontology chain based on the connectivity between the ontology exposure intensity and the explanation of missing segments, making the ontology chain a continuous representation of the student ID's true latent behavior.

[0061] After generating the substitute chain and the original chain, the server performs substitute traction identification processing on the substitute chain and the original chain to generate a set of substitute traction relationships. Substitute traction identification processing refers to the process of identifying the association methods between the substitute chain and the original chain in terms of temporal coverage, spatial encirclement, order inheritance, and neighborhood dilution, and expressing these association methods in a structured manner. The set of substitute traction relationships records how the substitute chain forms occlusion traction, inheritance traction, guidance traction, and dilution traction with the original chain. Occlusion traction refers to the substitute chain temporally encircling the original chain or spatially covering the original chain, thereby reducing the visibility of the original chain; inheritance traction refers to the substitute chain maintaining legitimate order inheritance to cover the original chain across multiple order segments; guidance traction refers to the substitute chain guiding the original chain to the outer edge of a sensitive area or the area before access control through a normal movement direction on a certain surface; dilution traction refers to the substitute chain reducing the behavioral salience of the original chain by using group accompaniment or group edge states. During implementation, the server calculates the temporal coverage, spatial enclosure, order succession guidance, and neighborhood dilution for each segment of the substitute chain and each segment of the original chain, and determines whether a substitute traction relationship exists based on these metrics. The substitute traction strength function can be expressed as:

[0062] in, Indicates the first The first substitute chain segment and the first The substitution traction strength between individual body segments is used to measure whether there is a significant substitution traction relationship between them; , , , and Represents the weighting coefficients of different traction factors; Indicates time coverage, used to reflect the degree to which the substitute chain surrounds the main chain in time; Indicates spatial enclosure, used to reflect the degree of adjacency or coverage of the ontology segment by the surrogate segment in spatial semantic location; The order acceptance guidance level indicates whether the substitute chain segment provides an entry or exit channel for the main chain segment by maintaining legitimate order acceptance. This indicates neighborhood dilution, used to reflect whether the surrogate segment reduces the salience of the original segment through accompanying states or group masking states; This represents the overall coverage enhancement term, used to characterize the synergistic amplification effect when temporal coverage, spatial enclosure, and path guidance coexist. , , and These represent the nonlinear amplification exponents of various traction factors. The server pairs substitute chain segments with traction strength exceeding the identification threshold with the original chain segments to form a set of substitute traction relationships, providing structural constraints for subsequent interpretation of gap identification processing.

[0063] After generating the set of substitute traction relationships, the server performs interpretation gap identification processing based on the substitute chain and the ontology chain to generate an interpretation gap set. Interpretation gap identification processing refers to the process of identifying which spatial locations, temporal intervals, and behavioral states, although enveloped by the substitute chain, still cannot be reasonably explained by the order mother current graph, provided that the substitute chain has already provided the outer normal order appearance and the ontology chain has already represented the inner true latent trend. The interpretation gap set refers to a set composed of multiple interpretation gaps. An interpretation gap refers to a local blank unit in the ontology chain that, under the coverage of the substitute chain, still cannot be explained by normal order logic, normal background constraint logic, and normal neighborhood evolution logic. In implementation, the server uses the substitute chain as the outer envelope and the ontology chain as the inner revealing main line, examining each segment of the ontology chain segment by segment, calculating its path gap quantity, background gap quantity, neighborhood gap quantity, and behavioral gap quantity in the order mother current graph, and combining this with the set of substitute traction relationships to determine whether the ontology chain segment belongs to a local behavioral region that is covered by the substitute chain but still cannot be explained. The notch strength function can be expressed as:

[0064] in, Indicates the first The strength of the explanatory gap corresponding to each ontology segment is used to measure the degree to which the ontology segment forms an explanatory gap; Indicates the first Weighting coefficients for gap-like factors; Indicates the first The first body chain segment in the... The magnitude of the gap factor, when The corresponding path gap amount, when The corresponding background gap amount, when When the corresponding neighborhood gap amount, The corresponding behavior gap amount; The nonlinear amplification index represents various types of gap quantities; This indicates the Stand traction enhancement term, used to characterize whether the body chain segment is strongly obscured by the Stand chain. The stronger the obscuration but the gap still exists, the more valuable the interpretation gap is. This represents the residual explanation suppression term, used to characterize how much of the ontology segment can still be barely explained by the normal order. The more residual explanations there are, the stronger the explanation gap should be appropriately reduced. and These represent the adjustment coefficients of the surrogate traction enhancement term and the residual explanation suppression term, respectively. The server extracts ontology segments with explanation gap strength exceeding the identification threshold as explanation gaps and organizes them into an explanation gap set, thereby completing the explicit extraction of the inner unexplainable behavior region.

[0065] After generating the set of explanatory gaps, the server performs cavity aggregation processing on the set to generate intention cavity regions. Cavity aggregation processing refers to the continuous merging and overall convergence of multiple explanatory gaps that are temporally adjacent, spatially contiguous, share the same target tendency, and are similar in the nature of their explanatory deficiencies, forming a stable region that can represent the concentrated and revealed area of ​​latent intent on the student ID card. The intention cavity region refers to the explanatory deficiencies region formed by the aggregation of multiple explanatory gaps, which continuously exists within the envelope of the substitute chain and is supported by the actual ontology chain. It is used to represent the intention blank positions continuously created by the student ID card within the normal order appearance. Intention blank positions are not simply missing data, but latent behavioral areas with continuous behavioral traces that cannot be explained by the logic of the normal order. In implementation, the server first classifies the set of explanatory gaps according to the student ID card master index, then clusters them based on the temporal continuity, spatial contiguity, target tendency consistency, and gap nature similarity between explanatory gaps, and further performs cross-window reconnection along the time axis to avoid the same latent intent being scattered into multiple isolated explanatory gaps. To quantify whether multiple explanatory gaps should converge into the same intended cavity region, a cavity convergence strength function can be constructed:

[0066] in, Indicates the first The aggregation strength of the clusters that form the intended cavity region is used to measure whether the clusters are sufficient to constitute an independent intended cavity region. Indicates the first Weighting coefficients of aggregation factors; Indicates the first The explanatory gap cluster in the first The magnitude of the aggregate factor, when Time corresponds to a continuous quantity in time, when The corresponding spatial adjacency quantity, when When the corresponding target tends to be consistent, when The time corresponds to the similarity of the gap properties; Indicates the nonlinear amplification index of various aggregation quantities; This indicates a recurring enhancement term, used to characterize whether the explanatory gap clusters recur across multiple time intervals; This represents a dense clustering enhancement term, used to characterize whether multiple explanatory gaps form high-density convergence in local regions; and These represent the adjustment coefficients for repetitive enhancement terms and dense clustering enhancement terms, respectively. The server identifies explanatory gap clusters with aggregation intensity exceeding the cavity formation threshold as intended cavity regions, thereby completing the structural uplift from the explanatory gap set to the intended cavity region.

[0067] S150. Perform circumstantial evidence recall processing around the intention cavity area to obtain the recall result, and perform latent event incubation processing based on the recall result to generate the latent event protoker.

[0068] Specifically, the server first performs recall boundary generation processing around the intent cavity region to form a supporting evidence recall domain. The intent cavity region refers to an area of ​​missing explanation formed by the aggregation of multiple explanatory gaps, located within the envelope of the substitute chain and truly supported by the ontology chain. It represents the latent intent gaps continuously created by the student ID card within the normal order appearance. The recall boundary generation process refers to the process of setting the scope of supporting evidence retrieval around the intent cavity region in four dimensions: time, space, neighborhood, and order. The supporting evidence recall domain refers to the limited area subsequently used to retrieve peripheral supporting traces. It is not a simple spatial demarcation result, but a composite retrieval domain that simultaneously includes time boundaries, spatial boundaries, neighborhood boundaries, and order boundaries. During implementation, the server first reads the dominant spatial semantic location, dominant temporal distribution, dominant occlusion method, dominant anti-entropy source, and associated substitute chain and ontology chain corresponding to the intention cavity region. Then, it extends forward along the time axis to the moment the substitute chain begins to form, and backward to the moment the ontology chain significantly converges. Along the spatial axis, it covers the access control front area, boundary zone, outer edge of sensitive areas, floor transition points, passage corner areas, and open convergence areas directly adjacent to the intention cavity region. Along the neighborhood axis, it covers student ID objects that have accompanying traction, occlusion traction, and guiding traction relationships with the intention cavity region. Along the order axis, it covers the order range corresponding to the order parent flow primitive where the intention cavity region is located and its adjacent order relationships. To prevent the recall range from being too large and causing invalid circumstantial evidence generalization, the server performs joint contraction and expansion of the four types of boundaries to construct a recall domain strength function:

[0069] in, Indicates the first The strength of the circumstantial recall domain corresponding to each intention cavity region is used to measure the recall scope to which the intention cavity region should be expanded. These represent weighting coefficients for different boundary dimensions, used to adjust the influence of temporal boundaries, spatial boundaries, neighborhood boundaries, and order boundaries in the generation of the recall domain; Indicates the first The intentional cavity region in the first The basic extension quantity at the class boundary dimension, when The corresponding time boundary expansion amount, when The corresponding spatial boundary expansion amount, when The corresponding neighborhood boundary expansion amount, when The corresponding order boundary expansion amount; Nonlinear amplification exponents representing various basic expansion quantities are used to enhance recall boundary responses under high latent risk conditions; This indicates the Stand traction enhancement term, which characterizes the degree of Stand chain obscuring the current intention cavity area, and the more the recall domain needs to be amplified. The term representing the anti-entropy source enhancement is used to characterize the current intention cavity region. If it originates from a strong anti-entropy unit, the recall domain should extend towards the high-sensitivity order boundary. and These represent the adjustment coefficients for the substitute traction enhancement term and the anti-entropy source enhancement term, respectively. The server completes the multidimensional boundary setting based on the strength of the circumstantial recall domain, thereby obtaining the circumstantial recall domain.

[0070] After forming the circumstantial evidence recall domain, the server performs peripheral fragment detection processing based on the circumstantial evidence recall domain to generate a set of candidate circumstantial evidence fragments. Peripheral fragment detection processing refers to the process of extracting peripheral behavioral fragments, neighborhood change fragments, access control change fragments, and order disturbance fragments that may have a supporting relationship with the intent cavity area within the circumstantial evidence recall domain. The set of candidate circumstantial evidence fragments refers to the set of candidate fragments that have been initially identified within the circumstantial evidence recall domain as potentially providing peripheral support to the intent cavity area. Peripheral fragments are not segments in the current student ID main chain, but rather peripheral traces located around the intent cavity area, consisting of other student ID-related record fragments, other student ID-related stream fragments, access control trigger record fragments, channel-to-departure record fragments, and peer-to-peer neighbor record fragments. During implementation, the server first extracts student ID card-related record fragments that are spatially or orally adjacent to the intent cavity area from the entire set of student ID card-related records within the time window corresponding to the circumstantial evidence recall domain. Then, it extracts edge-related and occlusion-related fragments from the set of student ID card-related fragments. Simultaneously, it extracts asymmetric access control fragments, abnormal lag fragments, short-term yielding fragments, group fragmentation fragments, and accompanying jump fragments from access control trigger records, channel arrival / departure records, and peer-neighbor records. Candidate screening results are then formed based on temporal overlap, spatial proximity, path coupling, and behavioral relevance. To quantify the candidate value of peripheral fragments, the server constructs a peripheral detection scoring function:

[0071] in, Indicates the first The peripheral segment is relative to the first The candidate scores for each intention cavity region are used to measure whether the peripheral fragment should be included in the set of circumstantial candidate fragments; Represents the weight coefficients of different candidate factors; Indicates the first The peripheral segment is relative to the first The intentional cavity region in the first The magnitude of the candidate factor, when The corresponding time overlap amount, when Time corresponds to spatial proximity quantity, when The corresponding path coupling quantity, when The corresponding neighborhood intervention amount, when Time-related behavior-related quantities; The nonlinear amplification index representing various candidate factors; This indicates a cross-domain coupling enhancement term, used to characterize whether the peripheral segment simultaneously spans multiple highly sensitive regions such as the access control front area, boundary zone, or channel corner area; This indicates a hysteresis residual enhancement term, used to characterize whether peripheral fragments have persistent residual phenomena before and after the formation of the intended cavity region; , and These represent the adjustment coefficients of the cross-domain coupling enhancement term, the hysteresis residual enhancement term, and the background difference suppression term, respectively; This indicates the magnitude of the background constraint difference between the peripheral fragment and the intended cavity region, used to suppress spurious candidates caused by excessively large background scene differences. The server includes peripheral fragments with peripheral detection scores higher than the detection threshold into the circumstantial candidate fragment set.

[0072] After generating the set of candidate supporting evidence fragments, the server performs a support relationship discrimination process on the candidate fragments to generate a valid set of supporting evidence fragments. The support relationship discrimination process refers to determining whether a genuine supporting relationship exists between each candidate supporting evidence fragment and the intended cavity area, explaining its formation, maintenance, or expansion. The valid set of supporting evidence fragments refers to the set of peripheral fragments selected from the candidate supporting evidence fragment set that have a clear supporting logic with the intended cavity area. Supporting relationships include occlusion support relationships, yielding support relationships, relay support relationships, residual window support relationships, crack support relationships, and avoidance support relationships. The support relationships are categorized into three types: **Coverage Support Relationship:** **Coverage Support Relationship:** Peripheral segments reduce the salience of the ontological chain corresponding to the intended cavity area by forming a collective coverage or localized high-density flow. **Yield Support Relationship:** Peripheral segments make way for the ontological chain by leaving early, creating localized empty spaces, or shifting their paths. **Relay Support Relationship:** Multiple peripheral segments appear consecutively in a time-staggered manner, providing multi-stage cover for infiltration behavior. **Residual Window Support Relationship:** Peripheral segments provide opportunities for the ontological chain to enter or attach through legitimate access control triggers, residual access control openings, or short-term authorization extensions. **Crack Support Relationship:** Peripheral segments create localized order gaps, allowing the ontological chain to traverse areas covered by normal order. **Avoidance Support Relationship:** Peripheral segments create avoidance support for infiltration behavior through abnormal avoidance, lateral movement, or dispersion in the neighborhood. The server calculates the support strength of each circumstantial candidate segment for the intended cavity area and distinguishes between primary and secondary support relationships. The support relationship strength function can be expressed as:

[0073] in, Indicates the first The candidate circumstantial evidence fragment is relative to the first... The strength of the support relationship in each cavity area is used to measure whether it constitutes effective support. Represents the weight coefficients for different support relationship categories; Indicates the first The candidate circumstantial evidence fragment is relative to the first... The intentional cavity region in the first The magnitude of the class support relationship, when When the corresponding shielding support amount, When the corresponding yield support amount, when The corresponding relay support amount, when When the corresponding window support amount, The corresponding crack support amount, when The corresponding support level should be avoided. Nonlinear amplification index representing various support quantities; This indicates a spatiotemporal synchronization enhancement term, used to characterize the degree of synchronization between the circumstantial candidate fragment and the intended cavity region in terms of occurrence time and spatial location; This indicates a continuation enhancement term, used to characterize whether candidate supporting evidence segments form continuous support within the preceding and following time windows; The logical deviation is used to characterize whether the supporting candidate fragments deviate from the formation logic of the intended cavity area. The greater the logical deviation, the more the support strength should be weakened. , and These represent the adjustment coefficients for the spatiotemporal synchronization enhancement term, the continuity enhancement term, and the logical deviation attenuation term, respectively. The server includes candidate supporting evidence fragments with a supporting relationship strength higher than the discrimination threshold into the set of valid supporting evidence fragments.

[0074] After forming a set of valid supporting evidence fragments, the server performs supporting evidence clustering processing based on this set to generate supporting evidence reinjection clusters as the recall result. Supporting evidence clustering processing refers to the process of organizing multiple valid supporting evidence fragments surrounding the same intention cavity area, having similar supporting relationship types, adjacent temporal distributions, and adjacent spatial distributions into structured clusters. Supporting evidence reinjection clusters refer to multi-source peripheral supporting evidence clusters formed for the same intention cavity area, used to reinject peripheral support back into the intention cavity area during subsequent latent event incubation processing. Reinjection is not a simple addition, but rather refers to re-injecting peripheral supporting traces into the interpretation structure of the intention cavity area, elevating the intention cavity area from a simple area lacking interpretation into a latent event precursor with internal and external closed-loop support. In implementation, the server first merges valid supporting evidence fragments according to the intention cavity area identifier, and then clusters them based on supporting relationship type, temporal continuity, spatial adjacency, target tendency consistency, and the direction of action on the ontology chain. If multiple valid supporting evidence fragments simultaneously form both residual window support and shading support relationships around the same access control front area, they aggregate into an access control shading-type supporting evidence recharge cluster; if multiple valid supporting evidence fragments form both yield support and crack support relationships around the same boundary zone, they aggregate into a boundary relay-type supporting evidence recharge cluster. To quantify the clustering effect, the server constructs a supporting evidence clustering strength function:

[0075] in, Indicates the first The strength of the sympathetic cluster of a valid sympathetic fragment cluster is used to measure whether the cluster should be identified as a sympathetic recharge cluster. The weighting coefficients representing different clustering factors; Indicates the first The cluster of valid supporting evidence fragments in the first... The magnitude of the clustering factor, when The corresponding support relationship is consistent, when Time corresponds to a continuous quantity in time, when The corresponding spatial adjacency quantity, when When the corresponding target tends to be consistent, when The amount corresponding to the direction of the main body support is consistent; The nonlinear amplification index representing various clustering factors; This represents the density enhancement term, used to characterize whether multiple valid supporting evidence fragments form a high density in a local region; This indicates a coverage enhancement term, used to characterize whether the cluster simultaneously covers multiple interpretation gaps in the intended cavity region; and These represent the adjustment coefficients for the density enhancement term and the coverage enhancement term, respectively. The server identifies clusters with sympathetic cluster strength exceeding the confirmation threshold as sympathetic re-feedback clusters and outputs them as recall results.

[0076] After generating the supporting evidence recharge cluster, the server performs cavity support fusion processing based on the intent cavity region and the supporting evidence recharge cluster to generate a set of latent components. Cavity support fusion processing refers to the process of matching and merging the explanation gap structures within the intent cavity region with the external support structures provided by the supporting evidence recharge cluster in a one-to-one manner. The set of latent components refers to a collection of multiple minimal latent behavioral components that have formed internal and external support loops. A latent component is the minimal latent event particle composed of explanation gaps, corresponding ontology segments, and at least one type of external support relationship, such as boundary yielding latent components, residual window retention latent components, occlusion and stagnation latent components, and crack traversal latent components. During implementation, the server analyzes each explanatory gap in the intended cavity area and each valid supporting fragment in the supporting evidence cluster to determine whether there is a correspondence in type, time, space, and support direction. If the explanatory gap manifests as a delayed stay in the access control pre-area, and there is a significant residual window support relationship in the supporting evidence cluster, then it is merged into a residual window retention latent component. If the explanatory gap manifests as repeated attachment of the boundary zone, and there is short-term yielding and local crack support in the supporting evidence cluster, then it is merged into a boundary yielding latent component. To quantify this fusion effect, the server constructs a cavity support fusion degree function:

[0077] in, Indicates the first The degree of fusion between the intention cavity substructure and the supporting evidence recharge cluster substructure to form a latent component is used to measure whether the two are sufficient to constitute an effective latent component. Represents the weighting coefficients of different fusion factors; Indicates the first The candidate fusion unit in the first The magnitude of the class fusion factor, when The corresponding type matching quantity, when Time corresponds to time quantity, when Time corresponds to space and quantity, when The corresponding amount corresponds to the support direction at that time; The nonlinear amplification index representing various fusion factors; This represents a closed-loop enhancement term, used to characterize whether a complete causal loop has been formed between the internal explanation of the cavity region's lack of explanation and the external support. This represents the amount of missing bias, used to characterize how many unsupported and unexplained residual gaps remain after fusion. The larger the residual gaps, the lower the degree of fusion should be. and These represent the adjustment coefficients for the closed-loop enhancement term and the missing bias attenuation term, respectively. The server includes candidate units with a fusion degree higher than the component formation threshold into the latent component set.

[0078] After forming a set of latent components, the server performs component succession and arrangement processing on the set to generate a latent evolution chain. Component succession and arrangement processing refers to the process of continuously organizing multiple latent components according to temporal succession relationships, spatial transfer relationships, target proximity relationships, and support continuity relationships. A latent evolution chain refers to a continuous chain structure formed by multiple latent components, representing the evolution of latent behavior from local manifestation to continuous advancement. Temporal succession relationship means that after the previous latent component ends, the next latent component continues to appear within an acceptable interval; spatial transfer relationship means that there is a real reachable path and latent migration direction between the spatial semantic locations of the previous and subsequent latent components; target proximity relationship means that the previous and subsequent latent components advance towards the same access control node, the outer edge of the same sensitive area, or the same floor transition point; support continuity relationship means that the supporting conditions in the previous latent component are continued, replaced, or strengthened in the subsequent latent component. During implementation, the server groups latent components according to the student ID master index or temporary collaborative objects, then calculates the bearing strength between any two latent components, and completes the chained arrangement accordingly. The latent bearing strength function can be expressed as:

[0079] in, Indicates the first The first hidden component and the first The bonding strength between latent components is used to measure whether they should be connected to the same latent evolutionary chain; Indicates the weighting coefficients of different supporting factors; Indicates the first The first hidden component and the first The lurking component is in the first The value of the class-inherited factor, when The corresponding time capacity, when The corresponding spatial transfer amount when When the target approximation quantity is reached, The corresponding support quantity; The nonlinear amplification index representing various receiving factors; This indicates a trend enhancement term, used to characterize whether the latent strength between two latent components is increasing. This indicates a support enhancement term, used to characterize whether there is a superposition or continuation effect in the external support between two latent components; and These represent the adjustment coefficients for the trend enhancement term and the support enhancement term, respectively. The server connects latent components with a strength exceeding the orchestration threshold to generate a latent evolution chain.

[0080] After generating the latent evolution chain, the server performs cooperative coupling identification processing based on the latent evolution chain to generate cooperative latent groups. Cooperative coupling identification processing refers to the process of identifying whether multiple student IDs form cooperative relationships such as substitute complementarity, circumstantial evidence sharing, cover relay, and probing division of labor around the same latent target. Cooperative latent groups refer to a group structure composed of multiple latent evolution chains that have cooperative coupling relationships with each other, used to characterize the cooperative form of multiple student IDs jointly participating in the gestation process of the same latent event. Substitute complementarity means that different student IDs undertake different surface normal order maintenance roles; circumstantial evidence sharing means that multiple latent evolution chains jointly rely on the same circumstantial evidence backflow cluster or share the same peripheral support conditions; cover relay means that different student IDs alternately form group cover, yield support, or crack support at different time windows; probing division of labor means that different student IDs undertake different latent tasks such as boundary probing, remaining window attachment, path return, or regional stagnation. In implementation, the server calculates the temporal overlap, spatial coverage, support sharing, and role complementarity between different latent evolutionary chains, using these quantities in combination for cooperative latent group identification. The cooperative coupling strength function can be expressed as:

[0081] in, Indicates the first The latent evolutionary chain and the first The strength of the cooperative coupling between latent evolutionary chains is used to measure whether they should be grouped into the same cooperative latent group; Represents the weighting coefficients of different synergistic factors; Indicates the first The latent evolutionary chain and the first The latent evolutionary chain in the first The magnitude of the synergistic factor, when The corresponding time interleaving amount, when The corresponding spatial coverage, when The corresponding shared quantity is supported when The corresponding complementary amount of roles; The nonlinear amplification index representing various synergistic factors; This indicates a relay enhancement term, used to characterize whether two latent evolutionary chains alternately advance at different times; This indicates a shared enhancement term, used to characterize whether two latent evolutionary chains jointly depend on the same access window, the same group shading, or the same spatial crack. The amount of role conflict is used to characterize whether there is significant role overlap or logical conflict between two latent evolutionary chains. The greater the role conflict, the lower the cooperative coupling strength should be. , and These represent the adjustment coefficients for the relay enhancement term, the sharing enhancement term, and the role conflict suppression term, respectively. The server groups latent evolution chains with cooperative coupling strength exceeding the grouping threshold into the same cooperative latent group.

[0082] After generating the collaborative latent groups, the server performs event nucleation determination processing based on the latent evolution chain and the collaborative latent groups to generate the latent event protocore. Event nucleation determination processing refers to the process of determining whether the current latent evolution structure has evolved from a locally suspicious state into a pre-event core with continuous growth and structural self-stabilization capabilities. The latent event protocore refers to a pre-event core composed of one or more latent evolution chains and optional collaborative latent groups, possessing a clear latent target, stable support relationships, a continuous evolutionary trend, and a collaborative structure. The protocore is not the event itself, but rather a sustainable incubation core formed before the event becomes explicit. During implementation, the server extracts evolution length, evolutionary increment intensity, target convergence degree, and support loop closure degree from the latent evolution chains, and extracts collaborative scale, collaborative stability, and role division completeness from the collaborative latent groups, then jointly calculates whether the current candidate structure meets the nucleation conditions. If the latent evolution chain only exhibits a single probing event or weak circumstantial recharge cluster, then a latent event pronucleus will not form. If the latent evolution chain progresses continuously across multiple time intervals, the intended cavity region appears repeatedly, the circumstantial recharge cluster continues to thicken, and the collaborative latent group has formed a shielding relay or probing division of labor, then a latent event pronucleus is determined to have formed. The event nucleation determination function can be expressed as:

[0083] in, This represents the event nucleation intensity of a candidate latent structure, used to measure whether the candidate latent structure meets the formation conditions of the latent event nucleus; The weighting coefficients representing different nucleation factors; Indicates the candidate latent structure in the first... The magnitude of the nucleation factor, when The corresponding latency evolution length is when The corresponding evolutionary enhancement measure, when The corresponding target convergence quantity, when The corresponding support closed-loop quantity, when The corresponding amount of repeated exposure of the cavity; The nonlinear amplification index representing various nucleation factors; This represents the co-enhancing term, used to characterize the amplification effect of the presence of co-latency groups on nucleation intensity; This represents the closed-loop enhancement term, used to characterize whether a stable closed loop has been formed between the intended cavity region, the circumstantial recharge cluster, and the set of latent components. It represents the amount of random disturbance deviation, used to characterize whether the current candidate latent structure may be caused only by occasional fluctuations in pedestrian flow, ordinary stops or temporary relocation. The greater the random disturbance, the more the nucleation intensity should be reduced. , and These represent the adjustment coefficients for the collaborative enhancement term, the closed-loop enhancement term, and the random perturbation attenuation term, respectively. The server identifies candidate latent structures with event nucleation strength exceeding the nucleation threshold as latent event protonuclei, thus completing the entire process from the intended cavity region to the recall result and then to the latent event protonuclei.

[0084] S160. Based on the latent event kernel, perform unpacking and wake-up processing and hierarchical response determination processing to output the student ID card hierarchical response result, and feed back the student ID card hierarchical response result to form a closed-loop evolution of the student ID card collaborative perception and hierarchical response process.

[0085] Specifically, in the implementation of the above technical solution, the server first performs kernel parsing processing on the latent event kernel to generate a kernel structure expression. The latent event kernel refers to the pre-event core, composed of a latent evolution chain and optional cooperative latent groups, possessing a clear latent target, stable supporting relationships, a continuous evolutionary trend, and a cooperative structure. Kernel parsing processing refers to the deconstructive analysis of the spatial aggregation mode, temporal progression mode, supporting closed-loop mode, cooperative division of labor mode, and order shell dependency mode within the latent event kernel. The kernel structure expression refers to the structured representation of the internal composition of the latent event kernel, used to clarify which order shells, spatial semantic locations, supporting paths, and cooperative roles should be targeted for perturbation in subsequent shell removal and wake-up processing. During implementation, the server first reads the intent cavity region, circumstantial backflow cluster, latent component set, latent evolution chain, and cooperative latent group corresponding to the latent event protoker. Then, it jointly models its evolutionary length in the time dimension, convergence center in the spatial dimension, dominant support mode in the support dimension, and role division in the cooperative dimension, and extracts dominant order shells such as the group occlusion shell, access control window shell, channel receiving shell, boundary fallback shell, and scattering residual flow shell. To ensure that the protoker structure representation can simultaneously reflect the maturity, support level, and cooperative level of the latent event protoker, the server constructs a protoker analytical strength function:

[0086] in, Indicates the first The prokaryotic resolution intensity of a latent event prokaryote is used to measure the structural clarity and intervention targeting of the latent event prokaryote. The weight coefficients represent the different analytical factors, which are used to adjust the influence of different internal structural dimensions on the prokaryotic analytical strength. Indicates the first The latent event nucleus is in the first The magnitude of the class analysis factor, when The corresponding time advance amount, when The corresponding spatial convergence when When the corresponding support closed-loop quantity is, The corresponding collaborative division of labor, when The corresponding target stable quantity; The nonlinear amplification index represents various analytical factors and is used to enhance the identification effect of pronuclei of highly mature latent events. This represents a multidimensional structural coupling enhancement term, used to reflect the overall amplification effect when multiple structural factors are simultaneously significant. This represents the co-enhancement term, used to characterize the contribution of the co-latency group to the prokaryotic stability of latent events; This represents the shell enhancement term, used to characterize the degree of dependence of the latent event pronucleus on the order shell; and These represent the adjustment coefficients for the synergistic enhancement term and the shell enhancement term, respectively; It represents the random disturbance deviation, which is used to characterize the random component introduced into the current latent event nucleus by occasional congestion, ordinary stoppage or temporary avoidance. The stronger the randomness, the lower the nucleus resolution should be. The suppression coefficient represents the amount of random perturbation deviation. Based on the prokaryotic resolution strength and its composition results, the server generates a prokaryotic structure representation that includes the dominant spatial semantic location, dominant support mode, dominant cooperative structure, dominant maturity stage, and dominant order shell type.

[0087] After generating the proto-kernel structure representation, the server performs unpacking and perturbation processing based on the proto-kernel structure representation to generate perturbation execution results. Unpacking and perturbation processing refers to the process of selectively perturbing the normal order shell to which the latent event proto-kernel depends, based on the order shell dependencies identified in the proto-kernel structure representation, thereby weakening its disguise support conditions and causing its true behavior to tend to be revealed. The perturbation execution result refers to the perturbation process results recorded by the server after implementing unpacking and perturbation processing, including the perturbation object, perturbation method, perturbation duration, perturbation intensity, and perturbation coverage. During implementation, the server does not perform uniform intervention on all regions simultaneously, but rather applies customized perturbations according to the dominant order shell type in the proto-kernel structure representation. For the group occlusion shell, the occlusion effect is weakened by increasing the perception resolution density of the target area, reducing the tolerance for group fusion, and increasing the weight of individual trajectory separation. For the access control window shell, the support of the window is weakened by shortening the continuous opening time after access control is triggered, increasing the frequency of access control re-identification, and compressing the dwell tolerance window of the access control front area. For the channel receiving shell, the rationality of borrowing is weakened by changing the channel guidance rhythm, increasing the granularity of channel node behavior recognition, and enhancing the dwell sensitivity of the channel edge. For the boundary fallback shell, the edge-hugging camouflage is weakened by increasing the salience of boundary zone patrol, enhancing boundary stopping prompts, and increasing the frequency of boundary area behavior marking. For the residual flow shell, the rationality of delayed de-flow is weakened by shortening the residual aggregation time after the event ends, increasing the event clearing speed, and compressing the residual flow fallback window. To quantify the effect of the unpacking and wake-up processing on the stripping of the order shell, the server constructs an unpacking and wake-up action function:

[0088] in, Indicates the first The strength of the unpacking and wake-up process corresponding to each latent event nucleus is used to measure the degree to which the current unpacking and wake-up process weakens the camouflage support structure of the latent event nucleus. Represents the weighting coefficients of different disturbance factors; Indicates the first The latent event nucleus is in the first The applied value on the perturbation factor, when When the corresponding group shading disturbance amount, when When the corresponding door access control window disturbance amount, The corresponding channel receives the disturbance amount when When the corresponding boundary fall-back disturbance amount, when The corresponding residual current disturbance quantity in the time field; The nonlinear amplification index represents various disturbance quantities and is used to enhance the identification effect of directional strong disturbances; This indicates a focused enhancement term, used to characterize whether the disturbance is concentrated on the dominant spatial semantic location and the dominant support path; This indicates a cascaded enhancement term, used to characterize the cascaded weakening effect on the camouflage capability of latent event progenitors when multiple types of order shells are simultaneously disturbed; and These represent the adjustment coefficients for the focusing enhancement term and the cascaded enhancement term, respectively; This represents the amount of safety disturbance deviation, used to characterize whether the current disturbance excessively affects the normal campus order. If the deviation is larger, the intensity of the unpacking and wake-up effect should be suppressed. This represents the suppression coefficient for the amount of security disturbance deviation. The server performs disturbance operations based on the strength of the unpacking and wake-up action, and records the wake-up execution results.

[0089] After generating the wake-up perturbation execution result, the server performs wake-up perturbation feedback parsing processing on the result to generate wake-up perturbation feedback results. Wake-up perturbation feedback parsing processing refers to the process of re-collecting the change states of the surrogate chain, ontology chain, intention cavity region, circumstantial recharge cluster, latent evolution chain, and co-emergent group related to the latent event progenitor after the de-shelling wake-up perturbation processing is implemented, and jointly analyzing these change states to determine the true response mode of the latent event progenitor after losing the support of the order shell. The wake-up perturbation feedback result refers to the feedback object after the structured expression of the changes after the de-shelling wake-up perturbation processing, used to describe whether the surrogate chain has collapsed, whether the ontology chain has converged, whether the intention cavity region has migrated or shrunk, whether the circumstantial recharge cluster has broken or thickened, whether the latent evolution chain has been interrupted or strengthened, and whether the co-emergent group has decoupled or recombined. During implementation, the server continuously monitors a feedback window after the disturbance, comparing the structural differences before and after the de-shelling and wake-up process. If the substitute chain disappears rapidly while the original chain significantly strengthens, it indicates that the original latent event's core is highly dependent on the order shell; if the substitute chain remains stable while the original chain does not expand significantly, it indicates that the original latent event's core may have strong self-consistency; if multiple student IDs exhibit synchronous transfer, synchronous exposure, or alternating occlusion, it indicates that the collaborative latent group has reconstructed after being disturbed. To quantify the degree of structural change before and after the wake-up, the server constructs a wake-up feedback increment function:

[0090] in, Indicates the first The perturbation feedback intensity corresponding to each latent event pronucleus is used to measure the structural response amplitude of the latent event pronucleus after unpacking and perturbation processing; The weighting coefficients representing different feedback change factors; Indicates the first The latent event nucleus is in the first The amount of change in the feedback-like change factors, when The change in the Substitute Chain at that time When the corresponding change in the ontology chain, The change in the cavity area corresponding to the time, when The corresponding circumstantial evidence is the change in the recharge cluster, when The corresponding change in the latent evolutionary chain, when The corresponding change in the co-occurring latent group; The nonlinear amplification index representing various feedback changes; This indicates the core structure linkage enhancement term, used to reflect the overall amplification effect when the substitute chain, the body chain, and the intention cavity region undergo significant changes simultaneously; This indicates a recombination enhancement term, used to characterize whether new co-occurring latent group recombination or latent evolutionary chain reconnection occurs after arousal. The adjustment coefficient representing the recombinant enhancement term; This represents the normal recovery inhibition term, used to characterize whether the structural changes after the disturbance quickly return to normal order. The faster the return, the more unstable the latent event nucleus is, and the feedback intensity should be appropriately reduced. This represents the adjustment coefficient for the normal recovery inhibition term. The server generates the wake-up feedback results based on the wake-up feedback strength and the changes in various parameters.

[0091] After generating the wake-up feedback result, the server performs a tiered response judgment process based on the wake-up feedback result and the latent event kernel to generate a tiered response result for the student ID card. The tiered response judgment process refers to the process of classifying the risk level of the student ID card by comprehensively considering the maturity level of the latent event kernel before the disturbance, its manifestation level after the disturbance, its dependence on the order shell, and its degree of collaborative latentness. The tiered response result for the student ID card refers to the structured expression of the latent risk level, response level, and subsequent handling priority corresponding to the student ID card or student ID card group. The risk level can be represented as a low-level attention state, a medium-level warning state, and a high-level handling state; the response level refers to the strength level of the prompts, inspections, reviews, interceptions, or coordinated handling that the system should trigger subsequently; the handling priority refers to the resource allocation order of the corresponding objects when multiple latent event kernels exist simultaneously. During implementation, the server jointly inputs the kernel resolution strength, unpacking wake-up effect strength, wake-up feedback strength, and collaborative coupling characteristics of the latent event kernel into the tiered judgment model. If the disturbance feedback results show that the substitute chain collapses, the body chain converges at the outer edge of the sensitive region, the intention cavity region persists, and the latent evolution chain remains unbroken, it is judged as a high-level risk; if the disturbance feedback results show that the substitute chain fluctuates partially, the body chain repeatedly probes, the intention cavity region reappears multiple times, and the circumstantial backfill cluster continues to thicken, it is judged as a medium-level risk; if the disturbance feedback results show that the substitute chain remains stable, the body chain does not significantly expand, and the intention cavity region gradually disappears, it is judged as a low-level risk. To ensure that the graded response judgment considers both static structure and dynamic feedback, the server constructs a graded response judgment function:

[0092] in, Indicates the first The graded response judgment value corresponding to the original kernel of a latent event is used to measure what response level should be assigned to the student ID card corresponding to the original kernel of the latent event; Represents the weighting coefficients of different risk factors; Indicates the first The latent event nucleus is in the first The magnitude of risk factors, when The corresponding amount of prokaryotic maturation, when When the corresponding shell dependency, when The corresponding feedback exposure amount, when When the target convergence is corresponding to the time, The corresponding support closed-loop quantity; A non-linear amplification index representing various risk factors; This indicates a synergistic enhancement term, used to characterize the effect of synergistic latent groups on improving the risk level; This indicates the exposure enhancement term, used to characterize the effect of the degree of exposure of the real ontology behavior after unpacking and wake-up processing on the risk level. and These represent the adjustment coefficients for the synergistic enhancement term and the explicit enhancement term, respectively; This represents the reference threshold for risk demarcation, used to distinguish between low-level, medium-level, and high-level risks. This represents the compression mapping coefficient, used to control the sensitivity of grade transitions when the prokaryotic resolution strength and the wake-up feedback strength work together. The server outputs the student ID grade response result based on the range of the grade response judgment value.

[0093] After generating the student ID card tiered response results, the server performs a result backfeeding process to feed the results back into the order flow graph, shell extraction process, and intent cavity mining process. The server then performs a backfeeding consistency check on the fed-back order flow graph, shell extraction process, and intent cavity mining process to form a closed-loop evolutionary student ID card collaborative perception and tiered response process. Result backfeeding refers to the process of reversing the actual risk state and structural evolution results obtained during the tiered response judgment process and injecting them into the preceding recognition structure to update the order boundary, shell discrimination boundary, and cavity recognition boundary. Backfeeding consistency check refers to the process of re-verifying the stability and consistency of the fed-back order flow graph, shell extraction process, and intent cavity mining process after the result backfeeding is completed. During implementation, the server re-injects the set of naked intention fragments, reverse-order threads, intention cavity regions, circumstantial evidence re-injection clusters, and latent event proto-structures corresponding to high-risk levels into the order mother stream map. This is used to correct the boundary conditions of the order mother stream primitives and the constraints of order relationships. The server re-injects the set of shell-borrowing candidate fragments and the set of shell-dependent identifiers corresponding to medium-risk levels into the shell-borrowing extraction process to enhance the sensitivity of identifying concealed shell-borrowing behavior. The server re-injects the set of interpretation-consistent segments and substitute chains corresponding to low-risk levels into the intention cavity mining process to correct the interpretation boundaries of the normal order appearance and reduce the probability of misidentification. After the results are re-injected, the server reuses the re-injected order mother stream map to interpret the set of historical student ID card appendage fragments, re-applies the re-injected shell-borrowing extraction process to the set of historical student ID card appendage fragments, and re-applies the re-injected intention cavity mining process to the historical reverse-order threads. Changes in the normal structure interpretation rate, shell-borrowing candidate expansion rate, and intention cavity region drift rate before and after re-injection are compared to determine whether the re-injection is stable. To quantify the overall consistency after the power-back, the server constructs a power-back consistency function:

[0094] in, It indicates the overall consistency of the results after backfeeding, and is used to measure whether the order mother flow map, shell extraction processing and intention cavity mining processing remain stable and coordinated after closed-loop evolution; The weighting coefficients representing different consistency factors; Indicates the first The magnitude of the class consistency factor, when The time corresponds to the order interpretation of the quantity, when The corresponding stable quantity for identifying backdoor listings, when The corresponding cavity identification stability quantity, when The corresponding response backtracking consistency quantity; The nonlinear amplification index representing various consistency factors; This indicates the pre-structure synergistic enhancement term, used to reflect the overall amplification effect when the order mother flow map, shell extraction processing, and intention cavity mining processing are all kept stable. This represents the closed-loop stability enhancement term, used to characterize the consistency of the system's performance across multiple historical windows after refeedback. This represents the adjustment coefficient of the closed-loop stability enhancement term; This represents the drift penalty term, used to characterize whether abnormal drift occurs in the intended cavity region, shell candidate fragment, or order relationship after reflow; The expansion penalty term is used to characterize whether the set of candidate segments for shell borrowing or the number of intended cavity regions expands abnormally after refilling. and These represent the adjustment coefficients for the drift penalty term and the expansion penalty term, respectively. When the server meets the consistency requirements for backfeeding, it considers the current backfeeding result as valid, thus forming a closed-loop evolutionary student ID card collaborative perception and hierarchical response process.

[0095] This application also provides a student ID collaborative sensing and hierarchical response device for campus security incidents, referring to... Figure 2 , Figure 2 This application provides a schematic diagram of a student ID card collaborative sensing and hierarchical response device for campus security incidents. The device is a server, comprising an acquisition module 21 and a processing module 22. The acquisition module 21 acquires location observation records, access control trigger records, passage arrival / departure records, neighboring records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to the student ID card, forming a set of student ID card-related records for campus security incidents. The processing module 22 constructs an order flow graph based on the student ID card-related record set and maps the student ID card-related record set to the order flow graph to generate a set of student ID card-related flow segments. The processing module 22 is also used to process the student ID card-related flow segments. The segment set is subjected to shell extraction processing to obtain the extraction result, and the extraction result is subjected to order anti-entropy weaving processing to generate reverse-order silk chains; the processing module 22 is also used to perform substitute projection processing on the reverse-order silk chains to obtain the projection result, and perform intention cavity mining processing on the projection result to generate intention cavity regions; the processing module 22 is also used to perform circumstantial recall processing around the intention cavity regions to obtain the recall result, and perform latent event incubation processing based on the recall result to generate latent event kernels; the processing module 22 is also used to perform shell removal and wake-up processing and hierarchical response judgment processing based on the latent event kernels to output student ID card hierarchical response results, and perform backfeeding on the student ID card hierarchical response results to form a closed-loop evolution of student ID card collaborative perception and hierarchical response process.

[0096] This application also provides an electronic device, with reference to... Figure 3 , Figure 3This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: at least one processor 31, at least one network interface 34, a user interface 33, a memory 35, and at least one communication bus 32.

[0097] The communication bus 32 is used to enable communication between these components.

[0098] The user interface 33 may include a display screen and a camera. Optionally, the user interface 33 may also include a standard wired interface and a wireless interface.

[0099] The network interface 34 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0100] The processor 31 may include one or more processing cores. The processor 31 connects to various parts of the server via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in the memory 35, and calling data stored in the memory 35 to perform various server functions and process data. Optionally, the processor 31 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 31 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content to be displayed on the screen; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 31 and may be implemented as a separate chip.

[0101] The memory 35 may include random access memory (RAM) or read-only memory. Optionally, the memory 35 may include a non-transitory computer-readable storage medium. The memory 35 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 35 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 35 may also be at least one storage device located remotely from the aforementioned processor 31. Figure 3 As shown, the memory 35, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for a student ID collaborative perception and hierarchical response method for campus security incidents.

[0102] exist Figure 3 In the electronic device shown, the user interface 33 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 31 can be used to call the application stored in the memory 35, which is a student ID card collaborative perception and hierarchical response method for campus security events. When executed by one or more processors, the electronic device executes one or more methods as described in the above embodiments.

[0103] This application also provides a non-transitory computer-readable storage medium storing instructions. When executed by one or more processors, these instructions cause an electronic device to perform one or more of the methods described in the above embodiments.

[0104] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A student ID collaborative perception and hierarchical response method for campus security incidents, characterized in that, The method includes: Acquire location observation records, access control trigger records, passage arrival and departure records, companion neighborhood records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to student ID cards to form a set of student ID card-related records for campus security incidents; Based on the student ID card associated record set, an order mother stream graph is constructed, and the student ID card associated record set is mapped to the order mother stream graph to generate a student ID card attached stream fragment set; The student ID card attached fragment set is subjected to shell extraction processing to obtain the extraction result, and the extraction result is subjected to order anti-entropy weaving processing to generate reverse order silk chain; The reversed silk chain is subjected to a substitute projection process to obtain a projection result, and the projection result is subjected to an intention cavity mining process to generate an intention cavity region. A corroborating evidence recall process is performed around the intent cavity area to obtain a recall result, and a latent event incubation process is performed based on the recall result to generate a latent event protoker. Based on the latent event, the kernel performs unpacking and wake-up processing and hierarchical response determination processing to output the student ID card hierarchical response result, and the student ID card hierarchical response result is fed back to form a closed-loop evolution of the student ID card collaborative perception and hierarchical response process.

2. The student ID collaborative perception and hierarchical response method for campus security incidents according to claim 1, characterized in that, The process involves acquiring location observation records, access control trigger records, passage entry and exit records, peer-to-peer records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to student ID cards, to form a set of student ID card-related records for campus security incidents. Specifically, this includes: The location observation record, access control trigger record, channel arrival and departure record, peer neighborhood record, class time record, dormitory schedule record, campus activity record, and space control record are all bound together under the student ID main index to form a unified collection of multi-source records belonging to the same student ID. Time alignment processing is performed on all records merged into the student ID master index to form a student ID time-series record chain under a unified time base; Spatial placement processing is performed on the student ID time sequence record chain to map various records to preset spatial semantic positions and form a student ID spatial record chain; Perform event semantic annotation processing on the student ID spatial record chain to generate a student ID behavior record chain with behavioral semantic state; Perform fragment aggregation processing on the student ID behavior record chain to generate student ID associated record fragments; The student ID card associated record fragments are subjected to continuous succession processing to form a set of student ID card associated records with temporal succession, spatial transfer, behavioral continuity, neighborhood evolution, and background switching relationships.

3. The student ID collaborative perception and hierarchical response method for campus security incidents according to claim 1, characterized in that, The step of constructing an ordered main stream graph based on the student ID card associated record set, and mapping the student ID card associated record set to the ordered main stream graph to generate a student ID card supplementary stream fragment set, specifically includes: The student ID card associated record set is subjected to order primitive extraction processing to generate an order mother stream primitive set; The order mother stream primitive set is subjected to order relation weaving processing to generate an order mother stream relation network; Graph shaping processing is performed based on the ordered mother stream relationship network to generate the ordered mother stream graph; The student ID card associated record fragments in the student ID card associated record set are mapped to the order mother flow graph to generate the corresponding mapping results; Based on the mapping result, the student ID card associated record fragment is subjected to additional stream segmentation processing to generate the student ID card additional stream fragment set.

4. The student ID collaborative perception and hierarchical response method for campus security incidents according to claim 1, characterized in that, The process of performing shell extraction on the student ID card-related fragment set to obtain the extraction result, and then performing order anti-entropy weaving on the extraction result to generate a reverse-order silk chain, specifically includes: Perform a shell-borrowing candidate screening process on the student ID card attached fragment set to generate a shell-borrowing candidate fragment set; Perform shell dependency decomposition processing on the set of candidate shell-borrowing segments to generate a set of shell dependency identifiers; Based on the shell dependency identifier set, the shell-borrowing candidate fragment set is subjected to ordered shell stripping processing to generate a naked fragment set as the extraction result; Perform naked-meaning association and recombination processing on the naked-meaning fragment set to generate a naked-meaning fragment chain; The naked fragment chain is subjected to order consumption identification processing to generate an anti-entropy unit set; The directional weaving process is performed based on the anti-entropy unit set to generate the reverse-order yarn chain.

5. The student ID collaborative perception and hierarchical response method for campus security incidents according to claim 1, characterized in that, The process of performing a substitute projection on the reversed silk chain to obtain a projection result, and then performing intention cavity mining on the projection result to generate an intention cavity region, specifically includes: The reversed silk chain is subjected to an interpretation consistency decomposition process to generate a set of interpretation-consistent segments and a set of interpretation-missing segments; Based on the aforementioned set of consistent segments, a substitute chain construction process is performed to generate a substitute chain; Based on the set of missing segments explained, an ontology chain construction process is performed to generate an ontology chain; Perform substitute traction identification processing on the substitute chain and the original chain to generate a set of substitute traction relationships; Based on the substitute chain and the ontology chain, perform interpretation gap identification processing to generate an interpretation gap set; Cavity aggregation processing is performed on the set of interpretation gaps to generate the intended cavity region.

6. The student ID collaborative perception and hierarchical response method for campus security incidents according to claim 1, characterized in that, The step of performing circumstantial recall processing around the intended cavity region to obtain recall results, and performing latent event incubation processing based on the recall results to generate latent event proto-nuclei, specifically includes: A recall boundary generation process is performed around the intent cavity region to form a supporting evidence recall domain; Based on the aforementioned circumstantial evidence recall domain, peripheral fragment detection processing is performed to generate a set of circumstantial evidence candidate fragments; The set of candidate supporting evidence fragments is subjected to support relationship discrimination processing to generate a set of valid supporting evidence fragments; Based on the set of valid supporting evidence fragments, support evidence clustering is performed to generate a support evidence re-feedback cluster as the recall result; Based on the intended cavity region and the supporting evidence recharge cluster, cavity support fusion processing is performed to generate a set of latent components; Perform component acceptance and arrangement processing on the set of latent components to generate a latent evolution chain; Based on the latent evolution chain, perform cooperative coupling identification processing to generate cooperative latent groups; Based on the latent evolution chain and the cooperative latent group, the event nucleation determination process is performed to generate the latent event pronucleus.

7. The student ID collaborative perception and hierarchical response method for campus security incidents according to claim 1, characterized in that, The process of performing de-packing and wake-up processing and hierarchical response determination based on the latent event kernel to output the student ID card hierarchical response result, and then feeding back the student ID card hierarchical response result to form a closed-loop evolutionary student ID card collaborative perception and hierarchical response process, specifically includes: The latent event prokaryote is subjected to prokaryote parsing to generate a prokaryote structure representation; Based on the prokaryotic structure expression, perform unpacking and wake-up perturbation processing to generate wake-up perturbation execution results; The wake-up execution result is subjected to wake-up feedback parsing processing to generate wake-up feedback result; Based on the disturbance feedback result and the latent event proto-nucleus, a graded response determination process is performed to generate the student ID graded response result; The student ID card graded response results are fed back into the order flow map, shell extraction processing, and intent cavity mining processing. The fed-back order flow map, shell extraction processing, and intent cavity mining processing are then subjected to a consistency verification process to form a closed-loop evolution of the student ID card collaborative perception and graded response process.

8. A student ID card collaborative sensing and hierarchical response device for campus security incidents, characterized in that, The device is used to execute the student ID collaborative perception and hierarchical response method for campus security incidents as described in any one of claims 1 to 7, wherein the device includes an acquisition module and a processing module, wherein... The acquisition module is used to acquire location observation records, access control trigger records, passage arrival and departure records, peer neighborhood records, class time records, dormitory schedule records, campus activity records, and space control records corresponding to the student ID card, so as to form a set of student ID card associated records for campus security incidents; The processing module is used to construct an order mother stream graph based on the student ID card associated record set, and map the student ID card associated record set to the order mother stream graph to generate a student ID card attached stream fragment set; The processing module is also used to perform shell extraction processing on the student ID card attached fragment set to obtain the extraction result, and to perform order anti-entropy weaving processing on the extraction result to generate reverse order silk chain; The processing module is further configured to perform substitute projection processing on the reversed silk chain to obtain projection results, and perform intention cavity mining processing on the projection results to generate intention cavity regions. The processing module is also used to perform circumstantial recall processing around the intent cavity area to obtain recall results, and perform latent event incubation processing based on the recall results to generate latent event proto-nuclei; The processing module is also used to perform unpacking and wake-up processing and hierarchical response determination processing based on the latent event kernel, so as to output the student ID card hierarchical response result and to feed back the student ID card hierarchical response result to form a closed-loop evolution of the student ID card collaborative perception and hierarchical response process.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores instructions that, when executed, perform the method as described in any one of claims 1 to 7.

Images