An intelligent park monitoring method and system based on end-to-end cooperation

Abstract

The application discloses an intelligent park monitoring method and system based on end-to-end cooperation, and particularly relates to the field of intelligent park monitoring management based on end-to-end cooperation, and comprises the following steps: acquiring a cooperative behavior structure of multi-source synchronous sensing data in an intelligent park, constructing an initial cooperative path set, establishing an end-to-end cooperative behavior chain, extracting a candidate trusted path in the park which meets terminal response synchronization and type cross characteristics; performing disturbance injection on the terminal in the preliminary trusted cooperative chain and extracting disturbance feedback results to generate an elastic response trajectory with quantifiable characteristics. By injecting multiple types of disturbances into the cooperative path and dynamically extracting the response elastic characteristics thereof, combined with behavior generation path inversion and trust recoverability determination, identification of pseudo-consistent cooperative behavior and structural reconstruction of a trusted path are realized, thereby solving the core problem of false cooperative paths being mistakenly identified as real behavior by existing systems, leading to monitoring misjudgment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of end-to-end collaborative smart park monitoring management, and more specifically, the present application relates to a smart park monitoring method and system based on end-to-end collaboration. BACKGROUND

[0002] In the existing management system of the smart park, the identity confirmation and event linkage of personnel behavior usually rely on the collaborative verification mechanism of multi-terminal information. Generally, the mechanism provides fusion judgment based on the data provided by multiple heterogeneous sensing sources such as access card swiping, face recognition, WiFi trajectory, and video image;

[0003] To improve the automation level and response efficiency, the system design usually tends to take the consistency of multi-terminal data as the precondition for triggering trust judgment and action execution, that is, when multiple data sources appear collaborative matching in time and semantics, the system is considered as legal behavior, and automatically passes the verification or completes the instruction operation. The collaborative verification mechanism in the prior art has been widely deployed in many smart park platforms and is used as the core criterion to replace manual intervention.

[0004] However, the above technical route is generally based on the default that each terminal data source is independent and reliable, and lacks verification of the authenticity of the generation path of the collaborative relationship. With the development of sensor simulation, image replacement, and communication timing manipulation technologies, attackers can construct a set of pseudo data that is synchronized in time and matched in content, making the system misjudge it as a real behavior and bypass the normal abnormal identification mechanism.

[0005] Since the system trust logic is essentially based on the assumption that "collaborative consistency equals authenticity", the higher the degree of collaboration, the greater the probability of system misjudgment, thus forming a structural paradox: the system relies on multi-terminal collaboration to improve security, while high collaboration is used by attackers as a means to avoid security judgment. SUMMARY

[0006] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide a smart park monitoring method and system based on end-to-end collaboration, which injects multiple types of disturbances into the collaborative path and dynamically extracts its response flexibility features, combines behavior generation path inversion and trust recoverability judgment, realizes the identification of pseudo-consistent collaborative behavior and the structural reconstruction of credible path, and solves the core problem of the existing system that misidentifies the pseudo-collaborative path as a real behavior, leading to monitoring misjudgment.

[0007] To achieve the above purpose, the present application provides the following technical scheme: a smart park monitoring method based on end-to-end collaboration, comprising:

[0008] Acquire the collaborative behavior structure of multi-source synchronous perception data in the smart park, construct an initial collaborative path set and establish an end-to-end collaborative behavior chain, and extract candidate trusted paths within the park that meet the terminal response synchronization and type intersection characteristics;

[0009] Perform disturbance injection on the terminals in the preliminary trusted collaborative chain and extract the disturbance feedback results to generate elastic response trajectories with quantifiable characteristics;

[0010] By comparing resilience indicators with abnormal conditions, we can identify abnormal resilience response chains that can be used to reversely derive generation paths.

[0011] By constructing a behavior generation trajectory graph and calculating the transfer logic factor in the path, it identifies unnatural behavior paths that do not meet the device response conditions and outputs a pseudo-consistent collaborative behavior chain;

[0012] Perform light perturbation verification and recovery feature reconstruction on the paths weakened by behavioral trust, identify trust-recoverable paths that meet multiple perturbation response constraints, and output the final trusted collaborative path diagram and pseudo-consistency chain risk label for subsequent system intervention decisions.

[0013] In a preferred embodiment, a perception behavior dataset is obtained during smart park monitoring. The perception behavior dataset includes a card swiping event sequence, a face recognition record sequence, and a movement trajectory sequence of the smart park. A timestamp alignment operation is performed on the perception behavior dataset to output a collaborative candidate data window. A cross-modal fusion operation is performed on each set of synchronized perception data in the collaborative candidate data window, and an initial collaborative path set is output through behavior matching calculation.

[0014] For each path in the initial collaborative path set, its collaborative terminal combination group is extracted; the collaborative terminal combination group includes a card swiping device, a face recognition terminal, and a trajectory sensing node.

[0015] In a preferred embodiment, a structural connectivity analysis is performed on the path structure of the collaborative terminal combination group to obtain the end-to-end collaborative behavior chain, and a consistent collaborative path graph is constructed. For each path chain in the consistent collaborative path graph, the corresponding terminal response time series is extracted, and response timing continuity detection is performed to output the behavior synchronization chain group;

[0016] If the terminal response time interval fluctuation in the behavior synchronization chain group is less than the preset interval fluctuation threshold, and the terminal type cross ratio is not lower than the preset cross ratio threshold lower limit, the corresponding chain in the behavior synchronization chain group is marked as a preliminary trusted collaborative chain.

[0017] In a preferred embodiment, a disturbance request is injected into each terminal in the preliminary trusted collaborative chain; the disturbance request includes path deviation disturbance, identity confidence disturbance and trigger timing disturbance;

[0018] By injecting the disturbed terminal behavior sequence, a disturbed response behavior trajectory set is extracted, and an end-end disturbed feedback flow is constructed; each terminal trajectory in the disturbed feedback flow is subjected to dynamic recovery period analysis, and a disturbed recovery path set is output;

[0019] For each path in the disturbed recovery path set, a recovery response time, a behavior offset amplitude, and a path adjustment frequency are extracted, and a disturbed elasticity index vector group is constructed.

[0020] In a preferred embodiment, if the recovery response time of any path in the disturbed elasticity index vector group exceeds the upper limit of the preset recovery period, and the combination value of the offset amplitude and the adjustment frequency exceeds the target elasticity distribution interval, the path is marked as an elastic response abnormal chain;

[0021] The terminal and path structure marked in the elastic response abnormal chain correspond to the path structure established in the cooperative behavior chain, and the path structure is used as the basis for generating a reverse behavior path.

[0022] In a preferred embodiment, for the terminal set in the elastic response abnormal chain, a trigger time flow and a behavior state transition flow are extracted, and a time behavior generation trajectory graph is constructed; for each behavior transition path in the time behavior generation trajectory graph, a transition logic factor and behavior start original data are extracted; the original data includes personnel identity authentication input, device trigger event, and task scheduling record;

[0023] If the transition logic factor of any behavior transition path does not meet the preset original device trigger condition, the path is determined to be a non-natural behavior path.

[0024] In a preferred embodiment, all non-natural behavior paths are aggregated to form an abnormal behavior chain group, and the abnormal behavior chain group corresponds to the sub-path set marked as the elastic response abnormal chain; for each behavior path in the abnormal behavior chain group, behavior reconstruction verification is performed, and behavior generation credibility factors are generated through joint determination of trajectory closure degree, behavior causal continuity, and identity operation consistency;

[0025] If the behavior generation credibility factor is lower than the preset trust score baseline, and the corresponding elastic recovery index meets the elastic response abnormal chain condition, the path is marked as a pseudo-consistent cooperative behavior chain.

[0026] In a preferred embodiment, for the terminal path structure in the pseudo-consistent cooperative behavior chain, behavior granularity backtracking is performed, and a behavior trust mapping between the path and the task behavior result is established; if the behavior trigger in the behavior trust mapping cannot find a behavior chain support with a response time within a preset time limit in a non-pseudo-consistent chain, the path is marked with a trust weakening weight;

[0027] Reinjection of light disturbance behavior to all paths marked by weight weakening, and reconstruction of disturbance trust evolution graph according to its disturbance recovery period and behavior stable trajectory.

[0028] In a preferred embodiment, if the recovery time variation amplitude of the recovery curve in the disturbance trust evolution graph within the specified disturbance window is lower than the preset time fluctuation threshold, the residual mean of the continuous behavior offset value is lower than the residual determination lower limit, and the number of path continuation segment interruptions does not exceed the preset path continuous interruption threshold, the path is considered as a trust recoverable path.

[0029] The trust recoverable path refers to the chain group in the constructed collaborative path that still satisfies the trust determination condition after disturbance elasticity analysis and behavior generation path inversion verification, and the chain group constitutes the final trusted collaborative path graph.

[0030] All chain groups that do not satisfy the trust recoverable path determination condition are output with a pseudo-consistency chain risk label, and the pseudo-consistency chain risk label is used for subsequent system behavior correction and alarm process.

[0031] A smart park monitoring system based on end-to-end collaboration includes a collaboration extraction module, a disturbance construction module, an anomaly identification module, a path inversion module, and a trust reconstruction module.

[0032] The collaboration extraction module is used to obtain the collaborative behavior structure of multi-source synchronous perception data in the smart park, to extract candidate trusted paths in the park that satisfy terminal response synchronization and type intersection characteristics by constructing an initial collaborative path set and establishing an end-to-end collaborative behavior chain.

[0033] The disturbance construction module is used to perform disturbance injection on the terminals in the preliminary trusted collaborative chain and extract disturbance feedback results to generate elastic response trajectories with quantifiable characteristics.

[0034] The anomaly identification module is used to identify elastic response anomaly chains that can be used for reverse derivation of the generated path by constructing and comparing with abnormal conditions.

[0035] The path inversion module identifies non-natural behavior paths that do not meet the device response conditions by constructing behavior generation trajectory graphs and calculating transition logic factors in the path, and outputs pseudo-consistency collaborative behavior chains.

[0036] The trust reconstruction module is used to perform light disturbance verification and recovery feature reconstruction on the behavior trust weakened path, to identify trust recoverable paths that satisfy multiple disturbance response constraint conditions, and to output the final trusted collaborative path graph and the pseudo-consistency chain risk label for subsequent intervention decision of the system.

[0037] The technical effects and advantages of the present application are as follows:

[0038] 1. This solution injects disturbances into the collaborative path and extracts the terminal's elastic response behavior, breaking the assumption that "collaboration is authenticity" and substantially verifying the authenticity of the collaborative path.

[0039] 2. By constructing a disturbance resilience indicator vector group and comprehensively analyzing behavioral deviations, recovery time, and path adjustment frequency, we can accurately identify pseudo-cooperative paths that are structurally unresilient.

[0040] 3. Based on the behavior transfer logic factor and time state trajectory diagram, it can identify the behavior path that does not conform to the device response law and realize the structural judgment of the unnatural path;

[0041] 4. Use a light-disturbance backtest and disturbance recovery scoring mechanism to conduct low-intervention verification of trust-weakening paths, build a final trusted collaborative path map, and mark pseudo-chain risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flowchart of the framework of the method steps of the present invention.

[0043] Figure 2 It is a schematic diagram of the system module structure of the present invention.

[0044] Figure 3 This is a flowchart of the collaborative behavior structure extraction and preliminary trusted path screening of the present invention.

[0045] Figure 4 This is a flow chart of the terminal disturbance injection and elasticity index extraction of the present invention.

[0046] Figure 5 This is a flow chart of the abnormal chain behavior path inversion and pseudo-consistent chain identification of the present invention.

[0047] Figure 6 This is a flowchart of the trust weakening and disturbance backtesting and re-verification process of the present invention.

[0048] Figure 7 This is the output trusted path diagram and risk label flow chart of the present invention. DETAILED DESCRIPTION

[0049] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0050] Refer to the instruction manual Figures 1-7 According to an embodiment of the present invention, a smart park monitoring method based on end-to-end collaboration includes:

[0051] The cooperative behavior structure of acquiring multi-source synchronous perception data in a smart park is obtained by constructing an initial cooperative path set and establishing an end-to-end cooperative behavior chain, and extracting a candidate trusted path in the park that meets the terminal response synchronization and type cross characteristics;

[0052] The disturbance injection is performed on the terminal in the preliminary trusted cooperative chain, and the disturbance feedback result is extracted to generate an elastic response trajectory with quantifiable characteristics;

[0053] By constructing an elastic index and comparing with an abnormal condition, an elastic response abnormal chain that can be used for reverse derivation to generate a path is identified;

[0054] By constructing a behavior generation trajectory graph and calculating the transfer logic factor in the path, a non-natural behavior path that does not meet the device response condition is identified, and a pseudo-consistent cooperative behavior chain is output;

[0055] The light disturbance verification and recovery feature reconstruction are performed on the path weakened by behavior trust, the trust recoverable path that meets the multi-disturbance response constraint condition is identified, and the final trusted cooperative path graph and the pseudo-consistent chain risk label are output for subsequent intervention decision of the system.

[0056] It should be noted that for the formula structure involved in the present scheme, the dimensionless term can be used as a proportional or structural adjustment factor. When combined with quantities with units, it only plays a numerical scaling role and does not introduce new physical dimensions, so it will not change or confuse the unit system of the whole expression; Such a combination of "dimensionless term and quantity unit term" can be understood as a composite structure expression commonly used in mathematical and physical modeling, which conforms to the principle of dimensional consistency and has a clear physical interpretation basis;

[0057] Secondly, in the formula structure of the present scheme, if it involves multiple variable terms with different physical units, including but not limited to time, mass or energy variables, their joint occurrence is to express the cooperative modeling relationship of multiple physical mechanisms. Each variable can be mapped by a function, combined by a ratio, or normalized to form a unified structure. The unit is clear, the meaning is clear, and the whole expression conforms to the principle of dimensional consistency and the common norm of engineering modeling;

[0058] In the present scheme, if a constant, weight, adjustment factor, threshold parameter, proportion coefficient, etc. is designed, it belongs to adjustable control parameters for different application environments. Its value depends on the target device configuration, data input characteristics and performance optimization target. In the implementation stage, it is set within a reasonable range through model verification, performance constraint or engineering calibration, etc. Such parameters, although not pre-set with a unique value, have a clear adjustment logic and calculation path, and belong to the deterministic setting process in engineering implementation. The purpose of such setting is to ensure that the scheme has both general adaptability and reproducibility and operability, without affecting its technical clarity and implementability;

[0059] The perception behavior data set is obtained in the smart park monitoring. The perception behavior data set includes the card swiping event sequence, face recognition record sequence and movement trajectory sequence of the smart park. Through the perception behavior data set, the timestamp alignment operation is performed to output the collaborative candidate data window; the cross-modal fusion operation is performed on each set of synchronous perception data in the collaborative candidate data window, and the initial collaborative path set is output through the behavior matching calculation; the calculation logic of the cross-modal fusion operation is: the identity identification value in the card swiping data is mapped to a discrete coding value, the image feature vector in the face recognition result is normalized, and the position coordinate sequence in the trajectory data is extracted. The three types of data are aligned by time slice and spliced ​​into a joint behavior feature vector group. The label matching score between the identity code and the image features, the displacement direction consistency score between the image features and the trajectory behavior, and the temporal synchronization score between the card swiping time and the trajectory time are calculated respectively. Finally, the weighted average of the above three scores according to the preset weights is used as the behavior matching value of the time slice. If there are multiple behavior matching values ​​exceeding the set threshold in consecutive time slices, the associated card swiping data, facial image data, and trajectory data in this time period are marked as a set of initial collaborative paths.

[0060] For each path in the initial collaborative path set, extract its collaborative terminal combination group; the collaborative terminal combination group includes a card swiping device, a face recognition terminal, and a trajectory sensing node;

[0061] As a further solution, each set of synchronous perception data in the collaborative candidate data window is subjected to a triple fusion of card swiping identity encoding, facial image features, and trajectory vectors to construct a joint behavior feature structure. The behavior matching degree is calculated as the basis for the initial collaborative path formation, and based on this, a cross-modal behavior matching degree joint model is constructed:

[0062]

[0063] The identity label alignment

[0064]

[0065] The displacement direction consistency

[0066]

[0067] Time synchronization strength

[0068]

[0069] Among them, M tis a behavior matching value, the behavior matching value represents whether the data set satisfies the condition of constructing a collaborative path in cross-modal consistency at time slice t; I t is a card swiping event identity label code sequence, the card swiping event identity label code sequence occurs in time slice t; F t is a face image deep feature tensor, the face image deep feature tensor is generated by an image acquisition terminal at time slice t; T t represents a trajectory position vector sequence in a corresponding time slice, the trajectory position vector sequence is collected from a trajectory perception terminal; is an identity matching function, in actual application, the identity matching function is used to evaluate the semantic matching degree of a card swiping identity and a face image; δ i (I t ) is a state indication function of the i-th bit in the identity code; is the i-th sub-feature region in the image tensor; is an identity semantic feature reference subspace; is a behavior direction consistency function; is a motion direction vector of the j-th region in the image frame; is a motion direction vector of the j-th time period in the trajectory segment; is a time synchronization function; is the occurrence time of the k-th event in the card swiping event sequence; is the sampling time of the k-th point in the trajectory sequence; ΔT max is a maximum allowed time deviation window; is a logical indication function for judging whether the card swiping time is within the corresponding time range of the trajectory path; respectively represent element-level coupling and function output-level fusion operators; log represents that the final behavior matching value is compressed to a comparable range; n represents I t the total number of identity feature bits participating in alignment calculation; m represents the total number of corresponding frame pairs between a set of images and trajectories used to construct direction vectors; q represents the total number of pairs of card swiping events and trajectory sampling points participating in time synchronization calculation within a time window; in addition, rank in the above formula represents the semantic matching ranking position in the pre-defined , used to measure the matching priority with the identity code; the symbol in the formula represents the mapping projection operation in , used to construct an image semantic representation relationship matched with the card swiping identity code; ∠ represents the included angle between , in the above formula, the symbol ∠ is used to measure their behavior direction consistency.

[0070] Perform structural connectivity analysis on the path structure of the collaborative terminal combination group to obtain the end-to-end collaborative behavior chain, build a consistent collaborative path graph, extract the corresponding terminal response time series for each path chain in the consistent collaborative path graph, perform response timing continuity detection, and output the behavior synchronization chain group;

[0071] If the terminal response time interval fluctuation in the behavior synchronization chain group is less than the preset interval fluctuation threshold, and the terminal type cross ratio is not lower than the preset cross ratio threshold lower limit, the corresponding chain in the behavior synchronization chain group is marked as a preliminary trusted collaborative chain.

[0072] Inject a disturbance request into each terminal in the preliminary trusted collaborative chain; the disturbance request includes path deviation disturbance, identity confidence disturbance and trigger timing disturbance;

[0073] By injecting the terminal behavior sequence after the disturbance, the disturbance response behavior trajectory set is extracted to construct the end-to-end disturbance feedback flow; for each terminal trajectory in the disturbance feedback flow, dynamic recovery cycle analysis is performed to output the disturbance recovery path set;

[0074] For each path in the disturbance recovery path set, the recovery response time, behavior deviation amplitude and path adjustment frequency are extracted to construct a disturbance resilience index vector group.

[0075] If the recovery response time of any path in the disturbance resilience index vector group exceeds the preset recovery period upper limit, and the combined value of the offset amplitude and adjustment frequency exceeds the target resilience distribution range, the path is marked as a resilience response abnormal chain;

[0076] The terminals and path structures marked in the elastic response anomaly chain correspond to the established path structures in the collaborative behavior chain, and the path structures are used as the basis for reverse derivation of the behavior generation path;

[0077] As a further solution, disturbances are injected into the terminals of the preliminary trusted collaborative chain. By analyzing the behavior deviation, trajectory recovery dynamics, and path state change frequency, a disturbance resilience index is constructed to determine whether it is an abnormal chain. Based on this, a disturbance resilience structure identification and abnormal chain judgment model is constructed:

[0078]

[0079] The first-order derivative of the recovery time curve is described as:

[0080]

[0081] The spatial curvature function Π of the behavioral offset i (t):

[0082]

[0083] Path i-th segment state transition difference value

[0084]

[0085] wherein ε i is the perturbation elasticity anomaly value of path i, the perturbation elasticity anomaly value is used to determine whether it is an elastic response anomaly chain; Y i (t) is the difference function of recovery process time before and after the perturbation; is the time function of path i reaching a stable behavior state; is the time function of path i receiving perturbation input; is the perturbation behavior offset vector of path i at time t; θ i (t) is the angle between the direction vectors before and after the perturbation; represents the time derivative and second-order partial derivative operation; L is the number of path segments; l represents the i-th behavior state segment of path i, which is used to represent the index number of each segment state transition of the path after segmentation in the perturbation recovery process; sup represents the maximum second-order change rate extracted in the recovery interval; represents the state number of the i-th state segment of the path at the current time, which is used to identify the real-time behavior state of the segment in the perturbation recovery process; represents the state number of the i-th state segment at the previous time frame, which is used to compare with the current state to determine whether a state transition occurs.

[0086] For the terminal set in the elastic response anomaly chain, the trigger time flow and behavior state transition flow are extracted, and a time behavior generation trajectory graph is constructed. For each behavior transition path in the time behavior generation trajectory graph, the transition logic factor and behavior start original data are extracted; the original data includes personnel identity authentication input, device trigger event and task scheduling record; wherein the transition logic factor refers to whether the state transition between each behavior node and its previous node in the behavior transition path conforms to the device trigger logic and behavior causal law, and the calculation logic is: the time interval between the adjacent two behavior nodes, the device activation state and the personnel identity continuity are jointly matched, if the time is within the device response window, the identity remains consistent and the behavior type meets the pre-defined state transition rule, then the transition logic factor of this segment is output as valid; otherwise, the factor is marked as violating the trigger logic, which is used to determine whether the behavior path is natural;

[0087] If the transition logic factor of any behavior transition path does not meet the preset original device trigger condition (the original device trigger condition includes but is not limited to the preset condition: personnel does not produce card swiping behavior but there is device response record), it is determined that the path is a non-natural behavior path.

[0088] The abnormal behavior chain groups are formed by summarizing all non-natural behavior paths, and the abnormal behavior chain groups correspond to a sub-path set marked as an elastic response abnormal chain; for each behavior path in the abnormal behavior chain group, behavior reconstruction verification is performed, and a behavior generation credibility factor is generated through joint determination of trajectory closure degree, behavior causal continuity and identity operation consistency;

[0089] If the behavior generation credibility factor is lower than the preset trust score baseline, and the corresponding elastic recovery index satisfies the elastic response abnormal chain condition, the path is marked as a pseudo-consistent cooperative behavior chain.

[0090] For the terminal path structure in the pseudo-consistent cooperative behavior chain, behavior granularity backtracking is performed, and a behavior trust mapping between the path and the task behavior result is established; if the behavior trigger in the behavior trust mapping cannot find a behavior chain support with a response time within a preset time limit in a non-pseudo-consistent chain, the path is marked with a trust weakening weight;

[0091] All paths marked with a weight weakening mark are re-injected with a light disturbance behavior, and a disturbance trust evolution graph is reconstructed according to the disturbance recovery period and the behavior stable trajectory.

[0092] If the recovery curve in the disturbance trust evolution graph has a recovery time variation amplitude within a specified disturbance window lower than a preset time fluctuation threshold, a residual mean of a continuous behavior offset value is lower than a residual determination lower limit, and a path continuation interruption number does not exceed a preset path continuous interruption threshold, the path is considered as a trust recoverable path;

[0093] The trust recoverable path refers to a chain group in the constructed cooperative path that still satisfies the trust determination condition after disturbance elastic analysis and behavior generation path inversion verification, and the chain group constitutes a final trusted cooperative path graph;

[0094] All chain groups that do not satisfy the trust recoverable path determination condition are output with a pseudo-consistent chain risk label, and the pseudo-consistent chain risk label is used for subsequent system behavior correction and alarm processes;

[0095] As a further scheme, it is necessary to inject a light disturbance into the path marked with a trust weakening mark and observe the whole disturbance recovery process, to construct an overall score through recovery time fluctuation, behavior error dynamics and path connectivity, to determine whether it can be recovered as a trusted path, and to construct a recovery strength comprehensive determination model of the trust recoverable path based on this, so as to output a final trusted path;

[0096]

[0097] wherein is the disturbance recovery strength score of path j; H j(t) is a disturbance recovery time curve function, the disturbance recovery time curve function representing a response stability of the path at time t; is a second-order time fluctuation in the recovery process; Ξ j (t) is a behavior error dynamic change rate of the path j at time t; is a theoretically appearing behavior state vector; is an actually observed behavior state vector; ρ j (t) is a structure connectivity function of the path segment; is the number of interruptions of the path j within the observation window; λ is a connectivity decay coefficient for adjusting the influence of the interruption on the score; τ0, τ1 are start and end boundaries of the recovery observation time window.

[0098] An intelligent park monitoring system based on end-to-end cooperation, comprising a cooperation extraction module, a disturbance construction module, an anomaly identification module, a path inversion module, and a trust reconstruction module;

[0099] The cooperation extraction module is used to obtain the cooperative behavior structure of multi-source synchronous perception data in the intelligent park, to extract candidate trusted paths in the park that meet the terminal response synchronization and type cross characteristics by constructing an initial cooperative path set and establishing an end-to-end cooperative behavior chain.

[0100] The disturbance construction module is used to perform disturbance injection on the terminals in the preliminary trusted cooperative chain and extract disturbance feedback results, to generate an elastic response trajectory with quantifiable characteristics.

[0101] The anomaly identification module is used to identify an elastic response anomaly chain that can be used to generate a path in reverse by constructing and comparing with abnormal conditions through elastic indicators.

[0102] The path inversion module identifies a non-natural behavior path that does not meet the device response conditions by constructing a behavior generation trajectory and calculating a transfer logic factor in the path, and outputs a pseudo-consistent cooperative behavior chain.

[0103] The trust reconstruction module is used to perform light disturbance verification and recovery feature reconstruction on the paths with weakened behavior trust, to identify trust-recoverable paths that meet multiple disturbance response constraint conditions, and to output an ultimate trusted cooperative path graph and a pseudo-consistent chain risk label for subsequent intervention decisions of the system.

[0104] It is necessary to make a general description, including but not limited to: in the existing intelligent park scene, identity verification and behavior judgment generally rely on multi-terminal cooperative mechanism, such as card swiping behavior, face recognition, WiFi trajectory, etc. After being aligned on the time axis, the system jointly participates in the judgment of the behavior legality; when these terminal data are cooperative in time and the contents match, the system will automatically identify the behavior chain as trusted, thereby triggering actions such as gate release, task authorization or exception exemption.

[0105] However, existing technologies assume that all terminal data sources are inherently trustworthy and lack the ability to model the authenticity of this "cooperative relationship itself." Attackers can construct a set of forged information with consistent time and matching data to simulate "legitimate behavior" patterns, thereby deceiving the system into misleading trust and bypassing monitoring. This logic, which relies on "superficial consistency," is more susceptible to mistrust when the attack is more intense and the forged behavior is more coordinated, creating a serious structural paradox.

[0106] Therefore, this invention does not optimize the accuracy of a single terminal, but rather proposes a full-link solution to the core issue of "whether the collaborative path is true";

[0107] This solution includes the collaborative behavior chain construction stage:

[0108] Acquire synchronized behavioral data collected by multiple sensing terminals in the smart campus, including card swipe event sequences, facial recognition results, and trajectory path sequences. Through timestamp alignment, spatial location synchronization, and identity semantic fusion, construct joint behavioral features within each time slice. Based on these features, calculate cross-modal behavioral matching and group highly matching segments into an "initial collaborative path set." This allows for further screening of "candidate trusted paths" that meet the requirements for terminal response synchronization and device type cross-talk.

[0109] The design intention at this stage is to not use single terminal data as the basis for behavior, but to use the structural strength and inter-modal consistency of the collaborative chain as the evaluation unit, providing a chain-level data foundation for subsequent disturbance judgment;

[0110] This solution includes the disturbance response behavior construction phase:

[0111] For the initially trusted path, the system injects interference requests into the corresponding terminal. Interference methods include not only trajectory perturbations (such as path deviations), but also identity confidence perturbations (simulating misidentification) and trigger timing perturbations (controlling response time differences). By recording the terminal's behavioral feedback trajectory under interference, the system extracts the response time, behavioral deviation trajectory, and state adjustment frequency of each path during the recovery process, and constructs a "disturbance resilience index vector group."

[0112] An attacker can construct consistent paths simultaneously, but cannot truly reproduce the path's resilience under interference conditions. By constructing a perturbation-recovery mapping relationship, it is possible to infer which paths exhibit only superficial coordination and lack true behavioral inertia, thereby identifying "elastic response anomaly chains."

[0113] This solution includes the reverse behavior generation path derivation stage:

[0114] For the identified abnormal chain path, the behavior generation process thereof is extracted reversely, and a "time-state trajectory graph" is constructed for analyzing whether each node conforms to the natural generation logic; the judgment is performed by calculating the transition logic factor between the behavior nodes, such as whether the response time difference between adjacent behaviors falls within the equipment response window, whether the identity is continuous, whether the behavior trigger is reasonable, and the like;

[0115] This part distinguishes the behavior chain actually executed by the user from the pseudo path spliced by the attacker through "generation path inversion", and strengthens the internal verification of the authenticity of the cooperative path structure;

[0116] The scheme includes a behavior trust decay and re-verification stage:

[0117] If a behavior path is identified as a "pseudo-consistent chain", the system will not immediately exclude it, but will perform "trust weakening processing" on it; that is, the path behavior is mapped to the actual task execution record, and if no behavior chain can be found to support the trigger in the non-pseudo-consistent chain, the path trust degree is reduced; thereafter, the system re-injects low-intensity disturbance to it and collects its recovery trajectory for further establishing a disturbance recovery strength score;

[0118] Here, the scheme does not rely on a rigid strategy of "one-step judgment", but has multi-round dynamic feedback verification capability, which improves the misjudgment recovery rate and system flexibility;

[0119] The scheme includes a trust path reconstruction and pseudo-chain marking stage:

[0120] A plurality of recovery constraint judgments are performed on the recovery score results, including disturbance recovery time fluctuation amplitude, behavior offset residual change trend, and path interruption number, and the like, are jointly judged; when the path maintains a stable recovery mode within all constraint conditions, it is determined as a "trust recoverable path" and is re-included in the trusted path atlas; otherwise, a "pseudo-consistent chain risk label" is output for subsequent monitoring system to perform alarm or intervention;

[0121] This stage constitutes the closing mechanism of the system security judgment, ensures that only the paths that have really experienced interference test and recovered stably are confirmed as trusted, and thus realizes the paradigm shift from "surface cooperative judgment" to "process trusted verification".

[0122] The above only describes the preferred embodiments of the present application and is not used to limit the present application, and any modification, equivalent replacement, improvement, and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A smart park monitoring method based on end-to-end collaboration, characterized in that: include: Acquire the collaborative behavior structure of multi-source synchronous perception data in the smart park, construct an initial collaborative path set and establish an end-to-end collaborative behavior chain, and extract candidate trusted paths within the park that meet the terminal response synchronization and type intersection characteristics; Perform disturbance injection on the terminals in the preliminary trusted collaborative chain and extract the disturbance feedback results to generate elastic response trajectories with quantifiable characteristics; By comparing resilience indicators with abnormal conditions, we can identify abnormal resilience response chains that can be used to reversely derive generation paths. By constructing a behavior generation trajectory graph and calculating the transfer logic factor in the path, it identifies unnatural behavior paths that do not meet the device response conditions and outputs a pseudo-consistent collaborative behavior chain; Perform light perturbation verification and recovery feature reconstruction on the paths weakened by behavioral trust, identify trust-recoverable paths that meet multiple perturbation response constraints, and output the final trusted collaborative path diagram and pseudo-consistency chain risk label for subsequent system intervention decisions.

2. The method for monitoring a smart park based on end-to-end collaboration according to claim 1, characterized in that: Acquire a perception behavior dataset from smart park monitoring. The dataset includes card swipe event sequences, facial recognition record sequences, and movement trajectory sequences. Perform a timestamp alignment operation on the perception behavior dataset to output a collaborative candidate data window. Perform a cross-modal fusion operation on each set of synchronized perception data in the collaborative candidate data window, and output an initial collaborative path set through behavioral matching calculation. For each path in the initial collaborative path set, its collaborative terminal combination group is extracted; the collaborative terminal combination group includes a card swiping device, a face recognition terminal, and a trajectory sensing node.

3. The method for monitoring a smart park based on end-to-end collaboration according to claim 2, characterized in that: Perform structural connectivity analysis on the path structure of the collaborative terminal combination group to obtain the end-to-end collaborative behavior chain, build a consistent collaborative path graph, extract the corresponding terminal response time series for each path chain in the consistent collaborative path graph, perform response timing continuity detection, and output the behavior synchronization chain group; If the terminal response time interval fluctuation in the behavior synchronization chain group is less than the preset interval fluctuation threshold, and the terminal type cross ratio is not lower than the preset cross ratio threshold lower limit, the corresponding chain in the behavior synchronization chain group is marked as a preliminary trusted collaborative chain.

4. The method for monitoring a smart park based on end-to-end collaboration according to claim 3 is characterized by: Inject a disturbance request into each terminal in the preliminary trusted collaborative chain; the disturbance request includes path deviation disturbance, identity confidence disturbance and trigger timing disturbance; By injecting the terminal behavior sequence after the disturbance, the disturbance response behavior trajectory set is extracted to construct the end-to-end disturbance feedback flow; for each terminal trajectory in the disturbance feedback flow, dynamic recovery cycle analysis is performed to output the disturbance recovery path set; For each path in the disturbance recovery path set, the recovery response time, behavior deviation amplitude and path adjustment frequency are extracted to construct a disturbance resilience index vector group.

5. The method for monitoring a smart park based on end-to-end collaboration according to claim 4 is characterized in that: If the recovery response time of any path in the disturbance resilience index vector group exceeds the preset recovery period upper limit, and the combined value of the offset amplitude and adjustment frequency exceeds the target resilience distribution range, the path is marked as a resilience response abnormal chain; The terminals and path structures marked in the elastic response exception chain correspond to the established path structure in the collaborative behavior chain, and the path structure is used as a basis for reverse deducing the behavior generation path.

6. The method for monitoring a smart park based on end-to-end collaboration according to claim 5, characterized in that: For the terminal set in the elastic response anomaly chain, its trigger time flow and behavior state transition flow are extracted to construct a time behavior generation trajectory diagram. For each behavior transition path in the time behavior generation trajectory diagram, its transfer logic factor and behavior initiation raw data are extracted. The raw data includes personnel identity authentication input, device trigger events, and task scheduling records. If the transfer logic factor of any behavior transfer path does not meet the preset original device trigger condition, the path is determined to be an unnatural behavior path.

7. The method for monitoring a smart park based on end-to-end collaboration according to claim 6, characterized in that: All unnatural behavior paths are aggregated to form abnormal behavior chain groups, which correspond to sub-path sets marked as elastic response abnormal chains. For each behavior path in the abnormal behavior chain group, behavior reconstruction verification is performed. The behavior generation credibility factor is generated through the joint judgment of trajectory closure, behavior causal continuity, and identity operation consistency. If the behavior generation credibility factor is lower than the preset trust score baseline and its corresponding elastic recovery index meets the elastic response anomaly chain condition, the path is marked as a pseudo-consistent collaborative behavior chain.

8. The method for monitoring a smart park based on end-to-end collaboration according to claim 7, characterized in that: For the terminal path structure in the pseudo-consistent collaborative behavior chain, behavior granularity backtracking is performed to establish a behavior trust mapping between the path and the task behavior result; if the behavior trigger in the behavior trust mapping fails to find a behavior chain support with a response time within the preset time limit in the non-pseudo-consistent chain, the path is marked with a trust weakening weight; Lightly perturbed behaviors are re-injected into all weight-weakened marked paths, and the perturbation trust evolution graph is reconstructed based on their perturbation recovery cycle and behavioral stability trajectory.

9. The method for monitoring a smart park based on end-to-end collaboration according to claim 8, characterized in that: A recovery curve in the perturbation trust evolution graph is considered a trustworthy and recoverable path if the recovery time variation within the specified perturbation window is lower than the preset time fluctuation threshold, the residual mean of the continuous behavior offset value is lower than the residual judgment lower limit, and the number of interruptions in the path continuation segment does not exceed the preset path continuous interruption threshold. The trust-recoverable path refers to the chain group in the constructed collaborative path that still meets the trust judgment conditions after disturbance resilience analysis and behavior generation path inversion verification. This chain group constitutes the final trusted collaborative path graph; For all chain groups that do not meet the trusted recoverable path judgment conditions, a pseudo-consistency chain risk label is output. The pseudo-consistency chain risk label is used for subsequent system behavior correction and alarm processes.

10. A smart park monitoring system based on end-to-end collaboration, comprising the smart park monitoring method based on end-to-end collaboration according to claim 9, comprising a collaboration extraction module, a disturbance construction module, an anomaly identification module, a path inversion module, and a trust reconstruction module, characterized in that: The collaborative extraction module is used to obtain the collaborative behavior structure of multi-source synchronous perception data in the smart park. By constructing an initial collaborative path set and establishing an end-to-end collaborative behavior chain, it extracts candidate trusted paths within the park that meet the terminal response synchronization and type intersection characteristics. The perturbation construction module is used to perform perturbation injection on the terminals in the preliminary trusted collaborative chain and extract the perturbation feedback results to generate elastic response trajectories with quantifiable characteristics; The anomaly identification module is used to identify elastic response anomaly chains that can be used to reversely deduce the generation path by comparing elasticity indicators with abnormal conditions; The path inversion module generates a trajectory diagram by constructing behaviors and calculating the transfer logic factors in the path, identifying unnatural behavior paths that do not meet the device response conditions and outputting a pseudo-consistent collaborative behavior chain; The trust reconstruction module is used to perform light perturbation verification and recovery feature reconstruction on paths where behavioral trust is weakened, identify trust-recoverable paths that meet multiple perturbation response constraints, and output the final trusted collaborative path diagram and pseudo-consistency chain risk label for subsequent system intervention decisions.

Images