Digital twin empowered scenic spot full-scene visual management and emergency command system

Abstract

The present application belongs to the technical field of digital management of scenic spots, and is used to solve the problem that the prior art cannot form dynamic coupling between manual information reporting and machine deduction chain, and specifically is a digital twin enabled full-scene visual management and emergency command system for scenic spots, comprising a data acquisition module, an information reading module, a pulse calculation module, a topology construction module and a cancellation analysis module; the present application realizes real-time linkage of manual constraint survival state and changes in field physical conditions, and the system takes the absolute value of the deviation between the actual cumulative number of people passing through the downstream section and the expected cumulative number of people calculated based on the upstream flow and the conductivity coefficient as the erosion criterion of the constraint survival logic; through spatial overlap area calculation and reconstruction of isolated barrier nodes in the topology construction module, the logical ablation of multiple point mutual exclusion instructions in the topology domain is realized; so that the path deduction conclusion of the twin system can continuously converge to a state consistent with the actual physical space.

Description

Technical Field

[0001] This invention belongs to the field of digital management technology for scenic areas, specifically a digital twin-enabled full-scene visualization management and emergency command system for scenic areas. Background Technology

[0002] In the digital twin emergency command system of the scenic area, automated sensing devices (such as video analysis and gate counting) constitute the main source of real-time situational data. The system drives the status update and inference of the digital twin based on these sensor data. At the same time, various handling records reported by on-site patrol personnel through mobile terminals—including information such as temporary road closures, equipment unavailability, and abnormal gathering of tourists—are usually stored independently as event logs in the existing system, or only provided to the commander for reference in the form of a large floating window, and do not participate in the situational calculation and inference process of the twin engine.

[0003] The potential problem with this data isolation design is that when the facts described in the aforementioned manually reported information are true, the sensor data upon which the digital twin relies will fail to reflect the corresponding changes in physical constraints. For example, a passage may actually be closed by a rope, but the sensor may still show it as open; or an evacuation door may be online, but the key may not have been delivered, yet the sensor may still mark it as available. In these situations, because the machine-generated chain does not incorporate the constraints confirmed by on-site personnel, the command instructions such as evacuation routes and diversion plans generated by the twin system will deviate from the actual executable conditions on-site. The cumulative effect of this deviation manifests as follows: the guide screen recommends routes that are actually closed; the broadcast guidance is inconsistent with the on-site security gestures; and crowds repeatedly turn back at key nodes due to conflicting instructions, ultimately leading to unexpected accumulation in local areas.

[0004] A deeper examination of the causes of the above deviations reveals that even if manually reported information is injected into the simulation link in real time, if the system only processes it by "applying constraints" without a corresponding lifecycle management mechanism, new failure modes will still occur. Specifically, firstly, after the on-site constraints are removed, if there is no corresponding manual revocation command, the applied constraints will permanently lock the topology resources, causing subsequent simulations to continuously deviate from the actual physical state; secondly, when multiple manual reports generate mutually exclusive constraints at the same spatial node, the path planning engine will frequently switch between the "open" and "closed" states, making it impossible for the execution end to obtain stable decision signals; thirdly, if a single manual judgment lacks a verification mechanism with sensor physical quantities, once the judgment is incorrect, the system will lose its ability to make safety decisions based on objective physical boundaries.

[0005] All three failure modes mentioned above point to the same structural defect: the existing system lacks a mechanism that enables dynamic coupling between manually reported information and machine deduction chain. This mechanism needs to simultaneously solve the problems of admission verification of manual constraints, conflict resolution, and autonomous elimination linked to changes in physical state. Summary of the Invention

[0006] The purpose of this invention is to provide a digital twin-enabled scenic area full-scene visualization management and emergency command system to solve the problem that existing technologies cannot achieve dynamic coupling between manually reported information and machine-generated deduction chains;

[0007] The technical problem to be solved by this invention is: how to provide a digital twin-enabled scenic area full-scene visualization management and emergency command system that enables dynamic coupling between manually reported information and machine-generated deduction chains.

[0008] The objective of this invention can be achieved through the following technical solutions:

[0009] A digital twin-enabled scenic area full-scene visualization management and emergency command system includes:

[0010] Data acquisition module: Acquires manual control work orders and real-time sensing data streams. The work order includes the spatial location identifier, restriction type, and submitter weight. The data stream includes the instantaneous throughput, velocity field, and density field of each node.

[0011] Information reading module: In response to work order reception, reads the upstream average throughput and downstream average throughput corresponding to the action spatial location identifier, and generates a disturbance object instance containing the initial potential energy base;

[0012] Pulse Calculation Module: Copy the velocity field and density field of the main thread to the shadow calculation space, set boundary conditions according to the constraint type and deduce the first maximum number of people staying, while the main thread deduces the second maximum number of people staying without constraints, and calculates the absolute value of the difference between the two as the dynamic pulse value;

[0013] Topology construction module: Generates the permissible fluctuation envelope threshold based on the dispersion of the number of people in the entire area grid. If the dynamic fluctuation value is not greater than the threshold by a preset multiple, the initial potential energy base is back-modulated according to the relative difference between the two to generate the initial dynamic potential energy value, and the passage constraint corresponding to the restriction type is written into the main thread topology connectivity matrix.

[0014] Cancellation Analysis Module: During the effective period, the actual cumulative number of people passing through downstream and the cumulative number of people entering upstream are obtained. Based on the cumulative number of people entering upstream and the conduction coefficient, the expected cumulative number of people passing through is generated. When the absolute value of the deviation between the actual and expected cumulative number of people passing through is greater than the preset ratio of the initial dynamic potential energy value, the passage constraint is cancelled.

[0015] The present invention has the following beneficial effects:

[0016] 1. By parallel deduction of the shadow thread and the main thread in the pulse calculation module and the generation of dynamic pulse values, physical consistency quantitative verification before the injection of artificial constraints is realized. After copying the velocity field and density field of the main thread at the current moment in the shadow space, the constraint to be verified is applied. By comparing the difference in the maximum number of people staying in the entire domain before and after the constraint is applied, a quantitative index characterizing the degree of deviation between the constraint and the current physical inertia is obtained. This index serves as the basis for the admission gate, so that artificial constraints that deviate significantly from the pedestrian flow conservation law reflected by the sensor data are rejected before injection, eliminating the possibility of single-point subjective misjudgment overwriting the underlying safety boundary from the data flow entry point. This mechanism places human experience on an equal footing with the physical conservation law, solving the problem in the existing system where the lack of objective physical quantity verification of human judgment leads to the twin losing its safety adjudication ability.

[0017] 2. By continuously monitoring the cumulative flux conservation difference between upstream and downstream sections and setting an asymmetric hysteresis threshold through the cancellation analysis module, the system achieves real-time linkage between the artificial constraint persistence status and changes in on-site physical conditions. The system uses the absolute value of the deviation between the actual cumulative number of people passing through the downstream section and the expected cumulative number of people calculated based on the upstream flow and conductivity coefficient as the erosion criterion for constraint persistence logic. When this deviation value exceeds the preset proportion of the initial dynamic potential energy value, the constraint automatically disappears without waiting for a manual cancellation command. On this basis, the system further adopts an asymmetric hysteresis configuration where the disappearance trigger threshold is higher than the recovery threshold, so that random fluctuations in sensor counts near the critical value will not lead to high-frequency switching of the control state. This mechanism makes the life cycle of the constraint completely driven by the actual throughput of objective pedestrian flow, eliminating the phenomenon of topological resources being permanently locked due to the lack of disappearance perception.

[0018] 3. By calculating the spatial overlap area and reconstructing the isolation barrier nodes in the topology construction module, the logical ablation of multi-point mutual exclusion commands in the topology domain is realized. When the spatial distance between two or more active disturbance object instances is less than the preset grid edge length threshold and the constraint type is marked as mutually exclusive in the logical mutual exclusion table, the system calculates the spatial overlap area of ​​the affected circles of each instance. When the overlap area exceeds the preset area threshold, the corresponding grid edge is simultaneously removed from the topology connectivity matrix for both passage status indicators and reconstructed into an isolation node that the path planning algorithm cannot recognize. This processing method transforms the binary oscillation in the signal domain into a static structural modification in the topology domain, solving the problem that multi-point mutual exclusion constraints cause the control signal to flip frequently and fail to form a stable decision output.

[0019] 4. Through the cascaded data flow between the data acquisition module, information reading module, pulse calculation module, topology construction module, and undoing analysis module, each state transition node of the artificial constraint object has an independently verifiable physical criterion throughout its complete lifecycle from generation to destruction. The dynamic pulsation value before constraint injection provides the physical basis for admission, the flux conservation difference during constraint effectiveness provides the physical basis for survival, and the hysteresis lag after constraint destruction prevents false state recovery from a physical perspective. Throughout the entire operation cycle, each state change in the main thread's topology connectivity matrix corresponds to a set of traceable physical measurements, enabling the path deduction conclusions of the twin system to continuously converge to a state consistent with the actual physical space without increasing the cost of on-site manual verification. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a system block diagram of Embodiment 1 of the present invention;

[0022] Figure 2 This is a flowchart of the system method of Embodiment 1 of the present invention. Detailed Implementation

[0023] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] In the digital twin emergency command system of scenic areas, an automated sensor network forms the data foundation for situational awareness. The system drives the twin's status updates and path deduction based on real-time video analysis and gate counting. However, information such as temporary road closures, equipment unavailability, and abnormal tourist gatherings reported by on-site patrol personnel via mobile terminals is usually isolated from the machine deduction link under the existing technical framework. It is only stored as an independent log or provided to the commander for visual reference in the form of a floating window. This data isolation architecture means that when the physical constraints described by the manually reported information are met, the sensor data on which the twin relies cannot reflect the corresponding topological changes. For example, a passage may actually be closed by a rope, but the video analysis still shows it as open; or an evacuation door may be online, but the key has not been delivered, yet the gate signal still marks it as available. A structural deviation occurs between the evacuation path generated by the twin system and the actual feasible conditions on site. Crowds repeatedly turn back at key nodes because the path recommended by the guide screen is inconsistent with the actual passage status, ultimately leading to unexpected accumulation in local areas.

[0025] A deeper examination of the generation mechanism of the above deviations reveals that even if manually reported information is injected into the simulation link in real time, if the system only processes it by applying rigid constraints without a corresponding lifecycle management mechanism, three concurrent failure modes will still be triggered. First, if there is no corresponding manual revocation command after the on-site constraints are released, the injected topology lock will remain permanently in the connectivity matrix of the path planning engine, causing all subsequent simulations to continuously deviate from the actual physical state until the system restarts. Second, when multiple inspection personnel simultaneously submit mutually exclusive control commands at adjacent spatial nodes, the path planning engine repeatedly flips the connectivity state of the target mesh edges between adjacent sampling periods, causing the execution end to... The system receives high-frequency oscillating control signals on a millisecond timescale, making it impossible to form a stable decision output. Third, if a single manual judgment lacks runtime verification with the physical quantities of the sensors, in extreme conditions such as fires, once the judgment contradicts the objective law of pedestrian flow conservation, the system loses its underlying decision-making ability based on physical boundaries. All three failure modes point to the same structural defect: the existing system lacks a dynamic coupling mechanism that can form a two-way game loop between subjective constraints and objective physical fields. This mechanism needs to simultaneously solve the problems of admission verification before constraint injection, conflict resolution after injection, and autonomous elimination linked to real-time pedestrian flow changes within the same time frame.

[0026] Example 1: As Figure 1-2 As shown, the digital twin-enabled scenic area full-scene visualization management and emergency command system includes:

[0027] Data acquisition module: Acquires manual control work orders and real-time sensing data streams. The work order includes the spatial location identifier, restriction type, and submitter weight. The data stream includes the instantaneous throughput, velocity field, and density field of each node.

[0028] In this embodiment, the data acquisition module serves as the system input, responsible for the synchronous collection and standardized preprocessing of real-time sensing data streams and manual control work orders. The physical space of the scenic area is pre-divided into discrete grids of equal size, each grid having a unique spatial coordinate identifier. Binocular vision cameras are deployed at each passage cross-section, and the instantaneous pass rate of each grid node is output after being analyzed by the back-end image processing unit. Velocity field and density field The sampling step size for all three is fixed at 1 second (the speed of pedestrian flow in scenic areas is usually in the range of 0.5 to 1.5 meters per second, and a 1-second sampling step size corresponds to a pedestrian movement of about 0.5 to 1.5 meters, which matches the spatial resolution of the discrete grid side length; this step size is sufficient to capture the macroscopic changing trend of pedestrian flow while avoiding excessive computational load); instantaneous throughput Defined as the number of pedestrians passing through the cross section within a unit time window, measured in persons per second; during cross-frame tracking, if a target reappears in a subsequent frame after being lost in the current frame, the system does not count it as a newly entered target. Instead, it uses cross-union matching combined with Kalman filtering prediction for target re-identification. That is, taking the target's position in the last frame before it was lost as the starting point, it uses Kalman filtering to predict its trajectory during the loss period. When the Euclidean distance between the re-detected target position and the predicted position is less than 0.5 meters, it is determined to be the same target and its tracking number continues; when the displacement of the same target between two adjacent frames exceeds 2.5 meters per second, the tracking result is determined to be abnormal, and the cross-frame matching result corresponding to the target is not included in the instantaneous pass rate statistics of the current frame; velocity field The vector average of the instantaneous velocity of the target is measured in meters per second; density field. The density field is the ratio of the instantaneous number of people in the grid to the grid area, measured in people per square meter. The system also maintains an independent sensor status flag for each grid node. When the number of target detections for a grid node is zero for ten consecutive sampling periods, and the arithmetic mean of the density of the eight adjacent grid nodes is higher than 0.1 people per square meter, the density field of that node is marked as "data missing" and written into the status flag. When subsequent calculation modules that rely on density field data read the data of that node, if they detect that the status flag is "data missing", they will skip the value of that node at the current moment and use the arithmetic mean of the density values ​​of the eight adjacent grid nodes as a substitute.

[0029] On-site inspectors submit manual control work orders via a mobile application. The application provides a standardized input interface with three required fields: spatial location identifier (selected from grid coordinates or a dropdown menu), event type identifier (selected from a preset list; the system internally maps "fully enclosed control" to a restriction type code). Mapping "traffic restriction control" to The congestion level is categorized into "light / medium / severe" (mapped to values ​​1, 2, and 3 respectively). When a work order is submitted, the submitter's identity is automatically appended. The system backend retrieves the weighted attribute from a pre-stored job grade coefficient lookup table. The on-site supervisor corresponds to Ordinary patrolman .

[0030] The data acquisition module continuously monitors the work order submission port; when a work order arrives, it records the reception time as the baseline time. Then, it reads the instantaneous throughput sequence of the upstream nearest section at the grid node corresponding to the spatial location identifier of the work order from the shared data area over the past 60 sampling periods. And the instantaneous throughput sequence corresponding to the nearest downstream section. Calculate the arithmetic mean of the two sequences to obtain the upstream average throughput. Compared with downstream average throughput Its expression is:

[0031] ;

[0032] ;

[0033] Both units are in persons per second. If data is missing within the 60-second window due to momentary communication interruption or camera obstruction, for cases where there are at least two valid sampling points before and after the missing point (linear interpolation requires at least two known points to determine the slope; when there are fewer than two valid points at the boundary, linear interpolation cannot construct a reliable slope, and in this case, it should be switched to extrapolation mode), the instantaneous throughput of the two valid sampling moments before and after the missing moment is used for linear interpolation completion; for cases where the missing point is located at the beginning boundary of the window and there are fewer than two valid sampling points before the missing point, only the two adjacent valid sampling moments after the missing point are used for linear extrapolation completion; for cases where the missing point is located at the end boundary of the window and there are fewer than two valid sampling points after the missing point, only the two adjacent valid sampling moments before the missing point are used for linear extrapolation completion; if there are more than five consecutive missing sampling moments within the entire 60-second window ( The 60-second window is a commonly used short-term average window in traffic flow engineering. It can effectively smooth out instantaneous fluctuations (such as random noise from traffic light switching and pedestrians stopping to take photos) and is sufficient to respond to changes in pedestrian flow trends within 1 minute. If the window is less than 30 seconds, the fluctuation noise is too large, and if it is more than 120 seconds, the response lag is obvious. 60 seconds is the balance point between the two. If the total number of valid sampling moments is less than 30, the data in this window is discarded, and the complete window from the 61st to the 120th second before the reference time is used as the alternative data source. If the alternative window still does not meet the requirement that the total number of valid sampling moments is not less than 30, then all missing values ​​in the window are uniformly assigned the median of the instantaneous pass rate of the valid sampling moments in the window (the median has a stronger resistance to interference from extreme values ​​than the mean. It serves as a last resort when data is severely missing, ensuring that the average pass rate calculation still has numerical output and does not affect the continuity of subsequent modules).

[0034] After completing the above data acquisition and preprocessing, the data acquisition module will process the structured fields of the manually controlled work order—including the spatial location identifier and the restriction type code. Submitter weight attribute —Including the average upstream throughput at that moment Compared with downstream average throughput Encapsulated into a data packet, it is transmitted to the information reading module to generate a perturbation object instance; among the above parameters, and This will be used as input for calculating the initial potential energy base. This will serve as a reference benchmark for calculating the expected cumulative number of successful applicants.

[0035] Information reading module: In response to work order reception, reads the upstream average throughput and downstream average throughput corresponding to the action spatial location identifier, and generates a disturbance object instance containing the initial potential energy base;

[0036] In this embodiment, the information reading module receives the data packet output by the data acquisition module and is responsible for generating a disturbance object instance carrying an initial potential energy base; the core function of this module is to bind the structured fields of the manual control work order with the section passing rate in the real-time perception data, so as to form a unified data object available for subsequent deduction and extinction judgment.

[0037] After the information reading module receives the data packet transmitted by the data acquisition module, it first parses four fields contained therein: action space position identifier, restriction type code , submitter weight attribute , upstream average passing rate and downstream average passing rate ; wherein the action space position identifier is used to determine the effective coordinate of the disturbance instance in the topological connection matrix, and the restriction type code is used for setting boundary conditions in subsequent shadow threads, and are used for calculating the initial potential energy base, and are temporarily stored in the instance as a reference baseline value for the subsequent calculation of the expected cumulative number of people passing through;

[0038] During the operation of the revocation analysis module, the system simultaneously monitors the instantaneous passing rate of the downstream section compared with the historical average passing rate of the same section in the corresponding period for the degree of deviation; which is defined as the moving average of the instantaneous passing rate of the section in the same time period (with an hour granularity) in the past seven days; when and the ratio of is lower than 0.3 or higher than 2.0, it is determined that the current downstream passing rate has abnormal fluctuation, and the system marks the calculated value of at this sampling moment as "low confidence"; the marked value does not participate in the extinction threshold comparison in the current sampling period, and waits for recalculation in the next sampling period; when three consecutive sampling periods are all marked as "low confidence", the system issues an abnormal data source alarm, but the conductivity coefficient value remains unchanged.

[0039] Regarding the generation of the initial potential energy base, the information reading module performs the following calculation; let the submitter weight attribute be , the upstream average passing rate be , the duration value of the preset duration window is (in this embodiment, seconds), the mapping value of the congestion degree magnitude is recorded as ("light" corresponds to , "medium" corresponds to , "heavy" corresponds to ); then the initial potential energy base Defined as:

[0040] ;

[0041] The dimensions on the right side of the equation are as follows: This is a dimensionless coefficient, ranging from 0.6 to 0.9. The weighting coefficient reflects the difference in the credibility of information reported by personnel in different positions. Supervisors usually have more on-site experience and a higher accuracy rate in judgment, so they are given a higher weight. The value range is controlled within the range of 0.6 to 0.9 to avoid extreme values ​​(too low and the constraint will hardly last; too high and the constraint will not be eroded for a long time). The dimension is human per second. The dimension is seconds, and the product result is... The dimension of this value is human; its physical meaning is: under the weighting of job weight coefficients, the upstream section at... The total number of people expected to enter the controlled area within the specified timeframe; this baseline value, after being reverse-modulated in subsequent steps, forms the initial dynamic potential energy value, serving as the total budget for the logical erosion that this disturbance instance can withstand, used to determine when the artificial constraint automatically disappears due to changes in physical conditions; when the upstream average throughput... When it is zero, The value is zero; in this case, the system will set the initial dynamic potential energy value of the disturbed object instance to zero. Setting it directly to zero means that after this instance is written into the topological connectivity matrix, the undo analysis module will determine this in the first sampling period. If established, the travel restriction shall be immediately revoked; The value is based on the on-site inspectors' intuitive judgment of the congestion level when submitting work orders, which reflects the urgency of the controlled area. "Medium" corresponds to... This increases the potential energy base by 30% for the "lighter" version and the corresponding value for the "heavier" version. This increases the potential energy base by 60%, meaning that areas with higher congestion levels require a larger flux deviation to be cleared in subsequent elimination steps. In other words, areas with more urgent control have higher logical persistence. When the data acquisition module does not submit the congestion level magnitude... Use the default value of 1.0.

[0042] The information reading module then creates a data structure for the perturbation object instance; this instance contains five fixed attribute fields: the identifier of the action space coordinates. (Directly taken from the location identifier in the work order), restriction type code Initial potential energy base Reference time (Timestamp taken from data acquisition module record), status flag (initially assigned "pending review"); this instance is stored in the active instance pool in memory as a structured record for subsequent use by the pulse calculation module; simultaneously, the instance's... Value and The value will be read by the pulse calculation module to perform shadow thread deduction and inverse modulation, in the example. The value will be read by the topology building module to determine the specific location where the passage constraints are written into the topology connectivity matrix.

[0043] For example, suppose the data packet output by the data acquisition module contains... (Submitted by the on-site supervisor) The information reading module calculates the number of people per second. People; This value indicates that the supervisor expects approximately 65 people to enter the controlled area within a 60-second window. This value will be used as the initial dynamic potential energy value for subsequent reverse modulation. The base number input; if the work order's (Fully enclosed control) In this case, the restriction type code is also recorded in the instance, which is used by the pulse calculation module to set the reflection boundary in the shadow thread.

[0044] After the above creation is completed, the information reading module pushes the perturbation object instance to the active instance pool in shared memory and updates the status flag of the instance to "pending review". At the beginning of the next sampling period, the pulse calculation module reads the instance with the status "pending review" from the pool and performs shadow thread parallel inference.

[0045] Pulse Calculation Module: Copy the velocity field and density field of the main thread to the shadow calculation space, set boundary conditions according to the constraint type and deduce the first maximum number of people staying, while the main thread deduces the second maximum number of people staying without constraints, and calculates the absolute value of the difference between the two as the dynamic pulse value;

[0046] The pulse calculation module reads the perturbation object instance with a status of "pending review" from the active instance pool and obtains the action space coordinate identifier of the instance. Restricted type codes Initial potential energy base and reference time .

[0047] This module allocates an independent shadow computing space in background memory; this space is physically isolated from the main thread, ensuring that any data modifications within the shadow space will not pollute the real-time situational field of the main thread. The system will then store the current two-dimensional density field of the main thread. With velocity field The entire data is copied to the shadow space. The copy operation uses a deep copy method, which copies the values ​​of all grid nodes, rather than just the reference addresses, to ensure that the shadow space has a data foundation for independent evolution.

[0048] Within the shadow space, the system uses the restricted type code. Set boundary conditions; when When this occurs, it indicates that the work order is under fully closed-loop management, and the system will assign corresponding coordinates to it in the shadow space. Apply a reflective boundary condition at the edge of the grid to reduce the flow rate at that edge to zero; when When the signal is displayed, it indicates flow control, and the system halves the throughput of that grid edge. The specific implementation of the halving is as follows: the system first calculates the boundary flux under unconstrained conditions based on the density and velocity values ​​of the grid nodes on both sides of the grid edge at the current moment. Then Multiply by 0.5 to obtain the boundary flux after flow limiting. In the density field update equation, The original flux is used instead in the inflow and outflow calculations of the boundary grid node density; to ensure the physical rationality of the numerical calculations, the system... Apply lower bound cutoff: if If the calculated value is less than zero, then... Forced to zero; if If the passage capacity exceeds the maximum physical passage capacity of the grid edge (in this embodiment, 1.5 people per meter width per second), then... Truncate to this maximum value; reflection boundary ( The specific implementation method for the case is as follows: In the discrete difference scheme, the flux term of the boundary grid edge is set to zero, that is... In the density field update equation, this boundary does not produce any inflow or outflow; flow control ( The halving treatment in the case of (case ) only applies to It takes effect when the value is positive; if If the value is zero or negative, the original value remains unchanged.

[0049] The shadow space is derived using a discrete-difference scheme of a macroscopic traffic flow model; the continuous form of this model is... The system describes the conservation relationships of pedestrian flow in the field. After discretization, the system iterates for 120 steps with a time step of 1 second, covering a 120-second prediction window (120 seconds is a common time window in scenic area emergency response from issuing instructions to the start of pedestrian flow; this iteration length is sufficient to cover the initial response stage without introducing unacceptable model error accumulation due to excessive iteration time). During the iteration process, the density calculation value of each grid node is limited to the physically permissible maximum value. People / square meter (This value is based on the generally accepted crowding threshold standard in the field of scenic area emergency management; exceeding this value may trigger a stampede risk; 5 people / square meter is the physically achievable upper limit; density projections below this value are physically reliable, while densities exceeding this value have no engineering reference value in pedestrian flow models); when the calculated density value of a certain grid node exceeds... When the excess is detected, it is redistributed to the adjacent grid nodes according to the density gradient direction between the affected grid node and its neighbors, ensuring the total number of people in the entire area remains constant. Simultaneously, the system generates a saturation alarm marker for the affected grid node, recording its coordinates and the time of exceeding the limit. This marker does not participate in any numerical calculations or logical judgments; it is only appended to the attribute list of the disturbed object instance and pushed to the log module of the command screen for display. Each disturbed object instance's attribute list can record multiple saturation alarm markers, each containing two fields: coordinates and the time of exceeding the limit. The maximum number of markers is 100; if this number is exceeded, the earliest marker is overwritten in chronological order. The relationship between the velocity field and the density field follows the pre-calibrated basic graph function of this embodiment. The function was obtained by fitting measured speed-density pairwise data from the past thirty days during non-congested periods using the least squares method, and the fitting form is a Greenshields linear model. ,in The free flow velocity is taken as 1.5 m / s in this embodiment. In each time step, the system calculates the flux between adjacent grids based on the density and velocity values ​​of each grid node at the current moment, and updates the density distribution at the next moment. During the simulation, the grid edges with applied reflection boundaries always maintain a zero flux state.

[0050] After the simulation is completed, the system extracts the identifier from the shadow space using the coordinates of the action space. The maximum instantaneous number of people remaining in a circular area with a radius of 100 meters is determined by the following method: The density field within the circular area is recorded at each of 120 time steps. The density values ​​are then integrated and summed over the spatial grid covered by the circular area to obtain an estimate of the total number of people remaining in the area at that moment. The maximum value among the 120 estimates is then selected and denoted as [value missing]. Meanwhile, the main thread independently executes the same deduction within this 120-second window. Its density and velocity fields are unaffected by the shadow space boundary conditions, and it outputs the estimated maximum number of people remaining in the entire domain under purely physical conditions, denoted as... .

[0051] The system then calculates the absolute value of the difference between the two to obtain the dynamic pulsation value:

[0052] ;

[0053] Its dimension is human; the physical meaning of this value is: the predicted offset of the maximum number of people remaining in the entire area if an artificial constraint is applied; the larger the offset, the more severe the deviation between the artificial constraint and the physical inertia described by the current sensor data; because and Coming from the same discrete difference scheme and the same initial density and velocity fields, the numerical errors of the two are homogeneous within the 120-second extrapolation time domain; the operation of taking the absolute value of the difference cancels out the systematic numerical errors generated by discretization and iterative calculation, and the remaining difference is mainly caused by the reflection boundary or the boundary condition of halving the passage capacity applied in the shadow space.

[0054] For example, suppose a work order implements a complete closure of the middle section of the East Corridor Bridge; the traffic flow at the edge of that grid in the shadow space is zero, and after 120 seconds of simulation, it is found that people are continuously accumulating at the entrance of the corridor bridge, with the maximum number of people remaining in the entire area. Human; while the unconstrained deduction result of the main thread is People; then The number of people indicates that if the lockdown order is implemented, the total number of people stranded in the entire region will increase by 140 compared to the current physical inertia.

[0055] Dynamic pulsation value The density field sequence at each moment within the 120-second window used in the shadow simulation, as well as the corresponding results of the unconstrained simulation in the main thread, are all written into the instance data structure as additional attributes of the perturbation object instance. After the pulse calculation module completes the above calculations, it updates the status flag of the instance from "pending review" to "simulated" and pushes the instance to the input queue of the topology construction module, which then performs the generation of the permissible fluctuation envelope threshold and dynamic gate arbitration.

[0056] Topology construction module: Generates the permissible fluctuation envelope threshold based on the dispersion of the number of people in the entire area grid. If the dynamic fluctuation value is not greater than the threshold by a preset multiple, the initial potential energy base is back-modulated according to the relative difference between the two to generate the initial dynamic potential energy value, and the passage constraint corresponding to the restriction type is written into the main thread topology connectivity matrix.

[0057] The topology building module reads a perturbation object instance with a "deduced" status from the input queue of the impulse calculation module; this instance now carries the action space coordinate identifier. Restricted type codes Initial potential energy base Dynamic pulsation value , and the density field sequence recorded at each moment during the shadow thread deduction process.

[0058] A permissible fluctuation envelope threshold is generated based on the dispersion of the population distribution across all spatial grids in the panoramic area. This threshold characterizes the acceptable upper limit of the system's prediction offset under the current population flow distribution. Specifically, the system traverses all discrete grids in the panoramic area and counts the instantaneous population in each grid at the current moment. , forming a set ,in This represents the total number of grid cells. Calculate the global standard deviation of this set. The expression is ,in The arithmetic mean of the number of people in all grids at any given moment; when the total number of grids in the entire area... When less than 10, the standard deviation The calculation is still performed according to the formula, but middle Since the value is relatively small, the system simultaneously calculates 30% of the instantaneous number of people in each grid as a dynamic tolerance term (30% is approximately one standard deviation coverage range commonly used in engineering). The larger value in the dynamic tolerance term is used as the final constant term, replacing the original one. participate The calculation.

[0059] Permissible fluctuation envelope threshold Defined as a linear function of the global standard deviation, in the form of: ;in To preset the first coefficient, This is a preset constant term; and The value is determined using historical data during system initialization. It reflects the mapping ratio of the degree of fluctuation to the acceptable deviation threshold. This serves as the baseline tolerance for the system under stable conditions. The specific calibration method is as follows: load complete pedestrian flow records from the past thirty days, apply simulated constraints to random spatial nodes in each historical record, and iterate through... Within the range of 0.1 to 1.0, with a step size of 0.05 (this range covers the possible mapping ratio between the permissible fluctuation envelope and the global standard deviation; the lower limit of 0.1 excludes the standard deviation pair), In cases where contributions are too low, a maximum of 1.0 can ensure [the desired outcome]. Will not because Too big and cover up The actual difference is that a step size of 0.05 strikes a balance between calibration accuracy and computational complexity. Within the parameter range of 10 to 100 and with a step size of 5, calculate the sum of the false rejection rate and the false accuracy rate for each parameter combination. The false rejection rate is the percentage of simulated constraints incorrectly suspended and rejected, and the false accuracy rate is the percentage of simulated constraints incorrectly allowed to be injected. Select the parameter combination that minimizes the sum of these two rates as the optimal parameter combination. and The calibration value.

[0060] The topology building module then performs dynamic gate arbitration; the system will then use dynamic pulsation values. and Comparison: If If the system determines that the artificial constraint deviates significantly from objective physical inertia, it updates the instance's status to "suspended and rejected," pushes the record to the invalid input pool at the edge of the large screen for review by backend command personnel, and simultaneously prevents the instance from entering the topological connectivity matrix, terminating the process; if If the constraint is deemed to be within an acceptable error range, injection is permitted. This multiplier corresponds to expanding the permissible fluctuation envelope threshold by 2.5 times as the rejection decision boundary, which is equivalent to setting a confidence interval boundary of approximately 98% (under the assumption of normal distribution). The basis for this value is that in traffic flow models, short-term fluctuations in pedestrian density usually do not exceed the mean plus or minus 2 standard deviations, and the probability of events exceeding 2.5 standard deviations is less than 1%, which provides sufficient grounds for rejection.

[0061] When injection is permitted, the system performs reverse modulation on the initial potential energy base of the instance to generate an initial dynamic potential energy value; the modulation relationship is as follows:

[0062] ;

[0063] in The initial potential energy base generated for the information reading module. and The ratio represents the degree of relative deviation between the constraint and the current physical inertia; the reciprocal relationship shown in the formula determines: Compared to The larger, The smaller the value of the artificial constraint, the more it conflicts with the sensor data, and the lower the threshold required for it to be eroded in the subsequent elimination steps; the status flag of this instance is updated to "active".

[0064] The topology construction module then writes the access constraints corresponding to this instance into the topology connectivity matrix of the main thread; the topology connectivity matrix is ​​the basic data structure of the scenic area path planning engine, and each element records the access weight of the corresponding grid edge in the path search; the system determines the access weight based on the constraint type code. Determine the write method: When At that time, the coordinates will be marked. The corresponding grid edge passage right is reset to zero; when At that time, the passability weight is halved; after writing, the grid edge will no longer be included as a passable edge in the candidate path in subsequent path search algorithms; the instance continues to remain in the active instance pool, carrying Enter the execution phase of the undo analysis module.

[0065] Cancellation Analysis Module: During the effective period, the actual cumulative number of people passing through downstream and the cumulative number of people entering upstream are obtained. Based on the cumulative number of people entering upstream and the conduction coefficient, the expected cumulative number of people passing through is generated. When the absolute value of the deviation between the actual and expected cumulative number of people passing through is greater than the preset ratio of the initial dynamic potential energy value, the passage constraint is cancelled.

[0066] The undo analysis module reads a perturbation object instance with a status of "active" from the active instance pool. This instance already carries the action space coordinate identifier after the topology construction module completed the passage constraint writing. Restricted type codes Initial dynamic potential energy value Reference time And the current status flag.

[0067] This module runs continuously during the instance's active period. Within each sampling period, the system reads the actual cumulative number of people passing through from the hard sensors deployed at the nearest downstream section of the constraint space. The cumulative number of people entering is read from the hard sensors at the upstream flow section. ; and All values ​​are monotonically increasing cumulative values ​​calculated from 00:00 on the current day, with the dimension being human. The hard sensors include the turnstile and the ground piezoelectric induction strip, which are redundant. The historical average is defined as the moving average of the sensor's output increment over the past 60 sampling periods. If the count value output by any sensor shows an increment exceeding three times the historical average in an adjacent sampling period and that increment is greater than 10 people, or if the increment in an adjacent sampling period is zero while the increment in the previous period is positive and this state continues for more than 5 sampling periods, it is considered a jump. After a jump is determined, the sensor's count data is downweighted for the next 60 seconds: the turnstile channel data weight is reduced to 0.3, and the piezoelectric induction strip data weight is reduced to 0.5. The weighted sum of the two is taken as the actual cumulative number of people passing through at that sampling moment. The value of .

[0068] The system calculates the theoretically expected cumulative number of people who should pass through the downstream section under the unobstructed assumption:

[0069] ;

[0070] in The coefficient is a dimensionless transmissibility coefficient, with a value range limited to 0.75 to 0.95. This coefficient is obtained by extracting measured throughput data from the upstream and downstream sections during non-congested periods over the past 7 days, performing linear regression on paired data points, and using the regression slope value as the transmissibility coefficient. The criterion for determining non-congested periods is: the downstream instantaneous throughput rate within any consecutive 60-second window during that period. The system will suspend periods when the timeframe is below the 75th percentile of the historical data for the same period (calculated based on hourly data from the past 30 days). If, due to weather or other reasons, the total duration of periods meeting the above non-congestion criteria is less than 2 hours in the past 7 days, the system will be suspended. The automatic update uses the calculations from the previous week. The value will be updated only after subsequent data accumulation meets the minimum 2-hour duration requirement; the system will recalculate and update this coefficient weekly; if the system is newly deployed and has no historical data, The default value is 0.85; 7 days cover the complete week cycle (the passenger flow patterns are different on weekdays and weekends), while ensuring that there is a sufficient sample size for linear regression (non-congestion periods usually account for 60% to 70% of the total day).

[0071] System calculation logic erosion margin:

[0072] ;

[0073] The dimension of this value is human; The symbol reflects the actual effect of the blockade: when the blockade is effective, the downstream throughput is lower than expected. The value is negative; when the blockade is bypassed or breached, the downstream throughput is close to or even exceeds expectations. Approaching zero or being positive.

[0074] Continuous system monitoring ;like Where 0.8 is a preset ratio; the logic for determining the continuation of this constraint is no longer valid; this condition covers two scenarios: firstly, A negative value with an absolute value exceeding the limit indicates that there is no sufficient demand for passage within the controlled area, and the constraint loses its controllable object; secondly, A positive absolute value exceeding the limit indicates that the constraint has been bypassed or broken by the actual flow of people. When any scenario is satisfied, the system restores the corresponding edge weight in the main thread topology connectivity matrix to the default physical passage value, updates the status flag of the instance to "dead", and outputs a record containing the instance identifier and the death trigger scenario code to the command screen log module. In the formula, the fixed value of 0.8 indicates that when the throughput deviation exceeds 80% of the initial budget, it means that the control effect of the constraint has seriously deviated from the expectation, triggering death. This value takes into account the balance between tolerance deviation (avoiding false death caused by normal random fluctuations) and timely response (releasing topology resources when the deviation is serious).

[0075] In addition, the system maintains an independent state counter for each effective constraint, recording the continuous duration for which the constraint has not triggered any significant flux deviation since it became effective; if consistently less than for 300 seconds And the instantaneous throughput of the downstream section during this period If the coefficient of variation is less than 0.05, it is determined that the constraint has neither caused effective blocking nor been breached by the flow of people in the current period, but the physical control effect of the constraint can no longer be confirmed by the throughput data; at this time, the system will... The threshold is temporarily reduced to And continue monitoring for the next 60 seconds; if after 60 seconds... Still below If the instantaneous throughput coefficient of variation is still below 0.05, the constraint is determined to be in the "existence status unverifiable" state. The passage constraint is actively revoked and the status flag is updated to "dead". At the same time, the scenario code "03-state unverifiable automatic death" is output to the log module. This mechanism ensures that in the silent scenario where the sensor signal is continuously stable but the physical constraint no longer exists, the constraint will not be permanently suspended due to the long-term indifference of throughput data.

[0076] For example, suppose a certain constraint people, Person; 180 seconds after taking effect people, ,but People; if People, people, If the scene is destroyed; People, people, This triggers the demise of Scene 2.

[0077] After the constraint is removed, the system prioritizes upstream flow determination: if the cumulative increase in the number of people entering the upstream section within a consecutive 60-second window is less than 10% of the historical average for the same period, the system determines that there is no actual passage demand for the controlled path during the current period, directly marks the instance as "permanently eliminated," and no longer attempts to restore the passage constraint. This determination is reset once at the end of each 60-second window; only when the above upstream flow conditions are not met does the system enter the low deviation recovery determination process: monitoring Has it fallen back to In the following, 0.6 is the preset recovery ratio. If the low deviation state continues for more than a continuous time window (10 seconds in this embodiment), the system will restore the status flag to "active" and re-restore the passage constraints in the topology connectivity matrix. If the extinction is triggered again within 60 seconds after restoration, the recovery threshold of the instance will be adjusted from... Upgraded to The increased threshold remains effective in subsequent recovery assessments. The maximum increase in the recovery threshold is [missing information]. That is, not exceeding the extinction threshold. The recovery threshold is maintained at 93.75%, ensuring it remains below the extinction threshold, meaning the recovery decision may still be physically possible. If no new extinction is triggered within two consecutive full 60-second windows after the instance's constraints are removed, the system resets the recovery threshold to its initial value. .

[0078] After the topology construction module writes the passage constraints into the main thread's topology connectivity matrix, the system periodically performs spatial conflict ablation scans on active perturbation object instances. The scan period is consistent with the system sampling period, both being 1 second. When two or more perturbation object instances are detected, the system will perform ablation scans on them. When the respective restriction type identifiers are marked as mutually exclusive pairs in the preset logical mutual exclusion table, the conflict resolution logic is initiated; the preset logical mutual exclusion table contains at least mutually exclusive pairs of "fully closed control" and "forced opening", and mutually exclusive pairs of "flow restriction control" and "forced opening"; The value is equal to the side length of a single grid cell in the scenic area's discrete grid; the influence radius is set with the spatial location identifier of each disturbed object instance as the center. Construct the circle of influence. The value is 3.0 meters, which is based on the spatial range typically affected by a single on-site patrol officer implementing verbal or rope-pulling control measures; calculate the spatial overlap area between each affected circle. , The dimension is square meters; when Greater than the preset area threshold When the system determines that there is an irreconcilable instruction conflict in the spatial node, it removes both the "passable" and "inaccessible" status markers from the corresponding grid edge in the topological connectivity matrix and reconstructs it into an isolation barrier node. That is, the weight of all associated edges of this node in the topological connectivity matrix is ​​considered to be infinite, and subsequent path search algorithms cannot identify it as a path endpoint or transit point. The value is set at 5.0 square meters. This value is based on the fact that when the overlapping area of ​​two influence circles with a radius of 3.0 meters exceeds 5.0 square meters, the influence range of the two control measures has a high degree of overlap in physical space. The actual execution area of ​​the mutually exclusive instructions cannot avoid conflict through spatial division, and the conflict node must be isolated as a whole.

[0079] Based on the systematic coordination of the aforementioned modular architecture, this invention transforms manually reported information from static instructions into dynamic constraint objects with autonomous lifecycles. Specifically, the system performs physical consistency verification before constraint injection through a pulse calculation module, quantifying the degree of deviation between the constraint and the current pedestrian flow inertia, and mapping the deviation inversely to the initial value of its logical persistence potential energy—the greater the deviation, the lower the potential energy, thus embedding a physical reliability self-decay mechanism at the logical level. Through spatial overlap detection and isolation barrier node reconstruction in the topology construction module, the oscillation of multi-point mutual exclusion instructions in the signal domain is transformed into static isolation in the topology domain, eliminating control signals from the underlying data structure of the path planning engine. The high-frequency flipping; through the continuous monitoring of the cumulative flux conservation difference between upstream and downstream sections by the cancellation analysis module, the elimination of constraints no longer depends on secondary manual confirmation, but is driven by the actual throughput of objective pedestrian flow—when the blockade is effectively excessive and there is no upstream flow, the constraint disappears because it loses its control object; when the blockade is bypassed or broken by pedestrian flow, the constraint disappears because its physical premise is overturned; the above mechanisms together constitute a two-way game loop that places human experience and physical conservation on an equal footing for verification, so that the path deduction conclusion of the twin system can be self-verified and self-corrected by objective flux data at any time, thereby converging to a state consistent with the actual physical space without increasing the cost of on-site manual verification.

[0080] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0081] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0082] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A digital twin-enabled scenic area full-scene visualization management and emergency command system, characterized in that: include: Data acquisition module: Acquires manual control work orders and real-time sensing data streams. The work order includes the spatial location identifier, restriction type, and submitter weight. The data stream includes the instantaneous throughput, velocity field, and density field of each node. Information reading module: In response to work order reception, reads the upstream average throughput and downstream average throughput corresponding to the spatial location identifier, and generates a disturbance object instance containing the initial potential energy base; Pulse Calculation Module: Copy the velocity field and density field of the main thread to the shadow calculation space, set boundary conditions according to the constraint type and deduce the first maximum number of people staying, while the main thread deduces the second maximum number of people staying without constraints, and calculates the absolute value of the difference between the two as the dynamic pulse value; Topology construction module: Generates the permissible fluctuation envelope threshold based on the dispersion of the number of people in the entire area grid. If the dynamic fluctuation value is not greater than the threshold by a preset multiple, the initial potential energy base is back-modulated according to the relative difference between the two to generate the initial dynamic potential energy value, and the passage constraint corresponding to the restriction type is written into the main thread topology connectivity matrix. Cancellation Analysis Module: During the effective period, the actual cumulative number of people passing through downstream and the cumulative number of people entering upstream are obtained. Based on the cumulative number of people entering upstream and the conduction coefficient, the expected cumulative number of people passing through is generated. When the absolute value of the deviation between the actual and expected cumulative number of people passing through is greater than the preset ratio of the initial dynamic potential energy value, the passage constraint is cancelled.

2. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, Manually managed work orders are submitted using a predefined combination format, which includes at least the grid coordinates corresponding to the action space location identifier, the event type identifier, and the congestion level identifier; the submitter's weight attribute is a job grade coefficient pre-stored in the system database.

3. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, The initial potential energy base is generated as follows: the product of the submitter weight attribute and the upstream average pass rate is obtained, the product is multiplied by the duration value of a preset duration window, and the result is used as the initial potential energy base.

4. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, The velocity and density fields of the main thread are copied in the shadow computing space. Boundary conditions are set according to the constraint type identifier, and deduction is performed to obtain the first maximum number of people remaining. Specifically, this includes: An independent shadow computing space is created in the background memory, and the two-dimensional density field and velocity field of the main thread at the current moment are completely copied to the shadow computing space. When the restriction type is identified as a closed class, a reflection boundary is set at the corresponding grid edge in the shadow computing space to make the traffic flow at the grid edge zero. The discrete difference format of the macro traffic flow model is used to perform fixed-duration pedestrian flow simulation. The integral sum of the density field of the whole region with respect to the spatial grid is calculated in each simulation time step, and the maximum value among all time steps is taken as the first maximum number of stranded people.

5. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, The permissible fluctuation envelope threshold is generated based on the dispersion of the population distribution in each spatial grid of the entire area, specifically including: The number of people in each spatial grid is counted separately, and the global standard deviation of the number of people in all grids of the entire area is calculated. The global standard deviation is multiplied by a preset first coefficient and then added to a preset constant term. The result is used as the permissible fluctuation envelope threshold.

6. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, Based on the relative difference between the dynamic pulsation value and the permissible fluctuation envelope threshold, the initial potential energy base is inversely modulated to generate an initial dynamic potential energy value, specifically including: Calculate the ratio of the dynamic pulsation value to the permissible fluctuation envelope threshold, add the ratio to a preset constant and take the reciprocal, multiply the initial potential energy base by the reciprocal, and use the result as the initial dynamic potential energy value. The larger the ratio of the dynamic pulsation value to the permissible fluctuation envelope threshold, the smaller the initial dynamic potential energy value.

7. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, After writing the passage constraints corresponding to the restriction type identifier into the topological connectivity matrix of the main thread, the process also includes a spatial conflict resolution step: Periodically scan all currently active perturbation object instances; when the physical distance between the action spatial location identifiers of two or more perturbation object instances is less than a preset grid edge length threshold, and their respective restriction type identifiers are marked as mutually exclusive pairs in a preset logical mutual exclusion table, construct influence circles with preset influence radii centered on the action spatial location identifiers of each perturbation object instance, and calculate the spatial overlap area between each influence circle; when the spatial overlap area is greater than a preset area threshold, simultaneously remove both passage status identifiers from the corresponding grid edge in the topological connectivity matrix, and reconstruct it into an isolation barrier node.

8. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, The conductance coefficient is generated in the following way: extract the measured throughput data of the upstream section and the downstream section during non-congested periods within a preset historical period, perform linear regression on the measured throughput data, and use the regression slope value as the conductance coefficient; If the system is in the initial operation phase without historical data, the conductivity coefficient is set to the preset default value.

9. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, When the absolute value of the deviation is greater than a preset proportion of the initial dynamic potential energy value, after the passage constraint is revoked, the process also includes an anti-oscillation step for recovery after the deviation disappears. After the passage constraint is revoked, if the absolute value of the deviation falls below the preset recovery ratio of the initial dynamic potential energy value, and the low deviation state is maintained for more than a preset continuous time window, then the perturbation object instance is re-marked as an active state and the passage constraint is restored; wherein the preset recovery ratio is less than the preset ratio.

10. The digital twin-enabled scenic area full-scene visualization management and emergency command system according to claim 1, characterized in that, The determination that the absolute value of the deviation is greater than a preset proportion of the initial dynamic potential energy value includes two scenarios: In the first scenario, when the actual cumulative number of people passing through is less than the expected cumulative number of people passing through and the absolute value of the difference between the two is greater than the preset ratio, it is determined that the controlled object under the current constraint no longer has the need for passage, and the existence logic of the passage constraint no longer holds. In the second scenario, when the actual cumulative number of people passing through is greater than the expected cumulative number of people passing through and the absolute value of the difference between the two is greater than the preset ratio, it is determined that the passage constraint has been breached by the actual flow of people, and the physical premise of the passage constraint has been overturned; when either of the two scenarios is satisfied, the passage constraint is revoked.

Images