An airport equipment operation safety intelligent monitoring and analyzing system

Abstract

The application relates to an airport equipment operation safety intelligent monitoring and analyzing system, in particular to the field of intelligent monitoring and data analysis, which realizes digital mapping of an operation state by fusing real-time airport multi-source operation data into a dynamic attribute network model; the network flow and constraint propagation algorithm is used to dynamically identify key bottleneck nodes and paths which restrict overall efficiency, and the comprehensive resilience index reflecting the overall pressure bearing and recovery capacity of the airport is calculated by focusing on the bottleneck points; when the index shows a warning, the scheme can simulate and deduce preset intervention measures, and quantitatively evaluate the effect of the strategy on bottleneck elimination and resilience improvement; thus, operation management is changed from passive monitoring of isolated states to active insight into networked correlations, and is upgraded from emergency disposal relying on experience to predictive decision-making based on simulation deduction, so that the overall robustness, anti-disturbance capacity and scientificity of decision-making of airport operation are finally improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring and data analysis, and more specifically, to an intelligent monitoring and analysis system for the safe operation of airport equipment. Background Technology

[0002] As a comprehensive transportation hub operating continuously around the clock, the operational efficiency and safety level of a modern large airport highly depend on the deep collaboration and dynamic balance among multiple physical and functional subsystems, such as the airside flight area, terminal passenger service area, and landside transportation area. These subsystems are tightly coupled through passenger flow, baggage flow, aircraft flow, and information flow, forming a typical complex mega-system. In daily operations, the system needs to cope with the constant pressure of densely overlapping flight schedules and fluctuating passenger flow. Under special circumstances, such as encountering extreme weather, major aviation incidents, sudden failures of key equipment clusters, or sudden large passenger flows, local disturbances can be rapidly transmitted and amplified in the network through existing process dependencies, easily triggering chain reactions. This can cause a bottleneck in a single link to evolve into widespread operational congestion or even partial paralysis, posing a serious threat to flight punctuality, passenger experience, and operational safety.

[0003] Currently, airport operations control centers primarily rely on large monitoring screens integrating multiple independent business subsystems for situational awareness. Their technical implementation focuses on real-time monitoring and alarming of the operational status of each independent functional module, such as flight status in the departure system, security checkpoint opening / closing status, and baggage system rotation speed. However, existing technologies have significant limitations: First, they lack the ability to quantitatively model and simulate the dynamic impact relationships between subsystems, failing to reveal the strength and path of chain reactions such as how a surge in check-in queues subsequently affects security check pressure, or how a decrease in security pass rates delays passenger boarding and ultimately leads to longer aircraft ground waiting times. Second, existing systems struggle to extract characteristics representing the overall system throughput from massive amounts of discrete state information. While macro-level resilience indicators of efficiency and robustness exist, managers cannot quantify and assess the "safety margin" of the system from overload or failure in real time. More critically, existing technologies exhibit significant lag in identifying system bottlenecks, typically only responding passively after physical queuing occurs or equipment alarms are triggered, failing to predictively identify and proactively intervene before the bottleneck effect fully manifests. This results in a lack of precise guidance in the allocation of emergency resources when dealing with operational disturbances, with decisions relying on experience-based judgments and hindering the implementation of optimization decisions based on full-system simulation. Therefore, there is an urgent need for an intelligent monitoring and analysis technology that can deeply integrate multi-source real-time data, construct lightweight system-level analysis models, and achieve proactive identification of key bottleneck nodes and dynamic assessment of operational resilience. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing an intelligent monitoring and analysis system for airport equipment operation safety. The system utilizes a data fusion and modeling module, a bottleneck dynamic identification module, a resilience index calculation module, and a strategy deduction and evaluation module to solve the problems mentioned in the background.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: Specifically, it includes: a data fusion and modeling module, a bottleneck dynamic identification module, a resilience index calculation module, and a strategy deduction and evaluation module connected in sequence;

[0006] Data fusion and modeling module: Real-time access to airport operation data, including flight schedules, passenger flow, equipment status and location information. Based on the preset airport topology, the module abstracts airport entities into network nodes and the processes between entities into edges, constructing and continuously updating a dynamic attributed network model. The attribute values ​​of network nodes and edges are dynamically assigned based on the real-time access data.

[0007] Bottleneck dynamic identification module: running on a dynamic attributed network model, it applies network flow analysis algorithms to calculate the current network's flow capacity, and through constraint propagation and marginal contribution analysis algorithms, identifies the critical path that restricts flow capacity and a bottleneck node located on that critical path, generating and outputting a list of bottleneck nodes containing bottleneck node identifiers and their bottleneck strength ratings.

[0008] Resilience Index Calculation Module: Receives a list of bottleneck nodes and extracts the real-time operating parameters of each bottleneck node. The real-time operating parameters include load rate and queuing change trends. Based on the preset multi-indicator fusion rules, the module performs weighted and composite calculations on the real-time operating parameters to generate a comprehensive resilience index that characterizes the overall operational robustness of the airport.

[0009] Strategy simulation and evaluation module: Triggered when the comprehensive resilience index is lower than the preset threshold, based on the dynamic attributed network model and the comprehensive resilience index, the preset intervention strategy is imported, the embedded simulation engine is called to simulate the network state evolution after the implementation of the intervention strategy, the comprehensive resilience index after simulation is recalculated, and the changes in the comprehensive resilience index and the distribution changes of bottleneck nodes before and after simulation are compared, and an evaluation report is output.

[0010] In a preferred embodiment, the specific process of constructing and continuously updating the dynamic attributed network model in the data fusion and modeling module is as follows:

[0011] First, based on a pre-defined airport topology knowledge base, the airport is defined as a set of network nodes, including check-in counters, security checkpoints, boarding gates, baggage claim carousels, aircraft stands, and runway entrances. The physical relationship between entities is defined as a set of directed edges connecting network nodes, including edges representing passenger flow from check-in counter network nodes to security checkpoint network nodes, edges representing security checkpoint network nodes to boarding gate network nodes, and edges representing aircraft flow from boarding gate network nodes to runway entrance network nodes.

[0012] For each of the network nodes, a dynamic attribute tuple is calculated, which includes static theoretical processing capacity, dynamic processing rate, current queue length, and binary service status.

[0013] The dynamic processing rate is obtained by multiplying the average rate obtained by dividing the actual number of processes flowing through network nodes within a preset time window by the window length by a real-time performance coefficient that is determined based on the real-time device status and has a value between zero and one.

[0014] Meanwhile, a dynamic edge weight is calculated for each directed edge, which integrates the baseline travel time and the real-time congestion delay. The process of determining the dynamic edge weight is as follows: first, add an adjustment coefficient to the baseline time of movement between the network nodes at both ends of the connecting edge, then multiply it by the current average additional delay calculated based on real-time location information, and add the two to obtain the final value.

[0015] Ultimately, the complete structure containing the set of network nodes, the set of directed edges, the dynamic attribute tuples of all network nodes, and the dynamic edge weights of all edges will be used as the dynamic attributed network model.

[0016] In a preferred embodiment, the mechanism for continuously updating the dynamic attributed network model is as follows:

[0017] Set up an update trigger that is triggered at fixed time intervals or based on events that cause changes in critical operational data;

[0018] When the update trigger is triggered, based on the latest accessed airport operation data, the dynamic processing rate, current queue length, binary service status, and dynamic edge weight of each directed edge in the dynamic attributed network model are recalculated, thereby refreshing the dynamic attribute tuples and dynamic edge weights in the dynamic attributed network model, and outputting the refreshed dynamic attributed network model.

[0019] In a preferred embodiment, the specific process of identifying critical paths and bottleneck nodes on the dynamic attributed network model in the bottleneck dynamic identification module is as follows:

[0020] First, based on the dynamic attribute tuple of each network node in the dynamic attributed network model, the current effective throughput capacity of each network node is calculated. The effective throughput capacity is the dynamic processing rate of the network node plus a buffer component. This buffer component is obtained by dividing the positive value of the static theoretical processing capacity of the network node by the dynamic processing rate by one plus the current queue length of the network node.

[0021] Next, feasible flow paths from the source to the sink are searched in the dynamic attributed network model, and the contribution and resistance of each feasible flow path to the current network's flow capacity are evaluated. The contribution of a feasible flow path is quantified by the proportion of the flow that the feasible flow path can carry in the dynamic attributed network model to the current network's flow capacity.

[0022] The resistance to a feasible path is quantified by calculating the sum of the ratios of the dynamic edge weights of each edge on the feasible path to the average dynamic edge weight of all edges, and then taking the reciprocal. Feasible paths with high contribution and relatively low resistance are marked as critical paths.

[0023] Then, for each network node in the dynamic attributed network model, a comprehensive evaluation is performed based on the following three indicators: The first indicator is the network node's own saturation, which is the dynamic processing rate of the network node divided by the static theoretical processing capacity of the network node.

[0024] The second metric is the structural criticality of network nodes, which is the sum of the contributions of each feasible circulation path marked as a critical path to the network node.

[0025] The third indicator is the impact of network node failure, which is calculated by subtracting the simulated current network throughput capacity from the calculated current network throughput capacity. The simulated current network throughput capacity is the value recalculated after temporarily invalidating the network node in the dynamic attributed network model.

[0026] Finally, network nodes are filtered and sorted based on a bottleneck strength index, which is calculated based on the first, second and third indicators, generating a bottleneck node list that includes the bottleneck node identifier, bottleneck strength index and information of the critical path to which it belongs.

[0027] In a preferred embodiment, the bottleneck strength index is calculated as follows:

[0028] First, the hyperbolic tangent function is applied to the first indicator, namely the saturation of the network nodes themselves, for mathematical transformation.

[0029] The second indicator, namely the structural criticality of network nodes, is normalized by dividing it by the maximum value of the second indicator among all network nodes.

[0030] For the third indicator, namely the impact of network node failure, its calculated value is used directly;

[0031] Then, the first, second, and third indicators that have undergone the above processing are multiplied by their corresponding first, second, and third preset weights, respectively. The three products are then added together, and the sum is the bottleneck strength index.

[0032] In a preferred embodiment, the specific process for generating the comprehensive toughness index in the toughness index calculation module is as follows:

[0033] First, for each bottleneck node in the bottleneck node list, calculate a node vulnerability score, which is obtained by adding the first and second items;

[0034] The first term is the quotient obtained by dividing the dynamic processing rate of the bottleneck node by its static theoretical processing capacity.

[0035] The second item is a preset trend weight coefficient, multiplied by a transformation value, which is the result of applying the hyperbolic tangent function to a ratio. The ratio is the difference between the current queue length of the bottleneck node at the end of a preset time window and the queue length at the beginning of the time window, and then divided by a preset trend sensitivity parameter.

[0036] Next, a dynamic influence weight is assigned to each bottleneck node in the bottleneck node list. The calculation process for this dynamic influence weight is as follows: multiply the bottleneck strength index of the bottleneck node by a metric representing the size of the bottleneck node's influence range in the current network state to obtain an intermediate product; then divide the intermediate product by the sum of the intermediate products of all bottleneck nodes in the bottleneck node list, and the quotient is the dynamic influence weight of the bottleneck node; the sum of the dynamic influence weights of all bottleneck nodes calculated according to this rule is one; synthesize the comprehensive resilience index. The calculation process for the comprehensive resilience index is as follows: first, multiply the node vulnerability score of each bottleneck node by its dynamic influence weight to obtain the weighted vulnerability of the bottleneck node;

[0037] Then, the weighted fragility of all bottleneck nodes is summed to obtain a aggregate fragility; the aggregate fragility is multiplied by a positive constant, which is used to adjust the strength of the influence of the aggregate fragility on the fragility index decay, i.e., the system fragility decay coefficient, and then the negative exponential function value of the product is taken as the base of the natural constant; finally, the obtained negative exponential function value is multiplied by one hundred to obtain the final comprehensive fragility index, which ranges from zero to one hundred.

[0038] In a preferred embodiment, the method for determining the metric representing the size of the bottleneck node's influence range during the calculation of the dynamic influence weight is as follows:

[0039] In the dynamic attributed network model, the bottleneck node is simulated to be in an invalid state. Then, the number of downstream network nodes directly affected when the traffic originally planned to flow through the bottleneck node needs to be redistributed to other paths is identified and counted. This number is used as a metric.

[0040] In a preferred embodiment, the specific process of simulating the network state evolution after the implementation of the intervention strategy in the strategy deduction and evaluation module is as follows:

[0041] The imported preset intervention strategy is parsed into a series of explicit modification instructions. These modification instructions directly affect the elements in the dynamic attributed network model. The types of modification instructions include, but are not limited to, modifying the binary service status of network nodes and adjusting the dynamic processing rate of network nodes.

[0042] Create a copy of the dynamically attributed network model and apply the modification instructions to the copy to generate the initial inference network model; call the embedded simulation engine and, based on the principle of discrete event simulation, drive the initial inference network model to evolve its state within a preset future inference time window.

[0043] Based on real-time and predicted flight and passenger flow data, the simulation engine generates and schedules simulated passenger, baggage and aircraft entities within the initial simulation network model. These entities move between network nodes according to the network topology.

[0044] The service processing time of an entity at different network nodes is determined based on the stochastic service model built into the simulation engine. This stochastic service model makes the service time follow a specified probability distribution, the expected value of which is equal to the reciprocal of the current dynamic processing rate of the network node.

[0045] The transmission time of an entity on a directed edge is determined by the current dynamic edge weight on that directed edge, plus an additional random fluctuation value.

[0046] During the simulation, the current queue length and dynamic processing rate of network nodes are dynamically updated based on the arrival, service, and departure events of entities;

[0047] Finally, the simulation engine outputs the final network state at the end of the simulation time window, which is the simulated dynamic attributed network model.

[0048] In a preferred embodiment, the specific process of recalculating the simulated comprehensive resilience index and outputting the evaluation report is as follows:

[0049] First, based on the deduced dynamic attributed network model, the bottleneck dynamic identification module is executed to identify the bottleneck nodes and bottleneck strength index after recalculation. Then, based on the deduced dynamic attributed network model and the deduced bottleneck node list, the resilience index calculation module is executed to calculate the comprehensive resilience index after recalculation. A comparative analysis is then performed to calculate the difference between the comprehensive resilience index after recalculation and the comprehensive resilience index used when the strategy deduction and evaluation module is triggered, which is taken as the degree of resilience improvement.

[0050] The number of bottleneck nodes in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered is calculated, and the number of the intersection of the number of bottleneck nodes in the bottleneck node list after simulation is calculated. That is, the number of bottleneck nodes that appear in both lists at the same time. Then, the number of the intersection is subtracted from one and divided by the number of bottleneck nodes in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered. The result is used as the bottleneck mitigation rate.

[0051] Identify and list the new bottleneck nodes that appear in the bottleneck node list after the simulation, and that are not in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered, as new bottleneck risk nodes;

[0052] Finally, an evaluation report is generated, which includes resilience improvement, bottleneck mitigation rate, a list of new bottleneck risk nodes, and a comparison of the changes in at least one of the following before and after the simulation: network throughput, dynamic processing rate of network nodes, current queue length of network nodes, and dynamic edge weight of directed edges.

[0053] In a preferred embodiment, the simulation engine uses a modeling method that introduces random perturbations to simulate the service processing time of network nodes and the transmission time of entities.

[0054] The service time of a network node in processing an entity follows a specified probability distribution, which is an exponential distribution.

[0055] The random fluctuation value added to the transmission time of an entity on a directed edge is obtained by randomly sampling from a normal distribution with a mean of zero.

[0056] The beneficial effects of this invention are as follows: by integrating multi-source airport operational data into a dynamic attributed network model in real time, a digital mapping of operational status is achieved; it uses network flow and constraint propagation algorithms to dynamically identify key bottleneck nodes and paths that restrict overall efficiency, and focuses on these bottleneck points to calculate a comprehensive resilience index reflecting the airport's overall stress resistance and recovery capabilities; when the index indicates an alarm, the solution can simulate and extrapolate preset intervention measures to quantitatively evaluate the effectiveness of the strategy in eliminating bottlenecks and improving resilience; thus, operational management is transformed from passive monitoring of isolated states to proactive insight into networked connections, and from experience-based emergency response to predictive decision-making based on simulation and extrapolation, ultimately improving the overall robustness, anti-disturbance capability, and scientific nature of airport operations. Attached Figure Description

[0057] Figure 1 This is a flowchart of the method of the present invention;

[0058] Figure 2 This is a block diagram of the system structure of the present invention. Detailed Implementation

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0061] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0062] Example 1

[0063] This embodiment provides, for example Figure 1-2 The airport equipment operation safety intelligent monitoring and analysis system shown includes: a data fusion and modeling module, a bottleneck dynamic identification module, a resilience index calculation module, and a strategy deduction and evaluation module connected in sequence.

[0064] Data fusion and modeling module: Real-time access to airport operation data, including at least flight schedules, passenger flow, equipment status and location information. Based on the preset airport topology, the module abstracts airport entities into network nodes and the processes between entities into edges, constructing and continuously updating a dynamic attributed network model. The attribute values ​​of network nodes and edges are dynamically assigned based on the real-time access data.

[0065] Bottleneck dynamic identification module: running on a dynamic attributed network model, applying network flow analysis algorithms to calculate the current network's flow capacity, and using constraint propagation and marginal contribution analysis algorithms to identify the critical path that restricts flow capacity and at least one bottleneck node located on that critical path, generating and outputting a list of bottleneck nodes containing bottleneck node identifiers and their bottleneck strength ratings.

[0066] Resilience Index Calculation Module: Receives a list of bottleneck nodes and extracts the real-time operating parameters of each bottleneck node. The real-time operating parameters include at least the load rate and queuing change trend. Based on the preset multi-indicator fusion rules, the module performs weighted and synthetic calculations on the real-time operating parameters to generate a comprehensive resilience index that characterizes the overall operational robustness of the airport.

[0067] Strategy simulation and evaluation module: Triggered when the comprehensive resilience index is lower than the preset threshold or when an intervention command is received, based on the dynamic attributed network model and the comprehensive resilience index, the module imports the preset or user-defined intervention strategy, calls the embedded simulation engine to simulate the network state evolution after the implementation of the intervention strategy, recalculates the comprehensive resilience index after simulation, compares the changes in the comprehensive resilience index before and after simulation with the changes in the distribution of bottleneck nodes, and outputs an evaluation report.

[0068] In this embodiment, the specific process of constructing and continuously updating the dynamic attributed network model in the data fusion and modeling module is as follows:

[0069] First, based on a pre-defined airport topology knowledge base, airport physical and logical entities are defined as sets of network nodes. These sets include at least check-in counters, security checkpoints, boarding gates, baggage claim carousels, aircraft stands, and runway entrances. The physical or logical flow between entities is defined as a set of directed edges connecting network nodes. This set includes at least edges representing passenger flow from check-in counter network nodes to security checkpoint network nodes, edges representing security checkpoint network nodes to boarding gate network nodes, and edges representing aircraft flow from boarding gate network nodes to runway entrance network nodes. The airport topology knowledge base is stored in graph structure data, defining all network node types, spatial relationships between network nodes, and allowed flow directions, serving as the static framework foundation for constructing the network model.

[0070] Subsequently, a dynamic attribute tuple is calculated for each network node. This dynamic attribute tuple contains the network node's static theoretical processing capacity, dynamic processing rate, current queue length, and binary service status. The static theoretical processing capacity is the maximum service capacity of the network node under ideal conditions, such as the maximum number of people passing through a security checkpoint per hour. The binary service status is used to indicate whether the network node is available or unavailable. When the corresponding device of the network node fails or is scheduled to be shut down, this status is marked as unavailable.

[0071] The dynamic processing rate determination process is as follows: First, the actual number of processes flowing through the network node within a preset time window is obtained. The actual number of processes is divided by the time window length to obtain the average processing rate. Then, it is multiplied by a real-time performance coefficient determined based on real-time device status data, with a value ranging from zero to one. The preset time window length can be set according to the actual data refresh frequency and accuracy requirements, such as 60 seconds or 300 seconds. The real-time performance coefficient is obtained by querying the device health status and performance mapping table. This mapping table maps the device status (such as normal, degraded, fault) to specific coefficient values ​​(such as 1.0, 0.5, 0.1), thereby quantifying the device physical status as an impact factor on the processing capacity of the network node.

[0072] Simultaneously, a dynamic edge weight is calculated for each directed edge. This dynamic edge weight is a time-varying value that integrates the baseline travel time and real-time congestion delay. The dynamic edge weight determination process is as follows: first, the baseline travel time between the network nodes at both ends of the connecting edge is added to an adjustment coefficient, and then multiplied by the current average additional delay calculated based on real-time location information. The two are then added together to obtain the final value. The baseline travel time is the normal travel time estimated based on historical data or physical distance. The adjustment coefficient is used to adjust the degree of influence of real-time delay on the weight, and is usually set to 1. The current average additional delay is calculated based on location data (such as Wi-Fi probes, Bluetooth beacons, vehicle GPS trajectories), reflecting the additional time consumed due to congestion, scheduling, and other reasons.

[0073] Finally, the complete structure containing the set of network nodes, the set of directed edges, the dynamic attribute tuples of all network nodes, and the dynamic edge weights of all edges is used as the dynamic attributed network model. This dynamic attributed network model is maintained in memory in the form of a graph data structure, in which network node objects are associated with their dynamic attribute tuples, and edge objects are associated with their dynamic edge weights, forming a complete network representation whose attributes can be dynamically refreshed.

[0074] The mechanism for continuously updating the dynamic attributed network model is as follows:

[0075] Set an update trigger, which is triggered at a fixed time period or based on events that change key operational data. The fixed time period can be set according to the system's real-time requirements, such as 1 minute. Events that change key operational data include, but are not limited to: flight schedule status updates (such as takeoff, landing, and delays), critical equipment status alarms (such as baggage sorting machine stoppages), or sudden changes in passenger flow density in specific areas.

[0076] When the update trigger is activated, based on the latest airport operation data, the dynamic processing rate, current queue length, binary service status, and dynamic edge weight of each directed edge in the dynamic attributed network model are recalculated. This refreshes the dynamic attribute tuples and dynamic edge weights in the dynamic attributed network model, and outputs the refreshed dynamic attributed network model. The refresh is an atomic operation, ensuring that at any given time, the downstream module obtains a consistent and complete snapshot of the network model, thus avoiding logical errors caused by asynchronous data updates.

[0077] In this embodiment, the specific process of identifying critical paths and bottleneck nodes on the dynamic attributed network model in the bottleneck dynamic identification module is as follows:

[0078] First, based on the dynamic attribute tuples of each network node in the dynamic attributed network model, the current effective throughput capacity of each network node is calculated. The calculation process of effective throughput capacity is as follows: the dynamic processing rate of the network node is added to a buffer component. This buffer component is obtained by subtracting the dynamic processing rate from the static theoretical processing capacity of the network node to get a positive result. This positive result is then divided by one and the current queue length of the network node is added. The current effective throughput capacity comprehensively reflects the actual throughput limit of the network node under the instantaneous processing capacity and potential buffer capacity. The buffer component is suppressed by the queue length, which reflects the delay effect of the queue on the network node's recovery of idle processing capacity.

[0079] Next, feasible flow paths from the source to the sink are searched in the dynamic attributed network model, and the contribution and resistance of each feasible flow path to the current network's flow capacity are evaluated. The contribution of a feasible flow path is quantified by the proportion of the flow that the feasible flow path can carry in the dynamic attributed network model to the current network's flow capacity; the higher the contribution, the more important the path is to maintaining the overall network throughput.

[0080] The resistance to a feasible flow path is quantified by calculating the sum of the ratios of the dynamic edge weights of each edge on the feasible flow path to the average dynamic edge weight of all edges, and then taking the reciprocal. The larger the resistance value, the lower the expected time or cost of flowing through the path. Feasible flow paths with high contribution and relatively low resistance are marked as critical paths. Specifically, a contribution threshold (e.g., 0.05) and a resistance threshold can be set. A path is marked as a critical path only when its contribution is higher than the contribution threshold and its resistance is higher than the resistance threshold.

[0081] Then, for each network node in the dynamic attributed network model, a comprehensive evaluation is performed based on the following three indicators: The first indicator is the network node's own saturation, which is the dynamic processing rate of the network node divided by the static theoretical processing capacity of the network node; it directly reflects the real-time utilization rate of the network node's resources.

[0082] The second indicator is the structural criticality of a network node, which is the sum of the contributions of each feasible circulation path marked as a critical path to the network node; it reflects the structural importance of the network node in the critical circulation link.

[0083] The third indicator is the impact of network node failure, which is calculated by subtracting the throughput capacity of a simulated current network from the throughput capacity of the calculated current network. The throughput capacity of the simulated current network is the value recalculated after temporarily invalidating the network node in the dynamic attributed network model. It is used to quantitatively assess the potential impact of the network node failure on the overall system performance.

[0084] Finally, network nodes are filtered and sorted based on a bottleneck strength index. The bottleneck strength index is calculated based on the first, second, and third indicators, generating a bottleneck node list that includes the bottleneck node identifier, bottleneck strength index, and information about the critical path to which it belongs. Filtering can be done by setting a bottleneck strength index threshold (e.g., 0.6), and only network nodes with an index higher than the threshold are included in the final list as bottleneck nodes, and are sorted in descending order of index value to highlight the most critical bottleneck nodes.

[0085] The calculation process for the bottleneck strength index is as follows:

[0086] First, the hyperbolic tangent function is applied to the first indicator, namely the saturation of the network node itself, for mathematical transformation. The hyperbolic tangent function is used to convert the linear growth of saturation into a smooth S-shaped curve growth. Its output range is between 0 and 1, which can effectively distinguish the subtle differences of network nodes in the high saturation range (e.g., above 80%), while avoiding the problem of the indicator value increasing infinitely when the saturation is close to 100%.

[0087] The second indicator, namely the structural criticality of network nodes, is normalized by dividing it by the maximum value of the second indicator among all network nodes. The normalization process maps the structural criticality of all network nodes to the range of 0 to 1, making the structural importance of different network nodes comparable.

[0088] For the third indicator, namely the impact of network node failure, we directly use its calculated value; this value is already a normalized proportion that directly reflects the severity of performance degradation.

[0089] Then, the first, second, and third indicators processed as described above are multiplied by their corresponding first, second, and third preset weights, respectively. The three products are then added together, and the sum is the bottleneck strength index. The first, second, and third preset weights are used to adjust the relative weights of the three factors—the state of the network node itself, the importance of the network topology, and the global impact of the fault—in the final evaluation. Their values ​​can be set according to the actual management strategy. For example, a set of available weights can be 0.4, 0.3, and 0.3, with the sum of the weights being 1.

[0090] In this embodiment, it is necessary to specifically explain the process of generating the comprehensive toughness index in the toughness index calculation module as follows:

[0091] First, for each bottleneck node in the bottleneck node list, a node vulnerability score is calculated. The node vulnerability score is obtained by adding the first and second items. This node vulnerability score aims to comprehensively reflect the current immediate pressure status and recent deterioration trend of the bottleneck node. The higher the score, the more vulnerable the node is.

[0092] The first term is the quotient obtained by dividing the dynamic processing rate of the bottleneck node by its static theoretical processing capacity; this term directly describes the real-time proportion of the node's processing capacity being occupied.

[0093] The second term is a preset trend weighting coefficient, multiplied by a transformation value. The transformation value is the result of applying the hyperbolic tangent function to a ratio, which is the difference between the current queue length at the end of a preset time window and the queue length at the beginning of that time window, divided by a preset trend sensitivity parameter. The trend weighting coefficient is used to balance the importance of instantaneous load and changing trends in vulnerability assessment, and can be set to 0.2 for example. The trend sensitivity parameter is used to scale the queue change to fit the input range of the hyperbolic tangent function, and can be set to 10 people / minute for example. Applying the hyperbolic tangent function can compress and smoothly map the unbounded trend ratio to the (-1,1) interval, which can sensitively reflect the trend direction (increase or decrease) and avoid extreme changes from having too much impact on the score.

[0094] Next, a dynamic influence weight is assigned to each bottleneck node in the bottleneck node list. The calculation process of the dynamic influence weight is as follows: the bottleneck strength index of the bottleneck node is multiplied by a metric representing the size of the influence range of the bottleneck node in the current network state to obtain an intermediate product; the intermediate product combines the individual bottleneck strength of the node with the breadth of its influence in the network.

[0095] The intermediate product of each bottleneck node is then divided by the sum of the intermediate products of all bottleneck nodes in the bottleneck node list. The resulting quotient is the dynamic influence weight of that bottleneck node. The sum of the dynamic influence weights of all bottleneck nodes calculated according to this rule is one. This normalization operation ensures that the influence weights of different bottleneck nodes on the final resilience index constitute a probability distribution, making the aggregation calculation have clear mathematical meaning.

[0096] Finally, a comprehensive resilience index is synthesized. The calculation process of the comprehensive resilience index is as follows: First, the node vulnerability score of each bottleneck node is multiplied by its dynamic influence weight to obtain the weighted vulnerability of the bottleneck node; this realizes the weighting of vulnerability based on the importance of the node.

[0097] Then, the weighted fragility values ​​of all bottleneck nodes are summed to obtain a aggregate fragility value. This aggregate fragility value represents the overall system fragility level caused by all critical bottleneck nodes. The aggregate fragility value is then multiplied by a positive constant (greater than zero) used to adjust the strength of the aggregate fragility value's influence on the resilience index decay, i.e., the system resilience decay coefficient. Finally, the product is taken as a negative exponential function with a natural constant base. The system resilience decay coefficient determines the rate at which the resilience index decays with increasing aggregate fragility; its value can be set through historical data fitting or expert experience, for example, 0.1. The negative exponential function with a natural constant base is used. It is modeled based on the typical characteristic that system resilience decreases exponentially with increasing pressure. The function maps the input to a value in the interval (0,1]. After multiplying 100 by this value, an exponent of 0 corresponds to ideal resilience of 100, and an exponent approaching negative infinity corresponds to resilience approaching 0. Finally, the resulting negative exponential function value is multiplied by 100 to obtain the final comprehensive resilience index. The comprehensive resilience index ranges from zero to 100. The comprehensive resilience index can be interpreted intuitively. For example, a value above 80 indicates stable operation, 60-80 indicates that the pressure is controllable but needs attention, and a value below 60 indicates insufficient resilience that requires warning and may trigger the strategy inference module.

[0098] In the calculation of dynamic influence weights, the method for determining the metric representing the size of the bottleneck node's influence range is as follows:

[0099] In a dynamic attributed network model, the bottleneck node is simulated to be in an invalid state. Then, the number of downstream network nodes directly affected when the traffic originally planned to flow through the bottleneck node needs to be redistributed to other paths is identified and counted. This number is used as a metric. This method quantifies the ripple effect of node failure from the perspective of network flow. The affected downstream network nodes usually include their direct successor nodes and adjacent nodes whose load has increased significantly or whose paths have changed due to traffic rerouting. The statistics can be automatically identified by analyzing the changes in network flow distribution before and after the simulation.

[0100] In this embodiment, it is necessary to specifically explain the process of network state evolution after the implementation of the intervention strategy in the strategy deduction and evaluation module as follows:

[0101] First, the imported preset or user-defined intervention strategies are parsed into a series of explicit modification instructions. These instructions directly affect the elements in the dynamic attributed network model. The types of modification instructions include, but are not limited to, modifying the binary service status of network nodes, adjusting the dynamic processing rate of network nodes, or resetting the dynamic edge weights of directed edges. For example, a modification instruction can change the binary service status of a security checkpoint node from "available" to "unavailable" to simulate a closed checkpoint, or increase the dynamic processing rate of a check-in counter by 20% to simulate opening additional counters, or increase the dynamic edge weight of the directed edge connecting the boarding gate and the runway by a fixed value to simulate aircraft ground taxiing delays.

[0102] Then, a copy of the dynamically attributed network model is created, and the parsed modification instructions are applied to the copy of the dynamically attributed network model to generate an initial inference network model with the intervention strategy applied.

[0103] Next, the embedded simulation engine is invoked, and based on the principle of discrete event simulation, the initial inference network model is driven to evolve its state within a preset future inference time window; the length of the inference time window can be set according to the evaluation requirements, such as the next thirty or sixty minutes.

[0104] Based on real-time and predicted flight and passenger flow data, the simulation engine generates and schedules simulated passenger, baggage and aircraft entities within the initial simulation network model. These entities move between network nodes according to the network topology. The entities move in the network according to a preset process. For example, the passenger entity passes through the check-in counter network node, the security checkpoint network node, and finally arrives at the boarding gate network node.

[0105] The service processing time of an entity at different network nodes is determined by the stochastic service model built into the simulation engine. This stochastic service model makes the service time follow a specified probability distribution, the expected value of which is equal to the reciprocal of the current dynamic processing rate of the network node. This means that for a network node with a dynamic processing rate of ten people per minute, the average time to serve each entity is 0.1 minutes (six seconds).

[0106] The transmission time of an entity on a directed edge is determined by the current dynamic edge weight on that directed edge, plus an additional random fluctuation value, to simulate the uncertainty in actual transmission; the random fluctuation value is used to characterize the natural fluctuations in travel time, such as differences in walking speed or small changes in vehicle speed.

[0107] During the simulation, the current queue length and dynamic processing rate of network nodes are dynamically updated based on the arrival, service, and departure events of entities. When an entity arrives at a network node, the current queue length of that node increases; when an entity begins to be served, the queue length decreases; if the modification instruction adjusts the dynamic processing rate of a network node, the service time of subsequent entities will be calculated based on the new processing rate.

[0108] Finally, the simulation engine outputs the final network state at the end of the simulation time window, which is the simulated dynamic attributed network model. This dynamic attributed network model includes the final current queue length, dynamic processing rate, and final dynamic edge weight of all directed edges of all network nodes after the simulation ends.

[0109] The specific process for recalculating the simulated comprehensive resilience index and outputting the assessment report is as follows:

[0110] First, based on the deduced dynamic attributed network model, the bottleneck dynamic identification module is executed to identify the bottleneck node list and bottleneck strength index after recalculation.

[0111] Meanwhile, based on the deduced dynamic attributed network model and the deduced bottleneck node list, the calculation process of the resilience index calculation module is executed to recalculate the deduced comprehensive resilience index.

[0112] Then, a multi-dimensional comparative analysis is conducted to calculate the difference between the comprehensive resilience index after the deduction and the comprehensive resilience index based on when the strategy deduction and evaluation module is triggered, which is used as the resilience improvement degree; a positive resilience improvement degree indicates that the intervention strategy has improved the system resilience, while a negative value indicates that the system resilience has decreased.

[0113] The bottleneck mitigation rate is calculated by taking the number of bottleneck nodes in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered, and the number of bottleneck nodes in the bottleneck node list after simulation. The intersection of the number of bottleneck nodes in the bottleneck node list is calculated by subtracting the number of bottleneck nodes from the number of bottleneck nodes in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered. The bottleneck mitigation rate ranges from zero to one. The higher the value, the higher the proportion of the original bottleneck nodes that have been successfully eliminated.

[0114] Identify and list the new bottleneck nodes that appear in the bottleneck node list after the simulation, and that are not in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered. These new bottleneck risk nodes indicate the secondary problems or risk transfer that the intervention strategy may cause.

[0115] Finally, an evaluation report is generated, which includes resilience improvement, bottleneck mitigation rate, a list of new bottleneck risk nodes, and a comparison of the changes in at least one of the following before and after the simulation: network flow capacity, network node dynamic processing rate, current queuing length of network nodes, and dynamic edge weight of directed edges. The evaluation report is presented in a structured form, such as including tables and trend charts, to intuitively show the expected effects and potential risks of the intervention strategy and provide a quantitative basis for operational decisions.

[0116] In the simulation engine, the simulation of the service processing time of network nodes and the transmission time of entities adopts a modeling method that introduces random perturbations.

[0117] The probability distribution that a network node follows when processing a service for an entity is an exponential distribution. The exponential distribution is chosen because its memorylessness is suitable for describing the random intervals of many service processes, and its probability density function is uniquely determined by the dynamic processing rate.

[0118] The random fluctuation value added to the transmission time of an entity on a directed edge is obtained by randomly sampling from a normal distribution with a mean of zero. The standard deviation of the normal distribution can be set according to the fluctuation of historical transmission time data, for example, set to 10% of the dynamic edge weight value, so as to introduce reasonable randomness in the simulation.

[0119] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0120] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0121] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0122] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0123] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0124] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0125] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent monitoring and analysis system for airport equipment operation safety, characterized in that, Specifically, it includes: The modules are connected sequentially: data fusion and modeling module, bottleneck dynamic identification module, resilience index calculation module, and strategy deduction and evaluation module. Data fusion and modeling module: Real-time access to airport operation data, including flight schedules, passenger flow, equipment status and location information. Based on the preset airport topology, the module abstracts airport entities into network nodes and the processes between entities into edges, constructing and continuously updating a dynamic attributed network model. The attribute values ​​of network nodes and edges are dynamically assigned based on the real-time access data. Bottleneck dynamic identification module: running on a dynamic attributed network model, it applies network flow analysis algorithms to calculate the current network's flow capacity, and through constraint propagation and marginal contribution analysis algorithms, identifies the critical path that restricts flow capacity and a bottleneck node located on that critical path, generating and outputting a list of bottleneck nodes containing bottleneck node identifiers and their bottleneck strength ratings. Resilience Index Calculation Module: Receives a list of bottleneck nodes and extracts the real-time operating parameters of each bottleneck node. The real-time operating parameters include load rate and queuing change trends. Based on the preset multi-indicator fusion rules, the module performs weighted and composite calculations on the real-time operating parameters to generate a comprehensive resilience index that characterizes the overall operational robustness of the airport. Strategy simulation and evaluation module: Triggered when the comprehensive resilience index is lower than the preset threshold, based on the dynamic attributed network model and the comprehensive resilience index, the preset intervention strategy is imported, the embedded simulation engine is called to simulate the network state evolution after the implementation of the intervention strategy, the comprehensive resilience index after simulation is recalculated, and the changes in the comprehensive resilience index and the distribution changes of bottleneck nodes before and after simulation are compared, and an evaluation report is output.

2. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 1, characterized in that: The specific process of constructing and continuously updating the dynamic attributed network model in the data fusion and modeling module is as follows: First, based on a pre-defined airport topology knowledge base, the airport is defined as a set of network nodes, including check-in counters, security checkpoints, boarding gates, baggage claim carousels, aircraft stands, and runway entrances. The physical relationship between entities is defined as a set of directed edges connecting network nodes, including edges representing passenger flow from check-in counter network nodes to security checkpoint network nodes, edges representing security checkpoint network nodes to boarding gate network nodes, and edges representing aircraft flow from boarding gate network nodes to runway entrance network nodes. For each of the network nodes, a dynamic attribute tuple is calculated, which includes static theoretical processing capacity, dynamic processing rate, current queue length, and binary service status. The dynamic processing rate is obtained by multiplying the average rate obtained by dividing the actual number of processes flowing through network nodes within a preset time window by the window length by a real-time performance coefficient that is determined based on the real-time device status and has a value between zero and one. Meanwhile, a dynamic edge weight is calculated for each directed edge, which integrates the baseline travel time and the real-time congestion delay. The process of determining the dynamic edge weight is as follows: first, add an adjustment coefficient to the baseline time of movement between the network nodes at both ends of the connecting edge, then multiply it by the current average additional delay calculated based on real-time location information, and add the two to obtain the final value. Ultimately, the complete structure containing the set of network nodes, the set of directed edges, the dynamic attribute tuples of all network nodes, and the dynamic edge weights of all edges will be used as the dynamic attributed network model.

3. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 2, characterized in that: The mechanism for continuously updating the dynamic attributed network model is as follows: Set up an update trigger that is triggered at fixed time intervals or based on events that cause changes in critical operational data; When the update trigger is triggered, based on the latest accessed airport operation data, the dynamic processing rate, current queue length, binary service status, and dynamic edge weight of each directed edge in the dynamic attributed network model are recalculated, thereby refreshing the dynamic attribute tuples and dynamic edge weights in the dynamic attributed network model, and outputting the refreshed dynamic attributed network model.

4. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 3, characterized in that: The specific process of identifying critical paths and bottleneck nodes on the dynamic attributed network model in the bottleneck dynamic identification module is as follows: First, based on the dynamic attribute tuple of each network node in the dynamic attributed network model, the current effective throughput capacity of each network node is calculated. The effective throughput capacity is the dynamic processing rate of the network node plus a buffer component. This buffer component is obtained by dividing the positive value of the static theoretical processing capacity of the network node by the dynamic processing rate by one plus the current queue length of the network node. Next, feasible flow paths from the source to the sink are searched in the dynamic attributed network model, and the contribution and resistance of each feasible flow path to the current network's flow capacity are evaluated. The contribution of a feasible flow path is quantified by the proportion of the flow that the feasible flow path can carry in the dynamic attributed network model to the current network's flow capacity. The resistance to a feasible path is quantified by calculating the sum of the ratios of the dynamic edge weights of each edge on the feasible path to the average dynamic edge weight of all edges, and then taking the reciprocal. Feasible paths with high contribution and relatively low resistance are marked as critical paths. Then, for each network node in the dynamic attributed network model, a comprehensive evaluation is performed based on the following three indicators: The first indicator is the network node's own saturation, which is the dynamic processing rate of the network node divided by the static theoretical processing capacity of the network node. The second metric is the structural criticality of network nodes, which is the sum of the contributions of each feasible circulation path marked as a critical path to the network node. The third indicator is the impact of network node failure, which is calculated by subtracting the simulated current network throughput capacity from the calculated current network throughput capacity. The simulated current network throughput capacity is the value recalculated after temporarily invalidating the network node in the dynamic attributed network model. Finally, network nodes are filtered and sorted based on a bottleneck strength index, which is calculated based on the first, second and third indicators, generating a bottleneck node list that includes the bottleneck node identifier, bottleneck strength index and information of the critical path to which it belongs.

5. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 4, characterized in that: The calculation process for the bottleneck strength index is as follows: First, the hyperbolic tangent function is applied to the first indicator, namely the saturation of the network nodes themselves, for mathematical transformation. The second indicator, namely the structural criticality of network nodes, is normalized by dividing it by the maximum value of the second indicator among all network nodes. For the third indicator, namely the impact of network node failure, its calculated value is used directly; Then, the first, second, and third indicators that have undergone the above processing are multiplied by their corresponding first, second, and third preset weights, respectively. The three products are then added together, and the sum is the bottleneck strength index.

6. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 5, characterized in that: The specific process for generating the comprehensive resilience index in the resilience index calculation module is as follows: First, for each bottleneck node in the bottleneck node list, calculate a node vulnerability score, which is obtained by adding the first and second items; The first term is the quotient obtained by dividing the dynamic processing rate of the bottleneck node by its static theoretical processing capacity. The second item is a preset trend weight coefficient, multiplied by a transformation value, which is the result of applying the hyperbolic tangent function to a ratio. The ratio is the difference between the current queue length of the bottleneck node at the end of a preset time window and the queue length at the beginning of the time window, and then divided by a preset trend sensitivity parameter. Next, a dynamic influence weight is assigned to each bottleneck node in the bottleneck node list. The calculation process for this dynamic influence weight is as follows: multiply the bottleneck strength index of the bottleneck node by a metric representing the size of the bottleneck node's influence range in the current network state to obtain an intermediate product; then divide the intermediate product by the sum of the intermediate products of all bottleneck nodes in the bottleneck node list, and the quotient is the dynamic influence weight of the bottleneck node; the sum of the dynamic influence weights of all bottleneck nodes calculated according to this rule is one; synthesize the comprehensive resilience index. The calculation process for the comprehensive resilience index is as follows: first, multiply the node vulnerability score of each bottleneck node by its dynamic influence weight to obtain the weighted vulnerability of the bottleneck node; Then, the weighted fragility of all bottleneck nodes is summed to obtain a aggregate fragility; the aggregate fragility is multiplied by a positive constant, which is used to adjust the strength of the influence of the aggregate fragility on the fragility index decay, i.e., the system fragility decay coefficient, and then the negative exponential function value of the product is taken as the base of the natural constant; finally, the obtained negative exponential function value is multiplied by one hundred to obtain the final comprehensive fragility index, which ranges from zero to one hundred.

7. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 6, characterized in that: In the calculation of the dynamic influence weight, the method for determining the metric representing the size of the bottleneck node's influence range is as follows: In the dynamic attributed network model, the bottleneck node is simulated to be in an invalid state. Then, the number of downstream network nodes directly affected when the traffic originally planned to flow through the bottleneck node needs to be redistributed to other paths is identified and counted. This number is used as a metric.

8. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 7, characterized in that: In the strategy deduction and evaluation module, the specific process of simulating the network state evolution after the implementation of the intervention strategy is as follows: The imported preset intervention strategy is parsed into a series of explicit modification instructions. These modification instructions directly affect the elements in the dynamic attributed network model. The types of modification instructions include, but are not limited to, modifying the binary service status of network nodes and adjusting the dynamic processing rate of network nodes. Create a copy of the dynamically attributed network model and apply the modification instructions to the copy to generate the initial inference network model; The embedded simulation engine is invoked, and based on the principle of discrete event simulation, the initial inference network model is driven to evolve its state within a preset future inference time window. Based on real-time and predicted flight and passenger flow data, the simulation engine generates and schedules simulated passenger, baggage and aircraft entities within the initial simulation network model. These entities move between network nodes according to the network topology. The service processing time of an entity at different network nodes is determined based on the stochastic service model built into the simulation engine. This stochastic service model makes the service time follow a specified probability distribution, the expected value of which is equal to the reciprocal of the current dynamic processing rate of the network node. The transmission time of an entity on a directed edge is determined by the current dynamic edge weight on that directed edge, plus an additional random fluctuation value. During the simulation, the current queue length and dynamic processing rate of network nodes are dynamically updated based on the arrival, service, and departure events of entities; Finally, the simulation engine outputs the final network state at the end of the simulation time window, which is the simulated dynamic attributed network model.

9. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 8, characterized in that: The specific process of recalculating the simulated comprehensive resilience index and outputting the evaluation report is as follows: First, based on the deduced dynamic attributed network model, the bottleneck dynamic identification module is executed to identify the bottleneck nodes and bottleneck strength index after recalculation. Then, based on the deduced dynamic attributed network model and the deduced bottleneck node list, the resilience index calculation module is executed to calculate the comprehensive resilience index after recalculation. A comparative analysis is then performed to calculate the difference between the comprehensive resilience index after recalculation and the comprehensive resilience index used when the strategy deduction and evaluation module is triggered, which is taken as the degree of resilience improvement. The number of bottleneck nodes in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered is calculated, and the number of the intersection of the number of bottleneck nodes in the bottleneck node list after simulation is calculated. That is, the number of bottleneck nodes that appear in both lists at the same time. Then, the number of the intersection is subtracted from one and divided by the number of bottleneck nodes in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered. The result is used as the bottleneck mitigation rate. Identify and list the new bottleneck nodes that appear in the bottleneck node list after the simulation, and that are not in the bottleneck node list on which the strategy simulation and evaluation module is based when it is triggered, as new bottleneck risk nodes; Finally, an evaluation report is generated that includes resilience improvement, bottleneck mitigation rate, a list of new bottleneck risk nodes, and a comparison of the changes in at least one of the following before and after the simulation: network throughput, dynamic processing rate of network nodes, current queue length of network nodes, and dynamic edge weight of directed edges.

10. The intelligent monitoring and analysis system for airport equipment operation safety according to claim 9, characterized in that: In the simulation engine, the simulation of the service processing time of network nodes and the transmission time of entities adopts a modeling method that introduces random perturbations. The service time of a network node in processing an entity follows a specified probability distribution, which is an exponential distribution. The random fluctuation value added to the transmission time of an entity on a directed edge is obtained by randomly sampling from a normal distribution with a mean of zero.

Images