Subway access control system multi-specialty collaborative construction monitoring method and system

Abstract

The application relates to the technical field of construction monitoring, and discloses a metro access control system multi-specialty collaborative construction monitoring method and system. The method comprises the following steps: collecting the voltage, protocol and delay parameter between access control equipment, calculating the coupling strength and constructing a matrix; dividing an equipment installation unit based on the coupling strength matrix; constructing an access control circuit knowledge enhanced graph neural network, and outputting an equipment installation priority score matrix; monitoring a multi-specialty operation state, identifying a construction conflict, generating an early warning signal; dynamically adjusting an equipment installation time sequence in response to the early warning signal, and obtaining a construction scheduling scheme. The application solves the technical problems of lacking of equipment electrical topology analysis, integrated application of professional knowledge and real-time conflict detection in multi-specialty construction of the metro access control system. The application improves the intelligent monitoring level and construction efficiency of multi-specialty collaborative construction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of construction monitoring, in particular to a metro access control system multi-specialty collaborative construction monitoring method and system. BACKGROUND

[0002] The existing metro access control system construction monitoring technology mainly relies on traditional project management methods and independent professional operation modes, and supervises and manages the construction process through manual preparation of construction plans, on-site inspection, and regular reporting. In the traditional method, the construction activities of different specialties such as civil engineering, electromechanical engineering, communication, and debugging are relatively independent, and each specialty operates according to the predetermined time node and spatial distribution. The construction monitoring mainly focuses on the progress control and quality inspection within a single specialty, and lacks systematic monitoring and intelligent scheduling of collaborative work among multiple specialties.

[0003] However, the existing technology has significant deficiencies, mainly manifested in the lack of in-depth analysis and utilization of the electrical topology relationship among access control devices, the inability to scientifically group and plan based on the coupling strength of the devices, and the resulting unreasonable construction arrangement. At the same time, the existing method lacks integrated application of professional knowledge in the access control field, and fails to convert knowledge in the fields of device installation specifications, electrical safety requirements, etc. into computable intelligent decision-making basis, resulting in subjectivity and randomness in determining construction priorities. In addition, the traditional monitoring method lacks precise digital modeling and real-time detection capability for space occupation and time conflicts in the construction site, making it difficult to identify and prevent potential conflicts in multi-specialty cross-operation in advance. SUMMARY

[0004] The present application provides a metro access control system multi-specialty collaborative construction monitoring method and system to solve the technical problems of lack of device electrical topology analysis, integrated application of professional knowledge, and real-time conflict detection in multi-specialty construction of metro access control systems. The present application improves the intelligent monitoring level and construction efficiency of multi-specialty collaborative construction.

[0005] In a first aspect, the present application provides a metro access control system multi-specialty collaborative construction monitoring method, which comprises:

[0006] Step S1: Collect the voltage level, communication protocol type, and signal transmission delay parameters among the access controller, card reader, and electromagnetic lock, calculate the coupling strength between devices, and construct a device coupling strength matrix;

[0007] Step S2: Based on the device coupling strength matrix, divide the access control devices into multiple device installation units;

[0008] Step S3: Construct a metro access circuit knowledge enhanced graph neural network, input the physical attribute features and electrical connection relationship of the device installation unit, and output a device installation priority score matrix;

[0009] Step S4: monitoring the multi-specialty operation state of the access control system installation construction, identifying the inter-professional construction conflict under the guidance of the equipment installation priority score matrix, and generating a multi-specialty construction conflict warning signal;

[0010] Step S5: in response to the multi-specialty construction conflict warning signal, dynamically adjusting the access control equipment installation timing, and obtaining a construction scheduling scheme.

[0011] In a second aspect, the application provides a subway access control system multi-specialty collaborative construction monitoring system, which comprises:

[0012] A construction module is configured to collect the voltage level, communication protocol type and signal transmission delay parameter among the access controller, card reader and electromagnetic lock, calculate the coupling strength between devices, and construct a device coupling strength matrix.

[0013] A division module is configured to divide the access control equipment into a plurality of equipment installation units based on the device coupling strength matrix.

[0014] An input module is configured to construct an access control circuit knowledge enhanced graph neural network, input the physical attribute features and electrical connection relationship of the equipment installation unit, and output a device installation priority score matrix.

[0015] An identification module is configured to monitor the multi-specialty operation state of the access control system installation construction, identify the inter-professional construction conflict under the guidance of the device installation priority score matrix, and generate a multi-specialty construction conflict warning signal.

[0016] A response module is configured to respond to the multi-specialty construction conflict warning signal, dynamically adjust the access control equipment installation timing, and obtain a construction scheduling scheme.

[0017] In a third aspect, a subway access control system multi-specialty collaborative construction monitoring device is provided, which comprises a memory and at least one processor, the memory has instructions stored therein; the at least one processor invokes the instructions in the memory, so that the subway access control system multi-specialty collaborative construction monitoring device executes the above-mentioned subway access control system multi-specialty collaborative construction monitoring method.

[0018] In a fourth aspect, a computer readable storage medium is provided, which has instructions stored therein, when it runs on a computer, it makes the computer execute the above-mentioned subway access control system multi-specialty collaborative construction monitoring method.

[0019] In the technical scheme provided in the application, the voltage level, communication protocol type and signal transmission delay parameter among the access control controller, card reader and electromagnetic lock are collected, and a device coupling strength matrix is constructed by calculating the coupling strength between devices, which overcomes the defect of ignoring the electrical correlation characteristics of access control devices in the prior art, realizes scientific grouping based on the electrical topology relationship of devices, and avoids the subjectivity and randomness of device grouping in traditional methods. The technical feature of dividing the access control devices into multiple device installation units based on the device coupling strength matrix considers the electrical dependency relationship and signal transmission characteristics between devices, so that the devices in the same installation unit have stronger electrical correlation, reducing the complexity and coordination difficulty of cross-unit construction. The technical feature of constructing an access control circuit knowledge enhanced graph neural network and inputting the physical property characteristics and electrical connection relationship of the device installation unit to output a device installation priority score matrix combines the professional knowledge in the access control field with the physical characteristics of the devices, overcomes the limitations of traditional methods relying on manual experience to determine the installation sequence, and improves the scientificity and accuracy of construction priority decision. The technical feature of monitoring the multi-specialty operation state of access control system installation and construction, and generating a multi-specialty construction conflict early warning signal by identifying the professional construction conflict under the guidance of the device installation priority score matrix solves the problem that the prior art cannot detect and warn multi-specialty cross-operation conflicts in real time, and improves the safety and controllability of the construction process.

[0020] The technical feature of responding to the multi-specialty construction conflict early warning signal and dynamically adjusting the access control device installation time sequence to obtain a construction scheduling scheme realizes the change of the construction management mode from passive response to active prevention, and overcomes the deficiency that the traditional static scheduling method cannot adapt to the dynamic changes in the field. In this specific application field of the subway access control system, the electrical topology mapping algorithm fully considers the voltage compatibility, protocol consistency and signal transmission characteristics of the access control devices. Compared with the general device grouping method, this algorithm is specially optimized for the electrical characteristics of the access control system, so that the device grouping is more in line with the engineering practice of the access control system. The access control circuit knowledge enhanced graph neural network embeds the field-specific knowledge such as installation specifications and electrical safety distance requirements of access control devices into the learning process of the neural network. Compared with the general graph neural network, this network is specially enhanced with knowledge for the technical characteristics of the access control system, improving the professionalism and practicality of the priority score. The particle swarm optimization algorithm fully considers the dependency constraints, spatial conflict penalties and other constraints specific to the access control system in the construction scheduling of the access control system. Compared with the general scheduling optimization algorithm, this algorithm is specially customized for the construction characteristics of the subway access control system, so that the optimization result is more in line with the construction requirements and safety standards of the access control system. BRIEF DESCRIPTION OF DRAWINGS

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of an embodiment of the multi-disciplinary collaborative construction monitoring method for subway access control systems in this application.

[0023] Figure 2 This is a schematic diagram of an embodiment of the multi-disciplinary collaborative construction monitoring system for subway access control in this application.

[0024] Figure 3 This is a schematic block diagram of the multi-disciplinary collaborative construction monitoring equipment for the subway access control system in this embodiment of the invention. Detailed Implementation

[0025] This application provides a multi-disciplinary collaborative construction monitoring method and system for a subway access control system. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0026] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the multi-disciplinary collaborative construction monitoring method for subway access control systems in this application includes:

[0027] Step S1: Collect the voltage level, communication protocol type, and signal transmission delay parameters between the access controller, card reader, and electromagnetic lock; calculate the coupling strength between the devices; and construct the device coupling strength matrix.

[0028] Step S2: Based on the device coupling strength matrix, divide the access control device into multiple device installation units;

[0029] Step S3: Construct an enhanced graph neural network for access control circuit knowledge, input the physical attribute features and electrical connection relationships of the device installation unit, and output the device installation priority scoring matrix;

[0030] Step S4: Monitor the multi-disciplinary work status of the access control system installation and construction, identify inter-disciplinary construction conflicts under the guidance of the equipment installation priority scoring matrix, and generate a multi-disciplinary construction conflict early warning signal;

[0031] Step S5: Respond to the multi-disciplinary construction conflict early warning signal, dynamically adjust the installation sequence of the access control equipment, and obtain a construction scheduling plan.

[0032] It is understood that the executing entity of this application can be a multi-disciplinary collaborative construction monitoring system for subway access control, or it can be a terminal or a server; the specifics are not limited here. This application's embodiment uses a server as an example for illustration.

[0033] Specifically, a device coupling strength matrix is ​​constructed by collecting voltage levels, communication protocol types, and signal transmission delay parameters between access controllers, card readers, and electromagnetic locks. Voltage levels reflect the electrical compatibility between devices, communication protocol types reflect the degree of data transmission compatibility, and signal transmission delay parameters characterize the time characteristics of device response. The device coupling strength is calculated using a weighted summation method, assigning weight coefficients to voltage matching degree, protocol compatibility, and the reciprocal of transmission delay. Voltage matching degree is determined by comparing the deviation between the device's rated voltage and the standard voltage; protocol compatibility is evaluated based on the communication protocol standards supported by the device; and the reciprocal of transmission delay is calculated by measuring the signal transmission time between devices. The calculated coupling strength values ​​are then filled into the corresponding positions in the matrix according to the device number order, forming a coupling strength matrix reflecting the degree of electrical correlation between devices. Normalization is then performed to eliminate the influence of dimensions.

[0034] Device grouping is performed based on a standardized device coupling strength matrix. First, a degree matrix for access control devices is constructed, where each diagonal element represents the number of connections between the corresponding device and other devices. The Laplace matrix is ​​calculated by subtracting the degree matrix from the coupling strength matrix, reflecting the connection structure characteristics between devices. Eigenvalue decomposition is then performed on the Laplace matrix, decomposing it into a product of eigenvalues ​​and eigenvectors. Eigenvalues ​​and their corresponding eigenvectors below the electrical coupling strength threshold are selected; these eigenvectors contain key information about device grouping. The selected eigenvectors are then weighted and summed with the spatial coordinates of the access control devices, including their three-dimensional location on the subway platform. This weighted summation combines electrical and spatial features to form a comprehensive spatial constraint feature representation. An improved k-means clustering algorithm is then executed based on this spatial constraint feature representation. During clustering, the number of devices within the same installation unit is limited to a preset access control component capacity threshold, ensuring a balanced workload for each installation unit.

[0035] A knowledge-enhanced graph neural network for access control circuits is constructed to generate a device installation priority scoring matrix. The location coordinates, device type code, power rating, installation height, and reserved space dimensions of each access control device in the installation unit are extracted and combined to form a device physical attribute feature vector. The device type code uses one-hot encoding to convert different types of access control devices into numerical vectors. The power rating reflects the electrical power requirements of the device, the installation height indicates the vertical installation position, and the reserved space dimension describes the space range required for device installation. Simultaneously, the standard installation specifications, electrical safety distance requirements, and commissioning and acceptance standards for access control devices are converted into vector representations to construct an access control circuit knowledge graph embedding vector. This knowledge graph embedding vector contains professional knowledge and installation specifications for the access control system, and vectorization allows this professional knowledge to participate in the neural network's computation process. Electrical connection weights are calculated based on the voltage difference, current capacity, and signal type parameters between access control devices. The voltage difference reflects the voltage compatibility between devices, the current capacity represents the current carrying capacity of the device, and the signal type describes the signal characteristics transmitted between devices. These electrical connection weights are then filled into the device adjacency matrix to construct the access control system device adjacency matrix. The knowledge graph embedding vector of the access control circuit is concatenated with the physical attribute feature vector of the devices through a matrix concatenation operation. This operation connects the two vectors along the feature dimension, forming a comprehensive feature vector that includes both physical attributes and professional knowledge. The embedding layer parameters of the graph neural network are initialized based on this comprehensive feature vector and the device adjacency matrix. The graph neural network learns the correlation features between devices through multiple layers of graph convolution operations. Access control device installation dependency constraint weights and spatial conflict penalty coefficients are set for the graph attention layer. The attention mechanism automatically learns the importance weights between devices, the dependency constraint weights ensure that dependent devices are installed in the correct order, and the spatial conflict penalty coefficient prevents multiple devices from being installed simultaneously in the same space.

[0036] This project monitors the multi-disciplinary work status of access control system installation and identifies construction conflicts. By collecting 3D spatial coordinate data of the subway platform construction area, the entire construction space is divided into regular grid cells, each with fixed length, width, and height dimensions. A space occupancy matrix is ​​constructed to record the occupancy status of each grid cell at different times, represented by binary values: 1 for occupied and 0 for idle. Based on an equipment installation priority scoring matrix, the installation time window and space requirement set for each access control device are determined. The time window defines the start and end times of equipment installation, and the space requirement set describes the range of grid cells required during equipment installation. A space requirement database for equipment installation is established, storing the space requirement information and time schedule for each device. Spatiotemporal conflict detection calculations are performed based on the space occupancy matrix and the equipment installation space requirement database. Conflict detection identifies potential conflicts by comparing the overlap of time windows and space requirements of different devices. When the time windows and space requirements of two or more devices overlap, a construction conflict is identified. The identified conflicts are classified according to their degree of impact: minor conflicts affect construction efficiency but not safety, moderate conflicts may lead to rework, and serious conflicts pose safety hazards.

[0037] The system responds to multi-disciplinary construction conflict early warning signals and dynamically adjusts the installation sequence of access control equipment. Upon receiving a conflict early warning signal, it analyzes the conflict type and its impact range. Conflict types include spatial, temporal, and resource conflicts, while the impact range describes the number of equipment involved and the affected construction area. Based on the conflict information, a set of conflict constraints is constructed, including time adjustment restrictions for conflicting equipment, space reallocation requirements, and personnel rescheduling needs. A multi-objective optimization function is established using an equipment installation priority scoring matrix. The optimization objectives include minimizing total construction time, maximizing installation quality scores, and minimizing inter-disciplinary waiting time. Weight coefficients are assigned to each optimization objective, reflecting the importance of different objectives. The multi-objective optimization function is input into a particle swarm optimization algorithm for iterative solution. The particle swarm optimization algorithm searches for the optimal solution by simulating bird flock foraging behavior. Each particle represents a possible scheduling scheme, moving in the solution space to find the optimal position. During the algorithm iteration, the position and velocity of the particles are continuously updated, guiding the particles towards the optimal solution through individual optimal positions and the global optimal position. The optimal adjustment scheme for the access control equipment installation sequence is calculated, including the new installation time and space allocation for each device. Based on the adjusted plan, the installation time window and professional operation sequence of the access control equipment were updated to form a construction scheduling plan.

[0038] In one specific embodiment, step S1 includes:

[0039] Establish a set of access control device nodes and a set of electrical connection edges, and construct a directed graph of the electrical topology of the access control system;

[0040] The device coupling strength is obtained by weighting the voltage matching degree, protocol compatibility and reciprocal of transmission delay between devices in the directed graph of the electrical topology.

[0041] The device coupling strength values ​​are filled into the corresponding positions in the matrix according to the device number order to construct the device coupling strength matrix;

[0042] The device coupling strength matrix is ​​normalized to obtain the device coupling strength matrix.

[0043] Specifically, the process of constructing a directed graph of the electrical topology of the access control system by establishing a set of access control device nodes and a set of electrical connection edges requires identifying all access control devices within the subway station and assigning them unique identifiers. The set of access control device nodes includes the access control server in the control center, the turnstile controllers at each entrance and exit, card readers, electromagnetic locks, door magnetic sensors, and emergency door opening buttons. Each device node has attribute information such as device type, device number, installation location, and electrical parameters. The set of electrical connection edges describes the physical connection relationships between devices, including power supply lines, signal transmission lines, and communication data lines. The construction of the directed graph determines the directionality of edges by analyzing the signal flow between devices. A controller sends a query command to a card reader, forming a directed edge from the controller to the card reader; the card reader returns card information to the controller, forming a directed edge from the card reader to the controller. The directed graph of the electrical topology is stored using an adjacency list data structure. Each node maintains a list of connected devices, containing the identifier and connection type information of the target device. The process of constructing the topology graph requires traversing the electrical interface definitions of all devices, determining the connection relationships between devices based on the input and output characteristics of the interfaces, and verifying the validity of the connections by considering the electrical compatibility requirements of the devices.

[0044] The process of calculating the device coupling strength value based on the weighted average of voltage matching degree, protocol compatibility, and reciprocal of transmission delay in a directed graph of electrical topology involves the quantification of three key parameters. Voltage matching degree is calculated by comparing the operating voltage ranges of two devices. When the output voltage range of device A overlaps with the input voltage range of device B, the matching degree is equal to the ratio of the length of the overlap to the length of the input voltage range of device B. Protocol compatibility is evaluated based on the communication protocol standards supported by the devices. Devices supporting the RS485 protocol have a compatibility of 1, devices supporting different protocols but with conversion interfaces have a compatibility of 0.8, and completely incompatible devices have a compatibility of 0. The reciprocal of transmission delay is calculated by measuring the signal propagation time between devices, including the propagation time of the electrical signal in the conductor and the processing time within the device. The weighted calculation uses a linear weighted summation method, multiplying the voltage matching degree by a voltage weighting coefficient, the protocol compatibility by a protocol weighting coefficient, and the reciprocal of transmission delay by a delay weighting coefficient, and then summing the three products to obtain the device coupling strength value. The weighting coefficients are determined based on the access control system's assessment of the importance of different parameters. The voltage matching weight is usually set to 0.4, the protocol compatibility weight is set to 0.4, and the transmission delay weight is set to 0.2.

[0045] The process of constructing the device coupling strength matrix by filling the device coupling strength values ​​into the corresponding positions in the matrix according to the device number order requires establishing a mapping relationship between device numbers and matrix indices. Device numbers use a hierarchical encoding method: the first layer represents the device type, the second layer represents the device's sequence number within the same type, and the third layer represents the specific functional module of the device. The row and column indices of the matrix correspond to the numbers of the source and target devices, respectively, and the values ​​of the matrix elements represent the coupling strength between the corresponding device pairs. The matrix filling process proceeds in lexicographical order of device numbers, first processing the coupling relationships of the device with the smallest number with all other devices, then processing the remaining coupling relationships of the device with the second smallest number, and so on, until the coupling relationships of all device pairs have been calculated and filled into the matrix. For device pairs without direct electrical connections, their coupling strength value is set to 0. The diagonal elements of the matrix represent the coupling relationship between the device and itself, typically set to 1 to indicate complete coupling. The asymmetry of the matrix reflects the directional characteristics of the connections between devices in the access control system; the coupling strength from the controller to the card reader differs from the coupling strength from the card reader to the controller.

[0046] The process of normalizing the equipment coupling strength matrix to obtain a standardized equipment coupling strength matrix employs a max-min normalization method to eliminate the influence of different parameter dimensions. The normalization process first calculates the maximum and minimum values ​​of all non-zero elements in the matrix. Then, for each matrix element, the minimum value is subtracted, and the result is divided by the difference between the maximum and minimum values. The normalization formula is: the new value equals the original value minus the minimum value, divided by the maximum value minus the minimum value. The normalized value ranges from 0 to 1. Zero elements in the matrix remain 0 during normalization, indicating no coupling relationship between the corresponding devices. Normalization ensures that the coupling strength of different types of parameters has the same numerical scale, facilitating subsequent cluster analysis and optimization calculations. The standardized matrix maintains the structural characteristics and relative size relationships of the original matrix while eliminating the influence of absolute numerical differences on the calculation results.

[0047] In one specific embodiment, step S2 includes:

[0048] Construct the access control device degree matrix based on the device coupling strength matrix, and calculate the Laplace matrix of the access control system.

[0049] The access control system's Laplace matrix is ​​subjected to eigenvalue decomposition, and eigenvalues ​​below the electrical coupling strength threshold are filtered out to obtain the access control device group feature vector.

[0050] The spatial constraint feature representation is obtained by performing a weighted summation operation between the grouped feature vectors of the access control device and the installation spatial coordinates of the access control device.

[0051] Based on the spatial constraint feature representation, an improved k-means clustering calculation is performed to limit the number of devices in the same device installation unit to no more than a preset access control component capacity threshold, thereby obtaining the multiple device installation units.

[0052] Specifically, the process of constructing the access control device degree matrix based on the device coupling strength matrix and calculating the access control system Laplace matrix involves fundamental mathematical operations in graph theory. The access control device degree matrix is ​​a diagonal matrix, where the diagonal elements represent the sum of the connection strengths between the corresponding device and other devices, and the off-diagonal elements are all zero. The degree matrix is ​​calculated by traversing each row of the device coupling strength matrix and summing all non-zero elements in that row to obtain the degree value of the corresponding device. The degree value reflects the density of connections of the device in the electrical topology of the access control system; a larger degree value indicates a tighter electrical connection between the device and other devices. The access control system Laplace matrix is ​​obtained by subtracting the device coupling strength matrix from the degree matrix. The Laplace matrix is ​​an important tool in graph theory for describing network structure characteristics. The diagonal elements of the Laplace matrix are equal to the degree values ​​of the corresponding devices, and the off-diagonal elements are equal to the negative values ​​of the corresponding elements in the device coupling strength matrix. The Laplace matrix has a positive semi-definite property; its eigenvalues ​​are all non-negative real numbers, and the distribution of these eigenvalues ​​reflects the connectivity and clustering structure characteristics of the network. The matrix calculation process follows the standard matrix subtraction rules, calculating the difference between the degree matrix and the corresponding element of the coupling strength matrix element by element.

[0053] The process of eigenvalue decomposition (EVD) of the Laplace matrix of the access control system and filtering eigenvalues ​​below the electrical coupling strength threshold to obtain the grouped eigenvectors of the access control devices employs the eigenvalue decomposition algorithm from linear algebra. Eigenvalue decomposition decomposes the Laplace matrix into a product of eigenvalues ​​and eigenvectors. Eigenvalues ​​represent the scaling factor of the matrix along the direction of the corresponding eigenvector, and eigenvectors represent the principal direction of the matrix transformation. Eigenvalue decomposition is calculated using the Jacobi iterative algorithm or the QR decomposition algorithm, which diagonalizes the Laplace matrix through repeated matrix transformations. The electrical coupling strength threshold is a pre-set value based on the electrical characteristics and construction requirements of the access control system, and its selection is based on the minimum acceptable coupling strength level between devices. The filtering process iterates through all calculated eigenvalues, selecting those less than or equal to the electrical coupling strength threshold and their corresponding eigenvectors. The eigenvectors corresponding to small eigenvalues ​​contain information about the weakly connected parts of the network, which helps identify relatively independent groups of devices. The grouped eigenvectors of the access control devices are a set of filtered eigenvectors, where the dimension of each eigenvector equals the total number of access control devices, and each element in the vector corresponds to a group weight for a specific device. The numerical values ​​of the elements of the feature vector reflect the tendency of the corresponding device to belong to that group dimension. The larger the value, the more likely the device is to belong to that group.

[0054] The process of obtaining spatial constraint feature representation by weighted summation of the access control device group feature vector and the installation spatial coordinates of the access control device combines electrical characteristics and spatial location information. The installation spatial coordinates of the access control device include the X, Y, and Z coordinates of the device in the three-dimensional space of the subway platform, and the coordinate values ​​are calibrated using the platform's architectural coordinate system. Spatial coordinates are obtained through architectural drawing analysis, on-site measurement, or export from 3D modeling software, with a coordinate accuracy requirement at the centimeter level to meet construction positioning needs. The weighted summation operation performs element-wise multiplication of each device's group feature vector with its spatial coordinate vector, and then adds the product results in the order of device number. The weighting coefficient reflects the relative importance of electrical and spatial characteristics in device grouping, and the determination of the weighting coefficient is based on the actual needs of access control system construction and expert experience. The spatial constraint feature representation is a multi-dimensional vector, with the dimension equal to the number of group feature vectors, and each dimension corresponding to a potential device grouping direction. Each element in the vector comprehensively considers the electrical correlation and spatial distribution characteristics of the device, and its value reflects the overall adaptability of the device in the corresponding group. The introduction of spatial constraints solves the problem of traditional electrical grouping methods ignoring construction space limitations, making the grouping results more in line with the operability requirements of actual construction.

[0055] The process of obtaining multiple equipment installation units by performing improved k-means clustering calculation based on spatial constraint feature representation and limiting the number of devices within the same equipment installation unit to no more than a preset access control component capacity threshold employs a constraint-optimized clustering algorithm. The improved k-means algorithm adds a capacity constraint mechanism to the traditional k-means algorithm, ensuring that the size of each cluster does not exceed a preset threshold limit. The preset access control component capacity threshold is determined based on the construction team's personnel configuration, tool and equipment capacity, and construction schedule. The threshold setting needs to balance the rationality of grouping and the operability of construction. The k-means clustering algorithm first randomly initializes k cluster centers, where k is equal to the expected number of equipment installation units. The initial positions of the cluster centers are randomly distributed in the data space of the spatial constraint feature representation. The clustering process adopts an iterative optimization approach, with each iteration including two steps: equipment allocation and center update. The equipment allocation step calculates the Euclidean distance from each device to each cluster center and assigns the device to the nearest cluster center that has not reached the capacity threshold. The center update step calculates the arithmetic mean of the spatial constraint feature representations of all devices in each cluster and uses this mean as the new cluster center position. A capacity constraint mechanism plays a role in the equipment allocation process. When the number of devices in a cluster reaches a capacity threshold, that cluster will no longer accept new equipment allocations, and the remaining devices can only be allocated to other clusters that are not yet full. The iterative process continues until the change in the location of the cluster center is less than a preset convergence threshold or the maximum number of iterations is reached. The clustering results form multiple equipment installation units, each containing a group of access control devices with high electrical correlation and relatively concentrated spatial locations.

[0056] In one specific embodiment, step S3 includes:

[0057] Extract the location coordinates, device type code, power level, installation height, and reserved space size parameters of each access control device in the multiple device installation units, and construct a device physical attribute feature vector;

[0058] The standard installation specifications, electrical safety distance requirements, and commissioning and acceptance standards for access control equipment are converted into vector representations, and an access control circuit knowledge graph embedding vector is constructed.

[0059] Based on the physical attribute feature vectors and electrical connection relationships of the devices, an adjacency matrix of the access control system devices is constructed, and an access control circuit knowledge-enhanced graph neural network is constructed by combining the embedding vectors of the access control circuit knowledge graph.

[0060] The physical attribute feature vector of the device is input into the knowledge-enhanced graph neural network of the access control circuit to perform graph attention mechanism calculation, and the hidden state vector of the device node is obtained.

[0061] The hidden state vector of the device node is mapped using a fully connected layer to output the device installation priority scoring matrix.

[0062] Specifically, the process of extracting the location coordinates, equipment type code, power rating, installation height, and reserved space dimensions of each access control device in multiple equipment installation units to construct the device's physical attribute feature vector involves the standardization of multi-dimensional data. Location coordinates include the X, Y, and Z axis values ​​of the device in the subway platform coordinate system. The origin of the coordinate system is set at the geometric center of the platform. The X-axis is along the train's direction of travel, the Y-axis is perpendicular to the X-axis pointing towards the platform's width, and the Z-axis represents the device's installation height. The equipment type code adopts a hierarchical coding system. The first layer of numbers represents the major equipment category, such as controller (1), card reader (2), and electromagnetic lock (3). The second layer of numbers represents the sub-category of the equipment, such as entry card reader (21) and exit card reader (22). The third layer of numbers represents the specific model and specifications of the equipment. The power rating describes the electrical power consumption level of the equipment. The power rating of controller equipment is typically between 50 watts and 200 watts, the power rating of card reader equipment is between 5 watts and 20 watts, and the power rating of electromagnetic lock equipment is between 10 watts and 50 watts. Installation height refers to the vertical distance of the device from the ground. The standard installation height for card readers is 1.2 meters, suitable for adults and children. The installation height for controllers is 2.5 meters to prevent unauthorized access. The installation height for electromagnetic locks matches the door frame structure. Reserved space dimensions include the length, width, and depth space required for device installation, as well as the necessary maintenance and operation space and safety distance around the device. Feature vector construction involves arranging the above five types of parameters in a fixed order to form a multi-dimensional array. The dimension of the array equals the total number of parameters, and each parameter occupies a specific position index in the vector. Data standardization processing uses the min-max normalization method to eliminate the influence of different parameter dimensions, scaling all parameter values ​​to the range of 0 to 1.

[0063] The process of constructing a knowledge graph for access control circuits by converting standard installation specifications, electrical safety distance requirements, and commissioning and acceptance standards for access control equipment into vector representations employs vectorization techniques for knowledge graphs. Standard installation specifications for access control equipment include requirements for installation location, installation direction, installation fixing methods, and wiring specifications. These specifications are converted into a set of numerical rules through text parsing and semantic analysis. Electrical safety distance requirements stipulate the minimum distances that must be maintained between different types of equipment; the safety distance requirements between high-voltage and low-voltage equipment are greater, and electromagnetic isolation distances must be maintained between high-voltage and low-voltage equipment. Commissioning and acceptance standards define the functional test items, performance indicators, and acceptance procedures after equipment installation. The standards include communication connection testing, electrical parameter testing, mechanical motion testing, and system integration testing. The knowledge graph is constructed using an entity-relationship-entity triple structure. Entities represent specific objects in the access control system, such as equipment, specifications, and standards; relationships represent logical connections between entities, such as installation relationships, testing relationships, and constraint relationships. Vector representation transforms triple structures into low-dimensional dense vectors using a knowledge graph embedding algorithm. The embedding algorithm employs the TransE model to map entities and relations to the same vector space. The access control circuit knowledge graph embedding vector contains prior knowledge of the access control system's professional domain. Each dimension of the vector corresponds to a feature direction in the knowledge space, and the vector value reflects the importance of the corresponding knowledge feature.

[0064] The process of constructing an access control system device adjacency matrix based on device physical attribute feature vectors and electrical connection relationships, and then combining this with the access control circuit knowledge graph embedding vectors to construct an access control circuit knowledge-enhancing graph neural network, involves neural network modeling of graph-structured data. The access control system device adjacency matrix is ​​a square matrix where row and column indices correspond to different access control devices, and matrix elements indicate whether there is an electrical connection between corresponding devices. The construction process of the adjacency matrix iterates through the electrical connection information of all device pairs. When there is a direct electrical connection between device A and device B, the element at the corresponding position in the matrix is ​​set to 1; otherwise, it is set to 0. Electrical connection relationships include power supply connections, signal transmission connections, data communication connections, and control command connections, etc., and different types of connection relationships are represented by different values ​​in the adjacency matrix. The access control circuit knowledge-enhancing graph neural network adopts a graph attention network architecture. The input layer of the network receives device physical attribute feature vectors and the adjacency matrix, the hidden layer learns the association features between devices through graph convolution operations, and the output layer generates high-level representation vectors of the devices. The knowledge enhancement mechanism is achieved by fusing the knowledge graph embedding vector of the access control circuit with the device feature vector. The fusion operation combines the two types of information by vector concatenation or vector addition. The parameters of the graph neural network are trained using the backpropagation algorithm, and the training objective is to minimize the loss function value between the predicted result and the true label.

[0065] The process of inputting the physical attribute feature vectors of devices into a knowledge-enhanced graph neural network for access control circuits to calculate the hidden state vectors of device nodes using a graph attention mechanism employs an attention mechanism to learn the importance weights between devices. The graph attention mechanism calculates the attention weights of each device node to its neighboring nodes, reflecting the contribution of neighboring nodes to the current node's feature learning. The calculation process first concatenates the feature vectors of the current node with those of its neighboring nodes, then inputs the concatenation result into a single-layer perceptron for nonlinear transformation, and finally normalizes it using a softmax function to obtain the attention weight values. The hidden state vectors of device nodes are obtained by weighted summing of the feature vectors of neighboring nodes according to their corresponding attention weights. This weighted summation operation maximizes the influence of important neighboring nodes on the current node. The multi-head design of the graph attention mechanism captures different types of node relationships by computing multiple sets of attention weights in parallel. The results of multi-head attention are merged through concatenation or averaging. The hidden state vectors contain the device's own physical attribute information, the association information of neighboring devices, and professional knowledge information from the access control domain. The dimension of the vector is determined by the size of the hidden layers of the graph neural network.

[0066] The process of mapping the hidden state vectors of device nodes using a fully connected layer to output the device installation priority score matrix generates the score results through linear transformation and nonlinear activation. The fully connected layer uses a linear transformation to map the hidden state vectors to the output space. This linear transformation is achieved through matrix multiplication of the weight matrix and the hidden state vectors; the parameters of the weight matrix are learned during network training. A nonlinear activation function processes the result of the linear transformation, using the sigmoid function to restrict the output values ​​to the range of 0 to 1. The output values ​​represent the installation priority score of the corresponding device. The device installation priority score matrix is ​​a two-dimensional matrix. The row indices correspond to different device installation units, the column indices correspond to the specific devices within those units, and the matrix elements represent the installation priority score of the corresponding device within that unit. Higher score values ​​indicate higher installation priority. The score calculation considers factors such as electrical dependencies, spatial constraints, and construction difficulty. The output format of the matrix facilitates subsequent construction scheduling algorithms, which determine the installation order and timing of devices based on the score matrix.

[0067] In one specific embodiment, the process of constructing the access control system device adjacency matrix based on the device physical attribute feature vector and electrical connection relationship can specifically include the following steps:

[0068] The electrical connection weights are calculated based on the voltage difference, current capacity, and signal type parameters between the access control devices, and the adjacency matrix of the access control system devices is constructed.

[0069] The knowledge graph embedding vector of the access control circuit is concatenated with the physical attribute feature vector of the device to obtain the knowledge enhancement node feature vector.

[0070] Based on the knowledge-enhanced node feature vectors and the access control system device adjacency matrix, the embedding layer parameters of the graph neural network are initialized, and a graph attention layer and a fully connected output layer are constructed.

[0071] By setting access control device installation dependency constraint weights and spatial conflict penalty coefficients on the graph attention layer, a knowledge-enhanced graph neural network for access control circuits with configured parameters is obtained.

[0072] Specifically, the process of constructing the access control system's device adjacency matrix by calculating electrical connection weights based on voltage difference, current capacity, and signal type parameters between access control devices employs a multi-parameter weighted fusion data processing method. The voltage difference parameter is calculated by comparing the operating voltage values ​​of two access control devices. The voltage difference equals the absolute value of the difference between the output voltage of device A and the input voltage of device B. A smaller voltage difference indicates better electrical compatibility between the two devices. The current capacity parameter reflects the matching degree of the devices' current carrying capacity. It is calculated by comparing the output current capacity of device A with the input current requirement of device B. When the output current capacity is greater than or equal to the input current requirement, the matching degree is 1; otherwise, the matching degree is equal to the ratio of the output current capacity to the input current requirement. The signal type parameter describes the compatibility of the signal characteristics transmitted between devices. The compatibility degree between digital signals is 1, between analog signals is 1, and between digital signals and analog signals is 0.5, indicating that an interface conversion is required. The electrical connection weight is calculated using a weighted summation method. The reciprocal of the voltage difference is multiplied by a voltage weighting coefficient, the current capacity matching degree is multiplied by a current weighting coefficient, and the signal type compatibility degree is multiplied by a signal weighting coefficient. These three products are then added together to obtain the electrical connection weight value. The weighting coefficients are set based on the importance assessment of the access control system's electrical design. The voltage matching weighting coefficient is set to 0.5 because voltage mismatch directly affects equipment operation; the current capacity weighting coefficient is set to 0.3 because insufficient current affects equipment performance; and the signal type weighting coefficient is set to 0.2 because signal incompatibility affects communication quality. The access control system's device adjacency matrix uses these electrical connection weight values ​​as matrix elements. The row and column indices of the matrix correspond to the source and target devices, respectively, and the values ​​of the matrix elements represent the electrical connection strength between the corresponding device pairs.

[0073] The process of obtaining the knowledge-enhanced node feature vector by matrix concatenating the access control circuit knowledge graph embedding vector and the device physical attribute feature vector involves the fusion of different types of information. The access control circuit knowledge graph embedding vector contains professional knowledge information about access control systems, such as expert knowledge in areas like equipment installation specifications, electrical safety distance standards, and system debugging procedures. The device physical attribute feature vector contains objective physical information about the device, such as measurable parameters like location coordinates, device type, power rating, installation height, and spatial dimensions. The matrix concatenation operation merges the two vectors into a longer vector through vector joining. The concatenation process follows a fixed order: first, all elements of the device physical attribute feature vector are placed, followed by all elements of the access control circuit knowledge graph embedding vector. The dimension of the concatenated knowledge-enhanced node feature vector is equal to the sum of the dimensions of the two original vectors. The first half of the vector retains the physical feature information of the device, while the second half incorporates professional knowledge information from the access control domain. This knowledge enhancement mechanism, through this information fusion method, allows the neural network learning process to consider both the objective attributes of the device and prior knowledge from the professional domain, improving the richness and accuracy of the feature representation. The concatenation operation is performed using an array concatenation function to ensure that the order and numerical precision of the vector elements remain unchanged during the concatenation process.

[0074] The process of initializing the embedding layer parameters of the graph neural network based on the feature vectors of knowledge-enhanced nodes and the adjacency matrix of access control system devices, and constructing the graph attention layer and fully connected output layer, adopts a layered network architecture design. The embedding layer of the graph neural network is responsible for converting the feature vectors of knowledge-enhanced nodes into the network's internal representation. The parameter initialization of the embedding layer uses the Xavier initialization method, which sets the initial value range of the parameters according to the size of the input and output dimensions, ensuring stable gradient propagation during network training. The dimension of the weight matrix of the embedding layer is determined by the dimension of the feature vectors of the knowledge-enhanced nodes and the dimension of the hidden layer; the dimension of the bias vector is equal to the dimension of the hidden layer. The graph attention layer uses a multi-head attention mechanism to learn the association weights between devices. The construction of the attention layer includes the parameter initialization of the query matrix, key matrix, and value matrix. The dimensions of these matrices are determined by the dimension of the hidden layer and the number of attention heads. The attention weights are calculated by performing a dot product operation between the query vector and the key vector, followed by normalization using the softmax function. The normalized weights are then used to perform a weighted summation of the value vectors. The fully connected output layer maps the output of the graph attention layer to the prediction result, and the weight matrix of the output layer maps the feature vectors of the hidden layer to the output space. The output dimension is set to the number of device installation priority scores according to task requirements. Residual connections and layer normalization techniques are used between network layers to improve training performance. Residual connections directly add the input to the output to avoid the gradient vanishing problem, and layer normalization standardizes the output of each layer to accelerate the convergence process.

[0075] The process of setting the access control device installation dependency constraint weights and spatial conflict penalty coefficients on the graph attention layer to obtain the parameter configuration of the access control circuit knowledge-enhancing graph neural network improves network performance through constraint optimization. The access control device installation dependency constraint weights reflect the installation sequence dependencies between devices. When device A must be installed before device B, the corresponding constraint weight is set to a large positive value; when there is no dependency between devices, the constraint weight is set to zero. The dependency constraint weights are set based on the electrical connection logic and construction process requirements of the access control system. Controller devices typically have the highest dependency constraint weights because other devices need to connect to the controller to function properly. Power supply devices also have high dependency constraint weights because other devices require power. The spatial conflict penalty coefficient is used to avoid conflicts when multiple devices are installed simultaneously in the same spatial location. The value of the penalty coefficient is calculated based on the spatial occupancy range of the devices and the degree of overlap in installation time. When the installation spaces of two devices overlap and their installation times also overlap, the corresponding penalty coefficient is set to a large negative value. The magnitude of the penalty coefficient is proportional to the spatial overlap area and the temporal overlap length. The constraint weights and penalty coefficients influence the network training process by modifying the loss function. The modified loss function consists of three parts: the original prediction loss, the dependency constraint loss, and the spatial conflict penalty loss. During network training, the backpropagation algorithm updates the network parameters based on the modified loss function, ensuring that the network's prediction results meet both accuracy requirements and the actual constraints of the access control system.

[0076] In one specific embodiment, step S4 includes:

[0077] Collect three-dimensional spatial coordinate data of the subway platform construction area, divide the construction space into grid units, and construct a space occupancy matrix;

[0078] Based on the equipment installation priority scoring matrix, determine the time window and space requirement set for the installation of access control equipment, and establish a database of space requirements for equipment installation.

[0079] Based on the space occupancy matrix and the equipment installation operation space demand database, spatiotemporal conflict detection calculations are performed to identify spatial overlap areas and time conflict periods of multi-disciplinary construction, and construction conflict identification results are obtained.

[0080] The construction conflict identification results are classified according to the degree of conflict impact to generate multi-disciplinary construction conflict early warning signals for minor, moderate and severe conflicts.

[0081] Specifically, the process of collecting three-dimensional spatial coordinate data of the subway platform construction area and dividing the construction space into grid cells to construct a spatial occupancy matrix adopts a spatial discretization data processing method. The three-dimensional spatial coordinate data is obtained through laser scanning equipment, total station measurement, and building information model (BIM) export. The coordinate data includes spatial information such as the subway platform's boundary outline, column positions, wall distribution, equipment reserved locations, and passageway areas. The coordinate system adopts the platform building coordinate system, with the platform's geometric center as the origin. The X-axis is along the train's running direction, the Y-axis is perpendicular to the train's running direction pointing towards the platform width, and the Z-axis is vertically upward representing the height direction. The construction space is divided using a regular gridding method, dividing the entire three-dimensional space into several cubic grid cells with fixed length, width, and height intervals. The size of the grid cells is set based on the minimum installation space requirements for access control equipment and the operating space needs of construction personnel, typically set to a 1-meter long, 1-meter wide, and 1-meter high cubic cell. Each grid cell is assigned a unique three-dimensional index identifier, represented by a triplet as the grid cell's sequence number in the X, Y, and Z directions. The space occupancy matrix uses a three-dimensional array structure to store the occupancy status of grid cells. The three dimensions of the array correspond to the grid indices in the X, Y, and Z directions, respectively, and the values ​​of the array elements represent the occupancy status of the corresponding grid cell. The occupancy status is encoded in binary, with 0 indicating that the grid cell is free and 1 indicating that the grid cell is occupied. The initial state of the occupancy matrix is ​​set according to the location of the station's fixed facilities such as columns, walls, and existing equipment.

[0082] The process of determining the installation time window and space requirement set for access control equipment based on the equipment installation priority scoring matrix and establishing a database of equipment installation space requirements involves the quantitative allocation of spatiotemporal resources. The scoring values ​​in the equipment installation priority scoring matrix directly affect the installation sequence; higher-scoring equipment receives an earlier installation time window, and equipment with the same score is assigned time windows according to its equipment number. The determination of the installation time window uses a sequential allocation algorithm. First, all equipment is sorted in descending order of priority score, and then consecutive time periods are allocated to each equipment according to the sorting order. The length of the time window is determined based on the equipment type and installation complexity. The installation time window for controller equipment is typically 4 hours because it requires complex wiring and debugging; for card reader equipment, it's 2 hours because installation is relatively simple; and for electromagnetic lock equipment, it's 1 hour because it mainly involves mechanical fixing operations. The space requirement set describes the grid cell range required during equipment installation. The calculation of space requirements is based on parameters such as the physical dimensions of the equipment, the operating space of the installation tools, and the safe operating distance. The physical dimensions of the equipment include the length, width, and height of the equipment itself; the operating space of the installation tools includes the required range of movement for construction personnel and tools; and the safe operating distance includes the minimum isolation distance required to prevent accidental injury. The spatial requirement set is represented as a list of grid cell indexes, containing the 3D indexes of all grid cells required during equipment installation. The equipment installation operation spatial requirement database uses a relational database structure. Database tables include fields such as equipment number, installation time window, spatial requirement set, specialty type, and dependencies. The database supports queries and statistical operations based on various conditions.

[0083] The process of detecting and identifying spatial overlap areas and time conflict periods in multi-disciplinary construction based on a space occupancy matrix and a database of equipment installation space requirements employs a spatiotemporal overlap analysis algorithm. The algorithm iterates through all equipment pair combinations in the database, checking the overlap of time windows and space requirements for each pair. Time conflict detection compares the installation time windows of two devices; if the start time of device A is less than the end time of device B, and the end time of device A is greater than the start time of device B, then the two devices are considered to have time overlap. Spatial conflict detection compares the space requirement sets of two devices; if the two sets share a common grid cell index, then the two devices are considered to have spatial overlap. The identification of multi-disciplinary construction conflicts focuses on the spatiotemporal overlap between equipment of different professional types, as equipment within the same professional field is usually handled by the same construction team, making coordination relatively easy. The identification of spatial overlap areas is achieved by calculating the intersection of the space requirement sets of two devices; the grid cell indices in the intersection constitute a description of the spatial overlap area. The identification of time conflict periods is achieved by calculating the intersection of the time windows of two devices. The start time of the intersection is equal to the maximum of the two start times, and the end time of the intersection is equal to the minimum of the two end times. The construction conflict identification results are stored in the form of a conflict record list. Each conflict record contains information such as the number of the conflicting device pair, the conflict type, the overlapping area, the conflict period, and the disciplines involved.

[0084] The process of classifying construction conflict identification results according to the degree of conflict impact and generating multi-disciplinary construction conflict early warning signals (minor, moderate, and severe conflicts) employs a quantitative assessment method for impact degree. The assessment of conflict impact degree is based on a comprehensive consideration of multiple factors, including the number of equipment involved, spatial overlap area, temporal overlap length, differences in professional types, and equipment importance level. The equipment quantity factor reflects the scale of the conflict; the more equipment involved, the wider the scope of the conflict's impact. The equipment quantity factor equals the total number of equipment involved in the conflict. The spatial overlap area factor is quantified by calculating the number of overlapping grid cells; the more overlapping grid cells, the more severe the spatial conflict. The area factor equals the product of the number of overlapping grid cells and the volume of a single grid cell. The temporal overlap length factor is quantified by calculating the duration of the overlapping time period; the longer the overlap time, the greater the sustained impact of the conflict. The time factor equals the number of hours in the overlapping time period. The professional type difference factor reflects the difficulty of coordination between different professional fields. Coordination of conflicts within the same professional field is relatively easy, while coordination of conflicts between different professional fields is more difficult. The difference factor is set according to the combination of professional types. Equipment importance level factors are set based on the criticality of equipment within the access control system. Controllers have the highest importance level, card readers have a medium level, and electromagnetic locks have the lowest. Conflict impact scores are calculated using a weighted summation method, multiplying each factor by its corresponding weight coefficient and then adding them together to obtain the total score. Classification is based on the numerical range of the impact score: conflicts with scores below a preset minor threshold are classified as minor conflicts; conflicts with scores between the minor and severe thresholds are classified as moderate conflicts; and conflicts with scores above the severe threshold are classified as severe conflicts. Multi-disciplinary construction conflict early warning signals include conflict level, conflict description, impact assessment, and recommended measures. These signals are sent to relevant construction managers and professional supervisors via message queues or notification interfaces.

[0085] In one specific embodiment, step S5 includes:

[0086] Receive the multi-disciplinary construction conflict early warning signal, analyze the conflict type and the scope of conflict impact, and construct a set of conflict constraint conditions;

[0087] A multi-objective optimization function is established based on the set of conflict constraints and the equipment installation priority scoring matrix, and weight coefficients are set for construction time, installation quality and professional waiting time.

[0088] The multi-objective optimization function is input into the particle swarm optimization algorithm for iterative solution to calculate the optimal adjustment scheme for the installation timing of the access control equipment and obtain the timing adjustment parameters.

[0089] The installation time window and professional operation sequence of the access control equipment are updated according to the timing adjustment parameters to generate the construction scheduling plan.

[0090] Specifically, the process of receiving multi-disciplinary construction conflict early warning signals and constructing a set of conflict constraints by analyzing conflict types and impact ranges employs structured data parsing and constraint modeling techniques. Multi-disciplinary construction conflict early warning signals are transmitted using structured data in JSON format. The signal content includes key information such as conflict identifiers, conflict levels, lists of involved equipment, conflict time periods, conflict spatial areas, and conflict type codes. Conflict type parsing identifies the specific nature of the conflict by interpreting the conflict type codes. Conflict types include four main categories: time conflicts, spatial conflicts, resource conflicts, and dependency conflicts. Time conflicts indicate overlapping installation time windows for multiple devices; spatial conflicts indicate overlapping installation space requirements for multiple devices; resource conflicts indicate competition for the same construction resources, such as personnel or tools; and dependency conflicts indicate violations of installation dependencies between devices. Conflict impact range parsing determines the scope of the conflict by extracting information from the list of involved equipment and conflict areas. The impact range includes directly conflicting equipment and related equipment that may be affected by cascading effects. The construction of the conflict constraint set converts the parsed conflict information into mathematical constraint expressions. The constraints describe the spatiotemporal relationship restrictions between equipment in the form of inequalities or equations. Time constraints specify the order in which conflicting devices must be installed; spatial constraints specify the minimum distance between their installation locations; and resource constraints limit the number of devices using the same resource within the same time period. The constraint set is stored in a list structure, with each constraint containing attributes such as constraint type, involved variables, constraint expression, and constraint weight.

[0091] The process of establishing a multi-objective optimization function based on a set of conflict constraints and an equipment installation priority scoring matrix, and setting weight coefficients for construction time, installation quality, and professional waiting time, involves mathematical modeling of a multi-objective optimization problem. The multi-objective optimization function comprises a linear combination of three optimization objectives: the first objective is to minimize the total construction time, expressed as the maximum completion time of all equipment installations; the second objective is to maximize the average installation quality score, expressed as the arithmetic mean of all equipment installation quality scores; and the third objective is to minimize the average waiting time between different professions, expressed as the arithmetic mean of all professional waiting times. The scores in the equipment installation priority scoring matrix serve as weighting factors in the installation quality objective function, with higher-scoring equipment receiving greater weight in the quality objective. The calculation of the construction time objective needs to consider the dependencies between equipment and resource constraints; the actual completion time of an equipment equals its start time plus the installation duration, and the total construction time equals the maximum completion time of all equipment. The calculation of the installation quality objective is based on the matching degree between the equipment installation priority score and the actual installation conditions; the highest quality score is achieved when equipment is installed in priority order and meets all technical requirements. The calculation of professional waiting time targets is based on the coordination and waiting situation between different professions. Waiting time occurs when a profession cannot start its work because it is waiting for other professions to complete their work. The weighting coefficients reflect the relative importance of different targets in construction management. Construction time is usually assigned a high weight because the construction period is a key factor for project success. Installation quality is assigned a medium weight because quality affects the long-term stable operation of the system. Professional waiting time is assigned a low weight because appropriate waiting time helps to ensure construction quality.

[0092] The process of inputting a multi-objective optimization function into a particle swarm optimization algorithm for iterative solution and calculating the optimal adjustment scheme for the installation sequence of access control equipment to obtain the timing adjustment parameters employs numerical solution techniques of swarm intelligence optimization algorithms. The particle swarm optimization algorithm searches for optimal solutions by simulating the foraging behavior of a flock of birds. Each particle in the algorithm represents a possible equipment installation sequence scheme, and the particle's position vector encodes the installation start time of all equipment. The particle swarm is initialized using a random generation method, randomly setting the position and velocity vector of each particle under the premise of satisfying basic constraints. The iterative process of the algorithm includes four steps: fitness evaluation, individual optimal update, global optimal update, and particle position update. Fitness evaluation assesses the merits of the scheme represented by each particle by calculating the value of the multi-objective optimization function; a smaller fitness value indicates a better scheme. Individual optimal update records the optimal position encountered by each particle during the search process; the individual optimal position is updated when the fitness value of the particle's current position is better than the historical best. Global optimal update records the optimal positions encountered throughout the entire particle swarm; the global optimal position is updated when the individual optimal position of any particle is better than the current global best. Particle position updates use a velocity-position update formula. The particle's new velocity equals its current velocity multiplied by its inertia weight, plus the acceleration term for its individual optimal direction plus the acceleration term for the globally optimal direction. The particle's new position equals its current position plus its new velocity. The algorithm's convergence is determined based on the degree of improvement in the globally optimal position and the iteration limit. The algorithm terminates when there is no significant improvement in the globally optimal position after multiple consecutive iterations or when the maximum number of iterations is reached. Timing adjustment parameters are extracted from the globally optimal position and include information such as the adjusted start time, adjustment duration, and adjustment priority for each device.

[0093] The process of generating a construction scheduling plan by updating the installation time windows of access control equipment and the sequence of professional tasks based on timing adjustment parameters employs scheduling plan reconstruction and formatted output technology. The equipment start time in the timing adjustment parameters is directly used to update the equipment's installation time window. The new time window start time is equal to the start time in the adjustment parameters, and the time window length remains unchanged or is fine-tuned according to the adjustment duration parameter. The update of the professional task sequence is based on reordering the adjusted time windows of all equipment within the same specialty, determining the task sequence within the specialty according to the order of start times. The construction scheduling plan is generated using a Gantt chart data structure. The horizontal axis of the Gantt chart represents the time axis, and the vertical axis represents equipment or specialty. The horizontal bars in the chart represent the installation time window for each piece of equipment. The scheduling plan includes a global schedule, a specialty coordination plan, a resource allocation table, and critical path analysis. The global schedule lists the installation time windows and critical time nodes for all equipment; the specialty coordination plan describes the handover times and coordination requirements between different specialties; the resource allocation table illustrates the usage of various resources within each time period; and the critical path analysis identifies key equipment and key activities affecting the overall project duration. The output format of the scheduling scheme supports multiple representation methods, including tabular, graphical, and textual description formats, making it easy for different types of users to understand and use.

[0094] The above describes the multi-disciplinary collaborative construction monitoring method for the subway access control system in this application embodiment. The following describes the multi-disciplinary collaborative construction monitoring system for the subway access control system in this application embodiment. Please refer to [link / reference]. Figure 2 One embodiment of the multi-disciplinary collaborative construction monitoring system for subway access control systems in this application includes:

[0095] The module is used to collect voltage levels, communication protocol types, and signal transmission delay parameters between access controllers, card readers, and electromagnetic locks, calculate the coupling strength between devices, and construct a device coupling strength matrix.

[0096] The partitioning module is used to divide the access control device into multiple device installation units based on the device coupling strength matrix.

[0097] The input module is used to construct a knowledge-enhanced graph neural network for access control circuits. It takes the physical attribute characteristics and electrical connection relationships of the device installation unit as input and outputs a device installation priority scoring matrix.

[0098] The identification module is used to monitor the multi-disciplinary work status of the access control system installation and construction, identify inter-disciplinary construction conflicts under the guidance of the equipment installation priority scoring matrix, and generate multi-disciplinary construction conflict early warning signals.

[0099] The response module is used to respond to the multi-disciplinary construction conflict early warning signal, dynamically adjust the installation sequence of the access control equipment, and obtain a construction scheduling plan.

[0100] above Figure 2 The multi-disciplinary collaborative construction monitoring system for subway access control in this embodiment of the invention is described in detail from the perspective of modular functional entities. The multi-disciplinary collaborative construction monitoring equipment for subway access control in this embodiment of the invention is described in detail from the perspective of hardware processing.

[0101] Reference Figure 3 This invention also provides a multi-disciplinary collaborative construction monitoring device for a subway access control system. This device can be a server, and its internal structure can be as follows: Figure 3 As shown, the multi-disciplinary collaborative construction monitoring equipment for the subway access control system includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor in this computer design provides computing and control capabilities. The memory of the multi-disciplinary collaborative construction monitoring equipment includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the multi-disciplinary collaborative construction monitoring equipment is used to store the data corresponding to this embodiment. The network interface of the multi-disciplinary collaborative construction monitoring equipment is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0102] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the multi-disciplinary collaborative construction monitoring equipment for the subway access control system to which the present invention is applied.

[0103] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the multi-disciplinary collaborative construction monitoring method for the subway access control system.

[0104] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0105] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a multi-disciplinary collaborative construction monitoring device for a subway access control system (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0106] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-disciplinary collaborative construction monitoring method for a subway access control system, characterized in that, The method includes: Step S1: Collect the voltage level, communication protocol type, and signal transmission delay parameters between the access controller, card reader, and electromagnetic lock; calculate the coupling strength between the devices; and construct the device coupling strength matrix. Step S2: Based on the device coupling strength matrix, divide the access control device into multiple device installation units; Step S3: Construct an access control circuit knowledge-enhanced graph neural network. Input the physical attribute features and electrical connection relationships of the equipment installation units, and output a device installation priority scoring matrix. This includes: extracting the location coordinates, device type code, power level, installation height, and reserved space size parameters of each access control device in the multiple equipment installation units to construct a device physical attribute feature vector; converting the standard installation specifications, electrical safety distance requirements, and commissioning and acceptance standards of the access control devices into vector representations to construct an access control circuit knowledge graph embedding vector; constructing an access control system device adjacency matrix based on the device physical attribute feature vectors and electrical connection relationships, and combining this matrix with the access control circuit knowledge graph embedding vector. The method for constructing a knowledge-enhanced graph neural network for access control circuits includes: calculating electrical connection weights based on voltage differences, current capacities, and signal type parameters between access control devices, and constructing the device adjacency matrix of the access control system; performing a matrix concatenation operation between the embedding vector of the access control circuit knowledge graph and the physical attribute feature vector of the device to obtain the feature vector of the knowledge-enhanced node; initializing the embedding layer parameters of the graph neural network based on the feature vector of the knowledge-enhanced node and the device adjacency matrix of the access control system, and constructing a graph attention layer and a fully connected output layer; setting the access control device installation dependency constraint weights and spatial conflict penalty coefficients on the graph attention layer to obtain the parameter-configured knowledge-enhanced graph neural network for access control circuits. The physical attribute feature vector of the device is input into the knowledge-enhanced graph neural network of the access control circuit for graph attention mechanism calculation to obtain the hidden state vector of the device node; the hidden state vector of the device node is processed by a fully connected layer to output the device installation priority scoring matrix. Step S4: Monitor the multi-disciplinary work status of the access control system installation and construction, identify inter-disciplinary construction conflicts under the guidance of the equipment installation priority scoring matrix, and generate a multi-disciplinary construction conflict early warning signal; Step S5: Respond to the multi-disciplinary construction conflict early warning signal, dynamically adjust the installation sequence of the access control equipment, and obtain a construction scheduling plan.

2. The multi-disciplinary collaborative construction monitoring method for subway access control systems according to claim 1, characterized in that, Step S1 includes: Establish a set of access control device nodes and a set of electrical connection edges, and construct a directed graph of the electrical topology of the access control system; The device coupling strength is obtained by weighting the voltage matching degree, protocol compatibility and reciprocal of transmission delay between devices in the directed graph of the electrical topology. The device coupling strength values ​​are filled into the corresponding positions in the matrix according to the device number order to construct the device coupling strength matrix; The device coupling strength matrix is ​​normalized to obtain the device coupling strength matrix.

3. The multi-disciplinary collaborative construction monitoring method for subway access control systems according to claim 1, characterized in that, Step S2 includes: Construct the access control device degree matrix based on the device coupling strength matrix, and calculate the Laplace matrix of the access control system. The access control system's Laplace matrix is ​​subjected to eigenvalue decomposition, and eigenvalues ​​below the electrical coupling strength threshold are filtered out to obtain the access control device group feature vector. The spatial constraint feature representation is obtained by performing a weighted summation operation between the grouped feature vectors of the access control device and the installation spatial coordinates of the access control device. Based on the spatial constraint feature representation, an improved k-means clustering calculation is performed to limit the number of devices in the same device installation unit to no more than a preset access control component capacity threshold, thereby obtaining the multiple device installation units.

4. The multi-disciplinary collaborative construction monitoring method for subway access control systems according to claim 1, characterized in that, Step S4 includes: Collect three-dimensional spatial coordinate data of the subway platform construction area, divide the construction space into grid units, and construct a space occupancy matrix; Based on the equipment installation priority scoring matrix, determine the time window and space requirement set for the installation of access control equipment, and establish a database of space requirements for equipment installation. Based on the space occupancy matrix and the equipment installation operation space demand database, spatiotemporal conflict detection calculations are performed to identify spatial overlap areas and time conflict periods of multi-disciplinary construction, and construction conflict identification results are obtained. The construction conflict identification results are classified according to the degree of conflict impact to generate multi-disciplinary construction conflict early warning signals for minor, moderate and severe conflicts.

5. The multi-disciplinary collaborative construction monitoring method for subway access control systems according to claim 1, characterized in that, Step S5 includes: Receive the multi-disciplinary construction conflict early warning signal, analyze the conflict type and the scope of conflict impact, and construct a set of conflict constraint conditions; A multi-objective optimization function is established based on the set of conflict constraints and the equipment installation priority scoring matrix, and weight coefficients are set for construction time, installation quality and professional waiting time. The multi-objective optimization function is input into the particle swarm optimization algorithm for iterative solution to calculate the optimal adjustment scheme for the installation timing of the access control equipment and obtain the timing adjustment parameters. The installation time window and professional operation sequence of the access control equipment are updated according to the timing adjustment parameters to generate the construction scheduling plan.

6. A multi-disciplinary collaborative construction monitoring system for subway access control, characterized in that, For implementing the multi-disciplinary collaborative construction monitoring method for a subway access control system as described in any one of claims 1-5, the multi-disciplinary collaborative construction monitoring system for the subway access control system comprises: The module is used to collect voltage levels, communication protocol types, and signal transmission delay parameters between access controllers, card readers, and electromagnetic locks, calculate the coupling strength between devices, and construct a device coupling strength matrix. The partitioning module is used to divide the access control device into multiple device installation units based on the device coupling strength matrix. The input module is used to construct a knowledge-enhanced graph neural network for access control circuits. It takes the physical attribute characteristics and electrical connection relationships of the device installation unit as input and outputs a device installation priority scoring matrix. The identification module is used to monitor the multi-disciplinary work status of the access control system installation and construction, identify inter-disciplinary construction conflicts under the guidance of the equipment installation priority scoring matrix, and generate multi-disciplinary construction conflict early warning signals. The response module is used to respond to the multi-disciplinary construction conflict early warning signal, dynamically adjust the installation sequence of the access control equipment, and obtain a construction scheduling plan.

7. A multi-disciplinary collaborative construction monitoring device for a subway access control system, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the multi-disciplinary collaborative construction monitoring method for the subway access control system as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the multi-disciplinary collaborative construction monitoring method for the subway access control system as described in any one of claims 1 to 5.

Images