An exhibition hall intelligent management system based on big data analysis

Abstract

This invention discloses an intelligent management system for exhibition halls based on big data analysis, belonging to the field of intelligent exhibition hall management technology. It includes a spatial topology module, a status acquisition module, a disturbance analysis module, a control decision module, and an instruction distribution module. The spatial topology module acquires architectural information model data of the exhibition hall and extracts the spatial boundaries and channel connections of each exhibition area, generating a global spatial topology network for the entire exhibition hall. The status acquisition module connects to multi-source sensing devices to collect real-time data on instantaneous visitor flow, ambient temperature and humidity, and equipment operating power, forming a time-series matrix of the exhibition area's operating status. The disturbance analysis module maps the operating status matrix to each node, calculates the node-related disturbance values, and superimposes and propagates them along the network, generating a global disturbance field distribution map of the exhibition hall. The control decision module identifies the extreme value regions of the disturbance field gradient and uses them as control anchor points, tracing back along the superimposed propagation path to determine the set of exhibition areas to be controlled and the control response coefficients of each area.

Description

Technical Field

[0001] This invention relates to the field of intelligent management technology for exhibition halls, specifically an intelligent management system for exhibition halls based on big data analysis. Background Technology

[0002] As a public space with high foot traffic and complex functional zoning, exhibition halls are affected by multiple factors, including fluctuations in visitor flow, changes in environmental parameters, and variations in equipment load. Existing exhibition hall management solutions typically rely on independently deploying sensors in each exhibition area to monitor thresholds for visitor density, temperature, humidity, or equipment power in individual areas. Local point-to-point control actions are triggered when any indicator exceeds a preset range. This approach only considers the isolated state of each exhibition area, ignoring the spatial coupling relationships formed between them through connecting passageways. When a disturbance occurs in one area, its impact spreads to adjacent areas via visitor flow, air circulation, or equipment load distribution. Existing technologies do not treat these areas as an interconnected dynamic system at the spatial topology level, causing control actions to lag behind the actual progress of disturbance propagation. A deficiency in existing technologies is the lack of means to incorporate the topological structure of the exhibition hall's physical space into operational status analysis, making it impossible to quantify the degree of influence of changes in one exhibition area on adjacent areas. The multidimensional time-series data collected by sensors is processed in a fragmented manner, failing to effectively extract the similarity and transmission relationship of operational status fluctuations between different zones, thus making it impossible to characterize the propagation intensity and diffusion path of disturbances in the spatial network. Another drawback is that, after identifying abnormal states in local areas, it is impossible to trace the source of the disturbance and sort out the influence transmission chain between upstream and downstream zones, and the selection of control targets and the allocation of control intensity lack a global basis. Typically, indiscriminate equalization regulation is applied to all related zones, or excessive intervention is only applied to the zone where the disturbance is first exposed. This easily leads to equipment response oscillations and energy waste, and it is also difficult to completely suppress disturbance factors originating from upstream.

[0003] To manage the exhibition areas of the exhibition hall as a whole intelligently, it is necessary to solve how to organically integrate the spatial connection relationship between the exhibition areas with the multi-source operational status time series data to generate a global situation map that can reflect the spatial distribution and propagation characteristics of disturbances. On this basis, it is necessary to solve how to locate the key areas of disturbance convergence from the global situation and decompose the optimal control path and control intensity in reverse. Summary of the Invention

[0004] This invention provides an intelligent management system for exhibition halls based on big data analysis. The purpose is to integrate and analyze the topological constraints of the exhibition hall's architectural space with multi-source sensor time-series data. By calculating the correlation disturbance values ​​of the operating states between nodes and superimposing and transmitting them along the spatial topological network, a global disturbance field distribution map that can characterize the spatial distribution and propagation characteristics of disturbances is generated. Furthermore, by identifying the extreme value regions of the disturbance field gradient and tracing back along the transmission path, the exhibition area to be controlled and the corresponding control response coefficient are accurately determined, thereby achieving linkage control of equipment parameters that match the disturbance transmission structure.

[0005] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides an intelligent management system for exhibition halls based on big data analysis, including a spatial topology module, a status acquisition module, a disturbance analysis module, a control decision module, and an instruction distribution module. The spatial topology module acquires architectural information model data of the exhibition hall and extracts the spatial boundaries and channel connection relationships of each exhibition area to generate a global spatial topology network for the entire exhibition hall. In a preferred embodiment, when generating the global spatial topology network, the enclosed areas of each exhibition area's spatial boundaries are used as topology nodes. The geometric center coordinates and access doorway coordinates of each topology node are extracted. The connection paths between adjacent topology nodes are determined based on channel connection relationships, and the path length and path width are used as edge weight attributes. The geometric center coordinates are used as node location identifiers, and the access doorway coordinates are used as port connection identifiers to construct the global spatial topology network for the entire exhibition hall. This accurately characterizes the constraint relationship between the exhibition hall's spatial structure and personnel flow, providing reliable spatial constraints for disturbance transmission analysis.

[0006] The status acquisition module connects to multi-source sensing devices deployed in each exhibition area to collect real-time data on instantaneous pedestrian flow, ambient temperature and humidity, and equipment operating power in each exhibition area, forming a time-series matrix of the exhibition area's operating status, providing multi-dimensional real-time status data for subsequent analysis.

[0007] The disturbance analysis module uses the entire exhibition hall's spatial topology network as a constraint basis, mapping the exhibition area's operational state time series matrix to each node in the spatial topology network. It calculates the associated disturbance value of each node's operational state relative to its neighboring nodes and generates a global disturbance field distribution map of the exhibition hall based on the superposition and propagation path of each node's associated disturbance value on the spatial topology network. When calculating the associated disturbance value, a local neighborhood set is formed by the target topology node and all its neighboring topology nodes. The operational state time series sub-matrices of each node within the local neighborhood set within the same time window are extracted. The dynamic time warping distance between the operational state time series sub-matrices of the target topology node and the operational state time series sub-matrices of each neighboring topology node is calculated. The weighted sum of all dynamic time warping distances is used as the associated disturbance value of the target topology node, where the weighting coefficient is the normalized value of the edge weight attribute between the target topology node and its corresponding neighboring topology nodes. In the aforementioned dynamic time warping distance calculation process, the state dimensions of each dimension in the operational state time-series submatrix are normalized, and the normalized values ​​of each dimension are weighted and corrected using the exhibition area and environmental capacity coefficient to eliminate measurement bias caused by differences in the scale of different exhibition areas, ensuring that the disturbance values ​​accurately reflect the degree of abnormal fluctuations in pedestrian flow and the environment. When generating the global disturbance field distribution map of the exhibition hall, the associated disturbance values ​​of each node are used as the initial disturbance source intensity, and the connecting edges of the spatial topology network are used as the transmission channels. The attenuation coefficient of the disturbance along the transmission channel is set to be inversely proportional to the edge weight attribute. Starting from each disturbance source, the disturbance intensity is propagated step by step to the outer nodes along each transmission channel. The nodes that are propagated to accumulate the received disturbance intensity with their own initial disturbance source intensity to generate the cumulative disturbance intensity of each node. Based on the cumulative disturbance intensity of each node and the geometric coordinates of the spatial topology network, spatial interpolation is performed to generate the global disturbance field distribution map of the exhibition hall. Preferably, the attenuation coefficient is also related to the ratio of the spatial boundary areas of adjacent nodes. When the disturbance propagates from a large area to a small area, the attenuation coefficient is scaled by the square root of the area ratio, thereby more realistically simulating the transmission attenuation characteristics of the disturbance at different spatial scales and improving the accuracy of the global disturbance field simulation.

[0008] The control decision module identifies the extreme value regions of the disturbance field gradient based on the global disturbance field distribution map of the exhibition hall. Using these extreme value regions as control anchors, it backtracks along the reverse direction of the superimposed transmission path to determine the set of exhibition areas to be controlled and the control response coefficients of each area. The process of identifying the extreme value regions of the disturbance field gradient includes: calculating the partial derivatives of the disturbance intensity along the horizontal and vertical directions for each grid point in the global disturbance field distribution map of the exhibition hall; using the square root of the sum of the squares of the partial derivatives as the gradient magnitude of that grid point; performing neighborhood aggregation on grid points with gradient magnitudes greater than a preset threshold to form a candidate region set; extracting the location of the maximum disturbance intensity of each grid point in the candidate region set as the extreme value point; and marking the spatial range containing the extreme value point and with the gradient direction pointing towards that extreme value point as the extreme value region of the disturbance field gradient, thereby automatically locating the area with the most severe disturbance. When determining the set of exhibition areas to be regulated and the regulation response coefficients, the regulation anchor point is used as the starting point for backtracking. The upstream nodes are traversed sequentially in the opposite direction of each transmission channel in the superimposed transmission path until the boundary node of the spatial topology network is reached or the cumulative backtracking steps reach a predetermined threshold. All traversed nodes are then included in the set of exhibition areas to be regulated. The disturbance transmission contribution rate of each node is determined based on the ratio of its cumulative disturbance intensity in the global disturbance field distribution map of the exhibition hall to the node's associated disturbance value. This contribution rate is then used as the regulation response coefficient for the corresponding exhibition area. The resulting regulation response coefficients quantify the contribution of each exhibition area to the global disturbance extreme value, providing a basis for differentiated regulation.

[0009] The instruction distribution module matches the equipment execution parameters of each control zone according to the control response coefficient, and converts the equipment execution parameters into a control instruction sequence and sends it to the terminal execution device of the corresponding zone. When matching equipment execution parameters, the module obtains the equipment type identifier and current operating parameter value of each control zone, and queries the preset adjustable range table of equipment parameters and the parameter response mapping table based on the equipment type identifier. It matches the control response coefficient with the response coefficient level in the parameter response mapping table to determine the target adjustment level for each control zone. Based on the target adjustment level and the adjustable range table of equipment parameters, it calculates the adjustment amount of the operating parameters for each execution device in each control zone, and adds the current operating parameter value and the adjustment amount to obtain the equipment execution parameter. Preferably, when the same control zone contains multiple execution devices, the total control response coefficient for the zone is allocated according to the rated power ratio of each execution device, and adjustments exceeding the adjustable range of equipment parameters are truncated step by step according to the device priority, thereby ensuring that the equipment operates within a safe range and maximizing the control execution effect, avoiding over-adjustment or equipment failure.

[0010] Through the collaborative work of the above modules, this invention closely integrates the physical space topology of the exhibition hall with multi-source sensor time-series data to construct a closed-loop control system for disturbance perception, transmission analysis, and reverse decision-making. It can intelligently identify the source of disturbance and generate refined equipment control strategies, effectively suppressing the spread of disturbances such as crowd congestion, environmental imbalance, and equipment malfunction within the exhibition hall, and significantly improving the initiative and accuracy of exhibition hall management.

[0011] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The spatial boundaries and passageway connections of each exhibition area in the exhibition hall are extracted into a global spatial topology network containing node location identifiers, port connection identifiers, and edge weight attributes. This network serves as the constraint basis for state analysis. Based on this, using the local neighborhoods formed by the target topology node and its adjacent nodes as processing units, dynamic time-warped distance calculations are performed on the temporal submatrices of the operating states of different nodes. The distance values ​​are then weighted and accumulated according to the edge weights to obtain the associated disturbance value for each node. This approach incorporates factors such as structural distances and passage widths within the exhibition hall space into the measurement of state differences, ensuring that the associated disturbance value not only reflects the magnitude of temporal morphological differences but also the differentiated impact of varying spatial connection tightness between exhibition areas on disturbance propagation. Subsequently, using the associated disturbance value of each node as the initial disturbance source intensity, the connecting edges as propagation channels, and setting an attenuation coefficient inversely proportional to the edge weight attribute, the disturbance propagates step-by-step along the network and accumulates to generate the cumulative disturbance intensity of each node. Combined with spatial coordinate interpolation, a global disturbance field distribution map of the exhibition hall is formed. This distribution map extends the disturbance transmission relationship between exhibition areas from local comparisons to a continuous field expression at the entire network level. It can characterize the spatial convergence areas and diffusion paths of disturbances, avoiding the loss of spatial propagation characteristics of disturbances caused by judging based on single-point thresholds alone. By calculating the gradient magnitude of each grid point in the global disturbance field distribution map, and after neighborhood aggregation and extreme value localization, the extreme value region of the disturbance field gradient is identified as the control anchor point. Starting from the control anchor point, the upstream nodes are traversed step by step in the opposite direction of the disturbance superposition transmission path until the boundary node is reached or the predetermined number of steps is reached. The traversed nodes are then included in the set of exhibition areas to be controlled. At the same time, the disturbance transmission contribution rate is determined according to the ratio of the cumulative disturbance intensity of each node to the associated disturbance value, and this is used as the control response coefficient of the corresponding exhibition area to be controlled. This reverse backtracking method starts from the gradient extreme value region of the disturbance field and traces the source of the disturbance transmission along the transmission link to the upstream exhibition area. This makes the control range no longer an isolated abnormal exhibition area or a uniform range of the entire venue, but rather covers the source of influence and key nodes along the real disturbance transmission chain. The control response coefficient is allocated according to the actual contribution ratio of each node in the entire transmission link, so that the control intensity matches the degree of influence of each zone on the convergence of disturbances. This can cut off or weaken the transmission of disturbances without blindly expanding the control range, thereby reducing repeated frequency adjustments and ineffective energy consumption of equipment. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0013] Figure 1 This is a schematic diagram of the structure of an intelligent management system for exhibition halls based on big data analysis; Figure 2 This is a flowchart of the calculation of topological node association perturbation values ​​based on dynamic time warping; Figure 3 This is a distribution map of the topological node association disturbance values ​​in the exhibition area; Figure 4 This is a schematic diagram showing the relationship between the control response coefficient and the target adjustment level for different equipment types. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] See Figure 1This invention provides an intelligent management system for exhibition halls based on big data analysis, comprising: a spatial topology module, a status acquisition module, a disturbance analysis module, a control decision module, and an instruction distribution module. The spatial topology module acquires architectural information model data of the exhibition hall and extracts the spatial boundaries and channel connections of each exhibition area, generating a global spatial topology network for the entire exhibition hall. The status acquisition module connects to multi-source sensors deployed in each exhibition area, collecting real-time data on instantaneous visitor flow, ambient temperature and humidity, and equipment operating power, forming a time-series matrix of the exhibition area's operating status. The disturbance analysis module uses the global spatial topology network as a constraint basis to map the time-series matrix of the exhibition area's operating status onto each node of the spatial topology network, calculates the associated disturbance value of each node's operating status relative to adjacent nodes, and generates a global disturbance field distribution map of the exhibition hall based on the superposition and propagation path of the associated disturbance values ​​of each node on the spatial topology network. The control decision module identifies the extreme value regions of the disturbance field gradient based on the global disturbance field distribution map of the exhibition hall. Using these extreme value regions as control anchor points, it backtracks along the reverse direction of the superimposed transmission path to determine the set of exhibition areas to be controlled and the control response coefficients of each area. The instruction distribution module matches the equipment execution parameters of each area to be controlled based on the control response coefficients and converts these parameters into a sequence of control instructions, which is then sent to the terminal execution devices of the corresponding exhibition areas.

[0016] Example 1: In practice, the BIM data for the exhibition hall is obtained as follows: Building component information, spatial attributes, and connection relationships are analyzed from the BIM files generated during the design or operation phase of the exhibition hall. The spatial boundaries of each exhibition area are identified from the BIM data. These boundaries are defined by the geometric information of the walls, partitions, floors, and ceilings enclosing the exhibition areas. By extracting the inner surfaces of all building components enclosing the exhibition areas, closed polygonal boundaries are formed; these polygonal boundaries represent the spatial boundaries of the exhibition areas. Simultaneously, the passageway connections between the exhibition areas are extracted from the BIM data. These connections include passable connections formed between exhibition areas through doorways, corridors, or open interfaces.

[0017] Using the areas enclosed by the spatial boundaries of each exhibition area as topological nodes, the geometric center coordinates and access doorway coordinates of each topological node are extracted. One method for extracting the geometric center coordinates is to obtain the coordinates of each vertex of the spatial boundary polygon enclosing the exhibition area, calculate the arithmetic mean of all vertex coordinates, and then obtain the geometric center coordinates. Let the spatial boundary polygon have... The vertex, the first The planar coordinates of the vertices are Then the geometric center coordinates of the topological node Calculate using the following formula: in, This represents the total number of vertices of the polygon enclosing the exhibition area. Indicates the index number of the vertex. Indicates the first The x-coordinates of each vertex. Indicates the first The ordinate values ​​of each vertex are extracted from the world coordinate system in the architectural information model data of the exhibition hall.

[0018] The method for extracting the location coordinates of passageway openings is as follows: identify the line segments belonging to the doorway components on the spatial boundary polygon, and take the midpoint coordinates of the doorway component line segments as the location coordinates of the passageway openings. If there are multiple passageway openings in an exhibition area, extract the midpoint coordinates of each passageway opening separately to form a set of passageway opening location coordinates.

[0019] The connectivity paths between adjacent topological nodes are determined based on the channel connection relationships. The path length and path width of the connectivity paths are extracted and used as edge weight attributes of the connecting edges. Specifically, the connectivity paths are determined as follows: if two exhibition areas are directly connected through a doorway, the width of the doorway is extracted as the path width, and the straight-line distance between the geometric centers of the two exhibition areas is used as the path length; if two exhibition areas are connected through a corridor, the length of the corridor's centerline is extracted as the path length, and the minimum width of the corridor is used as the path width; if two exhibition areas are connected through an open interface, the opening width of the open interface is extracted as the path width, and the straight-line distance between the geometric centers of the two exhibition areas is used as the path length.

[0020] A global spatial topology network for the exhibition hall is constructed, using the geometric center coordinates as node location identifiers, the access doorway coordinates as port connection identifiers, and edge weight attributes as connection edge attributes. The data structure of the global spatial topology network is stored using a node table and an edge table. Each record in the node table corresponds to a topology node, containing the exhibition area identifier, geometric center coordinates, and access doorway coordinates. Each record in the edge table corresponds to a connection edge, containing the start node identifier, end node identifier, and the path length and path width from the edge weight attributes. This generates an undirected weighted graph representation of the global spatial topology network for the exhibition hall.

[0021] Example 2: In specific implementation, please refer to Figure 2 Taking any topological node in the spatial topological network as the target topological node, query all topological nodes in the spatial topological network that have a connection edge with the target topological node, and take the queried topological nodes as adjacent topological nodes. The target topological node and all adjacent topological nodes together form a local neighborhood set.

[0022] From the exhibition area operation status time-series matrix generated by the status acquisition module, the operation status time-series sub-matrices of each node within the local neighborhood set within the same time window are extracted. The length of the time window is determined by a preset observation duration parameter, which is pre-configured to a fixed duration, such as 300 seconds. The operation status time-series sub-matrix of each node is a multi-dimensional vector sequence. Each time point in the multi-dimensional vector sequence corresponds to a state vector. The dimensions of the state vector include instantaneous pedestrian flow, ambient temperature, ambient humidity, and equipment operating power. Instantaneous pedestrian flow is the number of people in the exhibition area at the current moment; ambient temperature is the temperature value collected by the temperature sensor in the exhibition area; ambient humidity is the relative humidity value collected by the humidity sensor in the exhibition area; and equipment operating power is the total power of all executing devices in the exhibition area at the current moment.

[0023] For a target topology node and any adjacent topology node, calculate the dynamic time warping distance between the time sequence submatrix of the target topology node and the time sequence submatrix of the adjacent topology node. During the calculation of the dynamic time warping distance, the state dimensions of each dimension in the time sequence submatrix are normalized. The normalization process uses a min-max normalization method. For the instantaneous visitor flow dimension, obtain the maximum and minimum instantaneous visitor flow values ​​for all exhibition areas within the same time window. Subtract the minimum value from each value in the instantaneous visitor flow sequence of the target topology node and adjacent topology nodes, and then divide by the difference between the maximum and minimum values ​​to obtain the normalized instantaneous visitor flow sequence. The same method is used to obtain the normalized ambient temperature sequence for the ambient temperature dimension; the normalized ambient humidity sequence for the ambient humidity dimension; and the normalized equipment operating power sequence for the equipment operating power dimension.

[0024] The normalized values ​​for each dimension are then weighted and corrected using the exhibition area and environmental capacity coefficient. The exhibition area is calculated from the spatial boundary, and the environmental capacity coefficient is the ratio of the maximum capacity of the pre-set exhibition area to the average total capacity of the exhibition hall, reflecting the differences in the exhibition area's capacity to accommodate people. The weighting correction is as follows: for each value in the normalized instantaneous visitor flow sequence of the target topological node, multiply by the product of the target topological node's exhibition area and environmental capacity coefficient as a weighting coefficient to obtain the weighted and corrected instantaneous visitor flow sequence of the target topological node; for each value in the normalized instantaneous visitor flow sequence of adjacent topological nodes, multiply by the product of the adjacent topological node's exhibition area and environmental capacity coefficient as a weighting coefficient to obtain the weighted and corrected instantaneous visitor flow sequence of the adjacent topological nodes. The dimensions of ambient temperature, ambient humidity, and equipment operating power are also weighted and corrected according to their respective exhibition area and environmental capacity coefficient. The weighted and corrected multidimensional time series is used to perform subsequent dynamic time warping distance calculations.

[0025] The calculation process for dynamic time warping distance is as follows: A distance matrix is ​​constructed, where each row corresponds to the time index of the weighted corrected time series of the target topological node, and each column corresponds to the time index of the weighted corrected time series of adjacent topological nodes. Each element in the distance matrix represents the Euclidean distance between the multidimensional state vectors corresponding to two time points. The Euclidean distance between the multidimensional state vectors is the square root of the sum of the squares of the differences in each dimension of the two vectors. Following the recursive rule of dynamic time warping, starting from the initial point (1,1) of the distance matrix, the distance is calculated towards the endpoint (…). , The cumulative minimum distance is calculated in each direction. The recursive path allows movement in three directions: right, up, and upper right diagonal. After the recursion is complete, the cumulative distance value at the endpoint of the distance matrix is ​​the dynamic time warped distance between the weighted and corrected running state time sequence submatrices of the two topological nodes.

[0026] After obtaining the dynamic time-warped distance between the target topological node and each of its neighboring topological nodes, the associated perturbation value of the target topological node is calculated. The calculated dynamic time-warped distances between the target topological node and all its neighboring topological nodes are then weighted and summed according to corresponding weighting coefficients. These weighting coefficients are the normalized values ​​of the edge weight attributes between the target topological node and its corresponding neighboring topological nodes. The edge weight attributes include path length and path width. The normalized value of the edge weight attribute is calculated by taking the sum of the reciprocal of the path width and the path length as a comprehensive weight, and then normalizing this comprehensive weight between the target topological node and all its neighboring topological nodes so that the sum of all weighting coefficients is 1.

[0027] Associated perturbation value of target topology node Calculate using the following formula: in, Represents the target topology node The associated perturbation value, Represents the target topology node The total number of adjacent topological nodes, Indicates the index number of adjacent topological nodes. Represents the target topology node With the The weighted coefficients between adjacent topological nodes The value of is between 0 and 1, and satisfies , Represents the target topology node With the The dynamic time warping distance between the weighted, corrected running state time series submatrices of adjacent topological nodes. The associated perturbation value calculated according to the above formula. As the associated perturbation value of the target topology node.

[0028] By taking each topology node in the spatial topology network as the target topology node in turn, the above steps are repeated to obtain the associated perturbation value of each topology node.

[0029] See Figure 3 In the figure, the horizontal axis represents the node number of the spatial topology network, ranging from 0 to 500, and the vertical axis represents the associated perturbation value of the corresponding node. The data points are presented as scattered points, and the curve shows a monotonically increasing trend from the lower left to the upper right. In the initial stage (node ​​numbers from 0 to approximately 100), the associated perturbation value is low and changes gradually, mainly concentrated between 0 and 1, indicating that the dynamic time regularization distance between the operating state of the topology node and its neighboring nodes is small, and the degree of associated perturbation is low. In the middle stage (node ​​numbers from approximately 100 to 400), the associated perturbation value gradually increases, showing a steady upward trend, with most values ​​in the range of 1 to 2.5, reflecting a moderate degree of perturbation difference between the operating state of some topology nodes and their neighboring nodes. In the final stage (node ​​numbers from approximately 400 to 500), the associated perturbation value increases significantly, with many points showing a sharp increase in perturbation values ​​above 3, and some nodes exceeding 7, indicating a large dynamic difference between the operating state of these topology nodes and their neighboring nodes, representing strong associated perturbation.

[0030] Example 3: In practice, the disturbance analysis module uses the associated disturbance values ​​of each topology node as the initial disturbance source strength. Each topology node is assigned an initial disturbance source strength value, which is equal to the associated disturbance value of the corresponding topology node.

[0031] The connecting edges of the spatial topology network are used as disturbance propagation channels. For each connecting edge, an attenuation coefficient is set, which is inversely proportional to the edge weight attribute. The edge weight attribute includes path length and path width. The attenuation coefficient is calculated by taking the reciprocal of the path length, multiplying it by the path width, taking the reciprocal again, and then multiplying it by a preset base attenuation constant. The base attenuation constant is preset to a fixed value of 1.0, representing the baseline attenuation level when the disturbance propagates in an ideal channel of unit length and unit width.

[0032] The attenuation coefficient is also related to the ratio of the spatial boundary areas of adjacent nodes. For two topological nodes connected by a connecting edge, the ratio of the spatial boundary areas of the two topological nodes is calculated. The spatial boundary area is calculated from the area of ​​the polygon representing the spatial boundary of the exhibition area. When a disturbance propagates from a large-area exhibition area to a small-area exhibition area, the attenuation coefficient is scaled by the square root of the area ratio. Specifically, the originally calculated attenuation coefficient is multiplied by a scaling factor, which is the square root of the ratio of the spatial boundary areas of the two topological nodes, calculated by dividing the area of ​​the large-area exhibition area by the area of ​​the small-area exhibition area and then taking the square root. When a disturbance propagates from a small-area exhibition area to a large-area exhibition area, no scaling is performed, and the attenuation coefficient retains its original calculated value.

[0033] The disturbance propagation process employs a breadth-first, step-by-step propagation method. Starting from each topological node acting as a disturbance source, the disturbance intensity propagates along the disturbance propagation path to the outer topological nodes directly adjacent to the disturbance source node. The disturbance intensity propagated to the outer topological node is the disturbance source intensity of the disturbance source node multiplied by the attenuation coefficient on the corresponding connecting edge. The outer topological node to which the disturbance is propagated accumulates the received disturbance intensity with its own initial disturbance source intensity; the accumulated value is used as the temporary cumulative disturbance intensity of that outer topological node.

[0034] After all disturbance sources have propagated to their respective first-layer adjacent topological nodes, each outer-layer topological node becomes a new starting point for disturbance propagation. The propagation intensity of the new starting point is the initial disturbance source intensity of each topological node; disturbance intensities received during propagation in the previous layer do not participate in outward propagation. Propagation continues from the new starting point to outermost topological nodes, following the same propagation rules as the first layer. This step-by-step propagation process continues until all topological nodes in the spatial topological network have been traversed, or the propagation intensity of all topological nodes is lower than a preset propagation cutoff threshold. The propagation cutoff threshold is a pre-set minimum disturbance intensity value; when the propagation intensity falls below the threshold, propagation to subsequent topological nodes stops.

[0035] After propagation through each level, the final cumulative perturbation strength is obtained at each topology node. The cumulative perturbation strength is the sum of the initial perturbation source strength of the topology node itself and the strengths of all perturbations propagating from all other perturbation sources to the topology node. (The cumulative perturbation strength of the topology node is then calculated.) Calculate using the following formula: in, Represents topology nodes The cumulative disturbance intensity, Represents topology nodes Its own initial disturbance source strength. Represents the spatial topology network excluding topology nodes. The total number of other topology nodes besides Indicates excluding topology nodes The index numbers of other topological nodes, Indicates from topology node To topology node The number of effective transmission paths, The index number representing the effective propagation path. Represents topology nodes The initial disturbance source strength. Indicates from topology node To topology node The The total number of connected edges traversed by each valid propagation path. The index number representing the number of connected edge segments. Indicates from topology node To topology node The The first effective transmission path The basic decay coefficient on segment-connecting edges, which is inversely proportional to the edge weight attribute. Indicates from topology node To topology node The The first effective transmission path The scaling factor associated with the area ratio on the segment connecting edges, when the disturbance propagates from a large area to a small area. The value is taken as the square root of the area ratio, when the disturbance propagates from a small area to a large area. The value is 1.

[0036] After obtaining the cumulative perturbation intensity of all topological nodes, spatial interpolation is performed based on the cumulative perturbation intensity of each topological node and the geometric coordinates of the spatial topology network. The spatial interpolation uses the Kriging interpolation algorithm, taking the geometric center coordinates of each topological node as the known sampling point location and the cumulative perturbation intensity of each topological node as the attribute value of the known sampling point. A spatial interpolation model is obtained by fitting a variogram function, generating estimated perturbation intensity values ​​at regular grid points within the exhibition hall's planar space at a preset grid resolution. The grid resolution is preset to a 0.5m x 0.5m square grid cell. The estimated perturbation intensity values ​​of all grid points together constitute a global perturbation field distribution map of the exhibition hall.

[0037] Example 4: In practical implementation, the control and decision-making module acquires a global disturbance field distribution map of the exhibition hall. This map is a two-dimensional regular grid, with each grid point having a disturbance intensity value. For each grid point in the global disturbance field distribution map, the partial derivatives of the disturbance intensity along the horizontal and vertical directions are calculated. The partial derivatives are calculated using the central difference method: for each grid point... The horizontal partial derivative is the difference between the disturbance intensity of the grid point to the right and the disturbance intensity of the grid point to the left, divided by twice the grid spacing. The vertical partial derivative is the difference between the disturbance intensity of the grid point above and the disturbance intensity of the grid point below, divided by twice the grid spacing. Grid point gradient magnitude Calculate using the following formula: in, Represents grid points gradient magnitude, This is the row index of the grid points in the global disturbance field distribution map of the exhibition hall. For the column index of the grid points, This represents the partial derivative of the disturbance intensity along the horizontal direction. The subscript represents the partial derivative of the disturbance intensity along the vertical direction. This indicates the partial derivative at grid points. The value at that location.

[0038] The control decision module compares the calculated gradient magnitude of each grid point with a preset magnitude threshold. The preset magnitude threshold is determined by calculating the arithmetic mean and standard deviation of the gradient magnitudes of all grid points in the global disturbance field distribution map of the exhibition hall. The preset magnitude threshold is the arithmetic mean plus twice the standard deviation. Grid points with gradient magnitudes greater than the preset magnitude threshold are marked as high-gradient grid points.

[0039] High-gradient grid points are subjected to neighborhood aggregation using an eight-neighbor connected component analysis algorithm. Spatially adjacent high-gradient grid points are grouped into the same connected component, and any two high-gradient grid points within a connected component can be connected through a series of eight-neighbor adjacent high-gradient grid points. Each connected component forms a candidate region, and all candidate regions constitute a candidate region set.

[0040] In each candidate region set, the location of the maximum perturbation intensity of all grid points within the region is extracted as the extreme point. If multiple grid points within the same candidate region have the same maximum perturbation intensity, the geometric center of these grid points is taken as the extreme point. The control decision module further calculates the gradient direction of each grid point within the candidate region. The gradient direction is determined by the arctangent function of the partial derivatives in the horizontal and vertical directions. Grid points whose gradient directions point to the extreme points within a certain angular deviation range are included in the spatial range of the perturbation field gradient extreme region. The angular deviation range is preset to ±15 degrees. The spatial range containing the extreme points and whose gradient directions point to the extreme points is marked as the perturbation field gradient extreme region.

[0041] The control decision module uses the identified extreme value regions of the perturbation field gradient as control anchor points. During the generation of the global perturbation field distribution map of the exhibition hall by the perturbation analysis module, each perturbation propagation path is recorded. The records include the starting node, ending node, and the sequence of connecting edges traversed during propagation. All records form a superimposed propagation path dataset. Each propagation path in the superimposed propagation path dataset has a clear direction, pointing from the perturbation source node to the propagated node.

[0042] The control decision module starts from the topological node corresponding to the control anchor point, queries the superimposed transmission path dataset for all transmission channels terminating at the control anchor point, and extracts the starting nodes of these transmission channels as the upstream nodes of the control anchor point. Then, using each upstream node as a new backtracking point, it continues to query its respective upstream nodes, traversing level by level in the reverse direction of the transmission channel. The backtracking process continues until either of the following stopping conditions is met: the backtracking reaches the boundary node of the spatial topology network, where the boundary node is a topology node with only one connecting edge; or the accumulated backtracking steps reach a predetermined step threshold. The predetermined step threshold is determined based on the maximum diameter of the spatial topology network, which is obtained by calculating the maximum shortest path length between all pairs of topology nodes in the spatial topology network using the Floyd algorithm. The predetermined step threshold is set to half of the maximum diameter and rounded up. All topology nodes traversed during the backtracking process are added to the set of areas to be controlled, and the topology node corresponding to the control anchor point itself is also added to the set of areas to be controlled.

[0043] For each topological node in the set of exhibition areas to be regulated, the cumulative disturbance intensity and associated disturbance value of that topological node in the global disturbance field distribution map of the exhibition hall are obtained. The associated disturbance value is the associated disturbance value of the corresponding topological node calculated by the disturbance analysis module. The ratio of the cumulative disturbance intensity to the associated disturbance value of the topological node is calculated, and this ratio is used as the disturbance transmission contribution rate of the topological node. The disturbance transmission contribution rate of the topological node is directly assigned as the regulation response coefficient of the corresponding exhibition area to be regulated.

[0044] Example 5: In practical implementation, the instruction distribution module obtains the set of controllable zones output by the control decision module, as well as the control response coefficient corresponding to each controllable zone in the set. For each controllable zone in the set, the instruction distribution module obtains the device type identifier and current operating parameter values ​​for that zone. The device type identifier is used to distinguish the types of execution devices deployed within the zone, including air conditioning device identifiers, fresh air device identifiers, lighting device identifiers, and multimedia device identifiers. The current operating parameter values ​​are the actual operating parameters of each execution device within the controllable zone at the current moment. The operating parameters of the air conditioning device include the set temperature and airflow rate; the operating parameters of the fresh air device include the air exchange frequency; the operating parameters of the lighting device include brightness and color temperature; and the operating parameters of the multimedia device include volume and screen brightness. The current operating parameter values ​​are read from the multi-source sensing devices connected to the status acquisition module, which include operating status sensors connected to each execution device.

[0045] The instruction distribution module queries a preset adjustable range table of device parameters based on the acquired device type identifier. This table, using the device type identifier as the primary key, stores the lower and upper limits of allowable adjustments for each operating parameter corresponding to each device type. For example, the adjustable temperature range for air conditioning equipment is 16°C to 30°C, and the adjustable airflow speed is 0.5 m / s to 5.0 m / s; the adjustable air exchange rate for fresh air equipment is 1 to 10 times per hour; the adjustable brightness range for lighting equipment is 100 lumens to 2000 lumens, and the adjustable color temperature range is 2700 Kelvin to 6500 Kelvin; the adjustable volume range for multimedia equipment is 0 dB to 100 dB, and the adjustable screen brightness range is 50 nits to 500 nits. This adjustable range table is configured and stored by exhibition hall management personnel during system initialization based on the hardware specifications of the equipment in each exhibition area.

[0046] The instruction distribution module simultaneously queries a pre-defined parameter response mapping table based on the acquired device type identifier. The pre-defined parameter response mapping table uses the device type identifier as the primary key, and each primary key stores several mapping relationships between response coefficient ranges and target adjustment levels. The response coefficient ranges are pre-divided into multiple consecutive numerical intervals according to the value range of the control response coefficient, with each interval corresponding to a target adjustment level. The division method for the response coefficient ranges is as follows: for the entire value range of the control response coefficient, it is divided into five equally spaced ranges. The first range corresponds to the 0% to 20% range of the control response coefficient, arranged from smallest to largest; the second range corresponds to the 20% to 40% range; the third range corresponds to the 40% to 60% range; the fourth range corresponds to the 60% to 80% range; and the fifth range corresponds to the 80% to 100% range. Each gear interval maps to a target adjustment gear, which is an integer starting from 0 and incrementing. Specifically, the first gear interval maps to target adjustment gear 1, the second gear interval to target adjustment gear 2, the third gear interval to target adjustment gear 3, the fourth gear interval to target adjustment gear 4, and the fifth gear interval to target adjustment gear 5. The parameter response mapping table is configured and stored by the exhibition hall management personnel during the system initialization phase based on control strategies and experience values. The gear interval divisions corresponding to different equipment types may vary.

[0047] The instruction distribution module matches the control response coefficient of the area to be controlled with the corresponding response coefficient range in the parameter response mapping table. The matching process is as follows: the control response coefficient is normalized to the percentile within the range of control response coefficient values ​​to obtain the normalized control response coefficient. The normalized control response coefficient is then compared with each response coefficient range in the parameter response mapping table one by one to find the response coefficient range in which the normalized control response coefficient is located, and the target adjustment level corresponding to that response coefficient range is extracted.

[0048] After determining the target adjustment level for the area to be regulated, the instruction distribution module calculates the adjustment amount of the operating parameters for each executing device in the area based on the target adjustment level and the preset adjustable range table of equipment parameters. For each operating parameter, the adjustment amount is calculated as follows: subtract the lower limit of the allowable adjustment from the upper limit of the operating parameter in the adjustable range table to obtain the total allowable adjustment range. Multiply this total allowable adjustment range by the ratio of the target adjustment level to the total number of adjustment levels (5) to obtain the adjustment amount. The calculated adjustment amount is then added to the current operating parameter value to obtain the equipment execution parameter. If the added equipment execution parameter exceeds the upper limit of the allowable adjustment in the adjustable range table, the equipment execution parameter is truncated to the upper limit; if the added equipment execution parameter is lower than the lower limit of the allowable adjustment in the adjustable range table, the equipment execution parameter is truncated to the lower limit.

[0049] In some implementations, the same exhibition area to be regulated contains multiple execution devices. The instruction distribution module obtains a device list of all execution devices within the exhibition area to be regulated. The device list includes the device type identifier, rated power, and preset device priority for each execution device. The device priority is set by the exhibition hall management personnel based on the severity of the device's impact on the exhibition area environment; a smaller value indicates a higher priority. The instruction distribution module allocates the total regulation response coefficient of the exhibition area to be regulated according to the rated power ratio of each execution device, obtaining the allocated regulation response coefficient for each execution device. Calculate using the following formula: in, Indicates the first in the exhibition area to be regulated The allocation and control response coefficient of each execution device This represents the total regulatory response coefficient of the area to be regulated. Indicates the first The rated power of each actuator This indicates the total number of devices to be controlled within the exhibition area. Indicates the index number of the execution device. Indicates the first The rated power of each actuator This represents the sum of the rated power of all executing equipment within the exhibition area to be regulated.

[0050] The instruction distribution module calculates and truncates the operating parameter adjustments for each execution device in descending order of device priority. When the execution parameters calculated by a device according to the allocation control response coefficient exceed the adjustable range of the device parameters, the excess adjustment is truncated. The truncated adjustment is added to the allocation control response coefficient of the next-lower-priority execution device, and the process continues to process the next-lower-priority execution device. This step-by-step truncation process continues until all execution devices have completed processing, and all adjustments exceeding the adjustable range are accumulated and truncated, no longer being passed down.

[0051] The instruction distribution module converts the calculated device execution parameters of each execution device into a sequence of control instructions. Each control instruction sequence includes the execution device identifier, the target operating parameter name, and the target operating parameter value, and is encoded according to the exhibition hall's equipment communication protocol. The encoded control instruction sequence is then sent via wired or wireless communication interfaces to the corresponding terminal execution device in the exhibition area to be controlled. Upon receiving the control instruction sequence, the terminal execution device parses and executes it, completing the adjustment of the device's operating parameters.

[0052] See Figure 4 In the graph, the horizontal axis represents the control response coefficient, ranging from 0 to 1, and the vertical axis represents the target adjustment level, ranging from 1 to 5. The graph uses scatter dots of different shapes and colors to identify the corresponding data points for four equipment types: circles for air conditioning, squares for ventilation, triangles for lighting, and diamonds for multimedia equipment. The scatter dots are clearly distributed across five discrete horizontal lines, indicating a clear hierarchical correspondence between the control response coefficient and the target adjustment level.

[0053] Within the control response coefficient range of 0 to 0.2, the target adjustment levels for all equipment types are concentrated at level 1, indicating that the equipment operating parameter adjustment is at its lowest level. For control response coefficients ranging from 0.2 to 0.4, the corresponding target adjustment level is 2, and the equipment operating parameter adjustment increases to the second level. The control response coefficient range of 0.4 to 0.6 corresponds to level 3, 0.6 to 0.8 to level 4, and 0.8 to 1.0 to level 5, showing that the target adjustment levels increase with the control response coefficient, with clear divisions and no overlap between levels.

[0054] The scattered data points for each equipment type are relatively concentrated within each gear range, indicating that the instruction distribution module adopts a unified gear classification strategy based on the control response coefficient for different equipment types. The matching process between the control response coefficients of air conditioning equipment, fresh air equipment, lighting equipment, and multimedia equipment and the target adjustment gear shows consistency, reflecting the gear range division of different equipment type identifiers and the target adjustment gear mapping relationship in the preset parameter response mapping table.

[0055] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. An intelligent management system for exhibition halls based on big data analysis, characterized in that, The system includes: The spatial topology module acquires architectural information model data of the exhibition hall and extracts the spatial boundaries and passage connection relationships of each exhibition area to generate a spatial topology network of the entire exhibition hall. The status acquisition module connects to multi-source sensing devices deployed in each exhibition area to collect real-time data on instantaneous pedestrian flow, ambient temperature and humidity, and equipment operating power in each exhibition area, forming a time-series matrix of the exhibition area's operating status. The disturbance analysis module uses the exhibition hall's global spatial topology network as a constraint basis, maps the exhibition area's operating status time series matrix to each node of the spatial topology network, calculates the associated disturbance value of each node's operating status relative to its adjacent nodes, and generates a global disturbance field distribution map of the exhibition hall based on the superposition and propagation path of each node's associated disturbance value on the spatial topology network. The control decision module identifies the extreme value region of the disturbance field gradient based on the global disturbance field distribution map of the exhibition hall, and uses the extreme value region of the disturbance field gradient as the control anchor point to backtrack along the opposite direction of the superimposed transmission path to determine the set of exhibition areas to be controlled and the control response coefficient of each exhibition area to be controlled; The instruction distribution module matches the equipment execution parameters of each exhibition area to be regulated according to the regulation response coefficient, and converts the equipment execution parameters into a regulation instruction sequence and sends it to the terminal execution device of the corresponding exhibition area.

2. The exhibition hall intelligent management system based on big data analysis according to claim 1, characterized in that, Acquire architectural information model data of the exhibition hall and extract the spatial boundaries and passageway connections of each exhibition area to generate a global spatial topology network of the exhibition hall, including: Using the area enclosed by the spatial boundaries of each exhibition area as a topological node, the geometric center coordinates and the location coordinates of the passageway of each topological node are extracted; The connection path between adjacent topological nodes is determined based on the channel connection relationship, and the path length and path width of the connection path are used as edge weight attributes. Using the geometric center coordinates as node location identifiers, the passageway location coordinates as port connection identifiers, and the edge weight attributes as connection edge attributes, the exhibition hall's entire spatial topology network is constructed.

3. The exhibition hall intelligent management system based on big data analysis according to claim 1, characterized in that, The calculation of the correlation disturbance value of each node's operating state relative to its neighboring nodes includes: A local neighborhood set is formed by the target topology node and all its adjacent topology nodes. The time sequence submatrix of the running status of each node in the local neighborhood set within the same time window is extracted. Calculate the dynamic time warping distance between the time sequence submatrix of the target topology node's running state and the time sequence submatrix of the running state of each adjacent topology node, and use the weighted sum of all dynamic time warping distances as the associated perturbation value of the target topology node, where the weighting coefficient is the normalized value of the edge weight attribute of the target topology node and its corresponding adjacent topology node.

4. The exhibition hall intelligent management system based on big data analysis according to claim 1, characterized in that, The step of generating a global disturbance field distribution map of the exhibition hall based on the superposition and propagation paths of the associated disturbance values ​​of each node on the spatial topology network includes: The associated perturbation value of each node is used as the initial perturbation source strength. The connection edge of the spatial topology network is used as the propagation channel. The attenuation coefficient of the perturbation along the propagation channel is set to be inversely proportional to the edge weight attribute. Starting from each disturbance source, the disturbance intensity is propagated to the outer nodes level by level along each propagation channel. The nodes that are propagated to the outer nodes accumulate the received disturbance intensity with their own initial disturbance source intensity to generate the cumulative disturbance intensity of each node. Spatial interpolation is performed based on the cumulative disturbance intensity of each node and the geometric coordinates of the spatial topology network to generate a global disturbance field distribution map of the exhibition hall.

5. The exhibition hall intelligent management system based on big data analysis according to claim 1, characterized in that, Identify the extreme value regions of the perturbation field gradient based on the global perturbation field distribution map of the exhibition hall, including: Calculate the partial derivatives of the disturbance intensity at each grid point in the global disturbance field distribution map of the exhibition hall along the horizontal and vertical directions, and take the square root of the sum of the squares of the partial derivatives as the gradient magnitude of that grid point; Grid points with gradient magnitudes greater than a preset magnitude threshold are subjected to neighborhood aggregation processing to form a candidate region set; Extract the location of the maximum perturbation intensity of grid points in each candidate region set as the extreme point, and mark the spatial range containing the extreme point and whose gradient direction points to the extreme point as the extreme region of the perturbation field gradient.

6. The exhibition hall intelligent management system based on big data analysis according to claim 1, characterized in that, Determining the set of regions to be regulated and the regulation response coefficient of each region by tracing back along the opposite direction of the superimposed transmission path includes: Using the control anchor point as the starting point for backtracking, the upstream nodes are traversed step by step in the opposite direction of each transmission channel in the superimposed transmission path until the boundary node of the spatial topology network is reached or the cumulative backtracking steps reach a predetermined number of steps threshold. All the nodes traversed are then included in the set of the controllable exhibition area. The disturbance transmission contribution rate of each node is determined based on the ratio of the cumulative disturbance intensity of each node in the set of exhibition areas to be regulated to the associated disturbance value of the node in the global disturbance field distribution map of the exhibition hall. The disturbance propagation contribution rate of each node is used as the control response coefficient of the corresponding control zone.

7. The exhibition hall intelligent management system based on big data analysis according to claim 1, characterized in that, Matching the equipment execution parameters of each exhibition area to be regulated according to the aforementioned regulation response coefficient includes: Obtain the equipment type identifier and current operating parameter value of each exhibition area to be regulated, and query the preset equipment parameter adjustable range table and parameter response mapping table based on the equipment type identifier; The control response coefficient is matched with the response coefficient level in the parameter response mapping table to determine the target adjustment level for each control zone. Based on the target adjustment level and the adjustable range table of equipment parameters, calculate the adjustment amount of the operating parameters of each executing device in each controllable exhibition area, and then add the current operating parameter value and the adjustment amount of the operating parameter to obtain the equipment execution parameter.

8. The exhibition hall intelligent management system based on big data analysis according to claim 3, characterized in that, In the process of calculating the dynamic time warping distance, the state dimensions of each dimension in the running state time sequence submatrix are normalized, and the normalized values ​​of each dimension are weighted and corrected by the exhibition area and environmental capacity coefficient.

9. The exhibition hall intelligent management system based on big data analysis according to claim 4, characterized in that, The attenuation coefficient is also related to the ratio of the spatial boundary areas of adjacent nodes. When the disturbance propagates from a large area to a small area, the attenuation coefficient is scaled by the square root of the area ratio.

10. The exhibition hall intelligent management system based on big data analysis according to claim 7, characterized in that, When the same controllable area contains multiple actuators, the total control response coefficient of the area is allocated according to the rated power ratio of each actuator, and the adjustment amount exceeding the adjustable range of the device parameters is truncated step by step according to the priority of the devices.

Images