Intelligent visual energy-saving power distribution system, method, electronic device and storage medium

Abstract

The application discloses an intelligent visual energy-saving power distribution system and method, an electronic device and a storage medium, and relates to the field of energy management.The method comprises the following steps: acquiring facial data, sound data and motion data of all audiences in a region; calculating a facial consistency vector based on the facial data; calculating an acoustic event vector based on the sound data; calculating an emotional intensity value according to the facial consistency vector and the acoustic event vector; dividing each motion data into continuous frames, calculating the motion vectors of all audiences in the corresponding continuous frames; analyzing the direction of each motion vector to obtain direction consistency, and selecting a crowd flow with direction consistency greater than a preset threshold; determining a collective directional movement trend and determining a target region that needs to be preheated; predicting the power demand of each target region according to the emotional intensity value and each target region; and distributing power and displaying the result on a visual interface. The technical scheme provided by the application can improve the intelligent level of the power distribution system.

Description

Technical Field

[0001] This application relates to the technical field of energy management, specifically to an intelligent visual energy-saving power distribution system, method, electronic device, and storage medium. Background Technology

[0002] Large venues, such as stadiums, concert halls, and exhibition centers, are an important part of modern cities, consuming enormous amounts of energy, especially during large-scale events. The electrical systems of these venues need to support various equipment, including lighting, temperature control, and audiovisual systems. Efficiently and rationally managing the energy use of these systems while ensuring a positive audience experience is a core challenge in venue operation.

[0003] To address this challenge, existing technologies typically employ power control schemes based on static zone division. Specifically, the venue is pre-divided into multiple fixed, isolated zones, each with an independent infrared counting sensor deployed at its physical entrance or exit. These sensors are used only to count the instantaneous flow of people passing through their monitoring points and compare the cumulative number to a fixed, pre-set threshold. When the cumulative number of people in a zone exceeds this threshold, the logic controller triggers a power state switch for all associated equipment (such as air conditioning and lighting) within that zone, for example, switching from "standby" to "operation".

[0004] However, the inherent mechanism of this control method limits its application. Because the control logic is entirely based on discrete, static judgments of "quantity," its sensors and controllers lack the ability to analyze collective crowd behavior patterns. Therefore, it treats all crowds as homogeneous, static loads, unable to distinguish whether they are waiting quietly, moving in an orderly fashion, or engaging in loud, collective cheering. When a large number of spectators rapidly move from one area to another, because the controls between areas are isolated, the system in the target area only begins to respond after a large influx of people has triggered local sensors. This results in spectators entering a dimly lit, stuffy environment, significantly diminishing their experience and indicating a low level of intelligence in the venue's power distribution system. Summary of the Invention

[0005] This application provides an intelligent visual energy-saving power distribution system, method, electronic device, and storage medium, which can adjust power distribution according to actual activity status and improve the intelligence level of the power distribution system.

[0006] The first aspect of this application provides an intelligent visual energy-saving power distribution system, characterized in that the system comprises: a data acquisition module, an emotion intensity value calculation module, a preheating zone determination module, a power prediction module, and a visual power distribution module, wherein:

[0007] The data acquisition module is used to acquire facial data, voice data, and motion data of all spectators in at least one area of ​​the target venue at the current time.

[0008] The emotion intensity value calculation module is used to calculate a facial consistency vector based on the facial data of all viewers in each region. The facial consistency vector is used to characterize the overall emotional state of the viewers in each region.

[0009] The emotion intensity value calculation module is used to extract acoustic features from the sound data of all audience members in each of the regions, and calculate acoustic event vectors based on the acoustic features. The acoustic event vectors are used to characterize the overall sound environment of the audience in each of the regions.

[0010] The emotion intensity value calculation module is used to calculate the overall emotion intensity value of the audience in each of the aforementioned regions based on the facial consistency vector and the acoustic event vector.

[0011] The preheating area determination module is used to divide each of the motion data into consecutive frames and calculate the motion vectors of all viewers in the corresponding consecutive frames.

[0012] The preheating area determination module is used to analyze the direction of each motion vector in each area to obtain directional consistency, and mark the flow of people whose directional consistency is greater than a preset collective movement threshold as the target flow of people;

[0013] The preheating area determination module is used to determine the collective directional movement trend of each of the target crowds, and to determine the target area that needs to be preheated based on the collective directional movement trend.

[0014] The power prediction module is used to predict the power demand of each target area within the time range of the arrival of each target flow of people in the corresponding target area, based on the emotion intensity value and each target area.

[0015] The visualization power distribution module is used to distribute power to each target area according to the power demand of each target area and display the power distribution results on a preset visualization interface.

[0016] By employing the aforementioned technical solution, the data acquisition module collects real-time facial and audio data of the entire area, as well as the movement data of each audience member. The emotion intensity calculation module calculates the overall emotion intensity value of the audience based on facial consistency vectors and acoustic event vectors. Simultaneously, the preheating area determination module identifies the target crowd flow and its movement trend by analyzing the directional consistency of movement vectors. Furthermore, the power prediction module can predict the power demand of the target area in advance. This proactive perception mechanism based on multi-dimensional data overcomes the limitations of traditional static area counting, enabling the system to preheat and distribute power to the target area before the crowd arrives. More importantly, by quantifying the audience's emotional state and collective behavior patterns into specific power demands, the system achieves differentiated power supply for different activity states. Finally, the intuitive management interface provided by the visualized power distribution module further enhances the intelligence level of the venue's power distribution system.

[0017] Optionally, the emotion intensity value calculation module is used to calculate the overall emotion intensity value of the audience in each of the aforementioned regions based on the facial consistency vector and the acoustic event vector, including:

[0018] The emotion intensity value calculation module is specifically used for any of the aforementioned regions:

[0019] When the facial consistency vector is greater than or equal to the facial threshold and the acoustic event vector is greater than or equal to the acoustic threshold, the facial consistency vector and the acoustic event vector are weighted and summed to obtain the emotion intensity value.

[0020] When the facial consistency vector is greater than or equal to the facial threshold and the acoustic event vector is less than the acoustic threshold, or when the facial consistency vector is less than the facial threshold and the acoustic event vector is greater than or equal to the acoustic threshold, the primary and secondary vectors are determined by comparing the rate of change of the facial consistency vector and the acoustic event vector, and the primary and secondary vectors are weighted and summed to obtain the emotion intensity value.

[0021] When the facial consistency vector is less than the facial threshold and the acoustic event vector is less than the acoustic threshold, the historical facial consistency vector and historical acoustic event vector of the region are obtained, and the historical facial consistency vector and the historical acoustic event vector are weighted and summed to obtain the emotion intensity value.

[0022] By employing the aforementioned technical solution, the system achieves precise quantification of audience emotional states through the collaborative analysis of facial consistency vectors and acoustic event vectors. Specifically, when both vectors exhibit strong consistency and directional alignment, the system uses a weighted summation method to integrate the data from both dimensions. When one vector is prominent while the other is relatively weak, the system can identify the primary and secondary vectors by comparing their rates of change and adjust the weighting accordingly, ensuring that the calculation results better reflect the actual emotional intensity. More importantly, to address potential data quality issues in complex scenarios, the system establishes a degradation processing mechanism based on historical data. When the quality of facial consistency vectors or acoustic event vectors is poor, corresponding historical data is introduced for compensation calculations, ensuring the continuity and reliability of emotional intensity value calculations. This multi-layered adaptive calculation strategy effectively overcomes the limitations of traditional techniques that treat the crowd as a homogeneous static load, providing a more accurate input basis for subsequent power demand prediction.

[0023] Optionally, the preheating area determination module is used to determine the collective directional movement trend of each of the target pedestrian flows, and to determine the target area requiring preheating based on the collective directional movement trend, including:

[0024] The preheating area determination module is specifically used to calculate the average vector of all the motion vectors in each of the target pedestrian flows, and to determine the average vector as the collective directional movement trend.

[0025] Construct a spatial connectivity graph, wherein each region is a node and the path between regions is an edge.

[0026] Select the edge closest to the target pedestrian flow as the target edge, and aggregate all motion vectors on the target edge to calculate the instantaneous flow rate and mainstream direction of the target edge;

[0027] Based on the instantaneous flow rate and the mainstream direction, calculate the net flow divergence of each node;

[0028] The nodes on the spatial connectivity graph are filtered to obtain target nodes, and the passage cost of each target node is calculated based on the collective directional movement trend.

[0029] By combining the absolute value of the net flow dispersion of each target node with the passage cost of each target node, the attractiveness score of each target node is calculated. One or more target nodes whose attractiveness score is greater than the preset preheating score are identified as target areas that need to be preheated.

[0030] By employing the aforementioned technical solution, the overall movement trend is first obtained by calculating the average vector of the movement vectors within the target pedestrian flow. Then, combined with the topological structure of the spatial connectivity graph, the movement vectors on the target edges are aggregated and analyzed to obtain the instantaneous flow rate and mainstream direction. Based on this, by calculating the net flow divergence of nodes and combining it with the passage cost calculated based on the collective directional movement trend, the system can quantitatively evaluate the attractiveness score of each target node, thereby scientifically identifying target areas requiring preheating. This predictive mechanism, built upon the actual physical spatial structure and real-time pedestrian flow dynamics, overcomes the limitations of traditional technologies where areas are isolated from each other. It allows the system to identify potential gathering areas in advance based on the movement intentions of the audience before their actual arrival, providing accurate spatial dimensional guidance for proactive power adjustment and effectively avoiding the environmental adaptation lag problem caused by traditional passive responses.

[0031] Optionally, the preheating area determination module is used to filter each of the nodes on the spatial connectivity graph to obtain target nodes, and calculate the passage cost of each of the target nodes according to the collective directional movement trend, including: the preheating area determination module is further specifically used to mark the nodes with positive net flow divergence and absolute value of net flow divergence greater than a preset sink threshold as human flow sinks.

[0032] Calculate the vector angle between each of the aforementioned pedestrian flow convergence points and the collective directional movement trend, and designate pedestrian flow convergence points whose vector angle is less than a preset first angle threshold as target nodes;

[0033] The shortest path search algorithm is executed on each of the target nodes to calculate the basic passage cost of each target node;

[0034] The cost correction amount is calculated based on the vector angle between each target node and the collective directional movement trend, and the basic passage cost is added to the cost correction amount to obtain the passage cost of each target node.

[0035] By employing the aforementioned technical solution, potential pedestrian convergence points are first identified based on net flow divergence. Then, by analyzing the vector angles between these convergence points and the collective directional movement trend, nodes more likely to become crowd gathering targets are selected. Building upon this, the system not only considers the shortest path in the spatial connectivity graph as the basic access cost but also introduces a cost correction based on the vector angle, thus superimposing the influence of pedestrian movement direction on top of traditional distance metrics. This access cost calculation method, which considers both spatial structure and movement trend constraints, enables the system to more accurately assess the actual accessibility of each target node. This provides more realistic input parameters for subsequent attraction score calculations, thereby improving the accuracy of target area prediction and laying the foundation for precise preheating and power distribution.

[0036] Optionally, the power prediction module is used to predict the power demand of each target area within the time range of the arrival of each target pedestrian flow in the corresponding target area, based on the emotion intensity value and each target area, including:

[0037] The power prediction module is specifically used to obtain the base power of electrical equipment in each target area under the condition of no audience; obtain the number of people in the target flow, and determine the arrival time range based on the number of people and the passage cost of the target flow to the target area;

[0038] Calculate the upper limit of the power of the basic equipment in the electrical equipment based on the number of people;

[0039] Based on the power limit and the arrival time range, the first power of the infrastructure within the arrival time range is calculated; the passage cost and attraction score of the target area are normalized and then weighted and summed to obtain the emotion transfer coefficient;

[0040] Based on the emotional intensity value and the emotional transfer coefficient, predict the expected emotional intensity value of the target crowd after it arrives at the target area;

[0041] Based on the expected emotional intensity value and the arrival time range, calculate the second power of the atmosphere device in the electrical equipment after the target flow of people begins to move;

[0042] The base power, the first power, and the second power are time-series superimposed to obtain the power demand of each target area within the time range of the target flow arriving at the target area.

[0043] By adopting the above technical solution, a three-tiered power architecture was first established, comprising basic power, crowd-carrying power, and atmosphere-creating power. A minimum guarantee was established by acquiring the basic power under no-audience conditions. The power requirements of basic equipment were dynamically modeled based on the number of people and passage costs of the target crowd, yielding the first power reflecting crowd-carrying needs. Simultaneously, an emotion transfer coefficient was calculated based on passage costs and attraction scores, and the expected emotion intensity after the arrival of the target crowd was predicted by combining the current emotion intensity value, thus guiding the calculation of the second power for atmosphere-creating equipment. This predictive method, which maps multi-dimensional information such as the number of people, movement characteristics, and emotional states to the power requirements of different types of equipment, overcomes the limitations of traditional technologies that uniformly manage all equipment, achieving differentiated power supply based on scenario requirements. By time-series superimposing these three types of power, the system ultimately obtains a complete power curve that reflects both the physical presence of the crowd and matches the emotional atmosphere requirements, providing a reliable basis for precise and efficient energy allocation.

[0044] Optionally, the power prediction module is used to calculate the second power of the atmosphere device in the electrical equipment after the target crowd begins to move, based on the expected emotional intensity value and the arrival time range, including:

[0045] The power prediction module is further specifically used to divide the arrival time range into multiple time segments according to a preset time interval; to perform fluctuation analysis on the flow speed of people in each time segment, and to obtain the acceleration period and deceleration period of the target flow of people;

[0046] Calculate the percentage of emotional activation during the acceleration period and the percentage of emotional decay during the deceleration period;

[0047] Based on the emotional intensity value, the expected emotional intensity value, and the proportion of emotional activation or the proportion of emotional decay in each time segment, an asymptotic curve of the emotional intensity value is constructed to obtain the instantaneous emotional intensity value corresponding to each time segment; the instantaneous emotional intensity value of each time segment is mapped to the corresponding power adjustment coefficient, and the power adjustment coefficient is multiplied by the rated power of the atmosphere device to obtain the target power value in each time segment;

[0048] The target power values ​​of each time segment are combined in a time sequence to generate the second power of the atmosphere device after the target flow of people begins to move.

[0049] By adopting the above technical solution and analyzing the fluctuations in pedestrian flow speed at different time segments, the system can accurately identify the acceleration and deceleration phases of the target pedestrian flow, thereby gaining a deeper understanding of the dynamic behavioral characteristics of the crowd movement process. Based on the quantitative calculation of the proportion of emotional activation during the acceleration phase and the proportion of emotional decay during the deceleration phase, the system establishes an intrinsic correlation mechanism between changes in pedestrian movement speed and fluctuations in emotional state. According to the emotional intensity value, the expected emotional intensity value, and the proportion of emotional activation or emotional decay at each time segment, the system constructs a progressive curve of emotional intensity value. The system achieves smooth transition modeling from the current emotional state to the expected emotional state, avoiding abrupt errors in emotional prediction and significantly improving the prediction accuracy of the instantaneous emotional intensity value corresponding to each time segment. The technical approach involves mapping the instantaneous emotional intensity values ​​of each time segment to the corresponding power adjustment coefficients and multiplying them by the rated power of the atmosphere equipment to obtain the target power value within each time segment. By combining the target power values ​​of each time segment in a time sequence, the second power of the atmosphere equipment is generated after the target flow of people begins to move. The system forms a continuous and consistent power adjustment strategy, ensuring that the atmosphere equipment can provide an atmosphere effect that is highly matched with the changes in the emotional state of the crowd throughout the entire flow of people, thereby improving the energy efficiency optimization level of the equipment operation.

[0050] Optionally, the preheating area determination module is used to analyze the direction of each motion vector within each of the said areas to obtain directional consistency, and to mark the flow of people whose directional consistency is greater than a preset collective movement threshold as target flow, including:

[0051] The preheating area determination module is further specifically used to perform cluster analysis on the motion vectors in each of the said areas to obtain multiple motion vector clusters;

[0052] Calculate the angle between each motion vector within each motion vector cluster, and count the number of vector pairs with an angle less than a preset second angle threshold;

[0053] Based on the ratio of the number of vector pairs to the total number of vector pairs within the motion vector cluster, the directional consistency of each motion vector cluster is calculated.

[0054] The motion vector clusters whose directional consistency is greater than a preset collective movement threshold are selected as the target pedestrian flow.

[0055] By employing the aforementioned technical solution, the system first performs cluster analysis on the motion vectors within the region, classifying individuals with similar motion characteristics into different motion vector clusters, thus achieving a preliminary division of crowd movement patterns. Based on this, the system establishes a method for quantitatively evaluating the consistency of group movement by calculating the angles between motion vectors within clusters and counting the number of vector pairs with angles less than a preset second angle threshold. By using the ratio of the number of qualified vector pairs to the total number of vector pairs as a directional consistency index, the system can accurately identify crowds with obvious collective movement characteristics. This crowd behavior recognition method based on vector angle statistics overcomes the limitations of traditional techniques that rely solely on simple density or speed thresholds, enabling the system to accurately filter out target crowds with clear movement intentions in complex crowd movement scenarios, providing reliable input data for subsequent pre-warming area prediction.

[0056] A second aspect of this application provides an intelligent visual energy-saving power distribution method, applied to an intelligent visual energy-saving power distribution system, the method comprising:

[0057] Acquire facial data, voice data, and motion data of all audience members in at least one area of ​​the target venue at the current time; calculate a facial consistency vector based on the facial data of each audience member, the facial consistency vector being used to characterize the overall emotional state of the audience in each area;

[0058] Acoustic features are extracted from the sound data of each audience member, and acoustic event vectors are calculated based on the acoustic features. The acoustic event vectors are used to characterize the overall sound environment of the audience in each area.

[0059] The overall emotional intensity value of the audience in each region is calculated based on the facial consistency vector and the acoustic event vector; the motion data is divided into consecutive frames, and the motion vectors of all audience members in the corresponding consecutive frames are calculated.

[0060] The direction of each motion vector in each region is analyzed to obtain directional consistency, and the flow of people whose directional consistency is greater than a preset collective movement threshold is marked as the target flow of people;

[0061] Determine the collective directional movement trend of each of the target pedestrian flows, and determine the target area that needs to be preheated based on the collective directional movement trend;

[0062] Based on the emotional intensity value and each of the target areas, predict the power demand of each target area within the time range of the arrival of each target pedestrian flow in the corresponding target area;

[0063] Power is distributed to each target area according to its power demand, and the distribution results are displayed on a preset visualization interface.

[0064] By employing the aforementioned technical solution, the data acquisition module collects real-time facial and audio data of the entire area, as well as the movement data of each audience member. The emotion intensity calculation module calculates the overall emotion intensity value of the audience based on facial consistency vectors and acoustic event vectors. Simultaneously, the preheating area determination module identifies the target crowd flow and its movement trend by analyzing the directional consistency of movement vectors. Furthermore, the power prediction module can predict the power demand of the target area in advance. This proactive perception mechanism based on multi-dimensional data overcomes the limitations of traditional static area counting, enabling the system to preheat and distribute power to the target area before the crowd arrives. More importantly, by quantifying the audience's emotional state and collective behavior patterns into specific power demands, the system achieves differentiated power supply for different activity states. Finally, the intuitive management interface provided by the visualized power distribution module further enhances the intelligence level of the venue's power distribution system.

[0065] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any of the foregoing.

[0066] A fourth aspect of this application provides a computer-readable storage medium storing instructions that, when executed, perform the method described in any of the preceding descriptions. Attached Figure Description

[0067] Figure 1This is a schematic diagram of a module of an intelligent visual energy-saving power distribution system provided in an embodiment of this application;

[0068] Figure 2 This is an exemplary spatial connectivity graph provided in the embodiments of this application;

[0069] Figure 3 This is an example diagram of a visual interface provided in an embodiment of this application;

[0070] Figure 4 This is a flowchart illustrating an intelligent, visual, energy-saving power distribution method provided in an embodiment of this application.

[0071] Figure 5 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application.

[0072] Figure labeling: 10, Data acquisition module; 20, Emotion intensity value calculation module; 30, Warm-up area determination module; 40, Power prediction module; 50, Visual power distribution module; 901, Processor; 902, Communication bus; 903, User interface; 904, Network interface; 905, Memory. Detailed Implementation

[0073] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0074] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0075] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0076] Please see Figure 1 , Figure 1 This is a schematic diagram of a module of an intelligent visual energy-saving power distribution system provided in an embodiment of this application.

[0077] The intelligent visual energy-saving power distribution system includes a data acquisition module 10, an emotion intensity value calculation module 20, a preheating zone determination module 30, a power prediction module 40, and a visual power distribution module 50, wherein:

[0078] The aforementioned data acquisition module 10 is used to acquire facial data, voice data, and motion data of all spectators in at least one area of ​​the target venue at the current time.

[0079] Specifically, the data acquisition module 10 adopts a multimodal perception architecture, acquiring three types of key data in each area of ​​the target venue through different types of sensing devices.

[0080] The target venue comprises multiple zones, each housing a varying number of spectators. Each zone is equipped with high-definition cameras, sound acquisition devices, and motion sensors. This diverse sensor configuration comprehensively monitors the behavior and environmental changes of all spectators within each zone, determining their overall emotional state and individual movement patterns. This provides accurate data support for subsequent equipment control and power configuration.

[0081] Facial data refers to the overall facial expression features extracted from all audience members within a region using image recognition technology. This includes comprehensive features such as the distribution of facial expressions and emotional expression trends within the audience group, used to analyze the overall emotional state of the audience in that region. Sound data refers to the sound environment information generated by all audience members within a region, including sound intensity, spectral distribution, and acoustic atmosphere characteristics, used to assess the overall sound environment and atmosphere of the region. Motion data refers to the individual positional changes and movement trajectories of each audience member within the region, including their movement speed, direction, and temporal changes in their position coordinates, used to identify individual audience member behavior patterns and movement trends.

[0082] Specifically, for acquiring facial data, the data acquisition module 10 identifies all audience faces in the overall image captured by the cameras in each area and extracts the comprehensive facial expression features of the audience group in the area to form facial data that reflects the overall emotional state of the area.

[0083] For acquiring sound data, the data acquisition module 10 collects the overall sound environment signal of the area in real time through sound acquisition devices distributed in various areas, and acquires comprehensive sound data reflecting the acoustic state of the area.

[0084] For acquiring motion data, the data acquisition module 10 uses motion detection sensors and video analysis technology to track the position changes of each audience member in the area and record the coordinate position and movement trajectory of each individual audience member in the time series.

[0085] The aforementioned emotion intensity value calculation module 20 is used to calculate a facial consistency vector based on the facial data of all viewers in each area. The facial consistency vector is used to characterize the overall emotional state of the viewers in each area.

[0086] To quantify the overall consistency and emotional tendency of audience emotions across different regions, a facial consistency vector is introduced as a mathematical representation of emotional states. The facial consistency vector is a multidimensional numerical vector obtained by analyzing the uniformity and emotional tendency of facial expressions within a region. The direction of the vector indicates the positive or negative nature of the emotion, and the magnitude of the vector indicates the intensity of the emotion.

[0087] Specifically, the direction calculation of the facial consistency vector is based on the overall emotional polarity analysis of the region, and the emotional intensity calculation module 20 uses a bipolar emotion recognition algorithm to quantify the positive and negative tendencies of the audience's emotions. By analyzing the overall color distribution and motion patterns of the regional image, the statistical proportions of positive and negative emotional features are identified. Positive emotional features include an increase in the overall brightness of the regional image, an increase in high-frequency motion areas in the image, and a forward shift in audience gathering density. Negative emotional features include a decrease in the overall brightness of the regional image, an increase in static areas in the image, and a backward shift in audience gathering density. To quantify the intensity of positive emotional features, the increase in the overall brightness of the regional image compared to the baseline value, the proportion of high-frequency motion pixels detected in the image, and the forward shift of the audience gathering center of gravity relative to the stage direction are calculated respectively. After normalizing these three regional feature intensities, a weighted sum is obtained to obtain the overall intensity value of the positive emotional features. To quantify the intensity of negative emotion features, the following calculations were performed: the decrease in overall brightness of the region image compared to the baseline value, the proportion of stationary pixels detected in the image, and the backward shift of the audience's center of gravity relative to the stage direction. These three regional feature intensities were normalized and then weighted and summed to obtain the overall intensity value of the negative emotion features. The intensity difference between positive and negative emotion features was calculated. When the intensity of the positive emotion feature was greater than that of the negative emotion feature, the facial consistency vector was taken in the positive direction; when the intensity of the negative emotion feature was greater than that of the positive emotion feature, the facial consistency vector was taken in the negative direction.

[0088] Furthermore, the calculation of the facial consistency vector's modulus is based on the overall expression synchronicity analysis of the region. The emotion intensity calculation module 20 uses a spatiotemporal consistency algorithm to quantify the coordination and intensity of the audience's emotions. It analyzes the overall change pattern of the regional images within consecutive time frames, calculating the temporal synchronicity and spatial uniformity of expression changes. Temporal synchronicity is calculated by analyzing the correlation of overall pixel changes in regional images between adjacent time frames. Specifically, the overall pixel intensity distribution of two consecutive frames of regional images is extracted, and the overall pixel intensity correlation coefficient between the two image frames is calculated. The pixel intensity correlation coefficient is calculated as follows: first, all pixel intensity values ​​of the previous frame's regional image are arranged in row and column order into a one-dimensional array X; then, the corresponding pixel intensity values ​​of the next frame's regional image are arranged into a one-dimensional array Y; and finally, the Pearson correlation coefficient formula is used to calculate the linear correlation between the two arrays.

[0089] The following calculation formula is used:

[0090]

[0091] in and Let X be the mean of the two arrays. i and Y i The pixel intensity value at the corresponding location is calculated, with the result 'r' ranging from -1 to 1. A value closer to 1 indicates a more similar pixel change pattern between the two frames, meaning a higher temporal synchronicity of viewer facial expression changes. Spatial uniformity is calculated by analyzing the variance of pixel activity at different spatial locations within the region image. The region image is divided into several grid cells, and the variance of pixel change intensity in each grid cell is calculated. A variance close to 0 indicates a similar emotional state among viewers within the region. The temporal synchronicity correlation coefficient is weighted and fused with the reciprocal of the spatial uniformity variance, and then multiplied by the absolute value of the intensity difference between the aforementioned positive and negative emotional features to obtain the magnitude of the facial consistency vector. Finally, the direction and magnitude of the facial consistency vector are combined to obtain the facial consistency vector itself.

[0092] The aforementioned emotion intensity value calculation module 20 is used to extract acoustic features from the sound data of all audience members in each area, and calculate acoustic event vectors based on the acoustic features. The acoustic event vectors are used to characterize the overall sound environment of the audience in each area.

[0093] To quantify the overall activity level and acoustic atmosphere characteristics of the audience's sound environment in each area, an acoustic event vector is introduced as a mathematical representation of the sound environment. An acoustic event vector is a multidimensional numerical vector obtained by analyzing the spectral distribution, intensity variation, and temporal characteristics of the audience's sound within a region. The direction of the vector indicates the positivity or negativity of the sound environment, the magnitude of the vector indicates the intensity of sound activity, and each component of the vector reflects environmental characteristics in different acoustic dimensions.

[0094] Specifically, the direction calculation of acoustic event vectors is based on the overall sound atmosphere analysis of the region, and the emotion intensity value calculation module 20 uses a bipolar acoustic recognition algorithm to quantify the positive or negative tendencies of the sound environment. By analyzing the overall spectral characteristics and energy distribution of the regional sound signal, the statistical proportions of positive and negative acoustic features are identified. Positive acoustic features include an upward trend in overall sound energy, an increase in the proportion of energy in the mid-to-high frequency band, and an increase in the frequency of sound activity; negative acoustic features include a downward trend in overall sound energy, an increase in the proportion of energy in the low frequency band, and the appearance of long periods of silence.

[0095] To quantify the intensity of positive acoustic features, the following calculations are performed: the increase in overall sound energy in a region compared to a benchmark, the proportion of mid-to-high frequency energy in the total spectral energy, and the detection frequency of sound activity events per unit time. The overall intensity value of the positive acoustic features is obtained by normalizing the positive feature intensities of these three regions and then performing a weighted sum. Similarly, to quantify the intensity of negative acoustic features, the following calculations are performed: the decrease in overall sound energy in a region compared to a benchmark, the proportion of low-frequency energy in the total spectral energy, and the duration of silent periods per unit time. The overall intensity value of the negative acoustic features is obtained by normalizing the negative feature intensities of these three regions and then performing a weighted sum. Finally, the difference between the intensity of positive and negative acoustic features is calculated. When the intensity of positive acoustic features is greater than that of negative acoustic features, the acoustic event vector is in a positive direction; when the intensity of negative acoustic features is greater than that of positive acoustic features, the acoustic event vector is in a negative direction.

[0096] The magnitude calculation of the acoustic event vector is based on the overall sound activity analysis of the region. The emotion intensity calculation module 20 uses a multi-dimensional acoustic statistical algorithm to quantify the intensity of sound activity. It analyzes the overall energy change pattern of the regional sound signal and calculates the temporal volatility and spectral complexity of the sound intensity. The temporal volatility of sound intensity is calculated by analyzing the standard deviation of sound energy within a continuous time window; a larger standard deviation indicates more active sound activity. Spectral complexity is calculated by analyzing the entropy value of the sound signal spectrum; a larger entropy value indicates richer sound content. The temporal volatility and spectral complexity of the sound intensity are weighted and fused, and then multiplied by the absolute value of the difference between the aforementioned positive and negative acoustic feature intensities to obtain the magnitude of the acoustic event vector.

[0097] Finally, the direction and magnitude of the acoustic event vector are combined to obtain the acoustic event vector.

[0098] The aforementioned emotion intensity value calculation module 20 is used to calculate the overall emotion intensity value of the audience in each area based on the facial consistency vector and the acoustic event vector.

[0099] The emotion intensity value is a comprehensive numerical index obtained by fusing facial consistency vectors and acoustic event vectors. This index is used to quantify the intensity and tendency of the overall emotional state of the audience in a given area. The magnitude of the emotion intensity value reflects the intensity of the audience's emotional expression, and the sign of the value reflects the positive or negative nature of the audience's emotion. This value provides a quantitative basis for subsequent emotion analysis and feedback control.

[0100] Specifically, for any given region: the emotion intensity calculation module 20 first obtains the facial consistency vector and acoustic event vector magnitude values ​​for each region, and then performs classification processing based on preset facial and acoustic thresholds. When the facial consistency vector is greater than or equal to the facial threshold and the acoustic event vector is greater than or equal to the acoustic threshold, it indicates that the audience in that region exhibits significant emotional expression in both visual and auditory dimensions. At this time, the system performs a weighted summation calculation on the magnitude values ​​of the facial consistency vector and the acoustic event vector, balancing the contribution of the two dimensions by setting the same weight coefficient, ensuring that the comprehensive emotion intensity value can fully reflect the audience's true emotional state.

[0101] When the facial consistency vector is greater than or equal to the facial threshold and the acoustic event vector is less than the acoustic threshold, or vice versa, it indicates that there is inconsistency in the audience's emotional expression between the two dimensions. This may be due to environmental factors or behavioral habits that suppress the expression of one dimension. In this case, the dominant emotional expression dimension is determined by calculating the rate of change of the facial consistency vector and the acoustic event vector. The rate of change is obtained by dividing the difference between the magnitude of the facial consistency vector or the magnitude of the acoustic event vector at the current moment and the magnitude of the vector corresponding to the previous moment by the time interval. The system identifies the vector with the larger rate of change as the primary vector and the vector with the smaller rate of change as the secondary vector. The weight allocation is adjusted according to the ratio of the rates of change of the two vectors. That is, the weight of the primary vector is equal to its rate of change divided by the sum of the two rates of change, and the weight of the secondary vector is equal to its rate of change divided by the sum of the two rates of change. Through this dynamic weight allocation mechanism, the primary and secondary vectors are weighted and summed to obtain the overall emotional intensity value of the audience in each region.

[0102] When the facial consistency vector is less than the facial threshold and the acoustic event vector is less than the acoustic threshold, it indicates that the audience's emotional expression is relatively weak or they are in a calm state at the current moment. However, this situation may be a temporary emotional fluctuation. To avoid misjudgment caused by instantaneous data fluctuations, historical facial consistency vectors and historical acoustic event vectors within a preset time window are obtained for this region. A time decay weight is assigned to each historical data point, with the time weights distributed according to an exponential decay function. That is, the weight of the previous sampling moment is the largest, and the weight decreases exponentially with the increase of time intervals. The system calculates the sum of the products of each historical facial consistency vector and its corresponding time weight as the comprehensive value of the historical facial consistency vector, and calculates the sum of the products of each historical acoustic event vector and its corresponding time weight as the comprehensive value of the historical acoustic event vector. After obtaining the two comprehensive historical values, the system uses an equal-weighted approach to perform a weighted summation of the comprehensive values ​​of the historical facial consistency vectors and the comprehensive values ​​of the historical acoustic event vectors to obtain an emotion intensity value based on historical trends.

[0103] The aforementioned preheating area determination module 30 is used to divide each motion data into consecutive frames and calculate the motion vectors of all viewers in the corresponding consecutive frames.

[0104] To quantify the spatial displacement characteristics and velocity of each individual viewer within each region, a motion vector is introduced as a mathematical representation of the individual's motion state. A motion vector is a two-dimensional numerical vector obtained by analyzing the displacement and velocity changes of a single viewer's body within a continuous time frame. The direction of the vector represents the actual direction of movement of the individual viewer in the horizontal plane, and the magnitude of the vector represents the individual viewer's velocity. Each viewer has an independent motion vector to describe their individual spatial motion state.

[0105] Specifically, the preheating area determination module 30 first performs temporal preprocessing on the acquired motion data, dividing the continuous motion data stream into several time segments according to a preset time interval, with each time segment constituting a continuous frame. By analyzing the typical time scale of audience motion behavior, the system sets an appropriate frame length to ensure that each continuous frame can capture meaningful motion changes while maintaining the real-time requirements of the calculation.

[0106] After dividing the motion data into frames, the motion vectors of all viewers in the corresponding consecutive frames are calculated. The motion vector calculation process includes two key steps: position change detection and direction-velocity analysis. For each viewer, their actual displacement vector within the consecutive frames is calculated. By comparing the viewer's coordinates at the start and end of the frame, the net displacement distance in the horizontal and vertical directions is obtained. Vector subtraction is then used to subtract the viewer's starting coordinates from their ending coordinates to obtain the displacement vector representing the viewer's true direction and distance of movement.

[0107] The direction calculation of the motion vector is based on the spatial displacement direction analysis of each individual viewer. The preheating area determination module 30 uses coordinate geometry calculation methods to determine the angle of movement direction for each viewer. By analyzing the displacement vector of each viewer in consecutive frames, the angle of this vector relative to the reference coordinate system is calculated. The angle between the viewer's displacement vector and the positive direction of the horizontal axis is calculated, and the arctangent function is used to obtain the angle value, which ranges from 0 degrees to 360 degrees, corresponding to different spatial movement directions.

[0108] The magnitude of the motion vector is calculated based on the movement speed analysis of each individual viewer. The preheating area determination module 30 uses a speed calculation formula to quantify the movement speed of each viewer. The average movement speed of each viewer is obtained by calculating the total displacement distance of each viewer within a continuous frame time window and dividing this displacement distance by the time window length. The displacement distance is obtained by calculating the Euclidean norm of the viewer's displacement vector, which is the square root of the sum of the squares of the components of the displacement vector, specifically the square root of the sum of the squares of the horizontal and vertical displacements. The instantaneous movement speed of the viewer is obtained by dividing the calculated displacement distance by the duration of the continuous frame; this speed value is used as the magnitude of the viewer's motion vector.

[0109] Finally, the direction and magnitude of the motion vectors are combined to obtain the motion vectors of all viewers in the corresponding consecutive frames.

[0110] The aforementioned preheating area determination module 30 is also used to analyze the direction of each motion vector in each area to obtain directional consistency, and to mark the flow of people with directional consistency greater than the preset collective movement threshold as the target flow of people.

[0111] Specifically, the preheating area determination module 30 first calculates the similarity metric between any two motion vectors within the area. This metric comprehensively considers the differences in vector direction angles and motion speeds. By setting a similarity threshold standard, vectors with similar motion characteristics are grouped into the same cluster group. The clustering process adopts an iterative optimization mechanism. The system continuously adjusts the cluster center positions and cluster boundaries until the vector similarity within each motion vector cluster reaches the optimal state, ultimately forming multiple motion vector clusters.

[0112] The system then calculates the angle between the motion vectors within each motion vector cluster and counts the number of vector pairs whose angles are less than a preset second angle threshold. This step is used to quantitatively evaluate the degree of directional consistency within each motion vector cluster. The system iterates through all motion vectors within each motion vector cluster, calculates the angle between any two motion vectors, and obtains the precise angle using the vector dot product formula and inverse trigonometric functions. The system compares the calculated angle with the preset second angle threshold and counts the number of vector pairs that satisfy the condition that the angle is less than the preset second angle threshold.

[0113] Then, the system calculates the total number of vector pairs in each motion vector cluster. This number is equal to the result of the combination calculation of the total number of vectors in the motion vector cluster. Then, the number of vector pairs with an included angle less than a preset second angle threshold is divided by the total number of vector pairs to obtain the directional consistency of the motion vector cluster.

[0114] Finally, the system compares the directional consistency values ​​of each motion vector cluster with a preset collective movement threshold, filtering out motion vector clusters whose directional consistency exceeds the threshold standard. These motion vector clusters represent crowds with obvious collective movement characteristics. The system marks the motion vector clusters that meet the criteria as target crowds.

[0115] The aforementioned preheating area determination module 30 is used to determine the collective directional movement trend of each target flow of people, and to determine the target area that needs to be preheated based on the collective directional movement trend.

[0116] Specifically, the preheating zone determination module 30 first determines the collective directional movement trend by calculating the average vector of all movement vectors in each target crowd flow. The system sums the movement vectors of all spectators within each marked target crowd flow, then divides the sum by the total number of spectators in that flow to obtain the average vector representing the collective movement direction and average speed of the entire crowd flow. The direction angle of the average vector reflects the main movement direction of the target crowd flow, while the magnitude of the vector represents the overall intensity and degree of directional consistency of the crowd flow movement.

[0117] After obtaining the collective directional movement trend, the preheating area determination module 30 begins to construct a spatial connectivity map describing the spatial structure of the venue.

[0118] like Figure 2 As shown, Figure 2 This application provides an exemplary spatial connectivity graph. The spatial connectivity graph uses a node and edge structure from graph theory to abstractly represent the spatial layout of the venue. The system treats each region as a node in the graph (e.g., ...). Figure 2 The system uses nodes 1, 2, 3, and 4 in the graph as nodes, and the physical connection paths between regions are used as edges in the graph. After constructing the spatial connectivity graph, the system begins to analyze the flow characteristics of the target pedestrian flow within the graph structure. For each target pedestrian flow, the edge closest to the current position of the pedestrian flow is selected as the target edge, and the distance is calculated based on the Euclidean distance between the centroid of the pedestrian flow and the geometric center of the edge. After determining the target edge, the system performs aggregate analysis on all detected motion vectors on that edge, and calculates the instantaneous flow rate and mainstream direction of the target edge using statistical analysis methods. The instantaneous flow rate is obtained by counting the total number of people passing through the edge per unit time, and the mainstream direction is obtained by summing all motion vectors on the edge.

[0119] The preheating area determination module 30 further calculates the net flow divergence of each node on the spatial connectivity graph based on the instantaneous flow and mainstream direction information of each target edge. Net flow divergence reflects the tendency of population aggregation or dispersion in each node region, and the calculation process adopts the concept of divergence from fluid mechanics. The system first counts the inflow and outflow of each node. Inflow is the sum of the instantaneous flows of all edges pointing to that node, and outflow is the sum of the instantaneous flows of all edges emanating from that node. Net flow divergence equals inflow minus outflow; a positive value indicates a tendency for population aggregation at the node, while a negative value indicates a tendency for population dispersion. The absolute value reflects the intensity of aggregation or dispersion.

[0120] After calculating the net flow divergence, the preheating area determination module 30 filters nodes on the spatial connectivity graph to obtain target nodes. The filtering process comprehensively considers multiple factors, including the node's net flow divergence, spatial distance from the target pedestrian flow, and node capacity limitations. The system first excludes nodes with negative net flow divergence, as these nodes exhibit dispersed pedestrian flow characteristics and do not meet preheating requirements. Then, it filters nodes whose direction aligns with the collective directional movement trend of the target pedestrian flow. This is achieved by calculating the direction vector of the node's position relative to the current position of the pedestrian flow and comparing it with the average vector of the pedestrian flow. Nodes with angle differences within a preset range are selected as target nodes.

[0121] The preheating area determination module 30 calculates the passage cost of each target node based on the collective directional movement trend. The passage cost reflects the time and path complexity required for the target pedestrian flow to reach that node. In the specific implementation process, the system first executes a shortest path search algorithm on each target node to calculate the basic passage cost. The shortest path search uses Dijkstra's algorithm, taking the current centroid position of the target pedestrian flow as the starting point and the positions of each target node as the ending point, to find the optimal path in the constructed spatial connectivity graph. The calculation of the basic passage cost comprehensively considers factors such as the length and width of each edge in the path and the current pedestrian density. Specifically, the calculation method is to divide the length of each edge in the path by the effective passage width of that edge, multiply it by the pedestrian density correction factor, and then sum the costs of all edges in the path. The effective passage width considers the physical width of the passage and its current occupancy. The pedestrian density correction factor is determined based on the ratio of the current pedestrian density on the edge to the standard passage density.

[0122] After obtaining the basic passage cost for each target node, the system further calculates the cost correction based on the vector angle between each target node and the collective directional movement trend. The calculation process first determines the direction vector from the current position of the target flow towards each target node, and then calculates the angle between this direction vector and the average vector of the collective directional movement trend. The angle is calculated using the vector dot product formula, obtaining the angle value between the two vectors through the inverse cosine function. When the angle is zero degrees, it indicates that the target node is completely located in the direction of the flow, and the cost correction is negative, resulting in a path cost discount. When the angle is ninety degrees, it indicates that the target node is located in the perpendicular direction of the flow, and the cost correction is zero. When the angle is one hundred and eighty degrees, it indicates that the target node is located in the opposite direction of the flow, and the cost correction is positive and reaches its maximum, penalizing the path cost. The specific value of the cost correction is calculated using the cosine function; that is, the cost correction equals the basic passage cost multiplied by the negative cosine of the angle. The system adds the cost correction to the basic passage cost to obtain the passage cost for each target node.

[0123] Finally, the attractiveness score is calculated by combining the absolute value of the net flow divergence and the travel cost of each target node. The attractiveness score is calculated using a weighted combination method, where the absolute value of the net flow divergence reflects the node's potential for crowd gathering, with a positive weight, while travel cost reflects the difficulty of reaching the node, with a negative weight. The specific formula is: Attractiveness score = Absolute value of net flow divergence multiplied by the gathering weight coefficient minus travel cost multiplied by the distance weight coefficient. The weight coefficients are set based on historical data analysis and expert experience to ensure that the attractiveness score accurately reflects the preheating priority of the nodes. The system identifies one or more target nodes with an attractiveness score greater than the preset preheating score as target areas requiring preheating.

[0124] The aforementioned power prediction module 40 is used to predict the power demand of each target area within the time range of the arrival of each target flow of people in the corresponding target area, based on the emotional intensity value and each target area.

[0125] Specifically, the power prediction module 40 first obtains the base power of electrical equipment in each target area under conditions of no audience. The base power reflects the minimum energy consumption required for the target area to maintain basic operation, including but not limited to the standard brightness of the lighting system, the base air volume of the ventilation system, and the fixed loads such as the routine operation of security monitoring equipment. The system obtains the rated power parameters of various electrical equipment by querying the equipment power database and historical operation records, and calculates the base power based on the actual operation strategy during periods without audiences.

[0126] The system first uses a people counting algorithm to obtain the number of people in the target flow. This algorithm automatically counts the number of people in the target flow area based on image recognition technology, ensuring the accuracy of the headcount data. The system then converts the travel cost of the target flow to the target area into a forward arrival time based on a pre-established standard movement speed coefficient derived from statistical analysis of historical pedestrian movement data. Specifically, the system divides the travel cost by the standard movement speed coefficient, which represents the amount of travel cost consumed per unit of time. This conversion achieves a quantitative mapping from abstract cost values ​​to specific time values. After determining the arrival time of the forward of the target flow, the system calculates the diffusion time based on the number of people in the target flow. The diffusion time reflects the additional time required from the first person entering the target area to the last person fully entering. The calculation of the diffusion time takes into account the inverse relationship between the size of the flow and the throughput efficiency; that is, the larger the number of people in the target flow, the longer the diffusion time is required to completely pass through the passage and entrance to enter the target area. The system uses the arrival time of the target pedestrian flow front as the starting point of the arrival time range and the sum of the arrival time and diffusion time of the target pedestrian flow front as the ending point of the arrival time range, thus forming a complete arrival time range.

[0127] The system calculates the power requirement per person based on a preset baseline power requirement and the target number of people. The baseline power requirement per person represents the standard power demand of each visitor in the target area, taking into account the power consumption of basic services such as lighting, ventilation, and basic safety. The system multiplies the number of people by the baseline power requirement per person to directly obtain the upper limit of the power of the basic equipment.

[0128] The system calculates the initial power requirement of infrastructure within the arrival time range based on the power ceiling and arrival time limit. First, it analyzes the time span characteristics of the arrival time range. Since the arrival time range covers the entire process from the first person entering the target area to the last person entering, the flow of people within this time range exhibits a gradual growth distribution characteristic; that is, the number of people entering initially is small, gradually increasing over time until the total number of people in the target flow is reached. Based on this temporal distribution pattern of the flow of people, the system uses a time-weighted method to calculate the initial power requirement of infrastructure. Specifically, the power ceiling is multiplied by a time weighting coefficient. The time weighting coefficient is set based on the gradual nature of the cumulative arrival of people. At the beginning of the arrival time range, the system sets the initial time weighting coefficient to its minimum weight value. This minimum weight value reflects the minimum power requirement of infrastructure when only the first group of people enters the target area. The system uses a piecewise linear growth method to set the change pattern of the time weighting coefficient, dividing the entire arrival time range into several equal-length time periods. Within each time period, the time weighting coefficient increases linearly according to a predetermined growth rate, which is determined based on the expected speed curve of the incoming flow. The system obtains the average power demand of the infrastructure during the entire arrival time range by integrating the time weighting coefficients. This average power demand is the first power.

[0129] After completing the basic power analysis, the power prediction module 40 begins to process the impact of emotional factors on power demand. The system normalizes the access cost and attractiveness score of the target area and then performs a weighted sum to obtain the emotion transfer coefficient. The emotion transfer coefficient reflects the expected change in the viewer's emotional state as they move from their current location to the target area; higher access costs generally lead to a decrease in emotional state, while a higher attractiveness score may improve the emotional state.

[0130] The system predicts the expected emotional intensity of the target population after they arrive at the target area based on the emotional intensity value and the emotional transfer coefficient. The current emotional intensity value is multiplied by the emotional transfer coefficient to obtain the emotional change, which is then added to the emotional intensity value to arrive at the expected emotional intensity value.

[0131] Subsequently, the power prediction module 40 divides the arrival time range into multiple time segments according to preset time intervals to achieve refined time-series analysis of power demand. Fluctuation analysis is performed on the pedestrian flow speed in each time segment to obtain the acceleration and deceleration periods of the target pedestrian flow. The fluctuation analysis uses a speed gradient detection method, which identifies acceleration and deceleration periods by calculating the rate of speed change between adjacent time segments.

[0132] The power prediction module 40 calculates the percentage of emotional activation during the acceleration phase and the percentage of emotional decay during the deceleration phase, quantifying the impact of different movement phases on audience emotions. The percentage of emotional activation reflects the degree of emotional improvement during the acceleration phase, calculated as the proportion of acceleration duration to total movement time multiplied by a normalized value of the speed increase. The percentage of emotional decay reflects the degree of emotional decline during the deceleration phase, calculated as the proportion of deceleration duration to total movement time multiplied by a normalized value of the speed decrease.

[0133] The system first determines the starting and target points of the asymptotic curve of emotion intensity values. Using the current emotion intensity value as the starting point and the expected emotion intensity value as the target point, it establishes a path from the current emotional state to the expected emotional state. The system then analyzes the characteristics of emotion changes in each time segment. For time segments in the emotion-rising phase, the system uses the emotion activation percentage as the emotion growth weight within that time segment. The emotion activation percentage reflects the degree and speed at which the audience's emotions are aroused within that time segment; a higher percentage indicates a faster emotion rise, and a greater increase in the corresponding instantaneous emotion intensity value. For time segments in the emotion-falling phase, the system uses the emotion decay percentage as the emotion decay weight within that time segment. The emotion decay percentage reflects the degree and speed at which the audience's emotions fade within that time segment; a higher percentage indicates a more significant emotion fall, and a greater decrease in the corresponding instantaneous emotion intensity value. Based on the above weight information, the system uses a weighted interpolation method to calculate the trajectory of emotion intensity changes within each time segment. By combining the instantaneous emotion intensity value of the previous time segment with the emotion activation percentage or emotion decay percentage of the current time segment, the system calculates the instantaneous emotion intensity value at the end of the current time segment. The system repeats this calculation process until all time segments are covered, ultimately forming a complete asymptotic curve of emotional intensity values, which includes the instantaneous emotional intensity values ​​corresponding to each time segment.

[0134] The system divides the range of instantaneous emotional intensity values ​​into multiple emotional level intervals within a time segment according to an equal-interval principle. A baseline power adjustment coefficient is set for the boundary points of each emotional level interval; these baseline values ​​are determined through prior equipment performance testing and audience emotional response experiments. The system treats the interval between two adjacent boundary points as a linear interpolation segment, identifies the start and end boundary points of this interval and their corresponding baseline power adjustment coefficient values, and then calculates the relative position ratio of the current instantaneous emotional intensity value within that interval. This relative position ratio is multiplied by the difference in power adjustment coefficients between the two boundary points to obtain the power adjustment coefficient increment of the current position relative to the starting point. Finally, this increment is added to the baseline power adjustment coefficient of the starting boundary point to obtain the power adjustment coefficient corresponding to the current instantaneous emotional intensity value.

[0135] Then, the system multiplies the power adjustment coefficient corresponding to each interval with the rated power of the atmosphere equipment to obtain the target power value for each time segment.

[0136] The system sorts the time segments in chronological order and uses the target power value corresponding to each time segment as the power control benchmark for the atmosphere equipment within that time period. Discrete target power values ​​are converted into a continuous power scheduling sequence through time axis mapping. The combined power sequence is defined as the second power, which covers the power scheduling scheme for the entire time period after the target pedestrian flow begins to move.

[0137] Ultimately, the system first establishes a time alignment mechanism for three power types, unifying the time stamps of the base power, first power, and second power into the same time coordinate system. Using a time segment synchronization algorithm, the complete time range of the target pedestrian flow arriving at the target area is divided into standardized time units, each corresponding to a power superposition calculation cycle. Within each time unit, the system extracts the power values ​​of the base power, first power, and second power at that time node, and performs power superposition calculation by accumulating these values ​​to obtain the power demand of each target area within the time range of the target pedestrian flow arriving at the target area.

[0138] The aforementioned visualization power distribution module 50 is used to distribute power to each target area according to the power demand of each target area and display the power distribution results on a preset visualization interface.

[0139] Specifically, based on the power demand data of each target area calculated in the aforementioned steps, the system automatically performs power distribution calculations and allocation operations. The system uses the power demand of each target area as input parameters for power distribution control, allocating corresponding power resources to each target area to ensure that the power supply capacity can meet the actual power demand. Then, the execution results of the power distribution operation and the power supply status information of each target area are displayed in real time on a preset visualization interface. The interface displays the power values ​​of atmospheric equipment and basic equipment in each target area, divided into zones.

[0140] like Figure 3 As shown, Figure 3 This is an example diagram of a visual interface provided in an embodiment of this application, combined with... Figure 3 The power distribution of each area in the target venue is visualized on electronic devices. Assume that the target venue has four areas, for example, area A, atmosphere equipment power is a1, basic equipment power is a2; area B, atmosphere equipment power is b1, basic equipment power is b2; area C, atmosphere equipment power is c1, basic equipment power is c2; area D, atmosphere equipment power is d1, basic equipment power is d2.

[0141] In one embodiment, please refer to Figure 4 This paper presents a flowchart illustrating an intelligent visual energy-saving power distribution method. This method can be implemented using a computer program, a microcontroller, or run on an intelligent visual energy-saving power distribution system. The computer program can be integrated into an application or run as a standalone utility application. Specifically, in this embodiment, the method can be applied to a terminal device for intelligent visual energy-saving power distribution. The method includes steps 101 to 109, as follows:

[0142] Step 101: Obtain facial data, voice data, and motion data of all spectators in at least one area of ​​the target venue at the current time;

[0143] Step 102: Calculate the facial consistency vector based on the facial data of all viewers in each area. The facial consistency vector is used to represent the overall emotional state of the viewers in each area.

[0144] Step 103: Extract acoustic features from the sound data of all audience members in each area, and calculate acoustic event vectors based on the acoustic features. The acoustic event vectors are used to characterize the overall sound environment of the audience in each area.

[0145] Step 104: Calculate the overall emotional intensity value of the audience in each region based on the facial consistency vector and the acoustic event vector;

[0146] Step 105: Divide each motion data into consecutive frames and calculate the motion vectors of all viewers in the corresponding consecutive frames;

[0147] Step 106: Analyze the direction of each motion vector in each region to obtain the direction consistency, and mark the flow of people with direction consistency greater than the preset collective movement threshold as the target flow of people;

[0148] Step 107: Determine the collective directional movement trend of each target crowd, and determine the target area that needs to be preheated based on the collective directional movement trend;

[0149] Step 108: Based on the emotional intensity value and each target area, predict the power demand of each target area within the time range of the arrival of each target flow of people in the corresponding target area;

[0150] Step 109: Distribute power to each target area according to the power demand of each target area, and display the power distribution results on the preset visualization interface.

[0151] For details on the specific implementation, please refer to the description in the system section above; further details will not be provided here.

[0152] It should be noted that the system provided in the above embodiments is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0153] This embodiment also discloses an electronic device, as shown in the reference. Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. The electronic device 013 may include: at least one processor 901, at least one communication bus 902, a user interface 903, a network interface 904, and at least one memory 905.

[0154] The communication bus 902 is used to enable communication between these components.

[0155] The user interface 903 may include a display screen, and optionally, the user interface 903 may also include a standard wired interface or a wireless interface.

[0156] The network interface 904 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0157] The processor 901 may include one or more processing cores. The processor 901 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 905, and by calling data stored in the memory 905. Optionally, the processor 901 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array. The processor 901 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 901 and may be implemented as a separate chip.

[0158] The memory 905 may include random access memory (RAM) or read-only memory. Optionally, the memory 905 may include a non-transitory computer-readable storage medium. The memory 905 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 905 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 905 may also be at least one storage device located remotely from the aforementioned processor 901. (Reference) Figure 5 The memory 905, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application for intelligent visual energy-saving power distribution.

[0159] exist Figure 5In the electronic device shown, the user interface 903 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 901 can be used to call an application program for intelligent visual energy-saving power distribution stored in the memory 905. When executed by one or more processors 901, the electronic device 013 performs one or more methods as described in the above embodiments.

[0160] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0161] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0162] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some service interfaces; indirect couplings or communication connections between apparatuses or units may be electrical or other forms.

[0163] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0164] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0165] 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 device (CMD). Based on this understanding, the technical solution of this application, 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 memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0166] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Other embodiments of this disclosure will be readily apparent to those skilled in the art upon consideration of the disclosure in this specification. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. An intelligent, visual, energy-saving power distribution system, characterized in that, The system includes: a data acquisition module, an emotion intensity value calculation module, a preheating zone determination module, a power prediction module, and a power distribution visualization module, wherein: The data acquisition module is used to acquire facial data, voice data, and motion data of all spectators in at least one area of ​​the target venue at the current time. The emotion intensity value calculation module is used to calculate a facial consistency vector based on the facial data of all viewers in each region. The facial consistency vector is used to characterize the overall emotional state of the viewers in each region. The emotion intensity value calculation module is used to extract acoustic features from the sound data of all audience members in each of the regions, and calculate acoustic event vectors based on the acoustic features. The acoustic event vectors are used to characterize the overall sound environment of the audience in each of the regions. The emotion intensity value calculation module is used to calculate the overall emotion intensity value of the audience in each of the aforementioned regions based on the facial consistency vector and the acoustic event vector. The preheating area determination module is used to divide each of the motion data into consecutive frames and calculate the motion vectors of all viewers in the corresponding consecutive frames. The preheating area determination module is used to analyze the direction of each motion vector in each area to obtain directional consistency, and mark the flow of people whose directional consistency is greater than a preset collective movement threshold as the target flow of people; The preheating area determination module is used to determine the collective directional movement trend of each of the target crowds, and to determine the target area that needs to be preheated based on the collective directional movement trend. The power prediction module is used to obtain the basic power of electrical equipment in each target area under no-audience conditions; obtain the number of people in the target flow, and determine the arrival time range based on the number of people and the passage cost of the target flow to the target area; multiply the number of people by the average basic power demand value to obtain the upper limit of the power of the basic equipment, where the average basic power demand value represents the standard power demand of each audience member for the basic equipment in the target area; multiply the upper limit of the power by a time weighting coefficient, where the time weighting coefficient is set based on the gradual law of the cumulative arrival of the flow of people, and adopts a piecewise linear growth method to set the change mode of the time weighting coefficient, dividing the entire arrival time range into several time periods of equal length, and linearly increasing the time weighting coefficient according to a predetermined growth rate in each time period, the growth rate being determined according to the expected speed curve of the flow of people entering; and obtaining the first power of the basic equipment in that time period by integrating the time weighting coefficient over the entire arrival time range. After normalizing the passage cost and attractiveness score of the target area, a weighted sum is obtained to obtain the emotion transfer coefficient; the attractiveness score is equal to the absolute value of net flow divergence multiplied by the aggregation weight coefficient minus the passage cost multiplied by the distance weight coefficient, and the net flow divergence is equal to the inflow flow of each node minus the outflow flow; based on the emotion intensity value and the emotion transfer coefficient, the expected emotion intensity value after the target flow arrives at the target area is predicted; The arrival time range is divided into multiple time segments according to a preset time interval; fluctuation analysis is performed on the flow speed of people in each time segment to obtain the acceleration and deceleration periods of the target flow; the proportion of emotional activation during the acceleration period and the proportion of emotional decay during the deceleration period are calculated; based on the emotional intensity value, the expected emotional intensity value, and the proportion of emotional activation or emotional decay in each time segment, an asymptotic curve of the emotional intensity value is constructed, and the instantaneous emotional intensity value corresponding to each time segment is obtained through the asymptotic curve; the instantaneous emotional intensity value of each time segment is mapped to the corresponding power adjustment coefficient, and the power adjustment coefficient is multiplied by the rated power of the atmosphere device to obtain the target power value in each time segment; the target power values ​​of each time segment are combined in a time sequence to generate the second power of the atmosphere device after the target flow of people begins to move; The time markers of the base power, the first power, and the second power are unified into the same time coordinate system. The complete time range of the target flow arriving at the target area is divided into standardized time units. Each time unit corresponds to a power superposition calculation cycle. Within each time unit, the power values ​​of the base power, the first power, and the second power at that time node are extracted respectively. Power superposition calculation is achieved by accumulating the values ​​to obtain the power demand of each target area within the time range of the target flow arriving at the target area. The visualization power distribution module is used to distribute power to each target area according to the power demand of each target area and display the power distribution results on a preset visualization interface.

2. The intelligent visual energy-saving power distribution system according to claim 1, characterized in that, The emotion intensity value calculation module is used to calculate the overall emotion intensity value of the audience in each of the aforementioned regions based on the facial consistency vector and the acoustic event vector, specifically for: For any of the regions mentioned: When the facial consistency vector is greater than or equal to the facial threshold and the acoustic event vector is greater than or equal to the acoustic threshold, the facial consistency vector and the acoustic event vector are weighted and summed to obtain the emotion intensity value. When the facial consistency vector is greater than or equal to the facial threshold and the acoustic event vector is less than the acoustic threshold, or when the facial consistency vector is less than the facial threshold and the acoustic event vector is greater than or equal to the acoustic threshold, the primary and secondary vectors are determined by comparing the rate of change of the facial consistency vector and the acoustic event vector, and the primary and secondary vectors are weighted and summed to obtain the emotion intensity value. When the facial consistency vector is less than the facial threshold and the acoustic event vector is less than the acoustic threshold, the historical facial consistency vector and historical acoustic event vector of the region are obtained, and the historical facial consistency vector and the historical acoustic event vector are weighted and summed to obtain the emotion intensity value.

3. The intelligent visual energy-saving power distribution system according to claim 1, characterized in that, The preheating area determination module is used to determine the collective directional movement trend of each target pedestrian flow, and to determine the target area that needs to be preheated based on the collective directional movement trend, specifically for: Calculate the average vector of all motion vectors in each of the target pedestrian flows, and determine the average vector as the collective directional movement trend; Construct a spatial connectivity graph, wherein each region is a node and the path between regions is an edge. Select the edge closest to the target pedestrian flow as the target edge, and aggregate all motion vectors on the target edge to calculate the instantaneous flow rate and mainstream direction of the target edge; Based on the instantaneous flow rate and the mainstream direction, calculate the net flow divergence of each node; The process involves filtering the nodes on the spatial connectivity graph to obtain target nodes, and calculating the passage cost of each target node based on the collective directional movement trend. This includes: marking nodes with positive net flow divergence and absolute values ​​greater than a preset sink threshold as flow sinks; calculating the vector angle between each flow sink and the collective directional movement trend, and designating flow sinks with vector angles less than a preset first angle threshold as target nodes; performing a shortest path search algorithm on each target node to calculate its basic passage cost; calculating a cost correction based on the vector angle between each target node and the collective directional movement trend, and adding the cost correction to the basic passage cost to obtain the passage cost of each target node. By combining the absolute value of the net flow dispersion of each target node with the passage cost of each target node, the attractiveness score of each target node is calculated. One or more target nodes whose attractiveness score is greater than the preset preheating score are identified as target areas that need to be preheated.

4. The intelligent visual energy-saving power distribution system according to claim 1, characterized in that, The preheating zone determination module is used to analyze the direction of each motion vector within each zone to obtain directional consistency, and to mark the flow of people whose directional consistency is greater than a preset collective movement threshold as target flow. Specifically, it is used for: Cluster analysis is performed on the motion vectors within each of the aforementioned regions to obtain multiple motion vector clusters; Calculate the angle between each motion vector within each motion vector cluster, and count the number of vector pairs with an angle less than a preset second angle threshold; Based on the ratio of the number of vector pairs to the total number of vector pairs within the motion vector cluster, the directional consistency of each motion vector cluster is calculated. The motion vector clusters whose directional consistency is greater than a preset collective movement threshold are selected as the target pedestrian flow.

5. An intelligent, visual, energy-saving power distribution method, characterized in that, Applied to the intelligent visual energy-saving power distribution system as described in claim 1, the method includes: Acquire facial data, voice data, and motion data of all spectators in at least one area within the target venue at the current time; A facial consistency vector is calculated based on the facial data of all viewers in each of the aforementioned regions. The facial consistency vector is used to characterize the overall emotional state of the viewers in each of the aforementioned regions. Acoustic features are extracted from the sound data of all audience members in each of the aforementioned areas, and acoustic event vectors are calculated based on the acoustic features. The acoustic event vectors are used to characterize the overall sound environment of the audience in each of the aforementioned areas. The overall emotional intensity value of the audience in each of the aforementioned regions is calculated based on the facial consistency vector and the acoustic event vector. The motion data is divided into consecutive frames, and the motion vectors of all viewers in the corresponding consecutive frames are calculated. The direction of each motion vector in each region is analyzed to obtain directional consistency, and the flow of people whose directional consistency is greater than a preset collective movement threshold is marked as the target flow of people; Determine the collective directional movement trend of each of the target pedestrian flows, and determine the target area that needs to be preheated based on the collective directional movement trend; Based on the emotional intensity value and each of the target areas, predict the power demand of each target area within the time range of the arrival of each target pedestrian flow in the corresponding target area; Power is distributed to each target area according to its power demand, and the distribution results are displayed on a preset visualization interface.

6. An electronic device, characterized in that, The device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in claim 5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed, perform the method as described in claim 5.

Images