Method for dynamic integration of public space lighting loads based on edge node multiscale clustering

Abstract

This application relates to the field of lighting load technology, and in particular to a dynamic integration method for public space lighting load based on multi-scale clustering of edge nodes. The method includes: acquiring pedestrian flow distribution information and a fountain performance mode information set; based on the pedestrian flow distribution information and the fountain performance mode information set, analyzing the coupling relationship between the dynamic flow direction of pedestrians and the focal area of ​​the fountain art to obtain a pedestrian-scenery collaborative lighting demand information set; acquiring real-time environmental disturbance information; based on the real-time environmental disturbance information and the pedestrian-scenery collaborative lighting demand information set, analyzing the interference caused by environmental wind disturbance affecting the fountain water columns and thus the lighting effect to obtain an environmental disturbance information set; and generating a multi-scale collaborative control strategy based on the environmental disturbance information set, and outputting a dynamic integration log of public space lighting load. This enhances the expressive power of light and shadow art in public spaces and the immersive experience for audiences.

Description

Technical Field

[0001] This application relates to the field of lighting load technology, and in particular to a method for dynamic integration of public space lighting loads based on multi-scale clustering of edge nodes. Background Technology

[0002] Currently, the control of public space lighting (especially landscape lighting such as square fountains) mostly adopts preset programs or static strategies based on simple environmental sensing. Common methods include timer-based on-off control, zone dimming, or unified adjustment of lamp brightness and color through a central server according to fixed performance modes.

[0003] However, there are still significant shortcomings in dealing with the real-time changes in complex public spaces such as plaza fountains. First, current methods mostly adopt static or single-factor response modes, which cannot effectively handle the strong coupling dynamic relationship between the distribution of pedestrian density, fountain performance mode (water shape, music) and environmental conditions (such as wind speed and wind direction changing the water jet landing point, water mist affecting visibility), resulting in a mismatch between the lighting effect and the real-time scene. Second, in terms of ensuring the overall artistic effect, existing technologies lack an intelligent collaborative mechanism for multiple lamps on the edge side, making it difficult to adjust the light angle, focus range and color in real time according to the changes in the fountain water shape caused by wind disturbance, and also unable to dynamically balance the lighting load according to the area where people gather, which easily leads to visual disharmony and energy waste. Summary of the Invention

[0004] This application provides a method for dynamic integration of public space lighting load based on multi-scale clustering of edge nodes to solve the above-mentioned problems. The method includes: acquiring pedestrian distribution information and fountain performance mode information set; based on the pedestrian distribution information and the fountain performance mode information set, analyzing the coupling relationship between the dynamic flow direction of pedestrians and the focal area of ​​the fountain art to obtain a human-scenery collaborative lighting demand information set; acquiring real-time environmental disturbance information; based on the real-time environmental disturbance information and the human-scenery collaborative lighting demand information set, analyzing the interference caused by environmental wind disturbance of the square fountain water column, which affects the lighting effect, to obtain an environmental disturbance information set; and generating a multi-scale collaborative control strategy based on the environmental disturbance information set, and outputting a dynamic integration log of public space lighting load.

[0005] Through the above technical solutions, the dynamic integration of pedestrian flow, performance and environmental information enables intelligent collaboration between the lighting system and people, scenery and environment. This significantly enhances the expressive power of light and shadow art in public spaces and the immersive experience for the audience. It also proactively compensates for environmental factors such as wind disturbance to ensure stable effects, achieves refined energy saving through on-demand lighting, and provides data support for intelligent operation and maintenance management through integrated logs.

[0006] Optionally, based on the pedestrian distribution information and combined with the fountain performance mode information set, the coupling relationship between the dynamic flow direction of pedestrians and the focal area of ​​the fountain art is analyzed to obtain a human-scene coordinated lighting demand information set, including: the pedestrian distribution information includes crowd density and crowd movement direction; the fountain performance mode information set includes fountain performance art information and fountain human-scene interaction information; based on the crowd density and combined with the fountain performance art information, the attention priority of the core art presentation area of ​​the fountain corresponding to different crowd density intervals is analyzed to obtain density-art focus matching information; based on the crowd movement direction and combined with the fountain human-scene interaction information, the overlap range and movement sequence of the crowd flow trajectory and the interactive area of ​​the fountain are analyzed to obtain trajectory-interaction node adaptation information; based on the density-art focus matching information and combined with the trajectory-interaction node adaptation information, the coordinated adaptation relationship between the degree of crowd gathering, the focus of art presentation and the accessibility of interaction under different spatial locations over time is analyzed to obtain the human-scene coordinated lighting demand information set.

[0007] Optionally, the process of constructing the density-art focus matching information includes: dividing the crowd distribution area into different density levels based on the crowd density; analyzing the presentation intensity and spatial position of multiple core art elements in the fountain performance based on the fountain performance art information to obtain art focus area information; analyzing the changing patterns of the overall field of vision and attention concentration of the crowd at each density level based on the different density levels and in combination with the art focus area information to obtain crowd visual change information; and analyzing the priority sequence of lighting attention weights that should be allocated to each art focus area to adapt to the viewing effect of crowds at different density levels based on the crowd visual change information and in combination with the art focus area information to obtain the density-art focus matching information.

[0008] Optionally, the process of constructing the trajectory-interaction node adaptation information includes: based on the crowd movement direction and combined with the fountain human-scene interaction information, analyzing the spatial intersection and overlap between the crowd movement path and the interactive area of ​​the fountain to obtain trajectory-area spatial coupling information; based on the trajectory-area spatial coupling information, analyzing the temporal correspondence between the movement speed and the triggering mechanism of the interactive facility when the crowd moves to different intersection areas to obtain a spatiotemporal synchronization trigger sequence; based on the spatiotemporal synchronization trigger sequence, analyzing the dynamic interactive network and experience continuity formed by different flow trajectories triggering different interactive nodes in time sequence to obtain the trajectory-interaction node adaptation information.

[0009] Optionally, based on the density-art focus matching information and combined with the trajectory-interaction node adaptation information, the analysis of the collaborative adaptation relationship between crowd gathering degree, art presentation focus and interactive accessibility at different spatial locations over time, to obtain the human-scene collaborative lighting demand information set, includes: based on the density-art focus matching information and combined with the trajectory-interaction node adaptation information, analyzing the synchronous evolution trend of expected changes in crowd density, changes in art focus attention weight, and changes in the trigger probability of interaction nodes at each spatial location within a preset time period, to obtain spatiotemporal dynamic evolution information; based on the spatiotemporal dynamic evolution information, analyzing the demand fluctuations for lighting intensity, lighting color, and lighting response speed at different spatial locations over time due to crowd gathering, art focus shifts, and interactive activities, to obtain location-time-varying demand fluctuation information; based on the location-time-varying demand fluctuation information, constructing a dynamic map of human-scene collaborative lighting demand with time as the main axis and space as the layer, to form the human-scene collaborative lighting demand information set.

[0010] Optionally, based on the real-time environmental disturbance information and combined with the human-scene collaborative lighting demand information set, the interference caused by environmental wind disturbance affecting the square fountain water column, resulting in an impact on the lighting effect, is analyzed to obtain the environmental disturbance information set, including: the real-time environmental disturbance information includes wind speed and wind direction; based on the wind speed, the degree of suppression of the preset spray height of the fountain water column and the degree of aggravation of the water mist diffusion range are analyzed to obtain the wind strength disturbance factor; based on the wind direction and combined with the human-scene collaborative lighting demand dynamic map, the offset of the water column landing point pulled by the wind direction and the continuous spatial area covered by the main dispersion path of the water mist are analyzed to obtain... The wind direction influence area; based on the wind strength disturbance factor, the wind direction influence area, and the art focus attention weight, the dynamic attenuation law of the visual presentation integrity and color saturation of the art focus in the spatial area caused by changes in water column shape, drop point shift, and accompanying water mist is analyzed to obtain focus interference information; based on the focus interference information and the location time-varying demand fluctuation information, the dynamic correction and compensation of lighting intensity, color ratio, and lighting angle required in the wind direction influence area to offset the wind disturbance influence and maintain the expected human-scenery collaborative lighting effect is analyzed to obtain the environmental interference information set.

[0011] Optionally, the process of constructing the wind direction influence area includes: based on the wind direction influence area and combined with the dynamic map of human-scenery collaborative lighting needs, analyzing the degree of spatial deviation between the visual center and the original artistic focal area after the water column's landing point shifts, and obtaining focal point shift deviation information; based on the focal point shift deviation information, analyzing the main dispersion path of the water mist generated by the water column shift, and the diffusion profile of the water mist concentration changing over time along the main dispersion path of the water mist, and obtaining water mist diffusion dynamic information; based on the water mist diffusion dynamic information, analyzing the scattering and absorption effects of the continuous spatial area covered by the water mist on the lighting light, and the cumulative interference range on the satisfaction of regional lighting needs over time, and obtaining the wind direction influence area.

[0012] Optionally, the process of constructing the focusing interference information includes: based on the wind strength disturbance factor, analyzing the dynamic process of the fountain water column changing from a preset straight shape to a curved, dispersed, or fragmented state under the action of wind, and obtaining a water column deformation sequence; based on the wind direction influence area, combined with the water column deformation sequence, analyzing the continuous offset trajectory of the visual landing point of the water column due to deformation relative to the original artistic focal area, and obtaining a focal offset path; based on the focal offset path, combined with the artistic focal attention weight, analyzing the gradual decrease in the visual clarity of key artistic elements caused by water column shape distortion and landing point deviation along the offset path, and obtaining an integrity attenuation gradient; based on the integrity attenuation gradient, analyzing the synchronous attenuation trend of color purity and brightness of the illumination light in the water mist medium caused by enhanced water mist diffusion and water column shape scattering, and obtaining a color saturation attenuation curve; based on the integrity attenuation gradient, combined with the color saturation attenuation curve, fusing to form a complete law describing the dynamic attenuation of the visual presentation quality of the artistic focal point under wind disturbance with time and space, and obtaining the focusing interference information.

[0013] Optionally, the specific implementation of the dynamic correction compensation includes: based on the focus offset path and combined with the integrity attenuation gradient, analyzing the continuous spatial range where artistic expression fails due to water column deformation along the offset path to obtain the core lighting compensation area; based on the color saturation attenuation curve and combined with the water mist diffusion dynamic information, analyzing the different spectral intensities required to maintain color purity within the water mist coverage area to obtain the color compensation spectral formula; based on the core lighting compensation area, the color compensation spectral formula, and combined with the position time-varying demand fluctuation information, generating a collaborative control instruction set to synchronously adjust the lighting point and color within the wind direction influence area to complete the specific implementation of the dynamic correction compensation.

[0014] Optionally, the step of generating a multi-scale collaborative control strategy based on the environmental interference information set and outputting a dynamic integration log of public space lighting load includes: based on the core lighting compensation area and combined with the color compensation spectral formula, analyzing the differentiated control schemes required for lighting intensity enhancement and spectral correction at the spatial scale for the fountain art focal area and the continuous surrounding area affected by water mist diffusion, to obtain a spatial scale compensation strategy; based on the collaborative control instruction set and combined with the spatiotemporal dynamic evolution information, analyzing the timing of activation, duration of action, and fade-in / fade-out rhythm of each control instruction in the spatial scale compensation strategy according to the temporal characteristics of crowd flow, art focal point shift, and wind disturbance changes, to obtain a temporal scale arrangement strategy; based on the spatial scale compensation strategy and combined with the temporal scale arrangement strategy, analyzing the differences in the capabilities of different edge lighting nodes in terms of spatial coverage and control response speed, decomposing the composite control task and allocating it to an edge node cluster with corresponding scale response capabilities, generating the multi-scale collaborative control strategy, and outputting the dynamic integration log of public space lighting load. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application; Figure 2 A flowchart of a method for dynamic integration of public space lighting load based on multi-scale clustering of edge nodes, provided in an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0018] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0019] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0020] In the process of controlling lighting for fountains in public spaces, existing landscape lighting such as plaza fountains mostly adopts preset programs or static control strategies, relying on timed, zoned dimming, and uniform mode control. This approach cannot adapt to the coupled dynamic changes in pedestrian flow, fountain performances, and wind conditions, lacks intelligent edge coordination of lighting fixtures, and struggles to match scene requirements in real time, easily leading to visual disharmony and energy waste.

[0021] Based on this, this application provides a dynamic integration method for public space lighting load based on multi-scale clustering of edge nodes. It integrates data on pedestrian flow, performances and environment to drive intelligent collaboration between the lighting system and people, scenery and environment, enhance the artistic quality and immersive experience of light and shadow in public spaces, actively compensate for environmental interference such as wind disturbance, ensure stable output of light and shadow effects, and achieve high efficiency and energy saving through precise control on demand. The generated operation log provides full-dimensional data support for intelligent operation and maintenance.

[0022] Figure 1 This application provides an illustration of an application scenario. In the process of controlling the lighting of a public space fountain, the method provided in this application is applied to integrate data on pedestrian flow, performance, and environment to achieve intelligent coordination of the lighting system, enhance the light and shadow art and immersive experience of the public space, actively compensate for interference such as wind disturbance, ensure stable effects, control energy efficiently and on demand, and provide data support for intelligent operation and maintenance through the operation log.

[0023] Specifically, the method provided in this application can be applied to any server. The server interacts with smart cameras, smart control platforms, and wind speed and direction sensors to obtain information on pedestrian distribution provided by smart cameras, information on fountain performance modes provided by smart control platforms, and real-time environmental disturbance information provided by wind speed and direction sensors. It dynamically integrates pedestrian flow, performance, and environmental information to achieve intelligent collaboration between the lighting system and people, scenery, and environment. It generates and outputs a dynamic integrated log of public space lighting load to lighting equipment control personnel, thereby improving the targeting and humanization of lighting control.

[0024] For specific implementation details, please refer to the following examples.

[0025] Figure 2This is a flowchart illustrating a method for dynamically integrating public space lighting loads based on multi-scale clustering of edge nodes, provided in an embodiment of this application. The method of this embodiment can be applied to servers in the above scenario. Figure 2 As shown, the method includes: S201. Obtain pedestrian flow distribution information and fountain performance mode information set. Based on the pedestrian flow distribution information and the fountain performance mode information set, analyze the coupling relationship between the dynamic flow direction of pedestrian flow and the focal area of ​​fountain art to obtain the pedestrian-scenery coordinated lighting demand information set.

[0026] The pedestrian distribution information can be a set of information reflecting the spatial distribution and dynamic changes of people in different areas of the public space, such as the number of people, their movement trajectories, and the duration of their stay. This information is sourced from smart cameras deployed within the public space. The fountain performance mode information set can be a set of parameters reflecting the fountain's operational status, such as the water spray height, spray pattern, spray frequency, performance time, and color scheme. This information is sourced from the fountain's intelligent control platform. The human-scenery collaborative lighting demand information set can be a set of adaptive lighting demand information matching the human-scenery interaction needs of the public space.

[0027] Specifically, in the process of exploring the dynamic integration of lighting load in public spaces, the existing lighting control only adopts a fixed strategy and does not combine the dynamic flow of people with the coordinated planning of fountain performances, resulting in a disconnect between lighting and landscape and insufficient creation of artistic atmosphere. By coupling and analyzing the information of people flow distribution and fountain performance patterns, we can explore the real lighting needs of people-landscape coordination, provide accurate data support for subsequent control, and avoid blind control from the source.

[0028] S202. Obtain real-time environmental disturbance information. Based on the real-time environmental disturbance information and combined with the human-scenery collaborative lighting demand information set, analyze the interference caused by the environmental wind disturbance of the square fountain water column, which affects the lighting effect, and obtain the environmental disturbance information set.

[0029] Real-time environmental disturbance information can be a set of environmental parameters, such as wind speed and direction within the public space, that can affect the fountain's operation, with wind speed and direction sensors deployed within the public space serving as the data source. The environmental interference information set can be a collection of various interference characteristics that characterize the impact of environmental disturbances on lighting effects.

[0030] Specifically, in the interference analysis of dynamic integration of public space lighting load, wind disturbance in the open environment of public space can change the shape of fountain water columns, thereby damaging the lighting effect. Existing control measures ignore this environmental interference, which can easily cause misalignment of lighting focus. By combining the needs of human and landscape lighting to analyze the specific impact of environmental disturbance, the characteristics of interference can be clarified, providing a reliable basis for the subsequent formulation of targeted control strategies.

[0031] S203. Based on the environmental interference information set, generate a multi-scale collaborative control strategy and output a dynamic integrated log of public space lighting load.

[0032] A multi-scale collaborative control strategy can be a set of lighting load control measures formulated based on the lighting needs of public spaces at different scales and in response to environmental disturbances. A dynamic integration log of public space lighting load can be a collection of log information recording various data indicators during the dynamic integration process of lighting load.

[0033] Specifically, in the process of implementing and recording the strategy for dynamic integration of lighting load in public spaces, the existing fixed control strategies cannot flexibly adapt to actual scenarios due to lighting needs and environmental interference. By generating multi-scale collaborative control strategies through interference information, dynamic integration of lighting load is achieved. At the same time, control logs are output to ensure accurate implementation of the strategy and traceability of the process, thereby improving the intelligence of lighting control and the convenience of operation and maintenance.

[0034] The method provided in this embodiment dynamically integrates information on pedestrian flow, performance, and environment to achieve intelligent collaboration between the lighting system and people, scenery, and environment. This can significantly enhance the expressive power of light and shadow art in public spaces and the immersive experience for the audience. It actively compensates for environmental factors such as wind disturbance to ensure stable effects and achieves refined energy saving through on-demand lighting. The output integrated log provides data support for intelligent operation and maintenance management.

[0035] In some embodiments, the crowd distribution information includes crowd density and crowd movement direction; the fountain performance mode information set includes fountain performance art information and fountain-scenery interaction information; based on crowd density and combined with fountain performance art information, the attention priority of the core art presentation area of ​​the fountain corresponding to different crowd density intervals is analyzed to obtain density-art focus matching information; based on crowd movement direction and combined with fountain-scenery interaction information, the overlap range and movement sequence of crowd flow trajectory and interactive area of ​​the fountain are analyzed to obtain trajectory-interaction node adaptation information; based on density-art focus matching information and combined with trajectory-interaction node adaptation information, the collaborative adaptation relationship of crowd gathering degree, art presentation focus and interactive accessibility under different spatial locations over time is analyzed to obtain a human-scenery collaborative lighting demand information set.

[0036] Crowd density refers to the number of people per unit area of ​​public space, measuring the degree of crowd gathering. Crowd movement direction refers to the overall or partial direction of movement of people within a public space. Fountain performance art information refers to various information that reflects the artistic presentation effect during a fountain performance. Fountain human-scene interaction information refers to information representing interactive behaviors between fountain facilities and people within a public space. Density-art focus matching information refers to the corresponding matching relationship established between different crowd density intervals and the attention priority of the core artistic presentation area of ​​the fountain. Trajectory-interaction node adaptation information refers to the adaptation relationship formed by the overlap range of crowd flow trajectories and interactive areas of the fountain, the movement sequence, and the dynamic characteristics of different flow trajectories triggering interactive nodes.

[0037] Specifically, in the process of controlling the lighting of public space fountains, if the coupling relationship between the dynamic flow of people and the focal point of the fountain art is not analyzed, it will lead to a misalignment between the lighting focus and the crowd's attention, a disconnect between the interactive experience and the lighting resources, a waste of lighting resources, and an inability to adapt to the dynamic changes of people and scenery, causing the lighting control to lose its basis in meeting actual needs. To address the above problems, a hardware and software integrated analysis platform was constructed. First, on the hardware side, a camera array deployed around the square is used to extract the pixel coordinates and motion vectors of each individual by integrating real-time video streams from computer vision models such as YOLO-v5 or DeepSORT. Then, a kernel density estimation algorithm is used to generate a heat map of the crowd density of the entire venue, and the convergence direction of the vector field is analyzed by optical flow method to obtain structured crowd density level zoning (such as high, medium, and low) and mainstream movement direction data. At the same time, pre-programmed performance scripts (fountain performance art information) are read in real time from the central control server of the fountain. These scripts are in XML format. The format defines the height of the main water jet, the number of the accompanying water features, the intensity label of the background music, and the corresponding three-dimensional spatial coordinates for different time slots. Interactive information comes from pre-digitally modeled geofences of the interactive area (such as the boundary coordinates of a fountain water corridor whose lighting can be triggered by infrared sensors). In the software analysis layer, for "density-artistic focus matching," the DBSCAN algorithm is used to spatially cluster the crowd coordinates, combining this with heatmaps to classify density levels. Simultaneously, the performance script is parsed, extracting the top N coordinates with the highest water feature intensity within each time slot as artistic focuses. Subsequently, a visual attention mechanism (such as Saliency) is applied. The model (Detection) simulates the collective visual field and attention decay curves of crowds at different densities (e.g., visual attention radius compression at high densities). It calculates the visual saliency score of each art focus within the current density partition and normalizes it into lighting attention weights, thereby generating "density-art focus matching information." For "trajectory-interaction node adaptation," it smooths and predicts the motion trajectories extracted from consecutive frames (e.g., using Kalman filtering), performs spatial overlay analysis on the predicted trajectory lines and predefined interaction area geofences (using the ST_Intersects function of the spatial database), calculates the intersection probability and expected arrival time, and simultaneously... The average moving speed of the trajectory near the intersection point is analyzed and mapped to preset modes such as "fast passage - instantaneous trigger" and "slow stop - continuous interaction" to generate "trajectory-interaction node adaptation information" containing spatiotemporal trigger expectations. Finally, through a spatiotemporal fusion model, the above two sets of information are aligned and fused in the same spatiotemporal coordinate system (with the performance time axis as the horizontal axis and the square two-dimensional grid as the vertical axis). The composite demand fluctuations of each grid unit at different times due to the combined effects of crowd gathering expectations, art focus weight and interaction trigger probability on lighting intensity, color temperature and dynamic response speed are quantitatively calculated. Finally, the integrated output is a structured "human-scene collaborative lighting demand information set".

[0038] The method provided in this embodiment accurately matches the visual and interactive needs of the audience with the artistic presentation needs of the fountain, making the lighting requirements more in line with the actual scene, providing a scientific and accurate basis for subsequent lighting control, and improving the targeting and humanization of lighting control.

[0039] In some embodiments, the population distribution area is divided into different density levels based on population density; based on the fountain performance art information, the presentation intensity and spatial position of multiple core art elements in the fountain performance are analyzed to obtain art focus area information; based on different density levels, combined with the art focus area information, the variation pattern of the overall field of vision and attention concentration of the population under each density level is analyzed to obtain population visual change information; based on the population visual change information, combined with the art focus area information, the priority sequence of lighting attention weights that should be allocated to each art focus area to adapt to the viewing effect of the population at different density levels is analyzed to obtain density-art focus matching information.

[0040] Core artistic elements can be key components that constitute the artistic presentation of a fountain performance, such as combinations of water jets with specific shapes, layers of water curtains at varying heights, and the sequence of water jets changing in rhythm with the music. Presentation intensity refers to the visual prominence of each core artistic element in a fountain performance, which can be comprehensively characterized by indicators such as water jet height, water flow rate, and lighting coordination. Information on artistic focal areas can be a comprehensive set of information including the presentation intensity, spatial coordinate range, and visual core points of each core artistic element of the fountain performance. Crowd visual change information can be a systematic analysis of the overall visible range, visual focal points, and level of attention of crowds in public spaces at different density levels, showing how these changes correlate with crowd density. The lighting attention priority sequence can be an ordered sequence assigned to each artistic focal area, representing the priority of lighting resource allocation.

[0041] Specifically, in the process of controlling lighting for public space fountains, if density-artistic focal point matching information is not constructed, it will lead to a disconnect between lighting resource allocation and crowd density and artistic focal point needs. This results in insufficient lighting in high-density areas and wasted resources in low-density areas, significantly diminishing the viewing experience for the crowd and rendering subsequent lighting demand analysis unscientific. To address these issues: First, using a network of cameras deployed at computing nodes along the plaza's edge, video streams are processed in real-time using deep learning-based object detection and tracking algorithms (such as the YOLO series). The number of people in each grid area (e.g., dividing the plaza into 10m x 10m grids) is counted, continuously outputting a crowd density heatmap. Then, an unsupervised clustering algorithm (such as the DBSCAN density clustering algorithm) is used to analyze the values ​​in this heatmap, based on a preset threshold (e.g., more than 1 person / m²). 2 Designated as a high-density area, with a density ranging from 0.2 to 1 person / m². 2This is a medium-density area, less than 0.2 people / m². 2 The system automatically and dynamically divides the plaza space into zones of different density levels (for low-density areas). Simultaneously, it acquires real-time fountain performance art information from the central control system via standard protocols (such as OPC UA). This structured data explicitly describes all activated water features in the current performance (e.g., "sky-high water jets," "rotating showers"), their corresponding pump and valve numbers, preset lighting colors (RGB values), and dynamic change scripts. Based on this, a visual saliency calculation model (e.g., an improved application of the Itti-Koch visual attention model) is used to analyze the fountain's two-dimensional planar layout and three-dimensional spatial design. This quantitatively calculates the visual appeal weight of each artistic element within the current composition and its core coordinate range in space (i.e., the artistic focal area). Then, for each divided density level zone, an empirical model of crowd visual attention is introduced. This model integrates environmental psychology and visual perception. The research findings in geometry describe, in a parametric manner, the average effective visual field radius of the crowd at this density (e.g., the visual field radius may shrink to within 5 meters in high density), the range of the viewing angle, and the duration of attention to dynamic targets. Finally, the spatial location and visual weight of the above-mentioned "artistic focus area" are spatially superimposed and matched with the actual visible range and attention parameters within this density area as defined by "crowd visual change information". Through a multi-objective optimization algorithm (such as using a weighted summation method or a Pareto optimal search), with the objective function of "maximizing the total visual weight of the artistic focus that the crowd can perceive within the visible range of the current density crowd", a lighting attention weight priority sequence is dynamically calculated and output, which is the density-artistic focus matching information.

[0042] The method provided in this embodiment accurately matches the crowd density with the artistic focal lighting needs, realizes the rational allocation of lighting resources, conforms to the visual characteristics of crowds at different densities, enhances the fountain viewing experience, and lays a precise data foundation for subsequent human-scene collaborative lighting analysis.

[0043] In some embodiments, based on the direction of crowd movement and combined with fountain human-scene interaction information, the spatial intersection and overlap of the crowd movement path and the interactive area of ​​the fountain are analyzed to obtain trajectory-area spatial coupling information; based on trajectory-area spatial coupling information, the temporal correspondence between the movement speed and the triggering mechanism of the interactive facility when the crowd moves to different intersection areas is analyzed to obtain a spatiotemporal synchronous triggering sequence; based on the spatiotemporal synchronous triggering sequence, the dynamic interactive network and experience continuity formed by different flow trajectories triggering different interactive nodes in time sequence are analyzed to obtain trajectory-interactive node adaptation information.

[0044] Fountain-human interaction information can be a collection of relevant information within the fountain performance system that enables interaction between the crowd and the fountain. Trajectory-area spatial coupling information can be a quantified and descriptive information of the spatial positional relationships (intersection, overlap, adjacency, etc.) between the crowd's movement path and the interactive areas of the fountain in two-dimensional / three-dimensional space. Interactive facility triggering mechanisms can be the conditions and rules for activating various interactive facilities within the interactive area of ​​the fountain. Spatiotemporal synchronization triggering sequences can be ordered sequence information formed when the interactive facility triggering mechanisms match the timing of crowd movement as they arrive at different spatial intersection areas of the fountain at different movement speeds. Flow trajectories can be the continuous spatial displacement paths of a single or group of people in a public space, representing the continuous presentation of the crowd's movement direction in the time dimension. Interactive nodes can be the smallest spatial unit and facility combination within the interactive area of ​​the fountain that can independently realize an interactive experience. Dynamic interactive networks can be the interconnected and cooperative interactive relationship network formed between different interactive nodes after different crowd flow trajectories trigger different interactive nodes in a temporal sequence. Experience continuity can be the seamless and uninterrupted state characteristic of the experience process from one interactive node to another during the participation of the crowd in fountain-human interaction.

[0045] Specifically, in the process of human-scene interaction in public space fountains, if the adaptation relationship between crowd movement and interaction nodes is not analyzed, lighting will fail to follow the rhythm of crowd flow, resulting in a mismatch between lighting supply and demand in interactive areas, leading to insufficient lighting or waste of resources in key areas, severely reducing the lighting experience and collaborative effect of human-scene interaction. To address these issues: First, real-time acquired UWB positioning data or video trajectory tracking data is fitted into a smooth, continuous movement path using a cubic spline interpolation algorithm; simultaneously, based on the BIM model or CAD drawings of the fountain facilities, ArcGIS is used... The Engine's spatial data processing component precisely digitizes the interactive areas defined in the "fountain human-scene interaction information" (e.g., an interactive floor with a diameter of 2 meters and equipped with pressure sensors) into a polygonal feature layer. Then, applying buffer analysis and overlay analysis techniques from GIS, a dynamically changing buffer zone is generated, centered on points along each path (e.g., the buffer zone radius is adjusted between 1.5 and 3.5 meters based on crowd movement speed to simulate the potential real-time interaction range of individuals). This dynamic buffer zone layer is then spatially intersected with the interactive area polygon layer to accurately calculate the specific geographic coordinates, overlapping areas, and geometric center point sequences of all trajectory segments and each interactive area, forming structured trajectory-area spatial coupling information. Furthermore, to address the temporal synchronization problem, a Hidden Markov Model (HMM) is constructed for deduction: each spatial intersection area obtained in the previous step is defined as a "hidden state," and the average distance people pass through this area is... The movement speed (e.g., 1.2 m / s calculated by taking the time difference and distance between consecutive positioning points) and the fixed triggering delay of the interactive facilities in the area from perception to execution (e.g., after the infrared sensor detects a human body, the main control unit needs 0.8 seconds to drive the solenoid valve) are used together as the "observation state". The state transition probability matrix of the model is trained using historical data (e.g., the probability of a crowd moving from area A to area B). The forward-backward algorithm is used to calculate the most likely state sequence under the current observation sequence. Finally, the Viterbi decoding algorithm is used to predict the interactive events most likely to be triggered sequentially in the future and their precise time points, generating a spatiotemporally synchronized triggering sequence. Finally, based on this time series data, the temporal network modeling method in complex network analysis is adopted. A dynamic directed graph is constructed using Python's NetworkX library: each predicted triggering event is instantiated as a network node, and the node attributes include its spatial location and trigger timestamp; directed edges are established between consecutive triggering event nodes according to the actual flow of the crowd.Based on this, the topological indicators of the entire network are calculated. For example, the key hub interaction points connecting different flow paths are identified by calculating the "betweenness centrality" of nodes. At the same time, for a single flow path, the standard deviation of the time interval between all consecutively triggered events on the path is calculated (for example, the standard deviation threshold is set to within 1.5 seconds). This is used as the core indicator for quantitatively evaluating "experience continuity". Finally, spatial coupling, temporal synchronization and network continuity indicators are integrated to output multi-dimensional trajectory-interaction node adaptation information.

[0046] The method provided in this embodiment accurately captures the spatiotemporal adaptation patterns of crowd flow and fountain interaction, allowing the lighting strategy to match the interaction rhythm, achieving refined allocation of lighting resources, improving the accuracy of lighting in interactive areas, and ensuring the continuity of the lighting experience for human-scene interaction.

[0047] In some embodiments, based on density-art focus matching information and trajectory-interaction node adaptation information, the synchronous evolution trend of expected changes in crowd density, changes in the weight of art focus attention, and changes in the trigger probability of interaction nodes at various spatial locations within a preset time period is analyzed to obtain spatiotemporal dynamic evolution information. Based on the spatiotemporal dynamic evolution information, the demand fluctuations for lighting intensity, lighting color, and lighting response speed at different spatial locations in the time dimension caused by crowd gathering, art focus shifting, and interactive activities are analyzed to obtain location-time-varying demand fluctuation information. Based on the location-time-varying demand fluctuation information, a dynamic map of human-scene collaborative lighting demand with time as the main axis and space as the layer is constructed to form a human-scene collaborative lighting demand information set.

[0048] The collaborative adaptation relationship can refer to the mutual matching and adaptation among three factors—crowd density, fountain art presentation focus, and fountain interactive accessibility—as they change over time in different spatial locations. The preset time period can be a time range set for analyzing the spatiotemporal changes in the human-scenery collaborative state. It can be flexibly set according to the actual operational scenario of the public space and serves as the time benchmark for conducting temporal analysis. Spatial location can refer to various physical locations within the public space, centered on the fountain area. Expected changes in crowd density can be the estimated trend of crowd density at various spatial locations within the preset time period over time. Changes in the weight of artistic focus can be the trend of lighting focus weight in various artistic focus areas of the fountain changing with crowd status and performance progress within the preset time period. Changes in the probability of interactive nodes being triggered by the crowd can be the trend of the probability of interactive nodes being triggered by the crowd changing with crowd flow trajectory and performance progress within the preset time period. Spatiotemporal dynamic evolution information can be information integrating the synchronous changes in crowd density, artistic focus weight, and interactive node trigger probability at various spatial locations within the preset time period. Crowd gathering can be a state where crowd density significantly increases at a specific spatial location within the public space. Artistic focus shift can refer to the spatial change of the core artistic presentation area during a fountain performance, as the performance progresses. Interactive activities can occur when people enter the interactive area of ​​the fountain and trigger interactive facilities. Lighting intensity can refer to the brightness of the light emitted by public space lighting equipment. Lighting color can refer to the color and color ratio of the light emitted by public space lighting equipment. Lighting response speed can refer to how quickly public space lighting equipment adjusts lighting parameters according to changes in the human-scenery interaction. Location-varying time-dependent demand fluctuation information can be information on the fluctuations in lighting intensity, lighting color, and lighting response speed requirements at different spatial locations over time due to changes in crowds, fountain art, and human-scenery interaction.

[0049] Specifically, when analyzing the lighting needs of human-scenery coordination, if we do not explore the dynamic evolution of time and space and break down the dimensions of lighting needs, the lighting needs will be out of touch with the actual human-scenery coordination in dynamic scenarios such as the movement of people in public spaces and the shift of artistic focus. Subsequent wind disturbance analysis and lighting control will also lose their accurate basis, resulting in lighting effects that are seriously inconsistent with expectations.To address the aforementioned issues: Wide-angle cameras deployed on lampposts are used to extract real-time crowd density heatmaps and individual movement direction vector fields using a real-time video analysis model based on improved YOLOv5. The fountain control system provides the performance script in real-time via the OPC UA protocol, which includes the on / off status of each fountain pump, spray height, and preset artistic focal point coordinates (such as the center point of the main water curtain) at each moment. During the data processing stage, the DBSCAN spatial clustering algorithm is used to process the crowd heatmap to dynamically divide it into high, medium, and low density levels, constructing the "density-artistic focal point matching information." Simultaneously, in conjunction with the performance script, a visual saliency model based on attention mechanisms (such as Saliency Attentive) is employed. The model simulates the collective field of vision of crowds at different densities, calculates the dynamic lighting weight priority of each artistic focal point (such as multiple main water jets) in different density areas, and continuously tracks the movement trajectory of specific crowd clusters using an improved DeepSORT multi-target tracking algorithm. It also uses a geohash algorithm to perform real-time spatial overlay analysis of these trajectories with predefined interactive areas (such as ground LED interactive areas) in the fountain performance script. When the trajectory and interactive area overlap within a spatiotemporal grid (such as a 0.5-second time window and a 1-meter spatial grid), it is recorded as a valid "spatiotemporal synchronization trigger sequence." Subsequently, the above two types of information flows are input into a spatiotemporal sequence prediction model based on LSTM (Long Short-Term Memory). This model uses a fixed time window (such as the next 5 minutes) as the prediction length and inputs historical and current crowd density distribution and artistic focal point weights. The sequence and interactive trigger sequence output the evolution trend of these three types of indicators in each spatial grid in the future time period, namely "spatiotemporal dynamic evolution information". Based on this evolution information, a lightweight demand fluctuation analysis engine is further applied. This engine has a built-in lighting demand mapping rule library. For example, the rule "when the predicted density level in the grid increases and it is the main art focus area, the lighting intensity demand coefficient is increased by 0.3; if it is accompanied by an interactive trigger probability > threshold, the color switching response speed demand is increased to the 'fast' level"" generates quantitative "location time-varying demand fluctuation information". Finally, the graph database Neo4j is used to construct a "human-scene collaborative lighting demand dynamic map". The nodes of the map are spatial grids, the attributes contain their fluctuation demand data, and the edges represent the spatiotemporal correlation based on the predicted movement path of the crowd. Through real-time traversal and query of the map, a structured "human-scene collaborative lighting demand information set" can be efficiently organized for the downstream control strategy generation module to call.

[0050] The method provided in this embodiment accurately captures the spatiotemporal coordination relationship between people, scenery, and interaction, making lighting needs fit the actual dynamic scene, providing accurate and detailed data support for subsequent environmental interference analysis and lighting strategy formulation, and ensuring the scientific nature of lighting control.

[0051] In some embodiments, real-time environmental disturbance information includes wind speed and wind direction; based on wind speed, the degree of suppression of the preset spray height of the fountain water column and the degree of aggravation of the water mist diffusion range are analyzed to obtain the wind intensity disturbance factor; based on wind direction, combined with the dynamic map of human-scene collaborative lighting demand, the displacement of the water column landing point and the continuous spatial area covered by the main dispersion path of the water mist under the influence of wind direction are analyzed to obtain the wind direction influence area; based on the wind intensity disturbance factor and the wind direction influence area, combined with the art focus attention weight, the dynamic attenuation law of the integrity of the visual presentation of the art focus and color saturation in the spatial area caused by the change of water column shape, the displacement of the landing point and the accompanying water mist are analyzed to obtain the focus interference information; based on the focus interference information, combined with the location time-varying demand fluctuation information, the dynamic correction and compensation of lighting intensity, color ratio and lighting angle required in the wind direction influence area to offset the wind disturbance influence and maintain the expected human-scene collaborative lighting effect are analyzed to obtain the environmental disturbance information set.

[0052] Wind speed can be the real-time velocity of airflow in the plaza environment. Wind direction can be the real-time direction of airflow in the plaza environment. Wind intensity disturbance factor can be a quantitative indicator based on real-time wind speed, reflecting the degree to which wind inhibits the preset spray height of the fountain water column and aggravates the spread of water mist. The dynamic map of human-scene collaborative lighting demand can be a visualized and data-driven map constructed with time as the main axis and space as the layer, reflecting the fluctuation of lighting demand at different spatial locations in the time dimension due to crowd gathering, art focus shifts, and interactive activities. Water column landing point offset can be the spatial deviation between the actual landing position and the preset landing position of the fountain water column under the pull of the wind. The main dispersion path of water mist can be the main trajectory of water mist moving in the air with the airflow after the fountain water column is affected by the wind. The wind direction influence area can be the continuous spatial area of ​​the plaza jointly covered by the water column landing point offset range pulled by the wind direction and the main dispersion path of water mist. Art focus attention weight can be the specific weight value in the priority sequence of lighting attention weights assigned to each fountain art focus area to adapt to the viewing effect of different crowd density levels. Changes in water column shape can refer to the phenomenon where a fountain's water column transforms from a predetermined physical form to a non-preset form under wind disturbance. Color saturation refers to the vibrancy and purity of colors displayed when illumination shines on the focal point of a fountain's artwork. Dynamic attenuation patterns refer to the gradual decrease in the visual integrity and color saturation of the focal point's appearance over time and space under wind disturbance. Focus interference information is a collection of information reflecting the complete pattern of the dynamic attenuation of the visual quality of the focal point's appearance over time and space under wind disturbance. Location-dependent time-varying demand fluctuation information reflects the fluctuations in demand for lighting intensity, color, and response speed at different spatial locations over time due to crowd gatherings, shifts in the focal point, and interactive activities. Dynamic correction and compensation refers to the dynamic adjustments and compensation measures needed to adjust lighting intensity, color ratio, and lighting angle within the wind-affected area to offset the impact of wind disturbance and maintain the expected human-scenery coordinated lighting effect.

[0053] Specifically, in the process of presenting the fountain and lighting in a square, if the interference of wind disturbance on the lighting of the fountain water column is not analyzed, the lighting effect will be disconnected from the artistic presentation of the fountain, the visual integrity and color saturation of the artistic focus will be damaged, and the viewing needs of the audience will not be met. It will also result in the ineffective consumption of lighting resources.To address the aforementioned issues: The system begins with wind speed and direction sensors (e.g., Vaisala WMT52 models) deployed around the plaza. These sensors collect raw wind speed (e.g., 3.5 m / s) and wind direction (e.g., 30 degrees east of northeast) data in real time, several times per second. The data is then filtered and formatted via an edge gateway. After acquiring the data, a simplified empirical model based on computational fluid dynamics (CFD), such as the k-epsilon turbulence model, is used to quickly simulate the flow field near the nozzle. The wind speed value is input, and a wind intensity disturbance factor (a dimensionless coefficient, e.g., 0.7, characterizing the degree to which the water column shape deviates from the preset state) is calculated. The effective height of the water column is suppressed to 70% of the preset value. Simultaneously, wind direction data is combined with the precise geographic coordinates of the fountain nozzles. Using geometric projection and particle diffusion algorithms (such as simulating the movement of water mist particles in the wind field using a random walk model), the expected offset trajectory of the water column's visual landing point and the main dispersion path of the water mist concentration in three-dimensional space are deduced. This allows a dynamic wind-affected area (a time-varying triangular mesh) to be outlined in the digital twin model. Next, an established dynamic map of human-scene collaborative lighting needs is invoked to extract the attention weights of each artistic focal point within the affected area. For this area, an integrated optical distribution system is then activated. Analysis Module: First, based on the wind intensity disturbance factor and the water column physical model, a geometric description of the water column deformation sequence is generated. Then, this deformed geometry is compared with the original design model. Using the Structural Similarity Index (SSIM) algorithm in computer vision, the loss of visual integrity is calculated point by point along the offset path, forming an integrity decay gradient (e.g., the SSIM value decreases by 0.15 for every 1 meter). Simultaneously, using ray tracing simulation based on Mie scattering theory, the spectral changes of a specific spectrum (e.g., blue light with RGB values ​​[0, 100, 200]) after propagation in a simulated water mist field (whose concentration distribution is given by the aforementioned particle diffusion model) are calculated. The system fits a color saturation decay curve (e.g., in the core area, blue light saturation decays by 20% for every meter of water mist penetration). Finally, the gradient and curve are fused into structured focusing interference information and input into a real-time optimization controller (e.g., using a model predictive control (MPC) framework). This controller uses minimizing the decay described by the focusing interference information as the objective function, with the dimming range, color gamut, and mechanical rotation angle of the luminaire as constraints, and combines location-varying time-varying demand fluctuation information as a reference to solve for the percentage power adjustment, RGBW mixing ratio, and fine-tuning angle required for each affected smart luminaire (e.g., using Philips ColorKinetics iW Burst Powercore) within a future control cycle (e.g., the next 5 seconds). This series of instructions constitutes a dynamic correction and compensation scheme, which is encapsulated in the final environmental interference information set and sent to the edge lighting nodes for execution.

[0054] The method provided in this embodiment accurately quantifies the lighting interference caused by wind disturbance, provides a targeted basis for lighting compensation, ensures the stability of human-scenery collaborative lighting effects, and realizes refined allocation of lighting resources.

[0055] In some embodiments, based on the wind-affected area and combined with the dynamic map of human-scene collaborative lighting needs, the spatial deviation between the visual center and the original artistic focal area after the water column's landing point shifts is analyzed to obtain focal point shift deviation information; based on the focal point shift deviation information, the main dispersion path of the water mist generated by the water column shift is analyzed, as well as the diffusion profile of the water mist concentration over time along the main dispersion path, to obtain water mist diffusion dynamic information; based on the water mist diffusion dynamic information, the scattering and absorption effects of the continuous spatial area covered by the water mist on the lighting light, as well as the cumulative interference range on the satisfaction of regional lighting needs over time, are analyzed to obtain the wind-affected area.

[0056] Focus offset deviation information can be related to the degree of spatial deviation between the visual center and the original artistic focal area after the water column's landing point shifts. The main dispersion path of the water mist can be the main movement and extension trajectory of the water mist under wind action after the water column shifts. Water mist diffusion dynamics information can be a comprehensive set of information including the main dispersion path of the water mist, the concentration change over time, and the diffusion profile. Water mist concentration can be the amount of water mist per unit space. The diffusion profile can be the spatial coverage boundary and morphological characteristics formed by the water mist during diffusion. Scattering and absorption effects can be the collective term for the dispersion and attenuation of light intensity caused by the water mist on the illumination light. Lighting demand satisfaction can be the degree to which the actual lighting effect meets the preset requirements for human-scene coordinated lighting. The cumulative interference range can be the expansion range and final coverage area of ​​the spatial area where the water mist causes the lighting satisfaction to fall short over time.

[0057] Specifically, in the process of dynamically integrating the lighting load in public spaces, if the wind direction influence area is not precisely constructed, the dynamic interference of water mist diffusion will not be accurately captured, resulting in a mismatch between the lighting compensation area and intensity and the actual wind disturbance. This will fail to offset the lighting interference of water column deviation and water mist, ultimately destroying the lighting effect of human-landscape synergy. To address the aforementioned issues: Starting with an initial "wind direction influence area" based on a rough estimate of wind direction and speed, the process begins by employing spatial vector analysis, a geometric calculation method, to compare the predicted center coordinates of the water column impact point within this area with the pre-marked original artistic focal point coordinates in the dynamic map of human-scene collaborative lighting needs. This calculates the Euclidean distance and azimuth angle difference between the two in a two-dimensional plane. When this deviation (e.g., a distance greater than 3 meters or an angle deviation exceeding 15 degrees) is significant, quantified "focal point offset deviation information" is generated. Subsequently, using this information as the key input, a discrete phase model simulation based on computational fluid dynamics principles is initiated. This is a mature technology widely used in aerosol and particulate matter diffusion simulation. In this model, the fountain nozzle is set as the source releasing water mist particles with a specific particle size distribution (e.g., simulating main water droplet sizes in the 20-200 micrometer range). The offset direction obtained in the previous step is used as the initial momentum. The system simulates the movement, spatial distribution, and concentration changes of water mist particles over a future period (30 seconds of the next performance beat) by incorporating real-time wind speed and direction as a continuous phase wind field. This outputs "water mist diffusion dynamic information" containing a time series. Finally, based on this dynamic information, a physical model of light transmission in a medium is applied. Specifically, a simplified algorithm combining Snell's law and Mie scattering theory is used to calculate the attenuation coefficient of illumination light at various points in the water mist concentration field. A key optical attenuation threshold is set (for example, the light intensity at this point is reduced to less than 70% of its original value due to water mist scattering and absorption). Then, a spatial region clustering algorithm is used to aggregate all continuous three-dimensional voxels in the entire scene whose attenuation coefficient exceeds this threshold. This final spatial set, which represents the "actual range of significant optical interference," is the constructed "wind direction influence area" used to guide precise compensation.

[0058] The method provided in this embodiment accurately quantifies the focus offset deviation, captures the dynamic law of water mist diffusion, clarifies the scattering and absorption effect and cumulative interference of water mist on lighting, makes the determination of the wind direction influence area more realistic, and provides accurate spatial and interference data support for subsequent lighting compensation.

[0059] In some embodiments, based on the wind strength disturbance factor, the dynamic process of the fountain water column changing from a preset straight shape to a curved, dispersed, or fragmented state under the action of wind is analyzed to obtain a water column deformation sequence; based on the wind direction influence area, combined with the water column deformation sequence, the continuous offset trajectory of the visual landing point of the water column due to deformation relative to the original artistic focal area is analyzed to obtain a focal offset path; based on the focal offset path, combined with the artistic focal attention weight, the gradual decrease in the visual clarity of key artistic elements caused by water column shape distortion and landing point deviation along the offset path is analyzed to obtain an integrity attenuation gradient; based on the integrity attenuation gradient, the synchronous attenuation trend of color purity and brightness of the lighting light in the water mist medium caused by the enhanced diffusion of water mist and the scattering of water column shape is analyzed to obtain a color saturation attenuation curve; based on the integrity attenuation gradient, combined with the color saturation attenuation curve, a complete law describing the dynamic attenuation of the visual presentation quality of the artistic focal point under wind disturbance with time and space is formed to obtain focus interference information.

[0060] A water column deformation sequence can be a dynamic process sequence in which a fountain's water column transforms from a pre-set straight shape to a curved, dispersed, or fragmented state under the influence of wind. An integrity attenuation gradient can be the gradient change pattern of the visual clarity of key artistic elements gradually decreasing with space along the focal point offset path. A color saturation attenuation curve can be the trend curve of the synchronous attenuation of color purity and brightness of lighting light due to water mist diffusion and water column scattering. Focus interference information can be a set of information that integrates the integrity attenuation gradient and the color saturation attenuation curve, describing the spatiotemporal dynamic attenuation pattern of the visual presentation quality of the artistic focal point under wind disturbance.

[0061] Specifically, in the coordinated presentation of fountain lighting in outdoor public spaces, without constructing focusing interference information, it is impossible to quantify the dual interference of wind-induced water column deformation and water mist diffusion on the visual focus of the artwork. This leads to a lack of scientific basis for lighting compensation, resulting in blurred visual focus, color distortion, and a severe reduction in the interactive viewing experience. To address these issues, a series of feasible existing technologies are used to transform the physical process of wind disturbance into a calculable model of visual quality degradation. First, computational fluid dynamics (CFD) simulation software (such as ANSYS) is employed. Based on wind disturbance factors (wind speed and wind direction), the CFD simulation is used to transiently simulate the water column of a fountain. By setting a multiphase flow model (water-air), the simulation obtains a continuous deformation sequence of the water column under wind force, from a preset shape to bending and breaking. The simulation results of each frame include the three-dimensional geometric shape of the water column and the water droplet distribution field, which is the basis for analyzing physical deformation. Then, based on the three-dimensional surface data of the water column and the wind direction influence area output by CFD, feature point detection and tracking algorithms in computer vision (such as KLT optical flow method) are applied to track the two-dimensional projection coordinates of key feature points (such as the highest point and contour turning point) of the simulated water column surface in consecutive frames. The algorithm calculates the movement trajectory of the centroids of these feature points to generate a precise focus offset path. Then, along this path, combined with a pre-defined artistic focus weight map, an objective image sharpness evaluation algorithm (such as the Tenengrad gradient function) is used to process the simulated water column visual image sequence. This algorithm quantifies sharpness by calculating the sum of squared gray-level gradients in each local region of the image, thereby analyzing the gradient curve of sharpness decreasing along the offset path with distance and time, i.e., the integrity decay gradient. Simultaneously, based on the dynamic information of water mist diffusion (obtained from the water mist concentration field in CFD simulation), spectral analysis and color science models are used, for example, using CIE... The Lab color space was used, and a simplified model was constructed based on the Lambert-Beer law. The transmittance of different wavelengths of illumination light under a specific water mist concentration was analyzed, thereby calculating the curve of color saturation decay as water mist concentration and optical path increase. Finally, data fusion technology (such as Kalman filtering or multi-sensor information fusion framework) was used to fuse the temporally and spatially aligned integrity decay gradient (morphological information) and the color saturation decay curve (color information) to construct a unified multidimensional matrix containing spatiotemporal coordinates, decay intensity and spectral components, which is the final focusing interference information.

[0062] The method provided in this embodiment accurately quantifies the spatiotemporal attenuation law of the artistic focal point visual under wind disturbance, providing a scientific basis for dynamic correction and compensation of lighting, ensuring the integrity and color saturation of the fountain's artistic focal point visual presentation under wind disturbance environment, and improving the synergistic viewing effect of people and scenery.

[0063] In some embodiments, based on the focus offset path and combined with the integrity attenuation gradient, the continuous spatial range where artistic expression fails due to water column deformation along the offset path is analyzed to obtain the core lighting compensation area; based on the color saturation attenuation curve and combined with the water mist diffusion dynamic information, the different spectral intensities required to maintain color purity within the water mist coverage area are analyzed to obtain the color compensation spectral formula; based on the core lighting compensation area and the color compensation spectral formula, combined with the location time-varying demand fluctuation information, a collaborative control instruction set for synchronously adjusting the lighting point and color within the wind direction influence area is generated to complete the specific implementation of dynamic correction compensation.

[0064] The coordinated control command set can be a series of specific control commands formulated within the wind-affected area to synchronously adjust the lighting point and color, and to complete dynamic correction and compensation. The core lighting compensation area can be a continuous spatial range along the offset path where the fountain's artistic expression fails due to water column deformation.

[0065] Specifically, in the process of public space lighting where wind disturbance causes deformation of fountain water columns and diffusion of water mist, if dynamic correction and compensation are not implemented, the visual integrity and color saturation of the artistic focal point will continue to decline, the lighting needs will be seriously out of sync with the actual effect, the viewing experience of people and scenery will be greatly reduced, and the lighting resources will be wasted ineffectively.To address the aforementioned issues: First, relying on high-precision environmental perception and 3D simulation models, edge-deployed LiDAR and visual sensors are used to capture the morphological sequence of fountain water columns under wind conditions in real time. Through point cloud registration and skeleton extraction algorithms, a 3D curve describing the continuous movement trajectory of the water column's visual center is constructed (e.g., using the Iterative Closest Point (ICP) algorithm for sequence frame registration). Then, combining pre-defined artistic focus metadata calibrated by a computer vision model (such as a deep learning-based key feature point recognition model), spatial sampling is performed along this path. The deviation of the water column morphological features (such as outline sharpness and water splash dispersion) at each sampling point from the standard morphology is analyzed. A "visual integrity attenuation gradient" field representing the degree of visual integrity loss is generated through spatial interpolation (e.g., using Kriging interpolation to fit the deviation of discrete sampling points into a continuous spatial field). Based on this gradient field, "path buffer analysis" and "threshold propagation" techniques from the geographic information system are applied. With the offset path as the central axis, the buffer radius is dynamically determined according to the gradient value (e.g., for areas where the gradient value exceeds a preset threshold such as 0.7, the buffer radius is expanded to, for example, 2 meters). This delineates the polygonal boundary of the "core lighting compensation area" that needs priority lighting enhancement. At the same time, "water mist diffusion dynamic information" (such as water mist concentration distribution cloud map) is obtained through millimeter-wave radar and humidity sensor network. Spatial superposition analysis with the aforementioned gradient field identifies superposition regions with high gradients and high water mist concentrations. For these regions, a "radiative transfer model" is introduced for spectral analysis. This model simulates the scattering and absorption processes of different wavelengths of light in the water mist medium (e.g., using simplified localized calculations based on the principles of atmospheric radiative transfer codes such as MODTRAN). Combined with experimentally measured "color saturation decay curves" (characterizing the decay rate of brightness of each RGB channel at a specific water mist concentration), the "color compensation spectral formula" required to maintain the target color coordinates is calculated in reverse (e.g., calculating that at a certain water mist concentration, the red channel intensity needs to be increased to a baseline of, say, 130%, and the blue channel to a certain value). (e.g., 150%) Finally, the spatial coordinates of the "core lighting compensation area", the spectral parameters of the "color compensation spectral formula", and the "location time-varying demand fluctuation information" from the upper-level pedestrian flow analysis module (e.g., the current pedestrian flow in a certain area is dense, and the demand weight coefficient is 0.9) are input into an instruction synthesizer based on a "rule engine". This engine has built-in priority rules (e.g., artistic integrity compensation takes precedence over general ambient lighting) to perform multi-objective optimization and conflict resolution, and finally generate a batch of "collaborative control instruction sets" to be sent to each specific smart lamp. The instructions precisely include the lamp ID, adjustment timestamp, target spatial coordinates (for pan-tilt rotation), light intensity value and output ratio of each RGBW channel.

[0066] The method provided in this embodiment accurately counteracts the interference of wind on lighting, restores the visual presentation effect of artistic focal points, allows lighting adjustments to meet the actual needs of time and space, improves the utilization rate of lighting resources, and ensures a lighting experience that harmonizes people and scenery.

[0067] In some embodiments, based on the core lighting compensation area and combined with the color compensation spectral formula, the differentiated control schemes for lighting intensity enhancement and spectral correction required for the fountain art focal area and the continuous surrounding area affected by water mist diffusion are analyzed at the spatial scale to obtain a spatial scale compensation strategy. Based on the collaborative control instruction set and combined with spatiotemporal dynamic evolution information, the timing of the activation, duration of action, and fade-in / fade-out rhythm of each control instruction in the spatial scale compensation strategy are dynamically arranged according to the temporal characteristics of crowd flow, art focal point shift, and wind disturbance changes at the temporal scale to obtain a temporal scale arrangement strategy. Based on the spatial scale compensation strategy and combined with the temporal scale arrangement strategy, the differences in the capabilities of different edge lighting nodes in terms of spatial coverage and control response speed are analyzed. The composite control task is decomposed and allocated to the edge node cluster with corresponding scale response capabilities to generate a multi-scale collaborative control strategy and output a dynamic integration log of public space lighting load.

[0068] Spatial scale compensation strategies can be differentiated lighting control schemes, specifically targeting the focal area of ​​the fountain art and the surrounding areas affected by water mist diffusion, with separate plans for increasing lighting intensity and correcting the spectrum. Spatiotemporal dynamic evolution information can be the synchronous evolution trend information of expected changes in crowd density, changes in the focus weight of the art focal point, and changes in the trigger probability of interactive nodes at various locations in the public space within a preset time period. Temporal scale arrangement strategies can be time-dimensional lighting control schemes formed by dynamically arranging the activation timing, duration, and fade-in / fade-out rhythm of various control commands in the spatial scale compensation strategy based on the temporal characteristics of crowd flow, art focal point shifts, and wind disturbance changes. Edge lighting nodes can be lighting equipment nodes deployed at the edges of the public space, possessing independent lighting control and response capabilities. Spatial coverage can be the spatial area that a single edge lighting node can effectively illuminate. Control response speed can be the time it takes for an edge lighting node to adjust lighting parameters after receiving a lighting control command. Edge node clusters can be lighting control clusters composed of multiple edge lighting nodes with different scale response capabilities. Composite control tasks can be comprehensive lighting control tasks that require simultaneous consideration of spatial scale compensation and temporal scale arrangement during the dynamic integration of public space lighting loads. Scale response capability refers to the ability of edge lighting nodes or node clusters to adapt to lighting control needs at different scales in terms of spatial coverage and control response speed.

[0069] Specifically, in the implementation of dynamic integration of public space lighting load, if a multi-scale collaborative control strategy is not generated and logs are recorded, it will lead to a lack of unified lighting control scheme, chaotic task allocation at nodes, failure of wind disturbance compensation, inability to meet the needs of human-scenery collaborative lighting, and lack of data support for subsequent system optimization. To address the aforementioned issues, the approach begins with the spatial mapping and fusion of two fundamental compensation requirements: the core lighting compensation area and the color compensation spectral formula. A gridded spatial analysis technique based on Geographic Information System (GIS) is employed, dividing the plaza area into finely divided grid units (e.g., 0.5m x 0.5m). The compensation area and spectral formula data are then mapped onto these grids. A weighted superposition algorithm is used to generate the final lighting intensity correction value and RGB spectral adjustment coefficient required for each grid unit, thereby outputting differentiated spatial scale compensation strategies. Subsequently, for temporal orchestration, a hybrid model based on time series prediction and a rule engine is constructed. This model uses a collaborative control instruction set as its action library and spatiotemporal dynamic evolution information (pedestrian density time series, art focus switching schedule, wind speed and direction prediction sequence) as input. Algorithms such as dynamic window scheduling are applied to solve for the optimal control instruction sequence, determining the precise start time and duration of each spatial compensation action. The duration of the fade-in / fade-out brightness transition curves defined by Bézier curves are used to generate a time-scale orchestration strategy. Finally, to achieve task allocation, a capability profile including spatial coverage, dimming and color adjustment response speed, and control interface protocol is established for each edge node (such as a specific model of smart floodlight or wall washer controller). Using resource allocation models such as combined auction or improved Hungarian algorithm, the atomic control tasks decomposed from the aforementioned strategy (such as "at time T, increase the brightness of grid A to X and adjust the color temperature to Y") are treated as "goods", and the edge nodes are treated as "bidders" or "assigned objects" to achieve optimal matching between the capabilities of the tasks and the execution nodes. Finally, a multi-scale collaborative control strategy that can be distributed and executed is generated, and a JSON format log file recording all analysis parameters, decision logic, and allocation results is generated to complete the closed-loop generation of the entire control strategy.

[0070] The method provided in this embodiment enables precise formulation of spatial and temporal lighting control strategies, reasonable allocation of node tasks, effective mitigation of wind disturbance effects, and ensures the lighting effect of human-scenery collaboration. The log provides detailed evidence for system operation and maintenance optimization.

[0071] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A method for dynamic integration of public space lighting loads based on edge node multiscale clustering, characterized in that, include: Obtain pedestrian flow distribution information and fountain performance mode information set; based on the pedestrian flow distribution information and the fountain performance mode information set, analyze the coupling relationship between the dynamic flow direction of pedestrian flow and the focal area of ​​fountain art to obtain a set of information on the lighting requirements of human-scenery coordination. Real-time environmental disturbance information is obtained. Based on the real-time environmental disturbance information and combined with the human-scenery collaborative lighting demand information set, the interference caused by environmental wind disturbance of the square fountain water column, which affects the lighting effect, is analyzed to obtain the environmental disturbance information set. Based on the environmental interference information set, a multi-scale collaborative control strategy is generated, and a dynamic integrated log of public space lighting load is output.

2. The method according to claim 1, characterized in that, Based on the pedestrian distribution information and combined with the fountain performance mode information set, the coupling relationship between the dynamic flow of pedestrians and the focal area of ​​the fountain art is analyzed to obtain a set of information on pedestrian-scene coordinated lighting requirements, including: The pedestrian distribution information includes crowd density and crowd movement direction; The fountain performance mode information set includes fountain performance art information and fountain-human-scene interaction information; Based on the crowd density and the fountain performance art information, the attention priority of the core art presentation area of ​​the fountain corresponding to different crowd density intervals is analyzed to obtain density-art focus matching information; Based on the direction of crowd movement and combined with the fountain human-scene interaction information, the overlap range and movement sequence of the crowd flow trajectory and the interactive area of ​​the fountain are analyzed to obtain trajectory-interaction node adaptation information. Based on the density-art focus matching information and the trajectory-interaction node adaptation information, the collaborative adaptation relationship between the degree of crowd gathering, the focus of art presentation and the accessibility of interaction under different spatial locations over time is analyzed to obtain the human-scene collaborative lighting demand information set.

3. The method according to claim 2, characterized in that, The process of constructing the density-artistic focus matching information includes: Based on the population density, the population distribution area is divided into different density levels; Based on the aforementioned fountain performance art information, the presentation intensity and spatial position of multiple core artistic elements in the fountain performance are analyzed to obtain information on the artistic focal area. Based on the different density levels and combined with the information on the art focal area, the changes in the overall field of vision and attention concentration of the crowd are analyzed under each density level to obtain information on changes in crowd visuals. Based on the visual change information of the crowd and the information of the art focus area, the priority sequence of lighting attention weights that should be allocated to each art focus area to adapt to the viewing effect of crowds at different density levels is analyzed, and the density-art focus matching information is obtained.

4. The method according to claim 2, characterized in that, The process of constructing the trajectory-interaction node adaptation information includes: Based on the direction of crowd movement and combined with the fountain human-scenery interaction information, the spatial intersection and overlap between the crowd movement path and the interactive area of ​​the fountain are analyzed to obtain trajectory-area spatial coupling information. Based on the trajectory-regional spatial coupling information, the temporal correspondence between the movement speed and the triggering mechanism of interactive facilities when the crowd moves to different intersection areas is analyzed to obtain the spatiotemporal synchronization triggering sequence. Based on the spatiotemporal synchronization trigger sequence, the dynamic interactive network and experience continuity formed by different flow trajectories triggering different interactive nodes in time sequence are analyzed to obtain the trajectory-interaction node adaptation information.

5. The method according to claim 4, characterized in that, Based on the density-artistic focus matching information and the trajectory-interaction node adaptation information, the collaborative adaptation relationship between crowd gathering degree, artistic presentation focus, and interactive accessibility at different spatial locations over time is analyzed to obtain the human-scene collaborative lighting demand information set, including: Based on the density-art focus matching information and the trajectory-interaction node adaptation information, the synchronous evolution trend of expected changes in crowd density, changes in the weight of attention to art focus, and changes in the probability of triggering interaction nodes at each spatial location within a preset time period is analyzed to obtain spatiotemporal dynamic evolution information. Based on the aforementioned spatiotemporal dynamic evolution information, the demand fluctuations for lighting intensity, lighting color, and lighting response speed at different spatial locations over time are analyzed due to crowd gathering, shifts in artistic focus, and interactive activities, resulting in location-varying demand fluctuation information. Based on the location-varying demand fluctuation information, a dynamic map of human-scene collaborative lighting demand is constructed with time as the main axis and space as the layer, forming the human-scene collaborative lighting demand information set.

6. The method according to claim 5, characterized in that, Based on the real-time environmental disturbance information and combined with the human-scenery collaborative lighting demand information set, the interference caused by environmental wind disturbance affecting the square fountain water column, resulting in an environmental disturbance information set, is obtained, including: The real-time environmental disturbance information includes wind speed and wind direction; Based on the wind speed, the degree of suppression of the preset spray height of the fountain water column and the degree of aggravation of the water mist diffusion range are analyzed to obtain the wind intensity disturbance factor. Based on the wind direction and combined with the dynamic map of human-scenery collaborative lighting demand, the offset of the water column landing point and the continuous spatial area covered by the main dispersion path of water mist under the influence of wind direction are analyzed to obtain the wind direction-affected area. Based on the wind strength disturbance factor, the wind direction influence area, and the art focus attention weight, the dynamic attenuation law of the visual presentation integrity and color saturation of the art focus in the spatial area caused by changes in water column shape, drop point shift and accompanying water mist is analyzed to obtain focus interference information. Based on the focusing interference information and the location-varying demand fluctuation information, the analysis shows that in order to offset the wind disturbance and maintain the expected human-scenery coordinated lighting effect, dynamic correction and compensation of lighting intensity, color ratio and lighting angle are required in the wind direction-affected area, thus obtaining the environmental interference information set.

7. The method according to claim 6, characterized in that, The process of constructing the wind direction influence area includes: Based on the wind direction influence area and combined with the dynamic map of human-scene collaborative lighting demand, the spatial deviation between the visual center and the original artistic focal area after the water column landing point shifts is analyzed to obtain focal shift deviation information. Based on the focal offset deviation information, the main dispersion path of water mist generated by the water column offset is analyzed, as well as the diffusion profile of water mist concentration over time along the main dispersion path of water mist, to obtain dynamic information of water mist diffusion. Based on the dynamic information of water mist diffusion, the scattering and absorption effects of the continuous spatial area covered by water mist on the lighting light are analyzed, as well as the cumulative interference range on the satisfaction of regional lighting needs over time, to obtain the wind direction influence area.

8. The method according to claim 7, characterized in that, The process of constructing the focused interference information includes: Based on the wind strength disturbance factor, the dynamic process of the fountain water column changing from a preset straight shape to a curved, dispersed or fragmented state under the action of wind is analyzed to obtain the water column deformation sequence. Based on the wind direction influence area and the water column deformation sequence, the continuous offset trajectory of the visual landing point of the water column due to deformation relative to the original artistic focal area is analyzed to obtain the focal offset path. Based on the aforementioned focus offset path and combined with the artistic focus attention weight, the process of gradually decreasing visual clarity of key artistic elements caused by water column shape distortion and landing point deviation along the offset path is analyzed to obtain the integrity attenuation gradient. Based on the aforementioned integrity attenuation gradient, the synchronous attenuation trend of color purity and brightness of the illumination light in the water mist medium caused by enhanced water mist diffusion and water column morphology scattering is analyzed, and a color saturation attenuation curve is obtained. Based on the integrity attenuation gradient and the color saturation attenuation curve, a complete law describing the dynamic attenuation of the visual presentation quality of the artistic focus under wind disturbance with time and space is formed, thus obtaining the focus interference information.

9. The method according to claim 8, characterized in that, The specific implementation of the dynamic correction compensation includes: Based on the focal offset path and the integrity attenuation gradient, the continuous spatial range where artistic expression fails due to water column deformation on the offset path is analyzed to obtain the core lighting compensation area. Based on the color saturation decay curve and combined with the water mist diffusion dynamic information, the different spectral intensities required to maintain color purity within the water mist coverage area are analyzed to obtain a color compensation spectral formula. Based on the core lighting compensation area and the color compensation spectral formula, and combined with the location time-varying demand fluctuation information, a collaborative control instruction set is generated within the wind direction influence area to synchronously adjust the lighting landing point and light color, so as to complete the specific implementation of the dynamic correction compensation.

10. The method according to claim 9, characterized in that, The process of generating a multi-scale collaborative control strategy based on the environmental disturbance information set and outputting a dynamic integrated log of public space lighting load includes: Based on the core lighting compensation area and the color compensation spectral formula, we analyze the differentiated control schemes for lighting intensity enhancement and spectral correction that need to be implemented separately for the focal area of ​​the fountain art and the continuous surrounding area affected by water mist diffusion on a spatial scale, and obtain a spatial scale compensation strategy. Based on the aforementioned collaborative control instruction set and combined with the aforementioned spatiotemporal dynamic evolution information, the timing of activation, duration of action, and fade-in / fade-out rhythm of various control instructions in the spatial scale compensation strategy are dynamically arranged according to the temporal characteristics of crowd flow, shift of artistic focus, and wind disturbance changes on the time scale, in order to obtain a temporal scale arrangement strategy. Based on the spatial scale compensation strategy and the temporal scale orchestration strategy, the differences in the spatial coverage and control response speed of different edge lighting nodes are analyzed. The composite control task is decomposed and allocated to the edge node cluster with corresponding scale response capabilities, generating the multi-scale collaborative control strategy and outputting the dynamic integration log of the public space lighting load.

Images