Self-adaptive illumination dynamic regulation and control method and system for travel scene based on body perception

Abstract

The invention provides a text travel scene adaptive illumination dynamic regulation and control method and system based on personal perception, and relates to the technical field of scene perception, and the method comprises the steps: capturing visitor multi-dimensional behavior data and scene environment state data in real time through a distributed sensor network; constructing a body perception model fusing the visitor spatial position, the moving speed, the staying duration and the physiological wake-up degree, performing space-time correlation coding on the multi-dimensional behavior data, and generating a body perception feature vector; establishing a dynamic affinity matrix based on the feature vector and scene semantic information, and mapping the dynamic affinity matrix to a spatial topological structure of lighting equipment to form a regional lighting coupling map; according to the map node weight distribution and visitor perception state clustering result, generating a partition cooperative illumination regulation and control instruction through a multi-objective optimization algorithm; and carrying out online evolution on the model by utilizing the regulated and controlled visitor behavior track and physiological index change. According to the invention, accurate matching of the illumination environment and the visitor immersion state can be realized, and the experience quality of a text travel scene is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to scene perception technology, and in particular to a method and system for dynamically adjusting and controlling cultural and travel scene adaptive lighting based on embodied perception. BACKGROUND

[0002] With the rapid development of cultural tourism industry and the rise of experience economy, the environment creation of cultural and travel scenes has become a core element to improve the quality of visitor experience. Lighting, as a key means of scene atmosphere creation, its reasonable control directly affects the emotional perception, behavior pattern and overall satisfaction of visitors. The traditional lighting system of cultural and travel scenes mostly uses preset program control or simple time period switching mode, lacking dynamic response ability to the actual needs of visitors. In recent years, with the development of Internet of Things technology, sensor network and artificial intelligence algorithm, intelligent lighting systems have been gradually applied to museums, theme parks, historical blocks and other cultural and travel places, realizing automatic adjustment of lighting through environmental perception and data analysis.

[0003] Current cultural and travel scene lighting control technology mainly relies on environmental light sensors, human infrared detectors and other basic equipment to obtain scene state information, and combines pre-defined lighting rules for control. Some advanced systems introduce visual recognition technology to count pedestrian density, or collect visitors' subjective evaluation feedback through mobile applications. These methods have improved the intelligent level of the lighting system to some extent, but still face many challenges in practical application. Embodied cognition theory believes that human perception and cognition process is deeply dependent on the interaction between body and environment, and the experience of visitors in cultural and travel scenes is the result of the joint action of body movement, spatial position, sensory stimulation and psychological state. The complex correlation of multiple dimensions has not been fully reflected in the existing lighting control technology.

[0004] The existing lighting control system mainly relies on single-dimensional environmental parameters or pedestrian statistical data, and fails to establish a correlation model between the body state of visitors and psychological perception. Traditional sensors can only capture the position or movement facts of visitors, and cannot identify the visitors' intention to stay, attention focus and emotional arousal level, resulting in a significant disconnection between lighting control and visitors' real perception needs. This perception blind spot makes it difficult for the lighting system to accurately judge the immersion state of visitors, and it is unable to provide adaptive light environment support for different experience stages.

[0005] The existing method lacks a dynamic mapping mechanism between the perception state of visitors and spatial lighting resources. The control strategy of lighting devices is usually based on fixed spatial zoning or time rules, without considering the real-time distribution characteristics and perception differences of visitor groups in different areas. When multiple visitor groups are in different experience states at the same time, a unified lighting scheme cannot meet the differentiated needs, and simple independent control of different zones will cause harsh light environment breaks between adjacent areas, damaging the overall atmosphere coherence of the scene.

[0006] The existing lighting control system mostly adopts an open-loop control mode or a simple feedback mechanism based on a preset threshold, and lacks the ability of continuous evaluation of the regulation effect and model adaptive optimization. After adjusting the lighting parameters, the system cannot effectively track the response changes of the visitor behavior and physiological state, and it is difficult to determine whether the regulation strategy has truly improved the visitor experience. Such a static control mode lacks adaptability when facing different visitor group characteristics, seasonal changes and scene function conversion, and cannot realize the continuous evolution and optimization of the lighting strategy. SUMMARY

[0007] The embodiment of the application provides a self-adaptive lighting dynamic regulation method and system for a cultural and tourism scene based on embodied perception, which can solve the problems in the prior art.

[0008] In a first aspect, the embodiment of the application provides a self-adaptive lighting dynamic regulation method for a cultural and tourism scene based on embodied perception, comprising: A distributed sensor network deployed in a target cultural and tourism scene is used to capture multi-dimensional behavior data of visitors and scene environment state data in real time; an embodied perception model integrating visitor spatial position, moving speed, stay duration and physiological arousal is constructed, the multi-dimensional behavior data is spatio-temporally correlated and coded, and an embodied perception feature vector representing the immersion state of the visitors is generated; Based on the embodied perception feature vector and scene semantic information, a dynamic affinity matrix of the visitor perception state and the spatial region is established, the dynamic affinity matrix is mapped to the spatial topology structure of the lighting device, and a perception-driven regional lighting coupling graph is formed; According to the weight distribution of each node in the regional lighting coupling graph and the perception state clustering result of the visitor group, a multi-objective optimization algorithm is used to solve the spectral energy distribution scheme and the color temperature transition sequence of each lighting region, and a partitioned collaborative lighting regulation instruction is generated; The lighting device is driven to execute the lighting regulation instruction, and the changes in the visitor behavior trajectory and the fluctuations in the physiological indicators after regulation are monitored, the trajectory changes and the physiological indicator fluctuations are used as correction signals, and the spatio-temporal correlation coding weight in the embodied perception model and the calculation strategy of the dynamic affinity matrix are evolved online.

[0009] The embodied perception model integrating visitor spatial position, moving speed, stay duration and physiological arousal is constructed, the multi-dimensional behavior data is spatio-temporally correlated and coded, and an embodied perception feature vector representing the immersion state of the visitors is generated, comprising: A spatio-temporal trajectory representation of the visitor spatial position and moving speed is established, a motion state feature reflecting the spatial behavior continuity of the visitors is generated by establishing a corresponding relationship between the spatio-temporal trajectory representation and the stay duration; Based on the spatiotemporal distribution characteristics of the motion state features, system modeling is performed in combination with physiological arousal, dynamic variation rules of the physiological arousal under different motion states are analyzed, and a somatic perception model representing the collaborative relationship between the physical and mental states of the visitor is constructed. According to the mapping relationship established based on the somatic perception model, the spatial position of the visitor, the moving speed, the staying time, and the physiological arousal are organized in the spatiotemporal dimension, the state evolution characteristics of the visitor in the time dimension are obtained, and the regional response characteristics in the spatial dimension are determined. According to the internal relationship between the state evolution characteristics and the regional response characteristics, a unified mapping space is constructed, and a somatic perception feature vector representing the immersion state of the visitor is established in the unified mapping space.

[0010] Based on the spatiotemporal distribution characteristics of the motion state features, system modeling is performed in combination with physiological arousal, dynamic variation rules of the physiological arousal under different motion states are analyzed, and a somatic perception model representing the collaborative relationship between the physical and mental states of the visitor is constructed. The spatiotemporal distribution pattern of the motion state features is clustered, the motion state pattern of the visitor in different tourism and travel scene areas is identified, and an associated mapping relationship between the motion state pattern and the spatial semantic attribute is established. Based on the division of the motion state pattern, the fluctuation characteristics and variation trend of the physiological arousal in the corresponding period are analyzed, the physiological response mechanism of the visitor under different motion state patterns is described by constructing a physiological arousal response function, and the physiological response mechanism is annotated with scene semantics based on the associated mapping relationship. The physiological response mechanism and the physiological arousal response function are combined and operated to form a feature expression reflecting the spatiotemporal collaborative changes of the motion state and the physiological arousal, and the collaborative strength of the visitor's physical behavior and psychological state is quantified according to the associated mapping relationship. The collaborative strength is introduced into the reconstruction process of the feature expression, a new spatiotemporal association pattern is generated through spatial mapping transformation, and a somatic perception model representing the collaborative relationship between the physical and mental states of the visitor is constructed.

[0011] Based on the somatic perception feature vector and the scene semantic information, a dynamic affinity matrix of the visitor's perception state and the spatial area is established, the dynamic affinity matrix is mapped to the spatial topology structure of the lighting device, and a perception-driven regional lighting coupling graph is formed. Based on the somatic perception feature vector, semantic hierarchical decomposition is performed to generate emotion tendency features and behavior preference features reflecting the visitor's perception state, and in combination with the spatial function attribute and the environmental atmosphere attribute in the scene semantic information, the matching degree index of the emotion tendency features and the behavior preference features in each spatial area is calculated. constructing a dynamic affinity matrix of the visitor perception state and the space region according to the matching degree index, wherein a matrix element reflects a response intensity of the visitor perception state to a specific space region, and the dynamic affinity matrix is dynamically reconstructed through matrix iteration update to embody evolution trends of the sentiment tendency feature and the behavior preference feature; establishing a space structure map of the lighting devices, analyzing a space region coverage range of each lighting device and a topological connection rule between adjacent devices; constructing a state distribution of lighting device nodes using the response intensity in the dynamic affinity matrix, and generating a perception-driven regional lighting coupling map according to the topological connection rule.

[0012] solving a spectral energy distribution scheme and a color temperature transition sequence of each lighting region through a multi-objective optimization algorithm according to a weight distribution of each node in the regional lighting coupling map and a clustering result of the perception state of the visitor group, to generate a partitioned cooperative lighting regulation instruction including: performing clustering analysis on the perception state of the visitor group, identifying a visitor subgroup with similar perception state features, and statistically analyzing a distribution density of each visitor subgroup in different space regions to establish a distribution association relationship between the visitor subgroup and the space region; combining the weight distribution of each node in the regional lighting coupling map and the distribution association relationship, calculating a comprehensive service weight of each lighting region for different visitor subgroups, and constructing a multi-objective optimization function based on the comprehensive service weight; solving the multi-objective optimization function through a multi-objective optimization algorithm to determine a spectral energy distribution scheme of each lighting region at different time periods, calculating a color temperature gradient between adjacent lighting regions based on the spectral energy distribution scheme, and generating a color temperature transition sequence ensuring visual comfort according to the color temperature gradient and a moving track of visitors between adjacent lighting regions; converting the spectral energy distribution scheme and the color temperature transition sequence into control parameters executable by the lighting devices to generate a partitioned cooperative lighting regulation instruction.

[0013] calculating a color temperature gradient between adjacent lighting regions based on the spectral energy distribution scheme, and generating a color temperature transition sequence ensuring visual comfort according to the color temperature gradient and a moving track of visitors between adjacent lighting regions includes: calculating an equivalent color temperature value of each lighting region according to a spectral intensity distribution of each lighting region in the spectral energy distribution scheme, and identifying a pair of adjacent lighting regions, calculating a color temperature difference value between the pair of adjacent lighting regions, and forming a color temperature gradient representing color temperature spatial distribution non-uniformity; The color temperature gradient analysis visitor from the starting lighting area to the target lighting area moving trajectory and space-time path characteristics, calculate the color temperature change rate perceived by the visitor in the moving process, establish the mapping relationship between the color temperature change rate and the visual comfort, determine the color temperature change rate threshold to ensure the visual comfort; According to the color temperature change rate threshold, when the color temperature change rate exceeds the threshold, an intermediate color temperature regulation node is inserted between the starting lighting area and the target lighting area, and the color temperature setting value and the duration of each intermediate color temperature regulation node are calculated according to the visitor moving speed and the path length; Based on the equivalent color temperature value and the color temperature setting value, the color temperature values of the starting lighting area, the intermediate color temperature regulation nodes and the target lighting area are sorted and combined according to the visitor moving time sequence, and a color temperature transition sequence is generated to ensure the visual comfort.

[0014] Monitor the changes of the behavior trajectory and the physiological index fluctuation of the regulated visitor, and use the trajectory changes and the physiological index fluctuation as correction signals to online evolve the calculation strategy of the space-time correlation coding weight in the embodied perception model and the dynamic affinity matrix, including: Monitoring the behavior trajectory change rule of the regulated visitor, analyzing the real-time fluctuation characteristics of the physiological index in the behavior trajectory change process, generating dynamic feedback data reflecting the response effect of the visitor; using the fluctuation characteristics in the dynamic feedback data to adaptively adjust the space-time correlation coding weight of the embodied perception model, constructing an iterative optimization mechanism of model parameters; according to the adjustment result of the iterative optimization mechanism, reconstructing the calculation strategy of the dynamic affinity matrix, realizing the cooperative evolution of the perception model and the affinity matrix.

[0015] The second aspect of the embodiment of the application provides a self-adaptive lighting dynamic regulation system based on embodied perception in a travel and tourism scene, including: The first unit is used for capturing multi-dimensional behavior data and scene environment state data of visitors in real time through a distributed sensor network deployed in a target travel and tourism scene; an embodied perception model is constructed by fusing the spatial position, moving speed, stay duration and physiological arousal degree of the visitors, and the multi-dimensional behavior data is space-time correlation coded to generate an embodied perception feature vector representing the immersion state of the visitors; The second unit is used for establishing a dynamic affinity matrix of the visitor perception state and the space region based on the embodied perception feature vector and scene semantic information, mapping the dynamic affinity matrix to the space topology structure of the lighting device, and forming a perception-driven regional lighting coupling graph; a third unit configured to solve a spectral energy distribution scheme and a color temperature transition sequence of each lighting area by a multi-objective optimization algorithm according to a weight distribution of each node in the regional lighting coupling graph and a clustering result of the perception state of the visitor group, and generate a lighting regulation instruction with partition coordination; a fourth unit configured to drive the lighting device to execute the lighting regulation instruction, and monitor a change in a behavior trajectory and a fluctuation in a physiological index of the visitor after regulation, and take the change in the behavior trajectory and the fluctuation in the physiological index as a correction signal to perform online evolution on a spatiotemporal correlation coding weight in the embodied perception model and a calculation strategy of the dynamic affinity matrix.

[0016] In a third aspect, the embodiment of the present application provides an electronic device, comprising: a processor; a memory configured to store processor-executable instructions; wherein the processor is configured to invoke the instructions stored in the memory to execute the method described above.

[0017] In a fourth aspect, the embodiment of the present application provides a computer readable storage medium having stored thereon computer program instructions, which, when executed by a processor, implement the method described above.

[0018] The present application has the following beneficial effects: The embodiment of the present application can accurately capture the real immersion state of the visitor in the cultural tourism scene by constructing an embodied perception model fusing the spatial position, moving speed, stay duration and physiological arousal degree of the visitor, breaks through the limitation of the traditional lighting system which only depends on the environmental illuminance or time preset, realizes the paradigm shift of human-centered lighting perception, and significantly improves the response accuracy of the lighting system to the individual differences and dynamic needs of the visitor group.

[0019] The dynamic affinity matrix established by the embodiment of the present application realizes real-time correlation mapping of the visitor perception state and the spatial area, and realizes the spatial topology coordination of the lighting device through the regional lighting coupling graph, so that the lighting regulation is changed from isolated single-point control to networked partition coordination response, effectively avoids the visual discomfort caused by the sudden change of the lighting parameters of adjacent areas, and balances the rationality of spectral energy distribution and color temperature transition through a multi-objective optimization algorithm, which not only guarantees the visual comfort of the visitor but also meets the scene atmosphere creation demand.

[0020] The embodiment of the present application continuously monitors the change in the behavior trajectory and the fluctuation in the physiological index of the visitor after regulation, takes the actual feedback as a correction signal to perform online evolution on the embodied perception model and the dynamic affinity matrix, and constructs a closed-loop adaptive optimization mechanism, so that the lighting system has the ability of self-learning and continuous evolution, can adapt to the particularity of different cultural tourism scenes and the diversity of the visitor group, effectively improves the long-term operation stability and scene generalization ability of the system, and reduces the artificial maintenance cost. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the adaptive lighting dynamic control method for cultural and tourism scenarios based on embodied perception, as described in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the mind-body coordinated perception modeling of visitor movement state and physiological response in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0024] Figure 1 This is a flowchart illustrating the adaptive lighting dynamic control method for cultural and tourism scenes based on embodied perception, as described in an embodiment of the present invention. Figure 1 As shown, the method includes: Multidimensional behavioral data of visitors and scene environment status data are captured in real time by a distributed sensor network deployed in the target cultural and tourism scene; an embodied perception model that integrates visitor spatial location, movement speed, stay duration and physiological arousal is constructed, and the multidimensional behavioral data is spatiotemporally correlated and encoded to generate an embodied perception feature vector that represents the visitor's immersion state. Based on the embodied perception feature vector and scene semantic information, a dynamic affinity matrix between visitor perception state and spatial area is established. The dynamic affinity matrix is ​​then mapped to the spatial topology of lighting equipment to form a perception-driven area lighting coupling map. Based on the weight distribution of each node in the regional lighting coupling map and the clustering results of the visitor group's perception state, a multi-objective optimization algorithm is used to solve the spectral energy distribution scheme and color temperature transition sequence of each lighting area, and generate a zone-coordinated lighting control command. The system drives the lighting equipment to execute the lighting control command and monitors changes in visitor behavior trajectories and fluctuations in physiological indicators after control. The changes in trajectories and fluctuations in physiological indicators are used as correction signals to evolve online the spatiotemporal correlation encoding weights in the embodied perception model and the calculation strategy of the dynamic affinity matrix.

[0025] In an alternative embodiment, a somatic perception model is constructed by fusing the visitor spatial position, moving speed, staying duration and physiological arousal degree, the multi-dimensional behavior data is spatio-temporally correlated and coded to generate a somatic perception feature vector representing the visitor immersion state, which includes: A spatio-temporal trajectory representation of the visitor spatial position and moving speed is established, a motion state feature reflecting the continuity of the visitor spatial behavior is generated by establishing a corresponding relationship between the spatio-temporal trajectory representation and the staying duration; Based on the spatio-temporal distribution characteristics of the motion state feature, a system model is constructed in combination with the physiological arousal degree, the dynamic variation law of the physiological arousal degree under different motion states is analyzed, and a somatic perception model representing the coordination relationship between the visitor's physical and mental states is constructed; According to the mapping relationship established by the somatic perception model, the visitor spatial position, the moving speed, the staying duration and the physiological arousal degree are organized in spatio-temporal dimensions to obtain the state evolution feature of the visitor in the time dimension and determine the regional response feature in the space dimension; According to the internal relationship between the state evolution feature and the regional response feature, a unified mapping space is constructed, and a somatic perception feature vector representing the visitor immersion state is established in the unified mapping space.

[0026] The positioning device carried by the visitor collects spatial coordinate data at fixed time intervals. The positioning device interacts with the base stations distributed in the exhibition space through ultra-wideband wireless communication technology. The base stations calculate the three-dimensional spatial position of the visitor after receiving the signals. The positioning device records the spatial coordinates of the visitor every 0.5 seconds, including the horizontal coordinate, the vertical coordinate and the height coordinate, forming a continuous position sampling sequence. The spatial position change between the adjacent two samplings divided by the time interval is the instantaneous moving speed of the visitor, which contains two components of speed size and moving direction. When the position change of the five consecutive sampling points is less than 0.3 meters, the system determines that the visitor is in a staying state and starts to accumulate the staying duration. The coordinates of a visitor in front of exhibit A from time 10 seconds to time 45 seconds remain in the range of 3.2 meters horizontally and 5.8 meters vertically, and the system records the staying duration of the visitor in front of exhibit A as 35 seconds.

[0027] The spatial position data and the moving speed data are arranged in time sequence to form a space-time trajectory representation sequence, each data unit in the sequence containing three elements of time stamp, spatial coordinates and speed vector. After the visitor enters the exhibition hall from the entrance, at time 0 seconds, the visitor is located at the coordinate origin, moves at a speed of 1.2 meters per second towards the exhibit B, at time 8 seconds, reaches the coordinates near the exhibit B, which are 9.6 meters horizontally and 2.1 meters vertically, and the speed is reduced to 0.2 meters per second and starts to stay. The stay duration data is associated with the space-time trajectory representation, and the starting time, ending time, stay position and duration of each stay event are marked in the trajectory sequence. The visitor stays in front of the exhibit B from time 8 seconds to time 30 seconds, and after 22 seconds, the visitor leaves at a speed of 0.8 meters per second, which forms a complete motion state feature sequence containing moving-staying-moving.

[0028] The motion state features are divided into four types of fast moving state, slow moving state, short stay state and long stay state. The fast moving state corresponds to the trajectory segment with a speed greater than 1.0 meters per second and a duration of more than 3 seconds, the slow moving state corresponds to the trajectory segment with a speed between 0.3 meters per second and 1.0 meters per second, the short stay state corresponds to the position point with a stay duration between 5 seconds and 20 seconds, and the long stay state corresponds to the position point with a stay duration of more than 20 seconds. The trajectory data of a visitor shows that the visitor is in a fast moving state at time 0 seconds to time 6 seconds, in a slow moving state at time 6 seconds to time 12 seconds, and in a long stay state at time 12 seconds to time 50 seconds.

[0029] The physiological arousal degree is obtained by real-time collection of heart rate data and skin conductance data of the visitor through a wearable sensor. The sensor collects heart rate signals at a frequency of 10 times per second and collects skin conductance signals at a frequency of 5 times per second. The baseline heart rate of the visitor is 72 beats per minute in a resting state, and when the heart rate rises above 85 beats per minute, it is determined to be in a high arousal state, and when the heart rate remains between 60 beats per minute and 80 beats per minute, it is determined to be in a moderate arousal state. The skin conductance value rising from the baseline of 2.3 microsiemens to 4.1 microsiemens also indicates an increase in arousal. The physiological arousal value is calculated by weighting and combining the heart rate variation amplitude and the skin conductance variation amplitude, the weight coefficient of the heart rate variation is set to 0.6, and the weight coefficient of the skin conductance variation is set to 0.4.

[0030] Visitors exhibited significant differences in physiological arousal levels across different movement states. During rapid movement, the average heart rate was 88 beats per minute, skin conductance was 3.8 microsiemens, and the overall arousal score was 6.7. During prolonged stays, when the location corresponded to a highly attractive exhibit, the visitor's heart rate remained at 82 beats per minute, skin conductance reached 4.5 microsiemens, and the overall arousal score was 7.2. However, during prolonged stays facing less attractive exhibits, the heart rate dropped to 68 beats per minute, skin conductance was 2.6 microsiemens, and the overall arousal score was 3.9. The embodied perception model characterizes the synergistic relationship between mind-body states by constructing a mapping table between movement state feature types and physiological arousal ranges. This mapping table records the frequency distribution and typical numerical ranges of each movement state type across different arousal ranges.

[0031] When organizing visitor data by time, continuous sampling data is segmented into fixed time windows, each 10 seconds long. Within each time window, the average movement speed, range of position changes, number of dwell events, and average arousal level of visitors are statistically analyzed. For example, a visitor's average speed within a time window of 20 to 30 seconds is 0.4 meters per second, their position changes horizontally between 5.2 and 6.8 meters, they experience one dwell event, and their average arousal level is 5.3. The statistical data from continuous time windows are arranged chronologically to form a state evolution feature sequence, reflecting the complete state change process of a visitor from entering the exhibition hall to leaving.

[0032] Spatially, the exhibition space was divided into multiple grid areas, each with a side length of 2 meters by 2 meters. The total dwell time, average dwell time, number of visits, and average arousal level of all visitors within each grid area were statistically analyzed to form a description of the area's response characteristics. The grid area containing exhibit C had a cumulative dwell time of 450 seconds during the observation period, receiving 15 visitors with an average dwell time of 30 seconds per visit. The average arousal level of visitors in this area was 6.8, indicating that this area has high attractiveness and high arousal response characteristics. The grid area containing exhibit D had a cumulative dwell time of only 120 seconds, receiving 8 visitors with an average dwell time of 15 seconds and an average arousal level of 4.2, classifying it as a low-response area.

[0033] The state evolution feature and the regional response feature are determined to have an internal relationship through correlation analysis. The state evolution feature of the visitor when staying in a high response area shows that the arousal degree is continuously maintained at a high level, while the arousal degree shows a downward trend when staying in a low response area. The state evolution sequence of a visitor shows that the arousal degree rises from 5.1 to 7.4 in the time interval of 30 seconds to 60 seconds, and the corresponding spatial position data shows that the visitor is located in a high response grid area. The unified mapping space splices and combines the evolution feature vector in the time dimension and the response feature vector in the space dimension to form a comprehensive feature expression containing time evolution patterns and space distribution patterns.

[0034] The embodied perception feature vector includes six dimensional feature components, namely, an average moving speed component, a stay frequency component, a stay duration component, an arousal degree mean component, an arousal degree fluctuation amplitude component, and a high response area proportion component. The numerical values of the feature vector components of a visitor are: an average moving speed of 0.6 meters per second, a stay frequency of 1.8 times per minute, an average stay duration of 28 seconds, an arousal degree mean of 6.5, an arousal degree fluctuation amplitude of 1.9, and a high response area stay time proportion of 0.72. The feature vector comprehensively represents the immersion state intensity of the visitor in the exhibition space, and a higher value indicates a deeper immersion degree.

[0035] In an optional implementation, based on the spatiotemporal distribution characteristics of the motion state feature, a system model is established in combination with the physiological arousal degree, the dynamic change law of the physiological arousal degree under different motion states is analyzed, and an embodied perception model representing the coordination relationship between the visitor's physical and mental states is constructed, including: The motion state feature is clustered in a spatiotemporal distribution pattern, the motion state pattern of the visitor in different tourism and travel scene areas is identified, and an associated mapping relationship between the motion state pattern and the spatial semantic attribute is established; Based on the division of the motion state pattern, the fluctuation characteristics and change trend of the physiological arousal degree in the corresponding period are analyzed, the physiological reaction mechanism of the visitor under different motion state patterns is described by constructing a physiological arousal degree response function, and the physiological reaction mechanism is semantically annotated based on the associated mapping relationship; The physiological reaction mechanism and the physiological arousal degree response function are combined and operated to form a feature expression reflecting the spatiotemporal coordinated change of the motion state and the physiological arousal degree, and the coordination strength of the visitor's physical behavior and psychological state is quantified according to the associated mapping relationship; The coordination strength is introduced into the reconstruction process of the feature expression, a new spatiotemporal correlation pattern is generated through spatial mapping transformation, and an embodied perception model representing the coordination relationship between the visitor's physical and mental states is constructed.

[0036] As shown in Figure 2 The method includes: When clustering the spatiotemporal distribution patterns of motion state features, the motion data of visitors in the cultural tourism scene for 120 consecutive minutes is collected. The time axis is divided into 24 time segments with 5-minute units, and the spatial dimension is divided into different types of areas such as exhibition area, rest area, interactive experience area, and transition channel area according to the scene function attributes. For the motion state features in each time segment, basic parameters such as step frequency, moving speed, stay duration, and turning frequency are extracted, where the step frequency range is set to 80-140 steps per minute, the moving speed range is 0.5-2 meters per second, and the stay duration threshold is set to less than 0.3 meters of displacement change within 3 consecutive seconds. The system performs density peak clustering on these feature vectors, calculates the Euclidean distance between feature points, and determines adjacent points when the distance between two feature points is less than the set threshold distance of 1.2. Five typical motion state patterns are determined through iterative calculation, corresponding to fast walking mode, slow browsing mode, standing observation mode, frequent interaction mode, and random walking mode, respectively. The step frequency feature value of fast walking mode is more than 120 steps per minute, and the stay duration is less than 5 seconds. The moving speed of slow browsing mode is between 0.8-1.2 meters per second, and the turning frequency is 2-4 times per minute.

[0037] In the process of establishing the association mapping relationship between motion state patterns and spatial semantic attributes, the occurrence frequency and duration of each motion state pattern in different scene areas are recorded. Statistical data shows that the occurrence probability of fast walking mode in the transition channel area is 68%, with an average duration of 45 seconds; the occurrence probability of slow browsing mode in the exhibition area is 72%, with an average duration of 8 minutes; and the occurrence probability of standing observation mode in the interactive experience area is 81%, with an average duration of 3.5 minutes. The system assigns semantic label vectors to each scene area, with the exhibition area labeled as high information density, low interaction demand, and medium stay expectation, and the interactive experience area labeled as medium information density, high interaction demand, and long stay expectation. By calculating the similarity between motion state pattern features and spatial semantic labels, a mapping weight matrix is established. When standing observation mode appears in an area labeled as high interaction demand, the mapping weight value is set to 0.85, indicating a strong association relationship.

[0038] When analyzing the fluctuation characteristics of physiological arousal degree in the corresponding period, the heart rate variability index and skin conductance level data of the visitors are synchronously collected. In the fast walking mode, the standard deviation value of heart rate variability decreases from the baseline value of 50 milliseconds to 35 milliseconds, and the skin conductance level increases from the baseline value of 8 microsiemens to 15 microsiemens, and after about 40 seconds, it begins to fall. In the standing and observing mode, the standard deviation of heart rate variability shows a trend of first increasing and then stabilizing, from the baseline value of 50 milliseconds to 65 milliseconds and maintaining for about 2 minutes, and the skin conductance level appears three wave peaks, with peak values of 12 microsiemens, 14 microsiemens and 11 microsiemens, and the wave peak interval is about 50 seconds. The system calculates the change rate of physiological arousal degree, defined as the difference between the current time physiological index value and the previous time value divided by the time interval, and when the absolute value of the change rate is greater than the threshold value of 0.15, it is marked as a significant fluctuation event.

[0039] In the process of constructing the physiological arousal response function, the characteristic parameters of the movement state mode are taken as the input variables, and the change amplitude and change rate of the physiological arousal degree are taken as the output variables. For the fast walking mode, the response function is described as the physiological arousal degree showing a linear increase with the increase of the moving speed, and when the moving speed increases from 1 meter per second to 1.8 meters per second, the skin conductance level increases by 0.6 microsiemens per second. For the standing and observing mode, the response function is described as the physiological arousal degree rapidly rising at the beginning of the stay, with an increase rate of 2 microsiemens per 10 seconds, and after reaching the peak value, it enters the platform period, and the duration of the platform period is related to the complexity of the observation content, and for every 1 point increase in the complexity score, the platform period is extended by 15 seconds. For the frequent interaction mode, the response function is described as the physiological arousal degree showing a periodic pulse characteristic, with a peak value of arousal degree triggered once for each interaction action, with a peak amplitude of 1.4 times to 1.8 times of the baseline value, and a decay time constant of 25 seconds.

[0040] When performing scene semantic annotation on the physiological response mechanism based on the correlation mapping relationship, the spatial semantic attribute and the physiological response mode are coupled and analyzed. In the exhibition area annotated as high information density, when the visitor appears the standing and observing mode, the peak value of physiological arousal degree reaches an average of 16 microsiemens, and the annotation content is cognitive processing type arousal; in the experience area annotated as high interaction demand, when the visitor appears the frequent interaction mode, the pulse frequency of physiological arousal degree is 4 to 6 times per minute, and the annotation content is behavior participation type arousal. The system establishes a semantic annotation rule library, and when the mapping weight of the movement state mode and the scene semantic attribute is greater than 0.7 and the change rate of the physiological arousal degree is greater than 0.15, the automatic annotation process is triggered.

[0041] When the physiological response mechanism is combined with the physiological arousal response function, a multi-dimensional feature fusion matrix is constructed. The row vector of the matrix represents different time segments, and the column vector contains the motion state mode identifier, the physiological arousal value, the response function output value, and the scene semantic label. For the 8th time segment, the matrix records the slow browsing mode, the skin conductance value of 11 microsiemens, the response function output value of 0.72, and the high information density label of the exhibition area. The system calculates the cosine similarity of the feature vectors between adjacent time segments, and determines that the spatiotemporal collaborative change mode occurs when the similarity is greater than 0.82, indicating that the motion state transition and the physiological arousal change occur synchronously.

[0042] When the collaborative strength of the visitor's physical behavior and psychological state is quantified, the collaborative strength indicator is defined as the product of the motion state change amplitude and the physiological arousal change amplitude divided by the time interval. In a specific numerical case, the visitor transitions from the slow browsing mode to the observation mode, the step frequency decreases from 100 steps per minute to 20 steps per minute, the change amplitude is 80 steps, and the skin conductance level rises from 10 microsiemens to 14 microsiemens, the change amplitude is 4 microsiemens, and the time interval is 10 seconds. The calculated collaborative strength indicator value is 32. The system sets the collaborative strength classification standard, and the value less than 10 is weak collaboration, the value between 10 and 30 is moderate collaboration, and the value greater than 30 is strong collaboration.

[0043] When the collaborative strength is introduced into the reconstruction process of feature expression, a weighted fusion strategy is used to adjust the contribution proportion of each feature dimension. For strong collaborative time, the weight coefficient of the motion state feature is set to 0.6, and the weight coefficient of the physiological arousal feature is set to 0.4; for weak collaborative time, the weight coefficients are adjusted to 0.5 and 0.5, respectively. Through spatial mapping transformation, the original feature space is projected into a low-dimensional representation space, and the dimension of the projected feature vector is compressed from the original 48 dimensions to 12 dimensions, retaining the principal components with a cumulative variance contribution rate of 89%. When generating new spatiotemporal correlation patterns, the system identifies three key transition nodes in the entire tour of the visitor, located at the 15th minute, the 52nd minute, and the 87th minute, respectively. The collaborative strength indicators at these nodes are all greater than 35, and the corresponding scene transition paths are the exhibition area to the experience area, the experience area to the rest area, and the rest area to the exhibition area. The constructed embodied perception model encodes these spatiotemporal correlation patterns into state transition sequences, each state containing motion mode identifier, physiological arousal level, scene semantic information, and collaborative strength value, forming a complete representation framework of the visitor's psychosomatic state collaborative relationship.

[0044] In an alternative embodiment, based on the embodied perception feature vector and the scene semantic information, a dynamic affinity matrix of the visitor's perception state and the spatial region is established, and the dynamic affinity matrix is mapped to the spatial topology of the lighting device to form a perception-driven regional lighting coupling map, comprising: perform semantic hierarchical decomposition based on the embodied perception feature vector to generate sentiment tendency features and behavior preference features reflecting the visitor perception state, and combine spatial function attributes and environment atmosphere attributes in the scene semantic information to calculate matching degree indexes of the sentiment tendency features and the behavior preference features in each spatial region; construct a dynamic affinity matrix of the visitor perception state and the spatial region according to the matching degree indexes, wherein a matrix element reflects response intensity of the visitor perception state to a specific spatial region, and the dynamic affinity matrix is dynamically reconstructed through matrix iteration update to reflect evolution trends of the sentiment tendency features and the behavior preference features; establish a spatial structure map of the lighting devices, analyze spatial region coverage ranges of each lighting device and topological connection rules between adjacent devices, construct state distribution of lighting device nodes using the response intensity in the dynamic affinity matrix, and generate a perception-driven regional lighting coupling map according to the topological connection rules.

[0045] After receiving the embodied perception feature vector, the feature vector is subjected to multi-level semantic analysis processing. The processing process adopts a feature separation module to project the embodied perception feature vector to two independent representation dimensions of emotional semantic space and behavior semantic space. In the emotional semantic space, the system extracts sentiment tendency features of the visitor, and the features include numerical representations of three dimensions of emotional arousal, pleasure and dominance. In a specific example, when the facial micro-expression of the visitor presents a smiling state and the body posture presents a relaxed state, the system calculates that the emotional arousal value is 0.65, the pleasure value is 0.78, and the dominance value is 0.52. In the behavior semantic space, the system extracts behavior preference features of the visitor, and the features reflect activity type tendency and spatial use preference of the visitor. When the visitor maintains a sitting posture and the line of sight is concentrated in a specific direction, the system determines that the behavior preference feature is a static reading type, and the activity intensity value is 0.32.

[0046] The scene semantic information database is synchronously called to obtain the function attribute label and the environment atmosphere label of each space region. A space region is marked as a meeting negotiation function, and the function attribute vector of the space region includes a privacy index of 0.83, an interaction index of 0.71, and a formality index of 0.69. The environment atmosphere attribute vector of the space region includes a brightness expectation value of 0.75, a warmth expectation value of 0.48, and a concentration expectation value of 0.86. The system calculates the semantic similarity between the emotional tendency feature of the visitor and the environment atmosphere attribute of the space region, and performs point product operation and normalization processing between the feature vectors. When the joy degree 0.78 in the emotional tendency feature of the visitor and the warmth expectation value 0.48 of the space region are weighted and calculated, the weight coefficient is set to 0.6 according to the importance of the emotional dimension, and the emotional matching component 0.51 is obtained. At the same time, the matching component 0.67 of the dominance 0.52 of the visitor and the privacy index 0.83 of the space region is calculated.

[0047] The behavior preference feature of the visitor is matched with the function attribute of the space region. When the visitor presents a static reading behavior preference, the behavior feature has a negative matching effect with the interaction index 0.71 of the meeting negotiation function region, and the behavior matching degree value calculated by the system is 0.38. When the behavior preference feature of the visitor is matched with another space region marked as a quiet rest function, the behavior matching degree value calculated is 0.82 because the quietness index of the region is 0.89. The system integrates the emotional matching degree and the behavior matching degree, and generates a final matching degree index by using a weighted fusion strategy. The comprehensive matching degree index of the meeting negotiation region is 0.47, and the comprehensive matching degree index of the quiet rest region is 0.79.

[0048] Based on the matching degree index of each space region, a dynamic affinity matrix is constructed. The row dimension of the matrix corresponds to different visitor perception state types, and the column dimension corresponds to each space region identifier. The value of each element in the matrix represents the response strength of a specific perception state to a specific space region. The third row of the matrix corresponds to the visitor perception state detected at the current moment, and each element of the row is: the response strength of region A is 0.47, the response strength of region B is 0.79, the response strength of region C is 0.61, and the response strength of region D is 0.35. The system continuously monitors the change of the perception state of the visitor. When it is detected that the emotional tendency feature of the visitor changes, for example, the joy degree decreases from 0.78 to 0.54, the system recalculates the matching degree index of each region and updates the element value of the corresponding row in the dynamic affinity matrix.

[0049] The matrix iterative updating mechanism adopts a sliding time window strategy, and stores the dynamic affinity matrix data of the last ten time steps. At the Nth time step, the system calculates the weighted average of the current matrix element and the historical matrix element, and the weight coefficient of the historical data decays over time. The historical matrix weight one time step away from the current time is zero point eight, the historical matrix weight two time steps away from the current time is zero point sixty-four, and so on. Through this iterative updating process, the system can capture the evolution trend of the visitor perception state and avoid abnormal changes in response strength caused by instantaneous fluctuations.

[0050] A spatial structure map of the lighting devices is constructed, which abstracts the lighting devices as nodes and the spatial adjacency relationship as edge connection. In a scene containing twelve lighting devices, the system records the spatial coordinate position and irradiation range parameters of each lighting device. The position coordinates of lighting device No. 1 are 3.2 meters in the horizontal direction and 5.7 meters in the vertical direction, and its irradiation radius is 2.5 meters, covering area A and part of area B. The position coordinates of lighting device No. 2 are 6.8 meters in the horizontal direction and 5.4 meters in the vertical direction, and its irradiation radius is 3.1 meters, covering area B and part of area C. The system establishes the topological connection relationship between the lighting devices according to the overlap of the covered areas. When the irradiation ranges of two lighting devices have an intersection, an edge connection is established in the map, and the weight value of the edge reflects the overlap area ratio. The edge weight between lighting device No. 1 and No. 2 is 0.37, indicating that the overlap area accounts for 37% of the total coverage area of the two devices.

[0051] The response strength values of each space region in the dynamic affinity matrix are extracted, and these values are mapped to the lighting device nodes corresponding to the space regions. The response strength of area B is 0.79, which is covered by lighting devices No. 1, No. 2, and No. 5. The system allocates the response strength according to the coverage ratio of each device in the area. The coverage ratio of lighting device No. 1 in area B is 0.223, and the allocated node state value is 0.18. The coverage ratio of lighting device No. 2 in area B is 0.58, and the allocated node state value is 0.46. The coverage ratio of lighting device No. 5 in area B is 0.19, and the allocated node state value is 0.15.

[0052] Based on the state distribution and topology connection rule of the lighting device nodes, a perception-driven regional lighting coupling graph is generated. Each node in the graph carries a state value, and the edge connection between nodes retains the spatial topology relationship. The system applies a graph propagation mechanism, so that the state values of adjacent nodes influence each other. The initial state value of lighting device number two is zero point four six, and the state values of adjacent lighting devices number one and number three are zero point one eight and zero point three two respectively. The system calculates the updated state value of number two node after being affected by adjacent nodes as zero point four one. The update process considers the adjustment effect of edge weight, and the edge weight between number one and number two is zero point three seven, which determines the influence degree of number one on number two. Through multiple rounds of graph propagation iteration, the system finally forms a stable coupling graph state distribution, which directly drives the generation of control instructions of each lighting device, and realizes the dynamic coupling response of perception state and regional lighting.

[0053] In an optional embodiment, according to the weight distribution of each node in the regional lighting coupling graph and the clustering results of the perception state of the visitor group, the spectral energy distribution scheme and the color temperature transition sequence of each lighting region are solved by a multi-objective optimization algorithm, and the partitioned cooperative lighting control instruction is generated, including: The perception state of the visitor group is clustered and analyzed, the visitor sub-group with similar perception state characteristics is identified, and the distribution density of each visitor sub-group in different space regions is counted, and the distribution association relationship between the visitor sub-group and the space region is established; Combining the weight distribution of each node in the regional lighting coupling graph and the distribution association relationship, the comprehensive service weight of each lighting region to different visitor sub-groups is calculated, and a multi-objective optimization function is constructed based on the comprehensive service weight; The multi-objective optimization function is solved by a multi-objective optimization algorithm to determine the spectral energy distribution scheme of each lighting region at different time periods, based on the spectral energy distribution scheme, the color temperature gradient between adjacent lighting regions is calculated, and the color temperature transition sequence ensuring visual comfort is generated according to the color temperature gradient and the moving track of visitors between adjacent lighting regions; The spectral energy distribution scheme and the color temperature transition sequence are converted into control parameters executable by the lighting device to generate partitioned cooperative lighting control instructions.

[0054] In the clustering analysis of the perception state of the visitor group, the physiological and behavioral characteristic data of the visitors, such as the pupil diameter change rate, the blink frequency, the head orientation stability and the stay time, are collected. For the collected data, the system classifies the visitors with the pupil diameter change rate less than 5%, the blink frequency less than 15 times per minute into a comfortable perception state sub-group; the visitors with the pupil diameter change rate between 5% and 12%, the blink frequency between 15 and 25 times per minute into a neutral perception state sub-group; and the visitors with the pupil diameter change rate more than 12%, the blink frequency more than 25 times per minute into an uncomfortable perception state sub-group. In the actual application of a certain exhibition space, the system identifies that thirty-two visitors belong to the comfortable perception state sub-group, forty-five visitors belong to the neutral perception state sub-group, and eighteen visitors belong to the uncomfortable perception state sub-group.

[0055] The distribution density of each visitor sub-group in different space regions is counted, and the exhibition space is divided into several grid cells, and the side length of each grid cell is set to three meters. The system obtains the real-time position coordinates of the visitors through the positioning device, and calculates the number proportion of each visitor sub-group in each grid cell. In a certain period, the grid cell A in the entrance hall region detects eight visitors in the comfortable perception state sub-group, twelve visitors in the neutral perception state sub-group, and five visitors in the uncomfortable perception state sub-group; the grid cell B in the main display region detects fifteen visitors in the comfortable perception state sub-group, twenty visitors in the neutral perception state sub-group, and seven visitors in the uncomfortable perception state sub-group. The system establishes a distribution association relationship matrix of the visitor sub-group and the space region based on these statistical data, which records the number of each visitor sub-group corresponding to each space region and the proportion information thereof.

[0056] The weight distribution of each node in the regional lighting coupling graph is combined for comprehensive analysis, wherein the node weight reflects the influence degree of the lighting region on the overall visual experience. For the main display region node B, the weight value thereof in the regional lighting coupling graph is 0.75, indicating that the region has a greater influence on the overall perception of the visitors; for the transition corridor region node C, the weight value thereof is 0.42. The system performs weighted calculation on the node weight value and the distribution density of the visitor sub-group, to obtain the comprehensive service weight of each lighting region to different visitor sub-groups. In the specific calculation, the weight value 0.75 of the node B is multiplied by the comfortable perception state sub-group proportion, the neutral perception state sub-group proportion and the uncomfortable perception state sub-group proportion in the region respectively, to obtain the service weights of the region to the three sub-groups as 0.26, 0.35 and 0.12 respectively.

[0057] A multi-objective optimization function is constructed based on the comprehensive service weights, which considers three optimization objectives of improving the satisfaction of the comfort perception state sub-population, improving the perception state of the discomfort perception state sub-population, and reducing the overall energy consumption. In the construction process, the system takes the sum of the comprehensive service weights of each lighting area to the comfort perception state sub-population as the evaluation index of the first optimization objective, takes the product sum of the distribution density of the discomfort perception state sub-population in each area and the lighting improvement potential of the area as the evaluation index of the second optimization objective, and takes the total power consumption of all lighting devices as the evaluation index of the third optimization objective. The system sets different priority coefficients for the three objectives, which are 0.5, 0.35 and 0.15 respectively, reflecting the design intention of prioritizing the visitor's perception demand.

[0058] A multi-objective optimization algorithm based on the Pareto front is used to solve the optimization function, and a balanced solution between multiple objectives is found through iterative search. In each iteration, the algorithm generates a set of candidate spectral energy distribution schemes, each containing the energy distribution proportion of each lighting area in the red, green and blue light bands. In a certain iteration, for the main display area node B, the candidate scheme generated by the algorithm sets the red light band energy proportion to 35%, the green light band energy proportion to 45%, and the blue light band energy proportion to 20%. The system evaluates the contribution value of the scheme to the three optimization objectives and compares it with other candidate schemes, and retains the non-dominated solutions to form a Pareto front set. After two hundred iterations, the system selects the scheme with the highest comprehensive score from the Pareto front set as the final spectral energy distribution scheme.

[0059] The color temperature values of each lighting area are calculated according to the determined spectral energy distribution scheme, by substituting the energy proportions of the red, green and blue light bands into the color temperature calculation model. For the main display area node B, according to its energy distribution of 35% red light, 45% green light and 20% blue light, the calculated color temperature value is 4200K; for the adjacent transition corridor area node C, its spectral energy distribution is 40% red light, 42% green light and 18% blue light, and the calculated color temperature value is 3800K. The system calculates the color temperature gradient between adjacent lighting areas, i.e. the color temperature difference between node B and node C is 400K.

[0060] According to the moving trajectory data of the visitor between adjacent lighting areas, it is analyzed that the average moving speed of the visitor from the main display area to the transition corridor area is 0.8 meters per second, and the average time length through the area is 12 seconds. According to the color temperature gradient of 400 Kelvin and the moving time length of 12 seconds, the system generates a color temperature transition sequence, divides the transition corridor area into four transition sections, and sets the color temperature values of each transition section to 4,200 Kelvin, 4,000 Kelvin, 3,900 Kelvin and 3,800 Kelvin in turn, so as to ensure that the visitor experiences smooth color temperature change during the movement and avoid discomfort caused by visual mutation.

[0061] The spectral energy allocation scheme and the color temperature transition sequence are converted into control parameters executable by the lighting device. For the lighting device using adjustable spectrum, the system generates a parameter combination containing red light channel output power, green light channel output power and blue light channel output power. For the lighting device of the main display area, the red light channel output power is set to 21 watts, the green light channel output power is set to 27 watts, and the blue light channel output power is set to 12 watts. The system encapsulates these control parameters into a standardized instruction package containing device identification code, channel power value, execution timestamp, transition time length and other information, and sends them to the lighting devices in each area through the lighting control bus to realize partitioned and coordinated lighting control.

[0062] In an optional implementation, based on the spectral energy allocation scheme, a color temperature gradient between adjacent lighting areas is calculated, and a color temperature transition sequence ensuring visual comfort is generated according to the color temperature gradient and the moving trajectory of the visitor between adjacent lighting areas, comprising: According to the spectral intensity distribution of each lighting area in the spectral energy allocation scheme, the equivalent color temperature values of each lighting area are calculated, and the spatially adjacent lighting area pairs are identified. The color temperature difference value between the adjacent lighting area pairs is calculated to form a color temperature gradient representing the unevenness of the color temperature spatial distribution. The moving trajectory and spatio-temporal path characteristics of the visitor from the starting lighting area to the target lighting area are analyzed in combination with the color temperature gradient. The color temperature change rate perceived by the visitor during the movement is calculated, the mapping relationship between the color temperature change rate and the visual comfort is established, and the color temperature change rate threshold ensuring visual comfort is determined. According to the color temperature change rate threshold, when the color temperature change rate exceeds the threshold, an intermediate color temperature regulation node is inserted between the starting lighting area and the target lighting area, and the color temperature setting value and the duration of each intermediate color temperature regulation node are calculated according to the visitor moving speed and the path length. Based on the equivalent color temperature values and the color temperature setting values, the color temperature values of the starting lighting area, the intermediate color temperature regulation nodes and the target lighting area are sorted and combined according to the visitor moving sequence to generate a color temperature transition sequence ensuring visual comfort.

[0063] In the implementation of the color temperature transition sequence generation process, by analyzing the spectral intensity distribution data of each lighting area in the spectral energy allocation scheme, the spectral power distribution curve of each lighting area in the wavelength range of 380 nanometers to 780 nanometers is obtained. For the lighting area A1 of the exhibition hall A area, the spectral intensity in the 470 nanometer band is 0.35 watts per nanometer, the spectral intensity in the 555 nanometer band is 0.68 watts per nanometer, and the spectral intensity in the 650 nanometer band is 0.42 watts per nanometer. The system matches the spectral power distribution curve with the standard blackbody radiation curve, and determines the closest blackbody radiation temperature by comparing the spectral shape similarity point by point. When the similarity reaches 0.96, it is determined that the equivalent color temperature value of the lighting area A1 is 4200 kelvin. For the adjacent lighting area A2, the spectral power distribution curve has an intensity of 0.28 watts per nanometer at 470 nanometers, 0.52 watts per nanometer at 555 nanometers, and 0.58 watts per nanometer at 650 nanometers. Through the same matching calculation, the equivalent color temperature value is 3600 kelvin.

[0064] The spatial topology analysis module identifies the adjacent relationship of each lighting area in the exhibition hall, and constructs a spatial adjacency matrix based on the center point coordinates of the lighting area. The center coordinates of the lighting area A1 are X axis 12 meters and Y axis 8 meters in the exhibition hall coordinate system, and the center coordinates of the lighting area A2 are X axis 18 meters and Y axis 8 meters. The center distance between the two is 6 meters and there is no other lighting area separation, which is determined as an adjacent lighting area pair. The color temperature difference value of the lighting area pair is calculated, and the color temperature difference value of 600 kelvin is obtained by subtracting the color temperature of 4200 kelvin of the lighting area A1 from the color temperature of 3600 kelvin of the lighting area A2. The color temperature gradient of the region pair is 100 kelvin per meter, which is obtained by dividing the color temperature difference value by the center distance of 6 meters of the two lighting areas. The system iteratively calculates all adjacent lighting area pairs in the exhibition hall to form a complete color temperature spatial distribution map, in which the color temperature gradient between the lighting areas B3 and B4 is 75 kelvin per meter, and the color temperature gradient between the lighting areas C5 and C6 is 120 kelvin per meter.

[0065] The visitor tracking system collects visitor movement data through the depth perception device deployed on the ceiling of the exhibition hall, recording the complete trajectory of the visitor moving from lighting area A1 to lighting area A2. A certain visitor enters lighting area A1 at time T0, reaches the boundary between the two lighting areas at time T15 seconds, and completely enters lighting area A2 at time T28 seconds. The system extracts the visitor's moving speed as 0.4 meters per second and the moving path length as 11.2 meters. According to the visitor's position change on the path and the color temperature environment data at the corresponding time, the visitor's perceived color temperature change is calculated. Within 3 seconds of entering the boundary area, the ambient color temperature changes from 4200 Kelvin to 3900 Kelvin, with a color temperature change of 300 Kelvin. The color temperature change rate during this period is 100 Kelvin per second.

[0066] The built-in visual comfort assessment model is based on the physiological response characteristics of the human eye to color temperature changes. This model includes visual comfort score data under different color temperature change rates. When the color temperature change rate is 40 Kelvin per second, the visual comfort score is 8.5, when the color temperature change rate is 80 Kelvin per second, the score drops to 6.2, and when the color temperature change rate reaches 120 Kelvin per second, the score is only 3.8. The system establishes a mapping relationship curve between color temperature change rate and visual comfort by fitting these data points, and sets the acceptable threshold for visual comfort score as no less than 7 points. By reverse calculation, the color temperature change rate threshold is 60 Kelvin per second.

[0067] For the scenario of the visitor moving from lighting area A1 to lighting area A2, the actual color temperature change rate of 100 Kelvin per second exceeds the threshold of 60 Kelvin per second, and the intermediate color temperature control node insertion mechanism is started. The system calculates that the entire color temperature transition process needs to be extended to at least 10 seconds based on the difference between the starting color temperature of 4200 Kelvin and the target color temperature of 3600 Kelvin, combined with the threshold of 60 Kelvin per second and the visitor's moving speed of 0.4 meters per second. According to the visitor's moving path length of 11.2 meters and the moving time of 28 seconds, the system sets the first intermediate control node at the midpoint of the path, i.e. 5.6 meters from the starting point, corresponding to the visitor's moving to time T14 seconds. The color temperature setting value of this node is determined by the linear interpolation method. According to the position of the visitor in the middle of the starting and ending points, the color temperature is set to 3900 Kelvin, and the time period for maintaining this color temperature value is 3 seconds before and after the visitor passes through the node.

[0068] A second intermediate control node is inserted between the starting lighting area and the first intermediate node, located at a distance of 2.8 meters from the starting point, corresponding to time T7 seconds, with a color temperature set value of 4050 Kelvin, and a duration of 2 seconds before and after. Similarly, a third intermediate control node is inserted between the first intermediate node and the target lighting area, located at a distance of 8.4 meters from the starting point, corresponding to time T21 seconds, with a color temperature set value of 3750 Kelvin, and a duration of 2 seconds before and after. Through this multi-node segmented control mode, the maximum color temperature change rate perceived by the visitor when moving between any adjacent nodes is reduced to 53 Kelvin per second, meeting the visual comfort requirements.

[0069] The color temperature value of the starting lighting area A1, 4200 Kelvin, is marked as the starting point of the sequence, with a time marker of T0 to T5 seconds. The second intermediate control node color temperature value, 4050 Kelvin, is marked as the second item of the sequence, with a time marker of T5 to T9 seconds. The first intermediate control node color temperature value, 3900 Kelvin, is marked as the third item of the sequence, with a time marker of T11 to T17 seconds. The third intermediate control node color temperature value, 3750 Kelvin, is marked as the fourth item of the sequence, with a time marker of T19 to T23 seconds. The color temperature value of the target lighting area A2, 3600 Kelvin, is marked as the end point of the sequence, with a time marker of T26 seconds to the visitor leaving the area. This complete color temperature transition sequence ensures smooth and gradual changes in the visitor's color temperature perception during the entire movement, avoiding visual discomfort caused by sudden changes. The lighting control system dynamically adjusts the spectral output of the intelligent lighting devices in each area according to the sequence instructions, realizing synchronous color temperature environment adaptation with the visitor's movement.

[0070] In an optional implementation, the changes in the behavior trajectory of the visitor after regulation and the fluctuations in the physiological indicators are monitored, and the changes in the behavior trajectory and the fluctuations in the physiological indicators are used as correction signals to perform online evolution on the calculation strategy of the spatiotemporal correlation coding weight in the embodied perception model and the dynamic affinity matrix, which includes: The behavior trajectory change rule of the visitor after regulation is monitored, the real-time fluctuation characteristics of the physiological indicators in the behavior trajectory change process are analyzed, and dynamic feedback data reflecting the response effect of the visitor are generated; the fluctuation characteristics in the dynamic feedback data are used to adaptively adjust the spatiotemporal correlation coding weight of the embodied perception model, an iterative optimization mechanism of model parameters is constructed; and according to the adjustment result of the iterative optimization mechanism, the calculation strategy of the dynamic affinity matrix is reconstructed, realizing the collaborative evolution of the perception model and the affinity matrix.

[0071] In the specific implementation process of monitoring the changes in visitor behavior trajectories and the fluctuations in physiological indicators after regulation, a multi-sensor network deployed in the exhibition space continuously collects visitor movement data. The infrared positioning sensors arranged in the space record visitor position coordinates at a frequency of 10 times per second, while the physiological monitoring devices worn by the visitors continuously collect indicators such as heart rate, skin electrical response, and respiratory rate. When the system implements light regulation for a visitor, the visitor's stay time in the first exhibition area is extended from the original 45 seconds to 78 seconds, the moving speed is reduced from 0.8 meters per second to 0.5 meters per second, the heart rate value is decreased from 82 beats per minute to 74 beats per minute, and the skin conductance value is increased from 2.3 microsiemens to 3.7 microsiemens. These numerical changes are transmitted in real time to the data processing module, forming a multi-dimensional data stream containing timestamps, spatial coordinates, and physiological parameters.

[0072] The data processing module performs continuity analysis on the collected trajectory data to extract the differences in visitor movement patterns before and after regulation. The system identifies that the visitor's stop frequency in front of a specific exhibit increases, changing from a single pass before regulation to three round trips after regulation, and the eye gaze stay time is extended from 12 seconds to 34 seconds. At the same time, the physiological indicator analysis subsystem performs sliding window processing on the heart rate data, with a window length of 5 seconds, calculating the mean and variance within each window. It is detected that the visitor's heart rate shows peak fluctuations when viewing a certain exhibit, with the value rising from 74 to 89 in 3 seconds and then falling to 76, and this fluctuation pattern is marked as an emotional arousal event. After baseline correction processing of the skin electrical response data, the system identifies that the rising slope of the conductance value reaches 0.28 microsiemens per second, with a duration of 8 seconds, and this feature is classified as a strong interest response signal.

[0073] The dynamic feedback data generation module time-aligns and fuses the trajectory change data with the physiological fluctuation features, establishing a data fusion table based on the time axis, recording the corresponding spatial position, moving speed, heart rate value, skin conductance, and respiratory rate at each time point. When the visitor stops in front of a certain exhibit in the second exhibition area, the system captures a stationary state with position coordinates remaining within a radius of 0.3 meters for 25 seconds, simultaneously monitors a heart rate trend that first rises and then stabilizes, and skin conductance maintains a high level of 3.5 microsiemens or above. This combination of features is integrated into a strong positive response label, with a response intensity coefficient assigned as 0.85. Conversely, when the visitor quickly passes through the third exhibition area, the moving speed reaches 1.2 meters per second, the stay time is only 6 seconds, and the heart rate remains stable without significant fluctuations, this scenario is marked as a weak response, with an intensity coefficient assigned as 0.15.

[0074] The spatiotemporal correlation coding weight adjustment process of the body perception model is based on the response intensity coefficient in the dynamic feedback data. The model maintains a weight vector inside, which contains multiple components such as spatial distance sensitivity, time persistence weight, and physiological arousal influence factor. When a strong positive response signal is received, the system extracts the spatial feature code corresponding to the signal and adjusts the spatial sensitivity weight of the region from the initial value of 0.6 to 0.78. The adjustment amplitude is calculated by the deviation of the response intensity coefficient and the current weight value, and the deviation value is multiplied by the preset learning rate parameter 0.15 and then added to the original weight. For the scene of the second exhibition area producing strong response, the time persistence weight is updated from 0.55 to 0.68, reflecting that the long-time stay of visitors in this position has higher prediction value. The physiological arousal influence factor is adjusted according to the change rate of skin conductance, and when a rapid rising mode is detected, the factor is enhanced from 0.48 to 0.62, making the model pay more attention to the mutation signal of physiological indicators.

[0075] The iterative optimization mechanism of model parameters adopts a sliding update strategy, triggering a parameter update every time 100 sets of valid feedback data are accumulated. A historical response data cache pool is maintained to store the latest 500 visitor response records. During the update process, the system calculates the average response intensity in each spatial grid cell and identifies 8 high-value regions with response intensity exceeding 0.7. The coding weights of these regions are collectively enhanced, and the enhancement amplitude is positively correlated with the number of samples in the region. When a grid accumulates more than 50 high-response samples, the weight enhancement amplitude reaches 0.12, while the enhancement amplitude of a grid with less than 10 samples is only 0.03. The model also adjusts the time decay parameter. For the exhibit area that produces a response quickly, the decay coefficient is reduced from the standard value of 0.85 to 0.92, indicating that the behavior in the initial contact stage of visitors has a longer influence period.

[0076] The calculation strategy of the dynamic affinity matrix is reconstructed based on the optimized model parameters. The matrix dimension is set to the number of exhibition areas multiplied by the number of exhibit types, forming a 15 by 20 two-dimensional structure. Each element in the matrix represents the matching degree score of a specific exhibition area and a specific exhibit type. During the reconstruction process, the system reads the updated spatial sensitivity weight and amplifies the entire matrix row corresponding to the high-weight region. The values of all elements in the second exhibition area row are multiplied by the amplification coefficient 1.18, giving this region a higher priority in affinity calculation. For exhibit types that induce strong physiological responses, the system enhances the values of the corresponding matrix columns. The baseline value of the historical cultural relics column is adjusted from 0.58 to 0.71, reflecting that this type of exhibit has a significant attraction to visitors.

[0077] The adjustment of the matrix calculation strategy also involves dynamic modeling of cross-area relevance, analyzing the transfer mode of visitors between different exhibition areas, and finding that 68% of visitors moving from a first exhibition area to a second exhibition area generate high response behavior in the second exhibition area. Based on this finding, the system introduces a region transfer gain factor in the affinity matrix. When a visitor is currently located in a first exhibition area and the target is a second exhibition area, the affinity score of the second exhibition area is additionally increased by 0.14. The value of this factor is obtained by multiplying the statistical historical transfer success rate by the average response intensity. For the visitor group with a dramatic fluctuation in physiological indicators, the system allocates a dedicated affinity matrix version for them, which strengthens the weight proportion of the physiological arousal factor from 20% to 35% in the standard version.

[0078] The co-evolution of the perception model and the affinity matrix is achieved through a closed-loop feedback mechanism. After each regulation and decision execution, the system records the deviation value between the predicted affinity and the actual response intensity. When the predicted value of a certain exhibition area is 0.72 and the actual response intensity is only 0.54, the deviation value 0.18 is fed back to the model training module. The model adjusts the encoding weight of the corresponding feature according to the deviation direction to reduce the over-optimistic prediction tendency. After processing 200 regulation cycles, the overall prediction accuracy of the system is improved from 64% to 79%, and the average deviation value is reduced from 0.21 to 0.09, verifying the effectiveness of the evolution mechanism.

[0079] In a second aspect of the embodiments of the present application, a self-adaptive lighting dynamic regulation system based on embodied perception for a tourism and travel scene is provided, comprising: A first unit is configured to capture multi-dimensional behavior data of visitors and scene environment state data in real time through a distributed sensing network deployed in a target tourism and travel scene; an embodied perception model is constructed by fusing the spatial position, moving speed, stay duration and physiological arousal of visitors, the multi-dimensional behavior data is spatio-temporally correlated and encoded to generate an embodied perception feature vector representing the immersion state of visitors; A second unit is configured to establish a dynamic affinity matrix of visitor perception state and spatial area based on the embodied perception feature vector and scene semantic information, map the dynamic affinity matrix to the spatial topology of lighting devices to form a perception-driven regional lighting coupling graph; A third unit is configured to solve the spectral energy distribution scheme and color temperature transition sequence of each lighting area through a multi-objective optimization algorithm according to the weight distribution of each node in the regional lighting coupling graph and the perception state clustering result of the visitor group, and generate partitioned collaborative lighting regulation instructions; A fourth unit is configured to drive the lighting devices to execute the lighting regulation instructions, and monitor the changes in visitor behavior trajectory and physiological indicator fluctuations after regulation, and use the trajectory changes and physiological indicator fluctuations as correction signals to perform online evolution on the spatio-temporal correlation encoding weight in the embodied perception model and the calculation strategy of the dynamic affinity matrix.

[0080] In a third aspect, the present application provides an electronic device, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to invoke the instructions stored in the memory to perform the method described above.

[0081] In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method described above.

[0082] The present application can be a method, apparatus, system, and / or computer program product. Computer program products can include computer-readable storage media having computer-readable program instructions loaded thereon for performing various aspects of the present application.

[0083] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for dynamically adjusting and controlling a travel and tourism scene based on embodied perception, characterized in that, The method comprises the following steps: Real-time capture of multi-dimensional behavior data and scene environment state data of visitors through a distributed sensing network deployed in a target tourism scene; construction of an embodied perception model integrating visitor spatial position, moving speed, stay duration and physiological arousal degree; spatio-temporal correlation coding of the multi-dimensional behavior data to generate an embodied perception feature vector representing the immersion state of the visitors; Based on the embodied perception feature vector and scene semantic information, a dynamic affinity matrix of visitor perception state and spatial region is established, and the dynamic affinity matrix is mapped to the spatial topology of the lighting device to form a perception-driven regional lighting coupling graph; According to the weight distribution of each node in the regional lighting coupling graph and the perception state clustering result of the visitor group, the spectral energy distribution scheme and color temperature transition sequence of each lighting region are solved through a multi-objective optimization algorithm to generate a partitioned collaborative lighting control instruction; Driving the lighting device to execute the lighting control instruction, and monitoring the changes of visitor behavior trajectory and physiological index fluctuation after control, taking the trajectory changes and physiological index fluctuations as correction signals to online evolve the spatio-temporal correlation coding weight in the embodied perception model and the calculation strategy of the dynamic affinity matrix.

2. The method of claim 1, wherein, The construction of an embodied perception model integrating visitor spatial position, moving speed, stay duration and physiological arousal degree, spatio-temporal correlation coding of the multi-dimensional behavior data to generate an embodied perception feature vector representing the immersion state of the visitors comprises: Establishing a spatio-temporal trajectory representation of visitor spatial position and moving speed, generating a motion state feature reflecting the spatial behavior continuity of the visitors by establishing a corresponding relationship between the spatio-temporal trajectory representation and the stay duration; Based on the spatio-temporal distribution characteristics of the motion state feature, combined with physiological arousal degree, system modeling is performed to analyze the dynamic change law of the physiological arousal degree under different motion states, and an embodied perception model representing the coordinated relationship between the physical and mental states of the visitors is constructed; According to the mapping relationship established by the embodied perception model, the visitor spatial position, the moving speed, the stay duration and the physiological arousal degree are organized in spatio-temporal dimensions to obtain the state evolution feature of the visitors in the time dimension and determine the regional response feature in the spatial dimension; According to the internal relationship between the state evolution feature and the regional response feature, a unified mapping space is constructed, and an embodied perception feature vector representing the immersion state of the visitors is established in the unified mapping space.

3. The method of claim 2, wherein, Based on the spatio-temporal distribution characteristics of the motion state feature, combined with physiological arousal degree, system modeling is performed to analyze the dynamic change law of the physiological arousal degree under different motion states, and an embodied perception model representing the coordinated relationship between the physical and mental states of the visitors is constructed, which comprises: Spatio-temporal distribution pattern clustering of the motion state feature, identification of the motion state pattern of the visitors in different tourism scene regions, and establishment of an associated mapping relationship between the motion state pattern and the spatial semantic attribute; Based on the division of the motion state mode, a fluctuation characteristic and a change trend of the physiological arousal degree in a corresponding time period are analyzed, a physiological response mechanism of the visitor under different motion state modes is described by constructing a physiological arousal degree response function, and the physiological response mechanism is semantically annotated based on the correlation mapping relationship; The physiological response mechanism and the physiological arousal degree response function are combined and operated to form a characteristic expression reflecting the spatiotemporal coordinated changes of the motion state and the physiological arousal degree, and the coordinated strength of the visitor's body behavior and psychological state is quantified according to the correlation mapping relationship; The coordinated strength is introduced into the reconstruction process of the characteristic expression, a new spatiotemporal correlation mode is generated through spatial mapping transformation, and an embodied perception model representing the coordinated relationship between the visitor's body and mind is constructed.

4. The method of claim 1, wherein, Based on the embodied perception feature vector and the scene semantic information, a dynamic affinity matrix of the visitor's perception state and the space region is established, the dynamic affinity matrix is mapped to the spatial topology structure of the lighting device, and a perception-driven regional lighting coupling graph is formed, including: Based on the embodied perception feature vector, a semantic hierarchical decomposition is performed to generate emotion tendency features and behavior preference features reflecting the visitor's perception state, and the matching degree indexes of the emotion tendency features and the behavior preference features in each space region are calculated in combination with the spatial function attributes and the environment atmosphere attributes in the scene semantic information; According to the matching degree indexes, a dynamic affinity matrix of the visitor's perception state and the space region is constructed, wherein the matrix elements reflect the response strength of the visitor's perception state to a specific space region, and the dynamic affinity matrix is dynamically reconstructed through matrix iteration update to reflect the evolution trend of the emotion tendency features and the behavior preference features; A spatial structure graph of the lighting device is established, the spatial region coverage of each lighting device and the topological connection law between adjacent devices are analyzed, the state distribution of the lighting device nodes is constructed using the response strength in the dynamic affinity matrix, and a perception-driven regional lighting coupling graph is generated according to the topological connection law.

5. The method of claim 1, wherein, According to the weight distribution of each node in the regional lighting coupling graph and the perception state clustering result of the visitor group, the spectral energy allocation scheme and the color temperature transition sequence of each lighting region are solved by a multi-objective optimization algorithm to generate partition-coordinated lighting control instructions, including: The perception state of the visitor group is clustered and analyzed to identify a visitor subgroup with similar perception state characteristics, and the distribution density of each visitor subgroup in different space regions is counted to establish a distribution association relationship between the visitor subgroup and the space region; In combination with the weight distribution of each node in the regional lighting coupling graph and the distribution association relationship, the comprehensive service weight of each lighting region to different visitor subgroups is calculated, and a multi-objective optimization function is constructed based on the comprehensive service weight; The multi-objective optimization function is solved by a multi-objective optimization algorithm to determine the spectral energy allocation scheme of each lighting region at different time periods, the color temperature gradient between adjacent lighting regions is calculated based on the spectral energy allocation scheme, and a color temperature transition sequence ensuring visual comfort is generated according to the color temperature gradient and the moving track of the visitor between adjacent lighting regions; The spectral energy allocation scheme and the color temperature transition sequence are converted into control parameters executable by the lighting device, generating lighting regulation instructions for zoning coordination.

6. The method of claim 5, wherein, Based on the spectral energy allocation scheme, the color temperature gradient between adjacent lighting areas is calculated, and a color temperature transition sequence that ensures visual comfort is generated according to the color temperature gradient and the moving track of the visitor between adjacent lighting areas. According to the spectral intensity distribution of each lighting area in the spectral energy allocation scheme, the equivalent color temperature value of each lighting area is calculated, and spatially adjacent lighting area pairs are identified. The color temperature difference value between adjacent lighting area pairs is calculated to form a color temperature gradient representing the unevenness of the color temperature spatial distribution. The moving track and spatio-temporal path characteristics of the visitor from the starting lighting area to the target lighting area are analyzed in combination with the color temperature gradient. The color temperature change rate perceived by the visitor during movement is calculated, a mapping relationship between the color temperature change rate and visual comfort is established, and a color temperature change rate threshold that ensures visual comfort is determined. According to the color temperature change rate threshold, when the color temperature change rate exceeds the threshold, intermediate color temperature regulation nodes are inserted between the starting lighting area and the target lighting area. According to the visitor's moving speed and path length, the color temperature setting value and duration of each intermediate color temperature regulation node are calculated. Based on the equivalent color temperature value and the color temperature setting value, the color temperature values of the starting lighting area, the intermediate color temperature regulation nodes, and the target lighting area are sorted and combined according to the visitor's moving time sequence, generating a color temperature transition sequence that ensures visual comfort.

7. The method of claim 1, wherein, Monitoring the changes in the visitor's behavior track and the fluctuations in the physiological indicators after regulation, and using the track changes and the physiological indicator fluctuations as correction signals, the spatio-temporal correlation coding weights in the embodied perception model and the calculation strategy of the dynamic affinity matrix are evolved online. Monitoring the behavior track changes of the visitor after regulation, analyzing the real-time fluctuation characteristics of the physiological indicators during the behavior track changes, generating dynamic feedback data reflecting the visitor's response effect; using the fluctuation characteristics in the dynamic feedback data to adaptively adjust the spatio-temporal correlation coding weights of the embodied perception model, constructing an iterative optimization mechanism for model parameters; according to the adjustment results of the iterative optimization mechanism, reconstructing the calculation strategy of the dynamic affinity matrix, realizing the collaborative evolution of the perception model and the affinity matrix.

8. A dynamic lighting control system for a tourism and travel scene based on embodiment perception, for implementing the method of any one of the preceding claims 1-7, characterized in that, Comprising: A first unit for capturing multi-dimensional behavior data of visitors and scene environment state data in real time through a distributed sensing network deployed in a target tourism scene; constructing an embodied perception model that integrates visitor spatial position, moving speed, stay duration, and physiological arousal; spatio-temporally correlating and coding the multi-dimensional behavior data to generate an embodied perception feature vector representing the visitor's immersive state; A second unit for establishing a dynamic affinity matrix of visitor perception state and spatial area based on the embodied perception feature vector and scene semantic information, mapping the dynamic affinity matrix to the spatial topology of the lighting device, and forming a perception-driven regional lighting coupling graph. a third unit configured to solve a spectral energy distribution scheme and a color temperature transition sequence of each lighting region by a multi-objective optimization algorithm according to a weight distribution of each node in the regional illumination coupling graph and a clustering result of the perceived state of the visitor group, and generate a lighting regulation instruction with partition coordination; a fourth unit configured to drive the lighting device to execute the lighting regulation instruction, and monitor a change in a visitor behavior trajectory and a fluctuation in a physiological index after regulation, and take the change in the trajectory and the fluctuation in the physiological index as a correction signal to perform online evolution on a spatiotemporal correlation coding weight in the embodied perception model and a calculation strategy of the dynamic affinity matrix.

9. An electronic device, comprising: comprise: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to invoke the instructions stored in the memory to execute the method of any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon computer program instructions, wherein, The computer program instructions, when executed by the processor, implement the method of any one of claims 1 to 7.

Images