Personalized illumination adjusting system based on artificial intelligence

Abstract

The invention discloses a personalized illumination adjustment system based on artificial intelligence, and relates to the field of illumination adjustment, and the system comprises an identity recognition module which is used for providing user identity registration, distributing a user ID, and allowing login after identity verification; the attention recognition module is used for capturing a user face image through a camera part, judging the attention concentration direction of the user and generating current state data of the user; the behavior prediction module is used for collecting historical moving paths and dwell time data of the user in the control area, predicting future behaviors of the user through a time sequence prediction algorithm, and generating a behavior track diagram of the user; illumination is adjusted in advance to adapt to the current demand of the user, manual operation is reduced, convenience is improved, the moving path and stay time of the user are predicted, lamplight configuration can be actively adjusted, energy is saved, it is ensured that the user can obtain the optimal illumination effect in the place where illumination is most needed, and the adjusting scheme can adapt to complex and changing use scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of lighting adjustment, and particularly to a personalized lighting adjustment system based on artificial intelligence. Background Art

[0002] In modern life, personalized lighting experiences are increasingly valued. With the development of smart home and Internet of Things technologies, users have higher requirements for environmental comfort and convenience. Especially in various scenarios such as office, study, and leisure, the quality and adaptability of lighting directly affect users' concentration, mood, and overall experience. Therefore, a personalized lighting adjustment system based on artificial intelligence has emerged, aiming to optimize the lighting environment of users through intelligent means to adapt to different activities and situations.

[0003] The existing systems' permission control for lamps stays at the overall area level and does not achieve fine-grained permission management at the lamp level, making it difficult to support personalized needs in multi-user scenarios. Conventional solutions only rely on simple motion sensing and lack face orientation recognition and behavior trajectory prediction technologies, resulting in the inability to predict the lighting needs of the user's visual focus area.

[0004] Moreover, traditional prediction algorithms are only based on historical path analysis and do not combine time series prediction to generate dynamic behavior trajectory maps, causing the lighting adjustment to lag behind actual behavior changes. They ignore the dynamic impact of body movements and specific events on lighting parameters and do not establish a fusion calculation mechanism for body impact coefficients and event impact coefficients, resulting in the inability to automatically adapt to the best lighting mode in special scenarios such as reading, meetings, or rest. Summary of the Invention

[0005] In view of the above-mentioned drawbacks of the existing technology, the present invention provides a personalized lighting adjustment system based on artificial intelligence, which can effectively solve the problems of the existing technology.

[0006] To achieve the above objectives, the present invention is realized through the following technical solutions.

[0007] The present invention discloses a personalized lighting adjustment system based on artificial intelligence, including: An identity recognition module, used to provide user identity registration, assign user IDs, and allow login after identity verification;

[0008] A device management module, used to register and manage the positions, start / stop states, and lighting modes of all lamps in the controlled target area, and set the control permissions of different users for lamp start / stop and lighting modes;

[0009] A focus recognition module, used to capture the user's facial image through a camera component, judge the direction of the user's concentrated attention, and generate the user's current state data;

[0010] A behavior prediction module, which is used to collect historical movement paths and residence time data of users within the control area, predict the future behaviors of users through time series prediction algorithms, and generate a behavior trajectory map of users;

[0011] A region analysis module, which is used to analyze the attention time and frequency of users in each region according to the behavior trajectories predicted by the behavior prediction module, and perform priority sorting on the attention regions;

[0012] A control generation module, which is used to obtain the future residence order and residence time of each region according to the user's priority attention region and predicted behavior trajectory, and correspondingly generate a lamp start-stop timing and duration adjustment plan within the future cycle;

[0013] A body feedback module, which is used to collect body movement data of users within the control area in the current cycle, extract the behavior habit characteristics of users, establish a body model, input the behavior habit characteristics, and output the body influence coefficient of the evaluation behavior habit characteristics on the corresponding lighting requirements;

[0014] An event analysis module, which is used to record various events carried out by users within the control area, analyze the influence degree of different events on lighting requirements based on the event type and duration, and generate an event influence coefficient;

[0015] A strategy generation module, which is used to integrate and process the body influence coefficient and the event influence coefficient, and match the integration result with the corresponding lamp mode adjustment plan in the preset strategy library;

[0016] An adjustment implementation module, which is used to match the lamp start-stop adjustment plan and the lamp mode adjustment plan of the control generation module, and configure the corresponding lamps in the specified cycle.

[0017] Furthermore, in the stage where the attention recognition module determines the direction of the user's concentrated attention, by identifying the eye fixation characteristics and head deflection angle in the user's facial image, the target area currently being focused on by the user is obtained, the eye fixation characteristics and the head deflection angle data are aligned in time series, weighted calculation is performed according to the preset weight, the attention degree index is output, the attention state determination threshold is set, when the obtained attention degree index exceeds the preset threshold, a response signal for the target area is triggered, and the focus point of the user's gaze is matched with the lamps installed within the response signal target area to generate the user's current state data.

[0018] Furthermore, the working logic of the time series prediction algorithm run by the behavior prediction module includes:

[0019] Continuously collect the historical movement path data set and residence time data set of users within the control area. The historical movement path data set contains the spatial coordinate sequence of users at different timestamps, and the residence time data set contains the start time and duration of the user's residence at a specific coordinate point;

[0020] Perform spatio-temporal correlation processing on the historical movement path data set and the residence time data set to construct a three-dimensional spatio-temporal sequence data set;

[0021] Adopt a sliding window mechanism to segment the three-dimensional spatio-temporal sequence data set to generate multiple training subsequences of equal length time periods;

[0022] Input the training subsequences into a pre-constructed recurrent neural network model for time series pattern learning. The model extracts time-dependent relationships through a gating mechanism and establishes a user location transition probability matrix and a residence duration distribution function;

[0023] Based on the user state data at the current moment, call the user location transition probability matrix and the residence duration distribution function for multi-step iterative prediction to generate a predicted path sequence within a preset future period;

[0024] Perform trajectory smoothing processing on the predicted path sequence and adopt a cubic spline interpolation algorithm to generate a continuous user behavior trajectory map.

[0025] Furthermore, the construction process of the recurrent neural network model includes: defining the network structure, including an input layer, a hidden layer, and an output layer. The input layer receives the historical movement path and residence time features of the user, performs non-linear transformation through multiple hidden layers, and introduces non-linear factors through an activation function;

[0026] Set a loss function to quantify the difference between the predicted output and the actual result, and adopt a backpropagation algorithm for parameter optimization. Update the network weights through the gradient descent method to minimize the loss function;

[0027] Through batch input of historical sequence data for multiple rounds of iteration, adjust the network parameters until the model converges to obtain a recurrent neural network model for capturing the characteristics of time series data.

[0028] Furthermore, the region analysis module comprehensively measures the attention time and entry frequency of the user in each region, and integrates the information of the duration and event frequency through weighting. Its calculation formula is: ; In the formula, represents the comprehensive attention score of region , represents the attention time weight factor, represents the total observation time period, represents an indicator function, indicating whether the user is located in region at time point . If the user is located in region at time point If it is within the value of is 1. If the user is at time point is not located within the area then the value of is 0. represents the degree of the user's attention contribution within the time interval ; represents the weight factor of the attention frequency, represents the number of times the user enters the area during the observation period, represents the area the th time the entry event occurs, the time difference from the current moment, represents the area the th time the entry event occurs, the time difference from the current moment, represents the time decay coefficient, represents the number of times the user enters the area during the observation period, represents the total number of areas.

[0029] Furthermore, the operation logic of the control generation module is as follows:

[0030] Based on the user behavior trajectory graph, extract the time series of each predicted staying area and the corresponding staying time period in the future period, and combine with the sorting result of the priority attention areas to determine the pre-activation order and trigger time threshold of the lamps in each area;

[0031] Map the staying time period to the target duration interval of the lamps in the corresponding area, and dynamically adjust the lighting early start coefficient and delay decay coefficient according to the area priority;

[0032] According to the mapping relationship between the area lamp identifier and the physical position, convert the time sequence control parameters into a set of lamp start-stop instructions distributed in the spatial dimension, and form a structured adjustment plan including the lamp identifier, activation timestamp, duration, and shutdown condition.

[0033] Furthermore, the operation logic of the limb model in the limb feedback module includes:

[0034] Real-time collect the joint movement trajectory, limb displacement speed, posture holding duration, and action frequency parameters of the user in the control area, and extract the limb action direction preference coefficient, periodic swing amplitude, and static posture duration ratio characteristics;

[0035] A feature matching library is generated based on historical behavior data set training. The feature matching library contains the correlation mapping relationship between different behavior patterns and light sensitivity. Based on the current extracted feature input index, the behavior prediction of the associated feature and the corresponding light demand tendency value are obtained. The light demand tendency value is quantified as a limb influence coefficient, and strong, medium and weak lighting influence thresholds are set in stages. The lighting influence thresholds correspond to the lighting mode settings of the matching lamps.

[0036] Furthermore, in the process of generating the event impact coefficient, the event analysis module collects the user's activity type and corresponding timestamp data in the control area in real time through the preset event type weight library, extracts features and classifies events for the activity types, and simultaneously extracts the duration, frequency and time period distribution parameters of each event, presets weighted weights, and weightedly calculates the impact degree of a single event on the lighting demand, and maps the impact degree value to a standardized event impact coefficient through normalization processing.

[0037] Furthermore, the strategy generation module dynamically allocates the weight ratio of the limb influence coefficient and the event influence coefficient according to the user's historical behavior data, generates a comprehensive influence value through weighted calculation, and compares the comprehensive influence value with the threshold interval stored in the preset strategy library. When the comprehensive influence value falls into a specific threshold interval, the lamp mode adjustment scheme bound to the interval is automatically triggered. The lamp mode adjustment scheme includes a color temperature adjustment gradient, a light intensity change curve and a dynamic fill light strategy, and generates an adjustment instruction set including an execution time window and an execution priority parameter and submits it to the adjustment implementation module.

[0038] Furthermore, the identity recognition module is interactively connected with the device management module and the attention recognition module through a wireless network, the behavior prediction module is interactively connected with the attention recognition module, the area analysis module and the control generation module through a wireless network, and the strategy generation module is interactively connected with the limb feedback module, the event analysis module and the adjustment implementation module through a wireless network.

[0039] Compared with the known prior art, the technical solution provided by the present invention has the following beneficial effects:

[0040] 1. By capturing the user's facial image, analyzing the direction of attention, and combining historical data to predict the behavior trajectory, the lighting can be adjusted in advance to adapt to the user's current needs, reduce manual operations, and improve convenience. At the same time, through the historical movement path and time series prediction algorithm, the user's movement path and residence time can be predicted, and the lighting configuration can be actively adjusted instead of passively reacting. Lamps can be turned on or off in advance to save energy. Prioritization can be performed according to the attention time and frequency of different areas to ensure that users can get the best lighting effect where light is most needed.

[0041] 2. By collecting limb movements and event records, the system can analyze the impact of different events on lighting requirements, comprehensively calculate the influence coefficients of different factors, so as to achieve more accurate lighting adjustment, integrate the limb influence and event influence coefficients, and form a lighting mode adjustment plan based on comprehensive analysis to ensure that the lighting adjustment plan can adapt to complex and changing usage scenarios.

[0042] 3. By having the ability to adapt to multi-user environments, it can support the personalized needs of multiple people simultaneously and can be well adapted whether in a home, office or public space. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0044] Figure 1 It is a schematic framework diagram of the present invention.

[0045] The reference numerals in the figure respectively represent: 1. Identity recognition module; 2. Device management module; 3. Attention recognition module; 4. Behavior prediction module; 5. Area analysis module; 6. Control generation module; 7. Limb feedback module; 8. Event analysis module; 9. Strategy generation module; 10. Adjustment implementation module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In order to make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0047] The following further describes the present invention with reference to the embodiments.

[0048] Embodiment 1

[0049] The personalized lighting adjustment system based on artificial intelligence in this embodiment, as Figure 1 shown, includes:

[0050] An identity recognition module 1, used to provide user identity registration, assign user IDs, and allow login after identity verification;

[0051] The device management module 2 is used to register and manage the positions, start / stop states, and lighting modes of all lamps in the controlled target area, assign control permissions to each lamp, and set the control permissions of different users for lamp start / stop and lighting modes. The user permission grading mechanism is as follows: Administrator permission: Can globally modify lamp parameters; assign user permissions and set system rules; register user permissions; independently adjust the preset lighting mode within the authorized area; Visitor permission: Only trigger temporary area lighting; Ensure that different users implement differential control over parameters such as the start / stop state, lighting intensity, and color temperature mode of corresponding lamps within the specified time and space range through dynamic token verification. At the same time, set the priority arbitration rule when there is a permission conflict: Administrator instruction > Event-driven strategy > User habit preference;

[0052] The attention recognition module 3 is used to capture the user's facial image through the camera component, judge the direction of the user's concentrated attention, and generate the user's current state data. During the stage when the attention recognition module 3 judges the direction of the user's concentrated attention, by recognizing the eye gaze features and head deflection angle in the user's facial image, obtain the target area currently being focused on by the user, perform temporal alignment on the eye gaze features and head deflection angle data, perform weighted calculation according to the preset weight, output the attention index, set the attention state determination threshold, and when the obtained attention index exceeds the preset threshold, trigger the response signal of the target area, match the focus point of the user's gaze with the lamps installed within the response signal target area, and generate the user's current state data.

[0053] The behavior prediction module 4 is used to collect the historical movement path and residence time data of the user in the control area, predict the user's future behavior through time series prediction algorithms, and generate the user's behavior trajectory map;

[0054] The area analysis module 5 is used to analyze the attention time and frequency of the user in each area according to the behavior trajectory predicted by the behavior prediction module 4, and perform priority sorting of the attention areas;

[0055] The control generation module 6 is used to obtain the future residence order and residence time of each area according to the user's priority attention area and predicted behavior trajectory, and correspondingly generate the lamp start / stop timing and duration adjustment plan within the future cycle.

[0056] The operation logic of the control generation module 6 is as follows:

[0057] Extract the time series sequence and corresponding residence time periods of each predicted residence area within the future cycle based on the user behavior trajectory map, and combine the priority sorting results of the priority attention areas to determine the pre-activation order and trigger time threshold of the lamps in each area;

[0058] Map the residence time period to the target duration interval of the lamps in the corresponding area, and dynamically adjust the lighting early start coefficient and delay attenuation coefficient according to the area priority;

[0059] According to the mapping relationship between the area luminaire identifier and the physical location, the timing control parameters are converted into a set of luminaire start / stop instructions distributed in the spatial dimension, forming a structured adjustment plan that includes luminaire identifiers, activation timestamps, duration, and shutdown conditions. Among them, the lighting duration of the high-priority area is increased by a preset buffer duration on the basis of the predicted value and is preferentially arranged at the front end of the control sequence.

[0060] The limb feedback module 7 is used to collect the limb movement data of the user in the control area in the current cycle, extract the user's behavior habit characteristics, establish a limb model, input the behavior habit characteristics, and output the limb influence coefficient of the evaluated behavior habit characteristics on the corresponding lighting requirements. The operation logic of the limb model in the limb feedback module 7 includes:

[0061] Real-time collect the joint movement trajectory, limb displacement speed, posture holding duration, and action frequency parameters of the user in the control area, and extract the limb movement direction preference coefficient, periodic swing amplitude, and the proportion of static posture duration characteristics;

[0062] Based on the historical behavior data set, a feature matching library is trained and generated. The feature matching library contains the association mapping relationship between different behavior patterns and light sensitivity. Based on the input index of the currently extracted features, the behavior prediction of the associated features and the corresponding light demand tendency value are obtained. The light demand tendency value is quantified into a limb influence coefficient, and strong, medium, and weak lighting influence thresholds are set in stages. The lighting influence threshold corresponds to the matching luminaire lighting mode setting.

[0063] The event analysis module 8 is used to record various events that the user performs in the control area. Based on the event type and duration, analyze the influence degree of different events on the lighting requirements, and generate an event influence coefficient. In the process of generating the event influence coefficient by the event analysis module 8, through a preset event type weight library, the activity type and corresponding timestamp data of the user in the control area are collected in real time, the activity type is subjected to feature extraction and event classification, and the duration, occurrence frequency, and time segment distribution parameters of each event are synchronously extracted. A preset weighting weight is used to calculate the influence degree value of a single event on the lighting requirements. Among them, the event type weight library contains the basic weight factor and time sensitivity coefficient corresponding to different activity types. The influence degree value is mapped into a standardized event influence coefficient through normalization processing, and a time series association model is established with the historical event influence data to dynamically correct the coefficient parameters.

[0064] The policy generation module 9 is used to integrate and process the limb influence coefficient and the event influence coefficient, and match the integration result with the corresponding lamp mode adjustment scheme in the preset policy library; the policy generation module 9 dynamically allocates the weight ratio of the limb influence coefficient and the event influence coefficient according to the user's historical behavior data, generates a comprehensive influence value through weighted calculation, compares the comprehensive influence value with the threshold interval stored in the preset policy library, and when the comprehensive influence value falls into a specific threshold interval, automatically triggers the lamp mode adjustment scheme bound to this interval. The lamp mode adjustment scheme includes the color temperature adjustment gradient, the light intensity change curve and the dynamic supplementary light strategy, and generates an adjustment instruction set including the execution time window and the execution priority parameter and submits it to the adjustment implementation module 10.

[0065] The adjustment implementation module 10 is used to match the lamp start / stop adjustment scheme and the lamp mode adjustment scheme of the control generation module 6, and configure the corresponding lamps in the specified period.

[0066] The identity recognition module 1 is connected to the device management module 2 and the attention recognition module 3 through wireless network interaction. The behavior prediction module 4 is connected to the attention recognition module 3, the area analysis module 5 and the control generation module 6 through wireless network interaction. The policy generation module 9 is connected to the limb feedback module 7, the event analysis module 8 and the adjustment implementation module 10 through wireless network interaction.

[0067] Compared with the prior art, it realizes efficient user identity management, behavior prediction and lighting adjustment. The system provides hierarchical user permission management to ensure that different users can dynamically adjust the lamp status and mode according to their permissions. By capturing the user's facial images and limb movements through the camera, it accurately analyzes the user's attention direction and behavior habits, thereby generating user status data and behavior trajectory maps. Considering the user's current status and historical behavior data comprehensively, the system intelligently predicts the user's future staying area and time, and formulates the lamp start / stop and mode adjustment schemes accordingly.

[0068] By recording the events of the user in the control area and combining the impact of the events on the lighting requirements, a customized lighting setting is formed. The advantage is that it can provide a more refined and personalized lighting experience, intelligently respond to the user's real-time needs, improve the flexibility and comfort of lighting control, and greatly enhance the user's interactivity and satisfaction compared with the traditional fixed lamp control mode.

[0069] Embodiment 2

[0070] On other levels, this embodiment also provides another optimization mechanism based on Embodiment 1, specifically the working logic of a time series prediction algorithm, including:

[0071] Continuously collect the historical movement path data set and residence time data set of the user within the control area. The historical movement path data set contains the spatial coordinate sequences of the user at different timestamps, and the residence time data set contains the start time and duration of the user's residence at specific coordinate points;

[0072] Perform spatio-temporal correlation processing on the historical movement path data set and the residence time data set, construct a three-dimensional spatio-temporal sequence data set, and establish a mapping relationship between the user's current position coordinates and the time dimension;

[0073] Adopt a sliding window mechanism to segment the three-dimensional spatio-temporal sequence data set to generate multiple training subsequences of equal length time periods. Each training subsequence contains the user's movement path features and residence pattern features within a preset time length;

[0074] Input the training subsequence into a pre-constructed recurrent neural network model for time series pattern learning. The model extracts time-dependent relationships through a gating mechanism and establishes a user position transition probability matrix and a residence duration distribution function;

[0075] Based on the user state data at the current moment, call the user position transition probability matrix and the residence duration distribution function for multi-step iterative prediction to generate a predicted path sequence within a preset future period. The predicted path sequence contains the expected position coordinate set corresponding to the predicted time node set and the associated confidence parameter;

[0076] Perform trajectory smoothing processing on the predicted path sequence, and use the cubic spline interpolation algorithm to generate a continuous user behavior trajectory map. The user behavior trajectory map contains spatio-temporal heat distribution characteristics and visual expression of the path confidence interval.

[0077] As a preferred implementation manner in this embodiment, the construction process of the recurrent neural network model includes: defining the network structure, including an input layer, a hidden layer, and an output layer. The input layer receives the user's historical movement path and residence time characteristics, performs non-linear transformation through multiple hidden layers, and introduces non-linear factors through an activation function;

[0078] Set a loss function to quantify the difference between the predicted output and the actual result, and use the backpropagation algorithm for parameter optimization. Update the network weights through the gradient descent method to minimize the loss function;

[0079] Through batch input of historical sequence data for multiple rounds of iteration, adjust the network parameters until the model converges to obtain a recurrent neural network model for capturing the characteristics of time series data.

[0080] Compared with the prior art, through the temporal learning and multi-step prediction of the recurrent neural network model, it is possible to accurately predict the future location and residence time of users, improve the accuracy of prediction, conduct spatio-temporal association of the historical paths and residence times of users, combine user behavior characteristics, effectively capture the temporal and spatial dependence relationships of user behaviors, smooth the predicted path through cubic spline interpolation, ensure the continuity and naturalness of the trajectory, enhance the visualization performance of the user behavior trajectory map, and avoid incoherent or unrealistic predictions.

[0081] The generated predicted path sequence not only has high-confidence parameters, but also can be visually displayed through heat maps and path confidence intervals, making the prediction results more intuitive and easy to understand. The model can be dynamically updated and optimized. As more historical data accumulates, the model's prediction ability continues to enhance, so as to more accurately predict users' future behaviors and provide personalized lighting adjustment schemes.

[0082] Embodiment 3

[0083] This embodiment provides a method for weighted integration of duration and event frequency, and its calculation formula is: ; In the formula, represents the comprehensive attention score of the area , represents the attention time weight factor, represents the total observation time period, represents the indicator function, indicating whether the user is in the area at time point . If the user is in the area at time point , then has a value of 1. If the user is not in the area at time point , then has a value of 0. represents the degree of the user's attention contribution within the time interval , represents the weight factor of the attention frequency, represents the number of times the user enters the area during the observation period, represents the time difference between the th entry event of the area and the current moment, represents the time difference between the th entry event of the area and the current moment, represents the time decay coefficient, Represents the number of times the user enters the area during the observation period , represents the total number of areas.

[0084] The formula in this embodiment can not only reflect the staying duration of the user in the area, but also reflect the frequent attention of the user to a certain area, thus providing a basis for the intelligent adjustment of the lamp start / stop and lighting mode.

[0085] Working principle: After being carried, the system in the present invention allows the user to register and verify the identity through the identity recognition module 1, so as to obtain the user ID. The device management module 2 conducts system registration and manages the lamp positions and their control permissions. The attention recognition module 3 captures the user's facial image through the camera component, judges the direction of concentrated attention, generates user status data. The behavior prediction module 4 records the user's historical movement path and staying time, uses the time series prediction algorithm to predict the future behavior, and generates a behavior trajectory map. The area analysis module 5 analyzes the time and frequency of the user's attention to each area, conducts priority sorting, and the control generation module 6 formulates the start / stop timing and duration adjustment plan of the lamp.

[0086] In addition, the limb feedback module 7 collects the user's limb movement data, evaluates its influence on the lighting demand, the event analysis module 8 records the user's events, analyzes the degree of influence on the lighting demand, the strategy generation module 9 integrates the limb influence coefficient and the event influence coefficient, and the adjustment implementation module 10 configures the final lamp adjustment plan to the corresponding lamp. Through this process, the system can intelligently adjust the lighting according to the user's real-time needs and realize a personalized lighting experience.

[0087] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. The personalized lighting adjustment system based on artificial intelligence is characterized by: include: The identity recognition module is used to provide user identity registration, assign user IDs, and allow login after identity verification; The equipment management module is used to register and manage the position, start / stop status and lighting mode of all lamps in the target area, and set the control authority of different users over the start / stop and lighting mode of lamps; The attention recognition module is used to capture the user's facial image through the camera component, determine the direction of the user's attention, and generate the user's current state data; The behavior prediction module is used to collect the historical movement path and residence time data of users in the control area, predict the user's future behavior through the time series prediction algorithm, and generate the user's behavior trajectory map; The regional analysis module is used to analyze the user's attention time and frequency in each area according to the behavior trajectory predicted by the behavior prediction module, and prioritize the attention areas; The control generation module is used to obtain the future stay order and stay time of each area according to the user's priority areas of attention and predicted behavior trajectory, and to generate the start and stop timing and duration adjustment plan of the lamps in the future cycle accordingly; The limb feedback module is used to collect the limb movement data of the user in the control area in the current cycle, extract the user's behavioral habit characteristics, establish a limb model, input the behavioral habit characteristics, and output the limb influence coefficient; The event analysis module is used to record the type and duration of events performed by users in the control area, analyze the impact of different events on lighting needs, and generate event impact coefficients; A strategy generation module is used to integrate and process the limb influence coefficient and the event influence coefficient, and match the integration result with the corresponding lighting mode adjustment scheme in the preset strategy library; The adjustment implementation module is used to match the lamp start and stop adjustment plan with the lamp mode adjustment plan, and configure the corresponding lamps in the specified cycle.

2. The personalized lighting adjustment system based on artificial intelligence according to claim 1 is characterized in that: In the stage where the attention recognition module determines the direction in which the user's attention is focused, the target area that the user is currently focusing on is obtained by identifying the eye gaze features and the head deflection angle in the user's facial image, the eye gaze features and the head deflection angle data are time-series aligned, weighted calculation is performed according to preset weights, an attention index is output, and an attention state determination threshold is set. When the obtained attention index exceeds the preset threshold, a response signal of the target area is triggered, the focus point of the user's gaze is matched with the placed lamps in the target area of ​​the response signal, and the user's current state data is generated.

3. The personalized lighting adjustment system based on artificial intelligence according to claim 1 is characterized in that: The working logic of the time series prediction algorithm run by the behavior prediction module includes: Continuously collect the historical movement path data set and residence time data set of the user in the control area. The historical movement path data set contains the spatial coordinate sequence of the user at different timestamps, and the residence time data set contains the starting time and duration of the user's stay at a specific coordinate point; Perform spatiotemporal correlation processing on the historical movement path data set and the residence time data set to construct a three-dimensional spatiotemporal sequence data set; The sliding window mechanism is used to segment the three-dimensional spatiotemporal sequence data set to generate multiple training subsequences of equal length time periods; Input the training subsequences into a pre-built recurrent neural network model for temporal pattern learning, wherein the model extracts time dependencies through a gating mechanism and establishes a user location transition probability matrix and a dwell time distribution function; Based on the user status data at the current moment, the user location transfer probability matrix and the residence time distribution function are called to perform multi-step iterative prediction to generate a predicted path sequence within a preset period in the future; The predicted path sequence is smoothed and a cubic spline interpolation algorithm is used to generate a continuous user behavior trajectory map.

4. The personalized lighting adjustment system based on artificial intelligence according to claim 3 is characterized in that: The construction process of the recurrent neural network model includes: defining a network structure, including an input layer, a hidden layer and an output layer, wherein the input layer receives the user's historical movement path and residence time characteristics, performs nonlinear transformation through multiple hidden layers, and introduces nonlinear factors through an activation function; A loss function is set to quantify the difference between the predicted output and the actual result, and a back-propagation algorithm is used for parameter optimization to update the network weights by gradient descent to minimize the loss function; By batch-inputting historical sequence data for multiple rounds of iterations and adjusting network parameters until the model converges, a recurrent neural network model for capturing the characteristics of time series data is obtained.

5. The personalized lighting adjustment system based on artificial intelligence according to claim 1 is characterized in that: The regional analysis module comprehensively measures the user's attention time and entry frequency in each area by weighted integration of duration and event frequency information. The calculation formula is: ； In the formula, Representative area The comprehensive attention score of represents the attention time weight factor, represents the total time period of observation, Represents the indicator function, indicating that at the time point When the user is in the area If the user is at time Located in the area Inside, then The value is 1, if the user is at time Not in area Inside, then The value of is 0, Represents time interval The degree of attention contribution of internal users, represents the weight factor of attention frequency, Indicates that the user enters the area during the observation period The number of times Representative area No. The time difference between the first entry event and the current time, Representative area No. The time difference between the first entry event and the current time, represents the time decay coefficient, Indicates that the user enters the area during the observation period The number of times Represents the total number of regions.

6. The personalized lighting adjustment system based on artificial intelligence according to claim 1, characterized in that: The operation logic of the control generation module is: Based on the user behavior trajectory map, the time series of each predicted stay area and the corresponding stay time period in the future cycle are extracted, and the pre-activation order and trigger time threshold of the lamps in each area are determined in combination with the sorting results of the priority areas. Map the stay time period to the target duration interval of the corresponding area lamps, and dynamically adjust the lighting advance start coefficient and delay attenuation coefficient according to the area priority; According to the mapping relationship between the regional lamp identification and the physical location, the timing control parameters are converted into a lamp start and stop instruction set distributed according to the spatial dimension, forming a structured adjustment plan that includes the lamp identifier, activation timestamp, duration and shutdown conditions.

7. The personalized lighting adjustment system based on artificial intelligence according to claim 1, characterized in that: The operation logic of the limb model in the limb feedback module includes: Collect the user's joint motion trajectory, limb displacement speed, posture holding time and action frequency parameters in real time within the control area, and extract the limb movement direction preference coefficient, periodic swing amplitude and static posture duration ratio characteristics; A feature matching library is generated based on historical behavior data set training. The feature matching library contains the correlation mapping relationship between different behavior patterns and light sensitivity. Based on the current extracted feature input index, the behavior prediction of the associated feature and the corresponding light demand tendency value are obtained. The light demand tendency value is quantified as a limb influence coefficient, and strong, medium and weak lighting influence thresholds are set in stages. The lighting influence thresholds correspond to the lighting mode settings of the matching lamps.

8. The personalized lighting adjustment system based on artificial intelligence according to claim 1, characterized in that: In the process of generating the event impact coefficient, the event analysis module collects the activity types and corresponding timestamp data of users in the control area in real time through the preset event type weight library, extracts features and classifies events for the activity types, and simultaneously extracts the duration, frequency and time period distribution parameters of each event. The weighted weights are preset, and the impact degree of a single event on the lighting demand is weightedly calculated. The impact degree value is mapped to a standardized event impact coefficient through normalization processing.

9. The personalized lighting adjustment system based on artificial intelligence according to claim 1, characterized in that: The strategy generation module dynamically allocates the weight ratio of the body influence coefficient and the event influence coefficient according to the user's historical behavior data, generates a comprehensive influence value through weighted calculation, and compares the comprehensive influence value with the threshold interval stored in the preset strategy library. When the comprehensive influence value falls into a specific threshold interval, the lamp mode adjustment scheme bound to the interval is automatically triggered. The lamp mode adjustment scheme includes a color temperature adjustment gradient, a light intensity change curve and a dynamic fill light strategy, and generates an adjustment instruction set including an execution time window and an execution priority parameter and submits it to the adjustment implementation module.

10. The personalized lighting adjustment system based on artificial intelligence according to claim 1, characterized in that: The identity recognition module is interactively connected with the device management module and the attention recognition module through a wireless network, the behavior prediction module is interactively connected with the attention recognition module, the area analysis module and the control generation module through a wireless network, and the strategy generation module is interactively connected with the limb feedback module, the event analysis module and the adjustment implementation module through a wireless network.

Images