Energy-saving LED lamp intelligent control method

Abstract

The invention relates to an intelligent control method for an energy-saving LED lamp, and the method comprises the steps: building a simplified topology through the fusion of a space structure, a personnel position track and shielding object information, planning a candidate path leading to a region meeting an illumination threshold value, calculating a composite utility value through the combination of the path length and the illumination profit, and generating a lighting condition when a path utility threshold value is reached. And in combination with the sight direction of the personnel and the area function, determining a lamp group candidate area which can cover a vision field angle and meets the target illumination, calculating a brightness coverage index, and selecting a lamp group set which does not exceed a coverage threshold value for activation. After activation, according to the stay duration and behavior characteristics of the user, multi-stage illumination decline is implemented according to a non-fixed step length and a time interval; and if no user intervention or discomfort feedback exists in the decreasing process, continuously decreasing the energy. Meanwhile, a light supplementing area in linkage relation with the main lighting area is recognized in lighting control, linkage information is generated, a delay activation or power suppression strategy is executed on the light supplementing area, and collaborative optimization of lighting comfort and energy saving performance is achieved.

Description

Technical Field

[0001] This invention relates to an intelligent control method for energy-saving lamps, specifically an intelligent control method for energy-saving LED lamps. Background Technology

[0002] Existing technologies, such as Chinese patent CN109068440A ("An Intelligent Control Method and System for Energy-Saving Lighting"), still have significant systemic shortcomings in current practical engineering applications. Most existing tunnel LED energy-saving control schemes follow the technical approach outlined in that document. Their core control logic is primarily based on a single-point, instantaneous, passive triggering mode: a light sensor detects ambient brightness, and an infrared or microwave sensing module detects the presence of vehicles or personnel. That is, when the brightness of a tunnel section is detected to be below a threshold or a target is detected entering, the corresponding lamp or light group is directly switched on or dimmed. This mode has inherent flaws in the long-distance, highly linear, and continuous spatial environment of a tunnel. This method generally ignores the topological continuity and depth structure of the tunnel space itself, failing to treat the tunnel as a unified lighting system with entrance, transition, basic, and exit sections. Instead, it fragments it into multiple independent control units. The lack of overall coordination between light segments based on spatial paths and travel directions directly leads to frequent brightness abrupt changes or light-dark discontinuities during high-speed driving, easily causing repeated switching between light and dark adaptation, increasing the driver's visual burden and driving risks.

[0003] This document essentially remains at the level of binary or weakly continuous control, focusing on whether a target exists and whether lighting is necessary. Even when dimming is introduced, it's mostly done with fixed step sizes or preset time curves, failing to dynamically adjust based on the vehicle's actual speed, acceleration changes, traffic density differences, and the driver's visual focus area within the tunnel. This is particularly problematic in high-speed tunnel scenarios, easily leading to light changes lagging behind vehicle position, or overall brightness increases for safety, ultimately sacrificing energy efficiency. This method generally lacks the ability to model effective illumination within the visual field. Control targets are often based on the average illuminance of the tunnel cross-section or road surface illuminance, without considering the driver's primary gaze direction, gaze distance, and visual task differences at different stages of driving. The result is that many lights are turned on but not truly converted into effective visual information, creating a hidden waste of energy despite illuminance standards being met. Furthermore, this scheme lacks evaluation indicators for brightness coverage or illuminance utilization, failing to identify which lights have redundant coverage and which areas are true perception blind spots. It can only mitigate risks by conservatively increasing overall brightness, further amplifying energy consumption.

[0004] The document fails to establish a closed-loop learning mechanism for vehicle behavior or driving feedback. Whether a vehicle slows down in the tunnel, brakes frequently, or experiences abnormal stops or visual discomfort in a particular lighting section is not considered as feedback signals to correct the lighting strategy. The system lacks self-correction capabilities; if initial parameters are set incorrectly, long-term operation requires manual recalibration, making it difficult to adapt to changes in different time periods, traffic conditions, and driver demographics. Furthermore, regarding the depth of energy-saving strategies, this method focuses more on whether lights are on or off, or on overall dimming, without analyzing the spatial overlap of light fields between lamps. It also fails to consider the illuminance superposition between adjacent lamp segments, lamp sharing, and redundant supplementary lighting. Especially in tunnel curves, ramps, or auxiliary passageways, multiple lamp segments often operate at high power simultaneously with highly overlapping illuminance contributions, resulting in long-term hidden waste. Traditional methods lack the ability to quantitatively assess overlapping illuminance, let alone implement power suppression or delayed linkage based on overlap rates. In engineering implementation, this approach often uses fixed rules or empirical thresholds, ignoring the essential characteristic of tunnels as continuous travel paths. It fails to treat the vehicle's travel path as a dynamically evolving path of lighting demand, and it does not construct a utility model between path length, lighting benefits, and energy consumption. As a result, the system cannot predict upcoming lighting demands in advance and can only respond passively, which affects the driving experience and limits the potential for further energy savings. Summary of the Invention

[0005] The purpose of this invention is to provide an energy-saving intelligent control method for LED lights, thereby solving some of the drawbacks and shortcomings pointed out in the background art.

[0006] The present invention addresses the aforementioned technical problems by employing the following technical solution: an energy-saving intelligent control method for LED lights, comprising: acquiring spatial structure data, personnel location and activity trajectory, and combining them with obstruction information to form a simplified spatial topology; determining several candidate paths for users to reach a lighting area that meets a target illuminance threshold based on the topology; calculating the composite utility value of each candidate path obtained by combining path length and illumination benefit according to a preset rule; and generating lighting conditions when the composite utility value meets a preset path utility threshold.

[0007] When the lighting conditions are met, based on the direction of the person's line of sight and the functional attributes of the area, one or more candidate areas of lamp groups are determined that cover the user's field of vision angle range and can make the illuminance in the field of vision reach the target illuminance threshold. The brightness coverage index of each candidate area is calculated, and a set of lamp groups with a coverage index not higher than the coverage threshold is selected for activation control.

[0008] After the light group is activated, based on the user's dwell time and behavioral characteristics, a multi-level illuminance reduction with a non-fixed step size is executed according to a preset step size range and time interval. If no user intervention or discomfort feedback event is detected during the reduction process, the illuminance reduction continues to reduce energy consumption. During lighting control, supplementary lighting areas that have a linkage response relationship with the current main lighting area are identified, linkage relationship information is generated, and a delayed activation or power suppression strategy is applied to the supplementary lighting areas.

[0009] Furthermore, the illumination benefit is the integral result of the estimated illuminance value achievable within a unit path length; the path length comprehensively considers the actual movement distance of the user's restricted path, excluding areas blocked by obstructions; the composite utility value is calculated by weighting the path length and illumination benefit by assigning preset weighting coefficients to each.

[0010] Furthermore, when the composite utility values ​​of multiple candidate paths all meet the path utility threshold, the path lengths of these candidate paths are sorted in ascending order, and the candidate path with the highest ranking is selected to generate the lighting conditions; if there are candidate paths with the same path length, the candidate path with fewer path nodes is selected first; if they are still the same, the selection is made according to a preset priority table; the gaze direction is determined by combining the user's historical facing angle distribution statistics with the current head orientation information; the brightness coverage index of the lighting area is the ratio of the actual lighting area of ​​the lamp group set within the user's field of vision to the total field of vision area.

[0011] Furthermore, the brightness coverage index is calculated by taking into account the current ambient natural light intensity and performing illumination compensation calculations; wherein the activation control includes setting an upper limit value for the brightness of the selected light group to control the initial power consumption when lighting up; wherein the step size range of the illuminance decrease is dynamically adjusted according to the user's behavioral characteristics, such as action frequency, standing stability, and movement trend.

[0012] Furthermore, the discomfort feedback events include the user manually adjusting the brightness, repeatedly staying in the original area for more than the expected time, or the system detecting that the user is staring at the lighting boundary area; wherein the supplementary lighting area is an area where the spatial illuminance overlap rate with the main lighting area exceeds a preset threshold; wherein the linkage relationship information includes the spatial adjacency, functional similarity, and lamp group sharing degree between the main lighting area and the supplementary lighting area.

[0013] Furthermore, the user's manual brightness adjustment behavior includes sliding the brightness bar through the smart terminal control interface, repeatedly operating the physical switch more than a preset threshold, or having a voice command containing brightness enhancement semantics; wherein the system predicts the user's intention based on the change in the user's dwell time curve in the same area, and performs feedback preprocessing actions in advance before the dwell time approaches the threshold.

[0014] Furthermore, the system detects the user's gaze at the lighting boundary area by: identifying the user's eye focus point through a camera and performing spatial cross-analysis with the current lighting boundary; wherein, after detecting the user's gaze at the lighting boundary area, the system performs brightness fine-tuning or brightening expansion operations on the boundary area and records it as a local perception blind spot.

[0015] Furthermore, the calculation of the spatial illuminance overlap rate includes integrating the luminous flux distribution overlap between the main lighting area and the supplementary lighting area, and fitting and comparing it with the global spatial lighting demand surface; wherein after identifying the supplementary lighting area, the system first determines whether the area is in the user blind zone. If so, the area is set to a suppressed state and does not participate in the next round of activation scheduling.

[0016] The calculation of the spatial illuminance overlap rate includes constructing an integral and partial derivative fitting function based on the luminous flux overlap distribution between the main illumination area and the supplementary illumination area, in time slices. Inside, the main lighting area With fill light area Spatial illuminance distribution function Modeled as a continuous differentiable surface, the following overlapping fitting function is defined:

[0017]

[0018] in:

[0019] This indicates the dynamic illuminance overlap rate of a unit supplemental lighting area; Indicates time Time and Space Point Illuminance intensity function on; The two-dimensional spatial domain defining the main lighting area; Define the two-dimensional spatial domain of the supplementary lighting area; The area of ​​the supplementary lighting region is used for normalization; The second-order mixed partial derivative of illuminance intensity with respect to horizontal and vertical space represents the rate of change of the light spot gradient, reflecting the sharpness of the illuminance edge contour; the exponential term... Used to reduce the overlap weight in high gradient edge regions and emphasize the overlap contribution in continuous light field regions;

[0020] The The function derivation process includes:

[0021] Set any time The entire two-dimensional lighting space has a continuous illuminance distribution function. , indicating at point The instantaneous illuminance value at a location. The main lighting area is denoted as... The supplementary lighting area is denoted as The spatial overlap region is The intensity of illuminance overlap in this area is an important basis for assessing redundancy;

[0022] Therefore, starting from the effective illuminance coverage of the main lighting area to the supplementary lighting area, the following integral term is defined:

[0023]

[0024] This integral describes the total luminous flux projected by the main illumination onto the supplementary lighting area; however, using only luminous flux overlap can easily overestimate the contribution of the edge region. Therefore, a light field structure suppression function is introduced to enhance the weight of the central continuous illumination area and weaken the interference of the sharp edge transition area. The edge of the light spot usually exhibits abrupt changes in illuminance gradient, so a spatial second-order mixed partial derivative of the illuminance function is introduced. As a contour factor for illuminance variation, an exponential decay function was designed. Weighting is applied to suppress the impact of high-edge sharpness regions on the overlap rate.

[0025] Therefore, the illuminance-weighted overlap function is expressed as:

[0026]

[0027] The integral result reflects the main illumination's ability to project energy onto the supplementary lighting area in a spatially continuous, soft, and effective manner, i.e., the effective contribution of overlap. To facilitate comparisons between different area sizes, the integral result needs to be normalized to the area of ​​the supplementary lighting area, ultimately obtaining the average overlap weight per unit area, defined as the overlap rate. ...

[0028] Furthermore, the spatial adjacency is calculated based on the Euclidean distance between the geometric centers of the lamps, and areas below a preset adjacency threshold are judged to be adjacent areas; the functional similarity is determined by reading the area label information and matching it with the user-configured preset template, and assigning a functional similarity score to the area.

[0029] Furthermore, the degree of lamp sharing is calculated by analyzing the ratio of the number of shared lamps between the two areas to the total number of lamps, and is input into the lighting decision model as a linkage intensity coefficient; wherein the linkage relationship information is used to construct a regional response priority map, and prioritize the activation of supplementary lighting areas that have a high degree of sharing and high functional similarity with the main lighting area.

[0030] The beneficial effects of this invention are as follows: By constructing a path utility model based on spatial topology and occlusion information, deep coupling between user movement behavior and lighting response is achieved. This effectively reduces energy waste caused by redundant lighting and false triggering without relying on complex image recognition or large-scale sensor integration. By introducing a composite utility value judgment mechanism and a coverage threshold screening strategy, light group activation is triggered under high-intent, high-cost-performance path conditions. Furthermore, dynamic power compression is achieved through a multi-level decreasing dimming mechanism, significantly improving the unit energy efficiency of lighting behavior. In particular, by integrating multiple factors such as path length, occlusion structure, natural light compensation, user field of vision, and behavior dwell time into computable rules, the entire system possesses a high degree of structural awareness and spatial energy-saving scheduling capability. Compared to traditional LED lighting systems based on infrared sensing or time control, this invention can achieve more granular lighting behavior control while ensuring lighting quality.

[0031] To address individualized user response characteristics, an discomfort feedback event mechanism and a lighting boundary gaze detection strategy are introduced. The system can automatically identify whether the current lighting output matches the user's perceived needs based on multi-dimensional data, including manual adjustments, behavioral lingering trends, and visual focus points. It dynamically adjusts the illuminance curve based on feedback, significantly enhancing lighting comfort and subjective experience. Furthermore, by combining spatial illuminance overlap rate integration, light spot structure edge gradient calculation, and regional linkage response priority judgment, the system can automatically suppress brightness or delay activation in high-redundancy, low-interest areas, effectively reducing supplementary lighting energy consumption and extending luminaire lifespan. Through modeling the functional similarity and sharing degree between supplementary lighting areas and main lighting areas, cross-regional lighting scheduling strategies are optimized and updated, fully demonstrating the technological originality and practical application value in spatial cognitive control, user response prediction, and intelligent energy-saving strategy integration. Attached Figure Description

[0032] Figure 1 This is the main process of intelligent control for energy-saving LED lights according to the present invention.

[0033] Figure 2 This is a functional relationship diagram of the intelligent lighting system with composite utility value according to the present invention.

[0034] Figure 3 This is a flowchart illustrating the priority decision-making process for the intelligent linkage between the main lighting and supplementary lighting areas in this invention.

[0035] Figure 4 This is a schematic diagram of the intelligent lighting control process in the lobby of a smart office building, as described in Embodiment 1 of the present invention.

[0036] Figure 5 This is a diagram showing the relationship between the gaze at the boundary of the meeting area and the activation of intelligent supplemental lighting in Embodiment 2 of the present invention. Detailed Implementation

[0037] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0038] Combined with appendix Figure 1 This invention discloses an energy-saving intelligent control method for LED lights. It acquires spatial structure data, personnel location data, and personnel activity trajectory data. The spatial structure data includes the location coordinates, boundary contours, channel connectivity, and obstruction information of each functional area. Personnel location is determined through a sensor network or visual recognition device. The activity trajectory is obtained through the time-series changes of continuous location points. The system then fuses the above data to construct a simplified spatial topology representing the relationship between spatial layout and personnel behavior. This spatial topology is an abstract graph structure, where nodes represent the current location and target area of ​​the personnel, edges represent reachable paths, and edge weights are adjusted based on the degree of obstruction, passage width, and passage restrictions. Based on this simplified spatial topology, the system identifies the target lighting area for the current personnel state—that is, the area the user is about to reach or the area their behavior intends to point to—and determines multiple candidate paths leading to this target lighting area. These paths are all feasible, reachable, and unobstructed spatial channels from the current personnel location to the target area. The system then calculates a composite utility value for each candidate path. This value combines two factors: path length and illumination benefit. Path length is obtained by weighted summation of the geometric distances along the path's edges. Illumination benefit represents the expected increase in illuminance along the path based on the existing ambient light. This illuminance increase is estimated by integrating the illuminance gap at each point along the path under the current lighting configuration. Path length and illumination benefit are multiplied by preset weighting coefficients to reflect the balance between energy-saving priorities and user lighting experience. Finally, the weighted results are combined to form the composite utility value for that path. The system compares the composite utility values ​​of all candidate paths with a preset path utility threshold. If at least one path has a composite utility value higher than the threshold, the system generates a lighting condition, which serves as the input signal for the subsequent lighting response control module's decision-making. This triggers the lighting action when user behavior is significant and the expected lighting return is high.

[0039] Once the system detects that the lighting conditions are met, indicating that the current user's behavior path has satisfied the spatial path utility threshold, the system will enter the lighting activation phase. To avoid energy waste caused by ineffective or excessive lighting, the system will perform refined lighting range determination by combining the user's current gaze direction with the functional attributes of the area. It will acquire the user's head orientation data, which can be obtained by wearable sensors, ceiling-mounted visual acquisition devices, or other spatial behavior recognition modules. Through posture calculation or eye-tracking algorithms, the system will determine the user's current primary gaze direction. The system will further construct a spatial visual field model of the user based on the gaze direction. This model uses a cone or fan-shaped structure, projecting forward to form a perception area with a certain opening angle and depth at the user's location. Subsequently, the system will read the functional attribute information of the area where the user is located. This information is preset in the lighting control system, including office areas, passageways, rest areas, and display areas. Corresponding target illuminance thresholds will be set for different functional types of areas. These illuminance thresholds represent the visual illuminance of the functional area. The system first determines the minimum lighting level required under the given conditions. Then, it filters out all sets of luminaires that can provide lighting within the user's field of vision and divides these luminaires into several candidate luminaire groups. Each luminaire group is a set of luminaires that are pre-configured by the system or have physical connections in the deployment. For each candidate luminaire group, the system calculates its brightness coverage index, which is defined as the ratio between the lighting area provided by the luminaire group within the current field of vision and the total area of ​​the user's field of vision. The lighting area consists of the area that the luminaire group can illuminate and make the illuminance reach the target illuminance threshold under the current driving parameters. By modeling a light diffusion model and superimposing natural light distribution, the accurate lighting coverage area can be obtained. The system compares the above coverage index with the preset coverage threshold, eliminates luminaire groups with excessive coverage, and prioritizes the selection of luminaire groups with coverage indices not higher than the threshold for activation control. This activation control includes sending a start signal to the corresponding luminaire group and limiting its initial brightness upper limit to reduce instantaneous power surges.

[0040] After activating the target light group according to the aforementioned judgment mechanism, the illuminance is maintained at a constant value. An intelligent illuminance management strategy is executed based on the user's dwell time and behavioral characteristics to achieve deeper energy consumption optimization. The system records the user's dwell time within the current lighting area, which is continuously sampled by a personnel position sensor and dynamically analyzed in conjunction with the user's behavioral characteristics. These behavioral characteristics include multiple dimensions such as movement amplitude, movement frequency, standing stability, and posture persistence. A sensor fusion algorithm is used to determine the user's activity level or attention concentration within the current area. Subsequently, based on this information, the system invokes the corresponding illuminance reduction strategy. This strategy is executed based on a preset step size range and time interval, without using a fixed reduction rate. Instead, it dynamically adjusts the illuminance reduction amplitude in stages. A larger reduction step size is used when the user's movements are stable or their attention is concentrated, while the reduction speed is reduced when the activity is intense or the viewing angle changes frequently to maintain perception continuity. This non-fixed step size mechanism allows the system to gradually compress power output while ensuring lighting quality. During the illuminance reduction process, the system continuously detects whether the user engages in any intervention behavior or experiences discomfort. The intervention behavior includes... This includes user actions such as manually adjusting brightness, requesting increased illuminance via voice control, or frequently lingering at the lighting boundary. Discomfort feedback events include situations where the user stays longer than expected without reaching a stable state, or the system detects the user staring at an insufficiently lit area through visual recognition. If the system does not detect any intervention or feedback signal during the current reduction phase, it assumes the user is satisfied with the current brightness, and the system continues to execute the next level of illuminance reduction, thus forming a closed-loop adaptive illuminance compression process. During lighting control, supplementary lighting areas with a linkage response relationship to the current main lighting area will also be identified. Linkage information is generated by analyzing the spatial adjacency, functional similarity, and lamp sharing degree between lighting areas. This information is used to determine which supplementary lighting areas are mistakenly activated due to changes in the main lighting. Based on this linkage information, the system implements energy-saving optimization strategies for the supplementary lighting areas, including delayed activation and power suppression strategies. The delayed activation strategy sets a response delay period even after the supplementary lighting area meets the activation conditions, triggering lighting only when the activation logic is continuously met and the user's behavior is highly relevant. The power suppression strategy sets a brightness upper limit or a current start buffer for the lamps within the supplementary lighting area.

[0041] Combined with appendix Figure 2The system employs a composite utility value evaluation mechanism to measure the overall cost-effectiveness of each candidate path leading to the target lighting area, considering both energy saving and lighting efficiency. This mechanism defines illumination benefit as the integral result of the estimated illuminance value achievable per unit path length. It samples and accumulates the illuminance values ​​achievable at each point along the user's walkable path under current or predictable lighting conditions, ultimately forming a comprehensive evaluation of the illumination contribution to the entire path. Illuminance estimation considers ambient natural light intensity, existing lighting conditions, and the user's visual interaction, and uses an illuminance attenuation function for local compensation to ensure spatial continuity and physical rationality in the evaluation results. Path length is not merely calculated as the Euclidean distance between two points, but is measured based on the actual distance traveled along the user's restricted path. During path planning, the system excludes all obstructed areas, including walls, furniture, fixed structures, or temporary obstacles, ensuring that the path evaluation is based on physical accessibility. The aforementioned illumination benefits and path length are each assigned a preset weighting coefficient. This weighting coefficient is set or dynamically adjusted by the administrator before deployment based on lighting energy-saving priority, scene illumination requirements, and user comfort expectations. The two parameters are then combined linearly or non-linearly using a weighting function to generate a composite utility value for the candidate path. The higher the composite utility value, the better the lighting return per unit of energy consumption. The system uses this value to select the optimal path to trigger subsequent lighting decisions.

[0042] A multi-path composite utility value optimization mechanism is introduced when generating lighting conditions. Multiple candidate paths accessible to users are calculated, and their corresponding composite utility values ​​all reach or exceed a preset path utility threshold. This indicates that these paths are reasonable for triggering a lighting response under a weighted evaluation of illumination benefit and path length. To select the optimal path for generating lighting conditions, all candidate paths meeting the conditions are sorted in ascending order of path length, with the shortest path being prioritized as the final target path. This optimization logic is based on a combined consideration of path simplicity and energy efficiency for lighting requirements. If the sorting results show... If multiple candidate paths of the same length exist, the system further compares their number of nodes. The number of nodes represents the number of turning points or intermediate logical points in the path; fewer nodes indicate a more intuitive and concise path, and greater consistency in predicting user behavior. Therefore, the path with fewer nodes is prioritized as the final path. If multiple candidate paths still exist when both path length and number of nodes are the same, the system makes a decision based on a preset priority table. This priority table can predefine factors such as area security level, area functional importance, or historical usage frequency, assigning different logical levels to different paths, and ultimately determining the lighting path used to activate the lighting system. Regarding user orientation recognition, to further improve the accuracy of lighting target area recognition, the system integrates statistical data on the distribution of users' historical facing angles with currently acquired head orientation information to construct a multi-dimensional gaze direction judgment model. Historical facing data is obtained by statistically analyzing long-term user behavior records within the area to extract the probability density distribution of typical gaze directions. Current head orientation information is obtained through posture recognition or camera recognition. The combination of these two methods can accurately predict the area the user is about to focus on or move towards. For the evaluation of candidate lighting areas, the system introduces a brightness coverage index as a selection criterion. This index is defined as the ratio between the lighting area formed by the selected set of lamps within the current user's field of vision and the total area of ​​the entire user's field of vision. The lighting area is calculated by the optical simulation module based on the current brightness output, illumination angle, and occlusion conditions. The lower the ratio, the higher the energy concentration per unit illuminance and the stronger the effective coverage. The system will select the lamp combination with the highest energy efficiency ratio based on this index for subsequent activation control.

[0043] When evaluating the illuminance of candidate lighting groups, the system considers the lighting capacity of the luminaires themselves and the coverage of the user's field of vision. It also acquires the natural light intensity information of the current environment. The natural light intensity is collected in real time by an ambient light sensor and converted into a quantifiable illuminance level value. Based on this natural light value, the system performs illuminance compensation calculations for each area within the lighting coverage area. That is, the contribution of natural light is deducted from the target illuminance requirement, and only the luminaire lighting part makes up for the remaining illuminance difference. This allows the system to avoid redundant activation of luminaires when there is sufficient light and to perform accurate compensation when there is insufficient natural light, thereby optimizing the energy consumption structure and the user's visual experience. In terms of lighting activation control, to prevent power surges or excessively strong instantaneous brightness from causing user discomfort during the initial lighting phase, the system sets an upper limit value for the brightness of the selected light group when it is ready to be turned on. This upper limit value is set by the system in stages according to the area function type, time period, and user behavior background parameters, ensuring that the power consumption of the lights is within a safe range and the brightness change is smooth and controllable. After lighting, the system enters the illuminance adjustment phase, in which the illuminance reduction mechanism performs multi-level dynamic adjustment with energy saving as the goal. The reduction step size is not fixed, but dynamically adjusted based on the user's current behavior characteristics. The system comprehensively judges the user's action frequency, standing stability, and movement trend through multi-dimensional behavioral parameters. If the user exhibits low-frequency actions, high-stability standing, and no obvious movement trend, the system determines that the user is in a low-activity state and can use a larger step size to reduce illuminance quickly to reduce power consumption. Conversely, if the user moves frequently or is in an active state, the system uses a smaller step size to finely adjust the illuminance.

[0044] After implementing illuminance reduction or local lighting compression strategies, the system continuously monitors changes in user behavior to determine whether discomfort feedback events are triggered. These events include three types: First, manual brightness adjustment by the user. The system identifies user actions such as sliding the brightness bar on the smart terminal control interface, frequent switching of physical switches, and voice commands containing brightness enhancement semantics to identify user-initiated requests for increased brightness. Second, repeated lingering in the original area for more than the system's preset time threshold. The system determines that the user did not leave along the expected path based on the behavioral trajectory, inferring discomfort caused by insufficient illuminance or uneven lighting distribution, thus triggering a dimming response. Third, the system detects user behavior of staring at the lighting boundary area. The system obtains the user's gaze direction and eye movement characteristics through visual perception devices to determine whether the user's gaze point is within the boundary range of the current main lighting area. If the determination is correct, it indicates that the user is in a blurred illuminance zone, requiring boundary brightness compensation to improve visual continuity. To further enhance the linkage and spatial coherence of the lighting area response, the system identifies supplementary lighting areas outside the main lighting area that are related to it. These supplementary lighting areas are defined as areas where the spatial illuminance overlap rate with the current main lighting area exceeds a preset threshold. In other words, the luminous flux distribution in the two areas has a significant spatial intersection, and the illuminance distribution surface has a highly fitted interval, indicating that such areas are significantly affected by the activation of the main lighting and have a potential risk of insufficient illuminance. After identifying the supplementary lighting area, the system further generates linkage information. This information includes the spatial adjacency between the main lighting area and the supplementary lighting area, defined as an adjacency relationship when the physical distance between the geometric center points of the two areas is less than a set threshold. It also includes functional similarity, which is obtained by reading the area function labels and matching them with the user configuration template. Areas with high similarity indicate similar needs in terms of usage behavior. In addition, it includes the degree of lamp sharing, which is calculated by statistically analyzing the proportion of shared lamps in the two areas to their total number of lamps and calculating their sharing coefficient. The higher the coefficient, the more coupled the two areas are in terms of hardware resources. The above three types of indicators constitute the evaluation basis for the linkage strength between areas. Based on this, the system determines whether to apply a delayed activation strategy or a power suppression mechanism to the supplementary lighting area.

[0045] The user's manual brightness adjustment behavior is used as one of the important criteria for triggering the feedback mechanism. The user's manual brightness adjustment behavior includes the following three types: The first type is that the user actively increases the current lighting intensity by sliding the brightness adjustment bar through the control interface of the smart terminal. This behavior is recorded and analyzed in real time by the human-computer interaction system, and the adjustment range and frequency are analyzed. The second type is that the user performs a continuous trigger operation on the physical switch. When the number of such operations exceeds the system's preset action threshold in a short period of time, it is determined that the user wants the system to increase the brightness level of the current area. The third type is that the user issues a sentence with the meaning of enhancing illumination through voice commands, such as "increase the brightness" or "turn it up a little". The system extracts the intention of enhancing illumination in the speech through the semantic parsing module to complete the brightness adjustment behavior recognition. To enhance the system's ability to predict user lighting needs and avoid a decline in user experience due to response lag, the system introduces a behavioral trend modeling method based on user dwell time curves. After a user enters a certain lighting area, the system continuously records their dwell time and generates a dynamic time curve. When the curve shows a stable upward trend within a specific time period, the system calculates its slope. If the slope value reaches or exceeds a set threshold, it indicates that the user has a potential intention to be dissatisfied with the current lighting state. Based on this, the system predicts the user's behavioral tendency in advance and triggers feedback preprocessing actions before the dwell time exceeds the expected threshold of the area. These actions include measures such as slightly increasing local illuminance, activating edge supplementary lighting areas, or adjusting the softness of light color. This achieves an early response mechanism based on behavioral trend prediction, effectively avoiding the passive logic of users needing to explicitly trigger control in traditional lighting strategies.

[0046] Combined with appendix Figure 3The system is equipped with a gaze recognition and lighting boundary response module based on user eye movement behavior. This module continuously collects facial images of the user through a camera device installed in the lighting environment and performs eye tracking analysis. The system uses an embedded eye movement recognition algorithm to analyze the changes in the position of the user's pupils in real time and combines it with facial orientation data to calculate the spatial focal point of the current gaze, thereby determining the user's gaze area. After obtaining the gaze position, the system performs spatial cross-analysis with the illumination boundary information of the currently activated lighting area. The illumination boundary refers to the edge area of ​​the illumination distribution range formed by the current lighting group. This range is defined as an area with insufficient but perceptible external illumination based on a pre-determined spatial illuminance attenuation threshold. If the user's gaze point is within this boundary area, the system determines that the user is gazing at an area with insufficient illumination or blurred edges. To avoid visual discomfort and impact on users' spatial perception and task execution efficiency caused by such staring behavior, the system immediately performs brightness fine-tuning or brightening expansion operations on the boundary area after determining that a staring event has occurred. Brightness fine-tuning refers to slightly increasing the output power of the lamps in the boundary area to compensate for the lack of edge illumination without significantly increasing energy consumption. Brightening expansion refers to appropriately expanding the activation range of the lamps to cover the user's staring point and include it in the effective lighting area. The system marks the location corresponding to the staring event as a local perception blind spot and writes it into the historical lighting deviation record. Subsequently, in similar usage scenarios or the same area, this record can be used as a reference to perform lighting optimization presets, allowing the system to gradually build a user-individualized lighting adaptation model through multiple learning processes.

[0047] A illuminance overlap rate calculation mechanism is introduced. This mechanism considers the total illumination intensity of the overlapping luminous flux distribution area and incorporates the structural characteristics of spatial illumination variations to avoid misjudgment of high-gradient edge regions. A two-dimensional spatial mapping relationship between the main illumination area and the supplementary illumination area is established within time slice t, denoted as follows: and The system constructs a continuously differentiable illuminance intensity function in this space. , representing any point in space At any moment The system then defines the following overlap rate calculation function:

[0048]

[0049] in, This indicates the dynamic illuminance overlap rate of a unit supplemental lighting area; This represents the area of ​​the supplemental lighting region, used for normalizing the results; It is the second-order mixed partial derivative of illuminance intensity with respect to the horizontal and vertical directions, characterizing the rate of change of the light spot gradient and reflecting the edge sharpness of the illumination outline; the exponential function term This is used to suppress excessive overlap in the lighting edge area due to sharp gradient changes, thereby highlighting the effective projection ratio of continuous and high-quality light from the main lighting in the fill lighting area.

[0050] The spatial illuminance overlap rate calculation process sets an arbitrary time. In a two-dimensional lighting scene, there exists a continuous function. This reflects the illuminance distribution at each spatial point at that moment. The system identifies the spatial intersection of the main lighting area and the supplementary lighting area. The effective luminous flux coverage value is obtained by integrating the product of the luminance function and the edge suppression function within the intersection region, i.e.:

[0051]

[0052] Since this result is related to the spatial size of the supplementary lighting area, in order to make it comparable and facilitate the construction of scheduling decision logic, the system divides the integral value by the area of ​​the supplementary lighting area. Thus, the normalized illuminance overlap rate per unit area is obtained. .

[0053] The spatial adjacency between different areas is calculated by measuring the Euclidean distance between the geometric centers of luminaires in different lighting areas. This involves extracting the center coordinates of the luminaires within each area to obtain the two-dimensional spatial coordinates of their geometric centers. Then, the Euclidean distance between the center coordinates of any two luminaires in any given area is calculated. If the calculated result is less than a preset adjacency distance threshold, the two areas are determined to be spatially adjacent. This adjacency relationship is recorded as a structural connection item in the lighting scheduling model, indicating whether adjacent supplementary lighting areas have the physical conditions for synchronous linkage when the main lighting area undergoes a state change. Furthermore, after completing the spatial adjacency calculation... After the initial assessment, the system also performs a similarity analysis on the functional attributes of each area. This functional similarity is achieved by reading the tag information bound to each area during deployment. These tags include office area, rest area, corridor, and display space. The system performs semantic-level matching between the tags of the current area and the functional templates configured by the user during the setup phase. A functional similarity score is formed by combining keyword similarity and tag category weight. This score is used to evaluate the consistency of functional behavior between areas. In the scheduling logic, the system marks the supplementary lighting areas with true spatial adjacency and functional similarity scores higher than a set threshold as priority linkage areas and as candidate areas for the first round of scheduling response when the main lighting status changes.

[0054] The degree of light group sharing is introduced as a key parameter for evaluating the linkage intensity between the main lighting area and the supplementary lighting area. This degree of sharing is determined by statistically analyzing the ratio between the number of shared LED lights between the two areas and the total number of lights in the combined areas. In practice, the system extracts the light fixture lists for the main lighting area and the supplementary lighting area to be evaluated, performs an intersection operation to obtain the number of shared lights, and then calculates the ratio between this shared number and the sum of the total number of lights in both areas, thus obtaining the light group sharing degree index. This index reflects the degree of overlap and integration of lighting equipment between areas at the physical structure level. The light group sharing degree is quantified as a linkage intensity coefficient and used as an input parameter. The lighting decision model is used to guide the selection of subsequent lighting linkage scheduling strategies. At the same time, after collecting multi-dimensional information on the spatial adjacency, functional similarity, and lamp sharing degree between the main lighting area and each supplementary lighting area, the system integrates them to construct a regional response priority map. This map ranks the linkage value of each supplementary lighting area through a comprehensive scoring mechanism. In the priority scoring, the degree of lamp sharing and functional similarity have higher weights. Supplementary lighting areas with a high proportion of shared lamps with the main lighting area and strong consistency in regional functional behavior are selected as the first choice for linkage activation. When the main lighting status changes or an energy-saving strategy is triggered, the system will perform responsive scheduling of the supplementary lighting areas according to the priority order based on this map.

[0055] Example 1:

[0056] Combined with appendix Figure 4 Taking the lobby on the fifth floor of a smart office building as an example, this area has frequent passage and short-term stays, with multiple entrances and passageways leading to different functional areas. Multiple LED light sets are installed for localized lighting management. After collecting spatial structure data of this area, the system uses LiDAR and Building Information Modeling (BIM) technology to construct a simplified spatial topology map including information on obstructions and access restrictions. User Li enters from the northwest entrance, aiming for the conference room in the southeast corner. The system locates Li's current position and obtains his target area as he enters the entrance, then generates three candidate paths based on the spatial topology: paths P1, P2, and P3.

[0057] Path P1 is the main passageway, 18 meters long, but it suffers from glare from glass walls, affecting lighting efficiency; its estimated light benefit integral is 520 lux·m. Path P2 passes through the intermediate data area, but is obstructed by partitions, reducing the effective lighting area; its path length is 22 meters, with a light benefit of 570 lux·m. Path P3 detours through the open negotiation area, avoiding obstructions; its path length is 21 meters, with a light benefit of 640 lux·m. The system calculates the composite utility value according to the formula, assigning weight coefficients to path length and light benefit respectively. and The calculation is as follows:

[0058] The composite utility value of P1 is:

[0059] The composite utility value for P2 is:

[0060] The composite utility value for P3 is:

[0061] The system sets a path utility threshold of 0.3, therefore all three paths meet the lighting conditions and enter the candidate selection stage. With all paths meeting the utility threshold, the system first sorts the paths by length in ascending order, resulting in P1, P3, and P2. Although path P1 is the shortest, it has 5 nodes, including two turns, two crossing points, and one slow section. P3 has only 3 nodes, a relatively smooth path with no sharp turns, and P2 has 4 nodes. Therefore, based on the node count rule, the system selects P3 as the final path to be lit.

[0062] Upon entering the lighting response phase, the system identifies the lighting area based on Mr. Li's gaze direction. The gaze direction is inferred from his historical head-facing angle distribution after entering the room as southeast-east, and combined with the current head orientation identified by the camera, the orientation is confirmed to be 103 degrees. The system searches for light clusters intersecting with the viewing angle in this direction, and determines the target illuminance threshold to be 400 lux based on the lighting requirements at the conference room entrance. The brightness coverage of light cluster set L1 is calculated, setting the illumination area of ​​this set within the user's field of vision to 7.8 square meters. With the total user field of vision area being 10 square meters, the brightness coverage index is 0.78. The system's coverage threshold is set to 0.8, L1 meets the activation condition and is activated, with its initial brightness cap limited to 80% to reduce power consumption.

[0063] User Li chose to travel via path P3 to the southeast corner conference room. After illuminating the L1 light group in the path and the conference room entrance area, the system detected a natural illuminance of 90 lux in the current area based on the real-time light monitoring module, while the target illuminance at the conference room entrance was 400 lux. Therefore, the system calculated the illuminance compensation by including natural light as a partial contribution, setting the required compensation illuminance at 310 lux. The system selected five LED units with adjustable brightness within the L1 light group and, based on their illumination coverage area of ​​7.8 square meters within the user's field of vision, and a total field of vision of 10 square meters, derived a preliminary brightness coverage index of 0.78. Taking into account the contribution of natural light, the system dynamically adjusted this value, proportionally correcting the effective illumination area to 6.2 square meters. The updated brightness coverage index was 0.62, significantly lower than the system's set coverage threshold of 0.8. Therefore, the system confirmed that the area could be effectively activated.

[0064] To control initial energy consumption, the system set the brightness limit for each of the five light groups in L1 to 85% of their respective maximum output power, and adjusted the illuminance in stages based on Li's standing behavior. After entering the conference room, Li briefly stopped at the entrance to organize materials. The system, through infrared and visual behavior analysis modules, identified his low movement frequency, high standing stability, and no tendency to move in a short period of time. Therefore, the initial illuminance reduction step was set to 20 lux every 30 seconds, with a minimum reduction limit of 280 lux. If Li remained standing for more than 90 seconds without significant movement, the system continued to reduce the illuminance to 260 lux. At this point, Li did not exhibit any intervention behavior, and the system determined that his visual comfort state had not been disrupted.

[0065] At approximately 150 seconds, Li suddenly manually adjusted the lights via a swipe on his smart terminal, restoring the illuminance to 330 lux. The system recorded this action as an adverse feedback event, marked the corresponding light group as a priority sensing node, and regenerated the illuminance control curve based on the environmental conditions, pausing subsequent reductions. Simultaneously, the system searched for interconnected lighting areas around the main lighting area L1, identifying the auxiliary lighting area L2 on the east side. L2 shares two lights with L1, with the geometric center distance between the lights less than two meters, thus being determined as a spatially adjacent area. Furthermore, L2 belongs to the same meeting space and was assigned a functional similarity score of 0.9 based on the user-configured template. Further detection revealed that the spatial illuminance overlap rate between L2 and L1 in the current illuminance distribution reached 0.35, exceeding the system's set threshold of 0.3, therefore confirming L2 as a supplementary lighting area.

[0066] Considering that L2 is currently inactive, to avoid a sudden increase in power consumption, the system does not immediately activate L2. Instead, it adds L2 to the delayed activation queue and calculates the sharing degree as 0.4 based on the proportion of lights shared with L1. This value is used as the linkage strength coefficient and input into the decision model to determine the priority order in the delayed activation strategy. The system ultimately decides to activate L2 only when Li stays there for an extended period of time again or when a second discomfort feedback event occurs. The current linkage status information is then written into the regional response priority map for subsequent scheduling and use.

[0067] User Li's lingering behavior in the conference room entrance area further triggered the system's dynamic adaptation mechanism for his visual comfort. Approximately 180 seconds after light group L1 was activated, the system detected that Li manually adjusted the lighting through the touch interface of his smart terminal. He slid the brightness bar three times consecutively, adjusting the current brightness from 330 lux to 360 lux, then 390 lux, and finally stabilizing at 420 lux. The system recorded this repeated brightness adjustment as an active intervention behavior and, based on preset behavior recognition rules, marked it as a manual brightness adjustment type within an uncomfortable feedback event.

[0068] Meanwhile, the system's background behavior monitoring module analyzed Li's dwell time curve after entering the area. Using a sliding time window method, it recorded Li's dwell trajectory from entry to the current location. The analysis revealed a rapid increase in dwell time between 150 and 180 seconds, with the slope rising from 0.4 to 1.2, significantly exceeding the system's set slope threshold of 0.8. This indicates that the user is likely to extend their stay in the area and expect higher illumination. Based on the changing trend of the dwell time curve, the system activates the prediction module to infer user intent.

[0069] To verify the accuracy of the predictive response mechanism, the system cross-referenced the changes in the slope of the dwell time curve with historical user behavior data. It was found that Li manually adjusted the lighting after staying for more than 200 seconds on each of his four previous visits to the conference room area. Three of these adjustments were made via a smart terminal, and once via voice command to "turn the lights up." Based on this statistical behavior template, the system was set to initiate a feedback preprocessing action when the dwell time approached a threshold of 180 seconds.

[0070] The system proactively entered the prediction and preprocessing process when Mr. Li's dwell time reached 160 seconds. The preprocessing strategies included pausing the brightness reduction mechanism, adjusting the target illuminance limit to 390 lux in advance, and reducing the response delay of the two LED units closest to Mr. Li's position in LED group L1 to within 200 milliseconds to improve real-time brightness response efficiency. The system also preheated the voice recognition module, giving it higher command recognition sensitivity during this period to achieve low-latency response when the user issues enhanced voice commands.

[0071] After the control strategy was implemented, the system observed that Li's overall dwell time after manually adjusting the brightness increased to 300 seconds, and no further intervention occurred. This indicates that the system's early intervention and feedback preprocessing effectively met the user's lighting needs, avoided discomfort caused by insufficient illuminance, and at the same time limited the upper limit of illuminance through fine adjustment, controlling the overall power consumption within the energy-saving range.

[0072] Example 2:

[0073] Combined with appendix Figure 5Based on Example 1, after completing the organization of meeting materials, the employee waited temporarily outside the 10th meeting room for colleagues to meet. After standing still for approximately 90 seconds, the system actively identified the employee's eye movement trajectory using a wide-angle camera module deployed inside the ceiling light fixture. The system ran an eye-tracking focal point detection algorithm based on real-time acquired image frame data, determining that the coordinates of the area where the employee continuously gazed were concentrated in the lighting boundary area near the door, approximately 1.8 meters from the employee's position. This area was the edge illumination coverage zone of light fixture L1, and it was also located at the boundary between the main lighting and corridor lighting, where the ambient illuminance rapidly transitioned from 320 lux to 180 lux, forming a significant illuminance discontinuity. According to the system's illumination boundary space model, this focal point fell 3 to 5 centimeters outside the main lighting area's contour line, belonging to a standard boundary buffer zone.

[0074] The system executes a spatial cross-analysis process, fitting and comparing the coordinates of the eye-tracking focal point with the predefined illumination boundary illuminance curve. It determines that the gaze point is located in the transition zone of the main illumination boundary, triggering a local perception anomaly monitoring mechanism. Further analysis of Li's gaze duration reveals that he continuously focused on this area for more than 3 seconds without any head rotation or gaze shift during this period. Based on the default threshold setting rules, this is determined to be a valid gaze boundary event.

[0075] Based on the above detection results, the system implements a brightness fine-tuning strategy for the boundary area. The original illuminance gradient distribution map of the lighting area is retrieved, and the coverage area of ​​the smallest brightening unit is determined to be two LED nodes within a 1-meter radius of the boundary. Their brightness output is increased by 15% from the current setting, i.e., from the original current-corresponding output brightness of 250 lux to 288 lux. Simultaneously, the adjustment cycle is set to be completed within 2 seconds to ensure that the user receives soft light compensation before experiencing discomfort.

[0076] The system marks this boundary area as a local perception blind spot and records it as one of Li's gaze preference characteristics, storing it in the user's personalized lighting model. If Li subsequently enters the same area again, the system can use this marking to pre-execute a pre-lighting strategy, reducing the subjective discomfort caused by gazing at the boundary again. Furthermore, this blind spot marking will also be used in matching lighting area strategies. If the L1 main light group is activated and the linked lighting judgment involves this boundary area, the system will prioritize its special characteristics within the blind spot, triggering a delayed or prioritized lighting strategy to further optimize light field uniformity.

[0077] User Li enters meeting area A1 on the sixth floor of the building every morning at 8:00 AM. This area is the main lighting area, while the adjacent area A2, due to its proximity to the window, is often in auxiliary lighting mode during the morning when natural light is strong. To achieve energy-saving goals, the system needs to determine whether to activate the lights in area A2 based on the luminous flux coverage of main area A1 over A2.

[0078] The system collects the current time slice. Spatial illuminance distribution function An illuminance model is established using measured data on a two-dimensional plane. (Setting...) This indicates that the maximum illuminance of the main lighting area at its center point x=1, y=1 is 120 lx, and it decreases outward in a Gaussian manner. This area is defined as the main area. The overlapping area between the fill light area A2 and the main area is... That is, the two overlap at The region has an area of ​​1.

[0079] Calculate the illuminance integral term within the overlapping region:

[0080]

[0081] For ease of calculation, substituting the variable transformations u=x-1, v=y-1, the integration region becomes :

[0082]

[0083] Next, considering the spatial illuminance edge suppression factor, i.e. the mixed second-order partial derivative, we set:

[0084]

[0085] The weighted integral expression is then:

[0086]

[0087] exist Within the region, due to exponent term Therefore, the overall weighted result will be slightly lower than 87.66. Setting the average weight at 0.85, the integral value is approximately:

[0088]

[0089] The normalized overlap rate is calculated as follows:

[0090]

[0091] The system sets the illumination overlap rate activation threshold to 15. If this value is higher than 15, it indicates that area A2 is well covered by the main illumination at the current time and does not require supplemental lighting activation. However, further analysis of the user's eye movement data revealed that Li's gaze hotspot was concentrated at the boundary between A1 and A2 in the past 60 seconds, and there were records of insufficient illumination feedback events. This area has been marked as a perception blind spot by the system.

[0092] Therefore, although If the brightness exceeds the threshold, the system still determines that the user has potential discomfort in this area. Based on the perception blind zone mechanism, A2 is designated as the priority activation sequence. At the same time, the activation brightness is controlled within 50% and the illuminance increment step is set to 5 lx / min to provide responsive supplemental lighting and prevent energy waste.

[0093] During the event in meeting area A1, the system not only accurately calculated the overlap of spatial illuminance between the main area and adjacent areas, but also further incorporated factors such as spatial adjacency, functional similarity, and the degree of light group sharing to improve the determination of the linkage response mechanism for supplementary lighting areas. The following is an example illustrating this strategy and the calculation process. The main lighting area is meeting area A1, and its geometric center coordinates are set as follows: The adjacent candidate supplementary lighting area includes corridor area A2, whose geometric center is And the printing room A3, the center of which is The system's spatial adjacency threshold is set to 3 meters.

[0094] Calculate the Euclidean distance between A1 and A2:

[0095]

[0096] Next, calculate the distance between A1 and A3:

[0097]

[0098] Therefore, A2 and A1 are spatially adjacent because the distance between them is less than the adjacency threshold. A3 is not an adjacent area and does not participate in the first priority response. Next, the similarity of area functions is analyzed. The system identifies from the environmental label data that A1 is for meeting discussions, A2 is for personnel access, and A3 is for auxiliary office work. The user configuration template sets the similarity score between the meeting discussion area and the access area to 0.6, and the similarity score with the auxiliary office area to 0.3. Therefore, A2 scores higher.

[0099] The third step is to assess the degree of light sharing. The system identified a total of 6 lights used in area A1, of which 3 lights physically overlap with or intersect with the lighting boundary of A2, while only 1 light in A3 has coverage overlap with A1.

[0100] The degree of sharing between A1 and A2 is:

[0101]

[0102] Where 5 represents the number of independent lamps in A2, and minus 3 represents the number of duplicate counts.

[0103] The degree of sharing between A1 and A3 is:

[0104]

[0105] The system presets a response weight priority of 0.4 for the degree of sharing and a functional similarity priority of 0.6. Therefore, the system calculates a comprehensive linkage priority score for each region:

[0106] A2:

[0107]

[0108] A3:

[0109]

[0110] The results show that A2 is superior to A3 in terms of spatial adjacency, functional similarity and lamp group sharing. The system lists it as the first response level supplementary lighting area and prioritizes its activation and control during user gaze shift, boundary staring or main area brightness reduction.

[0111] When Li's meeting reached the 15th minute, the system detected that the second level of illumination reduction had been implemented in area A1. However, the user began to frequently adjust the screen position and turn his head to look at the notepad at the junction with the aisle. The camera detected that his gaze was focused on the edge area of ​​A2 and triggered discomfort feedback. The system quickly found the priority response area according to the linkage priority map. Because A2 had a priority of 0.51, which was higher than the system response threshold of 0.4, two lights were quickly activated to provide localized flexible supplementary lighting.

[0112] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A smart control method for energy-saving LED lights, characterized in that... include: Acquire spatial structure data, personnel location and activity trajectory, and combine them with obstruction information to form a simplified spatial topology; Based on the topology, several candidate paths are determined for the user to reach the lighting area that meets the target illuminance threshold. The composite utility value of each candidate path is calculated by combining the path length and the illumination benefit according to a preset rule. When the composite utility value meets the preset path utility threshold, the lighting condition is generated. When the lighting conditions are met, based on the direction of the person's line of sight and the functional attributes of the area, one or more candidate areas of lamp groups are determined that cover the user's field of vision angle range and can make the illuminance in the field of vision reach the target illuminance threshold. The brightness coverage index of each candidate area is calculated, and a set of lamp groups with a coverage index not higher than the coverage threshold is selected for activation control. After the light group is activated, based on the user's dwell time and behavioral characteristics, a multi-level illuminance reduction with non-fixed step size is performed according to a preset step size range and time interval. If no user intervention or adverse feedback event is detected during the illuminance reduction process, the illuminance reduction continues to reduce energy consumption. During lighting control, supplementary lighting areas that have a linkage response relationship with the current main lighting area are identified, linkage relationship information is generated, and a delayed activation or power suppression strategy is applied to the supplementary lighting areas.

2. The intelligent control method for energy-saving LED lamps according to claim 1, characterized in that... The illumination gain is the integral result of the estimated illuminance value that can be achieved within a unit path length; The path length is a comprehensive measure of the actual distance the user moves along the restricted path, excluding areas blocked by obstructions. The composite utility value is calculated by weighting the path length and illumination gain by assigning preset weighting coefficients to each.

3. The intelligent control method for energy-saving LED lamps according to claim 1, characterized in that... When the composite utility values ​​of multiple candidate paths all meet the path utility threshold, the path lengths of these candidate paths are sorted in ascending order, and the candidate path with the highest ranking is selected to generate the lighting conditions; if there are candidate paths with the same path length, the candidate path with fewer path nodes is selected first; if they are still the same, the selection is made according to a preset priority table; the gaze direction is determined by combining the user's historical facing angle distribution statistics with the current head orientation information; the brightness coverage index of the lighting area is the ratio of the actual lighting area of ​​the lamp set within the user's field of vision to the total field of vision area.

4. The intelligent control method for energy-saving LED lamps according to claim 1, characterized in that... The brightness coverage index is calculated by taking into account the current ambient natural light intensity and performing illumination compensation calculations; the activation control includes setting an upper limit value for the brightness of the selected light group to control the initial power consumption when lighting up; wherein the step size range of the illuminance decrease is dynamically adjusted according to the user's behavioral characteristics, such as action frequency, standing stability, and movement trend.

5. The intelligent control method for energy-saving LED lamps according to claim 1, characterized in that... The discomfort feedback events include the user manually adjusting the brightness, repeatedly staying in the original area for more than the expected time, or the system detecting that the user is staring at the lighting boundary area; wherein the supplementary lighting area is an area where the spatial illuminance overlap rate with the main lighting area exceeds a preset threshold; wherein the linkage relationship information includes the spatial adjacency, functional similarity, and degree of lamp sharing between the main lighting area and the supplementary lighting area.

6. The intelligent control method for energy-saving LED lamps according to claim 5, characterized in that... The user's manual brightness adjustment behavior includes sliding the brightness bar through the smart terminal control interface, repeatedly operating the physical switch more than a preset threshold, or having a voice command containing brightness enhancement semantics; wherein the system predicts the user's intention based on the change in the user's dwell time curve in the same area, and performs feedback preprocessing actions in advance before the dwell time approaches the threshold.

7. The intelligent control method for energy-saving LED lamps according to claim 5, characterized in that... The system detects the user's gaze boundary region by: identifying the user's eye focus point through a camera and performing spatial cross-analysis with the current illumination boundary; after detecting the user's gaze boundary region, the system performs brightness fine-tuning or brightening expansion operations on the boundary region and records it as a local perception blind spot.

8. The intelligent control method for energy-saving LED lamps according to claim 5, characterized in that... The calculation of the spatial illuminance overlap rate includes integrating the luminous flux distribution overlap between the main lighting area and the supplementary lighting area, and fitting and comparing it with the global spatial lighting demand surface. After identifying the supplementary lighting area, the system first determines whether the area is in the user's blind spot. If so, the area is set to a suppressed state and does not participate in the next round of activation scheduling.

9. The intelligent control method for energy-saving LED lamps according to claim 8, characterized in that... The spatial adjacency is calculated based on the Euclidean distance between the geometric centers of the lamps, and areas below the preset adjacency threshold are judged to be adjacent areas. Functional similarity is determined by matching regional label information with user-configured preset templates and assigning a regional functional similarity score.

10. The intelligent control method for energy-saving LED lamps according to claim 8, characterized in that... The degree of lamp sharing is calculated by analyzing the ratio of the number of shared lamps between two areas to the total number of lamps, and is input into the lighting decision model as a linkage intensity coefficient. The linkage information is used to construct a regional response priority map, prioritizing the activation of supplementary lighting areas that share a high degree of illumination with the main lighting area and have high functional similarity.

Images