Intelligent lighting control method and system for LED lamps

Abstract

This invention relates to the field of lighting technology and provides an intelligent lighting control method and system for LED lights. The method includes: deploying several sensing terminals integrating multi-dimensional sensing modules; the sensing terminals performing local edge preprocessing on the collected multi-dimensional data and extracting feature parameters to preliminarily determine lighting needs and generate basic control commands; a cloud-edge collaborative control platform receiving the multi-dimensional data, combining preset lighting scene modes, user permission configurations, and historical operating data to dynamically decide on the brightness, color temperature, and on / off status of the LED lighting terminals, generating precise control commands; and the cloud-edge collaborative control platform sending the precise control commands to the control drive modules of the corresponding LED lighting terminals, which then adjust the LED light source operating parameters according to the precise control commands. This invention can fuse data from all terminals within a region, breaking the isolation of single-point data, and through data association, complementarity, and verification, forming a comprehensive and accurate assessment of the overall lighting needs of the region.

Description

Technical Field

[0001] This invention relates to the field of lighting technology, and more specifically, to an intelligent lighting control method and system for LED lamps. Background Technology

[0002] With the rapid development of IoT, sensor, and intelligent control technologies, LED lighting, as a high-efficiency, energy-saving, and long-life lighting solution, has been widely used in various scenarios such as office buildings, industrial plants, public transportation, and residential communities, gradually replacing traditional incandescent and fluorescent lamps and becoming the mainstream development direction of the lighting industry. Currently, the core development needs of LED lighting have shifted from simple lighting functionality to intelligent control, precise adaptation, high energy efficiency, and convenient operation and maintenance. Especially in large public areas or industrial settings, higher requirements are placed on the intelligence level, real-time response, scene adaptability, and operational reliability of lighting systems.

[0003] Chinese invention patent application number CN121262705A discloses an intelligent lighting control method and system based on the Internet of Things (IoT). It deploys lightweight edge computing nodes on IoT terminals to collect pre-processed data in real time. The pre-processed data is then transmitted to a cloud platform. Based on an LSTM (Long Short-Term Memory) network, the system analyzes historical user behavior sequences and uses a GCN (Graph Convolutional Network) to model the spatial dependency between environmental parameters and user behavior, constructing a hybrid model. A dual temporal difference optimization algorithm, combined with a multi-objective reward function, dynamically adjusts the parameters of the hybrid model to obtain a target hybrid model. In essence, by collecting pre-processed data through lightweight edge computing nodes, the cloud platform uses an LSTM+GCN hybrid model combined with dual temporal difference optimization to generate control strategies, achieving on-demand lighting. However, this method lacks effective fusion and cross-validation of data from multiple terminals within the area, making it prone to misjudgments due to single-point sensor failures or data incompleteness. It suffers from low data fusion levels and insufficient accuracy in control strategies. Summary of the Invention

[0004] Based on this, in order to solve the problems of low data fusion and lack of precision in control strategies in existing technologies, the present invention provides an intelligent lighting control method and system for LED lights, the specific technical solution of which is as follows: A method for intelligent lighting control of LED lights includes the following steps: A number of sensing terminals integrating multi-dimensional sensing modules are deployed, with each sensing terminal corresponding to an LED lighting terminal. The multi-dimensional sensing modules are used to synchronously collect multi-dimensional data, including ambient light intensity, human presence status, and frequency of human activity. The sensing terminal performs local edge preprocessing on the collected multi-dimensional data and extracts feature parameters. Based on the feature parameters, it makes a preliminary judgment on lighting needs and generates basic control commands. At the same time, the preprocessed multi-dimensional data and feature parameters are uploaded to the cloud-edge collaborative control platform. The cloud-edge collaborative control platform receives multi-dimensional data uploaded by each edge node, and combines it with preset lighting scene modes, user permission configurations and historical operation data to make dynamic decisions on the brightness, color temperature and on / off status of LED lighting terminals, and generate precise control commands. The cloud-edge collaborative control platform sends precise control commands to the control and drive modules of the corresponding LED lighting terminals, and the control and drive modules adjust the working parameters of the LED light source according to the precise control commands.

[0005] The described intelligent LED lighting control method performs local edge preprocessing and extracts feature parameters from collected multi-dimensional data. Simultaneously, the preprocessed multi-dimensional data and feature parameters are uploaded to a cloud-edge collaborative control platform. This platform receives the multi-dimensional data from each edge node and, combined with preset lighting scene modes, user permission configurations, and historical operating data, dynamically determines the brightness, color temperature, and on / off status of the LED lighting terminal, generating precise control commands. It can fuse data from all terminals within the area, breaking the isolation of single-point data. Through data association, complementarity, and verification, it forms a comprehensive and accurate assessment of the overall lighting needs of the area. This solves the problems of existing technologies lacking effective fusion and cross-verification of multi-terminal data within the area, being prone to misjudgments due to single-point sensor failures or data bias, and exhibiting low data fusion levels and insufficient control strategy precision.

[0006] Preferably, the intelligent lighting control method for LED lights further includes the following steps: The LED lighting terminal collects its own working status data in real time and uploads it back to the cloud-edge collaborative control platform through the sensing terminal to form a closed-loop control.

[0007] Preferably, the intelligent lighting control method for LED lights further includes the following steps: The cloud-edge collaborative control platform regularly performs adaptive optimization of control strategies based on historical operational data, user behavior habits, and environmental change trends.

[0008] Preferably, the intelligent lighting control method for LED lights further includes the following steps: The cloud-edge collaborative control platform performs fault self-diagnosis based on the working status data uploaded by the LED lighting terminals. When a fault is detected, an alarm mechanism is immediately triggered, and maintenance prompts are pushed to the management personnel terminal. At the same time, the platform automatically switches to the backup lighting scheme.

[0009] Preferably, the specific method for initially determining lighting requirements and generating basic control commands through characteristic parameters includes the following steps: Obtain the maximum activity frequency threshold, and then determine the activity intensity based on the activity frequency and the maximum activity frequency threshold. Obtain the preset light compensation threshold, and then obtain the light difference compensation value based on the ambient light intensity and the light compensation threshold. The switching factor is obtained based on the human body's state and a preset probability threshold. Basic control commands are generated based on the switching factor, the intensity of human activity, and the light difference compensation value.

[0010] Preferably, the specific method for generating precise control commands includes the following steps: The system obtains the real-time brightness and target brightness of the LED light, calculates the comfort deviation based on the difference between the real-time brightness and the target brightness, and calculates the energy consumption value of the LED light based on the real-time brightness. The system obtains the real-time color temperature of the LED light and the preset color temperature of the scene, obtains the color temperature difference based on the real-time color temperature and the preset color temperature of the scene, and obtains the biorhythm matching value based on the color temperature difference and the ideal color temperature based on time. The objective function is obtained based on the comfort deviation, LED lamp energy consumption value, and biological rhythm matching value. Based on the objective function and the particle swarm algorithm, dynamic decisions are made on the brightness, color temperature, and on / off status of the LED lighting terminal to generate precise control commands.

[0011] Preferably, the specific method for performing fault self-diagnosis includes the following steps: Obtain the current harmonic distortion rate, temperature rise value, current change rate and flicker index of the LED lighting terminal, and obtain the fault feature vector based on the current harmonic distortion rate, temperature rise value, current change rate and flicker index; Obtain the fault baseline vector, obtain the fault prediction probability based on the fault feature vector and the fault baseline vector, and perform fault self-diagnosis based on the fault prediction probability.

[0012] An intelligent lighting control system for LED lights, used to implement the aforementioned intelligent lighting control method for LED lights, includes: A distributed sensing terminal network, which deploys several sensing terminals that integrate multi-dimensional sensing modules; Among them, the sensing terminal and the LED lighting terminal are deployed one-to-one. The sensing terminal is used to perform local edge preprocessing on the collected multi-dimensional data and extract feature parameters. The lighting needs are initially judged by the feature parameters and basic control commands are generated. At the same time, the preprocessed multi-dimensional data and feature parameters are uploaded to the cloud-edge collaborative control platform. The multi-dimensional sensing module is used to synchronously collect multi-dimensional data including ambient light intensity, human presence status, and frequency of human activities. The cloud-edge collaborative control platform is used to receive multi-dimensional data uploaded by each edge node, and combine it with preset lighting scene modes, user permission configurations and historical operation data to make dynamic decisions on the brightness, color temperature and on / off status of LED lighting terminals, generate precise control commands, and send the precise control commands to the control drive module of the corresponding LED lighting terminal. The control and drive module adjusts the operating parameters of the LED light source according to precise control commands.

[0013] Preferably, the cloud-edge collaborative control platform is also used to periodically adaptively optimize the control strategy based on historical operating data, user behavior habits and environmental change trends, and to perform fault self-diagnosis through working status data uploaded by LED lighting terminals. When a fault is detected, an alarm mechanism is immediately triggered, and maintenance prompts are pushed to the management personnel terminal, while automatically switching to the backup lighting scheme.

[0014] Preferably, the sensing terminal also integrates an Internet of Things (IoT) communication module, which adopts a dual-mode adaptive communication method of LoRa and NB-IoT, and automatically switches the communication mode according to the communication distance and network signal strength. Attached Figure Description

[0015] The invention will be further understood from the following description taken in conjunction with the accompanying drawings. The components in the drawings are not necessarily drawn to scale, but rather the emphasis is on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0016] Figure 1 This is a schematic diagram of the overall process of an intelligent LED lighting control method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a specific method for initially determining lighting needs and generating basic control commands in one embodiment of the present invention. Figure 3 This is a flowchart illustrating a specific method for generating precise control commands in one embodiment of the present invention; Figure 4 This is a schematic diagram of the overall process of an intelligent lighting control method for LED lamps according to another embodiment of the present invention; Figure 5 This is a flowchart illustrating a specific method for fault self-diagnosis in one embodiment of the present invention; Figure 6 This is a schematic diagram of the overall structure of an intelligent LED lighting control system according to an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to its embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of the invention.

[0018] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] In this invention, "first" and "second" do not represent a specific quantity or order, but are merely used to distinguish names.

[0021] Before describing the embodiments of the present invention, a brief introduction to the prior art will be given.

[0022] With the rapid development of IoT, sensor, and intelligent control technologies, LED lighting, as a high-efficiency, energy-saving, and long-life lighting solution, has been widely used in various scenarios such as office buildings, industrial plants, public transportation, and residential communities, gradually replacing traditional incandescent and fluorescent lamps and becoming the mainstream development direction of the lighting industry. Currently, the core development needs of LED lighting have shifted from simple lighting functionality to intelligent control, precise adaptation, high energy efficiency, and convenient operation and maintenance. Especially in large public areas or industrial settings, higher requirements are placed on the intelligence level, real-time response, scene adaptability, and operational reliability of lighting systems.

[0023] However, existing LED intelligent lighting control technology still has many technical defects and shortcomings in practical applications, making it difficult to fully meet the above application needs. These are mainly reflected in the following aspects: Firstly, the control mode is fragmented, resulting in poor regional coordination. Most existing LED lighting control solutions adopt a single-point independent control mode, meaning each LED lighting terminal corresponds to only one sensor, achieving simple on / off switching or brightness adjustment based on ambient light intensity or human presence signals collected at a single point. In this mode, the control logic of each lighting terminal is independent, lacking regional-level coordinated scheduling. This easily leads to uneven lighting issues such as bright spots and dark areas. Furthermore, frequent start-stop cycles may occur due to false detections by single-point sensors, affecting lighting comfort, shortening the lifespan of LED luminaires and driver modules, and increasing maintenance costs. For example, in office areas or corridors, sensors at adjacent points may detect "occupant" and "no one" signals respectively, causing adjacent lights to alternately turn on and off, severely impacting pedestrian traffic and work experience.

[0024] Secondly, data processing relies heavily on the cloud, resulting in insufficient real-time performance and reliability. Some existing IoT-based LED lighting control solutions upload all sensor data to a remote cloud platform for centralized analysis and decision-making, with control commands then being sent from the cloud to the terminal for execution. This centralized cloud control model is highly dependent on network communication. When network interruptions, signal delays, or cloud platform failures occur, the lighting system loses its control capabilities, failing to respond to real-time changes in personnel and the environment, leading to lighting paralysis or control failure. Furthermore, the remote transmission of large amounts of raw data not only consumes network bandwidth but also introduces data transmission delays, making it difficult to meet the real-time control requirements of scenarios such as industrial plants and transportation hubs.

[0025] Third, the data fusion level is low, and the control strategy lacks precision. In existing technologies, data such as ambient light intensity, human presence, temperature, and humidity collected by sensors are mostly processed independently, lacking effective fusion and cross-verification of data from multiple terminals within the area. This makes it prone to misjudgment in control due to single-point sensor failure or incomplete data. At the same time, control strategies are mostly based on fixed threshold settings, adopting a one-size-fits-all adjustment method. They cannot adaptively optimize lighting parameters (brightness, color temperature, etc.) according to dynamic characteristics such as the distribution of people, activity frequency, and environmental change trends within the area. This makes it difficult to balance lighting comfort and energy saving. Either excessive lighting is used to ensure comfort, resulting in energy waste, or lighting quality is reduced to pursue energy saving, affecting the user experience.

[0026] Fourth, the system has high operation and maintenance costs and lacks autonomous diagnostic and fault-tolerant capabilities. Existing LED lighting control systems largely rely on manual inspection and troubleshooting. When lamps or sensors malfunction, it is difficult to detect and issue warnings in real time. Repairs are often only possible after the lighting function has completely failed, affecting normal lighting use and increasing the workload and cost of manual maintenance. Furthermore, most systems lack backup lighting switching and offline emergency control mechanisms. If the core control unit or communication link fails, it cannot quickly switch to emergency lighting mode, further reducing the system's reliability and fault tolerance.

[0027] In other words, current LED intelligent lighting control technology still has significant shortcomings in terms of regional coordination, real-time response, control precision, and operational reliability, making it difficult to fully adapt to the actual needs of various complex application scenarios. Therefore, developing an IoT-based LED intelligent lighting control method and system that can achieve regional data fusion analysis, cloud-edge collaborative intelligent decision-making, adaptive optimization control, and possess high reliability and feasibility has become a pressing technical problem for those skilled in the art.

[0028] One of the objectives of this invention is to address the problems of low data fusion level and lack of control strategy accuracy in existing technologies, thereby improving the data fusion level and control strategy accuracy of intelligent lighting control for LED lights.

[0029] To achieve the above objectives, such as Figure 1 As shown, an embodiment of the present invention provides an intelligent lighting control method for LED lights, which includes the following steps: S1, deploy several sensing terminals integrating multi-dimensional sensing modules, with each sensing terminal corresponding to an LED lighting terminal; the multi-dimensional sensing modules are used to synchronously collect multi-dimensional data including ambient light intensity, human presence status, and frequency of human activities.

[0030] Specifically, a distributed sensing terminal network is established, deploying several sensing terminals that integrate multi-dimensional sensing modules and low-power IoT communication modules. Each sensing terminal is deployed in a one-to-one correspondence with an LED lighting terminal, and each sensing terminal has a built-in unique identifier. The multi-dimensional sensing modules synchronously collect data on ambient light intensity, human presence status, frequency of human activity, and ambient temperature and humidity. The low-power IoT communication module adopts a dual-mode adaptive communication method of LoRa and NB-IoT, automatically switching the communication mode according to the communication distance and network signal strength.

[0031] Ambient light intensity refers to the illuminance values ​​of natural light and surrounding light sources. Human presence can be understood as the confidence level of human presence, i.e., the probability of human presence, which is the result of multi-sensor fusion. Generally, a value greater than 0.4 is considered a valid presence. For example, the probability of human presence can be calculated by using one or more of the following: voiceprint information collected by a sound acquisition unit, environmental image information collected by an image sensor, and environmental thermal imaging information collected by an infrared thermal imager. The frequency of human activity refers to the number of human limb movements detected per unit time, which can be sampled using millimeter-wave radar, with typical values ​​between 2 and 15.

[0032] S2, the sensing terminal performs local edge preprocessing on the collected multi-dimensional data and extracts feature parameters. Based on the feature parameters, it makes a preliminary judgment on lighting needs and generates basic control commands. At the same time, the preprocessed multi-dimensional data and feature parameters are uploaded to the cloud-edge collaborative control platform.

[0033] Step S2 mainly involves data preprocessing and edge intelligent analysis. Specifically, the sensing terminal performs local edge preprocessing on the collected multi-dimensional data, filters out abnormal data and extracts feature parameters, preliminarily judges lighting needs through preset thresholds, and generates basic control commands. At the same time, the preprocessed data and feature parameters are uploaded to the cloud-edge collaborative control platform through the IoT communication module. Edge nodes perform fusion analysis on the data from multiple terminals in the area to optimize the control strategy.

[0034] Edge nodes integrate and analyze data from multiple terminals within a region to optimize control strategies. Their core function is to integrate, process, and intelligently analyze data from multiple devices within the region on the edge side, closer to the sensing terminals and LED lighting terminals, rather than relying entirely on the remote cloud. Ultimately, they output lighting control solutions that better meet the actual needs of the site, balancing real-time control, local autonomy, and energy-saving precision.

[0035] Generally, an edge node is not a single device, but rather a regional local processing unit defined by physical space, such as a floor of a building, an office area, a corridor, or a workshop section. Each edge node covers a certain area, and all sensing terminals and LED lighting terminals deployed within that area are managed by that edge node. Multi-terminal data refers to pre-processed data uploaded by all sensing terminals within the jurisdiction of the edge node, rather than raw data. This data has been filtered for outliers and has had its core features extracted. Specifically, it includes ambient light intensity collected by each terminal, such as the brightness values ​​of different points within the area; human presence status (e.g., whether there is someone at a certain point, and the number of people); frequency of human activity (e.g., the density of people in a certain area, and the duration of their stay); ambient temperature and humidity; and working status data fed back by each LED lighting terminal (e.g., current brightness / color temperature, operating power consumption, whether it is working normally, operating current and voltage, and fault information).

[0036] A single terminal can only achieve localized control at a single point, which can easily lead to fragmented control. For example, if someone turns on a light at point A, while no one turns off a light at the adjacent point B, it can cause uneven lighting in the area. Edge nodes do not process data from individual terminals individually, but rather integrate data from all terminals within the area. This breaks down the isolation of single-point data and, through data association, complementarity, and verification, forms a comprehensive and accurate assessment of the overall lighting needs of the area. Specifically, this includes the following three layers of processing: 1. Data Association and Integration: The spatiotemporal data of each sensing terminal in the area are associated (e.g., "At 10:00, sensing terminals No. 3 and No. 4 in the area detected people, and the average ambient light intensity in the area is 200 lux"), and the current working status of each LED lighting terminal is also associated (e.g., "The current brightness of LED lights No. 5 and No. 6 in the area is 50%, and the color temperature is 4000K"), forming a complete data profile of the area's lighting scene. This avoids misjudgments caused by the one-sidedness of single-point data (e.g., turning off the lights when a single point is mistakenly detected as "no one" may easily overlook the situation where there are people at other points in the area).

[0037] 2. Complementary Data Verification: Cross-validation using multi-terminal data enhances data reliability. For example, if a sensing terminal in the area detects "ambient light intensity 150 lux" (too low, needs brightening), but two or more surrounding sensing terminals detect "ambient light intensity 300 lux" (normal), the edge node, through fusion analysis, determines that it is a single-point sensor malfunction. It avoids blindly executing brightening commands, preventing control errors due to a single device failure and ensuring system stability.

[0038] 3. Demand Characteristic Analysis: Based on the integrated complete data, analyze the core demand characteristics of lighting in the area, such as determining the current scene of the area (e.g., whether the office area is in peak working hours, whether the corridor is during the period of personnel passage), personnel behavior habits (e.g., the area is densely populated with people at fixed times, and people stay at a certain point for a short time), and environmental change patterns (e.g., the overall light intensity of the area is low on cloudy days, and the light intensity gradually changes in the morning and evening), to provide accurate basis for subsequent control strategy optimization.

[0039] Optimized control strategies refer to the process by which edge nodes adjust, correct, and refine the original basic control commands (single-point control commands generated by a single sensing terminal or globally universal commands issued from the cloud) based on the results of fusion analysis, forming a personalized control scheme that is more suitable for the real-time scenario of the region and takes into account both energy saving and comfort. The core optimization directions include four points: 2. Avoid fragmented control and achieve regional collaborative control: Optimize single-point independent control logic and replace it with overall regional collaborative control. For example, if there are people at certain points in the area, the edge node will not only turn on the lights at those points, but will adjust the brightness of the surrounding LED lights according to the distribution of people (the core area is brighter, and the surrounding area is slightly dimmer, forming a gradual transition), avoiding "bright spots" and "dark areas" and improving lighting comfort; at the same time, when the last person leaves the area, the edge node will uniformly control all LED lights in the area to turn off after a delay (instead of turning off individual lights individually), reducing the wear and tear from frequent start-stop of equipment and extending the life of the lamps.

[0040] 2. Dynamically adapt to scene requirements and optimize brightness / color temperature accuracy: Based on scene characteristics obtained from fusion analysis, optimize the adjustment parameters of brightness and color temperature to avoid "one-size-fits-all" control. For example, if fusion data in the office area shows that "on weekdays from 10:00 to 11:00, the population is dense and the ambient light intensity is stable at 250 lux", the edge node optimization control strategy will uniformly adjust the brightness of LED lights in the area to 70% and the color temperature to 4500K (to meet the needs of focused office work); if the ambient light intensity is subsequently detected to rise to 400 lux, the brightness will be automatically reduced to 40%, which will both meet the lighting needs and maximize energy saving, thus breaking through the limitations of traditional fixed threshold control and achieving scene-based dynamic adaptation.

[0041] 3. Balancing real-time response and energy saving, optimizing control timing logic: Optimize the execution timing and priority of control commands to avoid frequent start-stop. For example, if the data on the frequency of personnel activity in the area shows "one person passes through a corridor every 5 minutes," the edge node optimization strategy can be adjusted from "lights off immediately after personnel leave" to "lights off after a 3-minute delay," while reducing the brightness during the delay period to 30%. This avoids the need for frequent light switching when people move around (to improve convenience) and is more energy-efficient than keeping the lights on all the time, thus better reflecting actual personnel flow scenarios.

[0042] 4. Adapt to local offline scenarios and optimize control autonomy: Edge nodes have independent decision-making capabilities. When the cloud network is interrupted (such as network outage or cloud failure), they do not need to wait for cloud instructions. They can independently optimize and execute control strategies (such as maintaining basic lighting in the area and responding to the needs of personnel movement) based on the fusion data of multiple terminals in the area. This avoids lighting paralysis caused by reliance on the cloud, ensures system continuity, and solves the pain point of network instability in IoT scenarios.

[0043] As a preferred technical solution, such as Figure 2 As shown, the specific method for initially determining lighting requirements and generating basic control commands through characteristic parameters includes the following steps: S21, Obtain the maximum activity frequency threshold based on personnel activity frequency. and the maximum activity frequency threshold Obtaining personnel activity intensity .in, These represent the activity sensitivity coefficient and the Sigmoid function, respectively. The activity sensitivity coefficient is used to amplify the impact of high-frequency activities on brightness requirements, and is generally between 1.2 and 1.5, which can be adjusted appropriately according to actual needs.

[0044] The Sigmoid function is specifically represented as follows: .in, These are represented as the slope parameter and the transition inflection point, respectively, both of which can be adjusted according to actual settings. The Sigmoid function is mainly used to normalize the activity frequency to the [0,1] interval, so that low-frequency activities (such as sitting) correspond to low demand (such as 0.2~0.4), and high-frequency activities (such as moving) correspond to high demand (such as 0.8~1.0).

[0045] The slope parameter is generally set to 0.2 to achieve a smooth transition and avoid abrupt changes in activity intensity. The transition inflection point can be adaptively adjusted according to different scenarios, such as setting it to 0.7 for office areas and 0.3 for fitness areas, to match the activity intensity distribution of different scenarios.

[0046] S22, obtain the preset light compensation threshold, and obtain the light difference compensation value based on the ambient light intensity and the light compensation threshold.

[0047] The light difference compensation value is represented as max(1-ambient light intensity / light compensation threshold, 0). It outputs a positive value only when the ambient light intensity is less than the light compensation threshold, indicating that supplemental lighting is needed to avoid ineffective lighting when the ambient light is sufficient.

[0048] S23: Obtain the switching factor based on the human body's state and a preset probability threshold for existence, and generate basic control commands based on the switching factor, the intensity of human activity, and the light difference compensation value.

[0049] Here, when the presence of a human being is less than a preset presence probability threshold, the switching factor can be set to 0, defining no one is present and the lights are off, and the lighting demand value is set to 0; when the presence of a human being is not less than the preset presence probability threshold, the switching factor can be set to 1, defining someone is present and the lights are on, and the lighting demand value is set to the weighted sum of the intensity of human activity and the light difference compensation value.

[0050] Based on the lighting demand value Kd, corresponding basic control commands are generated and preset decisions are executed, as detailed in the table below: Kd interval Lighting action Typical scenarios Kd=0 Turn off the lights. Unmanned 0<Kd≤0.20 Maintain minimum brightness (10% of maximum brightness). Corridor night patrol 0.2<Kd≤0.5 Medium brightness (30%~50% of maximum brightness) Meeting room break 0.5<Kd≤0.8 Standard office brightness (60%~80% of maximum brightness) Office daily work Kd>0.8 High brightness mode (90%~100% of maximum brightness) Precision machining work area In other words, the basic brightness of the LED light can be controlled according to the size of the lighting demand value Kd, so that the basic brightness = lighting demand value Kd × maximum brightness value.

[0051] In reality, for people performing high-frequency visual tasks (typing / note-taking), even if the ambient light difference is only 20 lux, standard / high brightness is still required to ensure visual comfort. Insufficient base brightness leads to inadequate illumination and increased visual fatigue. The human eye's lighting needs generally follow a physiological logic of prioritizing basic supplemental lighting, followed by incremental brightness adjustments based on activity intensity. When ambient light is insufficient, it needs to be supplemented to the light compensation threshold first, meaning light difference compensation is the basic requirement. The higher the activity intensity, the higher the brightness requirement added to the basic supplemental lighting, meaning activity intensity is the incremental requirement. Therefore, basic control commands can be generated through additive coupling, using the probability of human presence as the switching factor, light difference compensation as the basic term, and activity intensity as the incremental term, to align with the physiological logic of the human eye.

[0052] For example, the lighting requirement value Kd = (light difference compensation value + activity intensity × (1 - light difference compensation value)) × switching factor. Thus, the less ambient light there is, the higher the proportion of basic supplementary lighting and the lower the proportion of incremental supplementary lighting for activity intensity; the closer the ambient light is to the threshold, the higher the proportion of incremental supplementary lighting for activity intensity, which can both ensure basic visual needs and adapt to incremental needs for activity intensity.

[0053] To avoid abrupt changes in lighting demand values, a first-order low-pass filter can be applied to the continuously sampled lighting demand values ​​to achieve a slow transition in brightness, matching the visual adaptation speed of the human eye.

[0054] S3, the cloud-edge collaborative control platform receives multi-dimensional data uploaded by each edge node, and combines it with preset lighting scene modes, user permission configurations and historical operation data to make dynamic decisions on the brightness, color temperature and on / off status of LED lighting terminals, and generate precise control commands.

[0055] Specifically, step S3 mainly involves cloud-edge collaborative intelligent decision-making. Specifically, the cloud-edge collaborative control platform receives data uploaded by each edge node, combines it with preset lighting scene modes (including but not limited to office mode, energy-saving mode, emergency mode, and customized scene modes), user permission configurations, and historical operating data, and uses an improved particle swarm optimization algorithm to dynamically decide on the brightness, color temperature, and on / off status of the LED lighting terminal, generating precise control commands. The cloud-edge collaborative control platform supports both remote access and local offline control modes; in offline mode, the edge nodes independently complete control decisions.

[0056] As a preferred technical solution, such as Figure 3 As shown, the specific method for generating precise control commands includes the following steps: S31, obtain the real-time brightness and target brightness of the LED light, obtain the comfort deviation based on the difference between the real-time brightness and the target brightness, and obtain the LED light energy consumption value based on the real-time brightness.

[0057] The energy consumption value of the LED light here can be determined by constructing a mapping function between real-time brightness and power, and then conducting experimental tests to fit the value.

[0058] S32, obtain the real-time color temperature of the LED light and the scene preset color temperature, obtain the color temperature difference based on the real-time color temperature and the scene preset color temperature, and obtain the biorhythm matching value based on the color temperature difference and the time-based ideal color temperature.

[0059] S33 obtains the objective function based on the comfort deviation, LED lamp energy consumption value and biological rhythm matching value, and makes dynamic decisions on the brightness, color temperature and on / off status of the LED lighting terminal based on the objective function and the particle swarm algorithm to generate precise control commands.

[0060] Comfort deviation can be expressed as The LED energy consumption value is represented as the sum of the energy consumption values ​​of N LEDs, and the circadian rhythm matching value is represented as the norm of the difference between the color temperature difference and the time-based ideal color temperature. Here, N represents the total number of lights managed by the edge node. In the comfort deviation, the penalty value for brightness deviation can be amplified through secondary penalty, forcing high-precision tracking.

[0061] After obtaining the comfort deviation, LED light energy consumption value, and circadian rhythm matching value, the comfort deviation, LED light energy consumption value, and circadian rhythm matching value are normalized to [0,1] dimensionless to eliminate the difference in dimensions and numerical magnitude.

[0062] For example, the objective function can be expressed as Wherein, S, E, and C represent the normalized comfort deviation, LED light energy consumption value, and biological rhythm matching value, respectively. These are the weighting coefficients.

[0063] The aforementioned biorhythm matching values ​​use a global scalar (i.e., norm) between the color temperature difference and the ideal color temperature. However, in some cases, the color temperature of each luminaire in actual lighting has a spatial distribution (e.g., the color temperature of luminaires near windows needs to be higher than that of luminaires on the inside to offset the influence of natural light). Global scalar optimization will lead to spatial uniformity of color temperature, lose regional adaptability, and reduce the actual effect of rhythm matching.

[0064] Considering the color temperature-brightness matching law of human vision, such as high color temperature at low brightness leading to visual harshness and glare, and low color temperature at high brightness leading to visual dimness and fatigue, the biorhythm matching value can be defined as the weighted sum of the color temperature deviations of individual lamps, taking into account the influence of natural light on the spatial location of the lamps, i.e., the biorhythm matching value. .in, Let $i$ represent the spatial weight of the $i$-th LED, the color temperature difference of the $i$-th LED, and the ideal rhythmic color temperature of the $i$-th LED, respectively.

[0065] For weighting coefficients Dynamic adjustments can be made by combining natural light intensity and visual operation factors related to personnel activity. Specifically, when natural light intensity > 200 lux, the energy consumption value weight can be appropriately increased (e.g., +0.3), and the comfort deviation weight can be decreased (e.g., -0.2). When personnel are engaged in high-visual-range work, the comfort deviation weight can be increased (e.g., +0.3), and the energy consumption value weight can be decreased (e.g., -0.2). Alternatively, weighting coefficients can be combined based on different scenario modes such as office mode, energy-saving mode, and emergency mode, and then the corresponding weighting coefficients can be dynamically selected according to the actual scenario mode.

[0066] S4, the cloud-edge collaborative control platform sends precise control commands to the control drive module of the corresponding LED lighting terminal, and the control drive module adjusts the working parameters of the LED light source according to the precise control commands.

[0067] Step S4 mainly involves issuing commands and providing execution feedback. Specifically, the cloud-edge collaborative control platform sends control commands to the control drive module of the corresponding LED lighting terminal through the IoT communication module. The control drive module adjusts the operating parameters of the LED light source according to the commands. At the same time, the LED lighting terminal collects its own operating status data in real time (including but not limited to brightness feedback value, color temperature feedback value, operating current and voltage, and fault information), and uploads it back to the cloud-edge collaborative control platform through the sensing terminal to form a closed-loop control.

[0068] As a preferred technical solution, the intelligent LED lighting control method further includes the following steps: using an LED lighting terminal to collect its own operating status data in real time, and then uploading this data back to the cloud-edge collaborative control platform via a sensing terminal to form a closed-loop control. Here, the operating status data includes, but is not limited to, brightness feedback values, color temperature feedback values, operating current and voltage, and fault information.

[0069] In summary, the proposed intelligent LED lighting control method performs local edge preprocessing and extracts feature parameters from collected multi-dimensional data. Simultaneously, the preprocessed multi-dimensional data and feature parameters are uploaded to a cloud-edge collaborative control platform. This platform receives the multi-dimensional data uploaded by each edge node and, combined with preset lighting scene modes, user permission configurations, and historical operating data, dynamically determines the brightness, color temperature, and on / off status of the LED lighting terminal, generating precise control commands. It can fuse data from all terminals within the area, breaking the isolation of single-point data. Through data association, complementarity, and verification, it forms a comprehensive and accurate assessment of the overall lighting needs of the area. This solves the problems of existing technologies lacking effective fusion and cross-verification of multi-terminal data within the area, being prone to misjudgments due to single-point sensor failures or data bias, and exhibiting low data fusion levels and insufficient control strategy precision.

[0070] The maximum light difference compensation value (1 - ambient light intensity / light compensation threshold, 0) is a linear function. However, the human eye's perception of illuminance follows the Weber-Fechner law (logarithmic response), which means that the human eye is more sensitive to changes in brightness in the low-illuminance range and significantly less sensitive to changes in brightness in the high-illuminance range. Linear compensation can lead to insufficient supplementary light in low-illuminance conditions, causing visual fatigue, and excessive supplementary light in high-illuminance conditions, resulting in meaningless dimming and causing light flickering.

[0071] As a preferred technical solution, the light difference compensation term can be converted into a logarithmic form to match the Weber-Fechner law, i.e., light difference compensation value = max(1 - lg ambient light intensity / lg light compensation threshold, 0). In this way, the light difference compensation value increases faster in the low-illuminance range and approaches 0 in the high-illuminance range, perfectly matching the human eye's perception of light.

[0072] In one embodiment, such as Figure 4 As shown, the intelligent lighting control method for LED lights also includes the following steps: The S5 cloud-edge collaborative control platform regularly performs adaptive optimization of control strategies based on historical operating data, user behavior habits, and environmental change trends.

[0073] The S6 cloud-edge collaborative control platform performs fault self-diagnosis based on the working status data uploaded by the LED lighting terminal. When a fault is detected, it immediately triggers an alarm mechanism and pushes maintenance prompts to the management personnel terminal, while automatically switching to the backup lighting scheme.

[0074] As a preferred technical solution, such as Figure 5 As shown, the specific method for performing fault self-diagnosis includes the following steps: S61, obtain the current harmonic distortion rate, temperature rise value, current change rate and flicker index of the LED lighting terminal, and obtain the fault feature vector based on the current harmonic distortion rate, temperature rise value, current change rate and flicker index.

[0075] S62, obtain the fault baseline vector, obtain the fault prediction probability based on the fault feature vector and the fault baseline vector, and perform fault self-diagnosis based on the fault prediction probability.

[0076] Specifically, for current harmonic distortion rate, temperature rise, current change rate, and flicker index, a dimensionless normalization (e.g., maximum deviation normalization) is first performed on the [0,1] dimension to eliminate differences in dimensions and magnitudes, ensuring that the contribution of each feature to the fault distance matches the actual fault sensitivity. For example, the fault feature vector... ,in, These are represented, in order, the normalized current harmonic distortion rate, temperature rise, current change rate, and flicker index.

[0077] The probability of fault prediction is expressed as .in, These represent the equipment aging coefficient and the fault feature vector, respectively. This represents the Euclidean distance between the fault feature vector and the fault reference vector.

[0078] Generally, when the fault prediction probability is less than the normal fault probability threshold, the equipment is considered to be in normal condition; when the fault prediction probability is not less than the normal fault probability threshold but less than the fault probability threshold, the equipment is considered to be in a warning state; when the fault prediction probability is greater than or equal to the fault probability threshold, the equipment is considered to be in a fault state, and at this time, a response strategy of red alarm, switching to backup light source, and pushing maintenance work order can be implemented.

[0079] The aging factor of equipment can be defined as follows: .in, These represent the actual operating hours and the equipment's design life (e.g., 50,000 hours), respectively. The failure probability threshold can be dynamically adjusted based on the equipment aging coefficient, i.e., the adjusted failure probability threshold = initial failure probability threshold × (1 - equipment aging coefficient). In this way, the failure threshold of aging equipment is lower and more sensitive, allowing the aging equipment to have a lower tolerance for characteristic deviations, achieving an adaptive failure detection threshold throughout its entire lifecycle.

[0080] The driver power supply of lighting equipment is a component prone to failure. Its core faults (such as rectifier bridge damage, inverter failure, and capacitor aging) not only lead to abnormal current harmonic distortion rate but also directly cause an increase in voltage harmonic distortion rate and a significant drop in power factor. Therefore, voltage harmonic distortion rate and power factor can be obtained, and then the current harmonic distortion rate, temperature rise, current change rate, flicker index, voltage harmonic distortion rate, and power factor can be normalized to a dimensionless value of [0,1]. Finally, a fault feature vector is obtained based on the normalized current harmonic distortion rate, temperature rise, current change rate, flicker index, voltage harmonic distortion rate, and power factor.

[0081] In one embodiment, such as Figure 6 As shown, an embodiment of the present invention also provides an intelligent LED lighting control system for implementing the aforementioned intelligent LED lighting control method, including a distributed sensing terminal network and a cloud-edge collaborative control platform.

[0082] The distributed sensing terminal network is deployed with several sensing terminals integrating multi-dimensional sensing modules. Among them, the sensing terminals are deployed one-to-one with LED lighting terminals. They are used to perform local edge preprocessing on the collected multi-dimensional data and extract feature parameters. The feature parameters are used to make a preliminary judgment on lighting needs and generate basic control commands. At the same time, the preprocessed multi-dimensional data and feature parameters are uploaded to the cloud-edge collaborative control platform. The multi-dimensional sensing modules are used to synchronously collect multi-dimensional data including ambient light intensity, human presence status, and frequency of human activities.

[0083] The cloud-edge collaborative control platform receives multi-dimensional data uploaded by each edge node, combines preset lighting scene modes, user permission configurations, and historical operating data to dynamically make decisions on the brightness, color temperature, and on / off status of LED lighting terminals, generate precise control commands, and send these precise control commands to the control drive module of the corresponding LED lighting terminal; the control drive module adjusts the working parameters of the LED light source according to the precise control commands.

[0084] Specifically, the cloud-edge collaborative control platform first synchronously collects three core data elements: preset lighting scene modes, user permission configurations, and historical operation data. It then performs standardized preprocessing to ensure consistent data format and dimensionality, providing high-quality input for the particle swarm optimization algorithm. This includes: 1. Preset lighting scene mode data extraction: Extract the core parameters of various scene modes built into the system to provide a constraint benchmark for subsequent algorithm optimization. Different modes correspond to clear threshold ranges and priorities for lighting parameters.

[0085] 2. User Permission Configuration Data Verification: Extract the identity information and corresponding permission levels of currently connected users to the system, and verify whether users have the necessary permissions to switch scene modes, adjust lighting parameters, and configure decision-making strategies. For example, administrators can modify scene mode parameters and optimization algorithm thresholds for all areas, ordinary users can only switch the preset scene modes for their own areas, and visitors have no permission to adjust any parameters and can only passively accept system control commands. This step ensures that the decision-making process complies with permission control requirements and avoids control disruptions caused by unauthorized operations.

[0086] 3. Historical Operation Data Processing and Feature Extraction: Collect historical operation data of LED lighting terminals within the area (including historical brightness / color temperature adjustment records, switch status change logs, fault information, and energy consumption data), historical data collected by sensing terminals (historical trends in ambient light intensity, human presence status, and frequency of human activity), and user operation history (scene mode switching records and parameter customization records). Through data cleaning and filtering of outliers, and extraction of core features, data support is provided for algorithm optimization.

[0087] 4. Data standardization processing: The three types of element data are uniformly converted into a numerical format that the algorithm can recognize, eliminating the influence of different dimensions of data and ensuring the accuracy of algorithm calculation.

[0088] Then, to address the shortcomings of traditional particle swarm optimization algorithms, such as being prone to getting trapped in local optima and having slow convergence speed, the algorithm was improved in combination with the actual needs of LED lighting control (multiple constraints and multiple objectives). The core improvements include three aspects: objective function construction, constraint setting, and inertia weight optimization, to ensure that the algorithm can quickly and accurately output the optimal control parameters.

[0089] With optimal lighting comfort, lowest energy consumption, and minimal equipment wear as the core objectives, a weighted summation objective function is constructed. Combining preset scene modes, user permissions, and device hardware limitations, constraint boundaries are set for algorithm optimization to avoid outputting invalid or illegal control parameters. Core constraints include: (1) Parameter range constraints: brightness ∈ [lower limit of scene mode brightness threshold, upper limit of scene mode brightness threshold], color temperature ∈ [lower limit of scene mode color temperature threshold, upper limit of scene mode color temperature threshold], on / off state ∈ {0,1} (forced on / off state = 1 in emergency mode); (2) Permission constraints: If the current user does not have permission to customize parameters, the decision variables must not exceed the parameter range of the preset scenario mode; (3) Equipment hardware constraints: The brightness / color temperature adjustment range shall not exceed the maximum adjustment range of the LED driver module (e.g., the single brightness adjustment range ≤ 20%, to avoid damage to the equipment by inrush current). (4) Real-time scene constraints: Combine the real-time regional data uploaded by edge nodes (such as the presence of personnel and ambient light intensity) to set the preconditions for the switch status (such as the switch status of energy saving mode = 0 in the unmanned state, and the emergency mode is not subject to this constraint).

[0090] Traditional particle swarm optimization algorithms use fixed inertia weights, which can lead to slow convergence in the early stages and getting stuck in local optima in the later stages. This invention employs an inertia weight strategy that combines linear decreasing and adaptive adjustment, as shown in the following formula: ω=ωmax-(ωmax-ωmin)×t / T+α×(Favg-Fi) / Favg Where ω is the inertia weight, ωmax and ωmin are the maximum and minimum inertia weights (values ​​of 0.9 and 0.4 respectively), t is the current iteration number, T is the maximum iteration number, α is the adaptive adjustment coefficient (value of 0.1-0.3), Favg is the average objective function of all particles, and Fi is the objective function value of a single particle. This strategy achieves large weights in the early stage (enhancing global search capability and quickly traversing the feasible solution space), small weights in the later stage (enhancing local search capability and accurately locating the optimal solution), and dynamically adjusts the weights according to the particle fitness to avoid getting trapped in local optima.

[0091] Using standardized multi-factor data as input, an improved particle swarm optimization algorithm is used for iterative calculations to output the optimal brightness, color temperature, and on / off status parameters for each LED lighting terminal. The specific calculation process is as follows: 1. Particle Initialization: Each LED lighting terminal's control parameters (brightness, color temperature, on / off status) are treated as a "particle," and all particles form a particle swarm (the number of particles is set according to the number of LED terminals in the area; for example, if there are 20 terminals in the area, the number of particles is 20). The position of each particle is initialized (i.e., the initial control parameter value). The position value is based on the parameter range of the preset scene mode and the optimal value of historical operating data (e.g., in the historical office mode, the average optimal brightness is 65%, so the initial brightness position is randomly generated around 65%). At the same time, the particle speed (the adjustment step size of the control parameter) is initialized.

[0092] 2. Fitness calculation: Substitute the position (control parameter value) of each particle into the constructed multi-objective function F(x) to calculate the fitness value of each particle (fitness value = 1 / F(x), the larger the value, the better the control parameters). At the same time, check whether the particle satisfies all constraints. The fitness value of the particle that does not satisfy the constraints is set to 0 (considered as an invalid solution and excluded from the optimization range).

[0093] 3. Individual Optimal and Global Optimal Updates: Record the optimal fitness value and corresponding position of each particle during the iteration process (individual optimal pbest), and simultaneously record the optimal fitness value and corresponding position of the entire particle swarm (global optimal gbest). If the fitness value of a particle in the current iteration is better than its historical individual optimal, then update the individual optimal of that particle; if the optimal fitness value of the current particle swarm is better than the historical global optimal, then update the global optimal.

[0094] 4. Particle position and velocity update: Update the velocity and position of each particle according to the improved inertia weight strategy.

[0095] 5. Iteration Termination Judgment: Repeat the above fitness calculation, optimal update, and position / velocity update steps until the iteration termination condition is met (the number of iterations reaches the preset maximum number of iterations T, or the global optimal fitness value remains unchanged for 5 consecutive iterations, which is considered convergence). The output global optimal position at this time is the optimal control parameters (brightness, color temperature, and on / off status) for each LED lighting terminal in that area.

[0096] To ensure that the output control parameters meet the requirements of the actual application scenario and system stability, the optimal control parameters output by the algorithm are verified and corrected a second time to avoid control errors caused by algorithm calculation deviations or sudden changes in the real-time scenario. 1. Real-time Scene Matching Verification: Combining real-time regional data uploaded by edge nodes (current ambient light intensity, personnel presence status, and device operating status), the algorithm verifies whether the optimal control parameters are suitable for the real-time scene. For example, if the algorithm outputs a terminal brightness of 70%, but the real-time ambient light intensity has reached 500 lux (exceeding the optimal light intensity range for office mode), the brightness is corrected to 40%. If personnel are detected in the area in real time, and the algorithm outputs a switch status of 0 (off), the status is forcibly corrected to 1 (on).

[0097] 2. Secondary Permission Verification: Verifies whether the adjustment range of the current optimal control parameter is within the current user's permission range. For example, if a regular user switches to office mode and the algorithm outputs a brightness adjustment range that exceeds the preset range (e.g., adjusting from 50% to 85%, exceeding the regular user's maximum adjustment permission of 20%), then the brightness adjustment range is corrected to the permission range (e.g., adjusting from 50% to 70%), and the permission verification log is recorded.

[0098] 3. Equipment compatibility verification: Verify whether the optimal control parameters are compatible with the hardware of the LED lighting terminal and the driver module (e.g., whether the color temperature adjustment value is within the support range of the driver module, and whether the brightness adjustment step size is within the hardware's capacity). Incompatible parameters are corrected to the compatible range (e.g., if the driver module supports a maximum color temperature of 5500K, but the algorithm outputs 6000K, then it is corrected to 5500K).

[0099] The optimal control parameters, after secondary verification and correction, are converted into standardized control commands by the cloud-edge collaborative control platform (the command format is adapted to the control driver module of the LED terminal, supporting LoRa / NB-IoT dual-mode communication transmission). These commands are then precisely distributed to the corresponding LED lighting terminal's control driver module based on regional division and device identification. The control commands explicitly include: a unique terminal identification code, brightness adjustment value, color temperature adjustment value, on / off status command, and execution time (e.g., immediate execution or delayed execution). Simultaneously, the decision-making results (control command content, decision basis, verification records) are associated and stored with historical operating data, providing data support for subsequent algorithm optimization and adaptive adjustments, forming a closed-loop control logic of "decision-execution-feedback-optimization."

[0100] The cloud-edge collaborative control platform is also used to periodically adaptively optimize control strategies based on historical operating data, user behavior habits, and environmental change trends. It also performs fault self-diagnosis based on the working status data uploaded by LED lighting terminals. When a fault is detected, an alarm mechanism is immediately triggered, and maintenance prompts are pushed to the management personnel terminal. At the same time, it automatically switches to the backup lighting scheme.

[0101] The sensing terminal also integrates an IoT communication module, which adopts a dual-mode adaptive communication method of LoRa and NB-IoT, automatically switching the communication mode according to the communication distance and network signal strength. Specifically, it first obtains the current signal-to-noise ratio (SNR), the minimum SNR threshold, and the maximum SNR threshold. Based on these three thresholds, it obtains the SNR quality factor. When the current SNR is less than the minimum SNR threshold, the SNR quality factor is set to 0; when the current SNR is greater than the maximum SNR threshold, the SNR quality factor is set to 1; when the current SNR is greater than or equal to the minimum SNR threshold and less than or equal to the maximum SNR threshold, the SNR quality factor is calculated as (current SNR - minimum SNR threshold) / (maximum SNR threshold - minimum SNR threshold). In other words, the higher the current SNR, the closer the SNR quality factor is to 1.

[0102] Then, obtain the current packet loss rate and the maximum packet loss rate. Based on the current packet loss rate and the maximum packet loss rate, obtain the packet loss rate quality factor. When the current packet loss rate is greater than the maximum packet loss rate, the packet loss rate quality factor is set to 0. Otherwise, the packet loss rate quality factor = 1 - current packet loss rate / maximum packet loss rate. That is, the lower the packet loss rate, the higher the packet loss rate quality factor, and the closer it is to 1.

[0103] Finally, the channel quality index Q is obtained based on the weighted value of the signal-to-noise ratio quality factor and the packet loss rate quality factor. The communication mode is automatically switched according to the communication distance and network signal strength. When the channel quality index is greater than 0.8 and the distance is less than 1km, the LoRa mode is adopted. When the channel quality index is not greater than 0.8 or the distance is not less than 1km, the NB-IoT mode is adopted.

[0104] Alternatively, different decisions can be made based on the channel quality index. For example, when the channel quality index is greater than 0.8, a high-speed communication mode (such as 5G carrier aggregation) can be enabled; when 0.4 ≤ channel quality index ≤ 0.8, the current communication mode can be maintained; when the channel quality index is less than 0.4, a robust mode can be switched (such as downgrading to 2G / 3G).

[0105] In practical deployments, to buffer transient jitter, the channel quality index can be nonlinearly compensated. The compensated channel quality index = the channel quality index before compensation + historical Q value fluctuation amplitude × exp(-attenuation coefficient × t). The default value of the attenuation coefficient is 0.1 / second to suppress false handovers caused by short-term interference.

[0106] Basic control commands are generated by local edge computing at the sensing terminal, primarily based on real-time environmental data for rapid response, to meet basic lighting needs. For example, lights are immediately turned off when the corridor is empty, and the basic brightness is quickly turned on to 50% to meet emergency lighting needs when someone suddenly enters the meeting room.

[0107] Precise control commands are generated by the cloud-edge collaborative control platform. Its core function is to dynamically optimize by integrating global data to achieve a multi-objective balance of comfort, energy efficiency, and health. For example, for cross-regional office areas, the brightness in the east area is 68% + 4800K (active work), and the brightness in the west area is 42% + 3500K (video conferencing). In night mode, it automatically switches to 3000K color temperature + 20% brightness to follow the circadian rhythm.

[0108] Generally speaking, when commands conflict, the command is executed according to priority: safety commands (emergency mode) > precise commands > basic commands.

[0109] As a preferred technical solution, the LED lamp intelligent lighting control system shown includes a distributed sensing terminal network, a cloud-edge collaborative control platform, LED lighting terminals, and a control drive module; The distributed sensing terminal network consists of several sensing terminals. Each sensing terminal includes a multi-dimensional sensing module, a low-power IoT dual-mode communication module, and an edge preprocessing unit. The multi-dimensional sensing module is used to collect environmental and personnel-related data, the edge preprocessing unit is used for data filtering and feature extraction, and the dual-mode communication module is used for data uploading and command receiving. The cloud-edge collaborative control platform includes an edge node cluster and a cloud management platform. The edge node cluster is used for regional data fusion analysis and offline control decision-making, while the cloud management platform is used for global data management, scene mode configuration, remote control, adaptive optimization, and fault early warning. The control drive module is electrically connected to the LED lighting terminal and is used to receive control commands and adjust the working parameters of the LED lighting terminal, while collecting the working status data of the LED lighting terminal. The sensing terminal, cloud-edge collaborative control platform and control drive module realize bidirectional data interaction through the Internet of Things communication network, and support hierarchical management of device access permissions and encrypted data transmission.

[0110] In summary, the intelligent LED lighting control system of the present invention has the following technical advantages: 1. Innovative communication architecture: It adopts LoRa and NB-IoT dual-mode adaptive communication. Through this dual-mode switching mechanism, it can solve the problem of stable communication under different distances and signal strengths, and improve the system compatibility and reliability.

[0111] 2. Sensing and Control Synergy: Each sensing terminal corresponds to an LED terminal and has a unique built-in identifier, enabling precise positioning and independent control. Combined with cloud-edge collaboration, it ensures control continuity in offline states.

[0112] 3. Full-process closed-loop control: Covering the entire process from data acquisition to edge processing, cloud-edge decision-making, execution feedback, and self-diagnosis and maintenance, it can realize fault self-diagnosis, backup lighting switching, and adaptive optimization, thereby improving system stability and maintenance efficiency.

[0113] 4. Intelligent decision optimization: By using the particle swarm optimization algorithm and combining it with multi-scenario mode configuration, it is beneficial to optimize the fusion analysis of multi-dimensional sensor data, thereby improving the accuracy of decision-making and the adaptability of scenarios.

[0114] The technical features of the embodiments described can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0115] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for intelligent lighting control of LED lamps, characterized in that, The intelligent lighting control method for LED lamps includes the following steps: A number of sensing terminals integrating multi-dimensional sensing modules are deployed, with each sensing terminal corresponding to an LED lighting terminal. The multi-dimensional sensing modules are used to synchronously collect multi-dimensional data, including ambient light intensity, human presence status, and frequency of human activity. The sensing terminal performs local edge preprocessing on the collected multi-dimensional data and extracts feature parameters. Based on the feature parameters, it makes a preliminary judgment on lighting needs and generates basic control commands. At the same time, the preprocessed multi-dimensional data and feature parameters are uploaded to the cloud-edge collaborative control platform. The cloud-edge collaborative control platform receives multi-dimensional data uploaded by each edge node, and combines it with preset lighting scene modes, user permission configurations and historical operation data to make dynamic decisions on the brightness, color temperature and on / off status of LED lighting terminals, and generate precise control commands. The cloud-edge collaborative control platform sends precise control commands to the control and drive modules of the corresponding LED lighting terminals, and the control and drive modules adjust the working parameters of the LED light source according to the precise control commands.

2. The intelligent lighting control method for LED lamps as described in claim 1, characterized in that, The intelligent lighting control method for LED lamps also includes the following steps: The LED lighting terminal collects its own working status data in real time and uploads it back to the cloud-edge collaborative control platform through the sensing terminal to form a closed-loop control.

3. The intelligent lighting control method for LED lamps as described in claim 2, characterized in that, The intelligent lighting control method for LED lights also includes the following steps: The cloud-edge collaborative control platform regularly performs adaptive optimization of control strategies based on historical operational data, user behavior habits, and environmental change trends.

4. The intelligent lighting control method for LED lamps as described in claim 3, characterized in that, The intelligent lighting control method for LED lights also includes the following steps: The cloud-edge collaborative control platform performs fault self-diagnosis based on the working status data uploaded by the LED lighting terminals. When a fault is detected, an alarm mechanism is immediately triggered, and maintenance prompts are pushed to the management personnel terminal. At the same time, the platform automatically switches to the backup lighting scheme.

5. The intelligent lighting control method for LED lamps as described in claim 4, characterized in that, The specific method for initially determining lighting needs and generating basic control commands based on characteristic parameters includes the following steps: Obtain the maximum activity frequency threshold, and then determine the activity intensity based on the activity frequency and the maximum activity frequency threshold. Obtain the preset light compensation threshold, and then obtain the light difference compensation value based on the ambient light intensity and the light compensation threshold. The switching factor is obtained based on the human body's state and a preset probability threshold. Basic control commands are generated based on the switching factor, the intensity of human activity, and the light difference compensation value.

6. The intelligent lighting control method for LED lamps as described in claim 5, characterized in that, The specific method for generating precise control commands includes the following steps: The system obtains the real-time brightness and target brightness of the LED light, calculates the comfort deviation based on the difference between the real-time brightness and the target brightness, and calculates the energy consumption value of the LED light based on the real-time brightness. The system obtains the real-time color temperature of the LED light and the preset color temperature of the scene, obtains the color temperature difference based on the real-time color temperature and the preset color temperature of the scene, and obtains the biorhythm matching value based on the color temperature difference and the ideal color temperature based on time. The objective function is obtained based on the comfort deviation, LED lamp energy consumption value, and biological rhythm matching value. Based on the objective function and the particle swarm algorithm, dynamic decisions are made on the brightness, color temperature, and on / off status of the LED lighting terminal to generate precise control commands.

7. The intelligent lighting control method for LED lamps as described in claim 6, characterized in that, The specific methods for performing fault self-diagnosis include the following steps: Obtain the current harmonic distortion rate, temperature rise value, current change rate and flicker index of the LED lighting terminal, and obtain the fault feature vector based on the current harmonic distortion rate, temperature rise value, current change rate and flicker index; Obtain the fault baseline vector, obtain the fault prediction probability based on the fault feature vector and the fault baseline vector, and perform fault self-diagnosis based on the fault prediction probability.

8. An intelligent lighting control system for LED lamps, used to implement the intelligent lighting control method for LED lamps as described in any one of claims 1-7, characterized in that, The LED intelligent lighting control system includes: A distributed sensing terminal network, which deploys several sensing terminals that integrate multi-dimensional sensing modules; Among them, the sensing terminal and the LED lighting terminal are deployed one-to-one. The sensing terminal is used to perform local edge preprocessing on the collected multi-dimensional data and extract feature parameters. The lighting needs are initially judged by the feature parameters and basic control commands are generated. At the same time, the preprocessed multi-dimensional data and feature parameters are uploaded to the cloud-edge collaborative control platform. The multi-dimensional sensing module is used to synchronously collect multi-dimensional data including ambient light intensity, human presence status, and frequency of human activities. The cloud-edge collaborative control platform is used to receive multi-dimensional data uploaded by each edge node, and combine it with preset lighting scene modes, user permission configurations and historical operation data to make dynamic decisions on the brightness, color temperature and on / off status of LED lighting terminals, generate precise control commands, and send the precise control commands to the control drive module of the corresponding LED lighting terminal. The control and drive module adjusts the operating parameters of the LED light source according to precise control commands.

9. The intelligent lighting control system for LED lamps as described in claim 8, characterized in that, The cloud-edge collaborative control platform is also used to periodically adaptively optimize control strategies based on historical operating data, user behavior habits, and environmental change trends. It also performs fault self-diagnosis based on the working status data uploaded by LED lighting terminals. When a fault is detected, an alarm mechanism is immediately triggered, and maintenance prompts are pushed to the management personnel terminal. At the same time, it automatically switches to the backup lighting scheme.

10. The intelligent lighting control system for LED lamps as described in claim 9, characterized in that, The sensing terminal also integrates an IoT communication module, which adopts a dual-mode adaptive communication method of LoRa and NB-IoT, and automatically switches the communication mode according to the communication distance and network signal strength.

Images