Solar advertising lamp box intelligent light control system

Abstract

This invention belongs to the field of automatic lighting control technology and discloses an intelligent light control system for solar-powered advertising light boxes. It utilizes four sets of distributed brightness sensors symmetrically arranged at the four corners of the light box, combined with directional spectral sensors to filter non-visible light interference. Through a combination of multiple algorithms, transient light interference is eliminated, and temperature compensation and reverse calibration of drive control parameters are used to output ambient light characteristic parameters. Sensing and control form a closed loop, ensuring stable light output and simultaneously extending the lifespan of LED light sources and circuit components. A deep learning algorithm segments the key advertising information area from the background area, extracting features such as color and contrast, and establishing a four-dimensional linkage model of ambient light, power, visibility, and drive control parameters. This model combines real-time ambient light to calculate the minimum drive power and zone brightness weights. Brightness is allocated differently according to advertising content, reducing light energy consumption. Real-time data collection of photovoltaic power generation, energy storage battery SOC, SOH, and other operating parameters is also performed.

Description

Technical Field

[0001] This invention belongs to the field of automatic lighting control technology, specifically an intelligent light control system for solar-powered advertising light boxes. Background Technology

[0002] Solar-powered advertising light boxes have been widely used in outdoor settings such as municipal bus stops, commercial streets, community parks, and transportation hubs. As the core component of solar-powered advertising light boxes, the light control system directly determines the advertising display effect, energy consumption level, and the lifespan of the light source and energy storage battery. Existing solar-powered advertising light box light control solutions are severely out of sync with the needs of electronic control and intelligent dimming, resulting in the following technical problems: Conventional solutions often use a single photoresistor with a fixed brightness threshold, which can only achieve basic on / off control. They cannot adapt to complex outdoor lighting environments such as rain, tree shade, and transient light pollution at night. They are prone to problems such as false triggering and flickering, resulting in wasted energy and shortened light source lifespan. They also lack a perception-control closed-loop design with the lighting drive circuit.

[0003] The existing solution only performs general dimming for ambient light, without allocating brightness according to the color, contrast and key information area distribution of the advertising image. Indiscriminate dimming results in ineffective energy consumption, and the core information of the advertisement cannot be highlighted. It does not incorporate the concept of zone dimming and driving, and has poor synergy with the driving module.

[0004] The existing solution does not link the light control strategy with the photovoltaic power generation and the status of the energy storage battery. Under extreme weather conditions, the battery will still operate at full power when it is depleted, which can easily lead to over-discharge and damage to the battery. Furthermore, it cannot predict and optimize the lighting strategy based on the photovoltaic power generation, resulting in low energy storage utilization, insufficient battery life, and deviating from the core of lighting electronic control technology. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent light control system for solar-powered advertising light boxes to solve one or more of the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent light control system for solar-powered advertising light boxes, comprising the following modules: Furthermore, the light sensing module includes four sets of distributed brightness sensors, one set of directional spectral sensors with infrared cutoff and visible light narrowband filtering, an ambient temperature and humidity acquisition unit, a data preprocessing subunit, and a dimming drive feedback calibration subunit. The distributed brightness sensor synchronously collects ambient light data around the lightbox, and the directional spectral sensor filters out non-visible light interference, including vehicle lights and infrared radiation, and collects the brightness and spectral distribution data of effective visible light. The data preprocessing subunit uses an algorithm that combines moving average filtering, transient interference removal and AI adaptive threshold iteration to filter transient light interference, and at the same time, it combines ambient temperature to complete sensor temperature drift compensation. The dimming drive feedback calibration subunit collects the output current and voltage parameters of the dimming drive module in real time, reverse-calibrates the ambient light sensing data, and finally outputs the ambient light characteristic parameters, which are synchronously transmitted to the energy storage linkage module and the dimming drive module.

[0007] Furthermore, the content adaptation module includes an advertising screen feature storage unit, an image segmentation and feature extraction subunit, a visibility threshold modeling subunit, a brightness requirement calculation subunit, and a zone dimming drive adaptation subunit. The advertising images projected on the lightbox are preprocessed. Using a deep learning-based image segmentation algorithm, key information areas, including brand logos and core text, and non-key background areas are identified. The RGB color features, grayscale contrast, and minimum visible brightness threshold of each area are extracted. A four-dimensional linkage model is established, which includes ambient light brightness, light source output power, advertising content visibility compliance rate, and dimming drive control parameters. Based on the real-time parameters output by the ambient light sensing module, the minimum total output power of the light source that meets the visibility requirements of the advertising information area is calculated, and the brightness allocation weight parameters corresponding to each light strip channel of the light box are generated. The partitioned dimming drive adaptation subunit converts the brightness allocation parameters into drive control parameters and transmits them synchronously to the content adaptation module and the energy storage linkage module.

[0008] Furthermore, the energy storage linkage module includes a photovoltaic parameter acquisition unit, an energy storage battery status acquisition unit, an energy storage safety constraint modeling subunit, a multi-constraint light control decision subunit, and an energy storage and dimming drive collaborative control subunit; The photovoltaic parameter acquisition unit collects real-time power generation, daily cumulative power generation, and photovoltaic conversion efficiency of the photovoltaic module; the energy storage battery status acquisition unit collects real-time state of charge (SOC), state of health (SOH), battery temperature, and charge / discharge cycle count of the energy storage lithium battery; the energy storage safety constraint modeling subunit establishes a dynamic constraint model for battery over-discharge protection and cycle life optimization; the multi-constraint light control decision subunit integrates advertising content visibility requirements, spatiotemporal scene constraint rules, and energy storage safety constraint model to generate preliminary multi-channel dimming control parameters. The energy storage and dimming drive collaborative control subunit dynamically adjusts the drive control parameters of the dimming drive according to the energy storage battery status, and at the same time feeds back the light energy consumption data to the energy storage management unit, realizing the linkage of photovoltaic power generation prediction, energy storage status monitoring, light control strategy optimization, and dimming drive adjustment. Finally, the generated multi-channel dimming control parameters are sent to the dimming drive module.

[0009] After the light energy consumption data is fed back to the energy storage management unit, the energy storage management unit calculates the current energy consumption rate and remaining battery life in real time, and dynamically adjusts the charging priority and discharge limitation strategy. When the energy consumption is high, the charging efficiency is automatically optimized, and when the energy consumption is low, the brightness limitation is appropriately relaxed, so that the energy storage power allocation matches the actual energy consumption of the light box, maximizing the energy storage utilization efficiency and the equipment's battery life.

[0010] Furthermore, the dimming drive module includes a multi-channel constant current drive chipset, a linear and PWM dual-mode dimming switching unit, an LED lamp bead junction temperature detection unit, a drive protection subunit, and an AI adaptive drive control parameter calibration subunit. The system receives multi-channel dimming control parameters from the multi-constraint light control decision subunit and implements independent constant current drive for multiple backlight and fill light strips of the advertising light box. Based on the target brightness parameters, energy storage status, and LED junction temperature, it automatically switches dimming modes: linear dimming mode for low brightness and PWM dimming mode for high brightness. The LED junction temperature detection unit collects the operating temperature of the LEDs in real time, and automatically reduces the drive current when the junction temperature exceeds a preset threshold. The drive protection subunit monitors overcurrent, overvoltage, short circuit, and open circuit faults in real time and triggers corresponding protection actions. The AI ​​adaptive drive control parameter calibration subunit calibrates the output current, voltage, and PWM duty cycle of the drive chip in real time based on real-time changes in ambient light, advertising content requirements, and energy storage status, while recording the change data of drive control parameters.

[0011] Furthermore, the self-diagnosis and correction module includes a sensor self-correction subunit, a light source light decay compensation subunit, and a drive circuit self-diagnosis subunit; The sensor self-calibration subunit has a built-in standardized calibration process, which performs zero-point calibration regularly to correct the dark current and temperature drift deviation of the photosensitive sensor, and performs gain calibration regularly to correct the sampling gain deviation caused by factors such as dust obstruction and component aging. The light source light decay compensation subunit periodically collects the actual luminous flux feedback value of the light source and compares it with the factory reference parameters. When the light decay of the LED bead is detected to exceed the preset threshold, the driving current is automatically increased to compensate for the light decay. The self-diagnostic subunit of the drive circuit monitors the circuit parameters of the dimming drive module in real time, identifies potential faults in the drive circuit, provides early warnings and adjusts the drive control parameters, and feeds back the self-diagnostic data to the remote management module.

[0012] Furthermore, the remote control module includes an edge computing processing unit, a low-power wireless communication subunit, a local data storage unit, a cloud platform interaction subunit, and a remote parameter configuration subunit; Employing an edge computing architecture, the edge computing processing unit completes all real-time sensing, decision-making, and dimming control actions locally; and periodically uploads the system's operating data to the cloud management platform through the low-power wireless communication subunit. The edge computing is responsible for local real-time control, and only uploads information to the cloud when the network is normal and the data is normal. The cloud only issues intervention commands when the device is offline for a timeout, the parameters are abnormal, or the faults occur frequently. In daily operation, the local independent control is the main function, and the cloud only performs data statistics, parameter optimization, and batch configuration. The two work together without conflict. When the network is interrupted, the edge computing processing unit runs completely autonomously, and the data is automatically synchronized after the network is restored.

[0013] The local data storage unit is used to locally cache running data, calibration parameters, and fault records.

[0014] The remote parameter configuration subunit supports the cloud platform to issue dimming drive-related drive control parameters, and also supports unified configuration and iterative optimization of dimming drive control parameters for batch light boxes; it realizes unified operation and maintenance management of batch light boxes, and the cloud platform uses big data analysis to iteratively optimize the light control model and drive control parameters of light boxes in different scenarios and regions.

[0015] The present invention also provides a solar-powered advertising lightbox, based on the above-mentioned system, including an advertising display component, a photovoltaic power supply component, an energy storage component, and an intelligent control component; the intelligent control component integrates six modules: light sensing, content adaptation, energy storage linkage, dimming drive, self-diagnosis and correction, and remote control, and each module works together to achieve intelligent operation of the lightbox; The photovoltaic power supply module directly charges and manages the energy storage module through a dedicated charging line. The energy storage module, as the core of the whole machine's power supply, provides a stable and uninterrupted power supply to the intelligent control module and the advertising display module. The intelligent control module achieves a full-link connection with the LED light strip, various sensors, and drive circuits of the advertising display module through internal data signal lines. All control commands, collected data, and feedback signals are transmitted at high speed through the internal bus. The power supply line and the signal transmission line are independent and isolated from each other, and do not interfere with each other.

[0016] The photovoltaic power supply component is used to collect solar energy and convert it into electrical energy; the energy storage component is used to store electrical energy and power the advertising box; and the advertising display component is used to display advertising content. The light sensing module is used to collect ambient light data around the advertising box, filter out non-visible light and transient light interference, complete temperature drift compensation and calibrate the sensing data, and output ambient light characteristic parameters. The content adaptation module is used to preprocess the advertisement image, identify key information areas of the advertisement and extract image features, calculate brightness requirements and allocation parameters in combination with ambient light feature parameters, and convert them into dimming drive signals. The energy storage linkage module is used to collect the operating parameters of the photovoltaic power supply module and the energy storage module, establish a safety constraint model, generate dimming control parameters by integrating multiple factors, and dynamically adjust the drive control parameters in combination with the energy storage status. The dimming drive module is used to receive dimming control parameters, realize independent driving of the advertising display components for each channel, automatically switch dimming modes, detect the status of LED beads and circuit faults and trigger protection actions, and calibrate the drive control parameters at the same time. The self-diagnosis and correction module is used to calibrate the sampling deviation of the light sensing module, compensate for the light decay of the advertising display component, monitor potential faults in the driving circuit, and provide early warning and adjustment. The remote control module adopts an edge computing architecture to achieve local real-time control and uploading of operational data, and supports remote control, batch operation and maintenance, and parameter optimization in the cloud.

[0017] The beneficial effects of this invention are as follows: 1. This invention uses four sets of distributed brightness sensors symmetrically arranged at the four corners of the light box, combined with directional spectral sensors to filter non-visible light interference; it eliminates transient light interference through a combination of multiple algorithms, and outputs ambient light characteristic parameters by combining temperature compensation and reverse calibration of drive control parameters; the perception and control form a closed loop to ensure stable light output and extend the service life of LED light source and circuit components.

[0018] 2. This invention uses deep learning algorithms to segment the key information area and background area of ​​the advertisement, extracts features such as color and contrast, and establishes a four-dimensional linkage model of ambient light, power, visibility, and drive control parameters. It combines real-time ambient light to calculate the minimum drive power and the brightness weight of each zone; and allocates brightness according to the advertising content to reduce the energy consumption of the lights.

[0019] 3. This invention establishes a dynamic constraint model for battery over-discharge protection and lifespan optimization by real-time collection of photovoltaic power generation, energy storage battery SOC, SOH, and other operating parameters. Based on the energy storage status, it dynamically adjusts the dimming parameters and lighting plan, realizing the coordinated linkage of power generation prediction, energy storage monitoring, light control optimization, and drive adjustment. This not only avoids battery over-discharge damage and improves energy storage utilization and equipment endurance, but also extends the cycle life of the energy storage battery, enabling the solar advertising light box to operate stably in outdoor scenarios without external power, and reducing subsequent operation and maintenance costs. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the overall workflow of the system of the present invention; Figure 2This is a flowchart illustrating the intelligent light control dimming process of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] like Figures 1 to 2 As shown, this embodiment of the invention provides an intelligent light control system for solar-powered advertising light boxes, comprising the following modules: In this embodiment of the invention, the light sensing module includes four sets of distributed brightness sensors, one set of directional spectral sensors with infrared cutoff and visible light narrowband filtering, an ambient temperature and humidity acquisition unit, a data preprocessing subunit, and a dimming drive feedback calibration subunit. The distributed brightness sensors are symmetrically arranged 5cm from the edges of the light box, and the sensor acquisition angle covers a 120° range on the front of the light box, ensuring no blind spots. The directional spectral sensor has a built-in 850nm infrared cutoff filter and a 400-760nm visible light narrowband filter, which can completely filter out non-visible light interference such as infrared and ultraviolet light from the vehicle headlights, retaining only the effective visible light signal. The temperature drift compensation adopts a segmented linear compensation method, dividing the full temperature range from -20℃ to 60℃ into 10 temperature segments, each temperature segment is matched with a dedicated compensation coefficient to eliminate sampling deviations caused by temperature changes.

[0023] The distributed brightness sensor synchronously collects ambient light data around the lightbox and is deployed at the four corners of the lightbox to eliminate the influence of local occlusion. The directional spectral sensor filters out non-visible light interference, including stray light from vehicle lights and infrared radiation, and collects the brightness and spectral distribution data of effective visible light. The data preprocessing subunit uses an algorithm that combines moving average filtering, transient interference removal, and AI adaptive threshold iteration to filter transient light interference such as lightning and passing vehicle lights with a duration of less than 500ms. At the same time, it combines ambient temperature to complete sensor temperature drift compensation. The moving average filter continuously collects 10 sets of raw ambient light sampling data, smooths data fluctuations through mean calculation, and automatically removes abnormal sampling values ​​that deviate too much from the mean, avoiding data distortion caused by local strong light interference. The transient interference removal module monitors the duration of light signals in real time and judges short-term transient light signals such as lightning, car lights, and flashlights with a duration of less than 500ms as invalid interference, directly blocking them from participating in subsequent ambient light parameter calculations. The AI ​​adaptive threshold iteration algorithm combines historical ambient light change patterns, current scene illumination characteristics, and temperature and humidity conditions to correct filtering parameters and interference judgment criteria in real time. The three algorithms run in sequence and in tandem: first filtering and smoothing, then removing transient interference, and finally adaptively calibrating the threshold.

[0024] The AI ​​adaptive threshold iteration algorithm adopts a lightweight structure of a single hidden layer feedforward neural network, which includes an input layer, a 128-neuron hidden layer, and an output layer. It is designed for complex outdoor lighting environments and can effectively resist short-term transient light interference from vehicle lights, lightning, flashlights, etc. The specific algorithm training and execution process is as follows: 1. Collect ambient light sampling data for typical outdoor scenarios such as rain, fog, haze, tree shade, nighttime light pollution, and direct sunlight, and construct a training dataset containing 10,000 valid samples; 2. The algorithm takes the original sampled values ​​of the distributed brightness sensor, the visible light data of the directional spectrum sensor, the ambient temperature and humidity data, and the real-time voltage and current feedback values ​​of the dimming drive module as inputs, and the ambient light stability judgment threshold as the algorithm output. 3. Using mean squared error as the loss function, the network weights are iteratively updated using mini-batch gradient descent, with an adaptive threshold calibration performed every 5 minutes; 4. Set the threshold update step size to 0.5 lux, identify transient optical signals with a duration of less than 500 ms as interference and remove them directly, and combine the ambient temperature data to complete sensor temperature drift compensation.

[0025] The dimming drive feedback calibration subunit collects the output current and voltage parameters of the dimming drive module in real time, reverse-calibrates the ambient light sensing data, and finally outputs ambient light characteristic parameters such as effective visible light brightness, illumination uniformity, and spectral distribution characteristics, which are synchronously transmitted to the content adaptation module and the multi-constraint light control decision subunit.

[0026] The reverse calibration process uses the actual output electrical parameters of the dimming drive module as a reference. Every 100ms, a set of drive current and voltage feedback values ​​are acquired and compared with the theoretical drive control parameters corresponding to the current ambient light sensing parameters to automatically correct the sensor acquisition deviation. The calibration process is continuously executed in a closed loop. When the deviation between the sensing data and the feedback data exceeds the allowable range, the sensor calibration coefficient is immediately updated to ensure that the ambient light sensing results match the actual light emission state and drive output state of the light box.

[0027] The ambient light characteristic parameters are output in a standardized digital format, including three core data types: real-time ambient light brightness value, multi-directional illumination uniformity deviation value, and visible light center wavelength and spectral intensity value. The parameter output cycle is consistent with the sensor acquisition cycle. The output data is synchronously transmitted to the content adaptation module and energy storage linkage module through the internal high-speed bus. The data transmission process has a verification mechanism to ensure that there is no loss or error in the transmission.

[0028] In this embodiment of the invention, the content adaptation module includes an advertising screen feature storage unit, an image segmentation and feature extraction subunit, a visibility threshold modeling subunit, a brightness requirement calculation subunit, and a zone dimming drive adaptation subunit. The advertising images projected on the lightbox are preprocessed. Using a deep learning-based image segmentation algorithm, key information areas such as brand logos and core text are identified from non-key background areas. The RGB color features, grayscale contrast, and minimum visible brightness threshold of each area are extracted. A four-dimensional linkage model is established, which includes ambient light brightness, light source output power, advertising content visibility compliance rate, and dimming drive control parameters. The minimum visible brightness threshold is determined by classifying factors such as color depth, text size, pattern complexity, and viewing distance of the advertisement image. The smaller the text size, the weaker the color contrast, and the richer the pattern details, the higher the corresponding minimum visible brightness threshold is set. Non-critical areas such as solid color backgrounds and decorative patterns have the lowest threshold, while critical information areas such as brand logos, core copy, and product main body have the highest threshold. All threshold parameters are pre-entered into the advertisement image feature storage unit and directly called for matching during dimming control.

[0029] The four-dimensional linkage model uses ambient light brightness as the basic input variable and the visibility of advertising content as the core constraint. It bidirectionally binds the output power of the light source with the dimming drive control parameters. For every 1000 lux increase in ambient light brightness, the output power of the light source is increased by a corresponding proportion according to the weight of the key information area. For every 1000 lux decrease in ambient light brightness, the output power of the light source is decreased by a corresponding proportion according to the lowest visibility threshold. The dimming drive control parameters are synchronously matched with the power change, and the model parameters can be dynamically adjusted in real time according to different advertising screen types.

[0030] The deep learning image segmentation model adopts an Encoder-Decoder symmetric architecture, using a lightweight ResNet50 as the backbone for Encoder feature extraction, and a U-Net-structured Decoder to achieve pixel-level region reconstruction. The network includes a 5-layer downsampling convolution module, a 4-layer upsampling deconvolution module, and a 2-layer feature fusion and stitching module, which can adapt to the fast segmentation requirements of high-resolution images of outdoor advertising light boxes.

[0031] Model training follows a standardized process: 1. Construct a sample set of no less than 5,000 outdoor advertising images, covering various types such as text, logos, patterns, and solid color backgrounds, with professional personnel completing pixel-level annotations of key information areas and non-key background areas; 2. Perform data augmentation processing on the sample set, such as random cropping, brightness perturbation, and contrast adjustment, to improve the model's generalization ability in complex outdoor scenes; 3. The model input is RGB three-channel pixel data and grayscale distribution data of the advertisement image, and the output label is a binary classification segmentation mask of key information / background. 4. The cross-entropy loss function and the Dice loss function are used for joint optimization. The initial learning rate is set to 0.001, the batch size is 16, and the number of training iterations is no less than 1000. Training is stopped when the model segmentation accuracy reaches 95% or more. 5. During the model inference stage, the system receives real-time input from the advertising screen and quickly outputs screen feature parameters such as region segmentation results, color feature values, grayscale contrast, and minimum visible brightness threshold.

[0032] Based on the real-time parameters output by the ambient light sensing module, the minimum total output power of the light source that meets the visibility requirements of the advertising information area is calculated, and the brightness allocation weight parameters corresponding to each light strip channel of the light box are generated. The partitioned dimming drive adaptation subunit converts the brightness allocation parameters into drive control parameters including constant current drive current, PWM duty cycle and other parameters, and transmits them synchronously to the multi-constraint light control decision subunit and the dimming drive module.

[0033] The parameter conversion is based on the hardware specifications of the light box driver chip. Logical parameters such as zone brightness weight and minimum drive power are converted one by one into constant current drive current value, PWM signal duty cycle value and dimming mode switching command that the driver chip can directly execute. The conversion process automatically matches the rated range of each channel driver hardware and does not exceed the hardware safe operation limit. After conversion, the parameters are directly sent to the dimming driver module.

[0034] The brightness weighting of each zone is based on the visibility priority of the key information area of ​​the advertisement. The brand logo and core copywriting area are set to the highest weight, the product display area is set to the medium weight, and the background decoration and blank area are set to the lowest weight. The weighting is dynamically adjusted in conjunction with the ambient light brightness. The stronger the ambient light, the higher the weight of the key area. The weaker the ambient light, the lower the weight of the background area. All weighting calculations are based on the premise of meeting the visibility of the advertisement.

[0035] In this embodiment of the invention, the energy storage linkage module includes a photovoltaic parameter acquisition unit, an energy storage battery status acquisition unit, an energy storage safety constraint modeling subunit, a multi-constraint light control decision subunit, and an energy storage and dimming drive collaborative control subunit; The photovoltaic parameter acquisition unit collects real-time power generation, daily cumulative power generation, and photovoltaic conversion efficiency of the photovoltaic module; the energy storage battery status acquisition unit collects real-time state of charge (SOC), state of health (SOH), battery temperature, and charge / discharge cycle count of the energy storage lithium battery; the energy storage safety constraint modeling subunit establishes a dynamic constraint model for battery over-discharge protection and cycle life optimization; the multi-constraint light control decision subunit integrates advertising content visibility requirements, spatiotemporal scene constraint rules, and energy storage safety constraint model to generate preliminary multi-channel dimming control parameters. The multi-constraint light control decision-making system sets clear execution priorities. The first priority is energy storage safety constraints, ensuring that the battery is not discharged or damaged. The second priority is advertising visibility requirements, ensuring that core information is clearly displayed. The third priority is spatiotemporal scene constraints, adapting to the lighting requirements of different time periods and areas. The decision-making process matches the constraints one by one according to the priority order and automatically eliminates conflicting parameters.

[0036] The dynamic constraint model sets three levels of energy storage protection thresholds. When the SOC of the energy storage battery is below 30%, energy-saving constraints are activated, reducing the drive power by 20%. When the SOC is below 20%, endurance constraints are activated, retaining only the basic brightness of key information areas and turning off the light source in the background area. When the SOC is below 10%, over-discharge protection is activated, and the light box enters a low-power standby state, retaining only the light sensing and energy storage monitoring functions. At the same time, the depth of charge and discharge is adjusted according to the battery SOH and the number of cycles. The lower the SOH, the smaller the allowable depth of discharge.

[0037] The energy storage and dimming drive collaborative control subunit dynamically adjusts the drive control parameters of the dimming drive according to the energy storage battery status. The drive control parameters include reducing the drive current and optimizing the PWM frequency when the SOC is low, and matching the optimal brightness drive control parameters when the SOC is high. At the same time, the light energy consumption data is fed back to the energy storage management unit to realize the linkage of photovoltaic power generation prediction, energy storage status monitoring, light control strategy optimization, and dimming drive adjustment. Finally, multi-channel dimming control parameters are generated and sent to the dimming drive module. The multi-channel dimming control parameters include the target output power, dimming mode, and lighting time planning of each light strip channel.

[0038] The photovoltaic power generation prediction is updated hourly. It combines real-time photovoltaic power generation, historical weather and sunshine data for the same period, remaining sunshine duration, photovoltaic module conversion efficiency, and other multi-dimensional information to predict the remaining power generation and total power generation. The prediction results are synchronized to the multi-constraint light control decision subunit in real time. When the power generation is sufficient, it operates according to the optimal visibility parameters of the advertisement. When the power generation is insufficient, the brightness of the background area is reduced in advance, the lighting time planning is optimized, and sufficient energy storage capacity is reserved to ensure the lighting needs during core periods such as morning and evening peak hours.

[0039] In this embodiment of the invention, the dimming drive module includes a multi-channel constant current drive chipset, a linear and PWM dual-mode dimming switching pulse width modulation unit, an LED lamp bead junction temperature detection unit, a drive protection subunit, and an AI adaptive drive control parameter calibration subunit. The system receives multi-channel dimming control parameters from the multi-constraint light control decision subunit and implements independent constant current drive for multiple backlight and fill light strips of the advertising light box. Based on target brightness parameters, energy storage status, and LED junction temperature, it automatically switches dimming modes: linear dimming mode for low brightness and PWM dimming mode for high brightness, improving drive conversion efficiency. The LED junction temperature detection unit collects the operating temperature of the LEDs in real time, and automatically reduces the drive current when the junction temperature exceeds a preset threshold. When the preset threshold for junction temperature protection of LED beads is 85℃, when the junction temperature is detected to reach 85℃ in real time, the system immediately reduces the drive current in a stepwise manner, reducing the current output every 30 seconds until the junction temperature drops below 75℃; when the junction temperature stabilizes and falls back to a safe range, the system gradually restores the drive current to the normal operating parameters. The restoration process is slow and stable to avoid damage to the LED beads caused by rapid current changes, while recording abnormal junction temperature data.

[0040] The multi-channel driving division is strictly based on the key information area and non-key background area after deep learning segmentation of the advertising image. Key information areas such as brand logo and core copy correspond to independent dedicated driving channels, and the background area is divided into multiple independent driving channels according to the top, bottom, left and right positions. Each driving channel is independent of the others and does not interfere with each other.

[0041] The dual-mode dimming switching uses 30% of the lightbox output brightness as the dividing threshold. When the brightness is below 30%, it automatically switches to linear dimming to avoid flickering at low brightness and improve the softness of the image display. When the brightness is above 30%, it automatically switches to PWM dimming to improve drive efficiency and brightness control accuracy. The drive protection subunit is set with an overcurrent threshold of 1.2 times the rated current and an overvoltage threshold of 1.1 times the rated voltage. The short circuit and open circuit fault detection response time is less than 10ms. After a fault is detected, the corresponding channel drive power is immediately cut off, and a fault warning is triggered at the same time.

[0042] The drive protection subunit monitors overcurrent, overvoltage, short circuit, and open circuit faults in real time and triggers corresponding protection actions in the circuit. The AI ​​adaptive drive control parameter calibration subunit calibrates the output current, voltage, and PWM duty cycle of the drive chip in real time based on changes in ambient light, advertising content requirements, and energy storage status, while recording the data on changes in drive control parameters.

[0043] The AI ​​adaptive drive control parameter calibration model is a lightweight regression model with multiple inputs and a single output. It consists of three core sub-modules: a feature fusion layer, a parameter prediction layer, and an error correction layer, achieving matching of drive control parameters with operating conditions. The model input dimensions include ambient light characteristic parameters such as ambient light brightness, illumination uniformity, and spectral distribution; content adaptation parameters such as brightness weights of key advertising areas and background areas, and minimum visible power; energy storage status parameters such as battery SOC, SOH, and temperature; and hardware status parameters such as real-time junction temperature of LED beads and voltage and current of the drive circuit. The model output includes the target drive current, output voltage, and PWM duty cycle calibration value of the multi-channel constant current drive chip. The model employs a real-time closed-loop calibration mechanism with a fixed calibration cycle of 100ms and calibration error controlled within ±1%. When the LED junction temperature exceeds the threshold or the energy storage capacity is insufficient, the parameter protection and correction logic is automatically activated, synchronously matching the switching requirements of linear / PWM dual-mode dimming.

[0044] AI adaptive calibration forms a complete closed loop. First, it collects real-time data from multiple dimensions such as ambient light, energy storage, screen features, and LED status. Then, it calculates the optimal drive control parameters through a model. After execution, it immediately collects the actual output results and compares them with the calibration target. If a deviation occurs, it corrects the parameters in real time. The closed loop cycle is synchronized with the data acquisition cycle to ensure that the drive control parameters are always in the optimal state.

[0045] In this embodiment of the invention, the self-diagnosis and correction module includes a sensor self-correction subunit, a light source light decay compensation subunit, and a drive circuit self-diagnosis subunit; The sensor self-calibration subunit has a built-in standardized calibration process, which performs zero-point calibration regularly to correct the dark current and temperature drift deviation of the photosensitive sensor, and performs gain calibration regularly to correct the sampling gain deviation caused by dust obstruction and component aging. The light source light decay compensation subunit periodically collects the actual luminous flux feedback value of the light source and compares it with the factory reference parameters. When the light decay of the LED bead is detected to exceed the preset threshold, the driving current is automatically increased to compensate for the light decay.

[0046] Zero-point calibration is performed in a dark environment with the light box off and no external light, automatically clearing the sensor's dark current deviation and eliminating the basic acquisition error caused by temperature drift. Gain calibration uses a standard brightness light source as a reference, compares the actual sensor acquisition value with the standard value, automatically corrects the sampling amplification factor, and compensates for the acquisition attenuation caused by dust obstruction and component aging. Both calibrations are completed automatically, and the parameters are updated and saved immediately after calibration.

[0047] The light source light decay compensation test is performed every 7 days with full coverage testing of the entire LED strip. The actual luminous flux feedback value of each channel LED light source is collected in real time and compared with the factory benchmark luminous flux parameter. The light decay compensation adopts a step-by-step slow adjustment method, with the single drive current increase not exceeding 5% to avoid damage to the LED beads and drive circuit by sudden current changes. After the compensation is completed, parameters such as compensation time, compensation range, and luminous flux recovery value are automatically recorded to track the light decay rate and attenuation law of the light source over a long period of time.

[0048] The sensor zero-point calibration is performed daily, and the gain calibration is performed weekly. The calibration process is completed automatically without manual intervention. The preset threshold for light source decay compensation is 20%. When the actual luminous flux of the LED light source decreases by more than 20% compared to the factory reference, the driving current is automatically and gradually increased for compensation, with a single increase not exceeding 5% to avoid damage to the light source from sudden current changes. The driving circuit self-diagnoses by monitoring the current fluctuation range and chip operating temperature in real time to determine faults. When the temperature exceeds 85℃ or the current fluctuation exceeds ±5%, it is determined to be a potential fault and an early warning is triggered.

[0049] The self-diagnostic subunit of the drive circuit monitors the dimming drive module in real time, including circuit parameters such as drive current, voltage fluctuation, and chip temperature. It identifies potential faults in the drive circuit, provides early warnings, and adjusts the drive control parameters. At the same time, the self-diagnostic data is fed back to the remote management module.

[0050] When the system detects potential faults in drive circuits, sensors, light sources, etc., it sends out early warning information synchronously with the remote cloud via local status indicator lights. The early warning information includes three categories: fault location, fault type, and risk level. Parameter adjustment follows the corresponding strategy according to the fault level. Minor faults automatically correct operating parameters to maintain normal operation, moderate faults restrict some functions to ensure core operation, and severe faults immediately cut off the drive power and activate protection to prevent the fault from escalating and damaging the equipment.

[0051] In this embodiment of the invention, the remote control module includes an edge computing processing unit, a low-power wireless communication subunit, a local data storage unit, a cloud platform interaction subunit, and a remote parameter configuration subunit; The edge computing local closed-loop control cycle is 100ms. All dimming decisions, parameter calibration, and fault protection actions are completed locally. It can still operate independently and stably for more than 72 hours when the network is disconnected. The low-power wireless communication subunit uploads operating data every 30 minutes. When a fault occurs, it actively uploads fault information immediately. The local data storage unit can save more than 30 days of operating parameters, calibration records, and fault logs. The data can be completely saved after power failure and automatically restored after power is restored.

[0052] Employing an edge computing architecture, the edge computing processing unit completes all real-time sensing, decision-making, and dimming control actions locally, ensuring stable and independent operation of the system even during network interruptions. Through the low-power wireless communication subunit, system operation data, including energy consumption data, brightness parameters, ambient light data, energy storage status, and drive circuit status, are periodically uploaded to the cloud management platform. The low-power wireless communication subunit supports both 4G and NB-IoT communication modes.

[0053] The local data storage unit locally caches operating data, calibration parameters, and fault records to ensure that data is not lost when power is off. The remote parameter configuration subunit supports the cloud platform to send dimming drive-related drive control parameters, such as drive current threshold, PWM frequency, and dual-mode switching threshold, to achieve remote control. It also supports unified configuration and iterative optimization of dimming drive control parameters for batch light boxes, enabling unified operation and maintenance management of batch light boxes. The cloud platform uses big data analysis to iteratively optimize the light control models and drive control parameters of light boxes in different scenarios and regions.

[0054] Batch lightbox parameter configuration supports grouping operations by multiple dimensions such as geographical region, application scenario, device model, and operating status. After the configuration command is issued from the cloud, the local edge computing processing unit automatically verifies the legality and compatibility of the parameters. Once the verification is successful, the parameters take effect immediately without the need for manual on-site debugging. If the parameter verification fails or an anomaly occurs during the configuration process, the system automatically reverts to the previous stable operating parameters and simultaneously reports the fault information to the cloud.

[0055] The cloud-based big data iterative optimization model adopts an edge-cloud collaborative architecture, integrating K-Means scene clustering and XGBoost parameter regression algorithms to achieve intelligent iterative upgrades of batch lightbox parameters. The model input data includes: historical ambient light time-series data, real-time energy consumption data of the lightbox, energy storage battery charging and discharging data, advertising visibility compliance rate, driving circuit operating parameters, and scene characteristic data for different regions, such as municipal public transportation, commercial streets, and community parks. The model output consists of core control parameters such as the optimal dimming threshold for each scene, the upper limit of the driving current, PWM frequency modulation parameters, lighting time planning, and energy storage protection strategies. The specific model iteration rules are as follows: The local running data of the edge computing processing unit is summarized every day at midnight to complete the daily data cleaning and feature extraction; a full parameter iteration training is performed once a week, with the optimization goals of reducing energy consumption by ≥15%, achieving an advertising visibility rate of ≥98%, and extending the energy storage cycle life by ≥20%; after the iteration is completed, the optimal parameters are sent to the local device through the low-power wireless communication module, supporting unified configuration of batch light boxes and differentiated adaptation for different scenarios.

[0056] This invention also provides a solar-powered advertising light box. Based on the above system, it includes an advertising display component, a photovoltaic power supply component, an energy storage component, and an intelligent control component. The intelligent control component integrates a light sensing module, a content adaptation module, an energy storage linkage module, a dimming drive module, a self-diagnosis and correction module, and a remote control module. The modules work together to achieve intelligent operation of the solar-powered advertising light box. The intelligent control component adopts an integrated hardware design, integrating the control circuits, processing chips, and communication units of the six functional modules onto the same main control circuit board. The functional modules achieve high-speed data interaction through the onboard bus. The intelligent control component is connected to the photovoltaic power supply component, energy storage component, and advertising display component through standardized interfaces, which have reverse connection protection and overvoltage protection functions.

[0057] The photovoltaic power supply component is used to collect solar energy and convert it into electrical energy; the energy storage component is used to store electrical energy and power the advertising box; and the advertising display component is used to display advertising content. The light sensing module is used to collect ambient light data around the advertising box, filter out non-visible light and transient light interference, complete temperature drift compensation and calibrate the sensing data, and output ambient light characteristic parameters. The content adaptation module is used to preprocess the advertisement image, identify key information areas of the advertisement and extract image features, calculate brightness requirements and allocation parameters in combination with ambient light feature parameters, and convert them into dimming drive signals. The energy storage linkage module is used to collect the operating parameters of the photovoltaic power supply module and the energy storage module, establish a safety constraint model, generate dimming control parameters by integrating multiple factors, and dynamically adjust the drive control parameters in combination with the energy storage status. The dimming drive module is used to receive dimming control parameters, realize independent driving of the advertising display components for each channel, automatically switch dimming modes, detect the status of LED beads and circuit faults and trigger protection actions, and calibrate drive control parameters to ensure display accuracy and stability. The self-diagnosis correction module is used to calibrate the sampling deviation of the light sensing module, compensate for the light decay of the advertising display component, monitor potential faults in the driving circuit and provide early warning and adjustment, and provide feedback self-diagnosis data to support operation and maintenance and control decisions. The remote control module adopts an edge computing architecture to achieve local real-time control and uploading of operational data, and supports remote control, batch operation and maintenance, and parameter optimization in the cloud.

[0058] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0059] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart light control system for solar-powered advertising light boxes, characterized in that, Includes the following modules: The light sensing module is used to collect ambient light data around the light box, perform interference filtering, temperature drift compensation and data calibration of the ambient light data, and output standardized ambient light characteristic parameters. The content adaptation module is used to preprocess the advertising images projected by the lightbox, extract the visuality-related features of the images and establish a multi-dimensional parameter correspondence. It combines real-time ambient light parameters to calculate the light source power and brightness allocation parameters that meet the visibility requirements and converts them into drive control parameters that the driver can recognize. The energy storage linkage module is used to collect real-time operating data of photovoltaic power generation and energy storage equipment, establish dynamic constraint relationship between energy storage safety protection and cycle life optimization, generate multi-channel dimming initial control parameters by combining multiple constraint conditions, optimize parameters by combining real-time energy storage status, and output the final drive control parameters to the dimming drive module. The dimming drive module is used to receive dimming control parameters, implement independent constant current drive for the backlight and supplementary light components of the light box, automatically switch dimming modes according to operating conditions, and complete circuit fault detection and protection and real-time calibration of drive control parameters. The self-diagnostic calibration module is used to periodically calibrate the acquisition deviation of the light sensing module, compensate for the light decay of the LED light source, monitor the operating status of the drive circuit in real time, identify potential faults, and perform early warnings and parameter adjustments. The remote control module adopts an edge computing architecture to achieve local real-time dimming closed-loop control, supports the uploading of operating data, remote control in the cloud, and batch configuration of device parameters, and iteratively optimizes system operating parameters through big data analysis.

2. The intelligent light control system for a solar-powered advertising light box according to claim 1, characterized in that, The light sensing module includes a distributed brightness sensor, a directional spectral sensor with a filtering unit, an ambient temperature and humidity acquisition unit, a data preprocessing subunit, and a dimming drive feedback calibration subunit. The distributed brightness sensors are deployed at the four corners of the light box to synchronously collect ambient light data from multiple directions around the light box. The directional spectral sensor filters out non-visible light interference through infrared cutoff and visible light narrowband filtering, and collects effective visible light brightness and spectral distribution data. The data preprocessing subunit uses a combination of multiple algorithms to filter short-term transient light interference and combines ambient temperature data to complete sensor temperature drift compensation. The dimming drive feedback calibration subunit is used to collect the real-time output current and voltage parameters of the dimming drive module, reverse-calibrate the ambient light sensing data, and finally output ambient light parameters including illuminance, illumination uniformity, and spectral distribution.

3. The intelligent light control system for a solar-powered advertising light box according to claim 2, characterized in that, The content adaptation module includes an advertising screen feature storage unit, an image segmentation and feature extraction subunit, a visibility threshold modeling subunit, a brightness requirement calculation subunit, and a zone dimming driver adaptation subunit. The image segmentation and feature extraction subunit uses a deep learning image segmentation algorithm to distinguish key information areas from non-key background areas of the advertisement image and extracts color, grayscale contrast, and minimum visible brightness threshold features of each area. The visibility threshold modeling subunit is used to establish a four-dimensional linkage model of ambient light brightness, light source output power, advertising content visibility, and dimming drive control parameters; the brightness requirement calculation subunit calculates the minimum total output power of the light source to meet the visibility requirements of the key information area of ​​the advertisement, and the brightness allocation weight parameters corresponding to each light strip channel of the light box, in combination with real-time ambient light parameters. The partitioned dimming driver adapter subunit is used to convert brightness allocation parameters into driver-recognizable drive control parameters.

4. The intelligent light control system for a solar-powered advertising light box according to claim 3, characterized in that, The energy storage linkage module includes a photovoltaic parameter acquisition unit, an energy storage battery status acquisition unit, an energy storage safety constraint modeling subunit, a multi-constraint light control decision subunit, and an energy storage and dimming drive collaborative control subunit. The photovoltaic parameter acquisition unit is used to collect real-time power generation, daily cumulative power generation, and photoelectric conversion efficiency data of photovoltaic modules; the energy storage battery status acquisition unit is used to collect state of charge (SOC), state of health (SOH), battery temperature, and charge / discharge cycle count data of energy storage lithium batteries; the energy storage safety constraint modeling subunit is used to establish a dynamic constraint model for over-discharge protection and cycle life optimization of energy storage batteries; the multi-constraint light control decision subunit integrates advertising content visibility requirements, spatiotemporal scene constraint rules, and energy storage safety constraint models to generate preliminary multi-channel dimming control parameters; the energy storage and dimming collaborative control subunit dynamically adjusts the drive control parameters of the dimming drive according to the real-time status of the energy storage battery, and simultaneously feeds back the light energy consumption data to the energy storage status management unit inside the energy storage linkage module, ultimately generating and distributing the target power, dimming mode, and lighting plan for each light strip.

5. The intelligent light control system for a solar-powered advertising light box according to claim 4, characterized in that, The dimming drive module includes a multi-channel constant current drive chipset, a linear and pulse width modulation (PWM) dual-mode dimming switching unit, an LED junction temperature detection unit, a drive protection subunit, and an AI adaptive drive control parameter calibration subunit. The multi-channel constant current drive chipset is used to achieve independent constant current drive for multiple backlight and fill light strips in the light box; the dual-mode dimming switching unit is used to automatically switch the dimming mode according to the target brightness parameters, energy storage status and LED junction temperature, automatically adopting linear dimming mode under low brightness conditions and pulse width modulation dimming mode under high brightness conditions; the LED junction temperature detection unit is used to collect the operating temperature of the LED in real time, and automatically reduce the drive current when the junction temperature exceeds the preset threshold; the drive protection subunit is used to monitor overcurrent, overvoltage, short circuit and open circuit faults in real time and trigger corresponding protection actions; the AI ​​adaptive drive control parameter calibration subunit is used to calibrate the output parameters of the drive chip in real time and record the drive operation data.

6. The intelligent light control system for a solar-powered advertising light box according to claim 5, characterized in that, The self-diagnosis and correction module includes a sensor self-correction subunit, a light source light decay compensation subunit, and a drive circuit self-diagnosis subunit. The sensor self-calibration subunit incorporates a standardized calibration process, periodically performing zero-point calibration to correct the dark current and temperature drift deviations of the photosensitive sensor, and periodically performing gain calibration to correct sampling gain deviations caused by dust obstruction and component aging, compensating for LED light source light decay. The light source light decay compensation subunit periodically collects the actual luminous flux feedback value of the light source, compares it with the factory reference parameters, and automatically increases the drive current to complete light decay compensation when the light decay of the LED bead exceeds a preset threshold. The drive circuit self-diagnosis subunit monitors the circuit parameters of the dimming drive module in real time, identifies potential faults in the drive circuit and provides early warnings, synchronously adjusts the drive control parameters, and feeds back the self-diagnosis data to the remote management module.

7. The intelligent light control system for a solar-powered advertising light box according to claim 6, characterized in that, The remote control module includes an edge computing processing unit, a low-power wireless communication subunit, and a local data storage unit. The edge computing processing unit is used to complete all real-time sensing, decision-making, and dimming closed-loop control actions locally; the low-power wireless communication subunit is used to periodically upload the system's operating data to the cloud management platform; and the local data storage unit is used to locally cache operating data, calibration parameters, and fault records.

8. The intelligent light control system for solar-powered advertising light boxes according to claim 7, characterized in that, The remote control module further includes a cloud platform interaction subunit and a remote parameter configuration subunit; The cloud platform interaction subunit is used to receive control commands and parameter configuration information issued by the cloud management platform; The remote parameter configuration subunit supports the cloud-based distribution of dimming drive-related control parameters. It also supports unified parameter configuration and iterative optimization for batch light boxes. The cloud management platform can use big data analysis to iteratively optimize the operating parameters of light box systems in different scenarios and regions.

9. A solar-powered advertising light box, based on the system according to any one of claims 1-8, characterized in that, This includes advertising display components, photovoltaic power supply components, energy storage components, and intelligent control components; The intelligent control component integrates all functional modules of the solar advertising light box intelligent light control system, and the modules work together to realize the intelligent operation of the light box; The photovoltaic power supply component is used to collect solar energy and convert it into electrical energy, the energy storage component is used to store electrical energy and power the entire light box, and the advertising display component is used to display advertising content.

10. A solar-powered advertising light box according to claim 9, characterized in that, The intelligent control component collects ambient light data around the lightbox through a light sensing module and outputs ambient light characteristic parameters. It calculates brightness requirement parameters by combining advertising image features and ambient light parameters through a content adaptation module. It generates dimming control parameters by combining photovoltaic and energy storage operation status through an energy storage linkage module. It drives the advertising display component to complete zone dimming through a dimming drive module. It completes self-calibration and fault warning throughout the entire equipment lifecycle through a self-diagnosis and correction module. It realizes local closed-loop control and cloud-based batch operation and maintenance through a remote management and control module.

Images