LED illumination energy-saving control system based on data judgment

Abstract

The invention relates to the technical field of intelligent lighting control, and discloses an LED lighting energy-saving control system based on data judgment, which comprises a data sensing module, a scene recognition module, a strategy storage module and a driving control module, the data sensing module is used for collecting personnel dynamic behavior information and environment illumination intensity information in an illumination area; the scene recognition module outputs a corresponding scene recognition signal by receiving, fusing and analyzing the sensing data; a mapping relation table is stored in the strategy storage module; and the driving control module queries the mapping relation table through the scene identification signal to obtain the illumination control parameter of the target, and converts the illumination control parameter into an instruction signal for driving the LED lamp. According to the invention, through multi-dimensional data fusion and scene adaptive control, refined energy saving of the lighting system is realized on the premise of ensuring comfort.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent lighting control, and particularly relates to an LED lighting energy-saving control system based on data judgment. BACKGROUND

[0002] The LED lighting system has gradually replaced the traditional light source and become the mainstream lighting scheme due to its high light efficiency, long service life and other advantages. In order to achieve energy saving, the prior art mostly adopts simple timing control or switch control based on a single sensor. This kind of method reduces the waste of "long-lasting light" to a certain extent, but its control dimension is single, cannot perceive the complex changes of the lighting environment, the energy-saving effect is limited, and the user experience is poor.

[0003] With the development of Internet of Things technology, lighting control systems with simple grouping or preset scenes have appeared. However, these systems mostly rely on manual switching or fixed time table, lack of automatic and intelligent data judgment ability based on environmental and personnel activities. The core problems are: first, the system "perception" ability is insufficient, cannot obtain and fuse multi-dimensional data such as personnel dynamic behavior and environmental illumination; second, the system "cognition" ability is lacking, cannot accurately judge the current space use state based on these data; third, the system "execution" strategy is rigid, the control logic in the strategy storage module is fixed and cannot realize closed-loop intelligent decision from perception to execution. This leads to that the existing system cannot truly realize "lighting on demand", there is still significant energy waste in the complex scene of no one, low illumination or partial area use, and also cannot provide a comfortable and self-adaptive artificial light environment.

[0004] The LED lighting control system in the prior art has obvious shortcomings in the dimension of data perception, the intelligence of scene recognition and the adaptability of control strategy, and an integrated solution that can deeply fuse multi-source information, intelligently judge the scene and dynamically adjust the lighting strategy is urgently needed. Therefore, the present application proposes an LED lighting energy-saving control system based on data judgment. SUMMARY

[0005] The present application aims at solving the problems of single perception dimension, rigid scene recognition and static control strategy in the prior art, and proposes an LED lighting energy-saving control system based on data judgment.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: an LED lighting energy-saving control system based on data judgment, comprising: a data perception module, configured to collect personnel dynamic behavior information and environmental illumination intensity information in the lighting area, and output corresponding sensing data; The scene recognition module judges the current space use state by receiving and fusing analysis sensing data, and outputs a corresponding scene recognition signal; The policy storage module internally stores a mapping relationship table, which defines the corresponding relationship between different scene recognition signals and lighting control parameters; The driving control module queries the mapping relationship table to obtain the target lighting control parameter through the scene recognition signal, and converts it into an instruction signal for driving the LED lamp.

[0007] Compared with the prior art, the present application has the following advantages: The present application can realize all-round and fine perception of the space use state by setting the data perception module, providing rich and accurate data basis for intelligent decision-making, and fundamentally overcoming the single perception dimension defect of the prior art, and providing the possibility for precise energy saving and comfortable lighting.

[0008] The present application can enable the system to have the ability to understand the environmental intention, automatically distinguish different scenes such as office, meeting, lunch break, etc., realize the leap from passive switching to active cognition, and provide the core intelligent judgment basis for on-demand lighting.

[0009] The present application can provide the optimal lighting strategy for each identified scene by setting the policy storage module, ensuring the standardization and consistency of lighting actions, and enabling the intelligent judgment of the system to be accurately and reliably executed.

[0010] The present application can form a full-automatic and closed-loop intelligent control pipeline from scene judgment to strategy calling to lamp driving by setting the driving control module. Ultimately, the system can eliminate invalid lighting and excessive lighting to the maximum extent under the premise of meeting the lighting needs of different scenes, achieve significant energy saving effect, and greatly improve the comfort and intelligent level of the light environment. BRIEF DESCRIPTION OF DRAWINGS

[0011] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0012] Figure 1 is a system structure schematic diagram provided by the embodiment of the present application; Figure 2 is a signal conversion schematic diagram of the driving control module provided by the embodiment of the present application. DETAILED DESCRIPTION

[0013] In order to further clarify the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the specific implementation, structure, features and effects of a data judgment-based LED lighting energy-saving control system according to the present application are described in detail below in combination with the drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0014] 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 the present application belongs.

[0015] The following examples are for illustrative purposes only and are not intended to limit the scope of the present application.

[0016] Traditional and existing intelligent lighting systems often have rigid control strategies when facing complex indoor environments, and cannot achieve true on-demand lighting, resulting in energy waste and poor user experience.

[0017] For example, assume an open office area whose lighting system uses a traditional light sensor + timer control. In the afternoon of a working day, due to cloudy weather, the natural light suddenly weakens, and the system uniformly adjusts the brightness of the entire area's lamps to the highest level according to the light sensor signal. However, the actual staff distribution in this area at this time is: Zone A has employees working in concentration, Zone B employees have gone to the conference room for a meeting, and Zone C is empty. The system cannot sense the dynamic distribution of personnel and the actual use state of the space, resulting in full-power lighting for the unoccupied Zones B and C, causing energy waste. At the same time, when the employees in Zone B return to their workstations after the meeting, the system cannot predict or respond in a timely manner, and the lights will only be turned on after the sensor is triggered when the personnel enter the area with insufficient light, causing inconvenience and delay in use.

[0018] If the above problems are not solved, the lighting system will always be a blind energy consumer. Its fixed, rule-based control logic cannot adapt to dynamically changing environments and personnel activities, making it difficult to achieve fine energy-saving goals and providing users with a comfortable and seamless light environment experience.

[0019] The specific scheme of a data judgment-based LED lighting energy-saving control system provided by the present application is described in detail below in combination with the drawings.

[0020] Please refer to Figure 1 which shows a system structure schematic diagram of a data judgment-based LED lighting energy-saving control system provided by an embodiment of the present application, which includes: a data-aware module for collecting dynamic behavior information of people and ambient light intensity information in the lighting area, and outputting corresponding sensing data; a scene recognition module for judging the current space usage state by receiving and fusing the sensing data, and outputting corresponding scene recognition signals; a strategy storage module having a mapping relationship table stored therein, defining the corresponding relationship between different scene recognition signals and lighting control parameters; a driving control module for querying the mapping relationship table by the scene recognition signals to obtain the target lighting control parameters, and converting them into instruction signals for driving the LED lamps.

[0021] It should be noted that collecting dynamic behavior information of people and ambient light intensity information in the lighting area refers to obtaining physical quantities related to people's activities and ambient light conditions through various sensors deployed in the lighting area in a non-contact manner. For example, a microwave radar sensor can be used to detect the moving speed, distribution density and stationary state of people, and a photoelectric sensor can be used to measure the illuminance level in the environment. The purpose is to convert the dynamic environment in the real world into a digital, quantifiable data basis for analysis.

[0022] Receiving and fusing the sensing data to judge the current space usage state refers to integrating and interpreting various sensing data from different sensors, which may have different sampling rates and data formats. For example, a data fusion algorithm can be used to align the data in time and space, and a machine learning model can be used to classify the fused feature vectors. The purpose is to extract high-level semantic information from multi-source, heterogeneous sensing data, i.e., to understand that the current space is in an efficient work, group collaboration, rest or other specific state.

[0023] Having a mapping relationship table stored therein, defining the corresponding relationship between different scene recognition signals and lighting control parameters refers to pre-setting a data set of lookup table structure in the non-volatile memory of the system. The data set associates different scene recognition signals (such as a code representing "meeting mode") with a set of optimal lighting control parameters (such as target illuminance value, target color temperature value and gradual transition time parameter). The purpose is to serve as the "strategy brain" of the system, quickly and accurately mapping the recognized scene intention to specific, executable lighting action instructions.

[0024] The mapping relationship table is queried by the scene recognition signal to obtain the target lighting control parameter, and the target lighting control parameter is converted into an instruction signal for driving the LED lamp, which means that according to the input scene recognition signal, the corresponding preset parameter is obtained by searching the mapping relationship table as an index, and then the digital parameter is converted into a physical signal capable of directly adjusting the LED driver through a signal conversion circuit (such as a microcontroller generating a PWM signal). The purpose is to realize a closed loop from intelligent decision to physical execution, and finally realize the scene strategy of the upper layer into the optical output change of the terminal lamp.

[0025] I. Data perception module The data perception module includes a personnel perception unit, an environment perception unit, and a timing reference unit. The personnel perception unit includes a microwave radar sensor, which obtains real-time position, motion state, and distribution information of personnel by emitting a detection signal and analyzing phase and frequency change information in the echo signal, and encapsulates the information as a first sensing data substream. The environment perception unit synchronously performs quantitative measurement of ambient light intensity and waveform acquisition and feature analysis of ambient sound through parallel signal acquisition channels, outputs environment state data containing digital quantity of light intensity and spectral features of sound, and encapsulates the data as a second sensing data substream. The timing reference unit generates timing signals with absolute time stamps and fixed periods through a clock source, provides a unified time reference, and outputs the timing signals as a third sensing data substream.

[0026] Further, in the data perception module, the environment perception unit includes: The photoelectric sensing subunit includes a light-sensitive unit, which linearly converts incident light power into an electrical signal, and outputs a digital quantity representing ambient light intensity. The acoustic sensing subunit includes an array structure composed of multiple microphones, which enhances voice signal collection in a specific direction through beamforming technology, and extracts spectral features of the collected sound signals through time-frequency domain transformation.

[0027] It should be noted that the microwave radar sensor refers to a 60GHz frequency-modulated continuous wave radar chip of the model Infineon BGT60TR13C used in this embodiment. Its detection distance range is 0.1 to 20 meters, the static personnel detection distance is not less than 15 meters, the speed detection range is 0.1 to 5 meters / second, and the angle detection accuracy is ±1 degree. Its purpose in this embodiment is to accurately obtain real-time position, motion state, and distribution information of personnel in a non-contact manner by analyzing the phase and frequency changes of the echo signal.

[0028] The personnel dynamic behavior information refers to a data set about the real-time spatial coordinates, moving speed and direction of the personnel in the lighting area, and the distribution density in each sub-area, which is obtained by the microwave radar sensor by analyzing the phase and frequency changes of the echo signal. In the embodiment, it is to provide accurate quantitative personnel activity input for the scene recognition module to judge the state of the space, such as “concentrated office work”, “personnel flow” or “space idle”.

[0029] The ambient light intensity information refers to the digital quantity in lux units output by the photosensitive unit of the photoelectric sensing sub-unit after linear conversion and quantization of the incident light power. In the embodiment, it is combined with the personnel dynamic behavior information to be a direct basis for judging whether to turn on or adjust artificial lighting and evaluating the level of natural light utilization.

[0030] The sensing data refers to the standardized data set output by the data perception module, which is composed of the first, second and third sensing data sub-streams. In the embodiment, it is a unified format information interaction carrier agreed between the data perception module and the scene recognition module.

[0031] The first sensing data sub-stream refers to the special data stream generated by the personnel perception unit, which encapsulates the real-time position, motion state and distribution information of the personnel. In the embodiment, it is uniquely identified and transmitted in the system to transmit the personnel perception results from the microwave radar.

[0032] The second sensing data sub-stream refers to the special data stream generated by the environment perception unit, which encapsulates the digital quantity of ambient light intensity and the sound spectrum characteristics. In the embodiment, it is uniquely identified and transmitted in the system to transmit the environment perception results from the photoelectric and acoustic sensors.

[0033] The third sensing data sub-stream refers to the special data stream generated by the timing reference unit, which encapsulates the absolute time stamp and fixed period timing signal. In the embodiment, it provides a unified time mark for the first and second sensing data sub-streams, and provides a reference for the system to execute time-related lighting strategies.

[0034] The parallel signal acquisition path refers to the mutually independent hardware signal processing path set up in the environment perception unit to realize the synchronous measurement and acquisition of ambient light intensity and ambient sound. In the embodiment, it ensures that the acquisition of light and sound heterogeneous environment data is strictly synchronized in time, and provides a time-consistent data basis for subsequent judgment of the composite scene.

[0035] The light sensitive unit refers to the light sensitive core integrated in the AMS TSL2591 high dynamic range digital light intensity sensor in the embodiment. Its spectral response range covers visible light to infrared light, the illumination measurement range is 0Lux to 88000Lux, and the measurement accuracy is ±3% at 100Lux illumination. Its purpose in the embodiment is to linearly convert the incident light power into a digital electrical signal, and output a quantized digital quantity representing the intensity of the ambient light.

[0036] The array structure of multiple microphones refers to the array composed of four analog MEMS microphones arranged in a linear manner in the embodiment. The microphone spacing is 2.5 cm, the frequency response range is 100 Hz to 16 kHz, and the sensitivity at 1000 Hz is -26 dBFS. Its purpose in the embodiment is to enhance the collection of voice signals in a specific direction through the physical structure and beamforming algorithm, and the directional beam width is about 60 degrees.

[0037] Quantitative measurement refers to the process of converting the continuous physical analog quantity of ambient light intensity into discrete digital quantity by the internal analog-to-digital converter of the photoelectric sensing subunit. In the embodiment, a numerical value is generated that can be accurately compared with a preset illumination threshold to trigger deterministic lighting control logic.

[0038] Waveform acquisition and feature analysis refers to the process of continuously sampling and recording ambient sound signals by the acoustic sensing subunit, and further extracting spectral features from them through digital signal processing algorithms. In the embodiment, the original sound wave signal is converted into structured feature data that can be used to distinguish different acoustic scenes.

[0039] Sound spectrum feature refers to the data representing the energy distribution at different frequencies obtained by performing time-frequency domain transformation on the collected sound wave signal. In the embodiment, it is used as an acoustic basis to assist in determining whether the space is in a "meeting discussion", "quiet office" or "unoccupied" state.

[0040] Clock source refers to the temperature-compensated real-time clock chip with model Maxim DS3231M used in the embodiment. The chip has an accuracy of ±2ppm in the range of -40℃ to +85℃, and provides absolute time information of year, month, day, hour, minute and second. Its purpose in the embodiment is to generate a time sequence signal with absolute timestamp and fixed period, and provide a unified and high-precision time reference for all sensing data.

[0041] Absolute timestamp refers to time information based on the standard epoch time that can uniquely identify the time when a data point is generated. In the embodiment, it provides accurate time markers for all asynchronously collected sensing data, so that the system can implement calendar and clock-based lighting strategies.

[0042] Fixed cycle timing signal refers to clock pulse signal periodically generated at stable and accurate time interval. In the embodiment, it is used to coordinate the sampling rhythm within the data sensing module and serve as the synchronization heartbeat for the cooperation of various modules in the system.

[0043] The data sensing scheme of the present application overcomes the limitations of traditional lighting systems in single sensing dimension and poor data timeliness by constructing a multi-modal and asynchronous data synchronous acquisition and fusion framework. Specifically, first, the personnel sensing unit actively detects and analyzes the personnel dynamics in the space using microwave radar, which is equivalent to giving the system the ability to sense "who, where, and what" in the space. At the same time, the environment sensing unit quantifies the ambient light intensity and collects and analyzes the environmental sound characteristics in parallel and synchronously, which is equivalent to giving the system the ability to sense "visual light and dark" and "auditory noise". Since personnel activity and environmental factors are two key and interrelated dimensions that jointly determine the space usage state, for example, personnel concentration and environmental noise usually indicate a meeting state, while personnel sparsity and environmental quietness may indicate an office or departure state. Therefore, by synchronously acquiring these two types of information, more abundant and reliable joint features can be provided for subsequent scene recognition. The timing reference unit applies uniform and accurate timestamps to all data streams from different sensors collected at different times. This ensures that sudden movements of personnel, instantaneous changes in light, and sudden peaks in sound can be correctly associated on the time axis in subsequent data processing, thereby laying a solid data foundation for accurate scene judgment. The entire process forms a standardized preprocessing pipeline for multi-source heterogeneous data, enabling the subsequent intelligent analysis module to work based on a set of aligned in time and space, high-quality fusion perception data II. Scene recognition module The scene recognition module includes a data fusion unit, a feature extraction unit, a pattern classification unit, and a confidence evaluation unit. The data fusion unit uses a timestamp-based interpolation algorithm and coordinate system transformation to perform spatio-temporal alignment on the sensor data output by the data sensing unit, generating a unified fusion data frame. The feature extraction unit uses principal component analysis to extract low-dimensional feature vectors that can distinguish different space usage states from the fusion data frame, including efficient work state, group collaboration state, resting departure state, transition guarantee state, and security warning state. The pattern classification unit uses a classification model built by support vector machines to calculate the low-dimensional feature vectors, determine the current space usage state, and output the preliminary classification result. The confidence evaluation unit uses a confidence degree estimation algorithm based on probability distribution to quantify the reliability of the preliminary classification result, and outputs the scene recognition signal when the confidence degree is higher than the preset threshold. Further, in the scene recognition module, the sensing data output by the data sensing unit includes real-time position, motion state and distribution information of the personnel generated by the microwave radar, environmental sound spectrum features generated by the microphone array, and environmental illumination intensity generated by the illumination sensor; The data fusion unit is configured to align and splice the real-time position, motion state, distribution information of the personnel, environmental sound spectrum features, and environmental illumination intensity in the time dimension to generate a fusion data frame.

[0044] It should be noted that the space-time alignment refers to a data preprocessing process of unifying the time sequence of original sensing data from different sensors with different sampling time and physical coordinate systems through timestamp interpolation, and unifying the spatial reference system of the original sensing data through coordinate system transformation. In the embodiment, the space-time differences between the multi-source data are eliminated to provide a consistent reference for subsequent fusion analysis, and to ensure that the personnel position, environmental sound, and illumination intensity can be correlated and analyzed in time and space.

[0045] The fusion data frame refers to a structured data set formed by organizing the multi-source sensing data after space-time alignment in a unified time sequence, and is the output of the data fusion unit. In the embodiment, a data snapshot that can completely and synchronously represent the comprehensive state of a space at a time is formed, which is directly input to the feature extraction unit.

[0046] The principal component analysis method refers to a statistical method of converting a group of high-dimensional variables that may have correlations into a few linearly uncorrelated low-dimensional variables through orthogonal transformation, and the low-dimensional variables are called principal components. In the embodiment, the key features that have the largest information amount and can best distinguish different space states such as “office”, “meeting”, and “leaving” are extracted from the high-dimensional fusion data frame, and data dimensionality reduction is realized to improve the subsequent classification efficiency.

[0047] The low-dimensional feature vector refers to a feature numerical sequence extracted from the high-dimensional fusion data by the principal component analysis method, which has a significantly reduced dimension but retains most of the variance of the original data. In the embodiment, the core discriminant characteristics of the current space use state are accurately represented in a more compact data form, which is input to the support vector machine classification model.

[0048] The core features and judgment basis of the high-efficiency working state are that the personnel distribution is sparse, in a stationary or low-amplitude activity, the environment is quiet, the corresponding lighting strategy is to maintain a higher illuminance (such as 500 Lux) and a neutral color temperature (such as 4000 K), and to ensure the lighting quality of the work surface.

[0049] The core feature and judgment basis of the group collaboration state is high personnel gathering density and continuous language exchange (active sound spectrum feature). The corresponding lighting strategy is to use uniform overall lighting, appropriate illuminance (such as 300 Lux) and high color temperature (such as 5000K) to maintain concentration.

[0050] The core feature and judgment basis of the rest and leave state is that the personnel density is very low or zero and the activity intensity is weak within a preset period (such as lunch break). The corresponding lighting strategy is to greatly reduce the illuminance (such as 30%) and switch to low color temperature warm light (such as 2700K) to create a relaxed atmosphere and save energy.

[0051] The core feature and judgment basis of the transition security state is the instantaneous switching from no one to someone or the regular movement of personnel between different areas. The corresponding lighting strategy is to quickly respond (short fade time) to safe illuminance to avoid inconvenience and risk caused by darkness.

[0052] The core feature and judgment basis of the security warning state is to detect abnormal and unverified personnel movement during non-working hours. The corresponding lighting strategy is to light up part of the lamps to low brightness or start a specific flashing mode to play a warning and lighting role.

[0053] The support vector machine refers to a supervised learning model that finds a hyperplane in the feature space that maximizes the margin between different classes of samples, suitable for small sample, nonlinear and high-dimensional pattern recognition. In this embodiment, a classifier is constructed that can distinguish different space use states with high accuracy based on low-dimensional feature vectors.

[0054] The preliminary classification result refers to the preliminary judgment conclusion about the space use state calculated by the support vector machine model based on the current input low-dimensional feature vector. In this embodiment, a raw scene recognition output without reliability evaluation is given for the confidence estimation unit to make the final decision.

[0055] The confidence estimation algorithm based on probability distribution refers to a method of evaluating the reliability of the classification result by calculating the probability distribution of each pre-defined class, such as using the Softmax function to map the original output of the classifier to probability. In this embodiment, the certainty of the preliminary classification result is quantified, and those scene recognition results with ambiguous model judgment and low credibility are identified and filtered out.

[0056] The confidence is a probability value representing the reliability of the preliminary classification result calculated by the confidence estimation algorithm, usually in the range of [0, 1]. In this embodiment, it is used as a quantifiable key indicator to judge whether the current scene recognition result is reliable enough to trigger the subsequent lighting strategy execution.

[0057] The preset threshold refers to a preset minimum probability value used to determine whether the confidence is acceptable, for example, 0.9. In the embodiment, the confidence evaluation unit is provided with a clear and unified decision boundary, ensuring that only a high-confidence scene recognition signal can be output.

[0058] The scene recognition signal refers to a digital signal output by the confidence evaluation unit, representing the final confirmed space use state, only when the confidence of the preliminary classification result is higher than the preset threshold. In the embodiment, the scene recognition signal is the only and reliable output of the entire scene recognition module, directly triggering the calling of the corresponding lighting strategy in the strategy storage module.

[0059] The scene recognition scheme of the present application overcomes the limitations of traditional methods in terms of inaccurate judgment of complex scenes and unknown reliability by constructing a complete cognitive chain from multi-source data fusion to classification result reliability evaluation. Specifically, first, the data fusion unit performs spatio-temporal alignment on heterogeneous data from the microwave radar, microphone array, and light sensor to generate a unified fusion data frame. The feature extraction unit then extracts a low-dimensional feature vector that can effectively distinguish different space states from the data frame using principal component analysis. The pattern classification unit uses a support vector machine model to calculate these feature vectors to obtain a preliminary classification result about the current space use state. Finally, the confidence evaluation unit quantifies the confidence of the preliminary result using a probability distribution-based algorithm, and only when the confidence is higher than the preset threshold, the final scene recognition signal is output. This process ensures that the system's judgment of the scene is not only based on multi-dimensional information fusion, but also undergoes strict reliability verification, thereby providing a high-reliability decision basis for subsequent lighting strategy execution.

[0060] III. Strategy storage module The strategy storage module includes a basic strategy library unit and a strategy enhancement unit. The basic strategy library unit internally stores the initial mapping relationship between the associated scene recognition signal and the target brightness and target color temperature. The strategy enhancement unit is used to dynamically bind a set of transition time parameters for each set of target brightness and target color temperature in the initial mapping relationship, forming a lighting strategy set containing instantaneous target values and transition process parameters. Further, in the strategy storage module, the strategy enhancement unit dynamically binds in the following way: Establish a scene switching rule library to define the correspondence between different scene switching types and human visual comfort models. Query the scene switching rule library according to the scene recognition signal to determine the visual comfort model applicable to the current switching; Based on the calculation results of the visual comfort model, select a set of transition time parameters for the current switching; Further, in the strategy storage module, the visual comfort model is configured to: For the scenario of switching from the working state to the resting state, a first-order inertia delay model is used to calculate the brightness decay curve, and a long fade parameter of not less than 10 seconds is generated; For the scenario of switching from the unattended state to the attended state, an S-shaped acceleration function is used to calculate the brightness rise curve, and a short fade parameter of not more than 3 seconds is generated.

[0061] It should be noted that the initial mapping relationship refers to the static corresponding relationship table between the scene recognition signal and the target brightness and the target color temperature pre-stored in the basic strategy library unit. Its purpose in this embodiment is to establish the basic association between the scene and the final lighting effect, and to provide a benchmark configuration for strategy enhancement.

[0062] Dynamic binding refers to the process of matching the target brightness and target color temperature in the initial mapping relationship with the corresponding fade time parameter by the strategy enhancement unit according to the real-time scene switching characteristics. Its purpose in this embodiment is to realize the upgrade of the lighting strategy from static configuration to dynamic adaptation, so that the lighting control has intelligence in the time dimension.

[0063] The fade time parameter refers to the length of time required to smoothly transition the lighting output from the current state to the target state, measured in seconds. Its purpose in this embodiment is to define the rate of optical change during scene switching, which is a key control variable for achieving visual comfort.

[0064] The lighting strategy set refers to a complete set of lighting control instructions containing target brightness values, target color temperature values, and fade time parameters. Its purpose in this embodiment is to form a comprehensive control scheme that can be directly delivered to the driving control module for execution.

[0065] The scene switching rule library refers to a database that stores the mapping relationship between different scene switching types and corresponding human visual comfort models. Its purpose in this embodiment is to provide the strategy enhancement unit with intelligent decision-making basis based on scene switching semantics.

[0066] The human visual comfort model refers to a mathematical model that describes the physiological response of the human visual system to different brightness and color temperature change patterns. Its purpose in this embodiment is to provide a scientific basis for the selection of fade time parameters, ensuring that the lighting switching conforms to the comfort perception of the human eye.

[0067] The first-order inertia delay model refers to a differential equation model that simulates the inertia and delay characteristics of system response, with an exponential decay characteristic in the response curve. Its purpose in this embodiment is to generate a slow and smooth brightness decay curve for the transition from working to resting, avoiding the stimulation of the visual system caused by sudden changes in light.

[0068] The S-shaped acceleration function refers to a mathematical function whose output curve is S-shaped, having the characteristics of initial acceleration, intermediate approximate linearity, and final deceleration. Its purpose in the present embodiment is to generate a rapid but non-abrupt brightness increase curve for the transition from no one to someone, maintaining visual comfort while ensuring response speed.

[0069] The long fade parameter refers to a fade time parameter with a duration of no less than 10 seconds. Its purpose in the present embodiment is to provide a slow enough light change for scene transitions that require relaxation or rest, creating a soothing visual transition.

[0070] The short fade parameter refers to a fade time parameter with a duration of no more than 3 seconds. Its purpose in the present embodiment is to provide timely lighting for scene transitions that require rapid response, while ensuring the smoothness of the change process through an S-shaped curve.

[0071] The scheme of the present application overcomes the limitations of traditional lighting control system strategies, such as single strategy and lack of fine control in the time dimension, by constructing a double-layer strategy architecture containing a basic strategy and an enhanced strategy. Specifically, the basic strategy library unit provides the initial mapping relationship between scenes and target brightness, color temperature, and establishes the basic correspondence of lighting needs; the strategy enhancement unit dynamically binds the most suitable fade time parameter for different scene transition types through the scene transition rule library and the human visual comfort model. For transitions from a working state to a resting state, a first-order inertial delay model is used to generate a long fade parameter of no less than 10 seconds, simulating the natural light decay process; for transitions from an unoccupied state to an occupied state, an S-shaped acceleration function is used to generate a short fade parameter of no more than 3 seconds, achieving rapid but comfortable light increase. This dynamic parameter binding mechanism based on scene transition characteristics and visual comfort models makes lighting control not only focus on the final light environment state, but also pay attention to the visual experience during state transition, achieving comprehensive optimization of lighting strategies in the time and spatial dimensions.

[0072] IV. Drive control module Please refer to Figure 2 is the signal conversion schematic diagram of the drive control module provided by the embodiment of the present application.

[0073] The drive control module includes a strategy query unit, an instruction generation unit, and a signal output unit. The strategy query unit is configured to receive a scene recognition signal and query a mapping relationship table according to the signal to obtain corresponding target lighting control parameters. The instruction generation unit is configured to convert the obtained target lighting control parameters into PWM instruction signals that can drive LED lamps and lanterns. The signal output unit is configured to output the PWM instruction signals to the driver of the LED lamps and lanterns. Further, in the drive control module, the instruction generation unit converts the target lighting control parameter into a PWM instruction signal by performing the following operations: mapping the received target color temperature value to a pre-stored color temperature-brightness matching curve to obtain independent brightness values of the cold white and warm white LED channels; According to the independent brightness value, combined with the target brightness value, the PWM duty cycle of the two channels is calculated respectively; Based on the calculated duty cycle, two-way synchronous PWM pulse signals are generated.

[0074] It should be noted that the query mapping relationship table refers to the data operation process of retrieving and returning the corresponding lighting control parameter in the database of the strategy storage module according to the scene recognition signal as the input key value. Its purpose in this embodiment is to realize fast and accurate mapping from abstract scene to specific lighting parameter, which is the decision basis for obtaining the specific lighting action triggering step.

[0075] The target lighting control parameter refers to the complete control instruction set containing the target brightness value, the target color temperature value and the gradual change time parameter retrieved from the mapping relationship table. Its purpose in this embodiment is to serve as the direct input of the drive control module, which defines the final optical output target and its transition mode expected to be achieved.

[0076] The color temperature-brightness matching curve refers to a data table or function pre-stored in the system, which defines the corresponding relationship between the target color temperature value and the required brightness ratio of the cold white and warm white LED chips. Its purpose in this embodiment is to decouple the single target color temperature value into the brightness control reference of two independent LED channels, providing data support for accurate color temperature adjustment.

[0077] The independent brightness value refers to the brightness value required to be achieved by the cold white LED channel and the warm white LED channel respectively obtained by querying the color temperature-brightness matching curve. Its purpose in this embodiment is to provide independent brightness control targets for the subsequent generation of two independent PWM signals.

[0078] The PWM duty cycle refers to the percentage of the high level duration in the entire cycle in a complete pulse width modulation cycle, and its value has a specific functional relationship with the perceived brightness of the LED. Its purpose in this embodiment is to convert the digital "independent brightness value" into a physical quantity with specific time characteristics that can actually control the LED driver.

[0079] The two-way synchronous PWM pulse signal refers to two-way square wave signals generated by the instruction generating unit, which are strictly aligned in time and the duty cycle is independently controllable, and are used to drive the cold white and warm white LED channels. The purpose thereof in the embodiment is to mix the synthesized light with the target brightness and the target color temperature by the accurate timing control of the two-way LED power supply.

[0080] The driving control scheme of the present application overcomes the limitations of the traditional lighting drive which can only achieve single brightness adjustment or inaccurate color temperature control by constructing an accurate conversion chain from parameter analysis to signal generation. Specifically, first, the target lighting control parameter is obtained by querying the mapping relationship table. The instruction generating unit then performs the core conversion: according to the target color temperature value, the color temperature-brightness ratio curve is queried to obtain the independent brightness values of the cold white and warm white LED channels; combined with the total target brightness value, the required PWM duty cycle of the two channels is calculated respectively; finally, two-way synchronous PWM pulse signals with accurate duty cycle setting are generated. The signal output unit sends the two signals to the LED driver. This process ensures that the system can convert the upper layer lighting strategy into physical signals that can directly drive the dual-channel LED lamp without loss and high precision, and realizes the delicate, dynamic and comfortable control of the light environment.

[0081] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A data-driven LED lighting energy-saving control system, characterized in that, include: The data sensing module is used to collect information on the dynamic behavior of people and the intensity of ambient light within the lighting area, and output the corresponding sensing data. The scene recognition module receives, fuses, and analyzes sensor data to determine the current space usage status and outputs the corresponding scene recognition signal. The strategy storage module internally stores a mapping table that defines the correspondence between different scene recognition signals and lighting control parameters; The drive control module obtains the target's lighting control parameters by querying the mapping table through scene recognition signals, and converts them into command signals to drive the LED lights.

2. The LED lighting energy-saving control system based on data judgment according to claim 1, characterized in that, The data sensing module includes: The personnel sensing unit includes a microwave radar sensor, which acquires the real-time location, motion state, and distribution information of personnel by transmitting detection signals and analyzing the phase and frequency change information in the echo signals, and encapsulates it into a first sensing data sub-stream. The environmental sensing unit, through parallel signal acquisition paths, simultaneously performs quantitative measurement of ambient light intensity and waveform acquisition and feature analysis of ambient sound, outputs environmental state data containing digital light intensity and sound spectrum characteristics, and encapsulates it into a second sensing data sub-stream. The timing reference unit generates a timing signal with an absolute timestamp and a fixed period through a clock source, providing a unified time reference and using the timing signal as a third sensing data sub-stream.

3. The LED lighting energy-saving control system based on data judgment according to claim 2, characterized in that, The environmental sensing unit includes: The photoelectric sensing subunit includes a photosensitive unit, which linearly converts the incident light power into an electrical signal and outputs a digital quantity that characterizes the ambient light intensity. The acoustic sensing subunit includes an array structure consisting of multiple microphones. It enhances the acquisition of speech signals in a specific direction through beamforming technology and extracts the spectral features of the acquired sound wave signals by performing time-frequency domain transformation.

4. The LED lighting energy-saving control system based on data judgment according to claim 1, characterized in that, The scene recognition module includes: The data fusion unit uses a timestamp-based interpolation algorithm and coordinate system transformation to perform spatiotemporal alignment on the sensor data output by the data sensing unit, generating a unified fused data frame. The feature extraction unit uses principal component analysis to extract low-dimensional feature vectors from the fused data frame that can distinguish different spatial usage states, including efficient working state, group collaboration state, quiet departure state, transition protection state, and security warning state. The pattern classification unit uses a classification model built with support vector machines to calculate low-dimensional feature vectors, determine the current space usage status, and output preliminary classification results. The confidence assessment unit uses a confidence estimation algorithm based on probability distribution to quantify the reliability of the preliminary classification results and outputs a scene recognition signal when the confidence level is higher than a preset threshold.

5. The LED lighting energy-saving control system based on data judgment according to claim 4, characterized in that, The sensing data output by the data sensing unit includes the real-time location, motion state and distribution information of the personnel generated by microwave radar, the ambient sound spectrum characteristics generated by microphone array, and the ambient light intensity generated by light sensor. The data fusion unit is used to align and stitch together the real-time location, movement status, distribution information, environmental sound spectrum characteristics, and environmental light intensity of personnel in the time dimension to generate a fused data frame.

6. The LED lighting energy-saving control system based on data judgment according to claim 1, characterized in that, The policy storage module includes: The basic strategy library unit internally stores the initial mapping relationship between the associated scene recognition signals and the target brightness and target color temperature; The strategy enhancement unit is used to dynamically bind a set of gradient time parameters to each set of target brightness and target color temperature in the initial mapping relationship, forming a lighting strategy set that includes instantaneous target values ​​and transition process parameters.

7. The LED lighting energy-saving control system based on data judgment according to claim 6, characterized in that, The policy enhancement unit is dynamically bound in the following manner: Establish a scene switching rule base and define the correspondence between different scene switching types and the human visual comfort model; Based on the scene recognition signal, query the scene switching rule base to determine the visual comfort model applicable to the current switching; Based on the calculation results of the visual comfort model, a set of gradual time parameters are selected for the current switching.

8. The LED lighting energy-saving control system based on data judgment according to claim 7, characterized in that, The visual comfort model is configured as follows: For scenarios where the system switches from working state to resting state, a first-order inertial delay model is used to calculate the brightness decay curve and generate long gradient parameters of no less than 10 seconds. For scenarios transitioning from an unmanned to a manned state, an S-curve acceleration function is used to calculate the brightness enhancement curve and generate short gradient parameters of no more than 3 seconds.

9. The LED lighting energy-saving control system based on data judgment according to claim 1, characterized in that, The drive control module includes: The strategy query unit is used to receive scene recognition signals and query the mapping relationship table based on the signals to obtain the corresponding target lighting control parameters; The instruction generation unit is used to convert the acquired target lighting control parameters into PWM instruction signals that can drive LED lamps; The signal output unit is used to output PWM command signals to the driver of the LED lamp.

10. The LED lighting energy-saving control system based on data judgment according to claim 9, characterized in that, The instruction generation unit converts the target lighting control parameters into PWM instruction signals by performing the following operations: The received target color temperature value is mapped to the pre-stored color temperature-brightness ratio curve to obtain the independent brightness values ​​of the cool white and warm white LED channels; Based on the independent brightness value and the target brightness value, calculate the PWM duty cycle of the two channels respectively; Based on the calculated duty cycle, two synchronous PWM pulse signals are generated.

Images