Intelligent adjustment method and system for quantum dot light emitting lamp spectrum

Abstract

This invention belongs to the field of intelligent lighting technology, and discloses an intelligent adjustment method and system for the spectrum of quantum dot luminaires; aiming to solve the problem that existing lighting systems cannot meet the needs of personalized and healthy lighting. First, multi-dimensional data acquisition is used to obtain user biorhythms, environmental spectra, usage scenarios, user preferences, and physiological health data to construct a feature dataset; then, human-caused optical features are extracted and cross-contextual needs analysis is performed to establish a spectral response model; based on this model, the photoelectric properties of quantum dot materials are mapped and spectral synthesis modeling is performed to determine the luminescence parameter space; a spectral optimization model is constructed, comprehensively considering visual comfort, energy efficiency, and health impacts for multi-objective optimization, generating a dynamic spectral control strategy; based on real-time physiological feedback, closed-loop verification and adaptive optimization are performed to form an intelligent spectral adjustment scheme. This invention fully leverages the spectral tunability advantage of quantum dot materials, realizing truly personalized intelligent lighting.

Description

Technical Field

[0001] This invention relates to the field of intelligent lighting technology, and more specifically, to an intelligent adjustment method and system for the spectrum of a quantum dot luminaire. Background Technology

[0002] With the development of human society and changes in lifestyle, artificial lighting has become an indispensable part of modern life. Modern people spend most of their time indoors, much of which is under artificial lighting. Lighting quality not only directly affects visual comfort and work efficiency, but also has a profound impact on human physiological rhythms, emotional state, and long-term health.

[0003] While traditional LED lighting has made significant progress in energy efficiency, its fixed spectral characteristics make it difficult to meet the diverse needs of different scenarios, times of day, and user groups. Inappropriate lighting environments are a major contributing factor to health problems in modern people, such as visual fatigue, sleep disorders, and mood swings.

[0004] Quantum dots, as a novel nanomaterial for luminescence, exhibit excellent optical properties due to their unique quantum size effect. By controlling the size, composition, and structure of quantum dots, their emission wavelength can be precisely tuned, achieving full-spectrum coverage from ultraviolet to near-infrared. Quantum dots also possess advantages such as high luminous efficiency, good color purity, and strong stability, making them considered ideal materials for next-generation lighting technologies. Although quantum dot materials are regarded as the core of next-generation lighting technologies due to their excellent spectral tunability, current applications remain at the level of simple color temperature adjustment, far from realizing their enormous potential in precise spectral synthesis. Current control methods are crude and lack intelligence, failing to optimize in real time according to dynamic changes in usage scenarios—such as the transition from focused work to relaxation, or the natural transition from day to night, these frequent changes in daily life cannot be effectively responded to. Furthermore, existing systems often sacrifice one aspect for another in pursuit of a certain performance indicator, either sacrificing lighting effect for eye protection or causing a surge in energy consumption for high color rendering, failing to achieve synergistic optimization of visual comfort, health protection, and energy conservation.

[0005] More importantly, traditional lighting systems lack effective feedback verification mechanisms, failing to perceive the user's actual physiological response and comfort, resulting in a "blind output" open-loop control mode. This lighting method, detached from the actual needs of the human body, not only causes significant energy waste but may also unknowingly cause long-term negative impacts on user health. In the practical application of quantum dot devices, spectral shift issues caused by factors such as temperature drift and aging degradation have not been effectively resolved, severely affecting the stability and reliability of lighting quality. The accumulation of these technical defects means that existing lighting systems are far from meeting people's urgent needs for a healthy, comfortable, and intelligent lighting environment.

[0006] In view of this, the present invention proposes an intelligent adjustment method and system for the spectrum of quantum dot light-emitting lamps to solve the above problems. Summary of the Invention

[0007] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, this application provides, on the one hand, a method for intelligently adjusting the spectrum of a quantum dot light-emitting lamp, comprising:

[0008] Step S1: Multidimensional data collection and preprocessing are performed on user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings, and physiological health data to obtain a standardized intelligent spectral control feature dataset;

[0009] Step S2: Based on the standardized intelligent spectral control feature dataset, perform human-caused optical feature extraction and cross-contextual spectral demand analysis to obtain a personalized spectral response model;

[0010] Step S3: Based on the personalized spectral response model, perform photoelectric property mapping and nonlinear spectral synthesis modeling of quantum dot materials to obtain a tunable quantum dot luminescence parameter space;

[0011] Step S4: Based on the tunable quantum dot luminescence parameter space, construct a deep reinforcement learning spectral optimization model and perform multi-objective constraint optimization to obtain a dynamic spectral control strategy;

[0012] Step S5: Based on the dynamic spectral modulation strategy, the quantum dot luminescent material is subjected to real-time precise control of electrical parameters and thermal management compensation to obtain a high-fidelity spectral output control framework;

[0013] Step S6: Based on the high-fidelity spectral output control framework and the real-time collected user physiological feedback data, perform closed-loop verification and adaptive optimization to obtain the final execution scheme for intelligent quantum dot lamp spectral adjustment.

[0014] On the other hand, this application provides an intelligent adjustment system for the spectrum of a quantum dot light-emitting lamp, which is used to implement the intelligent adjustment method for the spectrum of the quantum dot light-emitting lamp, including:

[0015] The data acquisition and preprocessing module is used to perform multi-dimensional data acquisition and preprocessing of user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings and physiological health data to obtain a standardized intelligent spectral control feature dataset.

[0016] The human-cause optics analysis module is used to extract human-cause optics features and perform cross-context spectral demand analysis based on the standardized intelligent spectral control feature dataset to obtain a personalized spectral response model.

[0017] The quantum dot spectral mapping module is used to map the photoelectric properties of quantum dot materials and model nonlinear spectral synthesis based on the personalized spectral response model, so as to obtain a tunable quantum dot luminescence parameter space.

[0018] The intelligent optimization decision module is used to construct a deep reinforcement learning spectral optimization model based on the tunable quantum dot luminescence parameter space, and perform multi-objective constraint optimization to obtain a dynamic spectral control strategy.

[0019] The precision control execution module is used to perform real-time precise control of electrical parameters and thermal management compensation of quantum dot luminescent materials based on the dynamic spectral modulation strategy, so as to obtain a high-fidelity spectral output control framework.

[0020] The closed-loop optimization feedback module is used to perform closed-loop verification and adaptive optimization based on the high-fidelity spectral output control framework and real-time collected user physiological feedback data, so as to obtain the final execution scheme of intelligent quantum dot lamp spectral adjustment.

[0021] The technical effects and advantages of the intelligent adjustment method and system for the spectrum of quantum dot light-emitting lamps of the present invention are as follows:

[0022] This invention revolutionizes the traditional "one-size-fits-all" lighting model of traditional lighting systems, achieving a personalized and healthy lighting experience. By deeply understanding and responding to each user's unique physiological rhythms, visual characteristics, and health needs, it creates a truly human-centered intelligent lighting environment. This not only significantly improves user visual comfort and work efficiency, but more importantly, it actively influences the body's physiological rhythms through scientific spectral regulation, improving sleep quality, alleviating visual fatigue, and reducing the health risks associated with long-term improper lighting. In terms of energy utilization, this invention breaks the traditional dilemma that high-quality lighting inevitably leads to high energy consumption. Through intelligent optimization, it achieves a perfect balance between lighting quality and energy efficiency, significantly reducing energy consumption while providing a superior lighting environment, making a substantial contribution to green and low-carbon development. More importantly, this invention establishes an intelligent system capable of continuous learning and evolution. It continuously gathers feedback from user experience and optimizes itself, ensuring that the lighting effect becomes increasingly tailored to user needs over time, achieving a smart lighting experience. Furthermore, this invention fully unleashes the enormous potential of quantum dots in spectral regulation, driving the lighting industry towards high-end and intelligent transformation and upgrading. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of an intelligent adjustment method for the spectrum of a quantum dot light-emitting lamp according to the present invention;

[0024] Figure 2 This is a detailed flowchart illustrating step S2 of the present invention.

[0025] Figure 3This is a detailed flowchart illustrating the implementation steps of step S4 of the present invention.

[0026] Figure 4 This is a schematic diagram of an intelligent spectral adjustment system for a quantum dot light-emitting lamp according to the present invention. Detailed Implementation

[0027] 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.

[0028] This application provides an intelligent method for adjusting the spectrum of a quantum dot luminaire. The method is applied to a related system, whose execution entities include, but are not limited to, intelligent lighting control platforms, spectrum adjustment devices, healthy lighting management platforms, and human-centric lighting controllers, which can be considered general computing nodes of this application. The spectrum adjustment platform includes, but is not limited to, at least one of a quantum dot lighting control system, a healthy lighting management system, and a human-centric optical control system.

[0029] See Figure 1 This invention provides a method for intelligently adjusting the spectrum of a quantum dot light-emitting lamp, comprising the following steps:

[0030] Step S1: Multidimensional data collection and preprocessing are performed on user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings, and physiological health data to obtain a standardized intelligent spectral control feature dataset;

[0031] Step S2: Based on the standardized intelligent spectral control feature dataset, perform human-caused optical feature extraction and cross-contextual spectral demand analysis to obtain a personalized spectral response model;

[0032] Step S3: Based on the personalized spectral response model, perform photoelectric property mapping and nonlinear spectral synthesis modeling of quantum dot materials to obtain the tunable quantum dot luminescence parameter space;

[0033] Step S4: Based on the tunable quantum dot luminescence parameter space, construct a deep reinforcement learning spectral optimization model and perform multi-objective constraint optimization to obtain a dynamic spectral modulation strategy;

[0034] Step S5: Based on the dynamic spectral modulation strategy, the electrical parameters of the quantum dot luminescent material are precisely controlled in real time and thermal management compensation is performed to obtain a high-fidelity spectral output control framework.

[0035] Step S6: Based on the high-fidelity spectral output control framework and real-time collected user physiological feedback data, perform closed-loop verification and adaptive optimization to obtain the final execution scheme for intelligent quantum dot lamp spectral adjustment.

[0036] This invention collects user biorhythms and environmental spectral information through multi-dimensional data fusion, and constructs a personalized spectral response model by combining human factor optical feature extraction technology; it achieves precise spectral synthesis using quantum dot material photoelectric property mapping technology; it generates dynamic spectral control strategies using deep reinforcement learning optimization algorithms; it ensures spectral output stability by combining real-time electrical parameter control and thermal management compensation technology; and it achieves closed-loop verification and adaptive optimization through user physiological feedback data, ultimately realizing high-precision personalized spectral adjustment for different users and different scenarios.

[0037] In this embodiment of the invention, the steps of the intelligent adjustment method for the spectrum of quantum dot light-emitting lamps include:

[0038] Step S1: Multidimensional data collection and preprocessing are performed on user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings, and physiological health data to obtain a standardized intelligent spectral control feature dataset;

[0039] In this embodiment, wearable devices are first used to collect users' circadian rhythm indicators and sleep cycle data to obtain raw biorhythm data. Photoelectric sensors and accelerometers are used to continuously monitor users' activity patterns, sleep quality, and circadian rhythm markers (such as body temperature changes and activity intensity cycles). Time series analysis is performed on the collected raw biorhythm data, and Fourier transform and wavelet analysis are applied to extract periodic features, constructing an individual biorhythm model, including sleep-wake cycles, energy level fluctuation patterns, and markers for rhythm-sensitive periods. Then, a spectral sensor array is used to monitor ambient lighting conditions in real time, obtaining raw environmental spectral parameters. Color temperature analysis and light intensity calculations are performed on these parameters to construct an environmental spectral distribution map, extracting key features such as dominant light source type, color temperature range, light intensity gradient, and flicker index, forming structured environmental optical characteristic data. A context-aware system identifies the current usage scenario of the lighting fixtures, obtaining raw scenario information. This raw scenario information is then semantically classified and activity-labeled, establishing a scenario-activity association model and generating structured scenario data containing scenario type, activity nature, temporal attributes, and spatial features.

[0040] User lighting preference settings are collected through a user interface to obtain raw user preference data. This raw data is then processed into feature vectors, and principal component analysis and clustering algorithms are applied to extract user preference patterns in different scenarios, establishing a multidimensional user preference feature model. User visual comfort and emotional state information are collected through a non-invasive sensing system to obtain raw physiological health data. Physiological index correlation analysis is performed on this raw physiological health data, integrating pupil response, blink frequency, attention score, and emotional state indicators to generate a health status assessment result. Finally, multidimensional feature fusion and standardization are performed on the user's biological cycle model, environmental optical characteristic data, structured scene data, user preference feature model, and health status assessment result. Tensor decomposition and feature selection algorithms are applied to construct a unified feature representation framework, normalizing the data across all dimensions to eliminate dimensional differences, ultimately obtaining a standardized intelligent spectral control feature dataset.

[0041] Step S2: Based on the standardized intelligent spectral control feature dataset, perform human-caused optical feature extraction and cross-contextual spectral demand analysis to obtain a personalized spectral response model;

[0042] In this embodiment, principal component analysis (PCA) is first applied to the standardized intelligent spectral control feature dataset to extract the main factors that explain data variability, identify the variable combinations that have the most significant impact on spectral perception, and obtain key optical influence factors. Based on these key optical influence factors, a user visual sensitivity feature map is constructed. By analyzing the differences in user sensitivity to different wavelengths, brightness, and contrast, individual visual response curves are plotted, forming personalized visual response characteristics. Based on these personalized visual response characteristics, spectral sensitive regions are identified. Differential response analysis is used to determine the user's sensitivity to specific bands in the spectrum, obtaining key wavelength ranges. A refined analysis of the key wavelength range is performed using high-resolution spectral scanning and response measurement techniques to construct a wavelength-sensitivity function and generate wavelength sensitivity distribution curves. Cross-contextual data mining techniques are used to cluster spectral requirements in different scenarios, identifying spectral feature requirements for typical scenarios such as work, leisure, reading, and dining, obtaining scenario-based spectral requirement prototypes. Based on these scenario-based spectral requirement prototypes, a context-transitional spectral adaptability model is established. By modeling spectral gradation strategies during scenario transitions, visual comfort and functional adaptability are ensured, forming a context-adaptive spectral framework. Time-series data of user preferences is extracted from raw data, recording users' spectral preference adjustment behaviors under different time periods and environmental conditions, resulting in a record of user illumination preference changes. Pattern recognition analysis is performed on these records, and time-series pattern mining algorithms are applied to discover preference evolution patterns, identify stable preferences and changing trends, and obtain the regular characteristics of user spectral preferences. Based on these regular characteristics, a time-series prediction neural network is constructed, and a Long Short-Term Memory (LSTM) network is used to model the time dependence of user preferences, resulting in a preference trend prediction model. This model is then used to predict users' future spectral needs based on time, environment, and activity factors, generating dynamic spectral preference prediction results. Finally, personalized visual response features, wavelength sensitivity distribution curves, context-adaptive spectral frameworks, and dynamic spectral preference prediction results are fused and modeled at multiple levels. A hierarchical Bayesian network is used to integrate the sub-models, constructing a unified personalized spectral response model.

[0043] Step S3: Based on the personalized spectral response model, perform photoelectric property mapping and nonlinear spectral synthesis modeling of quantum dot materials to obtain the tunable quantum dot luminescence parameter space;

[0044] In this embodiment, the target spectral feature requirements are first extracted based on a personalized spectral response model to determine key parameters such as peak wavelength, full width at half maximum (FWHM), color temperature range, and color rendering index of the ideal spectrum, resulting in an ideal spectral parameter set. This ideal spectral parameter set is then converted into selection constraints for quantum dot luminescent materials, establishing a correspondence between spectral parameters and quantum dot material properties. Selection criteria such as material composition, size, and structure are determined, resulting in material screening standards. Based on these standards, a characteristic matching analysis is performed on the quantum dot material library, involving a multi-dimensional parameter space search to screen candidate material combinations that meet the spectral requirements, resulting in candidate quantum dot material combinations. The photoelectric properties of these candidate quantum dot material combinations are characterized by measuring parameters such as quantum yield, emission peak position, FWHM, excitation-emission relationship, and temperature stability, obtaining material photoelectric conversion characteristic data. Finally, the personalized visual response features in the personalized spectral response model are correlated and mapped with the material photoelectric conversion characteristic data, establishing a correspondence between user visual perception characteristics and quantum dot material luminescence characteristics, resulting in a visual perception-material characteristic correspondence. A parameterized photoelectric control model is established based on the correspondence between visual perception and material properties. Mapping functions between control parameters such as current density, voltage, and temperature and the output spectrum are constructed, yielding a precise spectral control mapping function. Using this precise spectral control mapping function, a synergistic luminescence scheme for multiple quantum dot materials is designed. The ratio and spatial arrangement of different materials are optimized, and a multilayer quantum dot luminescence structure is designed, resulting in a spectral synthesis strategy. The spectral synthesis strategy is modeled using a nonlinear spectral superposition effect, considering energy transfer, reabsorption, and quantum confinement effects between quantum dot materials. A mathematical model of complex spectral mixing is established, yielding a spectral mixing theoretical model. Based on the spectral mixing theoretical model and the context-adaptive spectral framework in the personalized spectral response model, a multidimensional control space encompassing parameters such as current density, driving voltage, pulse width, and modulation frequency is constructed, resulting in a continuously tunable quantum dot luminescence parameter domain. This continuously tunable quantum dot luminescence parameter domain is converted into a range of control parameters that can be realized by actual hardware. A mapping relationship between digital control commands and actual spectral output is established, resulting in a tunable quantum dot luminescence parameter space.

[0045] Step S4: Based on the tunable quantum dot luminescence parameter space, construct a deep reinforcement learning spectral optimization model and perform multi-objective constraint optimization to obtain a dynamic spectral modulation strategy;

[0046] In this embodiment, the tunable quantum dot luminescence parameter space is first converted into a high-dimensional spectral state representation. A state vector containing features such as spectral distribution, color temperature, color rendering index, light intensity, and wavelength distribution is constructed to obtain a quantum dot spectral state descriptor. Based on the quantum dot spectral state descriptor, a reinforcement learning state representation method is designed, defining a state transition function and a reward calculation mechanism to establish a mathematical description of the state space, resulting in a parameterized state evolution model. Based on the physical boundary of the tunable quantum dot luminescence parameter space, a quantum dot spectral control action space is defined. The continuous parameter space is discretized into a finite set of actions, including basic operations such as current adjustment, wavelength selection, and intensity adjustment, resulting in a discretized quantum dot control instruction set. A bidirectional mapping relationship is established between the discretized quantum dot control instruction set and the tunable quantum dot luminescence parameter space to ensure that instructions can be accurately converted into hardware control signals, and that execution instructions can be inferred from parameter changes, thus obtaining a quantum dot spectral control implementation method. Combining the parameterized state evolution model and the quantum dot spectral control implementation method, a state-action-reward interaction framework is constructed, and an environmental feedback mechanism is established to obtain a reinforcement learning interaction model. A multi-objective reward function is designed based on a reinforcement learning interactive model, considering three dimensions: visual comfort, energy efficiency, and health impact. Sub-reward functions are designed for each dimension and then weighted and fused to obtain a comprehensive evaluation index for quantum dot spectroscopy. A reinforcement learning value network is constructed based on this comprehensive evaluation index, employing a dual-network architecture (policy network and value network). The spectral optimization objective is transformed into maximizing cumulative reward, resulting in a spectral optimization objective function. Spectral simulations are performed using the tunable quantum dot luminescence parameter space to simulate the spectral output effects under different control parameters, generating a large-scale state-action-reward sample library for quantum dot spectroscopy. A deep neural network is trained based on this library and the spectral optimization objective function to implement the policy network and value network. Specific network structures and training algorithms (such as PPO and DDPG) are used, and network layer structures and activation functions are designed specifically for quantum dot characteristics, resulting in a deep reinforcement learning model structure adapted to quantum dot characteristics. The deep reinforcement learning model structure is simulated, tested, and iteratively optimized in a virtual environment constructed using the tunable quantum dot luminescence parameter space. Extensive simulated interactions improve model performance, yielding a basic quantum dot spectral control model. By utilizing online learning technology combined with real-time spectral feedback to continuously optimize the basic quantum dot spectral control model, the model parameters are dynamically updated to adapt to changes in user preferences and material property drift, resulting in an adaptive quantum dot spectral control strategy, and thus forming a complete dynamic spectral control strategy.

[0047] Step S5: Based on the dynamic spectral modulation strategy, the electrical parameters of the quantum dot luminescent material are precisely controlled in real time and thermal management compensation is performed to obtain a high-fidelity spectral output control framework.

[0048] In this embodiment, the dynamic spectral modulation strategy is first converted into a current-driven parameter set. The abstract strategy decision is then transformed into specific electrical control parameters such as voltage, current, pulse width, and frequency, resulting in control commands for the quantum dot driving circuit. Based on these control commands, a pulse width modulation scheme is designed, optimizing the PWM waveform, frequency, and duty cycle parameters to achieve sub-microsecond-level fine-grained current control, thus obtaining a fine-grained current control strategy. The temperature distribution of the quantum dot light-emitting device is monitored in real time using a micro-thermocouple array or infrared imaging technology to record temperature changes in various regions of the device with high spatiotemporal resolution, obtaining thermal field distribution data. Based on this data, hotspot identification and heat flow analysis are performed. A heat conduction model is applied to identify regions with abnormal temperature gradients, and the heat flow propagation path is analyzed to determine the device's thermal management requirements. An active heat dissipation control scheme is designed based on these requirements, combining technologies such as microchannel liquid cooling, phase change materials, or high-efficiency heat sinks to precisely dissipate heat in hotspot areas, resulting in temperature regulation commands. These commands are then converted into heat dissipation system control signals to control fan speed, liquid pump flow rate, or semiconductor cooler power, achieving precise temperature regulation and resulting in a thermal management execution scheme. The actual output spectrum is acquired in real time using a spectral feedback sensor. The output spectrum is compared with the target spectrum in real time, and the deviation value at each wavelength point is calculated to obtain spectral deviation data. Based on the spectral deviation data, the spectral error correction amount is calculated, and a compensation strategy is designed using a feedback control algorithm to dynamically adjust the driving parameters, thereby obtaining the spectral closed-loop correction parameters. The refined current control strategy, thermal management execution scheme, and spectral closed-loop correction parameters are integrated into a unified control framework. A multi-level control loop is designed, and a collaborative optimization mechanism between parameters is established to obtain a high-fidelity spectral output control framework.

[0049] Step S6: Based on the high-fidelity spectral output control framework and real-time collected user physiological feedback data, perform closed-loop verification and adaptive optimization to obtain the final execution scheme for intelligent quantum dot lamp spectral adjustment.

[0050] In this embodiment, an initial spectral configuration is first output using a high-fidelity spectral output control framework. A baseline spectral setting is then generated based on a personalized spectral response model and a dynamic spectral adjustment strategy, resulting in a baseline spectral state. Simultaneously, a physiological sensor array collects pupillary response and heart rate variability data of the user under the baseline spectral state, monitoring the user's physiological responses in a specific spectral environment and obtaining raw physiological feedback data. Signal processing and feature extraction are performed on the raw physiological feedback data. Wavelet transform and feature extraction algorithms are applied to identify the fluctuation patterns of physiological indicators caused by spectral changes, obtaining the spectral-affected physiological indicators. A correlation analysis is performed between the spectral-affected physiological indicators and the spectral closed-loop correction parameters in the high-fidelity spectral output control framework to establish a mapping relationship between spectral parameter adjustments and physiological response changes, constructing a physiological-spectral response relationship model. A spectral adjustment strategy is generated based on the physiological-spectral response relationship model. An optimization algorithm calculates the spectral parameter combination that maximizes physiological comfort, resulting in a physiologically driven optimization scheme. This physiologically driven optimization scheme is converted into parameter adjustment instructions for the high-fidelity spectral output control framework, updating the current density distribution, pulse width, and frequency parameters in the refined current control strategy. Simultaneously, the temperature control target and heat dissipation power allocation in the thermal management execution scheme are optimized, resulting in a physiologically sensitive control parameter set. By deploying a set of physiologically sensitive control parameters onto a high-fidelity spectral output control framework, precise spectral control based on physiological feedback is achieved. At the same time, a long-term physiological feedback database of users is established to record physiological response patterns under different spectral configurations. The spectral closed-loop correction parameters of the high-fidelity spectral output control framework are continuously optimized to obtain the final execution scheme for intelligent quantum dot lamp spectral adjustment.

[0051] In this embodiment, the detailed implementation steps of step S1 include:

[0052] By collecting users' circadian rhythm indicators and sleep cycle data through wearable devices, raw biorhythm data is obtained, and time series analysis is performed on the raw biorhythm data to obtain the user's biocycle model.

[0053] Ambient lighting conditions are monitored in real time using a spectral sensor array to obtain raw environmental spectral parameters. Color temperature analysis and light intensity calculation are then performed on these raw environmental spectral parameters to obtain environmental optical property data.

[0054] The current usage scenario of the lighting fixtures is identified by the context-aware system, the original scenario information is obtained, and the original scenario information is semantically classified and activity-tagged to obtain structured scenario data.

[0055] By collecting user lighting preference settings through the user interface, the raw data of user preferences is obtained, and the raw data of user preferences is processed into feature vectors to obtain the user preference feature model.

[0056] By collecting user visual comfort and emotional state information through a non-invasive sensing system, raw physiological health data is obtained, and physiological index correlation analysis is performed on the raw physiological health data to obtain health status assessment results.

[0057] Multidimensional feature fusion and standardization processing are performed on user biocycle model, environmental optical property data, structured scene data, user preference feature model and health status assessment results to obtain a standardized intelligent spectral control feature dataset.

[0058] In this embodiment, devices such as smart bracelets, optical sensor watches, or wearable patch sensors are used to continuously monitor the user's activity patterns, sleep status, and physiological parameters through multispectral photoelectric sensors, accelerometers, and temperature sensors. The collected circadian rhythm indicators include indirect indicators of melatonin levels (such as body temperature change curves), skin conductivity, activity intensity cycles, and sleep stage markers. The data collection frequency is 10-60 seconds per instance, forming a continuous stream of raw circadian rhythm data. Time series analysis methods, including Fourier transform, autoregressive moving average models, and wavelet analysis, are applied to the raw circadian rhythm data to extract the periodic characteristics and patterns of the user's circadian rhythm. An individualized 24-hour circadian rhythm model is constructed, including wake-sleep transition points, peak energy levels, light-sensitive periods, and mood fluctuation patterns, forming a complete user circadian rhythm model.

[0059] A distributed spectral sensor array is used to monitor ambient lighting conditions. The sensors employ a small spectrometer chip with a wavelength detection range of 380-780 nm and a resolution better than 2 nm. Deployed around lighting fixtures and in user activity areas, the sensors collect environmental spectral data at a frequency of 5-60 seconds per acquisition, recording spectral distribution, light intensity, and temporal variation trends to obtain raw environmental spectral parameters. Color temperature analysis is performed on these raw parameters to calculate parameters such as correlated color temperature (CCT), color coordinates (CIExy), and color rendering index (CRI). Simultaneously, light intensity is calculated, measuring illuminance level (lux), uniformity of brightness distribution, and temporal fluctuation characteristics. Features such as dominant light source type (natural light, fluorescent lamps, LEDs, etc.), flicker index, and blue light content are extracted to construct environmental optical characteristic data.

[0060] A context-aware system identifies the current usage scenario of lighting fixtures. This system integrates multiple sensing technologies, including environmental acoustic sensors, motion detectors, temperature and humidity monitoring, and IoT device status recognition. Sensors collect environmental sound characteristics, human activity patterns, indoor environmental parameters, and the status of associated devices to form raw scenario information. Semantic classification technology is used to process this raw scenario information, applying a deep learning classifier to map environmental features to predefined scenario categories (such as work, reading, entertainment, rest, etc.). Simultaneously, activity tagging is performed to identify the specific type of activity the user is currently engaged in, extracting the activity's temporal attributes (duration, repetition pattern) and spatial features (location, posture), ultimately forming structured scenario data containing scenario type, activity nature, purpose of use, and environmental conditions.

[0061] Personalized user preferences for lighting environments are collected through user interfaces such as smartphone apps, voice assistants, or smart home control panels. Manually adjusted spectral parameters by users in different scenarios are recorded, including brightness levels, color temperature selection, spectral distribution preferences, and dynamic changes, forming raw user preference data. This raw data is then processed into feature vectors, and principal component analysis and clustering algorithms are applied to extract core feature dimensions of user preferences. A scenario-preference mapping relationship is established to identify stable preference patterns and time-dependent characteristics in different scenarios. A multi-dimensional user preference feature model is constructed, including dimensions such as brightness preference range, color temperature preference distribution, spectral characteristic preferences, and tolerance for dynamic changes, forming a complete user preference feature model.

[0062] This system collects user visual comfort and emotional state information through a non-invasive sensing system, including a remote pupil monitoring camera, a facial expression analysis module, and a behavior pattern tracker. It monitors changes in pupil size, blink frequency, fixation behavior, and facial micro-expressions to assess visual fatigue and emotional state, obtaining raw physiological health data. Correlation analysis of physiological indicators is performed on this raw data, and a physiological-psychological model is applied to analyze the relationship between pupillary response and light comfort, the correlation between blink patterns and visual fatigue, and the matching of attention levels with the lighting environment. By integrating these indicators, a visual comfort score and emotional state mapping are constructed to obtain the health status assessment results.

[0063] Multidimensional feature fusion and standardization were performed on user biocycle models, environmental optical property data, structured scene data, user preference feature models, and health status assessment results. Tensor decomposition technology was applied to process multi-source heterogeneous data and extract cross-dimensional feature correlations. Feature selection and dimensionality reduction were performed to screen key features affecting spectral regulation. Data from each dimension were normalized to eliminate dimensional differences and achieve feature scale uniformity. A unified feature representation framework was constructed to form a structured and standardized intelligent spectral control feature dataset, providing a foundation for subsequent analysis.

[0064] See Figure 2 The diagram below illustrates the detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include:

[0065] Principal component analysis algorithm is applied to the standardized intelligent spectral control feature dataset to obtain key optical influence factors, and user visual sensitivity feature map is constructed based on the key optical influence factors to obtain personalized visual response features;

[0066] Based on personalized visual response characteristics, spectral sensitive regions are identified to obtain key wavelength ranges, and a refined analysis of the key wavelength ranges is performed to obtain wavelength sensitivity distribution curves.

[0067] Cross-scenario data mining techniques are used to cluster the spectral requirements under different scenarios to obtain scenario spectral requirement prototypes. Based on the scenario spectral requirement prototypes, a scenario-transition spectral adaptive model is established to obtain a scenario-adaptive spectral framework.

[0068] Time series data of raw user preference data are extracted to obtain records of changes in user lighting preferences. Pattern recognition analysis is then performed on these records to obtain the characteristics of user spectral preference patterns.

[0069] A time-series prediction neural network is constructed based on the characteristics of user spectral preferences to obtain a preference trend prediction model. The preference trend prediction model is then used to generate future spectral demand predictions, resulting in dynamic spectral preference prediction results.

[0070] By fusing personalized visual response features, wavelength sensitivity distribution curves, context-adaptive spectral frameworks, and dynamic spectral preference prediction results into a multi-level model, a personalized spectral response model is obtained.

[0071] In this embodiment, Principal Component Analysis (PCA) is applied to the standardized intelligent spectral control feature dataset to reduce the dimensionality of the feature space and extract the main factors that can explain data variability. The feature covariance matrix is ​​calculated, decomposing eigenvalues ​​and eigenvectors. The number of principal components is selected based on the cumulative explained variance ratio (>90%). Principal component loadings are analyzed to identify the variable combinations that have the most significant impact on spectral perception, determining key optical influencing factors, including wavelength-dependent factors, brightness sensitivity factors, and time modulation factors. Based on these key optical influencing factors, a user visual sensitivity feature map is constructed. Using controlled variable testing methods, the system evaluates the differences in user sensitivity to changes in wavelength, brightness gradient, and contrast. Combining pupil response data and subjective evaluation, individual visual response curves are plotted, marking high-sensitivity and low-sensitivity regions to form personalized visual response characteristics.

[0072] This study identifies spectrally sensitive regions based on personalized visual response characteristics, employing differential response analysis to perform spectral scanning tests within the 380-780 nm visible light range at 5 nm step sizes. It analyzes user sensitivity to specific wavelengths in the spectrum, identifies wavelength regions that elicit strong visual responses, and determines key wavelength ranges. Refined analysis of these key wavelength ranges is then conducted using high-precision spectral scanning at 1 nm resolution, combined with fine brightness gradient testing (10-1000 lux, 5 lux step size), to construct a three-dimensional wavelength-brightness-sensitivity mapping relationship. An interpolation algorithm is applied to generate a continuous sensitivity response surface, extracting wavelength sensitivity distribution curves under different lighting conditions.

[0073] This study utilizes cross-contextual data mining techniques to cluster and analyze spectral requirements across different scenarios, integrating user spectral preference data for typical scenarios such as work, study, leisure, dining, and bedtime. Applying K-means or hierarchical clustering algorithms, scenarios are grouped based on spectral parameter similarity. Spectral feature centers of each scenario group are extracted to identify characteristic spectral requirements for typical scenarios, forming scenario-based spectral requirement prototypes. Based on these prototypes, a context-transition spectral adaptability model is established, designing spectral gradient strategies during scenario transitions and modeling smooth transition curves for spectral parameters. Optimal transition times and paths are customized for different transition types (e.g., from work to rest, indoor to outdoor) to ensure visual comfort and functional adaptability, forming a context-adaptive spectral framework.

[0074] Time-series data of raw user preference data is extracted to construct a time-preference data matrix, recording users' spectral preference adjustment behavior under different time periods (morning, noon, afternoon, evening, and night) and different environmental conditions (sunny, cloudy, indoor, and outdoor), thus obtaining records of user lighting preference changes. Pattern recognition analysis is performed on these records, and time-series pattern mining algorithms (such as Dynamic Time Warping (DTW)) are applied to discover preference evolution patterns and identify periodic patterns and triggering conditions. Combining frequency analysis and variability detection, stable preferences and changing trends are distinguished, and the regularity characteristics of user spectral preferences are constructed.

[0075] A time-series prediction neural network is constructed based on the characteristics of user spectral preferences, employing a Long Short-Term Memory (LSTM) network structure to model the time dependence and long-term memory characteristics of user preferences. The network input includes time features, environmental conditions, activity type, and historical preference data, with the output being predicted spectral parameters (color temperature, brightness, and spectral distribution). The network is trained using supervised learning, with 80% of the historical data used as the training set and 20% as the validation set, resulting in a preference trend prediction model. This model, combined with current conditions and historical preferences, predicts spectral demand in different time periods and scenarios in the future, generating dynamic spectral preference prediction results with a timeline.

[0076] This paper presents a multi-level fusion model of personalized visual response features, wavelength sensitivity distribution curves, context-adaptive spectral framework, and dynamic spectral preference prediction results, integrating the sub-models using a hierarchical Bayesian network structure. A probabilistic graphical model is constructed to represent the dependencies between variables, establishing an inference chain from environmental conditions and user state to spectral requirements. The EM algorithm is applied to optimize model parameters, balancing the weight contributions of each sub-model under different conditions. Sensitivity analysis verifies the model's stability, ultimately resulting in a personalized spectral response model capable of adapting to various changing conditions.

[0077] In this embodiment, the detailed implementation steps of step S3 include:

[0078] Based on the personalized spectral response model, the target spectral feature requirements are extracted to obtain the ideal spectral parameter set. The ideal spectral parameter set is then converted into selection constraints for quantum dot luminescent materials to obtain material screening criteria.

[0079] Based on the material screening criteria, the quantum dot material library was subjected to property matching analysis to obtain candidate quantum dot material combinations. The photoelectric properties of the candidate quantum dot material combinations were then characterized to obtain photoelectric conversion characteristic data of the materials.

[0080] The personalized visual response features in the personalized spectral response model are correlated and mapped with the photoelectric conversion characteristics of materials to obtain the correspondence between visual perception and material properties. Based on the correspondence between visual perception and material properties, a parameterized photoelectric control model is established to obtain the accurate spectral control mapping function.

[0081] By using a precise spectral modulation mapping function, a synergistic luminescence scheme for multiple quantum dot materials is designed, resulting in a spectral synthesis strategy. The nonlinear spectral superposition effect of the spectral synthesis strategy is then modeled to obtain a spectral mixing theoretical model.

[0082] Based on the context-adaptive spectral framework in the spectral mixing theory model and the personalized spectral response model, a spectral dynamic control parameter space is constructed to obtain a continuously tunable quantum dot luminescence parameter domain. This continuously tunable quantum dot luminescence parameter domain is then mapped to the actual control parameter range to obtain a controllable quantum dot luminescence parameter space.

[0083] In this embodiment, the target spectral feature requirements are first extracted based on a personalized spectral response model. The optimal spectral features output by the model are analyzed to determine the key parameters of the ideal spectrum. Parameters such as peak wavelength (dominant and secondary wavelengths), full width at half maximum (FWHM), color temperature range (CCT), color rendering index (CRI and R9), spectral purity, blue light ratio, and flicker index are extracted to form a multidimensional ideal spectral parameter set. This ideal spectral parameter set is then converted into selection constraints for quantum dot luminescent materials, establishing a correspondence between spectral parameters and quantum dot material properties. Selection criteria such as material composition (e.g., CdSe / ZnS, InP / ZnS, PeNC), particle size range (2-12nm), core-shell structure (core-shell ratio, shell thickness), surface ligand type, and dispersion medium requirements are set to form a structured material screening standard.

[0084] A property matching analysis was performed on a quantum dot material library based on material screening criteria. The library contains property data for quantum dot materials with different compositions, sizes, and structures. A multi-dimensional parameter space search was conducted, and genetic algorithms or simulated annealing algorithms were applied to optimize material combination selection, screening candidate material combinations that meet spectral requirements. Considering material compatibility, stability, and fabrication feasibility, preferred candidate quantum dot material combinations were formed. Comprehensive photoelectric properties were characterized for the candidate material combinations, measuring parameters such as quantum yield (PLQY), emission peak position, full width at half maximum (FWHM), excitation-emission spectrum, fluorescence lifetime, temperature dependence (-20°C to 85°C), and photoelectric conversion efficiency. Aging tests were conducted to evaluate material stability and obtain complete photoelectric conversion characteristic data.

[0085] This paper establishes a correlation between personalized visual response features in a personalized spectral response model and photoelectric conversion property data of materials, thereby establishing a correspondence between user visual perception characteristics and the luminescence properties of quantum dot materials. The matching degree between user sensitivity to specific wavelengths and quantum dot emission peaks is analyzed to establish a correlation function between visual comfort scores and material spectral parameters, forming a visual perception-material property correspondence model. Based on this model, a parameterized photoelectric control model is established, constructing mapping functions between control parameters such as current density, voltage, and temperature and the output spectrum. A nonlinear mapping relationship is established using multilayer neural networks or high-order polynomial fitting methods, considering the electric field dependence and temperature dependence of the material, to obtain a precise spectral control mapping function.

[0086] A synergistic luminescence scheme for multiple quantum dot materials was designed using a precise spectral modulation mapping function, optimizing the ratio (e.g., blue:green:red) and spatial arrangement of materials with different emission peaks. A multilayer quantum dot luminescent structure was designed, considering energy transfer efficiency and light extraction efficiency, to form a material combination scheme capable of producing ideal spectra, thus obtaining a spectral synthesis strategy. The spectral synthesis strategy was modeled using a nonlinear spectral superposition effect, considering Förster resonance energy transfer (FRET), reabsorption effect, exciton-exciton interaction, and quantum confinement effect between quantum dot materials. A mathematical model of complex spectral mixing was established to predict the actual spectral output when multiple quantum dot materials co-emitte, resulting in a theoretical model of spectral mixing.

[0087] Based on the context-adaptive spectral framework of the spectral mixing theory model and the personalized spectral response model, a multi-dimensional control space is constructed, encompassing parameters such as current density (0.1-100 mA / cm²), driving voltage (2-12 V), pulse width (1 μs-continuous), and modulation frequency (0-10 kHz). For different contextual requirements, parameter control strategies are designed, and a scenario-parameter mapping relationship is established to obtain a continuously tunable quantum dot luminescence parameter domain. This parameter domain is then converted into a controllable parameter range that can be implemented in actual hardware. Considering the limitations of the driving circuit, heat dissipation capacity, and material stability boundaries, a mapping relationship is established between digital control commands (such as 8-16 bit PWM signals) and the actual spectral output, ultimately yielding the tunable quantum dot luminescence parameter space.

[0088] In this embodiment, a characteristic matching analysis of the quantum dot material library is performed according to the material screening criteria to obtain candidate quantum dot material combinations, including:

[0089] A quantum dot material property database containing quantum dot materials of different sizes, different component ratios and different core-shell structures was constructed to obtain a quantum dot material library. The quantum dot material library was then classified by emission peak position and quantum efficiency rating to obtain a structured material retrieval framework.

[0090] The material screening criteria are decomposed into multi-dimensional screening conditions, including emission wavelength requirements, color purity index, quantum efficiency threshold, and stability parameters, to obtain a parameterized screening vector. Based on the parameterized screening vector, a multi-level screening algorithm is designed to obtain a material matching search strategy.

[0091] Parallel retrieval is performed within a structured material retrieval framework using a material matching search strategy to obtain a preliminary set of materials, and a theoretical spectral combination scheme is obtained by applying a spectral superposition simulation algorithm.

[0092] Spectral coverage and color gamut range analysis were performed on the theoretical spectral combination schemes to obtain spectral performance evaluation results. Based on the spectral performance evaluation results, the schemes were ranked and screened to obtain the optimal combination scheme set.

[0093] The optimal combination scheme set is evaluated for material compatibility and co-processing feasibility to obtain a process compatibility score. Based on the process compatibility score and the spectral performance evaluation results, a comprehensive assessment is made to obtain the candidate quantum dot material combination.

[0094] In this embodiment, a database of quantum dot material properties with different sizes, component ratios, and core-shell structures is first constructed. This database covers common II-VI group quantum dot systems (such as CdSe / ZnS, CdS / ZnS) and III-V group quantum dot systems (such as InP / ZnS, InAs / ZnSe). For each material system, comprehensive data on particle size range of 2-12 nm (0.5 nm increments), core-shell ratio (1:0.5 to 1:5), surface ligand type (oleic acid, oleylamine, thioglycolic acid, etc.), and synthesis method (thermal injection, microfluidic, template method, etc.) are collected to form a quantum dot material library. The quantum dot material library is systematically classified and rated, and the materials are divided into four categories according to the emission peak position: blue light region (400-480 nm), green light region (480-550 nm), yellow light region (550-600 nm), and red light region (600-700 nm). Based on quantum yield (PLQY), materials are classified into five grades: premium (>90%), first-grade (80-90%), second-grade (70-80%), third-grade (60-70%), and basic grade (<60%). A multi-level index structure is constructed to establish a rapid query mapping between material properties and luminescence performance, forming a structured material retrieval framework.

[0095] The material screening criteria were decomposed into multi-dimensional screening conditions, including: emission wavelength requirements (center wavelength ±5nm), full width at half maximum (FWHM) limits (<30nm preferred), color purity index (color coordinate deviation <0.02), quantum efficiency threshold (>70% under operating conditions), thermal stability requirements (efficiency reduction <15% from -20°C to 85°C), photostability parameter (efficiency reduction <10% after 1000 hours of continuous illumination), and chemical stability index (no significant performance degradation after 1000 hours in a humid and hot environment). These conditions were quantified into a standardized parameter vector, with each parameter assigned an importance weight (emission wavelength 0.3, color purity 0.2, quantum efficiency 0.25, stability 0.25), forming a parameterized screening vector. Based on this vector, a multi-level screening algorithm was designed, employing a hierarchical screening strategy: the first level is hard screening (meeting emission wavelength and minimum quantum efficiency requirements), the second level is soft screening (sorted by color purity and FWHM), and the third level is comprehensive scoring (weighted calculation of scores for each parameter), resulting in a material matching search strategy.

[0096] Parallel retrieval is performed within a structured materials retrieval framework using a material matching search strategy. A distributed computing architecture is employed to simultaneously search multiple material categories, improving screening efficiency. A three-stage retrieval process is implemented: classification filtering, attribute matching, and similarity ranking. First, the required combination of emission peaks (e.g., requiring three main peaks at 450nm, 530nm, and 620nm) is determined based on the target spectral characteristics. Candidate materials meeting the criteria are then screened within the corresponding categories to form a preliminary material set. A spectral superposition simulation algorithm is applied to this set, using the Monte Carlo method to generate 1000 different material combinations and ratios, simulating the composite spectral output under each scheme. Considering energy transfer effects between quantum dots, reabsorption phenomena, and excitation spectral overlap, an accurate spectral superposition model is constructed to predict actual luminescence effects, yielding theoretical spectral combination schemes.

[0097] A comprehensive spectral performance evaluation was conducted on the theoretical spectral combination schemes, calculating the spectral coverage (overlap integral between the target spectrum and the actual spectrum, with an excellent standard of >90%) and color gamut (the proportion of the area covered on the CIE 1931 chromaticity diagram, with a target of >85% NTSC). The color rendering index performance (Ra and special color rendering indices R9-R15) was analyzed, and the adjustable color temperature range (e.g., 2700K-6500K) and adjustment accuracy (±100K) were evaluated. A multi-dimensional scoring model was constructed to quantify the spectral reproduction capability, color performance, and dynamic control potential of each scheme, yielding the spectral performance evaluation results. Based on these results, the Pareto ranking algorithm was applied to identify non-dominated solutions that performed well across multiple objectives. A weighted scoring method was then used to rank the schemes, selecting the top 30 schemes with the best overall performance to form the optimal combination scheme set.

[0098] Material compatibility and co-processing feasibility were assessed for the preferred combination schemes, and the implementation difficulty of each scheme was analyzed from the perspectives of materials science and process engineering. The dispersion stability of different quantum dot materials in the same matrix was evaluated (solvent compatibility, surface charge compatibility), and the chemical compatibility between materials was analyzed (to avoid performance degradation caused by ligand exchange and ion migration). The difficulty of co-processing was evaluated, including the control of mixing uniformity, the difficulty of multilayer structure fabrication, and the complexity of encapsulation processes, generating a process compatibility score of 1-10 for each scheme. Finally, a comprehensive evaluation function was constructed based on the process compatibility score and spectral performance evaluation results: S = 0.7 × P + 0.3 × C, where S is the comprehensive score, P is the spectral performance score (0-100), and C is the process compatibility score (0-100, converted from the original 1-10 score). The top 5-10 schemes with the highest comprehensive scores were selected as candidate quantum dot material combinations, ensuring that the selected schemes possess both excellent spectral performance and reliable process implementation paths.

[0099] See Figure 3The diagram below illustrates the detailed implementation steps of step S4. In this embodiment, the detailed implementation steps of step S4 include:

[0100] The tunable quantum dot emission parameter space is converted into a high-dimensional spectral state representation to obtain a quantum dot spectral state descriptor. Based on the quantum dot spectral state descriptor, a reinforcement learning state representation method is designed to obtain a parameterized state evolution model.

[0101] Based on the physical boundary of the tunable quantum dot luminescence parameter space, a quantum dot spectral modulation action space is defined to obtain a discretized quantum dot control instruction set. A two-way mapping relationship between the discretized quantum dot control instruction set and the tunable quantum dot luminescence parameter space is established to obtain a quantum dot spectral modulation implementation method.

[0102] By combining the parameterized state evolution model and the quantum dot spectral modulation method, a quantum dot spectral state-action transition framework is constructed to obtain a reinforcement learning interaction model. Based on the reinforcement learning interaction model, a multi-objective reward function is designed, taking into account three dimensions: visual comfort, energy efficiency, and health impact, to obtain a comprehensive evaluation index for quantum dot spectroscopy.

[0103] A reinforcement learning value network is constructed based on the comprehensive evaluation index of quantum dot spectroscopy, the spectral optimization objective function is obtained, and spectral simulation is performed using the adjustable quantum dot luminescence parameter space to generate a training dataset and obtain a quantum dot spectral state-action sample library.

[0104] Based on a quantum dot spectral state-action sample library and a spectral optimization objective function, a deep neural network is trained to realize the policy network and value network, resulting in a deep reinforcement learning model structure adapted to the characteristics of quantum dots. In a virtual environment constructed with a tunable quantum dot luminescence parameter space, the deep reinforcement learning model structure is simulated, tested and iteratively optimized to obtain a basic quantum dot spectral control model.

[0105] By utilizing online learning technology combined with real-time spectral feedback to continuously optimize the basic quantum dot spectral control model, an adaptive quantum dot spectral control strategy is obtained, thereby forming a complete dynamic spectral control strategy.

[0106] In this embodiment, the tunable quantum dot luminescence parameter space is first converted into a high-dimensional spectral state representation. A state vector containing features such as spectral distribution vector (380-780nm, 5nm step size, 81 dimensions in total), color temperature value, color rendering index (Ra and R9), chromaticity coordinates (CIExy), light intensity, blue light ratio, and spectral purity is constructed to form a quantum dot spectral state descriptor. Based on this descriptor, a reinforcement learning state representation method is designed. An autoencoder is used for dimensionality reduction, key features are extracted, a state transition function and reward calculation mechanism are constructed, and a mathematical description of the state space is established to obtain a parameterized state evolution model.

[0107] Based on the physical boundary of the tunable quantum dot luminescence parameter space, a quantum dot spectral control action space is defined, discretizing the continuous parameter space into a finite set of actions. The design includes basic operations such as current density adjustment (±1%, ±5%, ±10%), driving voltage variation (±0.1V, ±0.5V, ±1V), pulse width modulation (±1%, ±5%, ±10% duty cycle), and component ratio adjustment (±5%, ±10% of the proportion of each color quantum dot), forming a discretized quantum dot control instruction set. A bidirectional mapping relationship is established between this instruction set and the tunable quantum dot luminescence parameter space. Action execution functions and state feedback functions are constructed to ensure that instructions can be accurately converted into hardware control signals, and that execution instructions can be inferred from parameter changes, thus obtaining a method for quantum dot spectral control.

[0108] By combining a parameterized state evolution model and a quantum dot spectral modulation method, a state-action-reward interaction framework is constructed, and a state transition function is defined. and reward function The f function represents the state-action mapping relationship. It captures the physical response characteristics of quantum dot materials, describes how changes in specific control parameters affect the final spectral output, establishes an environmental feedback mechanism, and forms a complete reinforcement learning interaction model.

[0109] Based on this model, a multi-objective reward function is designed, considering three dimensions: visual comfort, energy efficiency, and health impact. The visual comfort sub-reward function evaluates the matching degree between the spectrum and the target spectrum, color temperature suitability, and color rendering performance; the energy efficiency sub-reward function calculates the effective luminous flux generated per unit power consumption and spectral utilization; and the health impact sub-reward function evaluates the melatonin suppression index, blue light hazard coefficient, and circadian rhythm matching degree. By fusing the evaluations of the three dimensions through weighted summation or Pareto optimality methods, a comprehensive evaluation index of quantum dot spectroscopy is obtained.

[0110] A reinforcement learning value network is constructed based on a comprehensive evaluation index of quantum dot spectroscopy, employing a dual-network architecture (policy network and value network). The value network estimates the long-term cumulative reward of state-action pairs, while the policy network selects the optimal action based on the current state. The spectral optimization objective is transformed into a problem of maximizing cumulative reward. ,in The discount factor is (0.9-0.99). Indicates the expected value. For time steps, for The system state at a given time is represented by the spectral state descriptor. for The actions taken at any given time represent spectral modulation operations. To perform the action The next state after that, From state Execute action Arrival Status The immediate reward (reward function) is obtained; the spectral optimization objective function is obtained. Spectral simulation is performed using the adjustable quantum dot luminescence parameter space to simulate the spectral output effect under different combinations of control parameters, generating a large-scale sample (>1 million records) containing state-action-reward-next state, forming a quantum dot spectral state-action sample library.

[0111] A deep neural network is trained based on a quantum dot spectral state-action sample library and a spectral optimization objective function to realize the policy network and value network. Advanced reinforcement learning algorithms such as PPO (Proximal Policy Optimization), DDPG (Deep Deterministic Policy Gradient), or SAC (Soft Actor-Critic) are employed, with network layer structures and activation functions designed specifically for quantum dot characteristics. The policy network uses a deep network structure including LSTM layers to capture the temporal dependencies of spectral modulation; the value network uses a deep residual network structure to improve evaluation accuracy, resulting in a deep reinforcement learning model structure adapted to the characteristics of quantum dots. In a virtual environment constructed using tunable quantum dot luminescence parameter space, the deep reinforcement learning model structure is simulated, tested, and iteratively optimized. Extensive simulated interactions (>100,000 rounds) are used to improve model performance, and experience replay and objective network techniques are applied to enhance training stability, resulting in a basic quantum dot spectral modulation model.

[0112] By leveraging online learning technology combined with real-time spectral feedback, a fundamental quantum dot spectral control model is continuously optimized, enabling dynamic updates to model parameters. An incremental learning method is employed to integrate newly acquired spectral control experience into the model, allowing it to adapt to changes in user preferences and material property drift. A priority experience replay mechanism is implemented, focusing on learning from high-reward and rare state samples to accelerate model optimization. An A / B testing strategy is applied to evaluate the effectiveness of different control strategies, automatically selecting the optimal solution to obtain an adaptive quantum dot spectral control strategy, thus forming a complete dynamic spectral control strategy.

[0113] In this embodiment, a multi-objective reward function is designed based on a reinforcement learning interaction model, considering three dimensions: visual comfort, energy efficiency, and health impact, to obtain a comprehensive evaluation index for quantum dot spectroscopy, including:

[0114] A spectral evaluation framework was constructed based on a reinforcement learning interaction model, resulting in a quantum dot spectral performance analysis structure. A visual comfort quantification model was established by combining human factors engineering theory, converting color temperature deviation, color rendering index, flicker index, and spectral smoothness into visual comfort scores, thus obtaining a visual experience evaluation function.

[0115] A model relating the luminous efficiency of quantum dot materials to driving parameters was established to obtain the energy consumption-light output function. Based on the energy consumption-light output function, the effective luminous flux per unit power and the matching degree of the target spectrum were calculated to obtain the energy efficiency evaluation function.

[0116] Based on human photobiological safety standards and research on the impact of physiological rhythms, a spectral health impact assessment system was constructed to obtain a health risk mapping matrix. Based on the health risk mapping matrix, the melatonin inhibition index, blue light hazard coefficient, and photobiological rhythm impact factor were quantitatively calculated to obtain a health impact evaluation function.

[0117] Design a context-aware weight allocation algorithm that dynamically calculates the weight coefficients of three evaluation dimensions based on the usage scenario type, usage time period, and individual user characteristics to obtain an adaptive weight vector, and ensures that the sum of each weight in the adaptive weight vector is 1.

[0118] A comprehensive evaluation index for quantum dot spectroscopy is obtained by weighting and combining visual experience evaluation function, energy efficiency evaluation function, health impact evaluation function and adaptive weight vector.

[0119] In this embodiment, a spectral evaluation framework is first constructed based on a reinforcement learning interaction model. A multi-dimensional evaluation index system is designed, and a mapping relationship between spectral parameters and performance indicators is established, forming a quantum dot spectral performance analysis structure. A visual comfort quantification model is established by combining human factors engineering theory, considering the sensitivity and preference of the human eye to spectral characteristics. Color temperature deviation (the difference between the actual color temperature and the target color temperature) is quantified into a standardized score, and a nonlinear penalty function is designed, with a more severe penalty for larger deviations. The color rendering index (Ra and special color rendering indices R9-R15) is evaluated, and a color rendering performance scoring function is established, focusing on the color rendering effect on skin tones and natural objects. The flicker index (measuring the frequency and amplitude of light intensity fluctuations) is quantified to assess the risk of visual discomfort caused by flicker. Spectral smoothness (the continuity and smoothness of the spectral curve) is analyzed, and peaks and discontinuities are penalized, as these characteristics typically lead to visual discomfort. A visual experience evaluation function is constructed by integrating these factors. ,in , , , These are the scoring functions for each sub-factor. to These are the weighting coefficients. The difference between the actual color temperature and the target color temperature, in Kelvin (K). The color rendering index value is 0-100; The flicker detection value is 0-100%. This is an index for the smoothness of the spectral curve.

[0120] A model relating the luminous efficiency of quantum dot materials to driving parameters was established. Through experimental data and theoretical analysis, a functional relationship between current density, voltage, pulse characteristics, and luminous efficiency was constructed. Considering the saturation, quenching, and thermal effects of quantum dots, an accurate energy consumption-light output function was established. Based on this function, the effective luminous flux per unit power (lm / W) was calculated, and energy utilization efficiency was evaluated. The matching degree between the output spectrum and the target spectrum was analyzed to quantify the effective spectral utilization rate (the proportion of actual useful light output to total light output).

[0121] Based on these indicators, construct an energy efficiency evaluation function: ,in , and For balance coefficient, Luminous efficiency is expressed in lm / W, and measures the luminous flux produced per unit power. Spectral efficiency is a measure of the proportion of effective light output to total light output (0-1). This represents the power consumption penalty factor, which increases with increasing power consumption.

[0122] Based on human photobiological safety standards (such as IEC 62471) and studies on the impact of physiological rhythms, a spectral health impact assessment system is constructed. A wavelength-health risk correspondence table is established, forming a health risk mapping matrix that quantifies the potential impact on the human body at different wavelengths, intensities, and exposure times. Based on this matrix, the melatonin inhibition index (MEDI) is calculated to assess the degree of inhibition of melatonin secretion by the spectrum, which is closely related to circadian rhythms and sleep quality. The blue light hazard factor (BHF) is calculated to assess the potential damage risk of short-wavelength blue light to the retina. The matching degree between the spectrum and the user's current circadian rhythm stage is analyzed, and photobiological rhythm impact factors are calculated. These indicators are integrated to construct a health impact evaluation function.

[0123] ,in , and These are the weighting coefficients. The melatonin inhibition index (0-1) indicates that the higher the value, the stronger the inhibition of melatonin secretion. The value represents the blue light hazard factor (0-1), with a higher value indicating a greater risk of blue light hazard. The value represents the biorhythm matching degree (0-1), with a higher value indicating a better match with the user's current biorhythm stage.

[0124] A context-aware weighting algorithm is designed to dynamically adjust the importance of different evaluation dimensions. Basic weight templates are set based on usage scenario type (work, study, leisure, bedtime, etc.), emphasizing visual comfort in work scenarios and health impact in bedtime scenarios. Weights are adjusted based on usage time (morning, noon, afternoon, evening, night), prioritizing energy efficiency and visual experience during the day and health impact at night. Weights are further refined based on individual user characteristics (age, occupation, vision condition, health needs), such as increasing health weight for elderly users and increasing visual comfort weight for professional designers. These factors are then considered to calculate the weight coefficients for the three evaluation dimensions. An adaptive weight vector is formed, and normalization is used to ensure that the sum of the weights is 1.

[0125] The overall reward value is calculated by weighting and combining the visual experience evaluation function, energy efficiency evaluation function, health impact evaluation function, and adaptive weight vector.

[0126] ;in, The total score for the comprehensive evaluation of the spectrum (overall reward value). Rate the visual experience. Rate energy efficiency Rate the health impact. For visual experience weight, As a weight for energy efficiency, Weights are assigned to the health impact. Each component is normalized to ensure scores across different dimensions fall within a similar numerical range. The Pareto optimization principle is applied to handle multi-objective conflicts, seeking the optimal balance between different objectives. Finally, a comprehensive quantum dot spectral evaluation index is formed, serving as a reward signal for the reinforcement learning algorithm and guiding the optimization of the spectral control strategy.

[0127] In this embodiment, the detailed implementation steps of step S5 include:

[0128] The dynamic spectral modulation strategy is converted into a current-driven parameter set to obtain the quantum dot driving circuit control command. Based on the quantum dot driving circuit control command, a pulse width modulation scheme is designed to obtain a refined current control strategy.

[0129] Real-time monitoring of temperature distribution in quantum dot light-emitting devices yields thermal field distribution data. Based on this data, hotspot identification and heat flow analysis are performed to determine the device's thermal management requirements.

[0130] Based on the thermal management requirements of the device, an active heat dissipation control scheme is designed to obtain temperature adjustment commands, which are then converted into heat dissipation system control signals to obtain a thermal management execution scheme.

[0131] The actual output spectrum is acquired in real time using a spectral feedback sensor to obtain spectral deviation data, and the spectral error correction amount is calculated based on the spectral deviation data to obtain the spectral closed-loop correction parameters.

[0132] By integrating the refined current control strategy, thermal management implementation scheme, and spectral closed-loop correction parameters into a unified control framework, a high-fidelity spectral output control framework is obtained.

[0133] In this embodiment, the dynamic spectral modulation strategy is first converted into a specific set of current driving parameters, and the strategy decision is then transformed into specific control parameters for the driving circuit. For different quantum dot luminescent materials, the required driving current density (0.1-100 mA / cm²), driving voltage (2-12 V), pulse frequency (0-10 kHz), and duty cycle (1-100%) are calculated to form control commands for the quantum dot driving circuit. Based on these control commands, a pulse width modulation scheme is designed, and for different wavelength quantum dot groups, the PWM waveform, carrier frequency (>200 Hz, to avoid visible flicker), and duty cycle accuracy (0.1% level accuracy) are optimized. Synchronous multi-channel PWM control technology is employed to achieve coordinated control of quantum dots of different wavelengths, ensuring spectral output stability. Predistortion compensation technology is applied to correct the nonlinear response characteristics of the quantum dots, ultimately forming a refined current control strategy capable of sub-microsecond-level fine control.

[0134] Real-time monitoring of temperature distribution in quantum dot light-emitting devices (QDs) employs high-precision miniature thermocouple arrays or miniature infrared imaging systems to monitor temperature changes at key locations on the device's surface and internal structure. Thermocouple arrays provide a temperature accuracy of ±0.1°C, while infrared imaging systems offer high spatial resolution (<0.5 mm) temperature distribution maps with a sampling frequency of 1-10 Hz, ensuring timely capture of temperature changes. The temperature distribution of the device under different operating conditions is recorded, forming thermal field distribution data. Based on this data, hotspot identification and heat flow analysis are performed, and thermal image processing algorithms are applied to identify regions with abnormal temperature gradients and hotspot locations. A heat conduction model is used to analyze heat flow propagation paths and thermal resistance distribution, assessing the heat dissipation requirements and temperature control challenges in hotspot areas, thus forming a structured set of device thermal management requirements.

[0135] An active heat dissipation control scheme is designed based on the device's thermal management requirements, selecting appropriate heat dissipation strategies for different heat loads and hotspot distributions. Passive heat dissipation measures are used for low heat load areas, while active heat dissipation technologies, such as microchannel liquid cooling, thermoelectric cooling, or forced air cooling, are employed for high heat load and critical hotspot areas. A temperature closed-loop control algorithm is designed to achieve precise temperature regulation (±1°C), forming a temperature regulation command set. The temperature regulation commands are converted into specific control signals for the heat dissipation system, including fan speed control signals (PWM, 0-100%), liquid cooling system flow rate control signals (0-2L / min), and thermoelectric cooler power control signals (0-100%). Multi-region differentiated temperature control is implemented, prioritizing the temperature stability of critical light-emitting areas, forming a complete thermal management execution scheme.

[0136] The actual output spectrum is acquired in real time using a spectral feedback sensor, and a miniature spectrometer is used to monitor the output spectrum in real time, with a sampling frequency of 5-20Hz, a wavelength range of 380-780nm, and a resolution better than 2nm. The acquired actual spectrum is compared with the target spectrum in real time, and the deviation value at each wavelength point is calculated to construct complete spectral deviation data. Based on the spectral deviation data, the spectral error correction amount is calculated, and a spectral compensation strategy is designed using a PID control algorithm or a model predictive control algorithm. For deviations at different wavelengths, the adjustment amount of the driving parameters is calculated to ensure the accuracy of the spectral output. Considering the response delay and temperature sensitivity of quantum dot materials, feedforward compensation is performed to improve the real-time performance and accuracy of the control, ultimately forming the spectral closed-loop correction parameters.

[0137] This system integrates refined current control strategies, thermal management implementation schemes, and spectral closed-loop correction parameters into a unified control framework, constructing a multi-layered control system architecture. The bottom layer implements hardware-driven precise control, the middle layer handles temperature management and spectral compensation, and the top layer implements strategy decision-making and optimization adjustments. A parameter co-optimization mechanism is designed to address the mutual influence and constraints between current control, temperature regulation, and spectral compensation. A control priority strategy is established to dynamically adjust the priority of each control element under different operating conditions, ensuring system stability and spectral output accuracy. Fault detection and fault tolerance mechanisms are implemented to maintain basic functionality even if some control elements fail, ultimately forming a robust and reliable high-fidelity spectral output control framework.

[0138] In this embodiment, the detailed implementation steps of step S6 include:

[0139] The initial spectral configuration is output using a high-fidelity spectral output control framework to obtain the baseline spectral state. At the same time, the pupil response and heart rate variability data of the user under the baseline spectral state are collected through a physiological sensor array to obtain the raw physiological feedback data.

[0140] Signal processing and feature extraction were performed on the raw physiological feedback data to obtain the spectral influence on physiological indicators. The correlation analysis between the spectral influence on physiological indicators and the spectral closed-loop correction parameters in the high-fidelity spectral output control framework was then performed to obtain the physiological-spectral response relationship model.

[0141] Based on the physiological-spectral response relationship model, a spectral adjustment strategy is generated to obtain a physiologically driven optimization scheme. The physiologically driven optimization scheme is then converted into parameter adjustment instructions for a high-fidelity spectral output control framework. The refined current control strategy and thermal management execution scheme in the framework are updated respectively to obtain a physiologically sensing control parameter set.

[0142] By deploying a set of physiologically sensitive control parameters into a high-fidelity spectral output control framework, precise spectral output adjustment is achieved. At the same time, a long-term physiological feedback database is established to continuously optimize the spectral closed-loop correction parameters of the high-fidelity spectral output control framework, resulting in the final execution scheme for intelligent quantum dot lamp spectral adjustment.

[0143] In this embodiment, an initial spectral configuration is first output using a high-fidelity spectral output control framework. Based on a personalized spectral response model and dynamic spectral adjustment strategy, a baseline spectral setting is generated, including initial configurations of parameters such as color temperature, color rendering index, light intensity, and spectral distribution, forming the baseline spectral state. Simultaneously, physiological response data of the user under this spectral environment is collected through a physiological sensor array. The sensor array includes a remote pupil monitoring camera (60-120 frames / second, capable of capturing pupil changes as small as 0.1 mm), a photoplethysmography (PPG) sensor (heart rate, blood pressure), and a skin conductance sensor (pressure, alertness). The focus is on monitoring pupil size changes, pupillary light reflection velocity, heart rate variability indices (RMSSD, SDNN, LF / HF ratio), and skin conductance responses, forming multi-dimensional raw physiological feedback data.

[0144] Signal processing and feature extraction were performed on the raw physiological feedback data. Preprocessing techniques such as wavelet denoising, artifact removal, and baseline correction were applied to improve signal quality. Time-frequency domain analysis was conducted to extract pupil dynamics features (pupil constriction rate, steady-state diameter, and light reflection recovery time) and heart rate variability features (sympathetic-parasympathetic balance index and stress index). Machine learning algorithms (such as random forest or support vector machine) were applied to identify the fluctuation patterns of physiological indicators caused by spectral changes, filtering out interference from non-spectral factors to obtain spectrally influenced physiological indicators. These physiological indicators were correlated with the spectral closed-loop correction parameters in the high-fidelity spectral output control framework to establish a mapping relationship between spectral parameter adjustments (such as color temperature changes, blue light ratio adjustments, and light intensity changes) and physiological response changes (such as pupil constriction, heart rate changes, and stress levels). A mathematical model was constructed to describe this correlation, forming a physiological-spectral response relationship model.

[0145] A spectral adjustment strategy is generated based on a physiological-spectral response model, setting maximizing physiological comfort as the optimization objective. Gradient descent or genetic algorithms are applied to search for the optimal combination of spectral parameters. Considering the comprehensive performance of multiple physiological indicators, a physiologically driven optimization scheme is formed by balancing short-term comfort and long-term health impacts. This optimization scheme is then converted into specific parameter adjustment instructions for a high-fidelity spectral output control framework, updating the current density distribution (driving current of quantum dots at each wavelength), pulse width (PWM duty cycle), and driving frequency parameters in the refined current control strategy. Simultaneously, the temperature control objective (optimal operating temperature for each region) and heat dissipation power allocation in the thermal management execution scheme are optimized to ensure that the quantum dot material operates under optimal conditions, forming a complete set of physiologically sensitive control parameters.

[0146] A physiologically-aware control parameter set is deployed into a high-fidelity spectral output control framework. Through a multi-layered control system, precise driving of the quantum dot luminescent material is achieved, generating a spectral output that best meets the user's physiological needs. A gradual transition strategy is implemented to avoid discomfort caused by abrupt spectral changes, smoothly transitioning from the current state to the target state. Simultaneously, a long-term user physiological feedback database is established to continuously record physiological response patterns and subjective evaluation results under different spectral configurations, accumulating individualized spectral-physiological mapping data. Incremental learning technology is applied to continuously optimize the spectral closed-loop correction parameters of the high-fidelity spectral output control framework, enabling the system to adapt to long-term changes in user physiological characteristics (such as changes in visual characteristics due to age) and short-term fluctuations (such as special needs under fatigue). Combining contextual awareness, physiological feedback, and user preferences, a multi-dimensional closed-loop optimized intelligent quantum dot lighting spectral adjustment final execution scheme is formed.

[0147] The above describes the intelligent spectral adjustment method for quantum dot light-emitting lamps in the embodiments of this application. The following describes an intelligent spectral adjustment system for quantum dot light-emitting lamps in the embodiments of this application. Please refer to [link to relevant documentation]. Figure 4 One embodiment of the quantum dot light-emitting lamp's spectral intelligent adjustment system includes:

[0148] The data acquisition and preprocessing module is used to perform multi-dimensional data acquisition and preprocessing of user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings and physiological health data to obtain a standardized intelligent spectral control feature dataset.

[0149] The Human Factors Optics Analysis Module is used to extract human factors optics features and perform cross-contextual spectral demand analysis based on a standardized intelligent spectral control feature dataset, thereby obtaining a personalized spectral response model.

[0150] The quantum dot spectral mapping module is used to map the photoelectric properties of quantum dot materials and model nonlinear spectral synthesis based on a personalized spectral response model, so as to obtain a tunable quantum dot luminescence parameter space.

[0151] The intelligent optimization decision-making module is used to construct a deep reinforcement learning spectral optimization model based on the tunable quantum dot luminescence parameter space, and perform multi-objective constraint optimization to obtain a dynamic spectral control strategy.

[0152] The precision control execution module is used to perform real-time precise control of electrical parameters and thermal management compensation of quantum dot luminescent materials based on dynamic spectral modulation strategy, so as to obtain a high-fidelity spectral output control framework.

[0153] The closed-loop optimization feedback module is used to perform closed-loop verification and adaptive optimization based on the high-fidelity spectral output control framework and real-time collected user physiological feedback data, so as to obtain the final execution scheme of intelligent quantum dot lamp spectral adjustment.

[0154] The quantum dot luminaire spectral intelligent adjustment method provided by this invention can construct a personalized spectral response model based on multi-dimensional data such as user circadian rhythms, environmental conditions, usage scenarios, and personal preferences; achieve high-precision spectral control through photoelectric property mapping of quantum dot materials and nonlinear spectral synthesis; optimize spectral parameters using deep reinforcement learning algorithms while taking into account visual comfort, energy efficiency, and health impact; ensure the stability and consistency of spectral output by combining real-time electrical parameter control and thermal management compensation; and achieve truly intelligent personalized spectral adjustment through closed-loop verification and adaptive optimization using user physiological feedback data.

[0155] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0156] It should be noted that, in this document, 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 a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0157] In the description of this invention, it should be understood that the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0158] In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0159] In the description of this invention, "several" means one or more, and "a large number" means two or more.

[0160] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0161] All formulas in this manual are dimensionless and calculated numerically. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0162] Although embodiments of the invention have been shown and described, those skilled in the art will understand 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 claims and their equivalents.

Claims

1. A method for intelligently adjusting the spectrum of a quantum dot light-emitting lamp, characterized in that, include: Step S1: Multidimensional data collection and preprocessing are performed on user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings, and physiological health data to obtain a standardized intelligent spectral control feature dataset; Step S2: Based on the standardized intelligent spectral control feature dataset, perform human-caused optical feature extraction and cross-contextual spectral demand analysis to obtain a personalized spectral response model, including: Principal component analysis algorithm is applied to the standardized intelligent spectral control feature dataset to obtain key optical influence factors, and user visual sensitivity feature map is constructed based on the key optical influence factors to obtain personalized visual response features; Based on the personalized visual response characteristics, spectral sensitive regions are identified to obtain key wavelength ranges, and refined analysis is performed to obtain wavelength sensitivity distribution curves. Cross-scenario data mining techniques are used to perform cluster analysis on spectral requirements under different scenarios to obtain scenario spectral requirement prototypes. Based on the scenario spectral requirement prototypes, a scenario-transition spectral adaptive model is established to obtain a scenario-adaptive spectral framework. Extract time-series data from raw user preference data to obtain records of changes in user illumination preferences, and perform pattern recognition analysis to obtain the regular characteristics of user spectral preferences; Based on the characteristics of the user's spectral preference patterns, a time-series prediction neural network is constructed to obtain a preference trend prediction model. The preference trend prediction model is then used to generate future spectral demand predictions, resulting in dynamic spectral preference prediction results. The personalized visual response features, wavelength sensitivity distribution curves, context-adaptive spectral framework, and dynamic spectral preference prediction results are fused and modeled at multiple levels to obtain a personalized spectral response model. Step S3: Based on the personalized spectral response model, perform photoelectric property mapping and nonlinear spectral synthesis modeling of quantum dot materials to obtain a tunable quantum dot luminescence parameter space, including: Based on the personalized spectral response model, the target spectral feature requirements are extracted to obtain the ideal spectral parameter set. The ideal spectral parameter set is then converted into quantum dot luminescent material selection constraints to obtain material screening criteria. Based on the material screening criteria, the quantum dot material library was subjected to characteristic matching analysis to obtain candidate quantum dot material combinations, and photoelectric properties were characterized to obtain photoelectric conversion characteristic data of the materials. The personalized visual response features are correlated and mapped with the photoelectric conversion characteristics data of the material to obtain the correspondence between visual perception and material characteristics. Based on the correspondence between visual perception and material characteristics, a parameterized photoelectric control model is established to obtain a precise spectral control mapping function. Using the precise spectral modulation mapping function, a synergistic luminescence scheme for multiple quantum dot materials is designed to obtain a spectral synthesis strategy. The spectral synthesis strategy is then modeled with a nonlinear spectral superposition effect to obtain a spectral mixing theoretical model. Based on the aforementioned spectral mixing theory model and context-adaptive spectral framework, a spectral dynamic control parameter space is constructed to obtain the quantum dot luminescence parameter domain. The quantum dot luminescence parameter domain is then mapped to the actual control parameter range to obtain an adjustable quantum dot luminescence parameter space. Step S4: Based on the tunable quantum dot luminescence parameter space, construct a deep reinforcement learning spectral optimization model and perform multi-objective constraint optimization to obtain a dynamic spectral modulation strategy, including: The tunable quantum dot emission parameter space is converted into a high-dimensional spectral state representation to obtain a quantum dot spectral state descriptor. Based on the quantum dot spectral state descriptor, a reinforcement learning state representation method is designed to obtain a parameterized state evolution model. Based on the physical boundary of the tunable quantum dot luminescence parameter space, a quantum dot spectral modulation action space is defined to obtain a discretized quantum dot control instruction set. A bidirectional mapping relationship between the discretized quantum dot control instruction set and the tunable quantum dot luminescence parameter space is established to obtain a quantum dot spectral modulation implementation method. Combining the parameterized state evolution model and the quantum dot spectral modulation method, a quantum dot spectral state-action transition framework is constructed to obtain a reinforcement learning interaction model. Based on the reinforcement learning interaction model, a multi-objective reward function is designed, taking into account three dimensions: visual comfort, energy efficiency, and health impact, to obtain a comprehensive evaluation index for quantum dot spectroscopy. Based on the comprehensive evaluation index of quantum dot spectroscopy, a reinforcement learning value network is constructed to obtain the spectral optimization objective function. Then, the spectral simulation is performed using the tunable quantum dot luminescence parameter space to generate a training dataset and obtain a quantum dot spectral state-action sample library. Based on the quantum dot spectral state-action sample library and the spectral optimization objective function, a deep neural network is trained to realize the policy network and the value network, resulting in a deep reinforcement learning model structure adapted to the characteristics of quantum dots. In the virtual environment constructed by the tunable quantum dot luminescence parameter space, the deep reinforcement learning model structure is simulated, tested and iteratively optimized to obtain a basic quantum dot spectral control model. By utilizing online learning technology combined with real-time spectral feedback to continuously optimize the basic quantum dot spectral modulation model, an adaptive quantum dot spectral control strategy is obtained, thereby forming a complete dynamic spectral modulation strategy. Step S5: Based on the dynamic spectral modulation strategy, the quantum dot luminescent material is subjected to real-time precise control of electrical parameters and thermal management compensation to obtain a high-fidelity spectral output control framework; Step S6: Based on the high-fidelity spectral output control framework and the real-time collected user physiological feedback data, perform closed-loop verification and adaptive optimization to obtain the final execution scheme for intelligent quantum dot lamp spectral adjustment.

2. The intelligent adjustment method for the spectrum of quantum dot light-emitting lamps according to claim 1, characterized in that, The process involves multi-dimensional data collection and preprocessing of user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings, and physiological health data to obtain a standardized intelligent spectral control feature dataset, including: By collecting users' circadian rhythm indicators and sleep cycle data through wearable devices, raw biorhythm data is obtained, and time series analysis is performed to obtain a user biocycle model. By using a spectral sensor array to monitor ambient lighting conditions in real time, the raw parameters of the ambient spectrum are obtained, and color temperature analysis and light intensity calculation are performed to obtain ambient optical property data. The current usage scenario of the lighting fixtures is identified by the context-aware system, the original scenario information is obtained, and semantic classification and activity tagging are performed to obtain structured scenario data; By collecting user lighting preference settings through the user interface, the raw data of user preferences is obtained, and then the data is processed into feature vectors to obtain a user preference feature model. By collecting users' visual comfort and emotional state information through a non-invasive sensing system, raw physiological health data is obtained, and correlation analysis of physiological indicators is performed to obtain health status assessment results. Multidimensional feature fusion and standardization processing are performed on user biocycle model, environmental optical property data, structured scene data, user preference feature model and health status assessment results to obtain a standardized intelligent spectral control feature dataset.

3. The intelligent adjustment method for the spectrum of quantum dot light-emitting lamps according to claim 1, characterized in that, The process of performing characteristic matching analysis on the quantum dot material library according to the material screening criteria to obtain candidate quantum dot material combinations includes: A quantum dot material property database containing different sizes, different component ratios and different core-shell structures is constructed to obtain a quantum dot material library. The quantum dot material library is then classified by emission peak position and quantum efficiency rating to obtain a structured material retrieval framework. The material screening criteria are decomposed into multi-dimensional screening conditions, including emission wavelength requirements, color purity index, quantum efficiency threshold, and stability parameters, to obtain a parameterized screening vector. Based on the parameterized screening vector, a multi-level screening algorithm is designed to obtain a material matching search strategy. The material matching search strategy is used to perform parallel retrieval in the structured material retrieval framework to obtain a preliminary set of materials, and a spectral superposition simulation algorithm is applied to obtain a theoretical spectral combination scheme. The spectral coverage and color gamut range of the theoretical spectral combination schemes are analyzed to obtain spectral performance evaluation results. Based on the spectral performance evaluation results, the schemes are ranked and screened to obtain a set of preferred combination schemes. The preferred combination schemes are evaluated for material compatibility and co-processing feasibility to obtain a process compatibility score. Based on the process compatibility score and the spectral performance evaluation results, a comprehensive assessment is made to obtain candidate quantum dot material combinations.

4. The intelligent adjustment method for the spectrum of quantum dot light-emitting lamps according to claim 1, characterized in that, The multi-objective reward function designed based on the reinforcement learning interaction model, considering three dimensions—visual comfort, energy efficiency, and health impact—results in a comprehensive evaluation index of quantum dot spectra, including: Based on the reinforcement learning interaction model, a spectral evaluation framework is constructed to obtain the quantum dot spectral performance analysis structure. A visual comfort metric model is also established, which converts color temperature deviation, color rendering index, flicker index and spectral smoothness into visual comfort scores, and obtains a visual experience evaluation function. A model relating the luminous efficiency of quantum dot materials to driving parameters was established to obtain the energy consumption-light output function. Based on the energy consumption-light output function, the effective luminous flux per unit power and the matching degree of the target spectrum were calculated to obtain the energy efficiency evaluation function. A spectral health impact assessment system was constructed to obtain a health risk mapping matrix. Based on the health risk mapping matrix, the melatonin inhibition index, blue light hazard coefficient, and photobiological rhythm influencing factor were quantitatively calculated to obtain a health impact evaluation function. Design a context-aware weight allocation algorithm that dynamically calculates the weight coefficients of three evaluation dimensions based on the usage scenario type, usage time period, and individual user characteristics to obtain an adaptive weight vector, and ensures that the sum of each weight in the adaptive weight vector is 1; The comprehensive evaluation index of quantum dot spectroscopy is obtained by weighting and combining the visual experience evaluation function, energy efficiency evaluation function, health impact evaluation function and the adaptive weight vector.

5. The intelligent adjustment method for the spectrum of quantum dot light-emitting lamps according to claim 1, characterized in that, The aforementioned dynamic spectral modulation strategy enables real-time precise control of electrical parameters and thermal management compensation for quantum dot luminescent materials, resulting in a high-fidelity spectral output control framework, including: The dynamic spectral modulation strategy is converted into a current driving parameter set to obtain quantum dot driving circuit control instructions. Based on the quantum dot driving circuit control instructions, a pulse width modulation scheme is designed to obtain a refined current control strategy. Real-time monitoring of temperature distribution in quantum dot light-emitting devices yields thermal field distribution data. Based on this data, hotspot identification and heat flow analysis are performed to determine the device's thermal management requirements. Based on the thermal management requirements of the device, an active heat dissipation control scheme is designed to obtain a temperature adjustment command, which is then converted into a heat dissipation system control signal to obtain a thermal management execution scheme. The actual output spectrum is acquired in real time using a spectral feedback sensor to obtain spectral deviation data, and the spectral error correction amount is calculated based on the spectral deviation data to obtain the spectral closed-loop correction parameters. The refined current control strategy, thermal management execution scheme, and spectral closed-loop correction parameters are integrated into a unified control framework to obtain a high-fidelity spectral output control framework.

6. The intelligent adjustment method for the spectrum of quantum dot light-emitting lamps according to claim 1, characterized in that, The closed-loop verification and adaptive optimization based on the high-fidelity spectral output control framework and real-time collected user physiological feedback data yield the final execution scheme for intelligent quantum dot lamp spectral adjustment, including: The high-fidelity spectral output control framework is used to output the initial spectral configuration to obtain the reference spectral state. At the same time, the physiological sensor array collects the user's pupil response and heart rate variability data under the reference spectral state to obtain the raw physiological feedback data. Signal processing and feature extraction are performed on the raw physiological feedback data to obtain the spectral influence on physiological indicators. The correlation analysis between the spectral influence on physiological indicators and the spectral closed-loop correction parameters in the high-fidelity spectral output control framework is then performed to obtain the physiological-spectral response relationship model. Based on the physiological-spectral response relationship model, a spectral adjustment strategy is generated to obtain a physiologically driven optimization scheme. The physiologically driven optimization scheme is then converted into parameter adjustment instructions for the high-fidelity spectral output control framework. The refined current control strategy and thermal management execution scheme in the framework are updated respectively to obtain a physiologically sensitive control parameter set. The physiologically sensitive control parameter set is deployed to the high-fidelity spectral output control framework, and a long-term physiological feedback database is established to continuously optimize the spectral closed-loop correction parameters of the high-fidelity spectral output control framework, thus obtaining the final execution scheme for intelligent quantum dot lamp spectral adjustment.

7. A smart adjustment system for the spectrum of a quantum dot light-emitting lamp, used to implement the smart adjustment method for the spectrum of a quantum dot light-emitting lamp as described in any one of claims 1 to 6, characterized in that, include: The data acquisition and preprocessing module is used to perform multi-dimensional data acquisition and preprocessing of user biorhythm data, environmental spectral parameters, usage scenario information, user preference settings and physiological health data to obtain a standardized intelligent spectral control feature dataset. The human-cause optics analysis module is used to extract human-cause optics features and perform cross-context spectral demand analysis based on the standardized intelligent spectral control feature dataset to obtain a personalized spectral response model. The quantum dot spectral mapping module is used to map the photoelectric properties of quantum dot materials and model nonlinear spectral synthesis based on the personalized spectral response model, so as to obtain a tunable quantum dot luminescence parameter space. The intelligent optimization decision module is used to construct a deep reinforcement learning spectral optimization model based on the tunable quantum dot luminescence parameter space, and perform multi-objective constraint optimization to obtain a dynamic spectral control strategy. The precision control execution module is used to perform real-time precise control of electrical parameters and thermal management compensation of quantum dot luminescent materials based on the dynamic spectral modulation strategy, so as to obtain a high-fidelity spectral output control framework. The closed-loop optimization feedback module is used to perform closed-loop verification and adaptive optimization based on the high-fidelity spectral output control framework and real-time collected user physiological feedback data, so as to obtain the final execution scheme of intelligent quantum dot lamp spectral adjustment.

Images