Intelligent control method for color temperature of LED lamp

Abstract

The application relates to the technical field of intelligent control of electric lighting, and discloses an intelligent control method for the color temperature of an LED lamp. A plurality of controllable LED groups and spectrum sampling nodes are arranged, unit power spectrum is collected, invalid data is removed, and a matrix is assembled; an absorption function is constructed and normalized based on a plant multi-peak absorption center and a half-width; a theoretical spectrum is discretized and normalized in a preset color temperature interval; each sample is overlapped and integrated with the absorption function based on a photosynthetic effective wave band to obtain a color temperature-efficiency sample; a continuous efficiency function is constructed by interpolation and derivation, an initial value and a step length are set, and the optimal color temperature is obtained by projection and cyclic elimination; current ratios are generated by segmented integration of RGB, and linear driving is performed; the average color temperature in the growth period is updated to execute whole-period driving. The process is data-driven, quantifiable and adaptive, photosynthesis and energy efficiency are improved, and the limits of single-point monitoring, empirical adjustment and single illumination are compensated.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for electrical lighting, specifically to an intelligent method for adjusting the color temperature of LED lights. Background Technology

[0002] With the widespread application of LED lighting technology, the requirements for the quality and controllability of artificial light sources in indoor plant cultivation are constantly increasing. LED lights, with their advantages of high efficiency, adjustability, and long lifespan, are gradually replacing traditional incandescent and fluorescent lamps in greenhouse agriculture, home plant care, and plant factories, becoming the mainstream plant supplemental lighting equipment. Currently, most common LED plant lights on the market use preset fixed combinations of red, blue, and white wavelengths, achieving rough control of the lighting spectrum and color temperature by manually selecting the light source chip or simply adjusting the drive current. These solutions mostly only provide limited manual color temperature adjustment or simple timer and intensity control, lacking the ability to dynamically adjust color temperature and spectral distribution according to the actual physiological needs of plants, making it difficult to achieve precise adaptation to the photosynthetic characteristics of different plant varieties and growth stages.

[0003] In existing technologies, color temperature control often relies on hardware-level grouping switches, PWM dimming, or adjusting the proportions of red, green, and blue light through multi-channel constant current sources. Some high-end LED lighting systems offer remote adjustment of color temperature and brightness, but these adjustments are largely based on preset programs or simple feedback, making it difficult to proactively optimize plant photosynthetic efficiency. Furthermore, in outdoor lighting, such as LED solar streetlights, current technologies focus primarily on brightness adjustment, light and time control switches, and battery energy management. The main objectives are energy saving and self-powered reliability, with less attention paid to fine-tuning of spectral distribution and color temperature, and neglecting plant photosynthetic needs. While methods exist for adjusting LED light source output based on sensor feedback, these are often limited to ambient brightness or simple timed supplemental lighting, lacking data-driven intelligent color temperature control algorithms based on the actual absorption spectrum of plants. Simultaneously, some solutions utilize empirical models or big data analysis to assist in light source output selection, but these suffer from black-box decision-making, lack of interpretability, and often fail to guarantee adaptability to new plant species and control precision.

[0004] Therefore, this case aims to propose an intelligent control method for LED light color temperature. By closely integrating spectral sensing with plant absorption experiment results, a continuous function reflecting the plant's efficiency in utilizing light energy in different wavelengths is constructed. Then, an efficient iterative algorithm is used to adaptively find the optimal color temperature. Finally, the optimal spectrum is converted into the current setting ratio of red, green and blue sub-light sources, realizing the precise supply of photosynthetically effective radiation to plants. In the middle of the growth period, the absorption characteristics are automatically updated and the color temperature is recalculated to ensure that the light source output always matches the needs of the plant throughout the entire growth cycle. Summary of the Invention

[0005] This invention provides an intelligent method for controlling the color temperature of LED lights, which helps to solve the problems mentioned in the background art.

[0006] This invention provides the following technical solution: an intelligent control method for the color temperature of LED lights, comprising:

[0007] Set up multiple LED light groups and spectral sampling nodes, collect unit power spectra, filter out invalid samples and assemble the effective node spectral matrix;

[0008] For the target plant, obtain the center of multiple absorption peaks and half-width at half-maximum, and establish and normalize the plant absorption function;

[0009] A theoretical spectrum is constructed within a preset color temperature range, and discrete sampling and normalization are performed to obtain discrete color temperature sample spectra.

[0010] In the photosynthetically effective band, an overlap integral is performed between each discrete color temperature spectrum and the plant absorption function to obtain discrete absorption efficiency samples.

[0011] A continuous absorption efficiency function is constructed using interpolation based on discrete samples, and its derivative is obtained.

[0012] Set the initial color temperature, step size, and iteration upper limit, perform interval projection and iterative elimination to obtain the optimal color temperature;

[0013] Based on the normalized spectrum corresponding to the optimal color temperature, partition integration is performed on the three bands of red, green and blue respectively to generate the current setting ratio of each lamp group and drive the output linearly;

[0014] The absorption function and optimal color temperature are updated at the beginning and middle of the plant growth cycle, the average color temperature of the stage is calculated, the whole cycle current ratio is generated and the drive is executed.

[0015] Optionally, the step of setting multiple LED light groups and spectral sampling nodes, collecting unit power spectra, filtering out invalid samples, and assembling a valid node spectral matrix specifically includes:

[0016] Multiple LED light groups are arranged in the indoor green plant cultivation area, each group containing independently controllable red, green, and blue sub-light sources;

[0017] A spectral sampling node is set at a predetermined height directly below each lamp group, and the spectral power density of each node is collected when sampling is started.

[0018] For each node, normalization is performed in the visible light band; if the total energy in that band is zero, it is marked as an invalid node.

[0019] Renumber all valid nodes sequentially and count the total number of valid nodes.

[0020] The normalized spectra of each effective node are assembled into a sampling spectral matrix by columns;

[0021] When the total number of valid nodes is zero, the output will show a result indicating that there is no valid measured spectral data.

[0022] Optionally, for the target plant, obtaining multiple absorption peak centers and half-widths (FWHMs), and establishing and normalizing the plant absorption function specifically includes:

[0023] Identify the plant species to be controlled;

[0024] Through absorption experiments of chlorophyll a, chlorophyll b, and carotenoids, the central wavelengths of multiple major absorption peaks of this variety were obtained, forming a set of absorption centers.

[0025] For each absorption peak, a peak shape is established based on its full width at half maximum (FWHM), and superposition is performed to obtain the unit response absorption function.

[0026] The absorption function is normalized across the entire wavelength range to obtain the normalized absorption density function.

[0027] Optionally, the step of constructing a theoretical spectrum within a preset color temperature range, performing discrete sampling and normalization to obtain discrete color temperature sample spectra specifically includes:

[0028] Set the theoretical color temperature and the corresponding ideal spectral shape;

[0029] Generate discrete color temperature samples within a preset color temperature range according to the sampling step size;

[0030] If the endpoints do not cover the upper limit, add endpoint samples to form the final sample set;

[0031] For each discrete color temperature, construct a theoretical spectrum and perform normalization, then output the corresponding normalized spectral density.

[0032] Optionally, the step of performing an overlap integral on each discrete color temperature spectrum and the plant absorption function in the photosynthetically effective band to obtain discrete absorption efficiency samples specifically includes:

[0033] Set upper and lower limits for the wavelength of photosynthetically active radiation;

[0034] For each discrete color temperature, its normalized spectrum and the plant's normalized absorbance density are overlapped and integrated in that band to obtain the photosynthetically active absorption overlap intensity.

[0035] The discrete color temperature and its corresponding overlap intensity are combined to form a set of sample pairs.

[0036] Optionally, the step of constructing a continuous absorption efficiency function based on discrete samples using interpolation and obtaining its derivative function specifically includes:

[0037] Within a predefined domain, an absorption efficiency function is constructed using Lagrange interpolation based on discrete sample pairs.

[0038] The derivative function is obtained by taking the derivative of the aforementioned absorption efficiency function.

[0039] Optionally, the step of setting the initial color temperature, step size, and iteration upper limit, and performing interval projection and iterative elimination to obtain the optimal color temperature specifically includes:

[0040] Set the initial color temperature, iteration step size, maximum number of iterations, and derivative stopping threshold;

[0041] In each iteration, candidate update values ​​are generated based on the relationship between the derivative sign and the derivative stopping threshold.

[0042] Perform interval projection on the candidate values ​​to keep the color temperature within a preset range;

[0043] When a round trip occurs between two points, the absorption efficiency is compared at the two points and the process is terminated directly based on the better one.

[0044] When a normal cycle occurs, select the color temperature with the highest absorption efficiency in the sub-set of the cycle state and terminate the cycle.

[0045] The process terminates at the current color temperature when the derivative meets the stopping threshold.

[0046] If the iteration limit is reached and the process does not terminate, then the color temperature with the highest absorption efficiency is selected as the optimal color temperature from the evaluated set.

[0047] Optionally, the step of performing partitioned integration on the red, green, and blue bands based on the normalized spectrum corresponding to the optimal color temperature to generate the current setting ratio of each lamp group and linearly drive the output specifically includes:

[0048] On the normalized spectrum corresponding to the optimal color temperature, integration is performed on the three fixed bands of red, green and blue respectively to obtain the spectral area ratio of each band;

[0049] The current setting ratio for each group of lights is generated according to the ratio of red, green, and blue.

[0050] When all three ratios are zero, set the current ratio of each channel to zero;

[0051] Each lamp group's drive circuit outputs a linear current according to the current ratio.

[0052] Optionally, the step of updating the absorption function and optimal color temperature at the beginning and middle of the plant growth cycle, calculating the average color temperature during the stage, generating the total cycle current ratio, and executing the drive specifically includes:

[0053] Set the plant growth cycle, collect the absorption function on day zero, and obtain the optimal color temperature for that stage;

[0054] Repeat the absorption test during the middle of the growth cycle, update the absorption function and obtain a new optimal color temperature;

[0055] Calculate the average color temperature of the two stages mentioned above, construct their theoretical spectrum, and perform normalization.

[0056] Integrating the average color temperature spectrum across the red, green, and blue bands yields the current setpoint ratio for the entire cycle.

[0057] Each lamp group is driven to output according to the full cycle current ratio in subsequent time periods.

[0058] The present invention has the following beneficial effects:

[0059] 1. This solution involves setting up spectral sampling nodes directly below each group of lamps to collect the band power density at unit power in real time from multiple points. Zero-response nodes are eliminated, and the valid data is then normalized across the entire visible light spectrum and assembled into a matrix using sequential numbering. This innovative process of matrixing multi-channel spectral data eliminates the interference of noise and invalid data on subsequent calculations while preserving the spectral characteristics of each sampling point, achieving high-precision characterization of the light source output. Compared to existing methods that rely solely on single-sensor data aggregation or simple averaging, this module provides a reliable large-scale sample foundation for subsequent refined models through a data matrix structure, improving the accuracy and robustness of the spectral response and laying a solid data foundation for the entire color temperature optimization process.

[0060] 2. This scheme conducts absorption experiments on chlorophyll a, chlorophyll b, and carotenoids in target plants, identifying multiple absorption peak centers and their corresponding full width at half maximum (FWHM) parameters. A peak superposition method is used to accurately construct the unit response absorption function, which is then normalized across the entire wavelength range to form an absorption density curve. This approach achieves three key benefits: first, it integrates real experimental data from multiple pigments to establish a multi-peak composite response model; second, the normalization strategy allows for direct quantitative comparison across different wavelengths; and third, the peak superposition method considers both overlap and peak width variations. This effectively overcomes the model errors caused by existing technologies that only focus on a single peak or approximate linear decay, making subsequent absorption efficiency calculations more closely reflect the actual photosynthetic absorption patterns of plants and improving the scientific basis and control precision of targeted lighting.

[0061] 3. This scheme designs a method to generate discrete color temperature samples at equal intervals within a preset color temperature range, and performs band normalization on each theoretical spectrum. If necessary, it automatically performs endpoint sampling to form a complete and seamless sample set. This decouples color temperature from theoretical spectrum construction, and normalization removes differences in total energy, ensuring that different color temperature spectra can be compared and applied under the same standard. Simultaneously, the discretization and endpoint sampling design ensures no color temperature gaps, thus laying the foundation for constructing a continuous efficiency function. Compared to existing methods that only provide a few discrete color temperatures or use trial-and-error adjustments, this module achieves automation, seamlessness, and scalability of the spectral library, improving the coverage and flexibility of the color temperature optimization space.

[0062] 4. While the overlap integral in the photosynthetically effective wavelength band is a common concept in existing research, this approach directly integrates this method with previous normalized spectra and absorption functions, calculating the overlap intensity at each discrete color temperature and generating a set of discrete sample pairs of color temperature-absorption efficiency. This achieves three key benefits: first, it places the coupling integral between the light source and the plant model under a unified standard; second, it directly obtains high-dimensional data pairs suitable for interpolation at the discrete sample level; and third, it is linked with subsequent interpolation optimization, achieving seamless integration of the computational process. Compared to commercially available methods that only output total illuminance or empirical gain ratios, this module quantifies the relative contribution of different color temperatures to plant photosynthesis, providing accurate and traceable efficiency indicators for color temperature optimization decisions, and realizing a direct mapping and closed-loop feedback between lighting effects and plant growth needs.

[0063] 5. This scheme applies a Lagrange interpolation method based on discrete samples to construct a continuous curve of photosynthetic absorption efficiency. Based on this, the derivative information of the efficiency function is obtained for subsequent iteration stopping and direction determination. Firstly, high-precision interpolation is used for the plant absorption efficiency curve, avoiding errors caused by oversimplification in traditional linear or low-order fitting. Secondly, derivative information is actively extracted, enabling rapid location of extreme value regions based on function trends during iteration. Thirdly, interpolation and derivative calculation are integrated, achieving modular algorithm design. Unlike existing methods that only use discrete value lookups or simple empirical interpolation, this module balances accuracy and operability, improving the efficiency and stability of optimal color temperature solution, and maintaining high convergence speed and accuracy under different plant types or environments.

[0064] 6. For the efficiency function, this scheme, for the first time, combines multiple strategies, including interval projection, two-point cyclic elimination, general cyclic elimination, and conventional stopping and non-convergence upper limit handling, to form a robust iterative solution process for color temperature. After each iteration, projection ensures that the solution does not exceed the predetermined range; for two-point or more complex cyclic cases, local optimization or subset optimization termination is adopted respectively; at the same time, derivative threshold stopping is used to improve the probability of fast convergence; once the iteration upper limit is reached, the color temperature with the highest efficiency is selected from the evaluated set. Unlike the traditional method that only uses a single steepest descent or Newton's method, this scheme integrates multiple cyclic elimination techniques, which can avoid getting stuck in small oscillations and quickly jump out of the extreme value region, improving the stability and adaptability of the solution, and ensuring that the optimal color temperature can be obtained efficiently and accurately even in the face of noise or complex efficiency curves.

[0065] 7. The normalized spectrum corresponding to the optimal color temperature is divided into red, green, and blue bands, and integrated separately. The area ratio of the three spectral segments is then directly mapped to the current setting ratio of each channel, and the output of the linear drive circuit is based on this. This achieves three key benefits: first, it precisely transmits spectral control to current control, realizing a closed-loop mapping from spectrum to electrical control; second, it avoids the empirical setting of traditional channel current that is disconnected from the effect, allowing for real-time adjustment of the current ratio through spectral integration data; and third, it automatically switches to a fully off state when the area of ​​all three spectral segments is zero, improving system safety. This module breaks through the existing mode of fixed power settings for multi-channel LEDs, achieving high-precision coupling between color temperature optimization and current drive, improving spectral utilization efficiency and energy efficiency.

[0066] 8. Addressing the dynamic changes in spectral requirements at different growth stages of plants, this solution re-acquires absorption functions and calculates the optimal color temperature for each stage during the early and middle growth phases. After calculating the average color temperature of the two stages, the spectrum is reconstructed, and a full-cycle current setpoint ratio is generated, achieving periodic closed-loop correction of plant growth requirements. Firstly, it breaks the static adjustment mode by incorporating differences in plant growth stages into the color temperature optimization process; secondly, it balances the initial activation and middle-stage steady-state requirements through two node acquisitions and average calculations; and thirdly, the full-cycle current setpoint ratio is driven by a single click, simplifying subsequent operation and management. Unlike existing methods that only require manual setting during installation or adjustment based on experience, this mechanism achieves full life-cycle spectral optimization, meeting photosynthetic needs at different stages while saving energy, improving management automation, and enhancing plant growth quality. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0068] 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.

[0069] Example, refer to Figure 1 A method for intelligently controlling the color temperature of LED lights, comprising:

[0070] Set up multiple LED light groups and spectral sampling nodes, collect unit power spectra, filter out invalid samples and assemble the effective node spectral matrix;

[0071] For the target plant, obtain the center of multiple absorption peaks and half-width at half-maximum, and establish and normalize the plant absorption function;

[0072] A theoretical spectrum is constructed within a preset color temperature range, and discrete sampling and normalization are performed to obtain discrete color temperature sample spectra.

[0073] In the photosynthetically effective band, an overlap integral is performed between each discrete color temperature spectrum and the plant absorption function to obtain discrete absorption efficiency samples.

[0074] A continuous absorption efficiency function is constructed using interpolation based on discrete samples, and its derivative is obtained.

[0075] Set the initial color temperature, step size, and iteration upper limit, perform interval projection and iterative elimination to obtain the optimal color temperature;

[0076] Based on the normalized spectrum corresponding to the optimal color temperature, partition integration is performed on the three bands of red, green and blue respectively to generate the current setting ratio of each lamp group and drive the output linearly;

[0077] The absorption function and optimal color temperature are updated at the beginning and middle of the plant growth cycle, the average color temperature of the stage is calculated, the whole cycle current ratio is generated and the drive is executed.

[0078] By deploying multiple independently controllable red, green, and blue sub-light sources and spectral sampling nodes, the system performs real-time, multi-point acquisition of unit power spectra of the indoor plant lighting environment. After removing invalid data, the data is sequentially assembled into a standardized spectral matrix, eliminating the problem of traditional single-point illumination monitoring being susceptible to location and sensor errors. Furthermore, based on experimentally obtained plant multi-peak absorption characteristics, a normalized absorption density function is constructed, enabling the system to accurately reflect the relative contribution of different wavelengths to plant photosynthesis, solving the absorption simulation bias caused by relying solely on empirical models from literature. Discrete sampling and normalization of theoretical spectra within a preset color temperature range lays the foundation for subsequent effects. A comprehensive sample library was established for efficiency calculation. By performing overlapping integrals on each discrete spectrum and absorption function in the photosynthetically effective wavelength band, a one-to-one correspondence between color temperature and absorption efficiency was obtained, avoiding the limitation of traditional methods that can only provide a single illuminance index. Then, through interpolation and derivative solving, the discrete samples were transformed into continuous high-precision efficiency curves. Based on this, multiple iterative strategies such as interval projection and cyclic elimination were used to obtain the optimal color temperature, ensuring that the solution process is both fast and robust. Finally, the piecewise integral of the optimal color temperature spectrum was mapped to the red-green-blue current setpoint ratio, and combined with dynamic corrections at the beginning and middle of the plant growth period, a closed-loop automatic adjustment for the entire cycle was achieved. This scheme, with its full-process data-driven, quantifiable, and dynamically adaptable approach, overcomes the contradiction between efficiency and energy saving that is difficult to balance in traditional fixed color temperature or empirical adjustment, thereby improving plant photosynthetic efficiency and energy utilization.

[0079] The process of setting up multiple LED light groups and spectral sampling nodes, collecting unit power spectra, filtering out invalid samples, and assembling a valid node spectral matrix specifically includes:

[0080] Multiple LED light groups are arranged in the indoor green plant cultivation area, each group containing independently controllable red, green, and blue sub-light sources;

[0081] A spectral sampling node is set at a predetermined height directly below each lamp group, and the spectral power density of each node is collected when sampling is started.

[0082] For each node, normalization is performed in the visible light band; if the total energy in that band is zero, it is marked as an invalid node.

[0083] Renumber all valid nodes sequentially and count the total number of valid nodes.

[0084] The normalized spectra of each effective node are assembled into a sampling spectral matrix by columns;

[0085] When the total number of valid nodes is zero, the output will show a result indicating that there is no valid measured spectral data.

[0086] Further specific implementation steps include:

[0087] Set up indoor green plant cultivation area indivual Light group, number Each group contains independently controllable red, green, and blue sub-light sources; among them, This represents the total number of light clusters. Number the light groups;

[0088] Directly below each light group A spectral sampling node is set at an altitude of meters, numbered as follows: Each sampling node The collected spectra are denoted as ;in, Number the spectral sampling nodes; For the first Each sampling node at time... spectral power density; Nanometers are wavelengths. This is the time to start sampling;

[0089] The normalized spectrum for each sampling point is calculated as follows:

[0090] ;in, For the first Each sampling node at time... Normalized spectral density; To perform definite integrals over the wavelength range;

[0091] All valid Perform sequential renumbering to obtain the total number of valid nodes, denoted as . The new index is ;in, The valid sampling node number;

[0092] Assemble the sampling spectral matrix ;in, It is a column vector field, where each column is the normalized spectral density function of an effective node; For the first The number of valid nodes at time [time] Normalized spectral density;

[0093] when If no valid measured spectral data is available, it is determined that there is no valid measured spectral data.

[0094] By arranging independent spectral sampling nodes at a preset height directly below each group of lamps, the spectral power density of each node is collected in real time and normalized across the entire visible band. This allows the system to eliminate invalid data caused by zero response or sensor failure, avoiding the data blind spots or biases that previously existed due to reliance on a single sensor. All valid nodes are renumbered according to the acquisition sequence and assembled into a column vector matrix, enabling structured storage of multi-point spectral data and providing a unified data entry point for subsequent batch calculations. The system automatically determines and alarms when there are no valid nodes, improving its robustness and maintainability. Compared to existing single-point illuminance or manual sampling methods, this solution, through matrix management of multi-channel spectral data, not only improves the spatial coverage of spectral representation but also reduces the impact of acquisition errors on the accuracy of subsequent control algorithms, providing a reliable and scalable data foundation and anomaly handling mechanism for the entire intelligent control process.

[0095] For the target plant, multiple absorption peak centers and half-widths (FWHMs) are obtained, and the plant absorption function is established and normalized. Specifically, this includes:

[0096] Identify the plant species to be controlled;

[0097] Through absorption experiments of chlorophyll a, chlorophyll b, and carotenoids, the central wavelengths of multiple major absorption peaks of this variety were obtained, forming a set of absorption centers.

[0098] For each absorption peak, a peak shape is established based on its full width at half maximum (FWHM), and superposition is performed to obtain the unit response absorption function.

[0099] The absorption function is normalized across the entire wavelength range to obtain the normalized absorption density function.

[0100] Further specific implementation steps include:

[0101] Let the plant variety to be controlled be numbered as follows: ,like It represents pothos. It represents ivy;

[0102] For plant species Through experiments on the absorption of chlorophyll a, b and carotenoids, the following results were obtained. The wavelengths of the main absorption peaks form a set of absorption centers: ;in, Plants The number of main absorption peaks; For containing plants The center wavelength of the absorption peak; Plants The The center wavelength of each absorption peak; For absorption peak indexing;

[0103] The chlorophyll a, b and carotenoid absorption experiment is existing technology;

[0104] Constructing plants The unit response absorption function: ;in, Plants For wavelength The relative absorption response; Plants The The full width at half maximum (FWHM) of each absorption peak;

[0105] Will Normalizing across the entire spectral range yields the normalized absorption function:

[0106] ;in, Plants The normalized absorption density function.

[0107] By conducting actual absorption experiments on target plants for chlorophyll a, chlorophyll b, and carotenoids, the centers of multiple major absorption peaks and their corresponding full width at half maximum (FWHM) parameters were obtained. A unit response absorption function was constructed based on peak superposition, and then normalized across the entire wavelength range. This system can accurately characterize the absorption characteristics of plants in each key wavelength range. This innovation overcomes the response bias caused by traditional methods that only use single-peak or empirical curve approximations. The combination of multi-peak superposition and normalization processing makes subsequent photosynthetic efficiency calculations closer to the actual absorption patterns of plants. Simultaneously, the introduction of the normalized absorption density function can be directly used for integral calculations with the normalized spectrum, eliminating dimensional interference caused by absolute energy differences. Compared with existing technologies, this scheme, based on experimental data, establishes a multi-peak composite model that improves the accuracy of the coupling between the spectrum and plant response, providing reliable plant physiological basis and high-precision input for intelligent control schemes.

[0108] The process of constructing a theoretical spectrum within a preset color temperature range, performing discrete sampling and normalization to obtain discrete color temperature sample spectra specifically includes:

[0109] Set the theoretical color temperature and the corresponding ideal spectral shape;

[0110] Generate discrete color temperature samples within a preset color temperature range according to the sampling step size;

[0111] If the endpoints do not cover the upper limit, add endpoint samples to form the final sample set;

[0112] For each discrete color temperature, construct a theoretical spectrum and perform normalization, then output the corresponding normalized spectral density.

[0113] Further specific implementation steps include:

[0114] The theoretical color temperature is set to Its corresponding ideal spectral shape is defined as:

[0115] ;in, The color temperature is Theoretical spectral density at that time; This is the normalized proportionality constant; ;

[0116] Normalized proportionality constant Specifically: ;

[0117] In color temperature range Inside, with step length Equal-interval sampling; where, , These are the minimum and maximum color temperatures, respectively. The color temperature sampling step size;

[0118] Set the initial number of samples before adding endpoints to... ;

[0119] Set the first Each discrete color temperature takes the value of ;in, For color temperature sample index;

[0120] like Then add endpoint nodes. The final sample size is obtained as follows:

[0121] ;in, This represents the final sample size.

[0122] For each The normalized spectrum is calculated as follows: ;in, For corresponding discrete color temperature The normalized spectral density.

[0123] By generating discrete color temperature samples at equal intervals within a preset color temperature range and normalizing the theoretical spectrum corresponding to each sample, this scheme not only achieves comprehensive coverage of the spectral shape within the continuous color temperature range, but also eliminates the "color temperature gap" caused by sampling loop incompleteness through automatic endpoint sampling. The introduction of the normalization ratio eliminates the interference of the total energy difference corresponding to different color temperatures on the comparison, enabling direct comparison of spectra at different color temperatures under the same standard. Compared with the traditional approach of providing only a few fixed color temperatures or based on a fixed spectral library, this scheme lays a data completeness and consistency foundation for subsequent efficiency curve interpolation and optimal color temperature solution by automatically constructing and normalizing a discrete spectral set, thereby improving the precision and flexibility of color temperature adjustment.

[0124] The step of performing an overlap integral between each discrete color temperature spectrum and the plant absorption function in the photosynthetically effective wavelength band to obtain discrete absorption efficiency samples specifically includes:

[0125] Set upper and lower limits for the wavelength of photosynthetically active radiation;

[0126] For each discrete color temperature, its normalized spectrum and the plant's normalized absorbance density are overlapped and integrated in that band to obtain the photosynthetically active absorption overlap intensity.

[0127] The discrete color temperature and its corresponding overlap intensity are combined to form a set of sample pairs.

[0128] Further specific implementation steps include:

[0129] Set the effective wavelength range of photosynthesis as ;in, , These are the lower and upper wavelengths of photosynthetically active radiation, respectively.

[0130] For each color temperature Calculate plants Photosynthetic absorption integral value:

[0131] ;in, Plants Color temperature The photosynthetic effective absorption overlap intensity;

[0132] Construct a set of discrete point pairs for the absorption efficiency function: ;in, For containing plants All Discrete sample pairs.

[0133] By performing overlapping integrals on the normalized theoretical spectrum and normalized absorption density function corresponding to each discrete color temperature within the photosynthetically effective wavelength range, this scheme quantitatively obtains discrete sample pairs of color temperature and photosynthetic absorption efficiency, realizing a direct mapping from spectral shape to plant absorption efficiency. This calculation method breaks through the traditional mode of only considering total light flux or empirical gain ratio, improving the coupling accuracy between spectrum and physiological response. Using the obtained sample pair set as input for subsequent interpolation function construction not only ensures data consistency but also enables the efficiency curve to closely follow the actual trend of discrete points. Compared with existing technologies, this scheme provides measurable and traceable efficiency indicators, providing high-confidence sample support for subsequent optimal color temperature iterative solutions, and achieving precise matching between spectral selection and actual plant needs.

[0134] The construction of a continuous absorption efficiency function based on discrete samples using interpolation, and the acquisition of its derivative function, specifically includes:

[0135] Within a predefined domain, an absorption efficiency function is constructed using Lagrange interpolation based on discrete sample pairs.

[0136] The derivative function is obtained by taking the derivative of the aforementioned absorption efficiency function.

[0137] Further specific implementation steps include:

[0138] In the domain The above is constructed using Lagrange interpolation. Specifically:

[0139] , ;in, Plants At any color temperature Continuous photosynthetic absorption efficiency; To and Different discrete color temperature sample indexes;

[0140] right Taking the derivative, we get the derivative function as:

[0141] , ;in, for right The derivative; , These are the product and summation operators, respectively.

[0142] By constructing a continuous efficiency function based on discrete color temperature-absorption efficiency sample pairs using a high-precision interpolation method, and further extracting the derivative information of this function, this scheme achieves a seamless transition from discrete samples to continuous curves, providing mathematical support for efficient extreme value localization. Obtaining the derivative avoids the accuracy loss caused by relying solely on table lookups or empirical judgments, allowing for rapid adjustment of direction and step size based on the function's changing trend during subsequent iterations. Compared to existing techniques that simply use discrete value lookups or simple curve fitting, this scheme maintains operability while ensuring computational accuracy. By combining continuous functions with derivatives, it provides a refined reference for the efficient solution of the optimal color temperature, improving the system's convergence speed and solution stability.

[0143] The process of setting the initial color temperature, step size, and iteration upper limit, performing interval projection and iterative elimination to obtain the optimal color temperature specifically includes:

[0144] Set the initial color temperature, iteration step size, maximum number of iterations, and derivative stopping threshold;

[0145] In each iteration, candidate update values ​​are generated based on the relationship between the derivative sign and the derivative stopping threshold.

[0146] Perform interval projection on the candidate values ​​to keep the color temperature within a preset range;

[0147] When a round trip occurs between two points, the absorption efficiency is compared at the two points and the process is terminated directly based on the better one.

[0148] When a normal cycle occurs, select the color temperature with the highest absorption efficiency in the sub-set of the cycle state and terminate the cycle.

[0149] The process terminates at the current color temperature when the derivative meets the stopping threshold.

[0150] If the iteration limit is reached and the process does not terminate, then the color temperature with the highest absorption efficiency is selected as the optimal color temperature from the evaluated set.

[0151] Further specific implementation steps include:

[0152] Set the initial color temperature value iteration step size Maximum number of iterations Derivative stopping threshold ;in, For discrete iterative indexes;

[0153] The process involves iterative updates, interval projection, iterative resolution, and determining the upper limit of non-convergence, specifically:

[0154] S601, Define the interval projection operator as follows:

[0155] ;in, For interval projection operators;

[0156] In the In the next iteration, candidate update values ​​are generated based on the relationship between the derivative sign and the derivative stopping threshold:

[0157] ;in, For the first Color temperature estimation in the next iteration; For the first Candidate values ​​before the next update;

[0158] make ;

[0159] S602, Two-point cyclic elimination:

[0160] If it exists make and Then in Choose the best: And terminate the iteration; where, Plants The optimal color temperature;

[0161] S603, General cyclic digestion:

[0162] If it exists make and ,make

[0163] , And terminate the iteration; where, It is a discrete iteration index used to refer to the sequence number of a specific iteration in history; A subset of loop states when a loop is detected;

[0164] S604, Normal Termination: If any make ,but ;

[0165] S605, Unconverged upper limit: If the iteration reaches... If the aforementioned termination is not triggered, then within the already evaluated set Selected from .

[0166] This method organically combines multiple iterative resolution strategies with interval projection operators, ensuring that the current color temperature estimate remains within a preset range after each iteration, effectively avoiding invalid searches caused by exceeding limits. Local optimization and subset optimization termination strategies are adopted for two-point and general loop cases, respectively, preventing loop trapping and enabling rapid convergence for complex curve structures. Further iterations can be terminated early when the derivative change meets the stopping condition, saving computational resources. If convergence is not achieved even after reaching the iteration limit, the most efficient value is selected from all evaluated values, guaranteeing the global optimality of the result. Unlike traditional single steepest descent or Newton's method, this method balances robustness and efficiency, enabling rapid and stable acquisition of the optimal color temperature on curves with noise or multiple extrema, solving the problems of existing methods easily getting trapped in local extrema or oscillating non-convergence.

[0167] The process involves performing partitioned integration on the red, green, and blue bands based on the normalized spectrum corresponding to the optimal color temperature, generating the current setpoint ratio for each lamp group, and linearly driving the output. Specifically, this includes:

[0168] On the normalized spectrum corresponding to the optimal color temperature, integration is performed on the three fixed bands of red, green and blue respectively to obtain the spectral area ratio of each band;

[0169] The current setting ratio for each group of lights is generated according to the ratio of red, green, and blue.

[0170] When all three ratios are zero, set the current ratio of each channel to zero;

[0171] Each lamp group's drive circuit outputs a linear current according to the current ratio.

[0172] Further specific implementation steps include:

[0173] Depend on Calculate the partition integral:

[0174] , , ;in, To correspond to the optimal color temperature Normalized spectral density; , , These represent the spectral area ratios for the red, green, and blue bands, respectively.

[0175] calculate Light source current setting ratio:

[0176] ;in, , , The first The current setting values ​​for the red, green, and blue channels;

[0177] Each lamp group's drive circuit outputs a linear current based on the aforementioned current ratio.

[0178] Based on the normalized spectrum corresponding to the optimal color temperature, the spectral area ratios of the red, green, and blue segments are calculated through piecewise integration. These ratios are then directly mapped to the current setting ratios of each channel, thus realizing the transformation from spectral optimization results to... The method features closed-loop conversion of the driving current; it automatically switches to full-off mode when the spectral allocation is all zero, improving safety; unlike existing methods that mostly use empirical settings or fixed power settings, this method precisely couples the lighting current with the spectral requirements, ensuring energy delivery to the wavelengths required by the plants while avoiding energy waste, thus improving the system's energy efficiency and the controllability of the lighting effect.

[0179] The process of updating the absorption function and optimal color temperature at the beginning and middle of the plant growth cycle, calculating the average color temperature during the stage, generating the total cycle current ratio, and executing the drive specifically includes:

[0180] Set the plant growth cycle, collect the absorption function on day zero, and obtain the optimal color temperature for that stage;

[0181] Repeat the absorption test during the middle of the growth cycle, update the absorption function and obtain a new optimal color temperature;

[0182] Calculate the average color temperature of the two stages mentioned above, construct their theoretical spectrum, and perform normalization.

[0183] Integrating the average color temperature spectrum across the red, green, and blue bands yields the current setpoint ratio for the entire cycle.

[0184] Each lamp group is driven to output according to the full cycle current ratio in subsequent time periods.

[0185] Further specific implementation steps include:

[0186] Assume the plant growth cycle is On day 0, its absorption function was collected. And obtain the initial optimal color temperature. ;in, This is the normalized absorption density function for stage 0; This represents the optimal color temperature corresponding to stage 0.

[0187] At the Daily repeatable spectral absorption tests were conducted to construct a new absorption function. and obtain ;in, This is the normalized absorption density function for the first stage; This is the optimal color temperature corresponding to stage 1;

[0188] Calculation stage average color temperature ;

[0189] based on Construct the corresponding spectral shape and calculate the ratio:

[0190] S801, First calculate the corresponding average color temperature Theoretical spectrum Specifically:

[0191] ;

[0192] S802, then normalized to ;in, This represents the normalized spectral density corresponding to the average color temperature.

[0193] S803, then calculate: , , ;in, , , In order to be in Normalized spectral area of ​​the three segments;

[0194] S804, Obtain the current ratio used for the entire lighting cycle:

[0195] ;in, , , For the first Group in the whole cycle Current set ratio;

[0196] Accordingly, each light group will proceed according to the following schedule in the subsequent time period. Ratio-driven output.

[0197] By collecting the absorption density function at the beginning and middle of the plant growth cycle and obtaining the corresponding optimal color temperature, the theoretical spectrum is reconstructed and normalized after calculating the average color temperature of each stage, and the current setpoint ratio for the entire cycle is obtained. This achieves dynamic closed-loop correction of the spectral requirements for different growth stages. This mechanism breaks the traditional static lighting mode, taking into account both the initial activation and the steady-state requirements in the middle stage. It can improve the early growth efficiency and continuously optimize the spectral environment in the middle stage. Dynamically adjusting the average color temperature of each stage and driving subsequent lighting with one click simplifies management, improves plant growth quality and system energy saving, and solves the pain point of existing technologies that cannot take into account the full cycle requirements and energy efficiency.

[0198] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0199] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligently controlling the color temperature of LED lights, characterized in that, include: Set up multiple LED light groups and spectral sampling nodes, collect unit power spectra, filter out invalid samples and assemble the effective node spectral matrix; For the target plant, obtain the center of multiple absorption peaks and half-width at half-maximum, and establish and normalize the plant absorption function; A theoretical spectrum is constructed within a preset color temperature range, and discrete sampling and normalization are performed to obtain discrete color temperature sample spectra. In the photosynthetically effective band, an overlap integral is performed between each discrete color temperature spectrum and the plant absorption function to obtain discrete absorption efficiency samples. A continuous absorption efficiency function is constructed using interpolation based on discrete samples, and its derivative is obtained. Set the initial color temperature, step size, and iteration upper limit, perform interval projection and iterative elimination to obtain the optimal color temperature; Based on the normalized spectrum corresponding to the optimal color temperature, partition integration is performed on the three bands of red, green and blue respectively to generate the current setting ratio of each lamp group and drive the output linearly; The absorption function and optimal color temperature are updated at the beginning and middle of the plant growth cycle, the average color temperature of the stage is calculated, the whole cycle current ratio is generated and the drive is executed.

2. The intelligent control method for LED lamp color temperature according to claim 1, characterized in that, The process of setting up multiple LED light groups and spectral sampling nodes, collecting unit power spectra, filtering out invalid samples, and assembling a valid node spectral matrix specifically includes: Multiple LED light groups are arranged in the indoor green plant cultivation area, each group containing independently controllable red, green, and blue sub-light sources; A spectral sampling node is set at a predetermined height directly below each lamp group, and the spectral power density of each node is collected when sampling is started. For each node, normalization is performed in the visible light band; if the total energy in that band is zero, it is marked as an invalid node. Renumber all valid nodes sequentially and count the total number of valid nodes. The normalized spectra of each effective node are assembled into a sampling spectral matrix by columns; When the total number of valid nodes is zero, the output will show a result indicating that there is no valid measured spectral data.

3. The intelligent control method for LED lamp color temperature according to claim 2, characterized in that, For the target plant, multiple absorption peak centers and half-widths (FWHMs) are obtained, and the plant absorption function is established and normalized. Specifically, this includes: Identify the plant species to be controlled; Through absorption experiments of chlorophyll a, chlorophyll b, and carotenoids, the central wavelengths of multiple major absorption peaks of this variety were obtained, forming a set of absorption centers. For each absorption peak, a peak shape is established based on its full width at half maximum (FWHM), and superposition is performed to obtain the unit response absorption function. The absorption function is normalized across the entire wavelength range to obtain the normalized absorption density function.

4. The intelligent control method for LED lamp color temperature according to claim 3, characterized in that, The process of constructing a theoretical spectrum within a preset color temperature range, performing discrete sampling and normalization to obtain discrete color temperature sample spectra specifically includes: Set the theoretical color temperature and the corresponding ideal spectral shape; Generate discrete color temperature samples within a preset color temperature range according to the sampling step size; If the endpoints do not cover the upper limit, add endpoint samples to form the final sample set; For each discrete color temperature, construct a theoretical spectrum and perform normalization, then output the corresponding normalized spectral density.

5. The intelligent control method for LED lamp color temperature according to claim 4, characterized in that, The step of performing an overlap integral between each discrete color temperature spectrum and the plant absorption function in the photosynthetically effective wavelength band to obtain discrete absorption efficiency samples specifically includes: Set upper and lower limits for the wavelength of photosynthetically active radiation; For each discrete color temperature, its normalized spectrum and the plant's normalized absorbance density are overlapped and integrated in that band to obtain the photosynthetically active absorption overlap intensity. The discrete color temperature and its corresponding overlap intensity are combined to form a set of sample pairs.

6. The intelligent control method for LED lamp color temperature according to claim 5, characterized in that, The construction of a continuous absorption efficiency function based on discrete samples using interpolation, and the acquisition of its derivative function, specifically includes: Within a predefined domain, an absorption efficiency function is constructed using Lagrange interpolation based on discrete sample pairs. The derivative function is obtained by taking the derivative of the aforementioned absorption efficiency function.

7. The intelligent control method for LED lamp color temperature according to claim 6, characterized in that, The process of setting the initial color temperature, step size, and iteration upper limit, performing interval projection and iterative elimination to obtain the optimal color temperature specifically includes: Set the initial color temperature, iteration step size, maximum number of iterations, and derivative stopping threshold; In each iteration, candidate update values ​​are generated based on the relationship between the derivative sign and the derivative stopping threshold. Perform interval projection on the candidate values ​​to keep the color temperature within a preset range; When a round trip occurs between two points, the absorption efficiency is compared at the two points and the process is terminated directly based on the better one. When a normal cycle occurs, select the color temperature with the highest absorption efficiency in the sub-set of the cycle state and terminate the cycle. The process terminates at the current color temperature when the derivative meets the stopping threshold. If the iteration limit is reached and the process does not terminate, then the color temperature with the highest absorption efficiency is selected as the optimal color temperature from the evaluated set.

8. The intelligent control method for LED lamp color temperature according to claim 7, characterized in that, The process involves performing partitioned integration on the red, green, and blue bands based on the normalized spectrum corresponding to the optimal color temperature, generating the current setpoint ratio for each lamp group, and linearly driving the output. Specifically, this includes: On the normalized spectrum corresponding to the optimal color temperature, integration is performed on the three fixed bands of red, green and blue respectively to obtain the spectral area ratio of each band; The current setting ratio for each group of lights is generated according to the ratio of red, green, and blue. When all three ratios are zero, set the current ratio of each channel to zero; Each lamp group's drive circuit outputs a linear current according to the current ratio.

9. The intelligent control method for LED lamp color temperature according to claim 8, characterized in that, The process of updating the absorption function and optimal color temperature at the beginning and middle of the plant growth cycle, calculating the average color temperature during the stage, generating the total cycle current ratio, and executing the drive specifically includes: Set the plant growth cycle, collect the absorption function on day zero, and obtain the optimal color temperature for that stage; Repeat the absorption test during the middle of the growth cycle, update the absorption function and obtain a new optimal color temperature; Calculate the average color temperature of the two stages mentioned above, construct their theoretical spectrum, and perform normalization. Integrating the average color temperature spectrum across the red, green, and blue bands yields the current setpoint ratio for the entire cycle. Each lamp group is driven to output according to the full cycle current ratio in subsequent time periods.

Images