Full-spectrum LED intelligent driving and dimming system for plant growth

Abstract

The invention relates to the technical field of plant illumination, and discloses a plant growth full-spectrum LED intelligent driving and dimming system which comprises a full-spectrum LED array module, a driving module, a control module, a sensing module and a feedback regulation and control module. The full-spectrum LED array module comprises five kinds of LED lamp beads and covers the ultraviolet A wave band and the spectral ranges of blue light, red light, far-red light and white light. And the driving module adopts a two-stage structure of a front-stage active PFC circuit and a rear-stage half-bridge LLC resonant converter. According to the plant growth full-spectrum LED intelligent driving and dimming system, a full-spectrum LED array module is optimized through an MATLAB exhaustion method to complete non-uniform layout, an illuminance sensor of a sensing module collects illumination data in a 100 ms period, a feedback regulation and control module calls a light uniformizing control algorithm to be combined with a Lambert formula and an improved Portrait equation to establish a precise mathematical model, and the precise mathematical model is used for driving the full-spectrum LED array module to conduct dimming on the full-spectrum LED array module. The double-closed-loop inner ring adjusts the PWM duty ratio of each LED branch in real time according to the PWM duty ratio, rapidly compensates illumination difference, and thoroughly solves the problem that the center is bright and the periphery is dark.

Description

Technical Field

[0001] This invention relates to the field of plant lighting technology, specifically to a full-spectrum LED intelligent driving and dimming system for plant growth. Background Technology

[0002] Plant factories, as an advanced form of facility agriculture, achieve efficient and high-quality crop production through the artificial control of environmental parameters such as light, temperature, humidity, water, and fertilizer. LED light sources, due to their advantages such as adjustable spectrum, high luminous efficacy, long lifespan, and low heat generation, have become the core lighting equipment in plant factories. However, existing LED lighting systems for plants still suffer from three major technological shortcomings: 1. Insufficient uniformity of illumination: Traditional LED arrays often adopt a uniform layout, resulting in a phenomenon where the center of the receiving surface is bright and the edges are dark, leading to poor uniformity of plant growth; although some systems have attempted to optimize the layout, they have not taken into account the dynamic needs of plant growth, so the effect of improving uniformity is limited, and there is a lack of precise layout optimization schemes for multiple monochrome LEDs. 2. Low precision of spectral regulation: The existing system has insufficient spectral fitting degree and cannot accurately match the light quality requirements of plants at different growth stages; spectral regulation is mostly a fixed mode switching, lacking dynamic adaptation ability based on plant physiological feedback, and it is difficult to meet the differentiated spectral requirements of seedling and growth stages in crop breeding. 3. Poor performance of the driving power supply: The driving power supply has a low power factor, serious harmonic pollution, and insufficient power utilization; the driving architecture is mostly a single-stage converter with low conversion efficiency, which cannot meet the energy consumption requirements of long-term stable operation of plant factories, and lacks efficient collaborative design with intelligent dimming systems. Therefore, a full-spectrum LED intelligent driving and dimming system for plant growth is proposed. Summary of the Invention

[0003] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a smart driving and dimming system for full-spectrum LEDs used in plant growth. This system offers advantages such as improved light uniformity, precise spectral control, and efficient driving coordination, solving the problems of insufficient light uniformity, low spectral control precision, and poor driving power performance in existing plant lighting LED systems.

[0004] (II) Technical Solution To achieve the aforementioned goals of improving light uniformity, precise spectral control, and efficient drive synergy, this invention provides the following technical solution: a full-spectrum LED intelligent drive and dimming system for plant growth, comprising a full-spectrum LED array module, a drive module, a control module, a sensing module, and a feedback control module; The full-spectrum LED array module contains five types of LED beads, covering the ultraviolet A band, blue light, red light, far-red light and white light spectrum range, and adopts a non-uniform layout design; The drive module adopts a two-stage architecture of a front-stage active PFC circuit and a rear-stage half-bridge LLC resonant converter. The front stage is a Boost topology, and the rear stage includes a resonant inductor, a resonant capacitor, and a transformer. The control module uses the UCS512G6 control chip and the STM32F103 main chip as its core to realize the PWM dimming function. The sensing module collects light data and plant physiological parameters.

[0005] The feedback control module is based on a spectral fitting algorithm and a uniform light control algorithm. It dynamically adjusts the spectrum, light intensity, and photoperiod according to the sensing data to achieve precise light control throughout the entire growth cycle of plants.

[0006] Preferably, the five types of LED beads in the full-spectrum LED array module are a UV-A band LED with a center wavelength of 395nm, a blue LED with a center wavelength of 450nm, a red LED with a center wavelength of 660nm, a far-red LED with a center wavelength of 730nm, and a white LED with a color temperature of 3000K. The non-uniform layout is optimized by exhaustive search using MATLAB, with the receiving surface size set to 900mm × 250mm and the distance between the receiving surface and the array module set to 200mm.

[0007] Preferably, the 395nm LEDs are arranged in rows 1, 3, and 5, with 14 LEDs in each row for a total of 42 LEDs; the 450nm LEDs are arranged in rows 2 and 5, with 14 LEDs in each row for a total of 28 LEDs; the 660nm LEDs are arranged in rows 1, 3, 4, and 6, with 14 LEDs in each row for a total of 56 LEDs; the 730nm LEDs are arranged in rows 2, 4, and 6, with 14 LEDs in each row for a total of 42 LEDs; and the remaining positions are filled with 672 3000K white LEDs. The array module has a row spacing of 9.2mm and a column spacing of 8mm.

[0008] Preferably, the front-end active PFC circuit operates in critical conduction mode, the switching device is a MOSFET, the drain-source breakdown voltage is not less than 540V, and the maximum continuous drain current is not less than 11.37A; the PFC boost inductor uses a TDK PC44PQ3535 magnetic core, with 69 turns in the main winding, 6 turns in the auxiliary winding, and an output capacitor capacity of 106μF.

[0009] Preferably, the resonant frequency of the subsequent half-bridge LLC resonant converter is 100kHz, which achieves the best balance between efficiency, electromagnetic interference suppression, and component size; the resonant inductance is 278μH, the resonant capacitor is 9.3nF, the transformer primary winding has 67 turns, the secondary winding has 23 turns, the auxiliary winding has 5 turns, the rectifier diode reverse breakdown voltage is 193.4V, and the maximum conduction current is 1.37A.

[0010] Preferably, the UCS512G6 control chip outputs 5 independent PWM signals, each with a duty cycle adjustment range of 0.1% to 100%, corresponding to an LED bead current adjustment range of 0 rated current. The chip has a built-in gamma correction module with a correction coefficient of 2.2, converting 256 grayscale levels to 65536 levels. The STM32F103 main chip's ADC interface is configured in continuous conversion mode, with the conversion results right-aligned and a sampling time of 1μs.

[0011] Preferably, the UCS512G6 control chip is compatible with the DMX512 protocol and interacts with the handheld addresser via an RS485 bus transceiver, with a data transmission rate of 250kb / s. The handheld addresser uses a 4.3-inch resistive touch LCD screen, supports 1512 channel parameter settings, and adjusts the LED brightness by inputting the self-test parameter value 0255. The relationship curve between the self-test parameter value and the current of a single LED exhibits a power function characteristic.

[0012] Preferably, the spectral fitting algorithm is based on the nonlinear regression least squares method, which uses the lsqnonneg function in MATLAB to solve for the non-negative least squares solution, and is used to match the target spectrum of the plant during the seedling stage and the growth stage.

[0013] Preferably, the uniform illumination control algorithm establishes a mathematical model of LED current and illuminance based on the Lambert formula and the improved Shockley equation; the Lambert formula is... and =2, this value was determined based on fitting experimental data of LED luminous characteristics. Testing verified that when n=2, the model showed the best fit with the actual illumination distribution; the improved Shockley equation is... And goodness of fit By adjusting the PWM duty cycle of each LED branch, uniform control of the illumination on the receiving surface and PPFD is achieved.

[0014] Preferably, the illuminance sensor of the sensing module is model RS-GZ-N01-2, with a measurement range of 0~200000 Lux, and communicates with the STM32F103 main chip via the Modbus-RTU protocol, with a communication distance of not less than 2000 meters; the plant physiological parameter acquisition unit detects the photosynthetic pigment content by ethanol extraction, with detection wavelengths of 665nm, 649nm and 470nm, and measures the fresh weight and dry weight of the aboveground parts of the plant by electronic balance.

[0015] (III) Beneficial Effects Compared with existing technologies, this invention provides a full-spectrum LED intelligent driving and dimming system for plant growth, which has the following beneficial effects: 1. This plant growth full-spectrum LED intelligent driving and dimming system uses a full-spectrum LED array module optimized by MATLAB exhaustive method to achieve a non-uniform layout, creating a physical basis for uniform light. The illuminance sensor of the sensing module collects light data at a 100ms cycle. The feedback control module calls the uniform light control algorithm and combines Lambert's formula and the improved Shockley equation to establish a precise mathematical model. The inner loop of the double closed loop adjusts the PWM duty cycle of each LED branch in real time according to this model, quickly compensating for light differences and stabilizing the light uniformity of the receiving surface at over 80%, completely solving the problem of bright center and dark periphery.

[0016] 2. This plant growth full-spectrum LED intelligent driving and dimming system features a full-spectrum LED array module equipped with five-band dedicated LED beads covering the key spectrum of the entire plant growth cycle. The control module's UCS512G6 control chip outputs five independent PWM signals. The feedback control module implements a spectral fitting algorithm based on nonlinear regression least squares method. Combined with plant physiological parameters collected by the sensing module, the dual closed-loop outer loop dynamically switches between the target spectrum for the seedling stage and the growth stage, achieving fitting degrees of over 92% and 94% respectively, thus achieving precise adaptation of the spectrum to the plant growth stage.

[0017] 3. This plant growth full-spectrum LED intelligent driving and dimming system adopts a two-stage architecture of a front-end active PFC circuit and a rear-end half-bridge LLC resonant converter. The front-end operates in critical conduction mode to improve the power factor to above 0.95 and reduce harmonic pollution. The rear-end achieves a conversion efficiency of over 92% and outputs low ripple at a 100kHz resonant frequency. It provides a stable voltage platform for the 0.1% precision PWM dimming achieved by the dual-chip collaboration of the control module, realizing efficient collaboration between the drive power supply and the intelligent dimming system. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the system structure of the present invention; Figure 2 This is a schematic diagram of the full-spectrum LED array module structure of the present invention; Figure 3 This is a flowchart of the feedback control module algorithm of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments and 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.

[0020] Example 1: This embodiment details the core physical structure (non-uniform layout) and overall architecture of the plant growth full-spectrum LED intelligent driving and dimming system of the present invention, as well as the collaborative workflow of each module.

[0021] The system consists of five core components: a full-spectrum LED array module, a driver module, a control module, a sensing module, and a feedback control module. Each module interacts with the other through hardware interfaces and communication protocols, with the following specific connections: the full-spectrum LED array module is connected to the driver module via a power line, receiving a stable current output from the driver module; the driver module is connected to the control module via a control signal line, receiving a PWM dimming signal output from the control module; the sensing module establishes a data transmission channel with the control module via communication protocols such as Modbus-RTU, uploading the collected physical parameters to the control module; and the control module and the feedback control module interact via an internal data bus, with the calculation results from the feedback control module serving as the basis for the control module's adjustments.

[0022] 1. The core design highlights of the full-spectrum LED array module are its non-uniform layout and precise band selection. It uses five types of dedicated LED beads, namely, ultraviolet A band with a center wavelength of 395nm, blue light at 450nm, red light at 660nm, far-red light at 730nm, and white light with a color temperature of 3000K. This selection is based on the absorption spectrum characteristics of plant photosynthetic pigments (chlorophyll a, chlorophyll b, carotenoids) and the response laws of photoreceptors (photosensitive pigments, blue light receptors, etc.), ensuring that the spectrum covers the key bands required for the entire growth cycle of plants. The LED layout adopts a non-uniform design. Optimization was performed using MATLAB exhaustive search on a scenario with a 900mm×250mm receiving surface and a 200mm distance between the array module and the LEDs. The results showed that 395nm LEDs were placed in rows 1, 3, and 5, with 14 LEDs per row for a total of 42; 450nm LEDs in rows 2 and 5, with 14 LEDs per row for a total of 28; 660nm LEDs in rows 1, 3, 4, and 6, with 14 LEDs per row for a total of 56; and 730nm LEDs in rows 2, 4, and 6, with 14 LEDs per row for a total of 42. The remaining positions were filled with 672 3000K white LEDs. The array module has a row spacing of 9.2mm and a column spacing of 8mm. This layout avoids the problem of a bright center and dark edges caused by traditional uniform layouts, providing a physical basis for subsequent uniform light distribution algorithms.

[0023] 2. The drive module adopts a two-stage architecture: a front-end active PFC circuit and a rear-end half-bridge LLC resonant converter. The front-end Boost topology is used for power factor correction, while the rear-end achieves efficient energy conversion through resonant inductors, resonant capacitors, and transformers. The control module is based on the UCS512G6 control chip and the STM32F103 main chip. The UCS512G6 is responsible for generating 5 independent PWM signals, and the STM32F103 is responsible for data processing and command distribution. The sensing module includes a light intensity sensor and a plant physiological parameter acquisition unit, which collect parameters such as light intensity, spectral distribution, photosynthetic pigment content, and plant fresh / dry weight. The feedback control module has built-in spectral fitting algorithm and light uniformity control algorithm, which dynamically adjust the spectral combination, light intensity, and photoperiod duration based on the sensing data.

[0024] 3. The overall workflow of the plant growth full-spectrum LED intelligent driving and dimming system of the present invention is as follows: Step 1: Initialization phase, the control module configures the parameters of the drive module and the sensing module, setting the output voltage range and resonant frequency of the drive module, and the sampling period and data transmission rate of the sensing module; Step 2: During the startup phase, the driver module converts the mains power into a stable current required by the LED array through a two-level architecture. The full-spectrum LED array emits light according to the initially set spectral ratio and light intensity. Step 3: During the operation phase, the sensing module collects light data and plant physiological parameters at a set cycle and uploads them to the control module in real time; Step 4: In the control phase, the feedback control module calls the spectral fitting algorithm and the light equalization control algorithm to process the sensed data, generate spectral adjustment commands, light intensity adjustment coefficients and light period parameters, and the control module drives the drive module to adjust the output through the PWM signal to realize the dynamic control of the LED array; Step 5: Stabilization phase. The system continuously cycles through the collection-processing-regulation process to maintain the optimal light environment required for plant growth.

[0025] In this embodiment, the five types of LED beads in the full-spectrum LED array cover key wavelength bands, providing a hardware foundation for precise spectral control. The non-uniform layout, through simulation optimization, constructs a preliminary spatial light intensity distribution adapted to plant growth. Its non-linear attenuation characteristics, combined with the physical constraints of 9.2mm row spacing and 8mm column spacing, determine the customization requirements of subsequent algorithms. The two-level architecture of the drive module ensures efficient power conversion, providing a stable energy supply for precise algorithm execution. The dual-chip collaboration of the control module enables precise command execution. The sensing module and the feedback control module form a preliminary closed loop. The collaborative work of each module addresses the core pain points of existing systems, such as insufficient light uniformity, low spectral control accuracy, and poor drive power performance, from an architectural perspective, ensuring the integrity and feasibility of the system.

[0026] Example 2: This embodiment details the hardware implementation scheme of the driver module, focusing on solving the technical pain point of "poor performance of the driver power supply" in existing systems. Through the optimized design of the front-end active PFC circuit and the rear-end half-bridge LLC resonant converter, the power conversion efficiency is improved, the power factor is improved, and harmonic pollution is reduced, providing stable and reliable hardware support for subsequent intelligent dimming algorithms.

[0027] The hardware implementation steps of the driver module are as follows: Step 1: Design the front-end active PFC circuit, adopting a Boost topology and operating in critical conduction mode. This mode enables the PFC circuit to maintain a high power factor over a wide input voltage range. The switching device is a MOSFET. Considering the fluctuation range of the mains input voltage and a 20% safety margin, the drain-source breakdown voltage of the MOSFET should be no less than 540V (preferred range 520-560V) according to engineering calculations. Combined with the maximum output current requirement of the circuit, its maximum continuous drain current is determined to be no less than 11.37A (preferred range 10-13A). This selection ensures that the switching device can work stably under extreme conditions. The boost inductor uses TDK's PC44PQ3535 magnetic core. Based on the electromagnetic coupling efficiency and inductance requirements, combined with the operating frequency and current ripple control target of the PFC circuit, the number of turns in the main winding is calculated to be 69 turns (preferred range 65-75 turns), and the number of turns in the auxiliary winding is 6 turns (preferred range 5-8 turns). This core material and turn count design can effectively reduce hysteresis loss and copper loss. The output capacitor is selected with a capacitance of 106μF (preferred range 100-120μF) to filter out the output voltage ripple, ensure the stability of the output voltage of the front stage, and provide a smooth input for the subsequent circuit.

[0028] Step 2: Design of the subsequent half-bridge LLC resonant converter. First, determine the resonant frequency as 100kHz (preferably within the range of 90-110kHz). This frequency is verified through simulation to achieve the best balance between efficiency, electromagnetic interference (EMI) suppression, and component size: too low a frequency will lead to an increase in component size, while too high a frequency will increase switching losses and EMI interference. 100kHz allows the converter to maintain high conversion efficiency under both full load and light load conditions. The resonant inductance value is set to 278μH (preferred range 260-290μH), and the resonant capacitor capacitance is 9.3nF (preferred range 8-11nF). This parameter combination is determined through LLC resonant characteristic simulation to ensure that the converter operates in the resonant frequency band, achieve zero voltage switching (ZVS), and reduce switching losses. The transformer primary winding has 67 turns (preferred range 62-72 turns), secondary winding has 23 turns (preferred range 20-25 turns), and auxiliary winding has 5 turns (preferred range 4-6 turns). The turns ratio is designed based on the rated operating voltage requirements of the LED array (combining the series and parallel connection of the LED beads and the forward voltage parameters), and precise voltage conversion is achieved through electromagnetic induction. The rectifier diodes selected have a reverse breakdown voltage of 193.4V (preferred range 180-210V) and a maximum conduction current of 1.37A (preferred range 1.2-1.5A). This is calculated based on the peak output current and voltage margin of the subsequent stage to ensure that the diodes are not broken down during rectification and can withstand the maximum operating current.

[0029] Step 3: Interface design between the drive module and the control module. Set up a PWM signal receiving interface in the drive module and connect it to the output terminal of the UCS512G6 control chip in the control module to receive the PWM dimming signal. At the same time, design voltage and current sampling interfaces to collect the output voltage and current data of the drive module in real time and feed them back to the STM32F103 main chip of the control module for overvoltage and overcurrent protection and dimming accuracy calibration.

[0030] Step 4: Electromagnetic compatibility (EMC) design of the driver module. Add a common-mode inductor and X / Y capacitor on the input side to suppress differential-mode and common-mode interference; set a shielding layer between the front and rear stage circuits to reduce electromagnetic coupling interference; adopt a power ground and signal ground separation design in PCB layout to reduce ground loop interference and ensure that the EMI index of the driver module meets relevant standards.

[0031] In this embodiment, the front-end active PFC circuit operates in critical conduction mode. With the help of MOSFETs and optimized inductor parameters, the power factor can be improved to above 0.95, and the harmonic distortion rate (THD) can be reduced to below 10%, solving the problems of low power factor and severe harmonic pollution in traditional drive power supplies. In critical conduction mode, the inductor current rises from zero to its peak value and then falls back to zero in each switching cycle, making the input current waveform track the input voltage waveform, thereby improving the power factor. The low loss characteristics of the PC44PQ3535 magnetic core and the reasonable number of winding turns reduce the energy loss of the inductor and further improve the circuit efficiency. The downstream half-bridge LLC resonant converter achieves a conversion efficiency of over 92% at a resonant frequency of 100kHz through parameter optimization of the resonant inductor, capacitor, and transformer. When the LLC resonant converter operates near the resonant point, the switching devices can achieve ZVS conduction, significantly reducing switching losses. The 100kHz frequency design balances efficiency and component size, avoiding increased switching losses due to excessively high frequencies, while also preventing excessively large inductor and capacitor sizes due to excessively low frequencies. The parameter selection of rectifier diodes and the optimization of transformer turns ratio ensure the stability of the downstream output voltage and current supply capability, with the output ripple voltage controlled within 5%, providing a stable voltage platform for dynamic dimming of the full-spectrum LED array. The overall design of the drive module, through the collaborative work of the front-end and back-end stages, not only solves the problems of low conversion efficiency, serious harmonic pollution, and poor coordination with the intelligent dimming system of existing drive power supplies, but also provides key hardware support for the subsequent 0.1% precision PWM dimming algorithm through the design of high power factor and low ripple output, providing efficient and reliable power support for the long-term stable operation of the system.

[0032] Example 3: This embodiment details the software algorithm implementation of the control module and the feedback regulation module, focusing on solving the technical pain point of "low spectral regulation accuracy" in existing systems. Through customized algorithm design based on hardware characteristics and collaborative work with the control chip, precise regulation of spectrum, light intensity, and light period is achieved. The algorithm and hardware structure are deeply coupled to ensure the irreplaceable regulation effect.

[0033] The implementation steps of the intelligent dimming algorithm are as follows: Step 1: Initialize the control module configuration. Configure the ADC interface of the STM32F103 main chip to continuous conversion mode, right-align the conversion results, and set the sampling time to 1μs. This configuration can improve the ADC sampling rate and accuracy, and ensure fast and accurate acquisition of the analog signals of the sensing module. Configure the parameters of the UCS512G6 control chip, enable the built-in gamma correction module, and set the gamma correction coefficient to 2.2. This value has complementary characteristics with the nonlinear light response curve of plant photoreceptors (such as phytochrome and phototropin), which can effectively improve the fineness of light intensity control in the low grayscale range, convert 256 grayscale levels to 65536 levels, and avoid dimming stuttering under low light intensity. Meanwhile, the UCS512G6 is configured to be compatible with the DMX512 protocol, and it interacts with external control devices through an RS485 bus transceiver. The data transmission rate is set to 250kb / s to ensure fast transmission of control commands. Most importantly, because the pre-amplifier PFC circuit improves the power factor to over 0.95 and the LLC resonant converter achieves low ripple output (ripple voltage ≤5%) at 100kHz, this provides an extremely stable voltage platform for the UCS512G6 chip to perform PWM dimming with 0.1% accuracy. This eliminates the interference of power grid fluctuations on light intensity control and ensures that the dimming signal can be accurately converted into the current change of the LED beads.

[0034] Step 2: Spectral fitting algorithm implementation, based on nonlinear regression least squares method embedded engineering implementation, specifically used to match the target spectrum of plant seedling stage and growth stage; Given the non-uniform array layout described in Embodiment 1, combined with the physical constraints of 9.2mm row spacing and 8mm column spacing and the difference in the light emission angle of the LEDs, the superposition of light intensity of each band on the receiving surface exhibits complex nonlinear convolution characteristics. Traditional linear dimming algorithms will cause distortion in light intensity compensation in the edge region and cannot accurately match the spatial distribution requirements of the target spectrum. The nonlinear regression least squares method used in this embodiment is specifically designed to model the spatial light intensity attenuation characteristics of this physical structure, and is implemented through the following steps: First, target spectral data is collected through the sensing module, and the spectral curve is discretized at 1nm intervals to obtain 401 data points, and the target spectral matrix is ​​constructed. Secondly, the absolute spectral power distribution of five types of LED beads under different currents was collected, and combined with the light intensity attenuation model of non-uniform layout, a spectral database adapted to this system was constructed. Then, the non-negative least squares spectral fitting model, which has been verified based on the MATLAB platform, is engineered and embedded in the STM32F103 embedded chip through fixed-point number operation optimization and lookup table method (LUT). Iterative solution is used to ensure that the number of LED beads is non-negative, avoiding the occurrence of physically unrealizable negative number configurations. Finally, based on the solution results, the driving current ratio of each LED bead is determined, spectral configuration parameters are generated, and the corresponding PWM signal is output through the UCS512G6 control chip.

[0035] Step 3: Implementation of uniform light control algorithm. A mathematical model of LED current and illuminance is established based on Lambert's formula and the improved Shockley equation. This model is also customized for the physical characteristics of non-uniform layout. Lambert's formula is used ,in =2, this value was determined by fitting experimental data on LED luminous characteristics with the spatial distribution law of non-uniform layout, and has been verified by testing. When the value is 2, the model has the best fit with the actual illumination distribution and can accurately describe the spatial light intensity distribution of the LED. The improved Shockley equation is goodness of fit This equation can accurately reflect the relationship between LED forward current and forward voltage, providing a mathematical basis for precise current control. The theoretical illuminance at each point on the receiving surface is calculated using this model and compared with the actual illuminance collected by the sensing module to generate an illuminance error matrix. Based on the error matrix, the PWM duty cycle of each LED branch is adjusted to achieve uniform control of the illuminance and PPFD on the receiving surface.

[0036] Step 4: PWM dimming signal generation and optimization. The UCS512G6 control chip outputs 5 independent PWM signals, each with a duty cycle adjustment range of 0.1%-100%, corresponding to an LED bead current adjustment range of 0-rated current. The STM32F103 main chip adjusts the PWM duty cycle parameters of the UCS512G6 in real time according to the control instructions of the feedback control module. At the same time, it collects the actual current of the LED branch through the ADC interface and compares it with the target current. The proportional-integral (PI) control algorithm is used to fine-tune the PWM duty cycle to ensure the accuracy of current control. The PI parameters are calibrated through simulation and experiments to further compensate for the current distribution differences caused by non-uniform layout.

[0037] In this embodiment, the spectral fitting algorithm is engineered using a nonlinear regression least squares method for non-uniform layout and physical constraints, which can achieve a target spectral fitting degree of over 92% during the seedling stage and over 94% during the growth stage. The construction of the target spectral matrix and the spectral database adapted to the non-uniform layout ensures the accuracy of the data source for fitting. Embedded engineering porting solves the problem of running complex algorithms on resource-constrained chips. The non-negative least squares solution avoids invalid configuration, so that the fitting results not only meet the physical implementation requirements, but also adapt to the nonlinear convolution characteristics of spatial light intensity, thereby accurately matching the spectral requirements of different growth stages of plants and solving the problems of insufficient spectral fitting degree and the disconnect between spectral adjustment and physical structure in existing systems. The uniform light control algorithm constructs an accurate mathematical model using the Lambert formula and the improved Shockley equation, achieving a goodness-of-fit of [missing information]. This ensures a high degree of consistency between the model and actual working conditions. The PWM duty cycle adjustment based on the error matrix can improve the uniformity of illumination on the receiving surface to over 80%. The Lambert formula accurately describes the spatial light intensity distribution of the LED, and the improved Shockley equation accurately reflects the relationship between current and voltage. The combination of the two can achieve accurate prediction of illumination. The error matrix provides a clear direction for control. By adjusting the PWM duty cycle of each branch, the illumination difference caused by the non-uniform layout is compensated, and the uniformity is improved. The dual-chip collaboration of the control module and the PI adjustment algorithm, combined with a stable hardware power supply platform, enable PWM dimming accuracy to reach 0.1%. The gamma correction module improves the grayscale level to 65536 levels, ensuring smooth dimming without flicker. The UCS512G6 focuses on PWM signal generation, while the STM32F103 is responsible for data processing and accuracy calibration. The clear division of labor between the two chips improves the response speed. The PI adjustment algorithm can compensate for current deviation in real time, and the gamma correction conforms to the response characteristics of plant photoreceptors. The hardware platform provides a stable foundation. The three work together to achieve precise control of spectrum and light intensity, solving the pain point of low spectral control accuracy in existing systems.

[0038] Example 4: This embodiment details the dual closed-loop feedback mechanism of the sensing module, feedback control module, and other modules, focusing on solving the technical pain point of "insufficient uniformity of illumination" in existing systems. By distinguishing between "fast and slow loops" in the dynamic control design, it achieves real-time optimization of the illumination environment and precise adaptation to long-term growth strategies, avoiding logical flaws in the real-time performance of the technical solution.

[0039] The steps to implement dual closed-loop feedback are as follows: Step 1: Data acquisition for the sensing module. The illuminance sensor selected is the RS-GZ-N01-2 model, with a measurement range of 0~200000 Lux. It communicates with the STM32F103 main chip via the Modbus-RTU protocol, with a communication distance of no less than 2000 meters. The sensor's measurement range and communication distance are designed to ensure accurate acquisition of illuminance data even in large-area lighting scenarios. The sensor's sampling period is set to 100ms, which can capture changes in illuminance intensity in real time, providing data support for real-time light environment adjustment. The plant physiological parameter acquisition unit is equipped with an automatic sampling unit (robotic arm) to acquire plant samples and transport them to the built-in micro biochemical analysis module. The photosynthetic pigment content is automatically quantified by ethanol extraction combined with a spectrophotometer (665nm, 649nm, 470nm wavelength detection). These wavelengths correspond to the characteristic absorption peaks of chlorophyll a, chlorophyll b and carotenoids, which can accurately quantify the photosynthetic pigment content. The fresh and dry weight of the aboveground parts of the plant were measured using an electronic balance. The sampling period was set to 24 hours, which was determined based on the rate of change of the plant's physiological growth parameters. This avoided frequent sampling from interfering with plant growth, while ensuring that physiological changes at key growth stages were captured, providing a basis for adjusting long-term growth strategies.

[0040] Step 2: Data transmission and preprocessing. The light data and plant physiological parameters collected by the sensing module are transmitted to the control module via the communication bus. The STM32F103 main chip preprocesses the received data. A moving average filtering algorithm is used on the illumination data to remove random interference noise and ensure the accuracy of real-time light feedback; By detecting outliers in plant physiological parameters and eliminating abnormal data caused by measurement errors, the reliability of data input to the feedback control module is ensured, providing a high-quality basis for adjusting growth strategies.

[0041] Step 3: Implementation of the dual closed-loop feedback control strategy. This system adopts a dual closed-loop control strategy: the inner loop is a real-time feedback loop based on the light intensity sensor, and the outer loop is a growth strategy loop based on plant physiological parameters. Inner loop (real-time light environment loop): sampling period of 100ms, used for instantaneous light uniformity control; the feedback control module calculates the light uniformity based on the real-time data of the illuminance sensor and calls the light uniformity control algorithm. If the uniformity is lower than the set threshold (80%), a light uniformity adjustment command is generated to adjust the PWM duty cycle of the LED branches in the edge area and the center area, increase the light intensity in the edge area and moderately reduce the light intensity in the center area, quickly compensate for the light difference, and solve the core pain point of insufficient light uniformity. Outer ring (growth strategy ring): sampling period 24h, used to correct the target function parameters of spectral fitting; the feedback control module determines the plant growth stage (seedling stage or growth stage) based on plant physiological parameters (photosynthetic pigment content, fresh weight / dry weight growth rate), calls the target spectral data of the corresponding stage, calculates the deviation between the current spectrum and the target spectrum through the spectral fitting algorithm, generates spectral adjustment instructions, and clarifies the current adjustment ratio of each LED bead. At the same time, the compatibility between light intensity and photoperiod is judged based on the growth rate. If the growth rate is lower than the set threshold, it indicates that the light intensity is insufficient or the photoperiod is unreasonable, and a light intensity increase command or a photoperiod adjustment command is generated (such as adjusting the photoperiod from 12h day / 12h night to 14h day / 10h night).

[0042] Step 4: Command Execution and Closed-Loop Optimization. The control module converts the spectral adjustment command, uniform light adjustment command, and light intensity / photoperiod adjustment command generated by the feedback regulation module into control signals. The drive module adjusts the output current according to the control signals, and the full-spectrum LED array emits light according to the new parameters. The sensing module continues to collect new light data and plant physiological parameters at a set period and transmits them to the control module for preprocessing and algorithm processing, forming a dual closed-loop process of "real-time acquisition-instantaneous regulation" and "periodic acquisition-strategy optimization". This ensures the real-time stability of the light environment and achieves dynamic adaptation between light strategy and plant growth.

[0043] Step 5: Dynamic adaptation and expansion; the dual closed-loop feedback system can dynamically adjust the control strategy according to changes in plant growth status. When plants transition from the seedling stage to the growth stage, the content of photosynthetic pigments increases significantly, and the growth strategy loop automatically switches the target spectrum to increase the ratio of red light to far-red light. When the plant growth rate increases, the strategy loop appropriately increases the light intensity to meet the needs of photosynthesis. When the real-time light environment loop detects that the uniformity of light on the receiving surface decreases due to plant growth (such as differences in plant height), it automatically adjusts the PWM duty cycle of the corresponding LED branch to maintain uniform light and ensure that the system adapts to the dynamic needs of the plant's entire growth cycle.

[0044] In this embodiment, the high-precision data acquisition of the sensing module provides reliable input for the dual closed-loop feedback; the wide measurement range, high resolution and 100ms sampling period of the RS-GZ-N01-2 sensor ensure accurate and efficient real-time illumination data; the integration of the automatic sampling unit and the micro biochemical analysis module realizes the automated quantification of photosynthetic pigment content; the measurement of fresh weight / dry weight by the electronic balance can directly reflect the plant growth status; the preprocessing algorithm removes noise and outliers, providing high-quality data support for dual closed-loop regulation; The dual closed-loop feedback mechanism enables dynamic optimization of the lighting environment, ensuring that the uniformity of illumination on the receiving surface remains stable at over 80%. The inner loop rapidly compensates for lighting differences through millisecond-level feedback. When the central area is too brightly lit, the PWM duty cycle of the corresponding LED branch is reduced; when the edge area is not bright enough, the duty cycle of the corresponding branch is increased, solving the problem of a bright center and dark periphery caused by the uniform layout of traditional systems. The outer loop dynamically adjusts the spectrum and photoperiod strategy through 24-hour cycle physiological parameter feedback, avoiding the problem of lag in physiological parameter measurement in real-time feedback, and achieving precise matching between the lighting environment and the plant growth stage. The dynamic adaptation strategy enables precise matching of light environment with the entire growth cycle of plants. The photosynthetic pigment content and growth rate of plants vary at different growth stages, and the corresponding requirements for spectrum, light intensity and photoperiod are different. The dual closed-loop feedback system ensures real-time uniform light through the inner loop and optimizes long-term strategies through the outer loop. The two work together to ensure that the seedling stage and the growth stage can obtain the optimal light conditions, further improving the consistency and efficiency of plant growth, and solving the problems of existing systems lacking dynamic adaptation capabilities and having a single feedback mechanism.

[0045] The dual closed-loop feedback mechanism in this embodiment not only solves the problem of insufficient light uniformity, but also realizes dynamic adaptation between the light environment and plant growth, making the system widely applicable and stable, and providing a guarantee for precise light regulation throughout the entire plant growth cycle.

[0046] In summary, this plant growth full-spectrum LED intelligent driving and dimming system utilizes a full-spectrum LED array module optimized using MATLAB exhaustive search to achieve a non-uniform layout, providing a physical foundation for uniform illumination. The illuminance sensor in the sensing module collects illumination data at a 100ms cycle. The feedback control module calls a uniform illumination control algorithm combined with the Lambert formula and an improved Shockley equation to establish a precise mathematical model. The inner loop of the dual closed loop adjusts the PWM duty cycle of each LED branch in real time accordingly, quickly compensating for illumination differences and ensuring that the illumination uniformity of the receiving surface is stabilized at over 80%, completely solving the problem of bright center and dark periphery.

[0047] Furthermore, this plant growth full-spectrum LED intelligent driving and dimming system features a full-spectrum LED array module equipped with five-band dedicated LED beads covering the key spectrum of the entire plant growth cycle. The control module's UCS512G6 control chip outputs five independent PWM signals, and the feedback control module implements a spectral fitting algorithm based on nonlinear regression least squares method. Combined with the plant physiological parameters collected by the sensing module, the dual closed-loop outer loop dynamically switches the target spectrum for the seedling stage and the growth stage, achieving fitting degrees of over 92% and 94% respectively, thus achieving precise adaptation of the spectrum to the plant growth stage.

[0048] Furthermore, this plant growth full-spectrum LED intelligent driving and dimming system employs a two-stage architecture in its driving module: a front-end active PFC circuit and a rear-end half-bridge LLC resonant converter. The front-end operates in critical conduction mode to improve the power factor to above 0.95 and reduce harmonic pollution, while the rear-end achieves a conversion efficiency of over 92% and low output ripple at a 100kHz resonant frequency. This provides a stable voltage platform for the 0.1% precision PWM dimming achieved by the dual-chip collaboration of the control module, enabling efficient collaboration between the driving power supply and the intelligent dimming system. This solves the problems of insufficient light uniformity, low spectral control accuracy, and poor driving power supply performance in existing plant lighting LED systems.

[0049] The relevant modules involved in this system are all hardware system modules or functional modules that combine computer software programs or protocols with hardware in the prior art. The computer software programs or protocols involved in these functional modules are technologies known to those skilled in the art and are not improvements to this system. The improvement of this system lies in the interaction or connection between the modules, that is, in improving the overall structure of the system to solve the corresponding technical problems that this system aims to address.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A smart LED driving and dimming system for plant growth across the entire spectrum, characterized in that, It includes a full-spectrum LED array module, a driving module, a control module, a sensing module, and a feedback control module; The full-spectrum LED array module contains five types of LED beads, covering the ultraviolet A band, blue light, red light, far-red light and white light spectrum range, and adopts a non-uniform layout design; The drive module adopts a two-stage architecture of a front-stage active PFC circuit and a rear-stage half-bridge LLC resonant converter. The front stage is a Boost topology, and the rear stage includes a resonant inductor, a resonant capacitor, and a transformer. The control module uses the UCS512G6 control chip and the STM32F103 main chip as its core to realize the PWM dimming function. The sensing module collects light data and plant physiological parameters; The feedback control module is based on a spectral fitting algorithm and a uniform light control algorithm. It dynamically adjusts the spectrum, light intensity, and photoperiod according to the sensing data to achieve precise light control throughout the entire growth cycle of plants.

2. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 1, characterized in that, The five types of LED beads in the full-spectrum LED array module are a UV-A band LED with a center wavelength of 395nm, a blue LED with a center wavelength of 450nm, a red LED with a center wavelength of 660nm, a far-red LED with a center wavelength of 730nm, and a white LED with a color temperature of 3000K. The non-uniform layout was optimized using the MATLAB exhaustive method, with the receiving surface size set to 900mm×250mm and the distance between the receiving surface and the array module set to 200mm.

3. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 2, characterized in that, The 395nm LEDs are arranged in rows 1, 3, and 5, with 14 LEDs in each row for a total of 42; the 450nm LEDs are arranged in rows 2 and 5, with 14 LEDs in each row for a total of 28; the 660nm LEDs are arranged in rows 1, 3, 4, and 6, with 14 LEDs in each row for a total of 56; the 730nm LEDs are arranged in rows 2, 4, and 6, with 14 LEDs in each row for a total of 42; and the remaining positions are filled with 672 3000K white LEDs. The array module has a row spacing of 9.2mm and a column spacing of 8mm.

4. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 1, characterized in that, The front-end active PFC circuit operates in critical conduction mode, with MOSFETs as the switching devices, a drain-source breakdown voltage of not less than 540V, and a maximum continuous drain current of not less than 11.37A. The PFC boost inductor uses a TDK PC44PQ3535 magnetic core, with 69 turns in the main winding, 6 turns in the auxiliary winding, and an output capacitor capacity of 106μF.

5. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 1, characterized in that, The resonant frequency of the subsequent half-bridge LLC resonant converter is 100kHz, which achieves the best balance between efficiency, electromagnetic interference suppression, and component size. The resonant inductance is 278μH, the resonant capacitor is 9.3nF, the transformer primary winding has 67 turns, the secondary winding has 23 turns, the auxiliary winding has 5 turns, the rectifier diode reverse breakdown voltage is 193.4V, and the maximum conduction current is 1.37A.

6. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 1, characterized in that, The UCS512G6 control chip outputs 5 independent PWM signals, each with a duty cycle adjustment range of 0.1% to 100%, corresponding to an LED bead current adjustment range of 0 rated current. The chip has a built-in gamma correction module with a correction coefficient of 2.2, converting 256 grayscale levels to 65536 levels. The STM32F103 main chip's ADC interface is configured for continuous conversion mode, with the conversion results right-aligned and a sampling time of 1μs.

7. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 6, characterized in that, The UCS512G6 control chip is compatible with the DMX512 protocol and interacts with the handheld addresser via an RS485 bus transceiver, with a data transmission rate of 250kb / s. The handheld addresser uses a 4.3-inch resistive touch LCD screen, supports 1512 channel parameter settings, and adjusts the LED brightness by inputting the self-test parameter value 0255. The relationship curve between the self-test parameter value and the current of a single LED exhibits a power function characteristic.

8. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 1, characterized in that, The spectral fitting algorithm is based on the nonlinear regression least squares method. It uses the lsqnonneg function in MATLAB to solve for the non-negative least squares solution, which is used to match the target spectrum of the seedling stage and the growth stage of plants.

9. The intelligent driving and dimming system for full-spectrum LEDs for plant growth according to claim 1, characterized in that, The uniform light control algorithm establishes a mathematical model of LED current and illuminance based on the Lambert formula and the improved Shockley equation; the Lambert formula is... and =2, this value was determined based on fitting experimental data of LED luminous characteristics. Testing verified that when n=2, the model showed the best fit with the actual illumination distribution; the improved Shockley equation is... And goodness of fit By adjusting the PWM duty cycle of each LED branch, uniform control of the illumination on the receiving surface and PPFD is achieved.

10. A plant growth full-spectrum LED intelligent driving and dimming system according to claim 1, characterized in that, The illuminance sensor of the sensing module is model RS-GZ-N01-2, with a measurement range of 0~200000 Lux. It communicates with the STM32F103 main chip via the Modbus-RTU protocol, with a communication distance of no less than 2000 meters. The plant physiological parameter acquisition unit detects the photosynthetic pigment content by ethanol extraction, with detection wavelengths of 665nm, 649nm, and 470nm. It also measures the fresh and dry weight of the aboveground parts of the plant using an electronic balance.

Images