Multi-band control method and device for full-spectrum LED lamp strip

Abstract

The invention relates to a multi-band control method and device for a full-spectrum LED lamp strip. The method comprises the steps that red fluorescent powder, green fluorescent powder and yellow fluorescent powder are mixed with organic silicon resin according to a preset proportion to form fluorescent glue, the surface of a single blue light LED chip is coated with the fluorescent glue for curing, and an LED lamp bead unit is prepared; an LED mounting bonding pad array with a power supply circuit and a signal circuit is manufactured on a double-sided copper-clad conductive substrate through a photoetching process, and LED lamp bead units are welded on the substrate to form a lamp strip array structure; the method comprises the following steps: acquiring spectral parameters such as peak wavelength, wavelength bandwidth and relative light intensity of each LED lamp bead under different driving currents, establishing spectral characteristic mapping data, and performing segmentation processing on a full-spectrum target curve based on the spectral characteristic mapping data to obtain frequency band control data to adjust the light emitting color of the LED lamp beads, thereby realizing full-spectrum light emitting.

Description

Technical Field

[0001] The present invention relates to the technical field of LED light strip control, and in particular to a multi-band control method and device for a full-spectrum LED light strip. Background Art

[0002] As a new generation of lighting source, full-spectrum LED light strips are widely used in plant growth, commercial lighting and display fields. Traditional LED light strips usually use single-wavelength LED chips and phosphors to achieve white light conversion, which is difficult to meet the refined requirements of spectral characteristics in different application scenarios. At the same time, since the performance of phosphors is significantly affected by temperature, and there are differences in luminous characteristics between different batches of LED lamp beads, the light strips will have spectral drift and unevenness problems during actual use.

[0003] At present, the industry generally adopts the RGB three-primary color mixing scheme to achieve full spectrum adjustment, but this scheme has problems such as insufficient color reproduction ability and low energy conversion efficiency. In addition, the existing LED control methods mostly use a single PWM dimming technology, which is prone to stroboscopic effects in high-precision dimming scenarios, and lacks a compensation mechanism for temperature changes and aging effects of LED devices, making it difficult to ensure long-term stable spectral output characteristics. Summary of the invention

[0004] The main purpose of the present invention is to provide a multi-band control method and device for a full-spectrum LED light strip to achieve full-spectrum luminescence of the LED light strip.

[0005] To achieve the above object, the present invention provides a multi-band control method for a full-spectrum LED light strip, comprising the following steps: Mixing fluorescent powder with silicone resin to obtain a fluorescent glue mixture, wherein the fluorescent powder includes red fluorescent powder, green fluorescent powder and yellow fluorescent powder mixed in a preset mass ratio; Applying the fluorescent glue mixture on the surface of a single blue light LED chip and performing a curing process to obtain an LED lamp bead unit; Performing photolithography on the double-sided copper-clad conductive substrate to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit; Performing array welding processing on the LED lamp bead units on the light strip substrate to obtain an LED light strip array structure; Collecting spectral parameters of each LED lamp bead in the LED light strip array structure under multiple sets of driving currents to obtain spectral characteristic mapping data, wherein the spectral parameters include peak wavelength, wavelength bandwidth and relative light intensity; The full spectrum target curve is wavelength segmented according to the spectral characteristic mapping data to obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence, wherein the frequency band control data includes the wavelength range and the light intensity contribution value.

[0006] The present invention also provides a multi-band control device for a full-spectrum LED light strip, comprising: A mixing module, used for mixing phosphor with silicone resin to obtain a phosphor glue mixture, wherein the phosphor includes red phosphor, green phosphor and yellow phosphor mixed in a preset mass ratio; A curing module, used for applying the fluorescent glue mixture to the surface of a single blue light LED chip for curing to obtain an LED lamp bead unit; A photolithography module, used for performing photolithography on a double-sided copper-clad conductive substrate to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit; A welding module, used for performing array welding processing on the LED lamp bead units on the light strip substrate to obtain an LED light strip array structure; An acquisition module is used to acquire spectral parameters of each LED lamp bead in the LED light strip array structure under multiple groups of driving currents to obtain spectral characteristic mapping data, wherein the spectral parameters include peak wavelength, wavelength bandwidth and relative light intensity; A processing module is used to perform wavelength segmentation processing on the full-spectrum target curve according to the spectral characteristic mapping data, obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full-spectrum luminescence, wherein the frequency band control data includes a wavelength range and a light intensity contribution value.

[0007] In summary, the present invention adopts a convolutional long short-term memory network with an integrated attention mechanism to achieve accurate modeling of spectral errors, captures temporal correlations through a bidirectional LSTM structure, and accurately extracts key spectral features in combination with spatial and channel attention mechanisms, thereby improving control accuracy by 30%; by establishing a relationship model between LED junction temperature and wavelength drift, temperature adaptive compensation is achieved, and stable spectral output characteristics are maintained in the range of -20°C to 85°C; a two-level PWM architecture is innovatively adopted to achieve 16-bit high-precision dimming, and visible stroboscopic effects are eliminated through time interleaving modulation technology, and multiple groups of staggered PWM signal designs significantly reduce power supply ripple; at the same time, real-time status monitoring and fault diagnosis of LED lamp beads are achieved, with overcurrent protection and short-circuit detection functions, and long-term stable spectral output is ensured through closed-loop control; in addition, the present invention also simplifies the production process, improves product yield, reduces manual debugging costs, and realizes intelligent control. The system has good scalability and is suitable for a variety of application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 1 is a schematic diagram of the steps of a multi-band control method for a full-spectrum LED light strip according to an embodiment of the present invention; Figure 2 It is a structural block diagram of a multi-band control device for a full-spectrum LED light strip in one embodiment of the present invention.

[0009] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0010] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0011] Reference Figure 1 This embodiment provides a multi-band control method for a full-spectrum LED light strip, comprising the following steps: S1, mixing phosphor and silicone resin to obtain a phosphor glue mixture, wherein the phosphor includes red phosphor, green phosphor and yellow phosphor mixed in a preset mass ratio; Among them, the wavelength of the red phosphor is measured to obtain a first phosphor material with a central wavelength in the range of 620 to 630 nanometers; the wavelength of the green phosphor is measured to obtain a second phosphor material with a central wavelength of 520-530nm; the wavelength of the yellow phosphor is measured to obtain a third phosphor material with a central wavelength of 580-590nm. The first phosphor material, the second phosphor material and the third phosphor material are mixed in a mass ratio of 2:3:1 to obtain a uniform mixed phosphor material. The selection of this ratio is intended to ensure that the prepared fluorescent glue has excellent luminous properties and color expression. The mixed phosphor material and the silicone resin are mixed in a preset mass ratio of 3:7 to form an initial fluorescent glue mixture. The initial fluorescent glue mixture is uniformly stirred to ensure the uniform distribution of all components to improve the optical properties and mechanical strength of the final product. In order to ensure the stable performance of the prepared fluorescent glue mixture in practical applications, its viscosity is tested. By measuring the viscosity parameters of the fluorescent glue mixture, its fluidity and processing adaptability are understood. If the viscosity is too high or too low, it will affect its coating effect on the surface of the LED chip. According to the measured viscosity parameters, the temperature of the fluorescent glue mixture is adjusted. By adjusting the temperature of the mixture, its viscosity is effectively changed to achieve an ideal flow state.

[0012] S2, applying the fluorescent glue mixture to the surface of the single blue light LED chip for curing to obtain an LED lamp bead unit; Specifically, a wavelength test is performed on a single blue LED chip to ensure that its central wavelength is within the range of 440 to 460 nanometers, confirming that the luminous characteristics of the chip meet the requirements. A power test is performed on the chip to be packaged to obtain its luminous power, which is between 0.2 and 1 watt. This test ensures that the selected chip has good light output capability. Only qualified chips that pass the power test can continue the subsequent packaging process. Metal pads are set for qualified chips to facilitate subsequent connection and packaging. The formed pad structure needs to be a copper pad with a surface silver plating and a thickness of 35 microns, which effectively improves the reliability and conductivity of welding. After the pad setting is completed, the dispensing equipment is appropriately set with parameters according to the viscosity parameters of the previous fluorescent glue mixture to ensure that the subsequent dispensing process can proceed smoothly. After the dispensing process parameters are set, the fluorescent glue mixture is dispensed. Ensure that the thickness of the formed fluorescent glue layer is between 0.3 and 0.5 mm. After the dispensing is completed, the formed fluorescent glue layer is subjected to a segmented temperature rise curing process to ensure its stability and uniformity during the curing process to form an initial LED lamp bead unit. Perform spectral detection on the initial LED lamp bead unit to obtain LED luminous spectrum data and provide information about the spectral characteristics of LED lamp beads under different working conditions. Screen and classify according to the LED luminous spectrum data to identify LED lamp bead units that meet the design requirements.

[0013] S3, performing photolithography on the double-sided copper-clad conductive substrate to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit; It should be noted that the thickness of the double-sided copper-clad conductive substrate is measured to confirm that its thickness is 1.6 mm, and the thickness of the copper foil should be 35 microns. These parameters are the key to ensuring that the substrate material has good electrical conductivity and mechanical strength. After the measurement is completed, the circuit pattern design of the substrate material is carried out, and a conductive circuit diagram including a power supply circuit, a signal circuit and an LED mounting pad is designed according to the required functions and layout. This design not only needs to consider electrical performance, but also needs to ensure the rationality and aesthetics of the wiring to achieve the best use effect. The conductive circuit diagram is input into the photolithography equipment for circuit exposure. After exposure, the development pattern of the conductive circuit is obtained by development treatment, and the designed circuit diagram is accurately transferred to the substrate material. Based on the development pattern, the substrate material is etched to remove the unnecessary copper layer to obtain the initial light strip substrate. The etching process needs to strictly control the time and conditions to ensure uniform etching and avoid damage to the circuit. After etching, the surface of the initial light strip substrate is coated with white solder resist ink to form a solder resist protective layer with a thickness of 15 to 20 microns. This protective layer can prevent solder short circuits generated during subsequent welding, improve the corrosion resistance and reliability of the substrate, and extend the service life of the LED light strip. After the solder resist ink is applied, it is cured to ensure that the ink is firmly attached to the surface of the substrate. The formed solder resist protective layer is processed for LED installation holes to obtain installation holes with a diameter larger than the diameter of the LED lamp bead base. Ensure that the LED lamp bead can be smoothly inserted and fixed on the substrate during installation, while avoiding installation difficulties caused by insufficient aperture. The back of the initial light strip substrate is patterned with a heat dissipation copper layer to ensure that the area of ​​the heat dissipation copper layer accounts for more than 80% of the total area of ​​the substrate. This design can effectively improve the heat dissipation performance of the LED light strip, avoid lamp bead failure due to overheating, and enhance the stability and durability of the overall structure. The optimization of the heat dissipation structure helps to improve the working efficiency and extend the service life of the LED light strip. After the heat dissipation copper layer is processed, the pad spacing of the installation through hole is measured to ensure that the center spacing between adjacent pads is 5 mm. The precise control of this parameter helps to achieve compact installation of LED lamp beads, ensure the reasonable layout of the circuit, and improve the overall performance and aesthetics of the light strip.

[0014] S4, performing array welding processing on the LED lamp bead units on the light strip substrate to obtain an LED light strip array structure; Specifically, the LED lamp bead unit and the light strip substrate are preheated to achieve a preheating temperature in the range of 150 to 180°C, reducing the thermal shock during the welding process, ensuring that the components to be welded can be subsequently operated at a more stable temperature, and improving the reliability of welding. After the preheating treatment is completed, the components to be welded will be reflow soldered, and the welding temperature should be controlled between 235 and 245°C to ensure that the solder is fully melted and a good solder joint connection is formed. After the reflow soldering is completed, the initial welding structure is optically scanned to obtain the position deviation data of the LED lamp bead. In this process, the installation accuracy of the LED lamp bead is ensured, and the position deviation data should be less than 0.1 mm to ensure that the lamp bead can emit light normally during operation and avoid beam scattering or unevenness. The solder joints of the initial welding structure are X-rayed to evaluate the quality of the solder joints to ensure that there is no cold soldering or false soldering, thereby improving the overall reliability of the product. Through these two tests, solder joint quality data can be obtained. After completing the optical scanning and solder joint detection, the interfaces are installed at the head and tail ends of the light strip substrate to form an LED light strip assembly with a power interface of a DC5V power supply terminal and a standard SPI communication interface. The setting of these interfaces can ensure that the light strip assembly can be easily connected to the external power supply and control system to provide support for the normal operation of the light strip. After completing the interface installation, a DC5V voltage power-on test is performed to obtain the current test data of each lamp bead, with the goal of reaching the current standard of 20mA / lamp bead. This test can verify the normal working state of the lamp bead. The quality judgment result of the LED light strip can be obtained by comprehensively analyzing the LED lamp bead position deviation data, solder joint quality data and current test data. Confirm the working state of each lamp bead and evaluate whether the performance of the overall light strip meets the design requirements. According to the quality judgment results, classification and screening are carried out to ensure that only LED light strip assemblies that meet the quality standards can enter the next production or sales process to obtain the LED light strip array structure.

[0015] S5, collecting spectral parameters of each LED lamp bead under multiple sets of driving currents for the LED light strip array structure to obtain spectral characteristic mapping data, where the spectral parameters include peak wavelength, wavelength bandwidth, and relative light intensity; Among them, the driving current gradient is set for each LED lamp bead in the LED light strip array structure to obtain multiple groups of test current values ​​in the range of 0 to 60 mA, and the spectral characteristics of the LED lamp bead under different current conditions are evaluated. Based on each group of test current values, the LED lamp bead is spectrally scanned to obtain the original spectral data with a wavelength range of 380 to 780 nanometers and a resolution of 1 nanometer, so as to fully capture the luminous characteristics of the LED lamp bead under different current driving. The peak wavelength is extracted from the original spectral data to help identify the central luminous wavelength value of each LED lamp bead under different driving currents, reflecting the relationship between the luminous characteristics of the LED lamp bead and the current. At the same time, the half-height width is calculated for the original spectral data to obtain the wavelength bandwidth value of each LED lamp bead under different driving currents. The half-height width is an important parameter to describe the shape of the spectral peak, which can provide important information about the luminous characteristics of the LED lamp bead. The energy integration operation is performed on the original spectral data to calculate the relative light intensity value of each LED lamp bead under different driving currents. This data helps to evaluate the light output capacity of the lamp bead. The extracted central emission wavelength value, wavelength bandwidth value and relative light intensity value are data-correlated to obtain the emission characteristic curve of each LED lamp bead, showing the performance changes of the LED lamp bead under different driving currents. At the same time, the emission characteristic curve is mathematically modeled to obtain the mapping relationship model between the driving current and the spectral parameters. The established mapping relationship model is stored in the database to form spectral characteristic mapping data.

[0016] S6, performing wavelength segmentation processing on the full spectrum target curve according to the spectral characteristic mapping data, obtaining the frequency band control data of each LED lamp bead and adjusting the luminous color of each LED lamp bead to achieve full spectrum luminescence, the frequency band control data including the wavelength range and the light intensity contribution value.

[0017] Specifically, the full spectrum target curve is divided into wavelength intervals so as to obtain multiple wavelength sub-intervals within the range of 380 to 780 nanometers. The light intensity of each wavelength sub-interval is calculated to obtain the target spectrum intensity distribution data. According to the spectrum characteristic mapping data, the LED lamp beads are allocated to the wavelength sub-intervals to ensure that each wavelength sub-interval can correspond to a suitable LED lamp bead group so as to achieve the expected spectrum effect in the actual light-emitting process. The luminous spectrum of each LED lamp bead group is superimposed to obtain the combined spectrum data of the LED lamp bead group, showing the spectrum characteristics when multiple lamp beads emit light together. Based on the difference calculation between the combined spectrum data and the target spectrum intensity distribution data, the spectrum error data is obtained to reveal the deviation between the combined spectrum and the target spectrum, providing a direction for further adjustment. According to the spectrum error data, the driving current of the LED lamp bead group is compensated and calculated to obtain the compensation value of the driving current. By adjusting the current, it is ensured that each LED lamp bead can still emit the expected spectrum characteristics under different driving conditions. The driving current compensation value is converted into a PWM signal to obtain the corresponding PWM control parameter. The compensation value is converted into a PWM signal suitable for LED lamp bead driving to achieve precise current regulation. Input the PWM control parameters into the LED control unit to adjust the driving current of each LED lamp bead in real time to achieve the desired luminous color adjustment and spectral output. Through the above steps, the frequency band control data of each LED lamp bead is obtained, and finally the luminous color of each lamp bead is adjusted to achieve a full-spectrum luminous effect. The frequency band control data includes the wavelength range of each LED lamp bead and the light intensity contribution value, ensuring that the light strip can show ideal spectral characteristics in different application scenarios.

[0018] Sliding window sampling is performed on the combined spectral data and the target spectral intensity distribution data to generate a time series feature matrix, which contains the characteristic information of the light intensity at each wavelength changing with time. The time series feature matrix is ​​input into a three-layer convolutional feature extraction network for processing. The convolutional neural network extracts local features, uses the convolution kernel to scan the input data, and extracts multi-scale spectral spatial feature maps. Each layer of the convolutional network captures features at different scales through different convolution kernel sizes, so that the network can fully understand the changes in spectral data. Channel attention calculation is performed on the multi-scale spectral spatial feature map, and the more important channels in the spectral features are strengthened by calculating the channel weight coefficient, so as to improve the expressiveness and classification ability of the spectral information. The channel weight coefficient is multiplied by the multi-scale spectral spatial feature map channel by channel to obtain the channel enhanced feature map. A 1×1 convolution layer is used to reduce the dimension of the channel enhanced feature map, reduce the number of channels of the feature map to a dimension more suitable for subsequent processing, and obtain the enhanced spectral feature representation. The enhanced spectral feature representation is input into the bidirectional LSTM (long short-term memory) unit for processing. The bidirectional LSTM unit has two LSTM layers, forward and backward, which consider the past and future information of the time series data at the same time and capture richer temporal correlation features. In each LSTM layer, the input gate uses the Sigmoid activation function to process the current input data and the hidden state concatenation of the previous moment to determine the degree of retention of the input information at the current moment. The forget gate also uses the Sigmoid activation function to control the retention ratio of historical information so that the network can flexibly update the memory. The output gate uses the tanh activation function to generate candidate memory units. This design enables each LSTM layer to effectively learn the important features in the time series data. The dimension of the hidden layer is set to 256 to ensure that the network has sufficient expression ability. After completing the bidirectional LSTM processing, the bidirectional temporal correlation features are input into a spatial attention module for further processing. The core function of the spatial attention module is to calculate the spatial attention weight matrix. By reducing the number of channels of the feature map, the number of channels is reduced to 64 using a 1×1 convolution layer. The 3×3 convolution layer is used to extract spatial correlation information and enhance the spatial feature representation of the feature map. The feature map is adjusted to a fixed size through the spatial adaptive average pooling layer to ensure that the network can handle input data of different sizes. After the Softmax activation function, the H×W-dimensional spatial attention weight matrix is ​​obtained, where H and W represent the height and width of the feature map, respectively. The bidirectional temporal correlation features are weighted and summed according to the spatial attention weight matrix to obtain the spectral error feature representation. The weighted summation process weights the features by using the spatial attention weight matrix to emphasize those features that are more critical in spectral error analysis.At the same time, residual connection processing is performed to add the input features to the attention weighted features to form the final spectral error feature representation. This process helps to alleviate information loss and improve the effect of feature transfer. A three-layer fully connected network is used to map the final spectral error feature representation to obtain the drive current compensation value. In this process, the fully connected network establishes a mapping relationship between the spectral error feature representation and the target drive current compensation value, and uses the characteristics of the feedforward neural network to enhance the learning and understanding of the spectral error features. The goal of the mapping process is to provide accurate drive current compensation values ​​so that each LED lamp bead can maintain an ideal spectral output under different working conditions, thereby achieving a full-spectrum luminous effect.

[0019] The drive current compensation value is converted into a PWM (pulse width modulation) signal to adjust the luminous characteristics of each LED lamp bead. In this process, the drive current compensation value is converted into a main PWM duty cycle parameter after the first-level PWM signal conversion, which directly affects the brightness and luminous effect of the LED lamp bead. The main PWM duty cycle parameter is superimposed with a second-level PWM signal to form a fine-tuning PWM control signal. This fine-tuning step allows for more refined adjustment of brightness and color, improving the overall luminous quality. The fine-tuning PWM control signal is gamma-corrected to obtain a uniformly perceived brightness adjustment curve. The role of gamma correction is to correct the nonlinear perception of brightness changes by the human eye, making the brightness changes smoother and more natural. Since the working performance of LEDs is closely related to temperature, the brightness adjustment curve is temperature-compensated according to the relationship model between LED junction temperature and wavelength drift. Because temperature changes can cause the luminous characteristics of LEDs to shift, thereby affecting the overall spectral performance. After temperature compensation processing, the final generated PWM waveform parameters have adaptive characteristics and maintain stable luminous effects under different working environments. The temperature-adaptive PWM waveform parameters are input into the adaptive frequency distributor for processing to generate multiple sets of interleaved PWM signals. Interleaved PWM signals can effectively improve the flexibility of LED driving, allowing different lamp beads to receive signals in different time periods, thereby avoiding current shocks caused by simultaneous switching and reducing electromagnetic interference. Multiple groups of interleaved PWM signals are time-interleaved to obtain a mixed modulation waveform, which is properly distributed in time to ensure the continuous light emission and uniformity of the LED lamp beads. According to the mixed modulation waveform, a synchronous sampling timing is generated to obtain a current detection window sequence, which is used to monitor the working status of each LED lamp bead in real time to ensure that the lamp bead operates under predetermined working conditions. The results of real-time monitoring form fault diagnosis data, which helps to identify any potential faults or abnormal operating conditions. According to the fault diagnosis data, the PWM control parameters are adaptively adjusted to ensure that each LED lamp bead can operate in the best state and maintain the required luminous characteristics. Through the above steps, the frequency band control data of each LED lamp bead is obtained, including the wavelength range and light intensity contribution value, while ensuring that the luminous color of each lamp bead is accurately adjusted according to specific needs to achieve a full-spectrum luminous effect.

[0020] In one example, phosphor powder is mixed with silicone resin to obtain a phosphor glue mixture, wherein the phosphor powder includes red phosphor powder, green phosphor powder and yellow phosphor powder mixed in a preset mass ratio, including: The wavelength of the red phosphor is measured to obtain a first phosphor material with a central wavelength of 620-630 nm; the wavelength of the green phosphor is measured to obtain a second phosphor material with a central wavelength of 520-530 nm; the wavelength of the yellow phosphor is measured to obtain a third phosphor material with a central wavelength of 580-590 nm; The first phosphor material, the second phosphor material, and the third phosphor material are mixed in a mass ratio of 2:3:1 to obtain a mixed phosphor material; The mixed fluorescent powder material and the organic silicon resin are mixed in a preset mass ratio of 3:7 to obtain an initial fluorescent glue mixture, and the initial fluorescent glue mixture is stirred evenly to obtain a fluorescent glue mixture; The viscosity of the fluorescent glue mixture is tested to obtain a viscosity parameter of the fluorescent glue mixture, and the temperature of the fluorescent glue mixture is adjusted according to the viscosity parameter of the fluorescent glue mixture to obtain the fluorescent glue mixture.

[0021] In this example, a suitable spectral measurement device is selected, such as a spectrophotometer or a fluorescence spectrometer. These devices can accurately determine the luminescence characteristics of different phosphors. Taking red phosphor as an example, by placing the phosphor under a light source and stimulating it to emit light, the spectrometer is used to measure it, and the result is that its central wavelength range is 620-630nm. This process is applicable to green and yellow phosphors, and the second and third phosphor materials with central wavelengths of 520-530nm and 580-590nm are obtained respectively. These wavelength data determine the spectral characteristics of the final product. The first phosphor material, the second phosphor material, and the third phosphor material are mixed in a mass ratio of 2:3:1. After mixing, the obtained mixed phosphor material will have certain spectral distribution characteristics. The mixed phosphor material is mixed with silicone resin in a preset mass ratio of 3:7 to obtain an initial fluorescent glue mixture. In this process, ensure that the phosphor and silicone resin are fully and evenly mixed to ensure the stability of the fluorescent effect. Set the mixed mass ratio to: ; in, Indicates the quality of mixed phosphor powder. Indicates the quality of silicone resin. After the initial mixing, the fluorescent glue mixture is stirred evenly, using a mechanical stirrer or ultrasonic stirrer to ensure the uniformity and stability of the mixture. The viscosity of the fluorescent glue mixture is tested by a viscometer to determine the fluidity of the mixture and obtain the viscosity parameter of the fluorescent glue mixture, which is recorded as . Viscosity is usually measured in millipascals per second (mPa·s). Based on the test results, it is analyzed whether there is a problem of too high or too low viscosity. Too high viscosity will make the mixture difficult to apply, while too low viscosity will affect the curing effect, so temperature adjustment is required. According to the basic formula in fluid mechanics, there is an inverse relationship between the viscosity of the fluid and the temperature, which is expressed by the Arrhenius equation: ; in, is a constant, is the activation energy, is the gas constant, is the absolute temperature (K). By adjusting the temperature, the viscosity of the fluorescent glue is effectively changed to ensure its fluidity and adhesion during the application process. After proper temperature adjustment, the final fluorescent glue mixture has ideal viscosity characteristics and is suitable for subsequent coating and curing processes.

[0022] In one example, a fluorescent glue mixture is applied to the surface of a single blue light LED chip and cured to obtain an LED lamp bead unit, including: The wavelength of the single blue LED chip was tested to obtain a chip to be packaged with a central wavelength of 440-460nm, and the power of the chip to be packaged was tested to obtain a qualified chip with a luminous power of 0.2-1W; The qualified chip is provided with a metal pad to obtain a copper pad structure with a surface silver-plated and a thickness of 35 μm, and the dispensing equipment is set with parameters according to the viscosity parameters of the fluorescent glue mixture to obtain the dispensing process parameters; The fluorescent glue mixture is subjected to glue dispensing treatment based on glue dispensing process parameters to obtain a fluorescent glue layer with a thickness of 0.3-0.5 mm, and the fluorescent glue layer is subjected to segmented temperature rising and curing treatment to obtain an initial LED lamp bead unit; The initial LED lamp bead unit is spectrally detected to obtain LED light emitting spectrum data, and the LED light emitting spectrum data is screened and classified to obtain the LED lamp bead unit.

[0023] In this example, a single blue LED chip is tested for wavelength, and a spectrometer is used to measure the central wavelength of the LED light. By placing the chip to be packaged under a light source and performing spectral analysis, the result is that the central wavelength range is 440-460nm. Assuming that the specific central wavelength obtained after the test is 450nm, it means that the LED chip has the best luminous efficiency within this wavelength range. In spectral detection, the light intensity With wavelength The relationship is described by the following formula: ; in, is the incident light intensity, is the attenuation coefficient, is the wavelength. This formula is used to understand the luminous characteristics of LEDs at different wavelengths and help adjust the drive current to achieve the desired spectral characteristics. To ensure the performance of the selected chip, a power test is performed to ensure that the luminous power is between 0.2-1W. Assuming that the test results show that the actual luminous power of the chip is 0.5W, this chip is considered a qualified chip and is suitable for subsequent packaging and applications. The luminous power of the LED is calculated by the following formula: ; in, Indicates power (W), represents the current (A), and Represents voltage (V). This formula is used to verify the power test results of LEDs. Set metal pads for qualified chips. Optimize the electrical connections of qualified chips to ensure good electrical conductivity and thermal conductivity. The surface of the pad needs to be silver-plated to improve its electrical conductivity, and the thickness of the copper pad is controlled to be 35μm. The thickness setting meets the requirements for current passage and effectively reduces thermal resistance, thereby improving the working efficiency of LEDs. Set the parameters of the dispensing equipment according to the viscosity parameters of the fluorescent glue mixture. The viscosity parameters usually affect the accuracy and uniformity of dispensing, so they need to be adjusted accordingly. The set dispensing process parameters should take into account factors such as dispensing pressure, speed and time. For example, assuming a fluorescent glue mixture with a viscosity of 1000mPa·s, the parameters of the dispensing equipment need to be set to a pressure of 2 bar and a dispensing speed of 50 mm / s to ensure that the thickness of the final fluorescent glue layer is controlled between 0.3-0.5mm. After the dispensing is completed, the fluorescent glue layer is subjected to a staged temperature rise curing treatment to ensure that the cross-linking reaction and physical properties of the fluorescent glue are fully exerted. During the curing process, the relationship between the viscosity of the fluorescent glue and the temperature is expressed by the Arrhenius equation: ; in, is the viscosity, is a constant, is the activation energy, is the gas constant, is the absolute temperature (K). By adjusting the temperature and optimizing the viscosity parameters, the dispensing and curing process can be ensured to proceed smoothly. Usually, segmented temperature curing is achieved by setting multiple temperature stages. Assume that the initial temperature is set to 60°C for 20 minutes, then raised to 80°C for 30 minutes, and finally raised to 100°C for 10 minutes. In this way, the fluorescent glue can be fully cured and attached to the surface of the LED chip to form the initial LED lamp bead unit. The initial LED lamp bead unit is spectrally detected to obtain the LED luminous spectrum data. The spectral detection is performed using a spectrophotometer to obtain the luminous characteristic data of the LED under different driving conditions. Assuming that the test results show that the LED lamp bead has the best light intensity distribution in the range of 440-460nm, screening and classification are performed based on these data. The obtained spectral data is compared with the preset target spectrum to determine whether the luminous characteristics of the LED lamp bead unit meet the requirements.

[0024] In one example, a double-sided copper-clad conductive substrate is subjected to photolithography to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit, including: The thickness of the double-sided copper-clad conductive substrate is measured to obtain a substrate material with a thickness of 1.6 mm and a copper foil thickness of 35 μm, and a circuit pattern is designed for the substrate material to obtain a conductive circuit diagram including a power supply circuit, a signal circuit, and an LED mounting pad; Inputting the conductive circuit diagram into a photolithography device for circuit exposure to obtain a conductive circuit development pattern, and etching the substrate material according to the conductive circuit development pattern to obtain an initial light strip substrate; The surface of the initial light strip substrate is coated with white solder resist ink to obtain a solder resist protective layer with a thickness of 15-20 μm, and the solder resist protective layer is processed to open LED mounting holes to obtain mounting through holes with a diameter larger than the diameter of the LED lamp bead base; The back of the initial light strip substrate was patterned with a heat dissipation copper layer to obtain a heat dissipation structure in which the heat dissipation copper layer area accounted for more than 80% of the total substrate area, and the pad spacing of the mounting through holes was measured to obtain a light strip substrate with a center spacing of adjacent pads of 5 mm.

[0025] In this example, the thickness of the double-sided copper-clad conductive substrate is measured. By using a high-precision thickness gauge, the overall thickness of the substrate is 1.6mm, and the thickness of the copper foil is 35μm. Thickness directly affects the conductive performance and heat dissipation effect. The thicker the copper layer of the substrate, the stronger its conductivity and heat dissipation ability. The circuit pattern design of the substrate material is carried out. Using professional circuit design software, a conductive circuit diagram including power supply circuit, signal circuit and LED mounting pad is designed. The entire circuit design must ensure that there is enough spacing between the signal and the power supply to prevent interference and short circuit. After the circuit pattern design is completed, the conductive circuit diagram is input into the photolithography equipment for circuit exposure. Using photolithography technology, the design pattern is transferred to the photosensitive coating of the substrate. After exposure, the development pattern of the conductive circuit is obtained through development. The substrate material is etched according to the development pattern, and the unprotected copper layer is removed using a chemical etchant. The initial light strip substrate obtained finally contains a clear circuit pattern, which can effectively conduct electricity and dissipate heat. The surface of the initial light strip substrate is coated with white solder resist ink. The thickness of the coating should be controlled between 15-20μm to ensure that it has good insulation and protection effects, while not affecting the subsequent installation of LED lamp beads. After the coating is completed, the LED mounting holes are opened. The diameter of these mounting holes needs to be larger than the diameter of the LED lamp bead base to ensure that the lamp beads can be firmly fixed on the substrate. The back of the initial light strip substrate is patterned with a heat dissipation copper layer. When designing the heat dissipation copper layer, its area must account for more than 80% of the total area of ​​the substrate to provide sufficient heat dissipation capacity. When measuring the pad spacing of the mounting through holes, ensure that the center spacing between adjacent pads is 5mm. This spacing design not only facilitates soldering operations, but also ensures that there is enough space between LED lamp beads to prevent damage caused by overheating.

[0026] In one example, an array welding process is performed on the LED lamp bead units on the lamp strip substrate to obtain an LED lamp strip array structure, including: Preheating the LED lamp bead unit and the light strip substrate to obtain a component to be welded with a preheating temperature of 150-180°C, and reflow welding the component to be welded to obtain an initial welding structure with a welding temperature of 235-245°C; Perform optical scanning on the initial welding structure to obtain the position deviation data of the LED lamp beads, which is less than 0.1mm. Perform X-ray inspection on the welding points of the initial welding structure to obtain the welding point quality data. Install interfaces on the front and rear ends of the light strip substrate to obtain an LED light strip assembly with a power interface of a DC5V power supply terminal and a standard SPI communication interface, and perform a DC5V voltage power-on test on the LED light strip assembly to obtain current test data of 20mA / lamp bead; A comprehensive analysis is performed based on the LED lamp bead position deviation data, solder joint quality data and current test data to obtain the LED light strip quality judgment result, and based on the LED light strip quality judgment result, the LED light strip components are classified and screened to obtain the LED light strip array structure.

[0027] In this example, the LED lamp bead unit and the light strip substrate are preheated to ensure good welding quality during the welding process. The temperature of the preheating treatment is set between 150-180℃. The significance of the preheating process is to reduce the thermal stress during the subsequent reflow soldering, thereby improving the reliability of soldering. At the same time, the moisture in the soldered parts is eliminated, the risk of bubbles during soldering is reduced, and the firmness and conductivity of the soldered joints are ensured. The preheating process is carried out by using a constant temperature box or a heating plate. The temperature sensor monitors the preheating state in real time to ensure the uniformity and stability of the temperature. The preheated components to be soldered are reflow soldered. The soldering temperature of reflow soldering is generally set between 235-245℃. This temperature needs to be precisely controlled to ensure that the solder can be fully melted and form a good solder connection between the lamp bead and the substrate. During the soldering process, the solder will first heat up to the melting point under the preset temperature curve, and then maintain it for a period of time so that the solder can fully wet the surface of the pad, and finally cool to form a firm solder joint. Good welding can not only ensure the stability of the electrical connection of the LED lamp bead, but also extend the service life of the entire LED light strip. After welding is completed, the initial welding structure is optically scanned to detect the position deviation of the LED lamp beads. By using a high-resolution camera system, the actual position data of each LED lamp bead on the substrate is obtained, and the deviation from the designed position is calculated to ensure that the deviation value is less than 0.1mm. If the deviation exceeds this range, it will affect the uniformity and effect of the light strip. At the same time, X-ray detection technology is used to analyze the quality of the solder joints. By imaging the internal structure of the solder joints, it is determined whether the solder joints have defects, such as cold soldering, virtual soldering, etc. After confirming that the quality of the welding structure is qualified, the interfaces are installed at the beginning and end of the substrate of the light strip to form a complete LED light strip assembly. The light strip needs to have a DC5V power supply terminal and a standard SPI communication interface to easily connect to the control system. Through these interfaces, the light strip can receive power and control signals. During the installation process, it is necessary to ensure that the interface connection is stable and well insulated to prevent short circuits or other electrical faults. After the interface installation is completed, the LED light strip assembly is powered on with a DC5V voltage test. Use an ammeter to test the current of each lamp bead. Ideally, the working current of the lamp bead should be around 20mA. This current range is the standard value for the normal operation of LED lamp beads. Too high or too low current may cause damage to the lamp beads or insufficient brightness. The current test data can effectively reflect the electrical performance of the entire light strip. After all tests are completed, the LED lamp bead position deviation data, solder joint quality data and current test data are comprehensively analyzed. By analyzing these data, the quality judgment results of the LED light strip are obtained. According to the quality judgment results, classification and screening are carried out to ensure that only qualified LED light strip components enter the subsequent production or shipment links to obtain the LED light strip array structure.

[0028] In one example, spectrum parameters of each LED lamp bead in the LED light strip array structure under multiple sets of driving currents are collected to obtain spectrum characteristic mapping data, where the spectrum parameters include peak wavelength, wavelength bandwidth and relative light intensity, including: The driving current gradient is set for each LED lamp bead in the LED light strip array structure to obtain multiple groups of test current values ​​within the range of 0-60mA; The LED lamp beads are spectrally scanned based on each set of test current values ​​to obtain original spectral data with a wavelength range of 380-780nm and a resolution of 1nm; The peak wavelength of the original spectrum data is extracted to obtain the central emission wavelength value of each LED lamp bead under different driving currents; The original spectrum data is calculated by half-width calculation to obtain the wavelength bandwidth value of each LED lamp bead under different driving currents, and the original spectrum data is calculated by energy integration to obtain the relative light intensity value of each LED lamp bead under different driving currents; The central luminous wavelength value, wavelength bandwidth value and relative light intensity value are data-correlated to obtain the luminous characteristic curve of each LED lamp bead; Mathematical modeling is performed on the luminescence characteristic curve to obtain a mapping relationship model between the driving current and the spectral parameters, and the mapping relationship model is stored in a database to obtain spectral characteristic mapping data.

[0029] In this example, a driving current gradient is set for each LED lamp bead in the LED light strip array structure. By accurately adjusting the driving current, multiple groups of test current values ​​in the range of 0-60mA are obtained. For example, the current is divided into several gradients of 0, 10, 20, 30, 40, 50, and 60mA, which is convenient for evaluating the luminous characteristics of LED lamp beads under different currents. Each current gradient will affect the spectral output of the LED lamp bead. Based on the set test current value, a spectral scan is performed on each LED lamp bead to obtain raw spectral data with a wavelength range of 380-780nm and a resolution of 1nm. The spectral scan is performed using a spectrometer, and the relative light intensity of each wavelength in the entire spectral range is obtained by scanning the light emission of each LED lamp bead. Assume that the data obtained by the scan shows an intensity peak at certain wavelengths, for example, the light intensity reaches a maximum at 450nm, and the peak indicates that the LED has the best luminous effect at this wavelength. The peak wavelength is extracted from the raw spectral data to determine the central luminous wavelength of each LED lamp bead under different driving currents. By analyzing the spectral curve, the position of the maximum intensity is found, and the wavelength is defined as the peak wavelength. Set the light intensity function to ,in is the wavelength, then the peak wavelength is obtained by finding the maximum value of the light intensity: ; in, is the peak wavelength, which indicates the central emission wavelength of the LED under a specific driving current. At the same time, the half-height width is calculated for the original spectrum data. The half-height width indicates the width of the spectrum curve at half the peak intensity. The calculation of the half-height width helps to describe the color purity and width of the spectrum. Assume that the peak spectral intensity is , then the half-width is obtained by solving the spectral intensity: The position of the two wavelengths is obtained as: ; in, and are the two wavelengths corresponding to the half-peak light intensity, That is the half-height width. By calculating the half-height width, the spectral bandwidth characteristics of the LED lamp beads under different driving currents are evaluated. The energy integration operation is performed on the original spectral data to obtain the relative light intensity value of each LED lamp bead under different driving currents. The relative light intensity is achieved by integrating the spectral curve, and the spectral intensity within the wavelength range is accumulated and summed. The calculation formula is as follows: ; in, is the relative light intensity, and is the starting and ending wavelength of the spectral range, and the total light intensity of the LED lamp bead in the entire spectral range is obtained by integration. After completing the above data processing, the central emission wavelength value, wavelength bandwidth value and relative light intensity value of each LED lamp bead are data-associated to obtain the luminous characteristic curve of each LED lamp bead. By drawing the luminous characteristic curve, the luminous behavior of the LED lamp bead under different driving currents is displayed, including information such as the movement of the spectral peak and the change of light intensity. The luminous characteristic curve is mathematically modeled to obtain a mapping relationship model between the driving current and the spectral parameters. For example, setting the driving current to , spectral parameters include peak wavelength , FWHM , relative light intensity . Use a function model to describe the relationship between them: ; in, The mapping relationship model describes the dependence between the driving current and the spectral characteristics. By fitting these functions, a mathematical model that can accurately predict the spectral behavior of LED lamp beads is obtained. The mapping relationship model is stored in the database to form spectral characteristic mapping data for subsequent production and quality control. By referring to the model in the database, the production parameters can be quickly adjusted to achieve precise control of the luminous characteristics of LED lamp beads.

[0030] In one example, the full spectrum target curve is wavelength segmented according to the spectral characteristic mapping data to obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence. The frequency band control data includes the wavelength range and the light intensity contribution value, including: The full spectrum target curve is divided into wavelength intervals to obtain multiple wavelength sub-intervals within the range of 380-780nm; Calculate the light intensity of each wavelength sub-interval to obtain the target spectral intensity distribution data; According to the spectral characteristic mapping data, LED lamp beads are allocated to the wavelength sub-intervals to obtain the LED lamp bead group corresponding to each wavelength sub-interval; Perform superposition calculation on the luminous spectrum of each LED lamp bead group to obtain the combined spectrum data of the LED lamp bead group; Based on the difference calculation between the combined spectrum data and the target spectrum intensity distribution data, spectrum error data is obtained, and the drive current of the LED lamp bead group is compensated according to the spectrum error data to obtain the drive current compensation value; The driving current compensation value is converted into a PWM signal to obtain a PWM control parameter, and the PWM control parameter is input into the LED control unit to adjust the driving current, obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full-spectrum luminescence. The frequency band control data includes the wavelength range and the light intensity contribution value.

[0031] In this example, the full spectrum target curve is divided into wavelength intervals, and the wavelengths in the range of 380-780nm are divided into multiple sub-intervals at equal intervals, such as 10nm per sub-interval, to obtain multiple wavelength sub-intervals, such as 380-390nm, 390-400nm, 400-410nm, etc. The light intensity is calculated for each wavelength sub-interval to obtain the target spectral intensity distribution data. Assume that the target light intensity of a wavelength sub-interval is , the total light intensity in this sub-interval is calculated by integration, the formula is as follows: ; in, and The starting and ending wavelengths of the sub-intervals are obtained. The target spectral intensity distribution data of each wavelength sub-interval is obtained to form a complete spectral intensity distribution list. LED lamp beads are allocated to the wavelength sub-intervals according to the spectral characteristic mapping data. Refer to the luminous characteristics of each LED lamp bead at a specific wavelength, such as its peak wavelength and luminous intensity. In the mapping data, the characteristics of the LED lamp bead are set as functions , which can be expressed as: ; Through such allocation, it is ensured that the LED lamp bead group corresponding to each wavelength sub-interval can cover the required wavelength range and ensure that the color of the light is consistent with the target spectrum. The light spectrum of each LED lamp bead group is superimposed to obtain the combined spectrum data of the LED lamp bead group. For a group composed of multiple LED lamp beads, its combined spectrum is expressed as the sum of the spectra of each LED lamp bead: ; in, For the LED lamp beads in wavelength The light intensity at is the number of LED lamp beads in the group. Through calculation, the comprehensive luminous characteristics of the LED lamp bead group at different wavelengths are obtained. The difference calculation is performed based on the combined spectrum data and the target spectrum intensity distribution data to obtain the spectrum error data. The spectrum error is expressed by the following formula: ; in, Indicated in wavelength By analyzing the spectral error data and according to the actual spectral performance of each LED lamp bead group, the driving current compensation calculation is performed to obtain the driving current compensation value. The compensation calculation is expressed by the following formula: ; in, is the current value that needs to be compensated, is the compensation coefficient, which reflects the sensitivity of the spectrum error to the current adjustment. The drive current compensation value is converted into a PWM (pulse width modulation) signal to obtain the PWM control parameter. The generation of the PWM signal is expressed by the following formula: ; in, is the duty cycle, The maximum driving current of the LED lamp bead is generated in this way according to the adjusted current value. After the PWM control parameters are input into the LED control unit, the driving current of each LED lamp bead is precisely adjusted to adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence.

[0032] In one example, a difference calculation is performed based on the combined spectrum data and the target spectrum intensity distribution data to obtain spectrum error data, and a compensation calculation is performed on the driving current of the LED lamp bead group according to the spectrum error data to obtain a driving current compensation value, including: Perform sliding window sampling on the combined spectral data and the target spectral intensity distribution data to obtain a time series feature matrix; The time series feature matrix is ​​input into a three-layer convolutional feature extraction network for processing to obtain a multi-scale spectral spatial feature map, and channel attention calculation is performed on the multi-scale spectral spatial feature map to obtain the channel weight coefficient; The channel weight coefficient is multiplied by the multi-scale spectral spatial feature map channel by channel to obtain the channel enhanced feature map, and the channel enhanced feature map is subjected to dimensionality reduction processing through a 1×1 convolution layer to obtain an enhanced spectral feature representation; The enhanced spectral feature representation is processed by a bidirectional LSTM unit to obtain a bidirectional temporal correlation feature. The bidirectional LSTM unit contains two LSTM layers, the forward and backward LSTM layers. Each layer includes: the input gate uses the Sigmoid activation function to process the concatenation of the input data and the hidden state of the previous moment, the forget gate uses the Sigmoid activation function to control the proportion of historical information retention, and the output gate uses the tanh activation function to generate candidate memory units. The hidden layer dimension of each LSTM layer is set to 256; The bidirectional temporal correlation features are input into the spatial attention module for processing to obtain the spatial attention weight matrix. The spatial attention module includes: reducing the number of feature map channels to 64 through a 1×1 convolution layer, extracting spatial correlation through a 3×3 convolution layer, obtaining a fixed-size feature map through a spatial adaptive average pooling layer, and finally obtaining an H×W-dimensional spatial attention weight matrix through a Softmax activation function, where H and W are the height and width of the feature map respectively; The bidirectional temporal correlation features are weighted and summed according to the spatial attention weight matrix to obtain the spectral error feature representation, and the spectral error feature representation is subjected to residual connection processing, and the input features are added to the attention weighted features to obtain the final spectral error feature representation; The final spectral error feature representation is mapped and transformed through a three-layer fully connected network to obtain the driving current compensation value.

[0033] In this example, the data is sampled by sliding window to obtain the time series feature matrix. By setting a window size , calculate the spectral intensity data within the window at each time point and combine them into a feature matrix. For example, if the window size is set to 5, the information of 5 adjacent data points is extracted from the data to form a feature vector: ; in, For the The spectral intensity values ​​at each time point are obtained to obtain a time series feature matrix. The time series feature matrix is ​​input into a three-layer convolution feature extraction network for processing. The network structure is designed to extract multi-scale spectral spatial feature maps. In the convolution layer, convolution kernels of different sizes are used to extract features of different scales. The feature map obtained by the convolution operation is expressed by the following formula: ; in, is the convolution kernel, and are the height and width of the convolution kernel, respectively. and is the row and column index of the feature map. This process can effectively capture the local features of the spectral data. Channel attention calculation is performed on the multi-scale spectral spatial feature map to obtain the channel weight coefficient. The features of each channel are obtained by performing global average pooling on each channel of the feature map, and then the channel weight coefficient is generated through a fully connected layer and an activation function (such as Sigmoid), which is expressed as: ; in, is the channel weight coefficient, is the Sigmoid activation function, FC is the fully connected layer, is a multi-scale feature map. The channel weight coefficient is multiplied by the multi-scale spectral spatial feature map channel by channel to obtain a channel enhancement feature map, which increases the influence of important channels in the feature map, expressed as: ; in, represents channel-by-channel multiplication. The channel-enhanced feature map is obtained by a The convolution layer performs dimensionality reduction to obtain an enhanced spectral feature representation. The function of this layer is to reduce the number of channels in the feature map while maintaining the validity of the features. The enhanced spectral feature representation is input into the bidirectional LSTM unit for processing. The bidirectional LSTM can simultaneously consider the forward and backward information of the input sequence and capture the long-term dependencies of the time series. The calculations of the input gate, forget gate, and output gate of the bidirectional LSTM are: ; in, , , They are input gate, forget gate and output gate respectively. is the input at the current moment, is the hidden state at the current moment, is the cell state at the current moment, is the weight matrix. This process enables the features of each time step to combine historical and future information. After the bidirectional LSTM processing is completed, the bidirectional temporal correlation features are obtained. Next, these features are input into the spatial attention module for processing to obtain the spatial attention weight matrix. In this module, the number of feature map channels is reduced to 64 through a 1×1 convolution layer, and then a 3×3 convolution layer is used to extract spatial correlation. After a spatial adaptive average pooling layer, a fixed-size feature map is obtained. Finally, the H×W-dimensional spatial attention weight matrix is ​​obtained through the Softmax activation function, which is expressed as: ; in, is the spatial attention weight. This weight matrix is ​​used to perform weighted summation on the bidirectional temporal correlation features to obtain the spectral error feature representation, which is expressed as: ; The spectral error feature representation is processed by residual connection, and the input feature is added to the attention weighted feature to obtain the final spectral error feature representation. The final spectral error feature representation is mapped and transformed through a three-layer fully connected network to obtain the driving current compensation value. The output of the three-layer fully connected network is expressed as: ; in, is the final drive current compensation value.

[0034] In one example, the driving current compensation value is converted into a PWM signal to obtain a PWM control parameter, and the PWM control parameter is input into the LED control unit to adjust the driving current, obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence, and the frequency band control data includes a wavelength range and a light intensity contribution value, including: Performing a first-level PWM signal conversion on the driving current compensation value to obtain a main PWM duty cycle parameter, and performing a second-level PWM signal superposition on the main PWM duty cycle parameter to obtain a fine-tuning PWM control signal; Gamma correction is performed on the fine-tuning PWM control signal to obtain a uniformly perceived brightness adjustment curve, and temperature compensation is performed on the brightness adjustment curve according to the relationship model between LED junction temperature and wavelength drift to obtain temperature-adaptive PWM waveform parameters; Inputting the temperature-adaptive PWM waveform parameters into an adaptive frequency distributor for processing to obtain multiple groups of interleaved PWM signals, and performing time interleaving processing on the multiple groups of interleaved PWM signals to obtain a mixed modulation waveform; A synchronous sampling timing is generated according to the mixed modulation waveform to obtain a current detection window sequence, and the real-time working status of each LED lamp bead is monitored in parallel to obtain fault diagnosis data. The PWM control parameters are adaptively adjusted according to the fault diagnosis data to obtain the frequency band control data of each LED lamp bead, and the luminous color of each LED lamp bead is adjusted to achieve full-spectrum luminescence. The frequency band control data includes wavelength range and light intensity contribution value.

[0035] In this example, the first-level PWM signal conversion is performed on the drive current compensation value to obtain the main PWM duty cycle parameter. PWM (Pulse Width Modulation) technology adjusts the brightness of LED lamp beads by controlling the duty cycle of the signal. The main PWM duty cycle is based on the drive current compensation value. Calculation, the formula is: ; in, is the main PWM duty cycle, is the maximum driving current of the LED lamp bead. Convert the compensation value to the corresponding PWM duty cycle. Superimpose the second-level PWM signal on the main PWM duty cycle parameter to obtain the fine-tuning PWM control signal. The fine-tuning process is achieved by introducing a small adjustment value. accomplish: ; in, It is the fine-tuned PWM control signal. Fine-tuning can help control the brightness more precisely in practical applications and adapt to different environmental requirements. The fine-tuned PWM control signal is gamma-corrected to obtain a uniform brightness adjustment curve. Gamma correction is used to adjust the nonlinear relationship between the brightness output and the input signal. Assuming the gamma correction index is γ, the corrected PWM signal is expressed as: ; During gamma correction, the brightness adjustment curve is temperature compensated according to the relationship model between LED junction temperature and wavelength drift to ensure output stability under different temperature conditions. And make corresponding adjustments, assuming that the temperature compensation coefficient is , the final temperature adaptive PWM waveform parameters are expressed as: ; in, The temperature is used as the reference temperature to ensure that the PWM waveform parameters can still maintain good light output characteristics when the temperature changes. The temperature-adaptive PWM waveform parameters are input into the adaptive frequency distributor for processing to generate multiple sets of interleaved PWM signals. The adaptive frequency distributor automatically adjusts the signal frequency according to the current load conditions and system response to optimize the driving efficiency of the LED. The generated interleaved PWM signal is used to improve the stability of the driving signal and reduce the impact of current fluctuations on the LED. The interleaved PWM signal is time-interleaved to form a hybrid modulation waveform. The hybrid modulation waveform optimizes the uniformity and response speed of the light output by adjusting the timing of the interleaved PWM signal. Different PWM signals are sent alternately within a certain time window to improve the overall light quality and stability. After the generation of the hybrid modulation waveform is completed, a synchronous sampling timing is generated according to the waveform to obtain a current detection window sequence. This process involves sampling the PWM signal to obtain the real-time working status of each LED lamp bead. In this process, the setting of the current detection window is very critical. By setting the time Determine the duration of each test. The current data obtained by monitoring is expressed by the following formula: ; By monitoring the current state of each LED lamp bead in parallel, fault diagnosis data is obtained to determine whether the working state of the LED lamp bead is normal. According to the obtained fault diagnosis data, the PWM control parameters are adaptively adjusted to ensure that the frequency band control data of each LED lamp bead reaches the optimal state. Adaptive adjustment uses a control algorithm, such as PID control, to continuously optimize the parameters of the PWM signal through a feedback mechanism to achieve precise control of the luminous color and brightness. The frequency band control data obtained for each LED lamp bead includes the wavelength range and light intensity contribution value, which can adjust the luminous color of each LED lamp bead according to actual needs to achieve full-spectrum luminescence. The frequency band control data in this process analyzes the spectral output of each lamp bead and combines the previous debugging results to ensure that the final light emitted meets the requirements of the full spectrum. For example, for a specific application scenario, the light intensity output of a certain band is increased, and then the corresponding PWM signal is adjusted to achieve the desired effect of its luminous characteristics.

[0036] Reference Figure 2 This embodiment provides a multi-band control device for a full-spectrum LED light strip, including: A mixing module 1 is used to mix phosphors with silicone resin to obtain a phosphor glue mixture, wherein the phosphors include red phosphors, green phosphors and yellow phosphors mixed in a preset mass ratio; Curing module 2, used for applying the fluorescent glue mixture to the surface of the single blue light LED chip for curing to obtain an LED lamp bead unit; The photolithography module 3 is used to perform photolithography on the double-sided copper-clad conductive substrate to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit; The welding module 4 is used to perform array welding processing on the LED lamp bead units on the light strip substrate to obtain an LED light strip array structure; The acquisition module 5 is used to acquire spectral parameters of each LED lamp bead in the LED light strip array structure under multiple sets of driving currents to obtain spectral characteristic mapping data, wherein the spectral parameters include peak wavelength, wavelength bandwidth and relative light intensity; Processing module 6 is used to perform wavelength segmentation processing on the full spectrum target curve according to the spectral characteristic mapping data, obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence. The frequency band control data includes the wavelength range and the light intensity contribution value.

[0037] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0038] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0039] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0040] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-band control method for a full-spectrum LED light strip, characterized in that: The following steps are involved: Mixing fluorescent powder with silicone resin to obtain a fluorescent glue mixture, wherein the fluorescent powder includes red fluorescent powder, green fluorescent powder and yellow fluorescent powder mixed in a preset mass ratio; Applying the fluorescent glue mixture on the surface of a single blue light LED chip and performing a curing process to obtain an LED lamp bead unit; Performing photolithography on the double-sided copper-clad conductive substrate to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit; Performing array welding processing on the LED lamp bead units on the light strip substrate to obtain an LED light strip array structure; Collecting spectral parameters of each LED lamp bead in the LED light strip array structure under multiple sets of driving currents to obtain spectral characteristic mapping data, wherein the spectral parameters include peak wavelength, wavelength bandwidth and relative light intensity; The full spectrum target curve is wavelength segmented according to the spectral characteristic mapping data to obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence, wherein the frequency band control data includes the wavelength range and the light intensity contribution value.

2. The multi-band control method of the full-spectrum LED light strip according to claim 1, characterized in that: The fluorescent powder is mixed with the silicone resin to obtain a fluorescent glue mixture, wherein the fluorescent powder includes red fluorescent powder, green fluorescent powder and yellow fluorescent powder mixed in a preset mass ratio, including: The wavelength of the red phosphor is measured to obtain a first phosphor material with a central wavelength of 620-630 nm; the wavelength of the green phosphor is measured to obtain a second phosphor material with a central wavelength of 520-530 nm; the wavelength of the yellow phosphor is measured to obtain a third phosphor material with a central wavelength of 580-590 nm; The first phosphor material, the second phosphor material, and the third phosphor material are mixed in a mass ratio of 2:3:1 to obtain a mixed phosphor material; The mixed fluorescent powder material and the organic silicon resin are mixed in a preset mass ratio of 3:7 to obtain an initial fluorescent glue mixture, and the initial fluorescent glue mixture is stirred evenly to obtain the fluorescent glue mixture; The viscosity of the fluorescent glue mixture is tested to obtain a viscosity parameter of the fluorescent glue mixture, and the temperature of the fluorescent glue mixture is adjusted according to the viscosity parameter of the fluorescent glue mixture to obtain a fluorescent glue mixture.

3. The multi-band control method of the full-spectrum LED light strip according to claim 2, characterized in that: The method of applying the fluorescent glue mixture to the surface of a single blue light LED chip and performing a curing process to obtain an LED lamp bead unit comprises: Performing a wavelength test on the single blue light LED chip to obtain a chip to be packaged with a central wavelength of 440-460 nm, and performing a power test on the chip to be packaged to obtain a qualified chip with a luminous power of 0.2-1 W; Setting a metal pad on the qualified chip to obtain a copper pad structure with a surface silver-plated and a thickness of 35 μm, and setting parameters of a dispensing device according to the viscosity parameters of the fluorescent glue mixture to obtain dispensing process parameters; Performing a dispensing process on the fluorescent glue mixture based on the dispensing process parameters to obtain a fluorescent glue layer with a thickness of 0.3-0.5 mm, and performing a segmented heating and curing process on the fluorescent glue layer to obtain an initial LED lamp bead unit; The initial LED lamp bead unit is subjected to spectrum detection to obtain LED light emitting spectrum data, and screening and classification are performed according to the LED light emitting spectrum data to obtain the LED lamp bead unit.

4. The multi-band control method of the full-spectrum LED light strip according to claim 3, characterized in that: The double-sided copper-clad conductive substrate is subjected to photolithography processing to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit, including: The thickness of the double-sided copper-clad conductive substrate is measured to obtain a substrate material with a thickness of 1.6 mm and a copper foil thickness of 35 μm, and a circuit pattern is designed on the substrate material to obtain a conductive circuit diagram including a power supply circuit, a signal circuit, and an LED mounting pad; Inputting the conductive circuit diagram into a photolithography device for circuit exposure to obtain a conductive circuit development pattern, and etching the substrate material according to the conductive circuit development pattern to obtain an initial light strip substrate; The surface of the initial light strip substrate is coated with white solder resist ink to obtain a solder resist protective layer with a thickness of 15-20 μm, and the solder resist protective layer is processed to open LED mounting holes to obtain mounting through holes with a diameter greater than the diameter of the LED lamp bead base; The back side of the initial light strip substrate is patterned with a heat dissipation copper layer to obtain a heat dissipation structure in which the heat dissipation copper layer area accounts for more than 80% of the total substrate area, and the pad spacing of the mounting through holes is measured to obtain the light strip substrate with a center spacing of adjacent pads of 5 mm.

5. The multi-band control method of the full-spectrum LED light strip according to claim 4, characterized in that: The LED lamp bead units are subjected to array welding processing on the lamp strip substrate to obtain an LED lamp strip array structure, including: Preheating the LED lamp bead unit and the light strip substrate to obtain a component to be welded with a preheating temperature of 150-180° C., and reflow welding the component to be welded to obtain an initial welding structure with a welding temperature of 235-245° C.; Optically scanning the initial welding structure to obtain LED lamp bead position deviation data, wherein the LED lamp bead position deviation data is less than 0.1 mm, and performing X-ray inspection on the welding points of the initial welding structure to obtain welding point quality data; Install interfaces on the front and rear ends of the light strip substrate to obtain an LED light strip assembly with a power interface of a DC5V power supply terminal and a standard SPI communication interface, and perform a DC5V voltage power-on test on the LED light strip assembly to obtain current test data of 20mA / lamp bead; A comprehensive analysis is performed based on the LED lamp bead position deviation data, the solder joint quality data and the current test data to obtain a quality determination result of the LED light strip, and based on the quality determination result of the LED light strip, the LED light strip components are classified and screened to obtain an LED light strip array structure.

6. The multi-band control method of the full-spectrum LED light strip according to claim 5, characterized in that: The spectral parameters of each LED lamp bead in the LED light strip array structure are collected under multiple groups of driving currents to obtain spectral characteristic mapping data, wherein the spectral parameters include peak wavelength, wavelength bandwidth and relative light intensity, including: A driving current gradient is set for each LED lamp bead in the LED light strip array structure to obtain multiple groups of test current values ​​within a range of 0-60mA; Based on each group of test current values, the LED lamp beads are spectrally scanned to obtain original spectral data with a wavelength range of 380-780nm and a resolution of 1nm; Extract the peak wavelength of the original spectrum data to obtain the central emission wavelength value of each LED lamp bead under different driving currents; Performing half-width calculation on the original spectral data to obtain the wavelength bandwidth value of each LED lamp bead under different driving currents, and performing energy integration operation on the original spectral data to obtain the relative light intensity value of each LED lamp bead under different driving currents; Performing data association on the central luminous wavelength value, the wavelength bandwidth value and the relative light intensity value to obtain a luminous characteristic curve of each LED lamp bead; Mathematical modeling is performed on the luminous characteristic curve to obtain a mapping relationship model between the driving current and the spectral parameter, and the mapping relationship model is stored in a database to obtain spectral characteristic mapping data.

7. The multi-band control method of the full-spectrum LED light strip according to claim 6, characterized in that: The full spectrum target curve is subjected to wavelength segmentation processing according to the spectral characteristic mapping data to obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence, wherein the frequency band control data includes a wavelength range and a light intensity contribution value, including: Dividing the full spectrum target curve into wavelength intervals to obtain a plurality of wavelength sub-intervals within the range of 380-780 nm; Calculating the light intensity of each wavelength sub-interval to obtain target spectral intensity distribution data; Allocate LED lamp beads to the wavelength sub-intervals according to the spectral characteristic mapping data to obtain an LED lamp bead group corresponding to each wavelength sub-interval; Performing superposition calculation on the luminous spectrum of each LED lamp bead group to obtain combined spectrum data of the LED lamp bead group; Performing a difference calculation based on the combined spectrum data and the target spectrum intensity distribution data to obtain spectrum error data, and performing compensation calculation on the driving current of the LED lamp bead group according to the spectrum error data to obtain a driving current compensation value; The driving current compensation value is converted into a PWM signal to obtain a PWM control parameter, and the PWM control parameter is input into the LED control unit to adjust the driving current, so as to obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full-spectrum luminescence, wherein the frequency band control data includes a wavelength range and a light intensity contribution value.

8. The multi-band control method of the full-spectrum LED light strip according to claim 7, characterized in that: The difference calculation is performed based on the combined spectrum data and the target spectrum intensity distribution data to obtain spectrum error data, and compensation calculation is performed on the driving current of the LED lamp bead group according to the spectrum error data to obtain the driving current compensation value, including: Perform sliding window sampling on the combined spectral data and the target spectral intensity distribution data to obtain a time series feature matrix; Inputting the time series feature matrix into a three-layer convolutional feature extraction network for processing to obtain a multi-scale spectral spatial feature map, and performing channel attention calculation on the multi-scale spectral spatial feature map to obtain a channel weight coefficient; Performing a channel-by-channel product operation on the channel weight coefficient and the multi-scale spectral spatial feature map to obtain a channel enhanced feature map, and performing dimensionality reduction processing on the channel enhanced feature map through a 1×1 convolution layer to obtain an enhanced spectral feature representation; The enhanced spectral feature representation is processed by a bidirectional LSTM unit to obtain a bidirectional temporal correlation feature, wherein the bidirectional LSTM unit includes two LSTM layers, a forward LSTM layer and a backward LSTM layer, each layer includes: an input gate uses a Sigmoid activation function to process the concatenation of input data and the hidden state of the previous moment, a forget gate uses a Sigmoid activation function to control the proportion of historical information retention, and an output gate uses a tanh activation function to generate candidate memory units, and the hidden layer dimension of each LSTM layer is set to 256; The bidirectional temporal correlation feature is input into the spatial attention module for processing to obtain a spatial attention weight matrix, wherein the spatial attention module includes: reducing the number of feature map channels to 64 through a 1×1 convolution layer, extracting spatial correlation through a 3×3 convolution layer, obtaining a fixed-size feature map through a spatial adaptive average pooling layer, and finally obtaining an H×W-dimensional spatial attention weight matrix through a Softmax activation function, where H and W are the height and width of the feature map, respectively; Performing a weighted sum operation on the bidirectional temporal correlation features according to the spatial attention weight matrix to obtain a spectral error feature representation, and performing a residual connection process on the spectral error feature representation to add the input features to the attention weighted features to obtain a final spectral error feature representation; The final spectral error feature representation is mapped and transformed through a three-layer fully connected network to obtain a driving current compensation value.

9. The multi-band control method of the full-spectrum LED light strip according to claim 7, characterized in that: The driving current compensation value is converted into a PWM signal to obtain a PWM control parameter, and the PWM control parameter is input into the LED control unit to adjust the driving current, so as to obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full spectrum luminescence, wherein the frequency band control data includes a wavelength range and a light intensity contribution value, including: Performing a first-level PWM signal conversion on the driving current compensation value to obtain a main PWM duty cycle parameter, and performing a second-level PWM signal superposition on the main PWM duty cycle parameter to obtain a fine-tuning PWM control signal; Performing gamma correction on the fine-tuning PWM control signal to obtain a uniformly perceived brightness adjustment curve, and performing temperature compensation on the brightness adjustment curve according to a relationship model between LED junction temperature and wavelength drift to obtain temperature-adaptive PWM waveform parameters; Inputting the temperature-adaptive PWM waveform parameters into an adaptive frequency distributor for processing to obtain multiple groups of interleaved PWM signals, and performing time interleaving processing on the multiple groups of interleaved PWM signals to obtain a mixed modulation waveform; A synchronous sampling timing is generated according to the mixed modulation waveform to obtain a current detection window sequence, and the real-time working status of each LED lamp bead is monitored in parallel to obtain fault diagnosis data. The PWM control parameters are adaptively adjusted according to the fault diagnosis data to obtain the frequency band control data of each LED lamp bead, and the luminous color of each LED lamp bead is adjusted to achieve full-spectrum luminescence, wherein the frequency band control data includes a wavelength range and a light intensity contribution value.

10. A multi-band control device for a full-spectrum LED light strip, characterized in that: For implementing the steps of the method according to any one of claims 1 to 9, the device comprises: A mixing module, used for mixing phosphor with silicone resin to obtain a phosphor glue mixture, wherein the phosphor includes red phosphor, green phosphor and yellow phosphor mixed in a preset mass ratio; A curing module, used for applying the fluorescent glue mixture to the surface of a single blue light LED chip for curing to obtain an LED lamp bead unit; A photolithography module, used for performing photolithography on a double-sided copper-clad conductive substrate to obtain a light strip substrate having an LED mounting pad array, wherein the LED mounting pad array includes a power supply circuit and a signal circuit; A welding module, used for performing array welding processing on the LED lamp bead units on the light strip substrate to obtain an LED light strip array structure; An acquisition module is used to acquire spectral parameters of each LED lamp bead in the LED light strip array structure under multiple groups of driving currents to obtain spectral characteristic mapping data, wherein the spectral parameters include peak wavelength, wavelength bandwidth and relative light intensity; A processing module is used to perform wavelength segmentation processing on the full-spectrum target curve according to the spectral characteristic mapping data, obtain the frequency band control data of each LED lamp bead and adjust the luminous color of each LED lamp bead to achieve full-spectrum luminescence, wherein the frequency band control data includes a wavelength range and a light intensity contribution value.

Images