LED lamp bead luminous efficiency improvement process control method and device

Abstract

The invention relates to the technical field of LED lamp bead detection, and discloses a control method and device for an LED lamp bead luminous efficiency improvement process. The method comprises the following steps: monitoring a current change rate and a temperature change rate of an LED chip, triggering a transient control switching signal, and starting a spectrum analyzer to monitor red shift amount and intensity change of a fluorescent powder emission spectrum so as to obtain transient performance attenuation data; inputting the transient performance attenuation data into a thermal quenching kinetic model for quantum efficiency prediction to obtain quantum efficiency prediction data of the fluorescent powder layer; performing three-dimensional gradient distribution analysis on the surface fluorescent powder concentration of the LED chip to obtain a spatial concentration distribution scheme; and the deviation between the current coating thickness and the target thickness is measured, and coating parameters are adjusted. According to the invention, the dynamic optimization control of the LED fluorescent powder layer under the transient working condition is realized, the technical problem that the traditional static process cannot adapt to the transient working mode is solved, and the high-performance output of the LED lamp bead under the transient working condition is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of LED lamp bead detection, and particularly relates to an LED lamp bead light efficiency improvement process control method and device. BACKGROUND

[0002] At present, the performance of LED lamp beads under transient working modes such as start-up impact, temperature mutation and pulse driving is very unstable. Under these transient working conditions, the current and temperature of the LED chip will change dramatically, causing the phosphor layer to bear non-steady-state thermal stress and light excitation intensity, resulting in serious problems such as rapid decay of phosphor light-emitting efficiency and decline of color temperature stability. The existing LED manufacturing process is mainly designed and optimized for steady-state working conditions, and lacks in-depth understanding of the performance change rule of the phosphor layer under transient working modes and effective control means. In addition, the existing technology lacks effective means to monitor the change of the excited state density of the phosphor in real time, and also lacks a control method for dynamic process adjustment based on transient performance decay data, making it difficult to maintain high light efficiency output of the LED lamp bead under transient working conditions. SUMMARY

[0003] The main purpose of the present application is to provide an LED lamp bead light efficiency improvement process control method and device, which realizes dynamic optimization control of the LED phosphor layer under transient working conditions, solves the technical problem that the traditional static process cannot adapt to the transient working mode, and realizes high-performance output of the LED lamp bead under transient working conditions.

[0004] To achieve the above purpose, the present application provides an LED lamp bead light efficiency improvement process control method, comprising the following steps: monitoring the current change rate and temperature change rate of the LED chip and triggering a transient control switching signal; starting a spectrum analyzer to monitor the red shift amount and intensity change of the phosphor emission spectrum according to the transient control switching signal, and obtaining transient performance decay data; inputting the transient performance decay data into a thermal quenching kinetics model for quantum efficiency prediction, and obtaining quantum efficiency prediction data of the phosphor layer; performing three-dimensional gradient distribution analysis on the surface phosphor concentration of the LED chip based on the quantum efficiency prediction data of the phosphor layer, and obtaining a spatial concentration distribution scheme; measuring the deviation of the current coating thickness from the target thickness according to the spatial concentration distribution scheme, and adjusting the coating parameters through a PID control algorithm.

[0005] Optionally, in the first implementation manner of the first aspect of the present application, the monitoring of the current change rate and temperature change rate of the LED lamp bead and the triggering of the transient control switching signal comprises: The current time series data of the LED chip is collected, and the temperature time series data of the LED chip is collected by an infrared thermal imager; The current time series data is subjected to time difference operation to calculate a current change rate, and the temperature time series data is subjected to time difference operation to calculate a temperature change rate; The absolute values of the current change rate data and the temperature change rate data are multiplied, and then multiplied by a time window weight function to perform transient state discrimination to obtain a transient state operation mode discrimination value; The transient state operation mode discrimination value is compared with a preset threshold value, and a transient state control switching signal is generated when the transient state operation mode discrimination value is greater than the preset threshold value.

[0006] Optionally, in the second implementation manner of the first aspect of the present application, the redshift amount and intensity change of the phosphor emission spectrum are monitored by starting the spectrum analyzer according to the transient state control switching signal to obtain transient state performance decay data, which comprises: The spectrum analyzer is started to collect the phosphor emission spectrum based on the transient state control switching signal to obtain real-time spectral wavelength data and spectral intensity data; The emission peak of the real-time spectral wavelength data is identified to obtain emission spectrum redshift amount data; The decay degree of the phosphor luminous intensity under the transient state operation mode is calculated based on the spectral intensity data to obtain luminous intensity decay ratio data; The emission spectrum redshift amount data and the luminous intensity decay ratio data are weighted and summed to calculate the transient state performance decay data of the phosphor layer under the transient state operation mode.

[0007] Optionally, in the third implementation manner of the first aspect of the present application, the quantum efficiency prediction data of the phosphor layer is obtained by inputting the transient state performance decay data into the thermal quenching kinetics model for quantum efficiency prediction, which comprises: The redshift amount change curve and the intensity decay curve are constructed based on the transient state performance decay data, and the transient decay index time-varying parameter under the transient state operation mode is calculated according to the redshift amount change curve and the intensity decay curve; The transient decay index time-varying parameter is substituted into the thermal quenching kinetics model to calculate the transient transition quantum efficiency data of the Ce 3+ ion under the transient state operation condition in combination with the transient temperature data of the LED chip; The Ce 3+ ion 5d→4f transition saturation correction is performed on the transient transition quantum efficiency data to obtain saturated corrected quantum efficiency data and calculate the quantum efficiency prediction data of the phosphor layer under the transient state operation mode.

[0008] Optionally, in a fourth implementation form of the first aspect of the present application, the Ce 3+ The ion 5d→4f transition saturation correction factor, the saturation correction quantum efficiency data are obtained and the quantum efficiency prediction data of the phosphor layer in the transient working mode are calculated, including: The excitation light intensity value of the LED chip on the phosphor in the transient working mode is measured, the excitation light intensity value is subjected to a ratio operation with the preset saturation light intensity threshold, and the Ce 3+ The ion transition saturation correction factor; The Ce 3+ The ion transition saturation correction factor is subjected to a product operation with the transient transition quantum efficiency data, and the Ce 3+ The ion 5d→4f transition saturation inhibition effect on the quantum efficiency is obtained, and the saturation correction quantum efficiency data are obtained. Based on the saturation correction quantum efficiency data, a temperature-time coupling prediction model is constructed, the transient temperature change and time parameters of the LED chip are taken as input variables of the temperature-time coupling prediction model for multivariate nonlinear regression analysis, and the quantum efficiency prediction data of the phosphor layer in the transient working mode are calculated.

[0009] Optionally, in a fifth implementation form of the first aspect of the present application, the surface phosphor concentration of the LED lamp bead is subjected to three-dimensional gradient distribution analysis based on the quantum efficiency prediction data of the phosphor layer, and a spatial concentration distribution scheme is obtained, including: The quantum efficiency prediction data of the phosphor layer are multiplied by a preset adjustment coefficient, and the dynamic penetration depth parameter of the phosphor layer in the transient working mode is calculated; Based on the dynamic penetration depth parameter, a three-dimensional concentration distribution matrix of each position on the surface of the LED chip is calculated; The three-dimensional concentration distribution matrix is subjected to gradient optimization according to the surface temperature distribution data of the LED chip, and a gradient optimization concentration matrix is obtained; The gradient optimization concentration matrix is subjected to weighted integral operation with the surface temperature distribution data, the thermal stress dispersion index in the transient working mode is calculated and minimized, and a spatial concentration distribution scheme is obtained.

[0010] Optionally, in a sixth implementation form of the first aspect of the present application, the gradient optimization concentration matrix is subjected to weighted integral operation with the surface temperature distribution data, the thermal stress dispersion index in the transient working mode is calculated and minimized, and a spatial concentration distribution scheme is obtained, including: The concentration value of the phosphor at each spatial position in the gradient-optimized concentration matrix is multiplied point by point with the surface temperature distribution data at the corresponding position, and then three-dimensional integration is performed on the entire LED chip surface to obtain a temperature-concentration weighted integral value; The gradient-optimized concentration matrix is independently subjected to three-dimensional integration operation to calculate the total integral value of the phosphor concentration on the entire LED chip surface, and the temperature-concentration weighted integral value is ratioed with the total integral value to obtain a thermal stress dispersion index in the transient working mode; According to the thermal stress dispersion index, the concentration configuration that makes the thermal stress dispersion index reach a minimum value is found by adjusting the phosphor concentration distribution parameters, and an optimized thermal stress dispersion coefficient is obtained. Based on the optimized thermal stress dispersion coefficient, the phosphor concentration distribution of each region on the LED chip surface is recalculated to determine a spatial concentration distribution scheme for relieving thermal stress concentration in the transient working mode.

[0011] Optionally, in the seventh implementation manner of the first aspect of the present application, the deviation of the current coating thickness from the target thickness is measured according to the spatial concentration distribution scheme, and the coating parameters are adjusted through a PID control algorithm, including: The target light efficiency value is calculated according to the spatial concentration distribution scheme, and the current light efficiency output of the LED chip is measured in real time, and the deviation of the current coating thickness from the target thickness is obtained by difference operation of the target light efficiency value and the current light efficiency output; The deviation is input into the PID control algorithm to calculate the adjustment control parameters of the coating thickness; The coating pressure and coating time of the dispensing machine are adjusted based on the adjustment control parameters, the dispensing amount of the phosphor colloid is controlled through pressure adjustment, and the coating duration is controlled through time adjustment to obtain a dispensing process control signal; The dispensing process control signal is transmitted to the coating turntable control system to adjust the rotation speed of the turntable to realize uniform distribution of the phosphor colloid.

[0012] Optionally, in the eighth implementation manner of the first aspect of the present application, the LED lamp bead light efficiency improvement process control method further includes: The total coating thickness value of the phosphor layer is calculated, and the solidification temperature curve parameters corresponding to different depth positions are calculated based on the total coating thickness value; The solidification temperature curve parameters are input into the reinforcement learning Q-learning algorithm, and the solidification time allocation strategy of each depth layer is calculated through state space sampling, action space searching and reward function optimization; Set a temperature control sequence of a preheating stage, a fast curing stage and a slow cooling stage based on the curing time allocation strategy, control the curing process of the phosphor layer from the surface to the bottom layer through inter-stage temperature gradient adjustment, and obtain a staged curing control parameter; Convert the staged curing control parameter into timing coordination instructions of surface layer fast curing and bottom layer deep curing, ensure that the temperature deviation of each stage is controlled within a preset range through temperature control precision adjustment, and obtain a layered curing process scheme.

[0013] The application also provides an LED lamp bead light efficiency improvement process control device, which comprises: A triggering module is configured to monitor the current change rate and temperature change rate of the LED chip and trigger a transient control switching signal; A monitoring module is configured to start a spectrum analyzer to monitor the red shift amount and intensity change of the phosphor emission spectrum according to the transient control switching signal, and obtain transient performance attenuation data; A prediction module is configured to input the transient performance attenuation data into a thermal quenching kinetics model to predict the quantum efficiency, and obtain quantum efficiency prediction data of the phosphor layer; An analysis module is configured to perform three-dimensional gradient distribution analysis on the surface phosphor concentration of the LED chip based on the quantum efficiency prediction data of the phosphor layer, and obtain a spatial concentration distribution scheme; An adjustment module is configured to measure the deviation between the current coating thickness and the target thickness according to the spatial concentration distribution scheme, and adjust the coating parameters through a PID control algorithm.

[0014] To sum up, the technical scheme provided by the application realizes real-time identification of the transient working mode of the LED lamp bead by establishing a current change rate and temperature change rate product discrimination algorithm, and breaks through the limitation of the prior art which is designed only based on steady-state conditions. The spectrum analyzer is used to dynamically monitor the change of the phosphor emission spectrum at a high frequency, and a thermal quenching kinetics model with a transient attenuation index is introduced, so that the quantum efficiency change trend of Ce³⁺ ion 5d→4f transition under transient conditions can be accurately predicted. Based on the prediction data, a three-dimensional gradient distribution optimization strategy of the phosphor concentration is established, and through the minimization of the thermal stress dispersion index, the spatial configuration of central sparseness and edge denseness is realized, so that the problem of transient thermal stress concentration is effectively alleviated. The adaptive coating thickness control of the PID control algorithm and the depth-related layered differential curing process are combined, so that dynamic optimization control of the phosphor layer of the LED under transient working conditions is realized, the technical difficulty that the traditional static process cannot adapt to the transient working mode is solved, and high-performance output of the LED lamp bead under transient working conditions is realized. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a step schematic diagram of an LED lamp bead light efficiency improvement process control method in an embodiment of the application; Figure 2 is a process control device structure block diagram of LED lamp bead light efficiency improvement in an embodiment of the present application.

[0016] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0018] Referring to Figure 1 , the present embodiment provides a LED lamp bead light efficiency improvement process control method, comprising the following steps: S1, monitoring the current rate of change and temperature rate of change of the LED chip, and triggering a transient control switching signal; Among them, the continuous current and temperature sampling operation is implemented on the working state of the LED chip, wherein the acquisition of current data depends on the high-precision driving current acquisition module, which records the current driving current value of the LED chip in real time with a sampling frequency of 1 kHz, forming a current time sequence data sequence; At the same time, the surface of the LED chip is scanned by infrared thermal imager, and the infrared imaging device has a frame frequency of more than 500 Hz, which provides the temperature time sequence data distribution of the chip surface. In order to capture the dramatic fluctuation trend of current and temperature in time, time difference operation is performed on the above current and temperature time sequence data, and first-order difference calculation is performed on adjacent sampling points to obtain two transient sensitive indexes of current rate of change and temperature rate of change. In order to enhance the response sensitivity to sudden changes, the current rate of change and the temperature rate of change are taken as absolute values and multiplied, and a combined change factor is formed to reflect the synchronous dramatic change of the two. In order to enhance the weight evaluation of the influence degree of the current moment, a time window weight function is introduced, and an exponential decay form is used to apply time weight correction to the current data, so that the contribution of the latest change to the result is greater. Multiply the product of the current rate of change and the temperature rate of change by the time window weight function to obtain a real-time transient working mode discrimination value. The larger the value, the more the LED current working state deviates from the steady state operation. Compare the transient working mode discrimination value with the preset threshold value, and when the discrimination value is greater than the threshold value, it is judged that the current is in a significant transient state, and a transient control switching signal is generated immediately.

[0019] S2, starting the spectrum analyzer to monitor the red shift amount and intensity change of the fluorescent powder emission spectrum according to the transient control switching signal, and obtaining transient performance attenuation data; Specifically, when the system receives the transient control switching signal, the spectral analyzer automatically starts collecting the emission spectrum generated by the current LED lamp bead phosphor layer, and records the real-time data of the fluorescence emission wavelength and light intensity at a high frequency of 1000 Hz to form time series spectral wavelength data and spectral intensity data. After obtaining the real-time spectral wavelength data, the peak recognition algorithm is used to automatically analyze the main emission peak position in the spectral curve, and compare it with the initial emission peak wavelength recorded before the system starts, to obtain the current emission spectrum redshift, i.e. the offset of the wavelength relative to the initial wavelength. The offset reflects the change of the energy level transition center of Ce 3+ ion under the influence of thermal excitation and carrier density change. At the same time, based on the real-time spectral intensity data, the luminescence intensity of the phosphor under the current transient condition is calculated, and the ratio of the initial luminescence intensity is calculated to obtain the luminescence intensity decay ratio data. The lower the ratio, the more serious the luminescence efficiency decreases, reflecting the higher loss of excited state density of the phosphor layer under electrical and thermal shock. In order to comprehensively measure the influence of redshift and intensity decay on the overall luminescence performance, the redshift and intensity decay ratio are weighted and summed by setting a weight coefficient, so as to calculate the transient performance decay data of the phosphor layer under transient working mode.

[0020] S3, inputting the transient performance decay data into the thermal quenching kinetics model to predict the quantum efficiency, to obtain the quantum efficiency prediction data of the phosphor layer; It should be noted that the transient performance decay data includes the change trajectory of the redshift with time and the relative decay trend of the fluorescence emission intensity. Based on this data, a redshift change curve and an intensity decay curve are constructed, and the slope, change amplitude and response delay of the two curves are extracted by time series fitting method during the transient process, and a transient decay index time-varying parameter is derived to characterize the decay rate. The parameter reflects the response ability of the phosphor in response to sudden electrical and thermal load, and embodies the dynamic characteristics of the excited state stability changing with time. The obtained transient decay index time-varying parameter is used as the core dynamic factor, which is substituted into the modified thermal quenching kinetics model. The model introduces a time-dependent term based on the traditional quantum efficiency decay function, so that the calculation result can reflect the nonlinear change of Ce 3+ ion transition efficiency in dynamic non-equilibrium environment. At the same time, the model integrates the transient temperature data provided by the real-time temperature sampling system of the LED chip, so that the quantum efficiency calculation has high sensitivity to the influence of thermal excitation. Based on the above parameter fusion, the transient transition quantum efficiency data of the phosphor Ce 3+ ion at a specific time and under specific temperature conditions is calculated. In order to improve the accuracy and practicality of the prediction, after obtaining the preliminary transition quantum efficiency data, the Ce 3+The saturation phenomenon of energy level transition of ions under high intensity excitation condition is corrected. By introducing the ratio between excitation light intensity and saturation intensity, a saturation correction factor is calculated, and the transient transition quantum efficiency data is adjusted accordingly to obtain the corrected quantum efficiency data after saturation correction. The corrected data is further processed to output the prediction result of the luminescent efficiency trend of the LED lamp bead phosphor layer in the future time window under transient working conditions, i.e. the quantum efficiency prediction data.

[0021] S4, based on the quantum efficiency prediction data of the phosphor layer, a three-dimensional gradient distribution analysis is performed on the surface phosphor concentration of the LED chip to obtain a spatial concentration distribution scheme; Specifically, the quantum efficiency prediction data of the phosphor layer is multiplied by a set of preset adjustment coefficients set according to material properties and process experience, which is used to quantify the penetration depth trend of the phosphor material under the current transient working condition, so as to calculate the dynamic penetration depth parameter. This parameter reflects the change of the excitation light propagation and energy absorption ability in the phosphor layer of different regions. Based on the dynamic penetration depth parameter, a concentration change function in the radial, axial and depth directions is defined in the three-dimensional spatial coordinate system of the LED chip, and a three-dimensional concentration distribution matrix of each position on the surface of the LED chip is calculated using a mathematical mapping model. This matrix reflects the phosphor concentration value that should be allocated at different spatial points to match the light emission demand and quantum efficiency variation law at different positions. The surface temperature distribution data under the current working state of the LED chip is introduced, and the data is coupled with the initially generated three-dimensional concentration distribution matrix. Through numerical optimization algorithm, gradient adjustment is performed to appropriately reduce the phosphor concentration in high temperature area to reduce heat accumulation, and to appropriately increase the concentration in low temperature area to maintain the overall light output efficiency, so as to obtain the gradient optimized concentration matrix under the driving of thermal field. In order to evaluate and minimize the local thermal stress concentration effect, the gradient optimized concentration matrix is weighted and integrated with the corresponding surface temperature distribution data to construct a thermal stress dispersion index reflecting the spatial coupling relationship of thermal energy density. The index is an important indicator to quantify the coupling degree of thermal field and material distribution, and the smaller the value is, the more uniform the thermal load is distributed in space, and the more stable the phosphor thermal excitation response is. Therefore, the system adjusts each distribution parameter in the concentration matrix through iteration to make the thermal stress dispersion index tend to be the smallest, and finally outputs the spatial concentration distribution scheme that meets the thermal field balance and quantum efficiency stability.

[0022] S5, according to the spatial concentration distribution scheme, the deviation between the current coating thickness and the target thickness is measured, and the coating parameters are adjusted through PID control algorithm.

[0023] The target light efficiency value of each coating area is converted according to the spatial concentration distribution scheme, combined with the structure characteristics of the LED chip and the preset material light emission response model. The target light efficiency reflects the local or overall light intensity that should be output under the ideal thickness and ideal concentration conditions. At the same time, during the process execution, the real-time output light efficiency data of the LED chip is monitored by the online optical measurement system, and the real-time output value is compared with the target light efficiency value to calculate the difference, thereby indirectly deducing the deviation between the current coating thickness and the expected target thickness. The deviation is input as a feedback variable into the PID control algorithm, and the system calculates the required coating thickness adjustment control parameter according to the deviation. During the PID calculation process, the proportional term is used to quickly respond to the instantaneous amplitude of the deviation, the integral term is used to accumulate the deviation to correct the long-term residual error, and the differential term is used to respond to the rate of change of the deviation to suppress dynamic overshoot. The three together output the thickness correction instruction, which is converted into a control parameter executable by the dispensing equipment. Based on this adjustment control parameter, two key variables of the dispensing machine, i.e. coating pressure and coating time, are adjusted. The increase or decrease of the coating pressure controls the flow rate and glue output of the fluorescent powder colloid from the nozzle, and the length of the coating time determines the total thickness of the coating formed on the chip surface. The two work together to ensure continuous and controllable thickness on a micro scale. At the same time, in order to ensure that the fluorescent powder colloid can be evenly coated on the chip surface, the dispensing process control signal is transmitted to the coating turntable control system at the same time, and the rotation speed of the turntable is adjusted to optimize the lateral diffusion and balanced deposition process of the colloid on the surface. High-speed rotation is used to improve the backflow distribution effect of the edge colloid to the center, and low-speed rotation enhances the local steady-state deposition capacity, thereby realizing dynamic balance of the coating thickness among multiple regions. In the entire control process, the dispensing pressure, time and turntable speed constitute a multivariable coupled control mechanism to drive the colloid to be accurately distributed and form the desired thickness under the guidance of the spatial concentration scheme.

[0024] In one example, the current rate of change and temperature rate of change of the LED lamp bead are monitored, and a transient control switching signal is triggered, including: The driving current of the LED chip is collected to obtain current time series data, and the temperature time series data of the LED chip is collected by an infrared thermal imager; The current time series data is subjected to time difference operation to calculate the current rate of change, and the temperature time series data is subjected to time difference operation to calculate the temperature rate of change; The absolute values of the current rate of change data and the temperature rate of change data are multiplied, and then multiplied by a time window weight function to perform transient discrimination to obtain a transient working mode discrimination value; The transient working mode discrimination value is compared with a preset threshold value, and when the transient working mode discrimination value is greater than the preset threshold value, a transient control switching signal is generated.

[0025] In this example, the working state of the LED chip is collected by the original physical quantity, in which the current collection system uses a high-frequency current sensor embedded in the LED driving circuit, and the sampling frequency is set to 1 kHz, which is used to continuously capture the driving current value of the LED chip with millisecond-level precision, forming a data sequence containing a time stamp, and the current time series data represents the current dynamic characteristics of the LED in the running process. At the same time, the surface temperature field of the LED chip is detected by an infrared thermal imager, and the sampling frequency of the infrared device is set to 500 Hz or more, and continuous temperature scanning is performed on multiple feature positions on the chip surface to generate temperature time series data with adjustable resolution and time accuracy. The current time series data obtained above is subjected to time difference processing, that is, the first-order difference calculation is performed on the current values of the adjacent two sampling times, and the change amplitude of the current per unit time is obtained, which is the current change rate, which can effectively capture the current transient fluctuations caused by start-up impact, load disturbance or PWM modulation; Similarly, the same time difference processing is performed on the temperature time series data to obtain the temperature change rate, which directly reflects the degree of thermal shock on the chip surface caused by electrothermal heating or external thermal coupling. Considering that the physical meaning of the current change rate and the temperature change rate is not the same, in order to eliminate their influence on the sign and direction, the difference results of the two are respectively taken as absolute values and multiplied, a unified dimension coupling mutation factor is constructed, and the product value is used to quantify the intensity of the simultaneous occurrence of current disturbance and temperature transition of the LED chip at a certain time. In order to strengthen the response ability of the identification algorithm to "recent events" and reduce the sensitivity to outdated data, a time window weight function is introduced to weight the above product value in the time domain, and the weight function adopts an exponential decay structure, that is, the weight exponentially decreases with the growth of time delay, and the recent data contributes significantly to the overall state discrimination, thereby realizing the priority response of transient information. The product of the current change rate and the temperature change rate is multiplied point by point with the time weight function to obtain a dynamically evolving transient working mode discrimination value, and the change curve of the value on the time axis is used to reflect the whole process of the LED chip from steady state to non-steady state transition. The significance of the discrimination value lies in establishing an objective quantitative standard for the system working state, so a representative discrimination threshold is set, and the threshold is obtained by experiment calibration, and its value represents the critical point of the LED chip from the normal running state to the state that needs to intervene in the compensation mechanism. If the current calculated transient discrimination value is greater than the preset threshold, it is judged that the LED chip is currently in a non-steady state working mode, and the light efficiency fluctuation or performance degradation caused by the transient accumulation of carriers, the decrease of the thermal excitation state density or the response delay of the fluorescent powder emission, etc. At this time, the system generates a transient control switching signal.

[0026] In one example, according to the transient control switching signal, the spectral analyzer is started to monitor the red shift amount and intensity change of the fluorescent powder emission spectrum, and the transient performance attenuation data is obtained, including: The spectral analyzer is started to collect the fluorescent powder emission spectrum based on the transient control switching signal, to obtain real-time spectral wavelength data and spectral intensity data; The emission peak of the real-time spectral wavelength data is identified, to obtain emission spectrum redshift data; Based on the spectral intensity data, the decay degree of the fluorescent powder luminous intensity under the transient working mode is calculated, to obtain luminous intensity decay ratio data; The emission spectrum redshift data and the luminous intensity decay ratio data are weighted and summed, to calculate the transient performance decay data of the fluorescent powder layer under the transient working mode.

[0027] In this example, the spectral analyzer is started to collect the fluorescent powder emission spectrum based on the transient control switching signal. The device is based on an integrated array type photoelectric detection element and a high-speed spectral module, to collect the fluorescent spectrum emitted by the LED in real time at a sampling frequency of not less than 1000 Hz, especially focusing on the visible light band emission information generated after the excitation of the fluorescent powder material (such as Ce 3+ doped YAG type fluorescent powder), and synchronously outputting two core physical quantities: one is the wavelength distribution data of the emission spectrum, and the other is the luminous intensity data corresponding to each wavelength position. These two items constitute the basic original information set of the real-time spectral wavelength data and the spectral intensity data. The system implements emission peak identification processing on the collected real-time spectral wavelength data. In this process, the high-frequency noise interference is eliminated through smoothing filtering technology, then the peak search algorithm is used to find the local maximum value point in the spectral waveform, and the main emission peak position is identified. The emission peak corresponds to the Ce 3+Radiative process of ion transition from 5d to 4f state. By comparing the wavelength of the main emission peak identified at the current moment with the wavelength of the steady-state reference emission peak recorded when the system is initialized, the red shift amount of the current emission spectrum is calculated, that is, the drift amplitude of the wavelength to the long wave direction. The red shift phenomenon reflects the decrease of the emission energy of the fluorescent center due to the change of the energy level structure or the enhancement of the excited state coupling, so the red shift amount is an important physical parameter for evaluating the influence of the thermal disturbance or excitation saturation of the phosphor. At the same time, based on the real-time spectral intensity data, the system calculates the relative decay degree of the phosphor luminescence intensity of the LED lamp bead in the transient working mode. This calculation is obtained by ratio processing of the total luminescence intensity integral value at the current transient moment and the reference intensity value at the steady-state starting moment, to obtain a dimensionless decay ratio. The smaller the ratio, the more serious the decrease of the excited state density of the phosphor in the transient driving process, and the higher the weakening degree of the emission ability. Considering that the red shift amount and the intensity decay ratio reflect the transient degradation process from the emission center energy level change and the total luminous efficiency decrease respectively, the two parameters are fused to obtain a more comprehensive performance evaluation index. The system introduces a weighted summation mechanism to superimpose the emission spectrum red shift amount and the luminescence intensity decay ratio in proportion to form the transient performance decay data. Two weight coefficients a and β (for example, a = 0.6, β = 0.4) are set to represent the importance of red shift in energy state stability and the representativeness of intensity decay in luminous efficiency maintenance. The normalized red shift ratio is obtained by dividing the red shift amount by the steady-state wavelength value, and multiplied by a, while 1 minus the intensity decay ratio (indicating the intensity decay degree) is multiplied by β, and then the two are summed up to finally obtain the transient performance decay data. The data is a quantitative reflection of the performance loss suffered by the current state of the phosphor in the non-steady state working mode.

[0028] In one example, the transient performance decay data is input into the thermal quenching kinetics model to predict the quantum efficiency, to obtain quantum efficiency prediction data of the phosphor layer, including: Based on the transient performance decay data, a red shift amount change curve and an intensity decay curve are constructed, and a transient decay index time-varying parameter in the transient working mode is calculated according to the red shift amount change curve and the intensity decay curve; The transient decay index time-varying parameter is substituted into the thermal quenching kinetics model, combined with the LED chip transient temperature data to calculate the Ce 3+ Transient transition quantum efficiency data of ions under transient working conditions; The Ce 3+ Ion 5d→4f transition saturation correction, to obtain saturated quantum efficiency data and calculate the quantum efficiency prediction data of the phosphor layer in the transient working mode.

[0029] In this example, the transient performance decay data is structured and sorted, and the emission spectrum redshift amount and light intensity decay ratio corresponding to different time points are extracted. The two types of data are constructed as time series functions respectively to form the redshift change curve and the intensity decay change curve. To enhance the modeling accuracy of the transient response characteristics, the system uses a combination algorithm of sliding window weighted filtering and spline interpolation to smooth the original data, so that the time curve has continuous derivative characteristics while maintaining the trend change. Based on the redshift and decay curve, a multivariate regression model is used to analyze and fit the transient response curvature. Through the analysis of the time-varying behavior of the first and second derivatives of the curve, the transient decay index time-varying parameter representing the performance degradation rate of the fluorescent powder is calculated. This parameter is used to describe the change process of the performance degradation speed of the system in the non-steady state, and its time dependence reflects the dynamic performance degradation characteristics caused by factors such as start-up shock, thermal surge or pulse driving. The transient decay index time-varying parameter is substituted into the thermal quenching kinetic model. In this stage, based on the traditional thermal quenching theory, a dynamic evolution term is introduced into the model structure to couple the transient decay index and the time variable, and a transition efficiency function that better fits the non-steady state behavior is constructed. Another key input parameter required by the model is the transient temperature data of the LED chip, which is provided in real time by an infrared thermal imager. The surface temperature change trajectory at different time points during the operation of the chip is recorded at a high sampling rate, and the near-surface phosphor layer temperature distribution can be extrapolated through a thermal diffusion mapping algorithm. Under the joint driving of the above two variables, namely the transient decay index and the transient temperature, the thermal quenching model is expanded into a nonlinear expression structure containing a time function, so as to calculate the Ce 3+ ion 5d→4f transition quantum efficiency at different time nodes, i.e. the transient transition quantum efficiency data, which reflects the change in the proportion of non-radiative transitions of the phosphor layer caused by thermal excitation at each dynamic working stage. Under high-power excitation, phosphor materials, especially rare earth ion doped systems, also face the limitation of excited state saturation effect, i.e. when the excitation light intensity approaches the saturation point of the material, even if the excitation continues to increase, the transition efficiency tends to be saturated or even decreases. Therefore, after obtaining the preliminary transient transition quantum efficiency data, saturation effect correction is performed. A saturation correction function is introduced, which takes the ratio of excitation light intensity to material saturation intensity as input, and constructs a first-order inhibitory nonlinear response function to dynamically down-regulate the original transition efficiency, so that the model output is closer to the actual working condition of the fluorescent emission characteristics. The saturation intensity threshold is included in the parameters of the correction function, and its value is obtained through previous calibration experiments. For example, for the commonly used Ce 3+The doping YAG phosphor, whose saturation intensity is generally set at the level of 50 mW / cm2. After the correction process is completed, the system will collect the saturation corrected transient transition quantum efficiency data and generate a quantum efficiency prediction curve within the time window by combining the multi-point curve reconstruction algorithm. This curve covers a dynamic prediction period from the trigger point of the transient event to the end of the system stable time, which is used to evaluate the luminescence performance limit of the phosphor layer in different response stages, and finally obtains the quantum efficiency prediction data.

[0030] In one example, the Ce 3+ ion 5d→4f transition saturation correction, saturation corrected quantum efficiency data and calculation of the quantum efficiency prediction data of the phosphor layer in transient working mode, including: The excitation light intensity value of the LED chip to the phosphor in transient working mode is measured, and the excitation light intensity value is compared with the preset saturation light intensity threshold to obtain Ce 3+ ion transition saturation correction factor; The Ce 3+ ion transition saturation correction factor is multiplied by the transient transition quantum efficiency data to eliminate the Ce 3+ ion 5d→4f transition saturation on quantum efficiency, to obtain saturation corrected quantum efficiency data; Based on the saturation corrected quantum efficiency data, a temperature-time coupling prediction model is constructed, and the transient temperature change and time parameters of the LED chip are used as input variables of the temperature-time coupling prediction model for multiple nonlinear regression analysis to calculate the quantum efficiency prediction data of the phosphor layer in transient working mode.

[0031] In this example, in the actual working process, since the LED chip will produce much higher short-time strong excitation radiation than the steady-state average level under the condition of starting or high-frequency pulse driving, the excitation light intensity of the LED chip to the phosphor layer in this period is measured in real time. This measurement is realized by the built-in micro-light detector array or the lateral light energy feedback loop. The system continuously samples the incident light intensity of the excitation region at a frequency not less than 1000 Hz, forming a set of high-time-resolution excitation light intensity value sequences. The collected excitation light intensity value is compared with the preset saturation light intensity threshold, which is determined in advance according to the physical properties of the phosphor material, such as Ce 3+ Ion-doped YAG phosphor is often used at 50 mW / cm2 2 as a typical threshold. By dividing the actual excitation intensity by this threshold, Ce 3+The transition saturation correction factor describes the proportional relationship between the excitation intensity region of the material and its saturated luminescence response in a physical sense. The closer the value is to 1, the closer it is to the saturation state. A value less than 1 indicates that it is still in the linear excitation region. In order to eliminate the inhibitory effect of the saturation region on the transition efficiency, the above Ce 3+ The transition saturation correction factor is multiplied point by point with the transient transition quantum efficiency data obtained in the previous modeling. Through this correction operation, the nonlinear luminescence loss caused by strong excitation is compensated. The product is the quantum efficiency data after saturation correction. Based on the saturation corrected quantum efficiency data, a temperature-time coupling prediction model is constructed. The model takes the transient temperature change curve of the LED chip and the absolute time parameter as input variables, and jointly models the time domain and thermal domain evolution behavior of quantum efficiency by constructing a two-variable coupling function. The temperature input data comes from the thermal imaging measurement channel established in the previous step, which can provide continuous recording of the temperature rise, thermal relaxation and heat dissipation process of the LED chip during the entire transient operation with millisecond-level accuracy. The time variable is collected synchronously through the system clock and embedded as a basic parameter in the model regression logic. The model body adopts a multivariate nonlinear regression form, in which the core fitting structure is a high-order polynomial regression, a two-variable Logistic surface fitting or a kernel regression model based on a radial basis function network. The specific selection depends on the target LED packaging process, the type of phosphor and the thermal response characteristics of the device. Through the model, the saturation corrected quantum efficiency is taken as the response variable, and the temperature and time are taken as the explanatory variables. The fitting process is performed and a residual control strategy is introduced to control the influence of transient disturbance on the prediction stability. After training and parameter calibration, the model can output the quantum efficiency prediction value of the phosphor layer at a given time point and corresponding transient temperature input, forming a dynamic luminous efficiency trend data stream in the operation of the LED system.

[0032] In one example, based on the quantum efficiency prediction data of the phosphor layer, a three-dimensional gradient distribution analysis of the surface phosphor concentration of the LED lamp bead is performed to obtain a spatial concentration distribution scheme, including: The quantum efficiency prediction data of the phosphor layer is multiplied by a preset adjustment coefficient to calculate the dynamic penetration depth parameter of the phosphor layer in the transient working mode. Based on the dynamic penetration depth parameter, a three-dimensional concentration distribution matrix of each position on the surface of the LED chip is calculated. According to the surface temperature distribution data of the LED chip, the three-dimensional concentration distribution matrix is gradient optimized to obtain a gradient optimized concentration matrix. The gradient optimized concentration matrix is weighted and integrated with the surface temperature distribution data to calculate the thermal stress dispersion index in the transient working mode and perform minimization processing to obtain the spatial concentration distribution scheme.

[0033] In this example, the quantum efficiency prediction data reflects the luminescent ability of the phosphor layer at different time points and different thermal states, on the basis of which a preset adjustment coefficient obtained by experience or experimental calibration is introduced, which is used to convert the quantum efficiency, an optical performance index, into a penetration depth representation parameter at the material structure level. Specifically, the quantum efficiency value is multiplied by the adjustment coefficient to obtain the dynamic penetration depth parameter under transient working conditions. The penetration depth is an important quantitative index for describing the propagation ability of the excitation light in the phosphor layer, and its value determines the concentration of the phosphor to be set at different depths to ensure sufficient energy absorption but not saturation. The dynamic penetration depth is used as a control variable for spatial concentration calculation to construct a three-dimensional concentration distribution function on the surface of the LED chip. In the specific modeling process, a cylindrical coordinate system is constructed in the cross-section of the LED chip, and the radial distance, depth distance and angular direction are defined with the center of the chip as the axis. By setting the concentration distribution function to satisfy exponential growth in the depth direction and hyperbolic tangent function trend in the radial direction, i.e. the structure of low concentration in the center area and slightly high concentration in the edge area, a C(x, y, z) three-dimensional concentration distribution matrix is calculated, where x and y are surface coordinates, and z is the penetration direction. The concentration matrix is scaled and corrected by the penetration depth parameter in mathematics to ensure that the high quantum efficiency area corresponds to a shallow penetration depth and a high local concentration, while the low efficiency area is the opposite. In generating the preliminary three-dimensional concentration matrix, considering the thermal sensitivity of the phosphor and the obvious non-uniformity of the surface temperature distribution of the LED chip during operation, real-time temperature distribution data is introduced to gradient optimize the concentration distribution matrix. The temperature values of each point on the chip surface are collected by an infrared thermal imager at a high frame rate and are one-to-one corresponding to the spatial coordinates to form a temperature field T(x, y). Then the temperature data and the concentration matrix are gradient mapped to adjust the concentration in the high temperature area and moderately compensate in the low temperature area, thereby improving the overall thermal balance ability and thermal excitation stability of the system. This optimization process is based on the gradient descent algorithm for iterative adjustment, and the goal is to make the sensitive response of the concentration matrix to the temperature field optimal for thermal distribution matching, and output the gradient optimized concentration matrix C'(x, y, z). To verify the adaptability of the concentration matrix to the thermal field and complete the control closed loop in physical sense, a thermal stress dispersion index is defined as an evaluation function for spatial structure optimization. This index is obtained by weighted integral operation of the optimized concentration matrix and the temperature distribution field. The calculation method is to integrate the product of the temperature value and the corresponding concentration value of each point in the entire effective working area of the LED chip and normalize the total concentration integral, thereby obtaining an average thermal load index under unit concentration. The smaller the value, the more uniform the thermal energy distribution in the phosphor layer, the lower the degree of local overheating or thermal aggregation, and the higher the overall light efficiency stability. To optimize the structure, the system adjusts the adjustment factor of the penetration depth parameter to re-iterate the concentration function, so that the thermal stress dispersion index converges to a local minimum value, forming a globally acceptable optimal distribution structure.The concentration distribution result obtained after the minimization processing is determined as the spatial concentration distribution scheme.

[0034] In one example, the gradient-optimized concentration matrix is subjected to a weighted integral operation with the surface temperature distribution data, a thermal stress dispersion index in the transient working mode is calculated and subjected to a minimization processing, and a spatial concentration distribution scheme is obtained, including: The concentration value of the phosphor at each spatial position in the gradient-optimized concentration matrix is subjected to a point-by-point multiplication operation with the surface temperature distribution data at the corresponding position, and then a three-dimensional integral is performed on the entire LED chip surface to obtain a temperature-concentration weighted integral value; The gradient-optimized concentration matrix is subjected to an independent three-dimensional integral operation, the total integral value of the phosphor concentration on the entire LED chip surface is calculated, and the temperature-concentration weighted integral value is subjected to a ratio operation with the total integral value to obtain a thermal stress dispersion index in the transient working mode; According to the thermal stress dispersion index, the concentration configuration that makes the thermal stress dispersion index reach a minimum value is found by adjusting the phosphor concentration distribution parameters, and an optimized thermal stress dispersion coefficient is obtained; Based on the optimized thermal stress dispersion coefficient, the phosphor concentration distribution of each region on the LED chip surface is recalculated, and a spatial concentration distribution scheme that relieves the concentration of thermal stress in the transient working mode is determined.

[0035] In this example, based on the gradient-optimized concentration matrix, the current phosphor concentration value at each spatial coordinate position is obtained, while the temperature distribution on the surface of the LED chip is mapped, and the temperature field is extended and matched to the spatial structure of the entire three-dimensional concentration matrix. By multiplying the concentration value at each spatial position with the temperature value at the corresponding position, a temperature-concentration coupling expression is formed to describe the thermal load of the phosphor in the local temperature environment. This point-by-point multiplication constructs a weighted heat distribution that physically reflects the bearing distribution characteristics of the phosphor concentration distribution in response to the heat field, i.e., areas with high concentration but low temperature are beneficial to light efficiency improvement, while areas with high concentration and high temperature are potential hidden dangers of thermal stress accumulation. After completing the point-by-point multiplication operation, a three-dimensional integration is performed on the entire LED chip surface packaging space to sum the heat-concentration product values at all spatial points, obtaining the overall thermal load expression value under the joint action of phosphor concentration and temperature, which reflects the total strength of the thermal stress borne by the phosphor layer under the current concentration layout. At the same time, a three-dimensional integration is also independently performed on the gradient-optimized concentration matrix itself to obtain the total distribution amount of phosphor material in the entire LED chip. By calculating the ratio of the aforementioned total thermal load value to the total concentration value, the thermal stress dispersion index under the transient working state is obtained, which represents the diffusion efficiency of thermal energy in the phosphor layer. The smaller the value, the more conducive the concentration distribution is to the dispersion and release of thermal stress, and the better the thermal stability of the system. After obtaining the thermal stress dispersion index, the system enters the iterative optimization phase of the concentration structure. In this phase, the control factors of the concentration distribution are used as variables, such as adjusting the depth variation amplitude of the spatial concentration, the edge transition slope, the lower limit of the central concentration, etc. The algorithm continuously adjusts these control variables, and after each adjustment, a new concentration distribution matrix is reconstructed, and the multiplication and integration calculations are repeated to obtain a new thermal stress dispersion index. The system continuously compares the change trend of the new and old indexes, and in each iteration, it moves towards the direction of reducing the index value until the value stabilizes in the minimum interval or reaches the set convergence threshold, determining a set of optimal concentration distribution parameters. Based on the optimization results, a new phosphor concentration distribution scheme is generated, which maintains the stability of the total luminous flux and reflects the coordinated configuration logic of concentration and thermal environment in the spatial distribution. Specifically, the concentration in the high-temperature central region of the chip is appropriately reduced to prevent excessive thermal excitation from causing quantum efficiency to decrease, while the concentration in the low-temperature edge region is appropriately increased to compensate for the natural attenuation of the edge light intensity, ultimately forming a spatial concentration mode that balances thermodynamics and optics.

[0036] In one example, the deviation of the current coating thickness from the target thickness is measured according to the spatial concentration distribution scheme, and the coating parameters are adjusted through a PID control algorithm, including: According to the space concentration distribution scheme, the target light efficiency value is calculated, the current light efficiency output of the LED chip is measured in real time, the target light efficiency value is subtracted from the current light efficiency output, and the deviation between the current coating thickness and the target thickness is obtained; The deviation is input into a PID control algorithm to calculate the adjustment control parameter of the coating thickness; Based on the adjustment control parameter, the coating pressure and coating time of the dispensing machine are adjusted, the dispensing amount of the fluorescent powder colloid is adjusted and controlled through pressure adjustment, the coating duration is adjusted and controlled through time adjustment, and a dispensing process control signal is obtained; The dispensing process control signal is transmitted to the coating turntable control system to adjust the rotation speed of the turntable to achieve uniform distribution of the fluorescent powder colloid.

[0037] In this example, based on the optimized spatial concentration distribution scheme, the system is divided into discrete LED chip surface, obtain the required concentration of phosphor target value of each sub-region, and combined with the quantum efficiency prediction data, excitation light distribution model and fluorescence conversion efficiency formula, the function relationship between the regional light contribution and the concentration of phosphor is established, and then the theoretical light efficiency output value of each sub-region under the ideal concentration configuration is calculated. The theoretical light efficiency of all regions is accumulated to obtain the target light efficiency value corresponding to the whole LED chip. This target value is set as the criterion for the system to meet the optical performance standard. At the same time, the real-time light efficiency acquisition device carried by the system continuously monitors the light intensity output of the LED chip during the process execution. The measurement method uses an integrating sphere structure combined with a photodetection module or an external luminous flux sensor based on lens calibration. The device outputs real-time light efficiency data stream at a high sampling frequency. By comparing the current real-time measured light efficiency value with the target light efficiency value calculated by the spatial concentration distribution scheme, the light efficiency deviation is obtained. This deviation directly reflects the deviation between the actual thickness and the expected target thickness in the optical sense. If the current light efficiency is lower than the target value, it means that the actual thickness is insufficient or the colloid distribution is uneven. Otherwise, there is excessive accumulation or scattering loss. In order to dynamically adjust this deviation and convert it into executable coating process instructions, the system uses a PID control algorithm for processing. The PID controller takes the light efficiency deviation as the input error signal and calculates the required thickness adjustment amount at the current time according to the proportional, integral and derivative terms. The proportional term is used to respond to the deviation itself, the integral term is used to accumulate the long-term deviation of the system to avoid steady-state error, and the derivative term is used to predict the trend of the deviation to improve the response foresight. The three control amounts are weighted and superimposed to form the thickness adjustment control parameter, which is input to the dispenser control system as a control instruction to adjust two key variables of the dispensing process: the dispensing pressure and the coating time. The coating pressure adjusts the flow rate of the fluorescent powder colloid when it is extruded from the dispensing nozzle, directly affecting the deposition rate of the fluorescent powder per unit time. The coating time control defines the length of time the nozzle stays in the target area, determining the thickness accumulation level of the fluorescent powder coating in that area. The system decomposes the thickness adjustment control parameter value calculated by PID into pressure adjustment factor and time adjustment factor, which are applied to the pneumatic control module and time relay control channel respectively to generate the dispensing process control signal. This process control signal is transmitted to the coating turntable control system simultaneously to adjust the rotation speed of the turntable, realizing the flow distribution regulation of the colloid on the chip surface from the center to the outside. In actual operation, the rotation speed of the turntable is dynamically matched according to the viscosity of the colloid, the area of the chip and the thickness adjustment amount. Too fast speed will cause centrifugal loss of the colloid, and too slow speed will cause concentrated accumulation. The system establishes a mapping curve between speed and coating uniformity through experimental calibration to realize real-time balance between coating quality and adjustment response.The closed-loop control system driven by light efficiency deviation dynamically senses coating error and corrects process parameters in real time during LED chip processing, so that the deposition thickness and spatial concentration of phosphor colloid are consistent with the target, ensuring that the LED device has high conversion efficiency, high luminous flux output, and low thermal stress distribution under transient working conditions.

[0038] In one example, the LED lamp bead light efficiency improvement process control method further comprises: calculating the total coating thickness value of the phosphor layer, and calculating the curing temperature curve parameters corresponding to different depth positions based on the total coating thickness value; inputting the curing temperature curve parameters into the reinforcement learning Q-learning algorithm, and calculating the curing time allocation strategy for each depth layer through state space sampling, action space search and reward function optimization; based on the curing time allocation strategy, setting the temperature control sequence of the preheating stage, the rapid curing stage and the slow cooling stage, controlling the curing process of the phosphor layer from the surface to the bottom layer through the temperature gradient adjustment between stages, and obtaining the staged curing control parameters; convert the staged curing control parameters into time sequence coordination instructions for rapid curing of the surface layer and deep curing of the bottom layer, adjust the temperature control accuracy to ensure that the temperature deviation of each stage is controlled within the preset range, and obtain the layered curing process scheme.

[0039] In this example, the system calculates the overall thickness of the phosphor layer on the entire LED chip surface according to the structural morphology data after the phosphor coating operation is completed. This total coating thickness is derived from the data fusion of multiple local thickness points, reflecting the overall depth of the phosphor deposition on the LED surface. The thickness information is a basic parameter for judging the solidification needs and temperature conduction period, which determines the heat transfer path length from the surface to the bottom layer, and directly affects the required heat solidification time and target temperature at different depths. Based on the total coating thickness value, the time variation trajectory of the solidification temperature is established for multiple depth positions of the entire phosphor layer from the surface to the chip side. Each depth corresponds to a heat reaction requirement curve to describe the temperature rise, holding and cooling process required at that position during the solidification period. These initial curve parameters form the basis of the temperature control strategy for solidification control. The Q-learning algorithm in reinforcement learning is introduced to optimize these temperature control strategies. The system defines a complete state set, with each state representing the temperature, solidification degree and time progress of the coating at a certain depth at a certain time. At the same time, a set of executable operations is designed for each state, such as extending the solidification time, increasing the temperature or reducing the energy consumption, etc. The system continuously tries different combinations of operations in the simulation environment and records the final results. The pros and cons of the results are evaluated by the reward mechanism, and the operation path with higher reward value is preferentially retained. After a sufficient number of training, the system obtains a set of optimal solidification time allocation strategies, so that the entire phosphor layer completes the structural solidification from the surface to the bottom layer without heat concentration, overburning and overcooling. According to the optimized solidification time allocation strategy, the system divides the overall solidification process into three stages. The first stage is the preheating period, which mainly uses medium-speed heating to gradually raise the overall temperature of the phosphor layer to approach the reaction starting point; the second stage is the rapid solidification period, which maintains a high temperature level to make the surface material react quickly to form a strong structure; the third stage is the slow cooling period, which aims to reduce the temperature difference and release internal stress, so that the overall thermal stress of the coating gradually balances and eventually stabilizes. Each stage has a clear temperature variation trajectory and duration setting, and the transition control between adjacent stages is controlled by the temperature variation rate to prevent material internal cracking or quantum efficiency reduction due to sudden temperature difference. These stage control parameters will be converted into hardware control instructions and transmitted to the temperature control module in the heat solidification system. The system controls the surface solidification and bottom solidification synchronously and coordinately, so that the surface reaction quickly forms a structural layer while the bottom can obtain sufficient heat to complete the deep solidification with appropriate delay. In addition, through continuous collection of feedback data from the temperature control equipment, the system adjusts the heater output power and solidification time in real time to ensure that the temperature control deviation is maintained within the allowed range.

[0040] Referring Figure 2 The embodiment provides an LED lamp bead light efficiency improvement process control device, which comprises: A trigger module 1 is configured to monitor the current rate of change and the temperature rate of change of the LED chip and trigger a transient control switching signal; A monitoring module 2 is configured to start a spectral analyzer to monitor the red shift and intensity change of the phosphor emission spectrum according to the transient control switching signal, and obtain transient performance decay data; A prediction module 3 is configured to input the transient performance decay data into a thermal quenching kinetics model to predict quantum efficiency, and obtain quantum efficiency prediction data of the phosphor layer; An analysis module 4 is configured to perform three-dimensional gradient distribution analysis on the surface phosphor concentration of the LED chip based on the quantum efficiency prediction data of the phosphor layer, and obtain a spatial concentration distribution scheme; An adjustment module 5 is configured to measure the deviation of the current coating thickness from the target thickness according to the spatial concentration distribution scheme, and adjust the coating parameters through a PID control algorithm.

[0041] In the embodiment, the specific implementation of each unit in the above device embodiment is described above in the method embodiment, which will not be described here.

[0042] The application can accurately identify the transient working state of the LED lamp bead by establishing a product discrimination algorithm of the current rate of change and the temperature rate of change, combined with a time window weight function, which is different from the design method based on only the steady state condition in the prior art, and provides timely and accurate trigger signals for subsequent dynamic control. The spectral analyzer high-frequency sampling technology is used to monitor the red shift and intensity change of the phosphor emission spectrum in real time, establish a transient performance decay evaluation system, break through the limitations of traditional static spectrum measurement, introduce a transient decay index time-varying parameter, and construct a thermal quenching kinetics model considering the saturation correction of Ce 3+ ion 5d→4f transition, which can accurately predict the quantum efficiency change trend of the phosphor layer under the transient working mode. Based on the quantum efficiency prediction data, a three-dimensional gradient distribution function of the phosphor concentration is established, and through the minimization processing of the thermal stress dispersion index, the spatial configuration of the center sparseness and the edge denseness is realized, and the thermal stress concentration problem under the transient working mode is effectively alleviated. The PID control algorithm is used to realize the dynamic adjustment of the coating parameters, the point glue pressure, the coating time and the rotation speed of the turntable are cooperatively controlled to ensure the accurate matching of the phosphor layer thickness and the concentration gradient distribution, and the process control precision is improved. The depth-related solidification temperature curve and the three-stage time sequence control mechanism are established to realize the coordination of the surface layer rapid solidification and the bottom layer deep solidification, and solve the technical problem that the traditional single solidification process cannot adapt to the transient working condition.

[0043] It is to be understood that the terminology "including", "comprising", or any other variation thereof, is intended to cover a non-exclusive inclusion such that process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a... " does not, without more constraints, exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0044] The above description is merely the preferred embodiments of the present application, and is not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made according to the content of the present application specification and drawings, or directly or indirectly applied to other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A method for controlling the process of improving the luminous efficacy of LED beads, characterized in that, include: Monitor the rate of change of current and the rate of change of temperature of the LED chip and trigger transient control switching signals; The transient control switching signal is used to start the spectrometer to monitor the redshift and intensity changes of the phosphor emission spectrum, and to obtain transient performance degradation data. The transient performance degradation data is input into the thermal quenching kinetic model to predict the quantum efficiency, thereby obtaining the quantum efficiency prediction data of the phosphor layer. Based on the quantum efficiency prediction data of the phosphor layer, a three-dimensional gradient distribution analysis of the surface phosphor concentration of the LED chip is performed to obtain a spatial concentration distribution scheme. The deviation between the current coating thickness and the target thickness is measured according to the spatial concentration distribution scheme, and the coating parameters are adjusted by a PID control algorithm.

2. The LED lamp bead luminous efficacy enhancement process control method according to claim 1, characterized in that, The monitoring of the current change rate and temperature change rate of the LED beads, and the triggering of transient control switching signals, includes: The driving current of the LED chip is collected to obtain current timing data, and the temperature timing data of the LED chip is collected by an infrared thermal imager. The current time series data is subjected to time difference operation to calculate the current change rate, and the temperature time series data is subjected to time difference operation to calculate the temperature change rate. Multiply the absolute values ​​of the current change rate data and the temperature change rate data, and then multiply them by the time window weighting function to perform transient discrimination, thereby obtaining the transient operating mode discrimination value; The transient operating mode discrimination value is compared with a preset threshold. When the transient operating mode discrimination value is greater than the preset threshold, a transient control switching signal is generated.

3. The LED lamp bead luminous efficacy enhancement process control method according to claim 1, characterized in that, The step of activating the spectrometer based on the transient control switching signal to monitor the redshift and intensity changes of the phosphor emission spectrum, and obtaining transient performance degradation data, includes: Based on the transient control switching signal, the spectrometer is activated to collect the emission spectrum of the phosphor, thereby obtaining real-time spectral wavelength data and spectral intensity data. The emission peaks of the real-time spectral wavelength data are identified to obtain the redshift data of the emission spectrum; Based on the spectral intensity data, the degree of attenuation of the phosphor's luminescence intensity in transient operating mode is calculated to obtain the luminescence intensity attenuation ratio data; The emission spectrum redshift data and the luminescence intensity attenuation ratio data are weighted and summed to calculate the transient performance attenuation data of the phosphor layer under transient operating mode.

4. The LED lamp bead luminous efficacy enhancement process control method according to claim 1, characterized in that, The step of inputting the transient performance degradation data into the thermal quenching kinetic model for quantum efficiency prediction to obtain quantum efficiency prediction data for the phosphor layer includes: Based on the transient performance degradation data, a redshift variation curve and an intensity degradation curve are constructed, and the time-varying parameter of the transient degradation index under transient operating mode is calculated by fitting the redshift variation curve and the intensity degradation curve. Substituting the time-varying parameters of the transient decay exponent into the thermal quenching kinetic model, and combining this with the transient temperature data of the LED chip, the phosphor Ce was calculated. 3+ Transient quantum efficiency data of ions under transient operating conditions; Ce was used to analyze the transient transition quantum efficiency data. 3+ Saturation correction was applied to the 5d→4f transition of ions to obtain saturated corrected quantum efficiency data, and the predicted quantum efficiency data of the phosphor layer under transient operating mode was calculated.

5. The LED lamp bead luminous efficacy enhancement process control method according to claim 4, characterized in that, The transient quantum efficiency data is subjected to Ce 3+ Saturation correction for the ion 5d→4f transition was performed to obtain saturation-corrected quantum efficiency data, and the predicted quantum efficiency data of the phosphor layer under transient operating mode was calculated, including: The excitation light intensity of the LED chip to the phosphor is measured in transient operating mode. The ratio of the excitation light intensity to a preset saturation light intensity threshold is then calculated to obtain Ce. 3+ Ion transition saturation correction factor; Ce 3+ The ion transition saturation correction factor is multiplied by the transient transition quantum efficiency data to eliminate Ce under transient operating conditions. 3+ The effect of ion 5d→4f transition saturation on the suppression of quantum efficiency was investigated, and saturation-corrected quantum efficiency data were obtained. Based on the saturated corrected quantum efficiency data, a temperature-time coupled prediction model is constructed. The transient temperature change of the LED chip and the time parameter are used as input variables of the temperature-time coupled prediction model to perform multivariate nonlinear regression analysis, and the quantum efficiency prediction data of the phosphor layer under transient operating mode is calculated.

6. The LED lamp bead luminous efficacy improvement process control method according to claim 1, characterized in that, The quantum efficiency prediction data based on the phosphor layer is used to perform a three-dimensional gradient distribution analysis of the surface phosphor concentration of the LED chip, resulting in a spatial concentration distribution scheme, including: The predicted quantum efficiency data of the phosphor layer is multiplied by a preset adjustment coefficient to calculate the dynamic penetration depth parameter of the phosphor layer under transient operating mode; The three-dimensional concentration distribution matrix at each location on the surface of the LED chip is calculated based on the dynamic penetration depth parameter. The three-dimensional concentration distribution matrix is ​​gradient optimized based on the surface temperature distribution data of the LED chip to obtain the gradient-optimized concentration matrix. The gradient-optimized concentration matrix and the surface temperature distribution data are weighted and integrated to calculate the thermal stress dispersion index under transient operating mode and minimize it to obtain the spatial concentration distribution scheme.

7. The LED lamp bead luminous efficacy enhancement process control method according to claim 6, characterized in that, The step of performing a weighted integral operation on the gradient-optimized concentration matrix and the surface temperature distribution data to calculate and minimize the thermal stress dispersion index under transient operating mode, thereby obtaining a spatial concentration distribution scheme, includes: The phosphor concentration value at each spatial location in the gradient optimization concentration matrix is ​​multiplied point by point with the surface temperature distribution data at the corresponding location. Then, the entire LED chip surface is integrated in three dimensions to obtain the temperature-concentration weighted integral value. Perform independent three-dimensional integral operations on the gradient optimization concentration matrix to calculate the total integral value of phosphor concentration on the entire LED chip surface, and calculate the ratio of the temperature-concentration weighted integral value to the total integral value to obtain the thermal stress dispersion index under transient operating mode. Based on the thermal stress dispersion index, the optimal thermal stress dispersion coefficient is obtained by adjusting the phosphor concentration distribution parameters to find the concentration configuration that minimizes the thermal stress dispersion index. Based on the optimized thermal stress dispersion coefficient, the phosphor concentration distribution in each region of the LED chip surface is recalculated to determine a spatial concentration distribution scheme to alleviate thermal stress concentration under transient operating mode.

8. The LED lamp bead luminous efficacy enhancement process control method according to claim 1, characterized in that, The step of measuring the deviation between the current coating thickness and the target thickness according to the spatial concentration distribution scheme, and adjusting the coating parameters through a PID control algorithm, includes: The target luminous efficacy value is calculated based on the spatial concentration distribution scheme. At the same time, the current luminous efficacy output of the LED chip is measured in real time. The difference between the target luminous efficacy value and the current luminous efficacy output is calculated to obtain the deviation between the current coating thickness and the target thickness. The deviation is input into the PID control algorithm to calculate the adjustment control parameters for the coating thickness; Based on the aforementioned adjustment and control parameters, the coating pressure and coating time of the dispensing machine are adjusted. The amount of phosphor colloid dispensed is controlled by adjusting the pressure, and the coating duration is controlled by adjusting the time, thereby obtaining the dispensing process control signal. The dispensing process control signal is transmitted to the coating turntable control system, and the turntable speed is adjusted to achieve uniform distribution of phosphor colloid.

9. The LED lamp bead luminous efficacy enhancement process control method according to claim 8, characterized in that, The method for controlling the luminous efficacy improvement process of LED lamp beads also includes: Calculate the total coating thickness of the phosphor layer, and calculate the curing temperature curve parameters corresponding to different depth positions based on the total coating thickness. The curing temperature curve parameters are input into the reinforcement learning Q-learning algorithm, and the curing time allocation strategy of each depth layer is optimized by state space sampling, action space search and reward function. Based on the curing time allocation strategy, a temperature control sequence is set for the preheating stage, the rapid curing stage and the slow cooling stage. The curing process of the phosphor layer from the surface to the bottom layer is controlled by adjusting the temperature gradient between stages, and the staged curing control parameters are obtained. The phased curing control parameters are converted into timing coordination instructions for rapid surface curing and deep bottom curing. Temperature control precision is adjusted to ensure that the temperature deviation of each stage is controlled within a preset range, thus obtaining a layered curing process scheme.

10. A process control device for improving the luminous efficacy of LED beads, characterized in that, The LED lamp bead luminous efficacy enhancement process control device, used to implement the LED lamp bead luminous efficacy enhancement process control method according to any one of claims 1 to 9, comprises: The trigger module is used to monitor the rate of change of current and the rate of change of temperature of the LED chip and trigger transient control switching signals; The monitoring module is used to activate the spectrometer based on the transient control switching signal to monitor the redshift and intensity changes of the phosphor emission spectrum, and obtain transient performance degradation data. The prediction module is used to input the transient performance degradation data into the thermal quenching kinetic model to predict the quantum efficiency, thereby obtaining the quantum efficiency prediction data of the phosphor layer. The analysis module is used to perform a three-dimensional gradient distribution analysis of the surface phosphor concentration of the LED chip based on the quantum efficiency prediction data of the phosphor layer, and obtain a spatial concentration distribution scheme. The adjustment module is used to measure the deviation between the current coating thickness and the target thickness according to the spatial concentration distribution scheme, and to adjust the coating parameters through a PID control algorithm.

Images